i

EXPLANATORY
SUPPLEMENT
TO THEASTRONOMICAL
ALMANAC

A Revisioflto the
EXPLANATORY
SUPPLEMENT
TOTHEASTRONOMICAL
EPHEMEBIS
and
THEAMERICAN
EPHEMERIS
ANDNAUTICAL
ALMANAC

Preparedby
THENAUTICAL
ALMANAC
OFFICE,U.S.NAVALOBSERVATORY
WITHCONTRIBUTIONS
FROM
H.M.NAUTICAL
ALMANAC
OFFICE,BOYALGREENWICH
OBSERVATORY
JET PROPULSION
LABOMTORY
BUREAUDESLONGITUDES,
and
THETIMESERVICE
ANDASTROMETRY
OEPARTMENTS.
U.S,NAVALOBSERVATORY

Editedby
P KennethSeidelmann

UNIVERSITY SCIENCEBOOKS
MillValley,Califomia
 AbbrEViAtEd
CONTENTS
University ScienceBooks
20 Edgehill Road
Mill valley, CA 9494 I
F a x :( 4 1 5 )3 8 3 - 3 1 6 7

Productionnanager:MaryMiller
LlsToF FloUBEs wii
Copyedibr AidanKelly andMdSy KDnD
TexIdd jacketdesigner:Robenhhi LISTOFTABLESsi
TEXfomarer and illusrabr: Ed Sznyer FOREWOBoN
Pr@ftqder: Jm McDemon PBEFACErvii
Printerod biDder:The Maple Vail B@k Mmufacturingcrcup

1 INTRODUCTIONTO POSITIONALASTRONOMY1
2 TIME 39
3 CELESTIALREFERENCE SYSTEMS 95
4 TERRESTRIALCOORDINATES
CopyrighrO 1992by Unive6iryScience B@ks ANOTHE ROTATIONOF THE EARTH .I99
Reprcducrionor r.anslarionof dy pan of rhis wo* beyond
ftar pemited by Secrionl0? or IoE of lhe t9?6 UniredSrares 5 ORBITALEPHEMERIDES
CopyrighrAct withour lhe pemission of the copyrightowner OF THESUN,MOON,ANDPLANETS 279
is unlawful.Requestsfor pemissiono. funherinformaron
shouldbeaddress€d to rhePemissions
6 ORBITALEPHEMERIDES ANDRINGSOF SATELLITES325
Depanment,
Unive6ityScicnceBooks. 7 PHYSICALEPHEIUERIOESOF THESUN,I\,,IOON,
PLANETS,
AND
Libral' of CongrcssCatalogNmbe.: 9l -65j3 I SATELLITES383
ISBN0-915702,68-7 8 ECLIPSESOFIHE SUNANDMOON 421
9 ASTFONOMICALPHENOMENA475
10 STARSANDSTELLqRSYSTEMS 505
11 COMPUTATIONALTECHNIOUES541
12 CALENDARS575
12.10REFERENCES606
13 HISTORICAL
INFORMATION609
14 RELATEDPUBLICATIONS667
15 REFERENCEDATA 693
Prinredin theUnnedSlaresofADerica
10987654121 GLOSSAFY721

tNoEx 741
 Contents

LISTOF FIGURES xvii

LISTOFTABLESxxi
FOREWORDxxv
PBEFACE xxvii

1 / INTRODUCTIONTOPOSITIONALASTRONOMY
1
1.1 INTRODUCTION 1
1.11 Purpose 1
1,2 TIMESCALESAND CALENDARS 2
1.21 AtomicImescales 2
1.22 Dynamical'Iime 3
1.23 Rotational
Tirnescales3
1.24 CoordinatedUniversalTime(UTC) 6
1.25 The Enumerationof Dales 7
1.3 CELESTIALAND TERRESTRIALCOORDINATES 8
1.31 Coordinate SystemsandFrames 8
1.32 CelestialCoordinate Systems 11
1.33 TerrestrialCoordinateSystems 13
1.34 The Rotationof the Earth 'tI
1.35 ThgConnections betweenTerrestrial Coordinates20
andCelestial
1.36 Efiectsof the PositionandMotlonof ths ObjectandObserver 21
1.4 ORBITALMOTIONS 24
'1.41 Motionin Two-body Systems 24
1.42 Typesof Perturbations28
1.,+3 Perlurlrationsby and on ExtondedBodies 30
1.5 ASTRONOMICALPHENOMENA 32
'1.51 Rlslng,Setting,andTwilight 32
1.52 Meridian Transit 3it
v
 v l CONTENTS CONTENTS
1.53 Conjunction,Opposition, and Elongation33 3.22 Nutation |09
1.54 Eclipses, Occultations,
andTransits 34 3.23 SoaceMotion 121
1.55 SatellitePhenomena35 3.24 Palallax 123
1.56 PhysicalObservations of the Sun,t\roon,andplanets 96 3.25 Aberration '127
3.26 lightdeflection 135
Gravitational
1,6 REFERENCES 38
3.27 PolarMotion 139
3.28 Refraction 140
/ TIME 39
2,1 INTRODUCTION
39 3.3 APPABENTAND TOPOCENTRICPLACEALGORITHMS 145
for Planels 145
Algorithm
3.31 Apparent-Place
2.2 MEASURES OF TIMEANDTHEIRRELATIONSHIPS
40 for Slars 152
Algorithm
3.32 Apparent-Place
Time(TAt) 40
2.21 Atomic 3.33 TheComputerlmplementation Algorithms 154
of Apparent-Place
2.22 OynamicalTtme 41 3.34 Apparent-Places-Day'NumberTechnique'155
2.23 SiderealIime48 3.35 Topocentric-PlaceAlgorithm160
2.24 UniversalTime 50
2.25 The EphemerjsMeridjan 54 3,4 DIFFEBENTIAL ASTROMETRY 165
2.26 JulianDate 55 3.41 VirtualPlace 165
2.27 Ime Zones 56 3.42 LocalPlace 165
3.43 AstrometricPlace 166
2.3 PRACTICALDETERMINATIONS OF TIME 58
2.31 Frequency
Standards andClockperformance5g 3,5 TRANSFORIVIATION TO FK5 SYSTEMAND EPOCHJ2OOO.O167
2.32 Measurementot AtomicTime 60 3.51 FK4Zero-Point in RightAscension 167
Correction
2.33 EadhRotationMeasurement61 3.52 TheCorrection to the FK4Properl/otionSystemin RightAscension | 68
2.34 DynamicalTimeDetermjnations69 3.53 EllipticTermsin Aberration 169
2.4 METHODSOF TIME TRANSFER 64 3.54 Precession 173
3.55 The Proper-Motion System 174
2.41 BadioIme Signats 65 3.56 Equations lor theTransformation of Catalogsfrom81950.0to J2000.0 175
2.42 PortabteCtocks 65 3.57 Transformation Catalogs 179
ot Observational
2.43 LORAN_C 66 3.58 Numerical ExamPles 180
2.44 Television
Comparison Techniques67 3.59 MatrixMethod 180
2.45 Useof Satellites 68
2.46 Intercontinental
CtockSynchronization by VLB| 69 3.6 REFERENCES 187
2.47 Relativistlc
Efiectsin lime Transfer 70
2.5 HISTORICALDEVELOPMENT OF TIIVlEKEEPING73 4 / TERRESTRIAL
COORDINATES
2.51 Inkoduction73 OF THEEARTH 199
2.52 ApparentSolarTime,MeanSotar-Time,
ANDTHEROTATION
andthe Equation
of Tjme 74
2.53 Rotationof the Earth 76 199
4.1 INTRODUCTION
2.54 Universall"ime 77
2.55 Ephemeris l'ime 79 SYSTEMS2OO
COORDINATE
4,2 TERBESTRIAL
2.56 Historyof AtomicTjme g4
4.21 Ellipsoid 200
The Figureol the Earthandthe Reference
2.57 Historyof CoordinatedUnive.sat
Ime 85 4.22 Geodetic,
Geocentric, andAstronomicalCoordinates202
2.58 Historyot Transmitted
lime Sionats g6 4.23 LocalcoordinaleSystems 207
2.6 REFERENCES 88 4.24 GeodeticDatumsandReference Systems 216
4.25 The Conventional Terrestrial
ReferenceSystem 223
at CELESTIAL
REFERENCE
SYSTEMS 95 4.3 GRAVITYTHE TIDES,AND MOTIONSOF THE CRUST 224
3.1 CELESTIAL
BEFERENCE
SYSTEMS95 4.31 Modelingthe Earth'sGravityField 224
3.11 FundamentalReferenceSystems 97 4.32 A Represenlationof lhe Earth'sGravilyField 227
3.12 The DynamicalRelerenceSystem 98 4.33 SolidEarthTides 233
3.13 The ConventionalCelestialReferenceSystem 98 4.34 OceanTideModel 240
4.35 SateDisplacement Dueto OceanandAtmospheric Loading 244
3.2 BASISOF REDUCTION
OF CELESTIAL
COORDINATES
99 4.36 Platel'rotions 249
3.21 Precession 99 4.37 TidalEffectson UTI 250
 CONTENTS
CONTENTS
4.4 THEMONITORING
OFTHEROTATION
OFTHEEARTH 251 OFTHEMAJOR
FORTHEPOSITIONS
ELEMENTS
5.8 KEPLERIAN
4.41 LaserFianging 254 PLANETS315
4.42 VeryLongBaselineInterferometry255
4.43 Historical
Methods 256 5.9 BASISFORPRE-1984 EPHEMERIDES 317
4.44 Alternative
Techniques258 ol New 317
Constants
5.91 Introduction
4.45 International
Services 262
5 . 1 0 R E F E R E N C E3S1 9
4.5 DETERMINATION OF PASTVARIATIONSIN LENGTHOF DAY
AND THE POSITIONOF THE POLE 265 6 / ORBITAL ANDRINGSOF SATELLITES325
EPHEMERIDES
4.51 Historical
Variations
in UTl and Lengthof Day 265
4.52 Historical
Variations
in Polarlvlotion 269 6.1 INTBODUCTION 325
Elements325
6.11 Orbital
4.6 REFERENCES 272 6.12 Secular ottheOrbit 327
Pertulbations
dueto Commensurabilities
6.13 Perturbations 330
6.14 Long-PeriodPertubationsby OtherSatellites 332
ORBITAL EPHEMERIDES 333
6.15 PlanetocentricRectangularCoordinates
O F T H E S U NM
, OON,AND
P L A N E T S2 7 9 Orbit 336
6.'t6 The APParent
Tabulated
6.17 Calculating Values 340
5.1 FUNDAMENTAL
EPHEMEBIDES
279 6.18 Notation 342
5.11 Gravitational
Model 280 OF MABS 342
6.2 THE SATELLITES
5.12 The Ephemeris Reference
Frame 280
5.13 TheAstronomicalConstantsUsedin the Ephemerides280 6.3 THE SATELLITES OF JUPITER 345
6.31 TheGalilean Satellites345
5.2 COI\,4PUTATION OF EPHEMERIDES 281
Amalthea 349
6.32 The FifthSatellite,
5.21 Mathematical Model 281 Thebe 351
Satellite'
6.33 The Fourteenth
5.22 NumericalIntegration286 6.34 TheSixththroughThirteenthSatellites 352
5.23 Orientation
of Ephemerides288
6,4 THE RINGSAND SATELLITES OF SATURN 354
5.3 OBSERVATIONAL DATAFIT BY THE PLANETARY
AND LUNAR 6.41 The Ringsof Saturn 354
EPHEIVEHIDES290 of Saturn 356
6.42 The Satellites
5.31 OptjcalData 290 OF URANUS 368
5.32 Radar-Ranging 6.5 THE RINGSAND SATELLITES
Data 294
5.33 Spacecraft
RangePoinls 297 6.6 THE SATELLITES OF NEPTUNE 373
5.34 LunarLaserRangeData 299
6.61 Trilon 373
5.4 LEAST-SQUARES ADJUSTI\,IENT OF THE EPHEMEBIDES 3OO 6.62 Nereid 375
5.41 TheObservationalEquations 301 6,7 THE SATELLITEOF PLUTO 377
5.42 The SolutionParameters301
5.43 TheStandardDeviations303 6.8 REFEBENCES 378

5.5 NUMERICALBEPRESENTATION OF THE EPHEN/ERIDES 303 7 / PHYSICAL OFTHESUN,MOON,PLANETS

EPHEMERIDES
5.51 Chebyshev Polynomials305
5.52 Chebyshev Coetflcient
ceneration 305
ANDSATELLITES383
5.53 Interpolation
ErrorandPolynomial
Degree 306 383
7.1 INTRODUCTION
Elements
7.11 Rolational Coordinates383
andCartographic
5.6 COMPUTATION OF OBSERVATIONAL
EPHEMERIDES 307 7.12 Phases andMagnitudes388
5.61 AooarentPositions 308
5.62 Astrometric
Positions 308 7.2 PHYSICALEPHEMERISOF THE SUN 397
5.63 TransitEphemerides308
7,3 OF THE MOON 398
PHYSICALEPHEIV!ERIS
5.7 OBBITAND EPHEIVIERIDES OF OTHERBODIES OF THE PLANETS 401
7.4 PHYSICALEPHEMERIDES
WITHINTHE SOLARSYSTEM 310
7.41 Metc!ry 4O1
andComets 310
5.71 MinorPlanets 7.42 \enus 402
 CONTENTS
CONTENTS x l
xll
9.3 RISINGS. SETTINGS. ANDTWILIGHT482
7.43 Mars 403
7.44 Jupiter 403 9.31 Sunrise,
Sunset,
andTwilight483
7.45 Salurn 404 9.32 lvloonrise
andMoonset485
7.46 Uranus 405 9.33 FormulasAssociated
withRlsino
andSettino486
7.47 Neptun€ 405 9.34 lllumination
490
7.48 Pluto 406 9.4 oCCULTATTONS 494
PHYSICALEPHEMERIDES OF THE SATELLITES 407 9.41 Occultations
ol Stars 494
ot Mars 407
7.5'1 Satellites 9.42 Occultations
of Planets498
7.52 Satellitesof Jupiler 408 9.5 POLE-STARTABLES
498
7.53 Satellitesof Saturn 408
7.54 Satellitesof Uranus 411 9.51 De.ivationof the Pole Star Coefficients 501
7.55 Satellitesof Neptune 413 9.6 BEFERENCES502
7.56 TheSalelliteof Pluto 414
7.6 OFTHE ASTEROIDS 414
PHYSICALEPHEMERIDES 1O/STARSAND
STELLAR
SYSTEMS 505
7.7 REFERENCES 417 10.1 SOURCES
OF DATAON STARSANDSTELLAH
SYSTEMS505
'10.1'lCompiled
Catalogsof StellarPositions
andNrotions506
10.12StandardReference Catalogs 507
OF THESUNANDMOON 421
8 / ECLIPSES 10.13Observational
PositionalCatalogs 508
8.1 421
INTRODUCTION 10.14OtherCatalogsandLists 509
8.11 EclipseDataAvailablefromthe NauticalAlmanac
Oflice 422 10.15DataCentersandTheirFacilities 509
to the Ephemerides424
8.12 Corrections 10.2 STELLAR DATAIN IHE ASIFONOMICAL ALMANAC 5O9
8.2 THE OCCURRENCEOF LUNARAND SOLARECLIPSES 426 10.21BrightSlars 512
8.21 Overview 426 10.22VariableStars 513
8.22 Geocentric
LeastAngularSeparation426 10.23Double andMultioleStars 515
'10.24Photometric
Standards 516
8.23 Occurrence
of LunarEclipses 428
8.24 Occurrence
of SolarEclipses 431 10.25Badial-VelocityStandards518
10.26SpectralClassification519
8.3 SOLARECLIPSES 434 10.27Pulsars521
8.31 Fundamental Equations:
Introduction434 10.3 CLUSTERSAND GALAXIES 522
8.32 BesselianElements 435
8.33 Coordinatesof theObserver 441 10.31OpenClusterData 522
8.34 10.32GlobularStarClusterData 526
ConditionalandVariationalEquations446
8.35 Calculation 10.33BrightGalaxies528
of GeneralSolarEclipsePhenomena450
8.36 LocalCircumstances461 10.34Quasi-Stellar
Obiects 530

4.4 10.4 SOURCESCATEGORIZED BY WAVELENGTHREGION 531

LUNARECLIPSES 467
10.41Radio-SourcePositionalCalibrators
531
8.41 Introduction467
10.42Radio-Flux
Calibrators532
8.42 Computalions467
10.43X-RaySources535
8.5 TRANSITS 471
10.5 REFERENCES 534
8.6 REFERENCES 472
11/ COMPUTATIONAL
TECHNIQUES541
9 / ASTRONOMICAL
PHENOMENA475 11.1 INTRODUCTION
TO COMPUTING
TECHNIOUES541
9 . 1 GENERALASPECTSOF THE NIGHTSKY 475
11.2 INTEFIPOLATION ANDSUBTABULATION
546
9.2 CONFIGURATIONS OF THE SUN, MOON,AND PLANETS 476 1l.21 Introduction
andNotation546
ol the Sun,Earth,andMoon 477
11.22lnterpolation
Formulas547
9.21 Int€resting
Phenomena
11.23 Inverse lnte.oolation 548
9.22 GeocentricPhenomena478
11.24 PolynomialRepresentations 548
Phenomena481
9.23 Heliocentric
xiv CoNTENTS coNTENTs xv
11.3 PLANEANDSPHERICALTRIGONOMETRY 549 12.93Converting
Between
lstamic
Tabutar
catendar
Date
andJulianDavNumber 604
1i.4 MATRIXAND vEcroR TEcHNleuEs 552
RotationorAxesusinsMatrjces
1141 552 13.33
3llJ3[il3i3l][::li1,:l3:]:fi*3:?::::jl',ff3:ltili:l
:3:
Coordinates
11.42Spherical UsingVectors553
Coordinate
11.43Specific Transformations
555 12.10 REFERENCES606
11.5 NUMERICALCALCULUS560
Ditrerentiation
1r.s1Numericar s60 13 / HISTORICAL INFORMATION 609
11.52 NumericalIntegration 562
13.1 HISTORY OFTHEALMANACS609
11.6 srATlsTlcs 566 13..11
TheAmerican
Ephemeris 609
11.61TheAccumulation
ol Error 566 13.12TheA;erican
Ephemeris andNauticalAlmanac
613
11.62TheMethod
of LeastSquares568 19.13TheCooperative
British
andAmerican (1960to present)614
Almanacs
'117 REFERENOES 574 r3.2 HrsroRy oF TNTERNATToNAL
coopEBATroN 616
575
12/ cALENDARS 13:31
fi?ifl?lfllilliifl::""Hi"fU";".,,
12.1 INTRODUCTION 575 13.23 OtherInternational
Organizations621
12.11 AstronomicalBasesof calendars s76 13.3 HISTORICAL LIST OF AUTHORITIES 621
12.12 NonastronomicalBases of Calendarsrthe Week 577
13 31 Introduction621
12.13 calendarReformand Accuracy 578
13 32 The NauticalAlmanac(1767-1900) 622
12.14 Historical
Erasand chronology 579
13.33 The AmericanEphemeris(1855-1900) 631
12.2 THE GREGORIAN CALENDAR 580 13.34 The NauticalAlmanac,and The Ameican Ephemerls(1901-1983) 639
12.21 Rutesfor civit use 5g0 13.35 Systemof Constants(1968-1983) 656
12.22 EcclesiasticalRules 581
134 REFERENCES 657
12.23 Historyofthe cregoriancatendar 583

12.3rHE HEBREW
oALENDAR584 14 / RELATED PUBLICATIONS 667
12.31Rules 584
12.32History
of theHebrew
Calendar588 14.1 CURRENTPUBLICATIONS667
12.4 fHE ISLAMICCALENDAR 5gg 14.1'1JointPublications
of the Royalcreenwich andthe UnitedStates
Observatory
12.41Rules 589 Navalobservatory 667
12.42Historyof the lslamiccalendar 591 14 12 other Publications
of the L'nitedstatesNavalobservatory 668
14.13OtherPublications oftheRoyalGreenwich Observatory669
12.5 THE INDIANCALENDAR 591 14.14Publicationsof OtherCountries669
12.51RulesforCivilUse 591
12.52principresoftheRerigiouscarendars92 142 ASTRoNoMICALPAPERSPREPAREDFoRTHE UsE oFTHEAMERIoAN
12.53Historyof the IndianCatendar 594 EPHEIVIER|S
AND NAUTICALALI\,IANAC 670

12.6 THE CHINESECALENDAR 594 CIRCULARS 675

14.3 UNITEDSTATESNAVALOBSERVATORY
12.61Rules 595
'12.62History
oftheChinese
Calendar599 14.4 PUBLICATIONS
OFTHE UNITEDSTATESNAVALOBSERVATORY
12.7 JULIANDAYNUMBERSAND JULIANDATE 600 SECONDSERIES 677

12.8 THEJULIANoALENDAR 600 14.5 SELECTEDNAO TECHNICALNOTESOFTHE ROYALGREENWICH

'12.91Rutes 601 OBSERVATORY684
12.82Historyof theJulianCalendar 601
'12 14.6 LISTSOF APPENDICESAND SUPPLEMENTS 685
9 CALENDARcoNVERsloN ALGoRITHMS 603
14.61TheBrtishNauticarArmanac
685
12.91Converting Dayof theWeek 603 14.62TheAmericanEphemeris688
12.92Converting betweenGregorianCalendarDateandJulianDayNumber 604 andAoDendices691
14.63JointSuDotements
xvl CONTENTS

1 5 / R E F E R E N C E D A T6A9 3
15.1 Fundamental Constants (1986Recommended Values) 693
1s.z IAU (1976) System of AstronomicalConstants 696
'Iime
andStandardEpochs 698
15.3
15.4 Sun,Earth,andMoon 700
LISTOF FIGURES
15.5 GeodeticRelerence Systems 702
15.6 Planets:MeanEl€ments 704
15.7 Planets:Rotational
Data 705
'15.8 Planels:PhysicalandPhotometric Data 706
15.9 Satellites:OrbitalDala 708
't5.10Salellites:
PhysicalandPhotometric Data 710 I.23l.l Calculationof siderealtime 5
15.11Planetary Bings 712 L233.1 Variationin theequationof timethroughtheyear 6
15.12Constellation NamesandAbbreviations713
'15.'13Mathematical I .3I . I Representation coordinates I 0
of thevector/ in rectangular
Constants 714
I.32l.l Equatorialandeclipticreference planes 12
15.14EnergyConversion Factors 715
15.15Unitsof Length, Speed, andMass 716 1.333.1 Geocentric andgeodeticcoordinates 15
15.16GreekAlphabel717 1.33'7.1 Relationbetweengeographic latitudeandthealtitudeof thecelestialpole l8
15.17InternalionalSystem of Units(Sl) 718 1.361.1 Parallaxofan object 22
1.411.1 Geometricproperties ofconic sections 25
1 5 . 1 R E F E R E N C E S7 1 9
I.4l2.I Angularorbitalelements 27
GLOSSARY721
2.25.1 Meridianrelationsandtime 55
lndex 741 2.27.1 Worldmapof time zones 57

3.21.1 The eclipticandequatorat epochanddate l0O

3.21.2 Theprecession angles(r, z,r,and9e 102
3.222.1 The meanandtrueequators of date I 15
3.251.1 Light-timeaberration 128
3.252.1 Stellaraberration 128
3.26.1 Gravitationallight deflection 136
3.26.2 Light from theplanetsandstarsdeflectedby thesun 138

4.21.1 TheEafih'ssurfaceandreference surfaces 201

4.22.1 Geocentric andgeodeticlatitude 202
4.22.2 Asrronomiclatitudeandlongitude 203
4.22.3 Deflectionof theverticalon a unit sphere 204
4.22.4 Geocentric coordinates 205
cartesian
4.231.1 Altitudeandazimuth 208
4.233.1 UTM andUPSgrid zonedesignations2ll
4.242.I Majorgeodeticdatumlocations 219
4.51.1 Irngth of dayfrom I 656to 1988 267
4.51.2 Ar from A.D.700to 1600 268
4.51.3 Ar from 700B.c.to A.D.2N 268
4.51.4 logA, from 700B.c.to A.D.1980 269
4.52.1 Polarmotion,1980September 28 to l99Ohly 2'1 2'7O
4.52.2 Xand y components of polarmotionfrom theILS, 1899.9to1979.0 271
4,52.3 X andy components ofpola.motion,1846.0to1891.5 272

system,in whichfigure-induced
5.212.1 The{r7( coordinate arecalculated 283
accelerations
5.321.1 Diagramof geometryfor planetaryradarranging 295
xv
 LtsToF FtcuREs L I S TO F F I G U R E S
xviii XIX
9.331.1 Thehorizonat risingor setring 488
5.7I4.|TherclationshipbetweentheBl950.0andJ2000.0referenceframesandtheorbital
plane 314 9.34.1 Grcundilluminationfrom varioussources 491
9,41.1 Fundamental planeshowingpathof occultation 49't
9.5.1 PZStrianglefor Pola.is 499
The orbital elementsusedto describethe orbital planerelativeto a referenceplane 326
6.I 1. 1
Theend-onview of a greatcircleon the 9.51.1 Polarisnorthpolardistance2096-2105 502
6.ll .2An equivalent form of tle orbitalelements.
celestialsphereis reprcsented asa sraightline. 326
6.12.1 The Laplacianplane 328
plane 334 l1.l.l Sphericaltriangle 542
6.15.1 The satelliteorbitrcfenedto an intermediate
6.15.2 Coordinates ofthe satelliterelativeto theplanet 335 I1.3.1 Arc ofcircle 551
Planetocentric celestialsphere 337 ll.42.l Triangleon unit sphefe 555
6.16.1
6,16.2 Sphericaltriangleusedto computeU,B,andP 337
6.16.3 Sphericaltriangleusedto computeo andp P 337
6.16.4 Apparentorbitof a satelliteasFojectedon thegeocentdc sphere 339
6.2.1 Reference systemfor PhobosandDeimos 344
6.32-l Reference systemfor AmaltheaandThebe 350
6.421.1 R€ference planefor Mimas,Enceladus, Tethys,andDione 357
6.422.1 Reference systemfor Rhea,Titan,Hypedon,Iapetus,andPhoebe 361
6.422.2 Anglesneededto evaluatesolarperturbations on Titan 362
6.5.1 Reference systemfor the Uraniansatellites 372
6.61.1 Refercnce systemfor Triton 374
6.62.1 Reference systemfor Nereid 376

'7.Il.l
Thepositionof thenorthpoleandprimemeridianof a planetin Earthequatorial
coordinates 384
7.11.2 Planetocentric
andplanetographic coordinate systems 386
'7.12.1
The basicvectorsandtheangleof illumination 389
7.12.2 Thegeometricappearance of theapparent polarradius,,' 391
1.12.3 Thediskofa planetasseenby an obseNeron rheEanh 392
7.12.4 Planetocentric
unit vectorsfor pointsof intereston thediskof theplanet,andtheangles
betweentheEarth'sequatorofJ2000.0andtheplanet'sequatorof date 394

8.22.1 Geometricconstructfor determining whethereclipsewill occu. 426

8.231.1 Geomeficparameters ofthe Eanh'spenumbralshadow 428
8.231.2 Geometricpatameters ofthe Earth'sumbralshadow 429
8.2321.1 Limitingconditionsfor lunareclipses 430
8.242.1 Geometdcparameters for a partialsolareclipse,whenMoonis extemallytangent 432
8-2421.1 Geometricparameters for a partialsolareclipsewhenMoonis intemallytangent 434
8.321.1 Transformation oi geocentric equatorial coordinatesto thefundamentalplane 435
8.323.1 Components ofshadowconesin theBesselian fundamentalreference
system 438
8.323.2 Relationships amongangularsemidiametet apparent semidiameter,andparallax 439
8.323.3 Vertexangleofthe penumbra 440
8.323.4 Vertexangleofthe umbra 440
8.331.1 Angularquantities in thegeocentric equatorial plane 442
8.352.1 Relationship betweenrectangular andpolarreference systems 451
8.3556.1 Definitionof auxiliaryquanriry1, 458
8.3623.1 MagnitLrdeis thefractionofthe lineardiameterofthe Suncoveredby theMoon 4&
8.3623.2 Obscuration is thefractionof rheareaof thesolardiskobscuredbv theMoon 465

9.31l.l Theanalemmic
curve 485
 LISTOFTABLES

1.342-l Variationsin the Earth's Rateof Rotation 19

2.33.1 Techniquesfor MeasuringEanh's Rotation 62

2.45.1 Error Budgetsfor CommonView GPSTime Transfer 68
2.58.1 Time andFrequencyStepsby WWV andMC (USNO) 86

3.2ll.l Accumulated hecessionAngles 104

3.214.1 Precession Angles1984Januaryld0h 107
3.222.1 Nutation itr LongitudeandObliquity Refelredto the Mean Ecliptic of Date I l2
3.222,2 FundamentalAryuments I 14
3-224-l Conectionsto IAU 1980NutationScries I 16
3.224.2 PlanetaryTermsin Nutation,CombinedDrect andIndirect Effects I l8
3-253.1 Coeffrcients(','tl / c) for the Major Planets 132
3.26.1 ApparentDeflectionAngles 138
3.344.1 Second-OrderTerms 159
33M.2 Errors Due to NeglectingSecond-OrderTerms 160
3.58.1 SelectedStarPositionson FK4 alld FI(5 Systems l8l

4.242.1 EarthEllipsoids 220

4,242,2 GeodeticDatums 221
4.32.1 GEM-TI NormalizedCoefficients (x 106) 228
4.331.l Step2 Solid Tide CorrectionsWhen &z= 0. 30 in Step I (Using a Cutoff Amplitude of
9 x lO-\2 fot A^6k"H") 235
4-34.1 OceanTide Coefficientsfrom the SchwiderskiModel 242
4.35I . I DisplacementDue to OceanLoading (cm in amplitudeanddegee in phase) 245
4.36.1 CanesianRotationVectorfor EachPlateUsing KinematicPlateModel AM0-2 (No Net
Rotation) 250
4.37.1 Zonal Tide Termswith PeriodsUp to 35 Days 252
4.37.2 Zonal Tide Termswith PeriodsGreaterthan35 Days 253

5.214.1 Lunar Libration AnglesandRates 285

5.31l.l TraNit Cide Observationsfrom the U.S. Naval Observatorythat havebeenusedin the
JPL Ephemerides, DEI 18 291
5.322.1 Radar-RangingObservationsUsedDircctly in DEllS 297
5-332-l Mariner 9 miter Normal RangePointsto MaIs 298
5.41.1 The ObservationalData Usedfor the AdjusEnentof DE1l8/LE62 301
5.42.1 ValuesandFormal StandardDeviationsofthe ConstantsUsedin DEI l8/ LEl l8
(DE2OOil-E20o) 3O2
5.42.2 in AU andAu/day 304
The Initial Conditionsof the Ephemeridesat JED 2,140100.5 i
5.53.1 Gnnule Length andPolynomialDegreefor the 11 EphemerisBodies 307 I
xxi I
I
I
 xxii LtsToF TABLES LIST OF TABLES xxiii
5.8.1 at theepochJ2U)0(JED2451545,O)316
ClassicalKeplerianelements l2.4l.l Monrhsof TabularIslamicCalendar 590
5.8.2 Approximatemaximumerro6 ofthe Keplerianformulasovertheinterval l2.5l.l Monthsofthe IndianCivil Calendar 592
1800-2050316 12.52.1 SolarMonthsof theIndianReligiousCalendar 593
12.6l.l ChineseSexagenary Cycleof DaysandYea$ 595
6.13.1 AmongSatellites 332
The PrincipalCommensurabilities 12.6L2 ChineseSolarTerms 597
12.82.1 RomanDatingin theJulianCalendar 602
7.4L| PhysicalEphemerisPammeters fo. Mercury 402
1.42.1 PhysicalEphemerisParameters for Venus 402 13.348.1 AdoptedValuesfor Coefhcients
of Nutation 654
7.43.1 PhysicalEphemerisParameters for Mars 403
7,44.I PhysicalEphemerisPammeters for Jupiter 404
7,45.1 PhysicalEphemerisParameters for Saturn 404 15.1 FundamenralConstants (1986Recommended Values) 693
7,46.1 PhysicalEphemerisParameters for Uranus 405 15.2 IAU (1976)Systemof AstronomicalConstants 696
7.47.l PhysicalEphemerisParameterc for Neptune 406 15.3 TimeandStandard Epochs 698
1.48.1 PhysicalEphemeris
Parameters for Pluto 406 15.4 Sun,Earth,andMoon 700
7.51.1 RotationParametersfor Mars' Satellites 407 15.5 GeodeticReference Systems 702
7.52,1 RotationParameters
for Jupiter'sSatellites 409 15.6 Planets:MeanElements 7(X
7.52.2 StandardCanographic Longitudesfor Jupiter'sSatellires 409 15.'7 Planets:Rorarional
Dara ?05
'7.53.1 15.8 Planets:PhysicalandPhorometric
RotationParameters
for Saturn'sSatellites 410 Data 706
7.53.2 StandardCartographicLongitudesfor Satum'sSarellites 4l I 15.9 Satellites:
OrbitalData 708
7,54.1 RotationParamete.s
for Uranus'Satellites 412 15.10 Satellites:
PhysicalandPhotometric Data 710
7.55.1 RotationParameters
for Neptune's Satellites 413 15.ll PlanetaryRings 712
7.56.1 for pluto'sSatellite 414
RotationParameters 15.12 Constellation
NamesandAbbreviations 713
15.13 Mathematical Constants 714
8.1l3.l U.S.N.O.
SolarEctipse
Circutars424 15.14 EnergyConversion Factors 715
8,422.I Sequences
andConditionsfor ContactTimes 470 I5.15 Unitsoflrngth, Speed, andMass 716
15.16 GreekAlphabet 717
9.21I.I Timeof Commencement of theSeasons477 15.17 InremationalSystem of Units(SI) 718
9.22.I Geocentric Phenomena for whichf(r) = 0 479
9.22.2 ceocenrdcphenomena for which/,(r) = 0 479
9.22l.M sibilityCriteriafor Ceocentricphenomena 480
9.222.1 SynodicPeriodsof rheplanersandFirstAsteroids 480
9.222.2 long-PeriodCyclesofMercury Venus,andMars 4gl
9,23.1 HeliocentricPhenomena482
9.34.1 Coefficientsfor CalculatingCroundllluminatjon 4gz

Precision
of AngleandNumberof Decimalsfor TrigonometricFunctions 544
II I i
ll.l.2 The Methodof InverseUse 545
I I.1.3 RangeofPrecisionof theInverseDetermination
of an Ansle 545
I l.2l .l Differences
in TabularArsumenrs547
ll :
t Formulasfor planeandSfhericalTriangles 550
I l.5l.l Derivatives
to anOrd".oi I0 :ot
ll.61.l Accumulation ofEnor in Arithmetical
Operarions567

12.21-l Monthsofthe CregorianCalendar 5gl

l2.3l.I of yearsin the HebrewCalendar 584
Classification
12,31.2 Monthsofrhe HebrewCalendar 585
12.31.3 Terminology ofrhe HebrewCalendar 585
12.311.1 LunationConstants for DetenninirgTishri I 586
 Foreword

The Etplanatory Supplement to the Astronomical Ephemeris and the Americe,n

Ephemeris a,nd Nautical Almanac was first published in 1961 "to provide the user
of these publications with fuller explanations of these publications themselves."
This supplement was reprinted with amendments in 1972, 1974, and 1977. It was
allowed to go out of print because the International Astronomical Union decided to
introduce new astronomical constants, a new standard epoch and equinox, new time
axguments, a new astronomical reference frame, and new fundamental ephemerides,
all of which required major revisions to the supplement. In addition, The Astro-
nornical Ephemeris ar'd The American Ephem,erisand Nautical Almanoc serieswere
continued from 1981 with a new tille, The Astronomical Almanac, which contains a
revised content and arrangement and is printed only in the United States, The work
of computation, prepa,ration, proofreading, and production of reproducible material
is still shared between the United Kingdom and the United States of America.
Ma,ny changes in the arrangement of the almanarc were introduced in the edition
for 1981, and major changesin the basis ofthe ephemerideswere made in the edition
for 1984. The changes in 1981 included: the replacement of the hourly ephemeris
of the Moon by a tabulation of daily polynomial coefficientsl the introduction of
a new system of rotational elements for the planets; the extension of the scope of
the data on satellites; the inclusion of orbital elements and other data for minor
planets of general interest; the extension of the list of bright sta.rs;the inclusion of
new lists of data for other types of stars and nonstellar objectsl new explanatory
material; and a glossary of terms. The changes in 1984 included: the replacement
of the classical theories by the Einsteinian theories of special and general relativ-
ity; the replacement of ephemeris time by dynamical timescales; the adoption of
new fundamental heliocentric ephemerides based on a numerical integration of the
motion of all the planets and of the Moon; and the use of the IAU (1976) system
 FOBEWORD
xxvi
of astronomical constants, the standard equinox of J2000 0' and the FKS celestial
referencesystem.Anaccountofthesechangeswasgiveninthe39-pageSupplement
to the Astronomi.cal Almanac fot 1984, rvhich was bound with the Almanac, and is
also given here in great detail in Chapter 13'
Most ofthe text in this supplement has been written for readers who are familiar
with the principal concepts of spherical and dynamical astronomy but who require
detailed information about the data published in The Astronomical Almanac and Preface
about how to use the data for particular purposes. Similarly, the reference data
given in this supplement are presented in forms that are appropria.te to rrsers who
understand the significance of the quantities whose values are given. To a la,rge
extent, the chapters are independent of each other, but an introductory overview
has been given in Chapter 1 and a glossary of terms has been given at the end of
the volume. Referencesto textbooks and other sources of background information
a.regiven at the end of each chapter.
Preliminary proposals for the new edition of the Explanatory Supplement were
drawn up in 1979, and more detailed outlines of the Supplement were prepared in
1986 bv the staffs of Her Majesty's Nautical Almanac Office and the U.S. Nautical The primary purpose of this revised.Explanatory Supplement is to provide users of
Almanac Office. By 1988, it was evident that Her Majesty's Nautical Alnanac The Astronomi,cal Almanac with more complete explanations of the significance,
Office would not be able to participate as originally planned, and the U.S. Nautical sources, methods of computation, and use of the data given in the almanac than
Almanac Ofrce took over the entire project. can be included annually in the almanac itself. The secondary purpose is to provide
The supplement is organized by chapters and sections such that it can be up- complementary information that doesn't change annually, such as conceptual ex-
daied in the future. It is planned that future reprints will incorporate developments planations, lists of constants and other data, bibliographic references,and historical
and improvements. We hope that this new publication will prove to be even more information relating to the almanac. It is hoped that lhe Erplanatory Supplement
useful than its predecessor. will be a useful reference book for a wide range of users in the fields of astronomy,
geodesy, navigation, surveying, and space sciences, and also tcachers, historians,
and people interested in the field of astronomy.
Many users of the almanac are not the professional astronomers for whom it
is primarily designed, and so this supplement contains some explanatory material
at an elementary level; it is not, however, intended for use as a basic textbook
on spherical and dynamical astronomy. In some respects it does supplement such
textbooks since it is concerned with new concepts or new techniques.
This supplement differs in many respects from its predecessot, the Erplanatorg
Supplement to The Astronomical Epherneris and The Ameri,can Ephemeris and
Nauti,cal Almanac. Vector and matrix notations have been introduced and more
diagrams have been provided. Simple conversion tables and tables of quantities
that can be calculated directly from simple formulas have been omitted. Detailed
step-by-step examples have been omitted, and approximation methods have not
been given. Most of the text is new but historical material has been carried over for
the convenienceof those who do not have ready accessto the previous supplement.
This supplement has been prepa,red by the Nautical Almanac Office of the
United States Naval Observatory. Material has been contributed by scientists from
the Nautical Almanac Office of the Royal Greenwich Observatory, Jet Propulsion
xxvll
)oo/iii PREFACE

Laboratory, Bureau d.eeLongitudes, Time Service and Astrometry Departments

of the U.S. Naval Observatory, alxd other scientists. The authors of each chapter
have been indicated, but other individuals may have been involved in contributing,
inproving, and checkiug the material. The valuable assistatrcethat has been given
in many ways by other astronomereand Bcientistsis gratefully acknowledged.
Suggestionsfor improvement of this supplement, a.ndof The Ashpnomical Al-
monac itself would be welcomed. They should be sert to the Director, Orbital
Mecha.rricsDepartment, U.S. Nav'al Obeervatory, Washington, DC 20392.

Jomes B. Hagan
Captain, U.S. Nawy
Superinteudeut, U.S. Naval Obserratory
EXPLANATORY SUPPLEMENT
TO THE ASTRONOMICALALMANAC
 CHAPTER
1

Introduction
to Positional
Astronomy
by PK. Seidelmann
andG.A.Wilkrns

1,1 INTRODUCTION

1.11 Puroose

The Astronomical Almanac gives data on the positions and, where appropriate,
orientations of the Sun, Moon, planets, satellites, and staxs as they may be seen
from the surface of the Earth during the course of a year. A proper appreciation
of the significance of these data requires a basic understanding of the concepts of
sphericol astrononly, which explain how the varying directions of celestial objects
may be represented by positions on the surface of the celestial sphere.In addition, an
appreciation of why these celestial objects appear to move in the ways predicted in
The Astronomical Almanac reqrires an uuderstanding of the concepts of dynamical
astronorny, which provides a mathematica.l explanation of the objects' motions in
space under the influence of their mutual gravitational attractions. Spherical and
d1'namical astronomy together form what is referred to here as positional astronornA.
This text has been written for readers familiar with the principal concepts of
spherical and dynamical a€tronomy who require detailed information about the
computation and use of the data published, in The Astronomical Almanac. The
prima.ry purpose of this introductory chapter is to introduce the concepts, termi-
nology, and notation that are used throughoul The Astronomical Almanac and.lhis
supplement; rigorous definitions, formulas, and further explanatory information are
given in the later chapters of this supplement. The glossary gives concise definitions
of words pa,rticular to spherical and dynamical astronomy. The reference data are
presented in forms that are appropriate to users who uaderstand the significance
of the quantities whose values are given.

*
 EXPLANATORYSUPPLEMENT TOpOStTtONAL
I / TNTRODUCTTON ASTBONOMY 3

1.2 TIMESCALESAND CALENDARS by small periodic terms that depend on the form of the relativistic theory beiug
used.
In 1991 the International Astronomical Union (IAU) adopted resolutions in_
1.21 AtomicTimescales
troducing new timescales which all have units of measurement consistent with the
proper unit of time, the SI second. Terrestrial Time (TT) is the time reference
1.211 International Atomic Time For scientific, practical, and legal purposes the
for apparent geocentric ephemerides, has a unit of measurement of the SI second.
standard unit for the measurement of intervals of time is the SI second, which is
defined by the adoption of a fixed value for the frequency of a particular transition of and can be considered equivalent to TDT. Geocentric Coordinate Time (TCG) and
cesium atoms. Time can be measured in this unit by the use of time standards based Barycentric Coordinate Time (TCB) arc coordinate times for coordinate systerns
on processesof physics, Cesium frequency standards, hydrogen masers, ion storage that have their origins at the center of mass of the Earth and of the solar system,
devices, and other such devices are able to count seconds and subdivide them very respectively. Their units of measuremcnt are chosento be consistent with the proper
precisely. Thus, such a device can provide a timescale whose accuracy is dependent unit of time, the SI second. Becauseof relativistic transformations, these timescales
on the precision of the measurement and the stabiliiy of the device. Such a timescale will exhibit secular variations with respect to Terrestrial Time. For example, TDB
provides a measure of time for identifying the instants at which events occur; the and TCB n'ill differ in rate by approxirnately .19seconds per centurv.
interval of time between two events can be calculated as the differences between
the times of the events. The results of the intercomparisorr of about 200 frequency 1.22 Dynamical
Tlme
standards located around the world are combined to form a standard timescale
that can be used for identifying uniquely the instants of time at which events The equations of motion of the bodics of the solar system involve l,ime as the inde-
occur otr the Earth. This standard timescale is known as International Atomic penderrt variablc. Until 1960, mean solar tin.rc.q,'asused, but when it rvas recognized
Time (TAI). It is the basis for all timescales in general use. It is distributed by that the rotatiorr of the Earth u'as irrcgular, a new timescalc rvas introduced that
many difierent means, including radio time signals: navigation systems. such as the correspondedto the independentrariable. This was called ephomeristime (ET) and
Global Positioning System, LORAN C. and O\IEGA; communication satellites: u'as bascd on the motion of the Sun. Later, rvhen it I'as necessinf.to distinguish be-
and precise time standards. tween relativistic eflects rvhich cause diffcrences betrveen tirrrcscalesfrr the ccntcr
of the Ea,r'th ancl the center of the solar systcm the dynauri<:al time arguments,
Terrestrial Dynamical Time, and Barycerrtric Dynamical Titnc, u,ere introduced.
1.212 Relativistic Effects In high-precision timekeeping, and for some purposes TDT was defincd in a way that nzrintains continuity with ET. Since ET was not
in solar-system dvnamics and astrometry, it is necessary to take into account the specifiedas eithcr TDT or TDB. either can bc consideredto be the extensionof ET.
efiects of special and general relativity and to recognize, lor example, that the rate of Since TDT is defined in terms of TAI, which can be determined onlt.back to 1956.
an atomic clock depends on the gravitational potential in u4rich it is placed and that the determinatiorr of dynamical time prior. to 1g56 must be bascd on comparison
the rate will appear to depend on its motion relative to another clock rvith rvhich of observationsald theories of the urotions of the Sun, Moon, and plalets.
it is compared. In particular, one should recognizethat the independentvariable,
or timescale, of the equations of urotion of the bodies of the solar system (or of a 1.23 Rotational
Timescales
subset of thern) depends upon the coordinate system to which the equations refer.
The relationship between any such timescale and TAI, which is appropriate for use As the Earth rotates about its axis. it also moves iu its orbit around the Sun.
at sea-levelon the surface of the Earth, may be specified by an appropriate formula Thus, while thc Earth rotates once with respect to a fixed star, the Earth movcs
containing periodic terms and an arbitrary linear term. Trvo su<:htimescales have in its orbit so that additional rotation is necessarywith respc<:tto the Sun. The
been given special names, and these have been used in ?i.e Asl.rortomical Almanac rotation of the Earth with respect to thc equinox is called sidercal tirne (ST). The
since 198.1.Terrestrial Dynamical Timc (TDT) is used as the tirnescale for thc Earth's rotzrtiorrrvith respect to the Suu is the basis of Universal Time, also called
geocentric ephemerides (giving. for exarnple. apparent positiolrs $'ith respect to solar tirrx:. Since the rotatio[ of thc Earth is subject to irrcgular.forces,sidereal
the center of the Earth), whereasBarycentric Dynamical Tinre (TDB) is used as time and Univorsal Time are irregular u'ith respectto atornic timc. Sidcrealtime is
the tirnescale for the ephemerides that refer to the cerrter of lhc solar system. the hour anglc of the catalog cquinox and is subject to thc motion of the equinox
TDT differs from TAI by a constant offset, which t'as cltosen arbitrarily to give itself due to precessionand nutation. Otherwise, it is a direct measrrreof the diurnal
contimrity with ephemeris time (see Section 1.22), whereas TDB and TDT differ rotation of the Earth. Sidereal time reflects the actual rotation of the Earth and can
A SUPPLEMENT
EXPLANATORY 1 / INTROOUCTIONTO POSITIONALASTRONOMY

be determined by observations of stars, artificial satellites, and extragalactic radio

sources. On the other hand, the apparent diurnal motion of the Sun involves both
Ephemedsmeridian Local m€ridian ComputedMesn Sun (U)
the nonuniform diurnal rotation of the Earth and the motion of the Earth in its
orbit around the Sun, In practice, Universal Time is directly related to sidereal time
by means of a numerical formula. For each local meridian there is a corresponding
local sidereal tirne. The measure of the rotation of the Earth with respect to the
true equinox is called apparent sid,erealtime, Tbe measte with respect to the mean
equinox of date is referred lo as mean sidereal time. Apparent sidereal time minus
mean sidereal time is called the equation of the equinooes.

1.231 Sidereal Time The local hour angle ofthe equinox has a special significance
since it serves to specify the orientation of the celestial (equatorial) coordinate Longitude
system with respect to the local terrestrial coordinate system in which hour angle
AT
|.002738
is measured. The local hour angle of the equinox is known as the local sidereal time
(LST) and it increases by 24' in a sidereal day. The fundamental relation for some Grcenwichm€ridian X True Sun
celestial object X is Figure 1.231.1
Calculationof siderealtime
local hour angle of X = local sidereal time - right ascension of X. (1.231-1)

This mav be written in the abbreviated form

that moves around the celestial equator at a uniforrn rate. Hence, Universal Time
LHAX=LST-RAX. \r.23r-2) is defined in terms of Grecnrvich sidcreal time (GST) by an expression of thc form

The value of the local sidereal time is equal to the right ascension of the local
UT= GST RAU-12', (t.232,2)
meridian, and so may be determined by observing the meridian transits of stars
of known right ascensions,the positions of radio sources using Very Long Baseline
where the coefficients in the expression for RA U are chosen so that UT mzry for
Interferometry, or the distance from a location on Earth to a retroreflector on the
most purposes be regarded as mean solar time on the Greenlvich meridian.
Moon by Lunar Laser-Ranging. Local sidereal time can be calculated from the
sidereal time on the prime meridian (Greenwich) when the difference in geographic
longitude is known (seeFigure 1.231.1);thus 1.233 Equation offime The difierence between local mean time (LNIT) and local
apparent solar time is known as the equation of time, and the relationship is now
pwnr-."-n
local sidereal time = Greenwich sidereal time + east longitude. (1.231-3) in iLa f^rm

1.232 Solar and Univ€rsal Time The general form of the relationship between LAT = LHA Sun + l2n = LMT + equation of time, (t.233-I)
sidereal time and Universal Time may be derived by substituting the Sun for X in
the above equation. The local hour angle of the Sun is, by definition, 12 hours less although the equation of time used to be regarded as the correction to be applied to
than local apparent solar time (LAT) and so apparent time to obtain mean time. The principal contributions to the equation of
time arise from the eccentricity of the Ea,rth's orbit around the Sun (which causes
LAT = LST - RA Sun* 121 (r.232-r) a nonuniformity in the apparent motion of the Sun around the ecliptic) and the
inclination of the plane of the ecliptic to the plane of the equator. The equation of
The right ascensionof the Sun does not varv uniformly with time nor does the time varies through the year in a smoothly periodic manner by up to 16 minutes,
Sun move on the equator, but it is possible to introduce the concept of a point U as shown in Fieure 1.233.1
 EXPLANATORYSUPPLEI\,4ENT 1 / INTRODUCTIONTO POSITIONALASTRONOI!,IY 7

18 months. The difference betrveen UTC and TAI is always an integral nuurhel of
seconds.UTC is tra,irrtaincd irr close agreernent, to better than one second, r'ith UT
by introducing extra secondsl knou'n as leap seconds.to UTC, usually at the errclof
! Eb
e - Z E: E € the last day of Ju[e or December. The Earth is divided into standard-time zones
Eegnq"a9:P in which the timc kcpt is that of a standard meridian (multiples of 15" Iongitude).
<>:=<362d Thus, the local noon at any place in the zone is near twelvc noorr ol the standard
tl6 m l9s

t +l5m30s
+l4m0s

+l1m36s
time. These standard times usually differ from UTC by integral numbers of hours.
In summer, the tine may be adranced to increase the hours of dal.light in the
eve[ing.

t
+9m43s
1.241 Greenwich M€an Time (GMT) In the past, the term Greeuwich Nleau Time
(GMT) has been used. It is the basisof the civil time for the United Kingdorn ancl,
+ 5 ml 3 s as such, is related to UTC. However, in navigation terminology, GN,IT has becn used
+3m,15s +4m29s
^ as Universal Time. For precise purposes it is recommended that the term GNIT not
+2m43s + 2 m 3 3s
be used, since it is ambiguous.
0m0s Prior to 1925, GMT rvas measured for astronomical purposes from loon to
0m35s
2m39s noon, so that thc date would not change in the middle of a night for an observer in
-3 m 16s 3m23s Europe. In 1925 that practicc was discontinued, and GNIT was then measured from
4n2'7s
L , 1 m3 6s
5mzl7s midniglrt to midnight. Thus, care rnust be taken in using time refelences before

I
-6m21s 1925.
-9m 12s
1.25 TheEnumeration
of Dates
l2m,15s
13m24s 1.251 Civil Calendars The alterrration of day and night is a clcar physical phe
1 4 m1 4 s
nomenon that is repetitivc and countable; so the solar day is the basic unit of all
Figure 1.233.1
Variationin theequationof time throughtheyear calendars. Some calendars usc the lunar month and/or solar year for longcr units,
but these periods arc neither fixed nor made up of integral mrmbers of days. Nlore-
over, thc length of thc lunar month is not a simple rational fraction of the solar
year. However, the cycle of u'eeks, each of seven namcd days, is very widely used,
1.24 Coordinated Time(UTC)
Universal and it continues independently of the enumeration of days in the calendar'.
Although TAI provicles a continuous, uuiform, and precise tirncscalc lbr scientific The Gregorian calendar is now cornmonly used throughout most countlics ofthe
referencc purposlrs, it is not conl,enicrrt lbr general use. In cvcrl.clay life it is more world for the identification of solar days. It was derived from the .Julian calendar,
couvenicnt to use a svslem of tirnr:scalesthat r:orrcsporrd to the alternation of clav so named because its system of rnonths was introduced by Julirrs Caesar' Thc
and niglrt, apply over fairly widc areasj and carr bc easily related to each other and present system for the enumeration ol ycars in the Julian calcndar was introduced
to TAI. In these tirrrcscales,the numelical cxpression,or rneasure.of the time ol only gradually after A.D. 500: the extension of the present systernto carlier years
an eve[t is given in the convcntiona] forni of 1.cars,months. da1's, hours, minutes. is known as the Julian proleptic calcndar. The change from the Julian calendar
seconds, and decirnals of scconds i.e.J as a calcndar date aDcl time of day. 'Ihe to the Gregorian calendal took placc in most countries in rvestertt Europe (alld
standard tirrre on the prime rnericlian is knorvn as Coordinatcd Unil.ersal Time in their colonies) between 1583 and 1753. However, rnany other countries did not
(UTC). UTC is an al.omictimcscalethat is kept in closeagreenent with Urrivcrsal changc systems until much later, and at times both systems were used concurrently.
Tine (UT), u'hich is a measule of the rotatior of thr: Earth on its axis. The rate One rmrst be very careful when using astronomical data published at times and in
of rotation of the Flnrth is not uniform (with respcct to atomic time), and the places where the calendrical svstetn in use may be in doubt. Nlany other calendars
difference betq'cen TAI arrd tlT is incrcasing irregularly by about I second every have been used, and may still bc used for some purposesJin other countries Such
 EXPLANATORYSUPPLEMENT ASTRONOMY
TOPOSITIONAL
1 / INTRODUCTION 9
8

calendars often depend on various astronomical phenomena and cycles, but are noq' context in which the position is to be used. Each coordinate system corresponds to
primarily of religioussigniflcance a particular way of expressing the position of a point with respect to a coordinate
A particular instant of time may be identified precisely by giving the calendar frame, such as a set of rectangular axes. Often the coordinate frames with respect
date (assumed to be Gregorian unless otherwise specified) and the time of day in a to which observations are made difier from the coordinate frames that are most
specified time sYstem. convenient for the comparison of observational data with theory ln general, each
position is represented by a set of three coordinates that specify the position of the
1,252 JulianDate For many astronomical purposes it is more convenient to use a object with respect to a particular coordinate frame. In many cases,the distance of
continuous count of days that is known as the Julian day number. The day number the object is not knownl so two coordinates are sufficient to represent the direction
is extended by the addition of the time of day expressed as a decimal fraction of to the object, although three direction cosines may be used
dav, to give the Julian date (JD). It is intportant to note, however, that the integral In general, an object is moving with respect to the coordinate frame' and the
values of the Julian date refer to the instants of Greenwich mean noon (since the coordinate frame is moving and rotating in "space." Therefore, it is necessary to
system was introduced when the astronomical day began at noorr rather than at specify the time to which the three spatial coordinates refer and the time for s'hir:h
midnight); correspondingly, the Julian date for 0h UT always errds in ".5" This, the coordinate frame is defrned. These two times may be the same, but are often
and the fact that current Julian dates require seven digits to express the integral difierent. Times may be expressed in terms of different timescales. Both spatial
parts, make the Julian date system inconvenient for some prtrposes- The valtre of coordinates and times rnay be expressed in a variety of units As a further compli-
JD minus 2400000.5is sometimes used for current dates: it is knorvn as the modified cation, positions may be of several kinds' according to whether or not alloll'ances
Julian date (NIJD). The Julian date system may be uscd rvith TAI, TDT, or UTC, have been made for aberration, diurnal rotation, refraction, and other factors that
so when the difierence is significant, the particular timescale should be indicatcd. afiect the direction in which au object is observed. Ephemerides and catalogs that
represent high-precision positions and properties of astronomical objr:cts at given
1.253 B€ss€lian and Jutian Years For some astronomical purposes, such as thc times must always be accompanied by a precise statement that specilies all of these
specificationofthe epochofa celestialcoordinatesvstem,it is convenientto measulc various factors, if the numbers are to be used properly to their full precision'
time in )€ars and to identify an instant of time by giving the year and the decimal A coordinate lrame is usually represented by three mutually perpendicular
fraction of the year to a few places. Such a system was introduce(l by Bessel' and is (rectangular) axes, and is defined by specifying its origin, a fundamental teference
still in use, but it has two disadvantages. First, the Iength of the year varies slowly, plane (the .ry-plane), a direction in that plane (the r-axis), and the positive (or
and second.the instants at the beginning of the years (.0) do not couespond to north) side of the plane. Other fundamental planes or directions may be used; for
Julian dates, which are convenient for use in dynaurical a.stronomy. The Bessclian example, the direction of the a-axis (i.e., the normal to. or pole of. the xy-plane)
system has been replaced by a new system in which 100 years is exactly 36525 days may be specified,or the.t-axis may be specifiedas the line of intersett'ion'or node,
(or 1 Julian century) arrd in which 1900.0 corresponds exactly to the epoch 1900 of two reference planes.
January 0.5, lrom rvhich time interval q'as reckoned in the principal theories of the The vector r from the origin to the object may be represented by rectangular
motions of the Sun, \{oon, and planets. The old ancl ne$, systems are distinguished coordinates(x,r,;); that is, by the projections of the distance r on thc tlrree axes
when necessary by the use of the prefix letters "8" and "J." Thc standard epoch It can also be representecl b1' spherical coordinates, in u'hich directiotr is usualll'
that is now recommended for use for new star cataloss and theorics of motion is specified by the longitudinal angle (.\) in lhe ry-referencc plane and the latitudinal
angle (B) from the reference plare (see Figure 1.31 1). Less frequeutly thc polar
angle (90" p), or complement of the latitudinal angle, is used; the prefix "co" may
J2000.0= 2000 January 1.5 = JD 2451545.0TDB. ( 1.2s3-1)
be added to the name in such cases. The lorrgitudinal and latitudilral aDgles are
given difierent names and symbols in <lifferent coordinate fi'aures lfthe longitudinal
angle is measurcd in the negative (or left-hauded) sensewith respect to the z axis'
1.3 CELESTIAL
ANDTEBRESTBIAL
COORDINATES
the prefix "co" may be used to indicate thc complement to 360'
'I-31 The geometric transformatiotrs of coordinates between different coordinate sys-
Coordinate
Systemsand Frames tems, and the allowances for physical effects, may be calried out using the tech-
In astronomy, it is neccssary and convenient to represent the position of an object, niques of spherical trigonometry or those of vector and matrix algebra' Generally'
such as a star or a planet, in several dilTerent coordinate systems according to the matrices would be the preferred method, but some transformations are easier to
 10 EXPLANATORYSUPPLEMENT 1 / INTRODUCTION
TOPOSITIONAL
ASTRONOMY 'lt

(3) selenocentric: viewed or measured from the center of the Moon.

(4) planetocentric: viewed or measured from the center of a planet (with cor-
responding designations for individual planets).
(5) heliocentric: viewed or measured from the center of the Sun.
(6) barycentric: viewed or measured from the center of the mass of the solar
system (or of the Sun and a specified subset of planets).

The principal celestial reference planes through the appropriate origins are as
follows:

(1) horizon: the plane that is normal to the local vertical (or apparent direction
of gravity) and passesthrough the observer.
(2) local meridian: the plane that contains the local vertical and thc direction
of the axis of rotation of the Earth.
(3) celestial equator: the plane that is normal to the axis of rotation of the
Earth and passes through the Earth's center.
(4) ecliptic: the mean plane (i.e., ignoring periodic perturbations) of the orbit
of the Earth around the Sun.
(5) planet's meridian: a plane that contains the axis of rotation of thc planet
r..t.,
",j,""
Representation
of thevectorr in rectangular
coordinates
and passesthrough the observer.
(6) planet's equator: the plane that is normal to the axis of rotation of the
planet and passesthrough the planet's center.
(7) orbital plane: the plane of the orbit of a body around another (e.g., of a
explain using spherical trigonometry. The concept of the celestial sphere will be planet around the Sun or barycenter).
used for purposes of explanation, with the arcs on the sphere representing angles (8) invariable plane or Laplacian plane: the plane that is normal to the axis of
between directions. The center of the sphere can be Iocated in many diflcrent places, angular momentum of a system and passesthrough its center.
but in most casesthe sphere will be used to illustrate a particular reference frame (9) galactic equator: the plane through thc central line of thc local Galaxv
with a specific origin. It must bc remembered that r'hen an objcct represerrted (NIilky Way).
on the sphere is changing radial distances from thc center. these changes must be
incorporated into the mathematics of the computation.

1.32 Celestial
Coordinate
Systems
1.311 Coordinate Designations Many dilTerent <xrordinate systems arc used to
specify the positions of celestinl objects. Each system depends on thc choice of 1.321 Equatorial and Ecliptic Frames The line of intersection of the mean plane
coordinate frame and on thc way of specifying coordinates with rcspect to the of the equator and the ecliptic dcfines the direction of the equinox (^l). Using
frame. The term celestial coordirrate frame is rrsed here for a coordinatc framc that this direction as the origin. the right ascension (a) is measured in the plane of
does not rotate rvith the Etrrth (or other plauet), so that the coordin.rtcsof stars, the equator and celestial (or ecliptic) longitude (.\) is rneasured in the plane of the
for example, change only slou'ly with time. (Frames that rotate with thc Earth are ecliptic (seeFigure 1.321.1).Right ascension,like hour angle, is usually expressedin
discussedin Section 1.33.) time measure from 0h to 24h, and both right ascension and longitude are measured
The designations used to indicatc the principal origins of celestial coordinate in the positive (or right-handed) sense. The contplement of right ascension with
framcs are as follows: respect to 24h is known as sidereal hour angle (SHA); in navigational publications
it is usually expressedin degrees.Declination is measured from the equatorial plane,
(1) topocentric: viewed or measured from the surface of the Earth. positive to the north, from 0o to 90". Celestial latitude is measured from the ecliptic
(2) geocentric: viewed or measured from the center of the Earth. plane, positive to the North, from 0" to 9O'.
 EXPLANATORYSUPPLEMENT 1 / TNTRODUCTTON
TOPOSTTTONAL
ASTRONOMY 13
12
rr Poleof planeO,r,)r,equator The motion of the equinox due to precession gives rise to an almost steady
increase in the celestial longitudes of all stars at the rate of about 50" per year, but
the latitudes are hardly affected. The corresponding changes in right ascensionsand
declinations depend on the position of the stars in the sk1,.The nutational motion
has an amplitude of about 9"; it is usually represented as the sum of two componelts.
one in longitude and one in obliquity. There is also a small rotation of the plane
of the ecliptic that gives a precession of the equinox of about 12" per century and
a decrease of the obliquity of the ecliptic of about 47" per century; this is known
as planetary precession since it is due to the effects of planetary perturbations on
Equarorof zr the motions of the Earth and Moon. These motions of the equatorial and ccliptic
coordinate frames can be modeled; however, some parameters have to be determined
from observations, since, for example, the internal structure of the Earth is not
known well enough.
In principle it is possible to obtain a standard celestial coordinate frame that is
fixed in space by using the frame that is appropriate to an arbitrarily chosen instant
of time, which is kno*'n as the standard epoch. The positions of the equator and
equinox for such a standard epoch cannot be observed directly, but must be specified
by adopting a catalog of the positions and moiions of a set of stars or other celestial
Figure 1.321.1 objects that act as reference points in the sky. The standard cpoch is now J2000.0,
planes
Equatorialandeclipticreference although the epoch 81950.0 has been used for much of this ccntury. Therc can bc
diflerent standard catalogsl for example, one is based on the positions of very distant
The equinox is at the ascending node of the ecliptic on the equator; this is quasars that have been measured by the techniques of Very-Long Baseline Radio
the direction at which the Sun, in its annual apparent path around the Earth, Interferometry (VLBI); another is the FKS catalog of the positions and motions
crosses the equator frorn south to north. It is also referred to as "the first point of a much larger number of bright stars. The relative orientation of these diflerent
of Aries," and it is the vernal (spring) equinox for the Northern Hemisphere. (See frames is determined from observations. Standard expressions for precession and
Figure 1.321.1.)The inclination of the plane of the ecliptic to the plane of the nutation are used to connect positions for the standard frames (r'hich are for thc
equator is known as lhe obli,quitll of the ecliptic (e). standard epoch of J2000.0) with those for the mean and true fraures of other dates.
The equator and ecliptic are moving because of the efiects of perturbing forces A statement of the celestial coordinates of a star or other celestial object must
on the rotation and motion of the Earth. Hence, the equinox and obliquity change always specify precisely the coordinate frame to which they refer. It must also bc
as a function of time; so these celestial coordinate frames must be carefully defined recognized that the coordinates that represent the position (direction) of a celestial
in such a way that they can be related to a standard frame that may be regarded object depend on the position and motion of the observer as well as on the origin
as being fixed in space. and orientation of the coordinate frame to which thev are referred.

1.322 Precessionand Nutation The celestial pole, which is in the direction of the
axis of rotation of the Earth, is not flxed in space, but changes because of gravita- 1.33 Terrestrial
Coordinate
Systems
tional forces, mainly from the Moon and Sun, that act on the nonspherical distri-
bution of matter rvithin the Earth. The quasi-conical motion of the mean celestial For astronomical purposes it is necessary to be able to specify the position of an
pole around the pole of the ecliptic is known as lunisolar prece$ion. The much observer, or other object, at or near the surface of the Earth in a geocentric coor-
smaller motion of the actual, or true, celestial pole around the mean pole is known dinate frame that rotates with the Earth and that may be related to the geocentric
as nutation. The period of the precessional motion is about 26,000 years; the prin- celestial frame used for specifying the positions of the "fixed" stars. It is also nec-
cipal period of the nutation is 18.6 years. The motions of the celestial pole are essary to set up topocentric coordinate frames that may be regarded as fixed with
accompanied by corresponding motions of the celestial equator and of the equinox. respect to the surface of the Earth and that may be used in the measurement of the
14 EXPLANATORY
SUPPLEI,IENT 1 / INTRODUCTIONTO POSITIONALASTRONOI\/Y tc

positions ol celestialobjects. The direction of the axis of rotation of the Earth pro-
vides an observable reference axis for both geocentric and topocentric frames, and
the direction of the vertical (or apparent direction of gravity) provides an additional
observable reference axis for topocentric frames.

1.331 Horizon Reference The lines of intersection of the plane of the meridian to spheroid

with the planes of the horizon and equator define the directions from which azimuth
(A) and local hour angle (h) are measurcd. Azimuth is measured in the plane of
the horizon from the north. incrcasing in positive value tou'ard the ea.st.Local hour
angle is measured in units of time: t hour for each 15' positive to the west $'ith
respect to the local meridian.
The latitudinal angles t'ith respect to the horizorr and equator are known as Equator
altitude (a) and d,ecli.natiort.(6). Altitude is measured positivcly ton'ard the zenith;
in astronomv the zenith distance (z = 90o a) is more generally used. Declination
(6) is measured lrom the equator, positive toward the north pole of rotation. The
zenith dislance of the north pole, t'hich is the same as the codeclination of the local
vertical (or zenith), is equal to the geographic colatitude ofthe point of observatiorr.
This relationship is the basis of the astronomical methods lor the determination of
geographic latitude.
Figure 1.333.1
Geocentricandgeodeticcoordinates
1,332 Shape ofthe Earth For some astro[ornical purposes the Earth may be re-
garded as a sphere, but often one must treat the Earth as an oblatc ellipsoid of
revolution. or spheroid, and may need to takc into account other departurcs frorn
sphericity. In particular, the height above mcan sea Ievel of a point on the Earth is from I 180". Longitude may be expressed in time units from 0n to 24h for some
measured from an irregular surface that is known as the geoid (see Section 1.336). applications. Latitude is measured from the equator (0") to t 90', positively to the
Such heiglrts satisfy the condition that water flows dou'nhill under gravity. There arc north.
several svstcms of terrestrial coordinates; so any sct of precise coordinates should The Earth is not a rigid body, but is subject to secular and periodic changes
b e r c c o m p a n i p db r a p r e c i s cd e . c r i p t i o no f t h e i r b a . i s . in shape and in the distribution of mass. The crust of the Earth mav be consid-
The oblateness of the Earth is a consequenceof the rotation of the Earth the ered to consist of plates that move slowly over the mantle and that are subjcct to
axis of rotation coincides on average with the axis of principal momerrt of inertia, deformation. In these circumstances, the coordinatc frame is realized by the adop-
u'hich is usually referred to as the orls of f,gure. The axis of rotation and the axis tion of values for the positions and motions of a sct of primary reference points
of figure do not coincide exactly, but the axis of rotation moves slowly in thc Earth on the surface of the Earth. These values are chosen in an endeavor to satisfy cer-
around the axis of figurc in a quasi-periodic motion with a maximum amplitudc of tain conditions; for example, so that the prime meridian nominally passesthrough
about 013 (or about 9m on the surface of the Earth). This is referred lo as polar Greenwich and so that there is no net rotation of the primary points with respect to
motion. The equator of the Earih passesthrough the centcr of mass of the Earth. the frame. In practice, different techniques of observation and analysis give rise to
reference frames that differ from each other, and from the standard frame, in origin,
1.333 Geocentric Coordinates Geocentric coordi,natesare coordinates that refer orientation, and scale. Moreover, the z-axis of the standard geocentric terrestrial
to a coordinate frame whose origin is at the ccnter of mass of the Earth arrd whose frame is not aligned with the Earth's present axis of figure, since it was chosen to
fundamental planes are the equator and an (arbitrary) prin.re meridian through give continuity with the axis represented by the adopted coordinates of the original
"Greenwich" (seeFigurc 1.333.1).This frame definesa system of rectangular coor- set of monitoring observatories. This axis corresponds to the Conventional Inter-
dinates (;r,y,z) or of sphericalcoordinates,(longitudc (I/), latitude (rp'), and radial national Origin of the coordinates of the pole, and so may be referred to as the
distance (p)). Longitude is measured positively to the east and usually designated conaentional pole,
,16 EXPLANATORYSUPPLEMENT 1 / TNTRODUCTTON
TOPOSTTTONAL
ASTRONOMY 17

Geodetic coordinatesare coordinates (longitude, Iat- Geographic latitude is also defined in terms of the local vertical and the axis
1334 Geodetic Coordinat€s
a.nd height) that refer to an adopted spheroid whose equatorial radius and of rotation, and so is not equal to the geodetic latitude. The inclination of the
itude,
(or equivalent parameter) must be specified. Geodetic longitude (,\) is local vertical to the normal to the reference spheroid is known as the deflection
flattening
geocentric longitude (,\'), and these two coordinates of the vertical. The corresponding difierence between the geographic and geodetic
defined in the same way as
coordinates may be as much as l' (that is up to 4" in time-measure). The motion of
a,re equal to each other apart from any error in the adopted origin of geodetic
the axis of rotation within the Earth causes the values of geographic longitude and
longitude. Geodetic latitude (9) is equal to the inclination to the equatorial plane
latitude to va.ry by up to about 013 over a period of a few years. The daily tidal
of the normal to the spheroid (see Figure 1.333.1). Geodetic Iatitude may differ
variations in the direction of the vertical may reach about 0101.
from geocentric laiitude (9') by up to l0' in mid-latitudes. Geodetic height (h) is
the distance above the spheroid measured along the normal to the spheroid. (See
Section4.21.) 1.336 Height and Geoid The height (or altitude) of a point on (or near) the sur-
Geodetic coordinates mav be formally referred to a,n internationally adopted face of the Earth depends on the reference surface to which it is referred. Geode-
spheroid whose origin and fundamental planes are those specified for geocentric tic heights are measured with respect to a specified geodetic datum (or reference
coordinates. Generally, however, they are referred to a spheroid that has been fitted spheroid), but the more commonly used geographic heights are measured with re-
to the geoid over a particular region. The center of such a spheroid may difier spect to a surface that is known as the geoid. The geoid is a particular equipotential
significantly from the center of mass of the Ea"rth, and the fundamental planes surface in the gravity field of the Earth and is commonly referred to as the mean-
may not be parallel to those of the international spheroid. The orientation of the sea-level surface. The geoid is an irregular surface that may lie above or below
reference frame of a regional spheroid is specified by the adopted coordinates of the international-reference spheroid by as much as 100m; this difierence is known
the origin point, which is usually a geodetic leference station near the center of as the "undulation of the geoid" and should be taken into account in the precise
the region concerned. Geodetic coordinates may also be referred to globally defined derivation of geocentric coordinates when the height above mean sea level is given.
spheroids whose origins and fundamental planes may be found to difier from the For very precise studies, it is also necessary to take into account the variations in
ideal specification. Usually the term datum is used in geodesy for regional or global height due to Earth tides and ocean loading (see Section 4.3).
referenceframes such as the ones given in Section 4'24 for the international spheroid
and also for certain global and regional spheroids. In certain casesup-to-date values
of these parameters are given in The Astronomical Alm(Lnac. 1.337 Topocentric Coordinate Systems There are two main topocentric coordi-
nate frames. One is based on the direction of the local vertical, which defines the
1.335 Geographic Coordinates Tl;.e letm geographiccoordi'natesis trsedfor terres- plane of the horizon; the other is based on the direction of the Earth's axis of ro-
trial longitude and latitude when these are determined by astronomical observations tation, which defines the plane of the celestial equator. The local meridian, which
with respect to the celestial pole and the local meridian through the Iocal verticall contains the direction of the local vertical and the direction of the axis of rotation,
and for height above the geoid, commonly referred to as height above mean sea is common to both frames.
level. The letms astronomical coord,inatesarrd,teffestrial coordinates are also used The angular coordinates in the topocentric coordinate frame of the horizon
for such coordinates. and local meridian are known as azimuth and altitude. Azimuth (A) is measured
The geographic longitude of a point may be defined as the angle between the from 0o to 360" from north in the direction of east; that is, in the negative sense
plane of the astronomical meridian through the point and the plane of the prime with respect to the direction of the zenith. (In some circumstances azimuth may be
meridian through Greenwich. The astronomical meridian is the plane that contains measured from the south, and so the convention that is used should be stated to
the direction of the local vertical and the direction of the line through the point that avoid ambiguity.) Altitude (c) is rneasured positively from the horizon toward the
is pa,rallel to the axis of rotation of the Earth. The Iocal vertical is a,fiectedby local zenith; however, in astronomy, the zenith distance (z = 90o - a) is more commonly
gravity anomalies and by the varying gravity fields of the Sun, Moon, and oceans. used. Both the altitude and the azimuth of any celestial object change rapidly as
Therefore, the astronomical meridian is not precisely the same as the geodetic the Earth rotates, and therefore, for many purposes it is more convenient to use the
meridian that passes through the point and the actual axis of figure through the topocentric frame of the celestial equator and the local meridian. The direction of
center of the Earth. Consequently, the geographic and geodetic longitudes of a point the north celestial pole is pa,rallel to the Earth's axis of rotation, and the altitude
difier slightly. Geographic longitude is equal to the difierence between (observed) of the pole is equal to the geographic latitude of the point of observation. This is
local sidereal time and Greenwich sidereal time at any instant. illustrated in Figure 1.337.1, which shows the Earth as a sphere.
18 EXPLANATORY
SUPPLEIT,lENT TOpOStTtONAL
I / TNTRODUCTTON ASTRONOITY 19
in advance, the position of the pole is determined from regular observations as an
international service. In a celestial coordinate system, the axis of maximum moment
of inertia moves around the axis of rotation in a complex motion in which an annual
component and a 14-month (Chandler) component create beats. The amplitude of
North celestial
pole the quasi-circular motion varies between 0105 and 0'i25 in a six-year cvcle.

Zenith 1,342 Length of Day The variations in the Earth's rate of rotation are, at present,
unpredictable and are due to several different causes. There is a long-term slowing
down of the rate of rotation that appears to be due largely to ihe efiect of ,idal
of celestial equator
Earth's friction. Therc are irregular changes with characteristic periods of a few years that
axisof
rotation appear to be due to an interchange of angular momentum between the mantle and
the core of the Earth. There are also seasonal and short-period irregular changes
Plane ofhodzon that appear to be due largely to exchanges of angular momentum betrveen the
solid Earth and the atmosphere. The total effect on the orientation of the Earth in
space is determined from observations by an international monitoring service that
Planeof Earth'sequator provides future predictions based on the observations.
These variations are usually represented by the corresponding changes in the
Figure 1.337.1
latitudeandthe altitudeof thecelestialpole
Relationbetweengeographic Iength ol the (universal) day (with respect to a day of 86400 SI seconds) and
by the accumulated difference between UT and TAI (or ET before 1955). The
latter difference in time may be expressed as a corresponding difference in angular
orientation of the Earth. Table 1.342.1shows the orders ofmaenitude ofthe different
In this topocentric equatorial frame, the angular coordinates are known as Iocal
variations.
hour angle (LHA, or h) and declination (6). LHA is measuredfrom 0" to 360o,or
The accumulated efiect of the changes in the rate of rotation can be seen most
from 0h to 24h, from south in the direction of west (i.e., in the negative sensewith
clearly in the geographic longitudes and times at which certain phenomena, such
respect to the direction of the north celestial pole), and the LHA of a celestial object
as eclipses of the Srrn, were observed in early historv. The accumulated effect of
increases by about l5o, or ln, for every hour of time. Declination is measured from
the changes in the rate of rotation has amounted to two hours over the last two-
the equator to the celestial pole, positive toward the north. Codeclination, measured
thousand vears. This q,as a source of much confusion when the rotation of the
from the pole toward the equator, is sometimes used instead of declination. The
Earth provided the standard timescale for all purposes. It is now recognized that
declination of a celestial object is not changed by the diurnal rotation of the Earth.
sidereal time and Universal Time are best regarded as measures of angles that $erve
to specify the orientation of terrestrial coordinate systems with respect to celestial
1.34 The Botation of the Eadh

The connection between a local coordinate frame (defined with respcct to the Earth
at the poilt of observation) and a celestial coordinate frame is complicated by
Table 1.342.1
three aspects ol the rotation of the Earth: the precession and nutation of the axis Variationsin the Earth's Rate of Rotation
of rotation (Section 1.322), the motion of the pole of rotation within the Earth
(Section 1.341),and the variability of the rate of rotation (Section 1.342). Characterislic Changein Difference Difterence
Time Length ol Day in Tlme in Orientation

1.341 Polar Motion Measurements of the zenith distances of the celestial pole of a few days 0 . 1m s 1m s 0t01
'100
rotation at different places and times show that the Earth's axis of rotation is not days 1ms 0.2s 3',
fixed within the Earth, but moves around the axis of maximum moment of inertia. a few years 3ms 10s 3',
This phenomenon is known as polar molion. Since it cannot be predicted accurately
20
EXPLANATORYSUPPLEMENT To posrroNALAsrBoNoMy
1 / rNTRoDUcroN 21

coordinate systems. They must be determined from observations, although man- the translations of the origin may be treated as srnall rotations corresponding to
made clocks can be used for interpolation and extrapolation. small parallactic shifts in direction.
The standard values of the coordinates of the pole that are required in Stage 3
between
Terrestrial
andCelestial
Coordinates are published in arrearsl "quickJook" and predicted values are also made available.
1.35 TheConnections
Indicative values are tabulated at a wide interval in The Astronomical Alrnanac (p.
In studies of the distribution and motion of celestial bodies it is necessary to have K1o).
procedures for the transform allon, or reduct'ion, of "observed" topocentric coor- For Stage 4, the required value of the Greenwich sidereal angle at the time
dinates, which are referred to a terrestrial frame that rotates with the Earth, to of observation may be calcrrlated in three steps from the UTC of the instarrt of
(or from) "reduced" barycentric coordinates, which are referred to a celestial frame observation. The first step is to apply the dillerence between UT and UTC values
that is fixed in space.The generally adopted procedure uses an intermediate system to the value of UTC. Standard values are published in arrears, but a quick-look
of geocentric coordinates, which are referred to a celestial equatorial frame of date or predicted value may be of sufficient accuracy. The value of UT is then used
that is defined by the current directions of the Earth's axis of rotation and of the to compute the values of the Greenwich mean sidereal time from a polvnomial
equinox. The principal stages of the full reduction are as follows: expression. Finally, the equation of the equinoxes is applied to give the Grcenwich
apparent sidereal time, which is converted to angular measure to give the rcquired
Stage 1: Rotation of the terrestrial frame from the horizon to the equator and Greenwich sidereal angle.
from the local meridian to the prime meridian. The values of the nutation in longitude and obliquity that are required in Stage
5 may be computed from harmonic series that are based on an adopted theory
Stage 2: Translation of the origin of the terrestrial frame from the observer to the of nutation. In some circumstances it may be appropriate to appll' further srnall
geocenter-a change from UTC to TDT may be required. conections obtained from the analysis of reccnt VLBI observations.
The values of the precession that are required in Stage 6 may be computed
Stage 3: Rotation of the terrestrial frame for polar motionl i.e., from the conven- from adopted expressions for the angular quantities concerned. Stages 5 and 6 may
tional pole to the true celestial pole (or celestial ephemeris pole). be combined in a single rotation matrix; daily r,alues of the matrix arc published
in The Astronomical Almanac (pp. B45 B59).
Stage 4: Rotation of the terrestrial frame around the true pole from the Green- The rectangular coordinates ofthe geocenter with respect to thc barycenter ma5,
wich meridian to the celestial frame whose prime meridian passesthrough the true be obtained from published ephemerides, such as that printed in The Astronomical
equinox. (The angle of rotation varies rapidly as the Earth rotates, and is known Almanac (pp. Baa B58), which gives both position and velocity.
as the Greenwich apparent sidereal time.)

1.36 Effecls of the Position and Motion of the Object and Observer
Stage 5: Rotation of the celestial frame for nutation from the true pole and true
equinox to the mean pole and mean equinox of date. I.361 Parallax The direction in which an object is seen depends on the position
of the observer as well as on the position of the object. The change in the direction
Stage 6: Rotation of the celestial frame for precession from the mean equator and of a celestial object caused by a change in the position of the observer is knorvn as
equinox of date to the standard (mean) equator and equinox of J2000.0. a parallactic shift, The arnount of change in direction corresponding to a standard
linear change in the position of the observer, across the line of sight, is known as the
Stage 7: Translation of the origin from the geocenter to the barycenter of the Solar parallax of the object it is an inverse measure of thc distance of the object from
System a change from TDT to TDB may be required. the observer (see Figure 1.361.1). For objects within the solar system the standard
linear change is the equatorial radius of the Earth. For the stars and other such
Stage 1 requires a knowledge of the geographic longitude, latitude, and height distant objects, however, the standard linear change is lhe astronornical unit of
of the place of observation. If appropriate, the efiect of the tides on the deflection d'istance (tu\ or, in efiect, the mean radius of the Earth's orbit around the Sun.
of the vertical and on the height may have to be taken into account. The basis of the parallax can be deduced immediately from a knowledge of the
Stage 2 requires a knowledge of the geodetic longitude, Iatitude, and height for type of object. The parallax of an object in the solar system is usually referred to
reducing the coordinates of very close objects. For distant objects, such as stars, as a horizontal parallax, since the actual parallactic shift takes this value when the
 EXPLANATORYSUPPLEMENT
I 1 / INTRODUCTION ASTRONOMY
TOPOSITIONAL 23
Eanh obser\ation and the orbital motion of the Earth around the barYcenter.When tho
parallax is known, thc propcr motion and radial velocity can be combined to forrn
a space-motion vector that represents the motion of thc object with respect to the
standard reference frarne. This space-motion vector should be used when compul-
ing apparent positions for any date and transforming positions from one standard
epoch to another.

Eanh 1.363 Aberration Bccarrse the velocity of light is finite. the direction in u hich a
moving celestialobject is see[ br.a moving obserYcr(its apparcnt positiol) is not
Figure 1.361.1
Parallaxof an object the same as the direction of the straight line betli'eenthe observerand the objcct
(its geometric position) at thc same instant. This displacementof thc apparent
position from the geometric position ma1'be attributed iII part to the motiorr of
the object and in part to the orbital and rotatiortal ntotiorr of the observer.both
of these motions being referred to a standard coordirtatc frtrme. The former part.
object is seen on the horizon (from a point on the Earth's equator). The parallax
which is indcpcn<lcnt of tho rDotion of the observer, is rt:ft:rred lo as llte corre't:lion
of a star is often referred lo arsan annual paru,llo,r,silncethe parallactic shift varies
with a period of one year. By analogy, a horizontal parallax may be referred to as for light-time;thc latter part. which is independentof tht: urotion or distanceof tlttr
object. is referred to as slcllor aberration. (For stars thc totrection for ligirt-tinre
?t d,i,urnalparallax; lhe oblateness of the Earth must be takerr into account in the
is ignored.) The suru of the t$'o parts is called planetarll aberrationbecauseit is
precise computation of the parallactic shift of the N{oon.
applicableto planets atrclother mcnbers of the solar svstettr.The tern:' aberratxon
Distances $'ithin the solar system are conveniently and preciselv expressed in
is used in this supplcntr:trtto include the effectsof the tttotions of both object aud
astronomical units and may be expressedin meters using a conversion factor based
observer.but it is sonretimesused elsclhere for the effect (llle to the rnotion of the
on radar measurements to planets. Prior to 1960, this conversion factor rvas based
observer alone. For prccise astlometr]'. stellar aberration rlrust be calculated froDl
on determinations of the solar parallax. i.e., of the equatorial horizontal parallax of
the formulas of the spocial theoly ol relativitl'. but for llx)st classicill purposes thc
the Sun rvhen at a distance of 1 rru.
theory depending ol or)h' the first power in u / c, r'ltcrc r is the vclocit.Y of th(l
The catalog positions of stars and other objects outside the solar system are
obsen'er and c is tlrt: vr:locitv of light, is srrfllcient.
usually referred to the barycentcr of the solar svstem. The correction for annual
The motion of alr ol)scrvcrorr lhc Earth is the rosrtltallt of the Earth's cliur-
parallax may be applied with srrfficient accuracy for many purposes by assuming
nal rotation ancl tlrr: Flarth's artrtualorbital motion aroun(l tltc barvcentcr of thtr
that the orbit of the Eartl.r is circular. For full precision, however. it is usual to
solar system. Stellal abcrrtrtiott is tlterefore regardecl a.stltc suttt of diurnal aberra-
express the correction in terrrrs of the barycentric rectangular coordinates of the
tion and annua,lal)qration. for q'hich the rnaximum cfft:cts ttrc about 1'i3and 20"'
Earth that are tabulated 1t The Astronomical Almo,nac.These coordinatesare also
respectir-el1..
used to reduce the tirnes of observation of such phenonrena as pulsar emissions to
Annual abcrratiort affects all ob.iectsin the same patt of the skf in a similar
a common origin at the barycenter of the solar svstern.
wal'. and so in photoglnphic obser\?tions of stars it dot's ttot obscure the mu<:h
smaller efiect of annual palallax, t'hich afiects thc aPl).uclll positiorls of onll' the
1.362 Proper Motion and Radiat Velocity The motion of an object outside the so- nearby stars.
lar system is usually expressed in terms of its proper lnotion and radial velocit)'. In determining plecisell' the differencesbctlvoen thc celestial coordilratesof
Proper motion is expressed in terms of the secular rates of chalge in the right as- a solar systcm objcr:t nnd a fixed star, it is uecessarv to takc into account thc
cension and declination with respect to a standard coordirrate frame. It is necessary differential abcrration lrr:tweenthe object and the star. In <rlnstructing an ephemeris
to know the distance (or parallax) of the object before the rnotion transverse to the of a moving object that is to be compared directly with tlrc catalog positions ol
Iine of sight can be expressedin Iinear measure. Ra<lial velocity is the name given to stars at some epoch, it is ncccssary to take into account thc correction for light-time
the motion of the object along the line of sight and is derived from observations of but not stellar aberration; such an ephemerisis known as at astrontetricephemeris.
the Doppler shift in the radiation from the object. It is usually expresscd in km/s. An ephemeris in s'hich aberration is fully taken into account along with precession,
An observational value must be corrected for the rotational motion of the point of nutation, and light-bendiug is known as an app&rent ephemeris.
 24 EXPLANATORY
SUPPLEMENT

1.364 Deflection of Light In precise computations it is necessary to take into ac- Directrix
count the efiect of the gravitational field of the Sun (or other massive object) on
the velocity of propagation of light. The effect causesa bending of the path, so that E = eccentdcanomaly
the apparent angular distance from the Sun is greater, and an increase in the travel y = trueanomaly
time, so that pulsed radiation is retarded. This effect resembles refraction in the
Earth's atmosphere, although the latter efiect is usually much greater. a

.'){g) '
Focus(fr) Focus(fz) , k
,
1.365 Refraction Refraction in the Earth's atmosphere affects most directional
observations. It depends largely on the altitude of the object above the horizon,
and its precise value depends on the atmospheric conditions at the iime. Since the
,-/i
efiect depends on the position of the object with respect to the observer, it cannot
be included in ephemerides. Nominal allowances are made for refraction in the
,,2
computations of the times of rising and setting phenomena, for which the efiects of "2
r a ) E l l i p s e t<0 e < l l : \ + tO= t
refraction are greatest. Refraction increasesthe apparent altitude of a celestial body
and increases the travel times of laser pulses. Most analytical theories and tables
of refraction assume that the surfaces of equal refractive index in the atmosphere
are concentric and that it is sufficient to know the pressurer temperature, and
Directrix
relative humidity at the point of observation. Significant variation from such models
can occur, so that the angular refraction may not be exactly zero in the zenith.
Measurements of azimuth may also be aflected by refraction in the horizontal plane.
The refraction of radio waves depends largely on the properties of the iono-
sphere and is much more variable than reftaction for light waves. Nlultiple-frequency
measurements may be used to determine the refraction correction for radio waves.

1.4 ORBITAL
MOTIONS
(b) Parabola(e = I): y2 = 4q1
1.41 Motionin Two-body
Systems
1.411 Laws of Motion Over a short period of timethe motion ofa planet around
the Sun, or of a satellite around a planet, can usually be treated as if the two
objects are isolated in space and moving around each other in accordance with
Newton's laws of motion and gravitation (i.e., Keplerian or two-body motion). The
smaller, or secondary, body moves around the larger. or primary, body in an orbit
that lies in a plane passing through the primary. The orbit is in the shape of an
ellipse with the primary at a focus. Correspondingly, the Iarger body moves in a
smaller orbit around the center of mass of the pair. The orbits of the major planets
and most satellites are almost circular, whereas the orbit of Comet Hallev, for
example, is extremely elongated. The shape of an ellipse is usually characterized
by its eccentricity, e. A circle is a limiting case as s approaches zero, and e tends
to unity with increasing elongation. Comets may also move in parabolic (e = l)
or hyperbolic (e > l) orbits. The geometric properties of these conic sections are Figure 1.4ll.l
Geometric DroD€rtiesof conic sections
surnmarizedin Fisure 1.411.1.
 26 EXPLANATORYSUPPLEMENT TO POSITIONAL
1 / INTROOUCTION ASTRONOMY 27
Kepler discovered the elliptical character of the orbit of Mars, and found that
the speed of motion of Mars was such that the line joining Mars to the Sun swept
out equal areas in equal times. He also found that the squares of the periods of r, = true anomoly
revolution (P) of the planets are proportional to the cubes of the lengths of their
semi-major axes (a). Newton was the first to demonstrate that these properties
were consistent with the existence of a gravitational force of attraction between the
Sun and the planet. This attraction varies inversely as the square of the distance
(r) between the two bodies and directly as the product of their masses (M and m).
Newton defined the mass of a body as a measure of the amount of matter in the
body, and the momentum of a body as the product of its mass and its velocity.
His second law of motion states that the rate of change of momentum of a body is
proportional to the force acting on it and is in the same direction.
If the Sun and a planet are regarded as forming an isolated system, then the
system's center of mass can be regarded as beirrg at rest or moving at a uniform
velocity in space. This point can be treated as the origin of a set of nonrotating
coordinate axes, which are then used to form an inertial coordinate frame to which
Newton's laws of motion and gravitation apply. One can then write the equations Figure 1.412.1
of motion of the Sun and of the planet with respect to the frame. The equation for Angularorbitalelements
the motion of the planet relative to the Sun (i.e., treating the Sun as the origin of
a coordinate frame whose axes remain parallel to those of the inertial frame) mav
then be derived; it takes the simple forn
where & is knol u as thc Gaussian constant of gravitation. It is no$' the plactice to
treat the value of k as a fixed constant that serves to definc the astronomical unit
GIM + m)r
(l.4ll 1) of Iength (ru) rvhich is slightly less than the semi-major axis of the orbit of the
Earth around thc Sun (seeChapter l5).
where r and i are the position and acceleration of the planet with respect to the
Sun expressed as vectors, with Newton's dot notation to denote differentiation 1.412 Orbital Parameters Positions of a body moving in an olliptical orbit can
with respect to time, and lv'here G is the constant of grar.itation. If the vector r has be computed frorn a set of six orbital pararleters (or elements) that specify thr:
componentsr, ), z with Iespect to the coordinate axes. then Equation 1.,1111 is size, shape. and orientation of the orbit in spaceand the positiorrof the bodl'at a
equivalent to three sepalate equations of the form particular instant in time knot'n as thc cpoch. Commonll-used paranlctersalc:

GIM + mF c= the scrrri-majoraxis of the ellt)sc;

(r.4tr-2) e= of the ellipse;
the <:cr:entricity
13
I= the inclination of the orbital planc to the reference plaue;
Many of the basic properties of motion in an ellipse rnay be dcrived by consider- Ct = thc lolgitude of the node (N) of the orbital plane on tlte reft tenc, platte;
ing the case when the orbit lies in the xy-plane and expressing r and y in polar cu= tho argurnent of the pericenter (P) (angle from N to P);
coordinatesr and d. The principal results arc given in Figure 1.411.la. i'= the epoch at rvhich the body is at the pericenter.
It is convenient to use an astronornical systern of units in q4rich the uDit of
mass is equal to the mass of the Sun, and the unit of tirne is onc dav. When this The longitude of the pericenter, D, *'hich is the sum of f! and u.,,mal be used
is
done, Equation 1..1111 is writtet as: instead of ,.,. And the value of the mean zruomaly,M, or rnean longitude (L = 6 + M)
at an arbitrary epoch may be used instead of T. The reference planc is usually the
. k2(t+ m\r ecliptic, and the epoch for which the reference plane is specified must be stated.
r= --- ' (r.411-3) These angular parameters are shown irr Figure 1.412.1. The mean daily motion of
13
2A EXPLANATORYSUPPLEMENT 1 / INTRODUCTIONTO
POSITIONAL
ASTRONOI'Y 29

a plaret (z) in its orbit may be determined from independentvariable,time, in thc PPN equationfor the solar systcrnis the Barycen-
tric Dynamical Time. When converting the barycentric positions of the plalets io
2n - k,,4 + m geocentricpositions,one must convert thc time accordingto the relativistic metric
"- P a3t2
\t.4r2-l) used in the conlputation. Similarly, the optical observations bcing compared to the
ephemerides nust be corrected for gravitational light bending, and radar obsena-
where P is the period in mean solar days and n is in radians per day. tions must be corrected for the appropriate time delay involved. With the exception
The elements may be given as mean elements rvhere the values are a type of of the NIoon, the natural satellite ephemerides do not require the inclusiorr of rela-
average value over some period of time. The elements may be given as osculating tivistic elfects to the accuracies currently being used.
elements. which are the instantaneous values of the elements or the elements for The perturbation effects may be corrtputed either by mrmerical integration of
the elliptical motion that the body would follow if perturbing effects were removed. the motion of the body or by the use of a general theory that provides an expression
Since the solar system bodies are in well-behaved orbits, this set of elements may for the motion of the body as a function of time. Currently, it is uot possible to
be used. In caseswhere the eccentricity or inclination may be zero, a universal set achieve the accrrracy of observations by means of a general-theorv approach for
of parameters should be used instead. the motion of the planets and the floon. On the other hand. for rapidll' moving
The orbital parameters can be used to compute the velocity as well as the satellites.such as the Galilean satellites,it is not possibleto nurrrericall1-
integrate
position of the body in its orbii. Conversely, if the position and velocity of a body those bodies over an extended period of tirne and retain thc necessary accuracy
are known at a given instant, it is possible to calculate the parameters of the required for observations.
elliptical orbit. When a new minor planet is discovered, each observation gives two
coordinates (usually, right ascension and declination). At least threc observations, 1.422 Numerical Int€gration Nurnerical integration is the most acorrate rnethod
preferably more, are required to determine the parameters of an elliptical orbit that of calculating the rnotion of bodies in the solar system, particularly the \'Ioon and
woulC pass through the observed positions. Gauss' and Laplace's methods can be planets. There arc many rnethods of nurncrical integration. Textbooks on celestial
used for this procedure (Herget, 1948; Danby, 1989). mechanics describc such methods. Usuallr', Cou'ell's method is used u,ith an equa-
tion that includes the PPN relativistic fornlulation. The initial conditions for the
numerical integr.rtiorr are adjusted to fit available observational data aDd a sec-
1.42 Typesof Perturbations
ond numerical intcgration is performed based on that correction. This process is
It is only for applications where reduced accuracy is required that two-body motion repeated until thc numerical integratiou represents the observational data to thc
is adequate for the representation of the motion of a planet or satellite. For other required accuracy.
applications a number of perturbation effccts must be included:
1.423 Analytical Theories As an alternative to numerical integration, an analyt-
(a) The attractions of other planets or satellites must be included in all cases. ical theorl', usuallv in the forrn of a Fourier series or Chebyshev series. can be
(b) The nonsphericity of the distribution of mass in a primary body must be developedto representthe motion of the bodies. Historicalll, this $,asthe method
included in calculating the motions of satellites. of computation for the planets and thc Nloon. and the techniquc is representedb1'
(c) The tidal forces that arise when the gravitational attraction departs from
Newcomb's theories of the Sun, Nlercury, and !'enus; Hill's theoly of Jupiter and
simple inverse square form must be included in the motions of satellites.
Saturn; and Brown's lunar theory. Today this method is used only for planetary
(d) The nongravitational forces, or reactive forces, must be included when com-
satellites, whcre sufficient accuracy can be achieved by this method. These theories
puting the orbit of a comet.
permit determi[ation of the positiorrs of the bodies for any timc within the period
(e) The effect of atmospheric drag and solar wind must be included in the
computation of orbits of artificial satellites. of validity for the theory. Theories of thc planets and the Moon have been developed
(f) Radiation pressuremust be included in the motion ofartificial satcllitesand by Bretagnon, Simon, Chapront, and Chapront Touze, but they do not achieve the
in the motion of particles in the rings around planets. accuracies of nrrmerical integration.

1.421 Ephemerid€s For precise calculations of the motion of the planets and the l.4Z Observational Data The computation of accurate epherneridesof the Sun,
Moon, the equations of motion must include the relativistic terms. This is normally Moon, and planets depends upon the availability of accurate observations of their
done using the parameterized post-Newtonian (PPN) version of the equations. The positions at precisely known times. Optical observations of the Sun, Moon, and
 30 EXPLANATORY
SUPPLEMENT 1 / INTRODUCTION
TOPOSITIONAL
ASTRONOMY 31
planets have been made using meridian, or transit circles; with a person's eye; or coeflicients of such a series may bc determined by analyzing thc departures
ol tlrer
with electronic detectors. Obserrations of \{ercury. Venus, and N{ars can be madc by motion of a satellite from a fixecl ellipse. The effects are most noticeable for
close
radar ranging. Observations of the Nloon have been made using laser ranging since satellites. and the eflects of the higher_order tcrms decrease very rapidly
with <lis-
1969. Planetary radar-ranging and Lunar Laser-Ranging provide much more precise tance. The most noticeable effect is that of the second-order zonal harmoui<..
measurements than transit-circle observations. Radio observations of spacecraft on .J.
which causes the polc of a satellite orbit to precess around the polar
a-xis of the
a planet, in orbit around a planet. or encountering a planet also provide tery precisc planet. though the inclination to the equator remains u.altered.
The clircction of
positional data. the pericenter also precessesaround the orbit. (SeeSection 4.31.)
Observations of the satellites can be made by photographic or charge-coupled The equations of motion of a satellite mtry be refencd to a coordinate
devices(CCD) techniques,and by recordingthe times ofmutual phenomena(eclipses, svslcl[
that is fixed relativc to the planct if the appropriate Coriolis terms arc includcrl
transits, or occultations) of these bodies. Accurate observational data go back to to
take accountof the rotation of the planet. The expressionfor.the potential
only about 1830i so computations of the motions of bodies in our solar systenl are furr<tion
nrzw also bc rnodified in a similar way the arbitrarv constant in the potelrtial
limited by the 160 years of observational history. function is usrrally chosen so that the surface of zero potential corresponds
to nx,an
sea le'el (or the geoicl) in the casc of the Earth. This is the surferc,.
1.425 Differential Corrections The process of correcting a numerical integration frorrr *,hir.lr
hcight is measuredfor nrapping purposes.
or an analytical theory to give a better fit to the observational data is referred to as
lhe differential correction process,Ir\ this process, an equation of condition can be
written for each observation. This equation provides the relationship betu'een cor-
1.432 Forces on Extended Bodies The gravitational forces on an cxtcuded
rections to the initial adopted conditions of the motion and the diflerence betrveen bo<\.
nray he resolr-edinto a force ancl a couple. Thc force ac,tsthrouglr t1,,,..1t,.1 ,,f
the computed and observed position of the body. These equations of condition are
milss and causes an acccleratiolr of t,he body il its orbit. .Ihe couplc changes
combined into normal equations, which are then solved to determine the improved 1h:
rotational rlotiou of thc body. A simple exturrpleis tlrtrt of a two_l)odv slst(,lll.
initial conditions for the motions. This process is described in textbooks on celestial
such as the Sun and a planet. irr rvhich oue bodl. is iur oblate spheroiclthat
mechanics. is
rotating about its principal axis. The torquc. rvhich urtry be consirlcr.edto urct orr
the cquatorial bulge, has a gvroscopic effect. irn(l causest,hc principar axls ro pr(,(,css
1.426 Repres€ntative Ephemerides It is often advantageous to use computers to
arorrnd the normal to the r.rrbitpiarneat a coDstant inclirration. In rnore coruplcx
compute the positions of the bodies from formulas rather than to read ephemerides
systems!a r.rtational r'otirn is superimposeclorr the rnairrprccessio.zrltrtotion.^'.1
that were based on either numerical integratiots or analytical theories. A represcn-
an irregular bocll.rnay "trrrrrble"as it mo\cs alound its or.l)it.
tative ephemcris can bc prepared by evaluating a limited expression that represcnts
The folces on an extendcd boclv *,ill also c:rusea tjdal distortiorr cven if it is :r
the motion of the body to a specified level of accuracy over a limited time period.
solid, but sliglrtll' elastic. bodli Tho effect is similar to that of the o( can tidos orr
These expressions, which are usrrally power series or Chebyshev polynomials, pro-
the Earth. For a simple nrocleltherc are trvo "high tides,' and trlo .,low tides.. that
vide the most emcient rneans of generating thc ephemerides with computers.
remain stationar.ywith lespect to thc attracting body. .Ihe line of t hc high tides is
slightly displacedfrom the line joirring thc cclters of tlxr trvo bocii<rs. This causcs
'I.43
Perturbations
by and on ExtendedBodies a Det torque tllat tends to change the rate of rotation of the bodr.. On the otlror
hand, thc existcnceof thc tidal distortion changcsthe attraction on thc other trody
1.431 Gravity Field ofExtended Bodies Newton's law of gravitation lormally ap-
ancl gives a srrr:rll comporrenl,at right angles to l,he line of t:enters, tlurs causing tlrt:
plies to idealized point masses,but it can be shor.n that the gravitational force due
ot'bit to exJranclor contract, clepenciirrg orr the (.ircumsta|ccs.This effcct is rcferrc<l
to a spherically symmetric body is the same as if all the mass were concentrated at
to as tidal Jrict.ion.
the center. In practice, astr.onomical bodies are rarely sphcrically symmetric, and
The principal elastic properties of a spherical body may be r.eprescntcdin terrrrs
most are rotating at such a rate as to cause the bodies to be clearly oblate. In ad-
of pirrameters known as Loue numbers. Resonanr:e cffccts may or:cur if the tidarl
dition, surfaces of equal dcnsity are not spherical, and the gravitational attraction
forces contain a periodic compol(tlt tlrat matc.hesone of the natural frequencies
on an external point mass is not normally in the radial direction. Other departures
of vibration of the bodv. The dissipation of thc tidal energy ma], bc large enough
from purely radial variation in density occur; so the extcrnal gravity field of an
to heat the body significantly; this is bclieved to bc the case for Io of thc Galilean
extended body, such as a planet, is represented by a spherical harmonic series. The
satcllites of Jupiter.
 EXPLANATORYSUPPLEMENT
1 / INTRODUCTIONTO
POSITIONAL
ASTFONOMY 33
32
tidal distortion periods. When the Sun is continuously very near the horizon, accurare rrmes are
The forces that give rise to the nutation of the rotation and the
although in the past they have often difrcult to tabulate.
of the shape of a body are essentially the same,
been expressedin difierent forms, and with difierent notations. These forces can be
1.512 T\dlight Conditions Twilight is caused when sunlight is scattered by the
derived from a potential function that allows the various periodic terms in the
upper layers of the Earth's atmosphere, Twilight takes place before sunrise or after
nutation and in the tides to be identified and computed The accurate computation
sunset. There are three distinct types of tv/ilight. Civil twilight ends in the evening
of the Earth's nutation requires the adoption of a model of its elastic properties
or begins in the morning when the center of the Sun reaches a zenith distance of
and also of the viscous properties and shape of the fluid core.
96'. Nautical twilight ends or begins vrhen the Sun reaches a zenith distance of
102o. Astronomical twilight ends or begins when the Sun reaches a zenith distance
of 108'. In good weather conditions and in the absence of other illumination, the
1.5 ASTRONOMICALPHENOMENA
indirect illumination of the Sun at the beginning or end of civil twilight allows
the brightest stars to be visible and the sea horizon to be clearly defined. For the
The various motions involved in the solar system, including the rotation of the
beginning or end of nautical twilight, the sea horizon is generally not visible, and
Earth, the orbital motion ol the Earth and planets, and the motion of the Moon
it is too dark for observations of altitudes rvith reference to the horizon. At the
around the Earth, cause a number of important and/or interesting astronomical
beginning or end of astronomical twilight, the indirect illumination from the Sun
phenomena. Some of these phenomena are events that can be timed exactly, some
is less than the contribution from starlight and is of the same order as that from
are matters of arbitrary definitions, and others are not subject to either exact timing
the aurora, airglow, zodiacal light, and the gegenschein. In all cases the actual
or exact definitions.
brightness of the sky depends on direction as well as meteorological conditions.

.1.51 Rising,Setting,
andTwilight
1.52 Meridian
Transit
The rising and setiing times of an object are given with respect to sea level, a
Once each day at a given location the Earth's rotation causes a celestial body to
clear horizon, and refraction based on normal meteorological conditions. Rising
cross the meridian plane defined by the north and south poles and the zenith. When
and setiing times are strictly local phenomena depending upon the longitude and
this happens, the object will have its ma-ximum altitude with respect to the horizon.
latitude of the place on the Earth's surface. For the Sun and Moon, the times of
Generally, observations with a telescope are best made when the object has an hour
rising and setting refer to the instants when the upper limb of the body appears on
angle of less than two hours.
the apparent horizon.
For objects near the pole that can be observed as circumpolar objects from a
The actual times of rising and setting may differ considerably from the tabu-
given location, there can be a second meridian transit when the object is below
lated values, especially near extreme conditions when the altitude is changing very
the pole (lower transit). Observations when the object is above the pole are called
slowly. Thus, the computation of the time of moonrise and moonset for very north-
upper transit. These two observations permit an independent determination of the
ern or southern latitudes can be subject to large uncertainties. Also, the use of
location of the pole.
imprecise methods can lead to extremely large errors. In addition, the illumination
at the beginning or the end of twilight can va,ry greatly because of meteorological
conditions and the physical surroundings. Precise times have little real significance 1.53 Conjunction,
Opposition,
and Elongation
except in special circumstances, such as navigation at sea.
The times of conjunction and opposition of the planets are those at which the
difierence between the apparent geocentric longitudes of the planet and the Sun
1,511 Times ofSunrise and Sunset The calculated times of sunrise and sunset are
are 0o and 180", respectively. Conjunctions for the outer planets take place u'hen
based on the Sun having a geocentric zenith distarce of 90'50' (34' is allowed for
the planets are in the direction of the Sun from the Earth. Oppositions occur when
horizontal refraction using a standard atnrospheric model, and 16' for the semi-
the planets are in the opposite direction from the Sun with respect to the Earth. For
diameter). Corrections are necessary if the meteorological conditions differ signif-
Mercury and Venus, superior conjunctions take place wherr the planets are more
icantly from the standard model or to allow for the height of the observer or the
distant from the Earth than the Sun, and inferior conjunctions occur when the
elevation of the actual horizon. The times of rising and setting and associated phe-
planets are closer to the Earth than the Sun.
nomena change rapidly from day to day in polar regions or may not occur for long
34 EXPLANATORY
SUPPLEMENT 1 / INTRODUCTION
TO POSITIONAL
ASTRONONIY J5
If heliocentric geometric longitudes are used in place of geocentric longitudes, 1.542 Solar Eclipses Eclipses of the Sun or Moon occur when the centers of the
a small correction must be applied to conect for the effect of light-time. This Sun, Earth, and Moon are nearly in a straight line. This condition can be fulfilled
correction can be a maximum of 10 ninutes for the conjunctions of Venus and only when conjunction or opposition occurs in the vicinity of a node of the lunar
Mars. Due to the eccentricities and inclinations of the orbits and the efiect of orbit on the ecliptic. A total solar eclipse takes place when the disk of the Moon
perturbations, the times for these phenornena may be difierent from those at which appears to completely cover the disk of the Sun. Alternately, an annular solar eclipse
the geocentric distance is a minimum at inferior conjunction of Mercury or Venus or takes place when the apparent disk of the Moon is smaller than the apparent disk
opposition of a superior planet, or at which the dista"nceis a maximum at superior of the Sun, so that the light of the Sun is not completely blocked. A pa.rtial eclipse
conjunction of Mercury or Venus or conjunction of a superior planet. of the Sun takes place when the location on the Earth is not in the path of the
The elongation (E) of a planet, measured eastward or westward from the Sun, umbral shadow cast by the Moon. The exterior tangents to the surfaces of the Sun
is calculated as the angle formed by the planet, the Earth, and the Sun. The elonga- and Moon form the umbral shadow cone. The interior tangents form the penumbral
tions a.remeasured from 0o to 180' east or west of the Sun and are tabulated to the cone. The common axis of the cones is the axis of the shadow. The actual size of
nearest degree. Because of the inclination of the planetary orbits, the elongations the shadow will be rather small, Ieading to a narrow path and a short duration
do not necessarily pass through 0o or 180" as they change from east to west or west for a.n eclipse. The path of the shadow on the surface of the Ea.rth can be rather
to east. complicated, due to the orbital motions of the Earth and Moon, and the rotation
of the Earth. Seven minutes of total solar eclipse is a rather long period for totality.
The actual length of time and width of the path varies for sola"r eclipses. Solar
1.54 Eclipses, andTransits
Occultations,
eclipses corne in certain patterns that repeat after a long interval called a,Snros.
An eclipse lakes place when one body passes into the shadow of another body. An A Saros is approximately 223 lunations, which is approximately 19 passagesof the
occultation takes place when a larger body passes in front of a smaller body so Sun through a node or approximately 6b85 1/3 days (see Chapter 8).
that it cannot be seen by the observer. A solar eclipse takes place when the Moon
blocks the light of the Sun as seen by an observer on Earth, and so is really an 1.543 Lunar Eclipses A luna.r eclipse takes place when the Moon passesthrough
occultation. A transit takes place when a smaller body passes in front of a larger the shadow of the Earth. In this case, the times and circumstances are the same
body, e.g., when a satellite passes in front of a planet or when a planet passes in for all parts of the Earth from which the Moon is visible. There are three types
front of the Sun. A shodor,utransit takes place when the shadow of a smaller body of lunar eclipses:A penumbral eclipse, also called an appulse, takes place when the
passesin front of a larger body. This can occur when the shadows of satellites pass i Moon enters only the penumbra of the Earth. A parti,al eclipse occurs when the
in front of planets. Moon enters the umbra without being entirely immersed in it- A totat eclipse ta,kes
place when the Moon is entirely immersed within the umbra.
1.541 Tfansits Tbansits of Mercurv and Venus across the Sun's disk can occur
only when both the Earth and the planet are simultaneously very close to the
1 .55 Satellite Phenomena
same node of the planet's orbit on the ecliptic. Because of the neax constancy of
the longitude of the nodes and perihelia of the planets, the Earth will be near the The Galilean satellites of Jupiter undergo the phenomena of eclipses, occultations,
nodes on about the same dates each year, and the planets will be at the same point transits, and shadow transits. The times for beginning or ending ofthese phenomena
in their orbits when they pass through the ecliptic. For the highly eccentric orbit (i.e., the disappearance and reappearance for eclipses and occultations, and the
of Mercury, this means that the conditions and limits at the November transits are ingress and egress for transits and shadow transits) are given in The Astronom,ical
very different from those at the May transits. During the twentieth century there Almanac. When Jupiter is in opposition, the shadow may be hidden by the disk
are fourteen transits of Mercury-ten in November and four in May, including and no eclipse can be observed. In effect the satellite is occulted by the planet
grazing transits in 1937 May and 1999 November. Future transits will take place before it goes into eclipse. In general, eclipses may be observed on the vrestern
on 1993 November 6 and 1999 November 15. In the twenty-first century there will side of Jupiter before opposition and on the eastern side after opposition. Before
be five transits in May and nine in November. The transits of Venus are quite rare. opposition, only the disappearance of Satellite I into the shadow may be observed,
Generally there are two transits of Venus within an eight-yea.r period and then a since it is occulted before reappearance. After opposition, only the reappea,rances
gap of over 100 years until the next pair of transits. The next transits of Venus will from the shadow are visible. In general, the same is true for Satellite II, although
be 2004 June 8 and 2012 June 6. occasionally both phenomena can be seen. For Satellites III and IV, both phases
36 EXPLANATOFY
SUPPLEMENT TOPOSITIONAL
1 / INTRODUCTION ASTBONOMY 37

of the eclipses are usually visible except near certain oppositions. Similarly, the The elements of the rotational motion are the period of rotation. the position of
occultation disappearances and reappearances of the satellite cannot be observed the axis of rotation in space (represented either by the coordinates of the point on
if at the times concerned the satellite is eclipsed. the celestial sphere toward which the axis is directed; i.e., the pole of the rotation,
For Satellites I and II there are usually cycles of six phenomena consisting of or by the inclination and node of the equator on an adopted reference plane), ald
both phases of both transit and shadow transit, of one phase of the eclipse, and the the planetographic longitude of the central point of the apparent disk at an adopted
other phase of the occultation. For Satellite IV, none of the phenomena occur when epoch. This longitude defines the central meridian on the disk.
the plane of the satellite's orbit, which is essentially the same as that of Jupiter's In general, the apparent positions of points on the disk are represented most
equator, is inclined more than about 2o to the line from Jupiter to the Earth (for conveniently by the apparent distance and position angle relative to the central
occultations and transits) or to the Sun (for eclipses and shadow transits). Because point of the disk. Position angles are ordinarily measured from the north point of
of the finite disks of th€ satellites, the phenomena do not take place instantaneously. the disk toward the east. The central point of the apparent disk is the subterrestrial
The times given refer to the centers of the disks. At certain times similar phenomena point on the surface. Its position on the geocentric celestial sphere is diametrically
can take place for Saturn satellites. Predictions of these phenomena are made in opposite the apparent position of the Earth on the planetocentric celestial sphere.
special publications. The north point of the disk is on the apparent northern limb at its intersection with
the celestial meridian that passesthrough the north celestial pole and the center
of the disk. The vertex of the disk is on the apparent upper Iimb at its intersection
1.56 Physical ofthe Sun,Moon,and Planets
Observations with the vertical circle that passesthrough the zenith and the center of the disk.
Two types of coordinate systems are used for physical observations. Planeto-
In addition to needing information about the positions of the bodies and the phe- centric coordinates are for gerreral use where the z-axis is the mean axis of rotation,
nomena due to the relative positions of the bodies, one also needs information about the r-axis is the intersection of the planetary equator (norrnal to the z-axis through
the physical appearance of the bodies. A physical ephemeris gives information about the center of mass) and an arbitrary prime meridian, and the 1'axis completes a
the orientation, illumination, and appearance of the body; it also provides data that right-hand coordinate system. Longitude of a point is measured positive to the east
can be used in determining the cartographic coordinates of points on the surface. from the ephemeris position of the prime meridian as defined by rotational ele-
For each body, it is necessary to adopt a set of rotational elements and other pa- ments. Latitude of a point is the angle between the plarretary equator and a line to
rameters that specify the shape of the surface and identify the position of the prime the center of mass. The radial distance is measured fiom the center of mass to the
meridian. For the Sun, and for planets and satellites whose solid surfaces cannot surface point.
be seen, it is necessary to adopt conventional values of the rotational elements. It Planetographic coordinates are used for cartographic purposes and are depen-
also may be convenient to adopt different values of the rate of rotation for certain dent on an equipotential surface or an adopted cllipsoid as a reference surface.
bands of latitude. The position of the prime meridian is fixed by adopting a value Longitude of a point is measured in the direction opposite to the rotation (pos-
for the longitude of the central meridian (which passesthrough the pole ol rotation itive to the west for direct rotation) from the cartographic position of the prirne
and the apparent center of the disk) at some suitable arbitrary epoch. For a body meridian defined by a clearly observable surface featurc. Latitude of a point is the
with a solid surface, the prime meridian is usually defined with respect to a clear, angle between the planetary equator (normal to the z-axis and through the center
sharp, surface feature. The appropriate ephemeris data are tabulated for the Sun, of mass) and the normal to the reference surface at the point. The height of a point
Moon, and planets in The Astronomical Almanac. The basic values for the Sun, is specified as the distance above a point with the same longitude and latitude on
Moon, planets, and satellites are given in their respective sections. the reference surface.
The fraction of the area of the apparent disk of the Moon or planet that is illu-
minated by the Sun is called the phase.It depends on thc planctocentric elongation
of the Earth from the Sun, which is called lhe phase onglc. Ncglecting the oblate-
ness of the body, the apparent disk is circular, and tlrc termi,nator is the orthogonal
projection, onto a plane perpendicular to the line ol sight, of the great circle that
bounds the illuminated hemisphere of the body. The terrninator is therefore, in
general, an ellipse, reducing to a straight line at a phasc angle of 90o and becoming
a circle at a phase angle of 0' or 180".
 EXPLANATORYSUPPLEMENT
38
1 . 6 REFERENCES

Brouwer, D. and Clemence, G.M. (Lg6l). Methods oJ Celestial Mechanics (Academic Press,
New York and London). CHAPTER
Danby, J.M.A. (1962, 1988). Funilamentals of Celestial Mechanics (Willmann-Bell, Inc.,
Richmond, VA).
Eopldnatory Supplement to The Astronomical Ephemeris and' The American Ephemeris
and, Nautical Almanoc (1961). (Her Majesty's Statiorery Office, London). Time
Hagahara, Y. (1970). Celestial Mechanics Volumes I and II (MIT Press, Cambddge, MA
and London, England) Volumes III, IV and V (Japan Society for the Promotion of by P.K-Seidelmann,
B. Guinot,andL.E.Doggett
Science, Kojimachi, Cheyoda-ku, Tokyo).
Herget, P. (1948). Conputation of Orbits Published privately by the author' (Willmann-
Bell, Inc., Richmond, VA).
Mueller, I.L (lgig). Spherical and Ptncti'cal Astronomg as Applied to Geodesg (UrLgat,
New York).
Woolard, E.W. and Clemence, G.M. (1966). Spherical Astronomg (Academic Press, New
York and London).

2.1 INTRODUCTION

The methods and accuracy of timekeeping have changed dramatically since 1960.
The definition of the second-the most precisely reproducible unit in the Systdme
International (SI) was once based on the rotation or the motion ofthe Earth. Now
the second is based on natural frequencies of atoms of selected elements. These
cha.ngesare significant for three reasons:

(1) All time systems now refer to the SI second.

(2) Precise time is available on a real-time basis.
(3) The unit of distance, the meter, has been defined most precisely in terms
of the distance traveled by light in a given time interval.

To establish a system of time, one must define two quantities: the unit of
duration (for example, the second or the day), and the epoch, or the zero, of the
chosen time. In physics and astronomy there are four types of systems in common
use. Broadly speaking, they are the following:

(a) Atomic time, in which the unit of duration corresponds to a defined number
of wavelengths of radiation of a specified atomic transition of a chosen
rsotope.
(b) Universal Time, in which the unit of duration represents the solar day,
defined to be as uniform as possible, despite variations in the rotation of
the Earth.
(c) Sidereal time, in which the unit of duration is the period of the Earth's
rotation with respect to a point nearly fixed rvith respect to the stars.
(d) Dynarnical time, in which the unit of duration is based on the orbital motion
of the Ea,rth, Moon, and planets.
eo
40 EXPLANATORY
SUPPLEMENT 2 / rtME 41

Fast-rotating pulsars (millisecond pulsars) may in the future lead to a new time September 23 25), the Consultative Committee for the Definition of the Second
system. In each of these systems, various definitions of the unit of drrration and of (CCDS) recognized this need and proposed the following to the International Com_
the epoch have prevailed in different periods of recent astronomical history. The mittee of Weights and Measures (CIPM):
following section describes the details of these systems.
(a) that TAI is a coordinate timescale defined at a geocentric datum line and
having as its unit one SI second as obtained on the geoid in rotation;
OF TIMEANDTHEIRRELATIONSHIPS
2.2 MEASURES (b) that with the present state of the art, TAI may be extended with sufficient
accuracy to any fixed or mobile point near the geoid by applying the correc-
tions of the first-order of general relativity (i.e., corrections for differences
2.21 AtomicTime(TAl)
in gravitational potential and velocity and for the rotation of the Earth).
Atomic time is based on counting the cycles of a high-frequency electrical sig-
nal that is kept in resonance with an atomic transition. The fundamental unit of
atomic time is the Systime International (SI) second. It is defined as the duration
of 9,192,631,770periods of the radiation corresponding to the transition between 2.22 DynamicalTtme
two hyperfine levels of the ground state of the cesium-133 atom (see Section 2.312).
International Atomic Time (Temps Atomique International, or TAI) is a prac-
Dynamical time represents the independent variable of the equations of notion of
tical time standard that conforms as closely as possible to the definition of the
the bodies in the solar system. According to the theory of relativitv, this inde-
SI second. It fulfills the needs of accuracy, long-term stability, and reliability for
pendent va,riable depends on the coordinate system being used as the system of
astronomy. Although TAI was officially introduced in January 1972, it has been
reference. In modern astronomy, the equations of motion are often referred to the
available since July 1955. The SI second and International Atomic Time are used
barycenter of the solar system. As a result there is a family of barycentric dynarni-
as bases for interpolatio[ and prediction of the other timescales.
cal timescales that depend on various forms of relativistic theory. The independent
TAI results from the analysis by the Bureau International des Poids et Measures
variable of an apparent geocentric ephemeris is a terrestrial dl.namical time, w.hich
in Slvres, France, of individual timescales from commercial atomic-time standards
is also theory dependent. For a given relativistic theory there exists a trarlsforrna-
and the primary frequency standards in many countries. It is realized in two steps.
tion between the barycentric and terrestrial dynamical timescales. The arbitrary
First, an intermediatetime scale,denoted EAL (EchelleAtomique Libre), is formed
constants in the transformation can be chosen so that the tiruescales have onlv
by combining data from all available high-precision atomic standards. The data periodic variations with respect to each other.
from the individual clocks are combined using an algorithm that minimizes the
Since the transformation between barycentric and geocentric tirnescales de-
effect of adding and dropping clocks and that weights clocks appropriately. As the
pends upon the theory, the two types of scales cannot both be unique. Thus,
name implies, EAL is a free-running, data-controlled timescale. After analysis of
the dynamical timescale for apparent geocentric ephemerides was chosen to be
EAL, corrections are applied for known ellects to maintain a unit of time as close as
a unique timescale, independent of the theories, whereas the dynarnical timescales
possible to the definition of the SI second. This adjusted timescale is published as
for barycentric ephemerides form a family of timescales that are not unique and are
TAI. Since TAI is accessiblethrough the published corrections to each contributing theory^dependent.
timescale, it is often called a paper-scale clock. The recommendations for the dcfinition of dynamical time adoptcd at the IAU
A large number of commercial cesium standards and a few hydrogen masers General Assembly of 1976 in Grenoble, France are as follows:
provide availability, reliability, and short-term stability to the TAI system. Accuracy
and long-term stability are provided by laboratory cesium clocks or standards. (a) At the instant 1977 Jauuary 0l d00h00-00"TAI, the value of the neu' tinrescale
It is anticipated that improvements in the clocks or standards will permit future for apparent geoccntric ephemerides will be 1977 January 110003725
improvements in the International Atomic Time, but the fundamental principles (1"00'00'321184)exactly.
underlying the formulation of the TAI will remain the sarne. (b) The unit of this tinescale will be a day of 86,400SI secondsat mean sea
level.
2.211 Relativistic Effects Advances in the accuracy of time compaxisons require (c) The timescales for equations of rnotion referred to the barycenter of the solar
the adoption of a set of conventions and a coordinate reference frame to account system will be such that there will be only periodic variations between these
for relativistic effects in aI internally consistent manner. At its 9th session (1980 timescales and those of the apparent geocentric ephemerides.
42 EXPLANATORY
SUPPLEMENT 2 | rtME 43
In 1991, in the framework of resolutions encompassing both space and time and JD is the Julian date to two decimals of a day.
references, the IAU has defined new timescales TCB, TCG' and TT, as explained A more accurate expression is given by Moyer (1981):
in Section 2.223.In lhe 1976 and 1991 resolutions, the dynamical time of planetary
1 - i - -* , s i n D
motions is identical to the timescale of terrestrial atomic physics. The consequences T D B- T D T= L T e + ) 1 / r u , a r e s i n E ,+" 1
of this assumption are considered in Section 2.34.
- qi#q, - | +^ - ttr)
+^)+esin(ur
2.221 Tenestrill Dynamical Time Terrestrial Dynamical Time (TDT) is the the- [,r ]",,,'",ur1
oretical timescale of apparent geocentric ephemerides of bodies in the solar systern I ^
+ e' sin(UTl r t -zm]
TDT is specifled with respect to TAI in order to take advantage of the direct avail- , l
ability of Coordinated Universal Time (UTC), which is based on the SI second.
TDT is currentlv set equal to TAI * 32.184 seconds. - sc(1 jos€)a1.in(uTl+I+2a)+esin(uTl + ^+2L+ M)l
Since TAI is a statistical timescale based on a large number of clocks operat-
s{ l,Ltft:''
ing on the Earth, and TDT is an idealized uniform timescale, the two timescales * \in{urt + A - D)
zc'\t + u)
cannot be made identical. TAI is subject to systematic errors in the length of the
S61sin",1v , prcJ - lsAesA -
TAI second and in the method of forming TAI. For the near future, however, the cosL+.2sJ slnEl+ slnEs+
d .:gro
cumulative effect of such errors is likely to be significant only for evaluation of the
millisecond pulsars. Ptsra ., , /seSsqa ,
+-Sln{L-Lr)+ ^ Sln(L Lsa,
The relationship between TDT and TAI provides continuity with ephemeris c, p: c, p:
time (ET). The chosen offset between TDT and TAI is equal to the estimate of the
difierence between ET and TAI at the time that TDT was introduced. For years
*l!!ffi@,rin(urr+)+z zr)
prior to 1955, when atomic time was not available, TDT must be extrapolated p5aS5a(l+ cose)z ---.
backward, using the theory of a rapidly moving celestial body that has been fit to
+ - sin(l Il+^+Z lsn). \ 2 . 2 2 22 )
2c'p:
observations made since 1955. The Moon and Mercury are the most suitable bodies
for such purposes. The accuracy of the backward extrapolation will depend on the When numerical valuesare substituted into the expression,the relationship
length of the comparison after 1955 and will deteriorate as a function of the length becomes:
of time prior to 1955. Use of the Moon for this extrapolation is complicated by the
Moon's tidal acceleration. At this time there are uncertainties concerning the value TDB T D T = A I e + 1 . 6 5 8 x 1 0 3 s i n ' + 1 . 5 4 8x 1 0 - 6 s i n D
of the tidal acceleration and the systematic variations between the theoretically + 3 . 1 7 6 7 9 xl 0 - r o a s i n ( U T+ll ) + 5 . 3 1 2 x 1 0 r 2 a s i n ( U T+l)M)
calculated position of the Moon and the observations. - 2 M ) - 1 . 3 6 7 7 x10 "asin(UT+
+ 1 . 0 0x l 0 - r 3 r s i n ( U T+l ) l ^+2L)
- 2 . 2 9x l O r 3 as i n ( U T+l ^ + 2 L + l u t ) + 1 . 3 3 x1 0 r 3 a s i n ( U T 1 + . \ - D )
2.222 Buycentic Dynamical fime Barycentric Dynamical Time (TDB) is the in-
- 1 . 3 1 8 4 x1 0 l O v c o s Z + 5 . 2 x 11 0 6 s i n E ;
dependent variable of the equations ol motion with resper:t to the barycenter of the
solar system. In practice, TDB is determined from TDT by means of a mathe- +2.45 x 10 6sinEse+ 20.73 x 10 6sin(Z 4)
matical expression. This expression depends upon the constants, the positions and + 4 . 5 8 x 1 0 6 s i n ( Z - t 5 a ) +1 . 3 3x 1 0 l 3 z s i n ( U T+l) + L - L 1 )
motions of the bodies of the solar system, and the gravitational theory being used.
+ 2 . 9 x l 0 ' * a s i n ( U T 1+ . \ + l - t 5 a ) . t2.222-3)
An approximate formula is suf6cient in most cases for converting from TDT to
TDB. Such a formula is
The symbols in Equations 2.222 2 and 2.222 3 are defined as follows:

TDB = TDT + 0i001658 sin29,

sing+ 0i000014 (2.222-r) AZ4 = s6ns16.1 term in expression
c = speed of light
where rl, fr = position and velocity vectors of point i relative to pointj. The dot denotes
g = 357i53+ 0. 9856003(JD 2451545.O) differentiation with respect to coordinate time t
44 EXPLANATORY
SUPPLEMENT 2 / f IME 4J

pj = gravitational constant of body i h=Mt (2.222_s)

pJ = gravitational constant of the Sun augmented by the grawitaiional con- Es.c = Msl
stants of the planets and the Moon
ai, e, Ei semi-major axis, eccentricity, and eccentric anomalg respectively, of
=
2.223 Space-Time Coordinates In 1991 the IAU adopted resolutions clarifying ihe
heliocentric orbit of planet i; without subscript, the orbit of Earth Moon
relationships between space-time coordinates. These resolutions introduced Terres-
ba,rYcenter
trial Time (TT), Geocentric Coordinate Time (TCG), and Barycentric Coordinate
,3i= circular orbit velocity of pla"net i relative to Srn, y(1r3 pay a1
Time (TCB), and they permitted the continued use of Barycentric Dynamical Time
.\ =east longitude of atomic clock
(rDB).
z= distance from Earth's spin axis of atomic clock in km
y= The space-time coordinates (Xo = ct, x|,12, 13) would be defined in the frame-
coordinate of atomic clock in distance from Earth's equatorial plane
work of the general theory of relativity in such a way that in each coordinate system
(positive north) in km
centered at the barycenter of an ensemble of masses exerting the main action, the
aln semi-major axis of geocentric orbit of the Moon
=
j,n = circular orbit velocity of the Moon relative to Earth, t/(ps + pJ7 q^ interval dsz would be expressed at the minimum degree of approximation by
e = mean obliquity of the ecliptic, i.e., inclination of the ecliptic plane to / 1t1\ / ) \
mean Earth equator of date dr2=-c2d,2 = {t - 1}(aro)'* lr*:l ) l \ d x t ) 2 + t d x ? l + t d s 3 t 2 1t 2 . 2 2 3 -1t
\ c'/ \ c'/
du Greenwich mean sidereal time : Greenwich hour angle of mean equinox
=
of date where c is the velocity of light, r is the proper time, and U is the sum of the
Sc = circular orbit velocity of Earth Moon barycenter relative to the Sun, gravitational potentials of the ensemble of masses and a tidal potential generated
'vTp;+n;+VJn by the external bodies, the latter potential r,anishing at the barycenter.
Z = mean longitude of the Sun with respect to the Earth Moon barycentric The space-time coordinates are further constrained such that:
referred to the mean equinox and ecliptic date
D = the mean elongation of the Moon from the Sun (1) The space coordinate grids with origins at the solar system barycenter and
4 = heliocentric mean longitude of planet i at the Earth's center of mass show no global rotation with respect to a set
of distant extragalactic objects.
Subscripts and superscripts are as follows: (2) The time coordinate is derived from a timescale realized by atomic clocks
operating on the Earth.
A = location of atomic clock on Earth that reads International Atomic Time (r) (3) The basic physical units of space-time in all coordinate systems are the SI
m = Moon second for proper time and the SI meter for proper length (this is related
S=Sun to the SI second by the value ofthe velocity of Iight: c = 299192458rr.s-' ).
C = solar system barycenter
E = Earth It should be noted that the preceding kinematic constraint for the state of ro-
J = Jupiter tation of the geocentric reference system implies that when the system is used for
SA = Saturn dynamics (e.g., the motion of the Moon), one must take into account the time-
dependent geodesic precession in the motion of the Earth, and introduce the corre-
The absence of a subscript indicates the Earth.
sponding inertial terms in the equations. Also, if the previously defined baxycentric
The relation between the eccentric and mean anomalies is:
reference system is used for dvnamical studies, the kinematic efiects of the galactic
geodesic precession may have to be taken into account.
E = M + e sinM + (e2/ 2) sin2M +. t ) ))t -Ar
To be consistent with the proposed space-time coordinates, coordinate times
would be specified such that:
The eccentric anomalies of the Earth, Jupiter and Saturn can be computed to
sufrcient accuracv from ( I ) The units of measurement of the coordinate times of all coordinated systems
centered at the barvcenters of ensemblesof masses are chosen so that they
EB=Me+esinMa are consistent with the proper unit of time, the SI second.
46 EXPLANATORY
SUPPLEMENT 2 / rtME 47
(2) The readingofthese coordinatetimes is 1977January 1, 0h0'32i184exactly, dependiug upon their terrestrial coordinates are diurnal, u.ith a maximuqr
on January 1,0r'0'0" TAI exactly (JD :2143114.5. TAI), at the geocenter. amplitude of 2.1 ps.
(3) Coordinate times in coordinate systems that havc their spatial origins at (3) The origins of coordinatetimes have been arbitrarily set so that these tirnes
the center of mass of the Earth and at the solar systen barycenter and all coincidewith the TerrestrialTime (TT) at the geocenterof 1977January
thai meet the previously stated requirements are designated as Geocen- 1, 0"0.0" TAI.
tric Coordinate Time (TCG) and Barycentric Coordinate Time (TCB), (.1) When realizations of TCB and TCG are needed, it is suggested that these
resPectivelY. realizations be designatcd by expressions such as TCB(,rxr). r'here r,tx in-
dicates the source of the realized timescale (e.g., TAI) and the theory used
TCG and TCB rvere adopted b1' the IAU in 1991. Thev are in a wav com- for translbrmation into TCB or TCG.
plementary to TDT turd TDB, but it must be recognized that some astronomical
constants and quantities have different numerical values when using TDB. as op- In additiolr, it was agreed to delete the rvord "dynamical" from Terrestrial
posed to TDT, TCG, or TCB. It should further be noted conccrnirrgTCB and Dynamical Tirne (TDT) and just call it TerrestrialTime (TT). TT rvouldbc defined
TCG that: as follows:

(1) In the domairr comDlon to any t*o coordinate systenls, thc tcnsor transfor- (1) The time reference for apparent geocentric ephemerides is Terrestrial Time,
mation law applied to the metric tensor is valid without rescaling the unit TT.
of time. Therefore thc various coordinate tirncs cxhibit secular variations. (2) Terestrial Tirne is a tirnescale difiering from TCG uniquely by a constant
Recommendation5 (1976) of IAU Cornmissious4, 8, and 31. completed rate, its unit of measurement being chosen so that it agrees with the SI
by Recommerdation 5 (1979) of IAU Con.rmissions 4. 19, and 31, stated secondon the gcoid.
that Teuestrial Dynamical Time (TDT) ancl Btrrycentric Dynamical Time (3) At instant 1977 January 1. 0h0'0" TAI exactl"v,TT has the reading 1977
(TDB) should clifferonll'by periodic rariations. Thcrcfore TDB and TCB Januar;' 1, 0'0'32i 184 exactly.
difler in rate. The rclationship bet$een these tiruescalcsin secondsis given
br.: It should be noted <oncerningTT that:

TCB - TDB = Lax(JD 2443144.5)

x 86400 (2.223-2)
(1) The dir,'ergencebct$.een TAI and TT is a consequenceof the phJ.si(al de-
fects of atoruic time standards. In the irrtcrvnl 1977 1990, in addition to
The presentestimate of the value of 16 is 1.550505x l0 8 (t I x 10 ra).
the constant offset of 321184,the deviation probablv remained within the
(2) The relation TCB - TCG involves a firll forrr-dimcusional trzrnsformation:
approxirnatc limits of i l0/.s. It is expectedto increasemore slowly in the
future as erconseqlrenceof improvements in timc standards. In many cases,
.l I I especiallyfor the publication of ephemerides,this deviation is negligible.
TCB ICG=c-'I -
I t t ! r Z + U " , r { x " r ) d r + v " . r x x " 1 l. t 2 . 2 2 33 l
1J,,, I In such cases, it can bc stated that thc argunrent of the ephemerides is
TAI + 32:184.
x. and v. denote the barycentric position and vclocitl'of the Earth's center (2) Terrestrial .firne difiels from TCG in seconclsuniqrrely b1'a scaling factor:
of mass, and x is thc barycentric position of the obsert'er. The external
potential, U."s,is the Nervtonianpotential of all solar sl.stcm bodies apart
from the Eart.h. Thc extcrnal potential urust bc craluated at the geocenter. TCG - TT = Zc x (JD - 2443144.
5) x 86400 (2.223,s)
In the integral, / = TCB and 19is chosen to agree with the epoch of Ter-
restrial Time (TT). As an approximation to TCB TC(l in secondsone Tlre prcscnt cstinrate of the r.alueof Lc is 6.969291x 10 r0 (1 3 x l0 16).
might use: The two tirncs<:alcsare distinguished by diflcrent namcs tc avoid scaling
errors.
TCB-TCG = Lc x (JD -2443144.5) x 86400+c 2v" (x-x")+P. (2.22i4) (3) The unit of measurement of TT is thc SI scconcl on the geoid. The usual
multiples, such as the TT day of 86400 SI secondsorr the geoid, and the TT
The prcsent estinrate of the value of Lq is 1.480813x l0-8 (i I x l0-r4). Julian ccntury of 36525 TT days can be used, provided that the reference
The quantity P replescnts the periodic terrns which can be evaluated using to TT be clearly indicated whenever ambiguity may arise. Corresponding
analytical formulas. For observers on the surface of the Earth, the terms time intervals of TAI are in aqreement with the TT intervals within the
48 EXPLANATORV
SUPPLEMENT 2 | TIME 49
14in
utcertainties of the primary atomic standards (e.g., u'ithin t 2 x l0 constant. However, the conceivable change in the period of rotation is such that the
relative value during 1990). effect of a variation in the daily precessional urotion is inappreciable. The secular
(4) It is suggestedthat realizations of TT be designated by TT(x.rr), where xrr vaxiations are almost inappreciable (see Section 3.21). The apparent sidereal day,
is an identifier. In most cases,a convenient approximation is: nominally of 24 hours of apparent sidereal time, differs from the period of rotation
by a variable amount depending on the nntation.
TT(TAI)=TAI+321184 (2.2234) To each local meridian on the Earth there corresponds a local sidereal time,
connected with the sidereal time ofthe Greenwich meridian by means ofthe rela,tion:
However, in some applications it may be advantageous to use other realiza-
tions. The BIPM, for example, has issued timescales such as TT (BIPM90) ( 2 . 2 3 -)t
local sidereal time = Greenwich sidereal time + east lonsitude.

The concepts were adopted by IAU resolution and will probably be implemented
Sidereal time is conventionally measured in hours, minutes, and seconds. so that
into general use in the future. There is an Astronomical Standards Working Group
longitude is measured(positively to the east) in time at the rate ofone hour to l5'.
currently considering various details of interpretation and implementation.
An object transits over the local meridian when the local sidereal time is equal to
its right ascension.
T]me
2.23 Sidereal Classically, sidereal time was determined from observations of the transits of
stars, either over the local meridian or, with a prismatic astrolabe, over the small
In general terms, sidereal time is the hour angle of the vernal equin<.rx,which is also circle corresponding to a constant altitude. Becausethe reduction of these observa-
known as the first point of Aries. Apart from the inherent motion of the equinox, due tions depended on star positions taken from a standard star catalog, sidereal tinre
to precession and nutation, sidereal time is a direct measure of the diurnal rotation was actually based on the hour angle of the zero point of right ascension of the
of the Earth. Equal intervals of angular motion correspond to equal intervals of 'lhis poiDt
catalog, known as lhe catalog equinor. is distinct from (but close to) thc
mean sidereal time. so that sidereal tirne reflects the actual rotation of the Earth. dynamical equinox defined by the orbital motion of the Earth (seeSection3.1).
and can be determined by observations of celestial objects. Apparent sidereal time, because of its variable rate. is used only as a rneasure
The sidereal time measured by the hour angle of thc true equinox i.e., the of epoch; it is not used as a measure of time interval. Observations of the diurnal
intersection ofthe true equator ofdate with the ecliptic ofdate is apparent sid,ereaL motions of the stars provide a direct measure of apparent sidereal time, a^stheir
time. The position of the true equinox is afiected by the nutation of the axis of right ascensionsare measured from the true equinox. But in many practical methods
the Earth, which consequently introduces periodic inequalities into the apparent of determining timc, the right ascensions are diminished by the equation of tht:
sidereal time. The time measured by the diurnal motion of the mean equinox of equinoxes, so that mean sidereal time is deduced directly from the observations.
date, which is a,fiected by only the secular inequalities due to the precession of Today, sidereal time is determined most accurately by interferometer obscr-
the axis, is mean sid,eteal time. Apparent sidereal time lninus mean sidereal time vations of radio sources. It can also be determincd from lunar and satellitc laser-
is the equation of the equinoxes. In ephemerides immediately preceding 1960, the ranging. Since the rotation of the Earth is subjcct to irregular forces, sidereal tinte
equation of the equinoxes was called the nutation in right ascension. is irregular with respcct to atomic time. The practical determination of sidereal
The period between two consecutive upper meridian transits of the equinox is a time requires that short-period irregularities and polar urotion be evaluated at the
sid,ereald,ay.It is reckoned from 0h at upper transit, which is knolr'l as sidereal noon. same time.
Owing to precession, the mean sidereal day of 24 hours of rnean sidereal time is
shorter than the actual period of rotation of the Earth by about 0i0084, the amount 2.231 Greenwich Sider€al Date In order to facilitate the enumetation of succes-
of precession in right ascension in one day. The ratio of the length of the mean sive sidereal days, the concepts of Greenwich sidereal date and Greenwich sidereal
sidereal day to the period of rotation of the Earth is 0. 99999990290i - 59 x 10-12T, day number, analogous to those of Julian date an<l Julian day number (see Sec-
where Z is measured in Julian centuries from J2000.0. Correspondingly, the period tion 2.26), have been introduced. The Greenwich sidereal date is defrned as the
of rotation is l. 00000009?093+ 59 T x IO-t2 mean sidereal days. Thcse numbers are interval in sidereal days, determined by the equinox of date, that has elapsed on
not rigorously constant b€cause the sidereal motion of the equinox due to precession the Greenwich meridian since the beginning of the sidereal day that was in progress
is proportional to the length of the UT day, that is. to the period of the rotation at JD 0.0. Thc integral part of the Greenwich sidereal date is the Greenwich sidereal
of the Earth, whereas the angular measure of the complete rotation is, of course, day number, a count of transits of the equinox over the Greenwich meridian The
 II
I
50 EXPLANATORY
SUPPLEI,ENT 2 / rtME
+ 51
decimal part of the Greenwich sidereal date is sirnply the Greenwich sidereal tirrrc (a) UTl is proportional to the angle of rotation of the Earth irr spacc, reck-
expressedas a fraction of a sidereal day. These concepts can be applied equally well oned around the true position of the rotation axis. In other words. UT1 is
to mean or aPParent sidereal time. proportional to the integral of the modulus of the rotation vo<.torof the
Earth.
(b) The rate of UT1 is choscn so that the day of UTl is close 1,othe nrean
duration of thc solar day.
Time
2.24 Universal (c) The phase of UT1 is chosenso that l2h UTl correspondsapproxirnatcly.in
the average.to thc iDstant rvhen the Sun crossesthe meridian of Grcr.nulich.
Universal Time (UT) is the measure of tirne used as the basis for all civil time-
keeping. It conforms closely to the mean diurnal tnotion of the Sun. The apparent
Conditions (b) and (c) arc uot strictl)' compatible with condition (a). Bv adopt_
diurnal motion of the Sun invoh'es both the nonunilorrn diurnal rotation of the
ing (a), one accepts that UTl deviatcs secularly from solar time; hou,cver, the
Earth and the motion of the Earth in its orbit around the Sun. Although it u'ould
dir''ergenceis extremely srnall.
be possible to define a system of timc mcasurcnlent in terrns of the hour angle of
The constant teru and the rrunrerical coefficient of T, in Eclra.tion 2.24 1 u,.ere
thc Sun, such a system could never be related prcciselv to sidereal time and could
chosen so that UT1 would bc rrrrrtinuo's in value and rate through the t:ltiurgefro'r
not, therefore, be determined by observations of sttrr trartsits. As a result, Univcrsal
the old to the new system of constants (1984 Jamrary 1). The surall cliflcrc,rrccsdue
Time is directly rclated to sidereal time by rneturs of a rrurncrical formula. It docs
to the change in the theory of nutatiorr rvere negiected.
not reler to the motion of the Earth and is not preciselv related to the hour angle
The r.ariation of UT1 q'ith rcspect to uniform time truh, rcpr.eserrtsthc irrog_
of the Sun.
ularities of the rotation of thc Earth. The duration (A) of thc clal. of l_rTl in SI
Universal Time at any instant can be derivcd from observations of the diurnal
seconds is giren b1.
motion of the stars or radio sources. The uncorr-cctedobserved rotational timescale.
rvhich is dependenton the place of obser\.atiou,is designatedUTO. Correcting this I = 86400 (t,z - ut) I n, t ) )/l-) \

tirnescalefor the shift in longitude of the observingsl ati()n causedb1' polar motion
produces the UTI timescale. l'hich is independent of obsen'iug location. but is rvhere i-1 and L,2 are tlle lalrrcs of l,-Tl - TAi in seconds at rr-dar-intcrvals. I'hc
influencedby the slightll' rariable rotation of the Earth. angular velocitl-of tlrr: Earth. ;. is given bv
Sincc 198.1January 1 Greenrvichrnean siderealtirrrc (G\IST) has bccn relartcrl
to UT1 as follorvs: x 10 0rads I
. , = ( 8 6 4 0 0 /A ) x 7 2 . 9 2 1 1 5 1 4 6 1 t) )4-1\

G M S T I o f 0 ' U T l = 2 4 1 1 0 : . 5 4 8 4 18 + 6 ,6+ 0 : 0 9 3 1 0 4
6 4 0 1 8 4 : 8 1 2 8? 4 6.2x l0 6Ii For the dynamicsof thc solar systcrn,thc studv of the rotation ol thc Earth, arrtl tlrrl
(2.24-t) strrdy of the Galaxy, orrc neeclsthc relationship betneen the position arrri trrotion
of the equinox and thosc of a poirrt orr thc cquator. the horrl tr,rrgk: of whiclr trrrlv
where 7u = d, 136525, du is the nurnber of clays of Urivorsal Timc elapsed since
representsthe rotatiorr of thc Earth in space (siderr:alrotation). Thcrc ar.cscvcral
.ID 2451515.0U'Il (2000 January 1, l2hUT1), takirrg on values oft 0.5, I 1.5.
ways to estabiishthis rclationslrip,incluclingusing statistics on l llc prol)('1.nlot ions
X 2.5. X 3.5. .. .. This relation conforms to tlx' I)osition and motion of the equinox
of stars. thc d1'narnicalrncth<xl.arr<lobsr:rrationsof quasars;thest: ktr<l to siiglrtll.
specificdirt the IAU 1976Svstem of Astronomi<al (lolstants. the 1980IAU Theory
different realizationsof the c<luinoxand. therefore.of the siclerealtinrt:. TIrcr.t'forc.
of Ntttation. artrl the positions and proper motions of stars in the FK5 catalog.
an agleement has berrr rrrack't()takc as a rcferencethe cquinox an(l its rn()tidr as
E<luatior 2.2-11 is oftcn consideredto bo tlle de,finitiorrof [,rT1. becausen.hen
realizedb1'the IAU 1!176Slst.ur of Astr(nromicalConstants.the 1980IAII'l'hcon'
it rvas cstablished.the obserr-ationof trausits of thc FI{5 stars (ol of other stars
of Nutation. arrd the positiorrsaud proper nrotions of stars of the FI{5 catalog.
u'ith t]rcir positions and proper rnotiolls itr tltc l'l{j-r slstem) *'as thc best means
to obtain llT1. \\Ihur new techniqucsfor rrrcasurirrgUTl becane opcrational (see
2.241 Conversions bet\,yeenUniversal Timeand Sidereal Time 'Ihe
Section 2.32). the arbitrarl' constantsinvolvc<iin thoir irrrpleruentationwere set to nunrbcr ol sec-
align thc lJTl r'alucs u.ith thosr: obtained by obsr.n'arlionsof FK5 stars and to bc onds of siderealtirne in arJuliarr ccrrtury can be obtaincd bv dir.cct diflcrcrl.iation
consistcDtwilh Ecluation 2.2,1 1. of Equation 2.2,1 1:
The rnodcrrr clefinition of UTI was established il order to fulfill the follo*'ing
conditions: s / = 8 6 4 0 1 8 4 . 8 1 2 8 6 61+806.2 0I 8 1 . 8 6 x 1 0 5 7 2 . (2.241-t)
 EXPLANATORY
SUPPLEMENT 2 | ltME c.t
52
where EQ is the equation of the equinoxes.
Dividing by 36525 (the number of days in a Julian century) gives the number of
(b) Express the date in Greenwich mean sidereal time by
seconds of mean sidereal time in a day of Universal Time (UTl to be precise):

+ 5.098097x 10-67 - 5. 09 x l0-1022

s = 866361,55536790872 (2.24t-2) GMSTI=LMST-), () )t)_)\

where ) is the astronomical longitude of the instrument, positir,-etoward

where Z is the number of Julian centuries elapsed since JD 2451545.0. The quantity
the east. Since the pole of rotation is moving with respect to the surface
J represents the number of sidereal seconds in a day of Universal Time. Dividing
of the Earth, the instrumental meridian is subject to small variations. The
Equation 2.241-2 by 86400 gives the ratio of mean sidereal time to UTl:
position of the rotational pole (r,;y) with respect to the Conr.entional Inter-
national Origin (CIO) can be used to correct for this effect. This correctron
+ 5.9006x 10 | I 7 - 5.9 x l0-r51.
y' = |.00273'1909350795 (2.24r-3) of )s (referred to CIO) to ) is accomplished by applying the expression

Taking the inverse of this expression gives the ratio of UTl to mean sidereal time:
,\ = ,\o - ("rsin )o + ycos )6) tan @ t) )L)-\l

- 5.8684 x 10 rr7+5.9 xl}-tsTa.

I I rt = 0.997269566329084 Q.2414) where / is the latitrrde of the instrument.
(c) Compute the tirne d elapsed between 0h UTl and the everrt, expressed in
Although the lengths of a day of UTl and a day of mean sidereal time vary mean sidereal time by
slightly with variations in the Earth's rotation, the ratio of UT1 to mean sidereal
time is given by Equation 2.24I-4 ar,rJis unaffected by rotational variations. Thus,
d = GMSTI G M S T 1o f 0 " U T I . () )4t--4\
the established method of determining Universal Time by multiplying an interval
of sidereal time elapsed from 0" UT by a fixed conversion factor maintains the (d) Convert d, expressed as an interwal of sidereal time, to an interval of UT1:
constancy of the ratio of Universal Time to sidereal time, irrespective of variations
in the rate of rotation of the Earth.
UTI = r0 () )l)_\\
Roughly speaking, UTl can be regarded as solar time or dynamical time in
these equations. Then Equation 2.241-4 can be thought of as giving the length
where r is the ratio of mean solar time to sidereal time intervals, given in
of one sidereal day in units of solar days. Disregarding the inappreciable secular Equation 2.241 4.
variations, the equivalent measures of the lengths of the days are:
For instance when the hour angle, ll, of a star with riglrt asccnsiorro has been
mean sidereal day = 23r56'94t0n0524 of Universal Time,
measured, the corresponding UT1 value is
day of Universal Time = 24h03'56?5553678of mean sidereal time.
UTI = (H+d - EQ - l - GMSTI of 0"UTl)r. (2.2424)
The rotational period of the Earth is 86164.09890369732seconds of UTI or
23'56'04:09890369732. This gives a ratio of a UTl-day to the period of rotation The relationshipbetq'ecnLMST and local civil time (LCT) is given by
of 1.002737811906. The rate of rotation is 15.04106717866910secondsof arc per
second of time.
LMST of 0h LCT = G\'ISTI of 0n UT1 - (t - r\\ t r. (2.242,7)
2.242 Conversion of Local Apparrnt Sidereal Time to UTI The usual procedure
where LCT = UTt + ), and GMST1 of 0h UTI is obtained from Equation 2.24 1.
to express the UTl date of an event that occurred at some instant dated in local
apparent sidereal time (LAST) is as follows:
2.?43 Coordinated Universal Time Since 1972 January 1, most, and now all, broad-
(a) Express the date in local mean sidereal time (LMST) by ca€t time services distribute timescales based on the redefined Coordinated Univer-
sal Time (UTC), which difiers from TAI by an integer number of seconds. UTC is
LMST=LAST-EQ, (2.242-r) maintained within 0.90 second of UT1 by the introduction of one-secondsteps (leap
54 EXPLANATORY
SUPPLET\,,iENT c5

seconds) when necessary, usually at the end of June or December. DUT1, an ex-
trapolation of the difference UTI mimrs UTC, is transmitted in code on broadcast
Local meridian TYueSun
time signals.

2.25 The Ephemeris Meridian

Since hour angles are calcrrlated with reference to the meridian of Greenwich or
other geographic meridians, they are dependent on the rotation of the Eafth. For
solar-system bodies, whose motions are expressed as a function of dynamical time,
such calculations can only be carried out accurately after a value of AI ( = TDT
UT) is known. To facilitate practical calculations of phenomena that depend uport
hour angle and geographic location. the concept of an auxiliary reference meridian,
known as lhe ephemeris meridi,an, was introduced.
Although the ephemeris meridian was originally used in conjunction with ephern- Universalmeridian Mean SunofUT
(Greenwich)
eris time, its name has been retained for use rvith calculations in TDT. Its position
in space is conceived as being where the Greenwich meridian would be if the Earth Figure2.25.1
Meridianrelationsandtime
rotated uniformly at the rate implicit in the definition of dynal.rical time; i.e.,
l. 002738A2 east of the actual mcridian of Grecnwich on thc surface of the Earth.
Calculations in terms of TDT may be relerred to the ephemeris neridian us-
ing methods that are formally the same as those done in UT, when referred to
the Greenwich meridian. As soon as a sufficiently accurate value of AZ can be and conventional to convert TDT to UT by means of AI and to relate hour angles,
determined or extrapolated, the longitudes and hour angles can be referred to the longitudes, and times to the Greenwich meridian.
Greenwich meridian, and the times in TDT can be expressedin UT. This procedure The speed of rotation ol thc ephemeris meridian is such that it makcs one
is followed, for example, in predictions of the general circumstances of eclipses. complete revolution of 360', relative to the mean equinox, in 23h56'04i098904 of
Equation 2.24 1 gives the hour angle of the equinox at 0n UT, referred to the dynarnical time. The ephemeris meridian coincided rvith the Greem'ich meridian
actual geographic meridian of Greenwich. However, if 7 is expressed in dynamical at some date between 1900 and 1905.
time rather than Unir''ersal Time. Equation 2.24 1 gives the hour angle of the Apart from its practical adrantages, the concept of the ephemeris meridian
equinox at 0" TDT, referred to the ephemeris meridian. is valuable in providing a clear picture of the relation between dynamical time
The Universal Time of transit of the Sun, Moon! or a planet across thc rrreridian and Universal Time. Figure 2.25.1 shows the relationship betwecn the ephemeris
of Greem.ich can be found by subtracting AZ from the dynamical time of Green- meridian and the universal (or Greenwich) meridian. It also shows the relationships
wich transit. To determine the dynamical time of Greenwich transit, the tabulated between the local meridian, the equinox, and the truc and mean Sun.
time of transit across the ephemeris meridian must be interpolated from the ge-
ographic longitude of the ephemerismeridian (1.0027384? east) to longitude 0'.
2.26 JulianDate
The Uuiversal Time of Greenwich transit is given approximately bv
To facilitate chronological reckoning, astronomical days, beginning at Greenwich
UT of Greenwich transit = TDT of ephemeris transit + (AZl lz)Aa, (2.25-l) noon! are numbered consecrrtively from an epoch intended to be sufficiently Iar in
the past to precede the historical period. The number assigned to a day in this
where Aa is the first difference in tabular rieht ascension and i is the tabular continuous count is the Julian day number, which is defined to be 0 for the day
interval. starting at Greenwich mean noon on 1 January,1713 B.c., Julian proleptic calendar.
It is possible to define hour angles, sidereal time, and an cquation of time with The Julian day number therefore denotes the number of days that has elapsed, at
respect to the ephemeris meridian. But to relate such a system to observations Greenwich noon on the day designated, since the preceding epoch The Julian date
from the surface of the Earth, the value of AZ is required. Jt is more convenient (JD) corresponding to any instant is, by a simple extension of this concept, the
56 EXPLANATORY
SUPPLEMENT

Julian day number followed by the fraction of the day elapsed since the preceding
ooon.
Julian dates can be expressedin Universal Time or dynamical time, though the
timescale should be specified if precision is of concern A Julian date expressed in
UT will differ by AZ from the same date expressed in TDT. It is rot recessary to
:': '
know the exact difference between dynamical and Universal Time when specifying r
;[.\
--^-."
the Julian day number in dynamical time. The timescale, when it needs to be I
specified,should be given after the Julian date; e.g-, JD 24515'15.0 TDT. \1J? ltlt

Prior to 1984, the terrn Julian ephemeris date (JED) was used to distinguish the
Julian date in ephemeris time, with the day beginning at 12hET, from the Julian
date in Universal Time, with the day beginning at 12' UT. The Julian epherneris E

date is equivalent to the Julian date in Terrestrial Dynamical Time.

The fundamental epochs of celestial refetence coordinate systems are properly
on Barycentric Dynamical Time (TDB). Thus J2000.0 is 2000 January 1.5 TDB, q 699 1:-
which is JD 2451545.0TDB. a
In timekeeping laboratories, for many space activities, and for other applica-
tions, a Modified Julian Date is often used. It is defined by MJD = JD - 2400000.5.
Thus a day of MJD begins at midnight of the civil day. The timescale should be z
specified when precision matters, e.g.. NIJD 47479.25 TAI. Note that NIJD is used
in Table 2.58.1 (pagc 86).

2.27 Ttme Zones

Universal Time is equivalent to the standard time for 0' longitude, u'hich is defined
to be the meridian through Greenrvich, England. A rvorlds'ide system of standard
time zonesis basedon 15' (i.e., one hour) incrementsin longitude. These standard
time zones are identified with letter designations. in addition to appropriate names.
Thus the standard time zone for 0o longitude is laboled Z Thc standard time zone
for l5o east longitude is labeled ,4. Subsequent letters designatc zorres increasing
eastward in l5o increments, $'ith M designating the zone for 180" cast. The time
zone for west 15'is labeled 1y',1a'ithletters progressing to )/for 180" r'est longitude.
;i
In practice, however, each country selects the appropriate time zone, or zones, z ! !
o P =;
for itself. The exact zone boundaries within each country are determined through
that country's internal political process. In addition. sorne countries disregard the E
aforementioned convention by adopting a zone on a fraciion of an hour. Nrlanycoun- i - 9a')
tries also adopt daylight saving time, sometimes called summer time or advanced z ;
time, which is generally one hour in advance of the standard time normally kept :
in that zone. The dates for changing to and from daylight saving timc vary among
the countriesl a few countries retain advanccd tirne all year.
Figure 2.27.1 is a world map ofthe different time zones. Since the zonesadopted
by individual countries are subject to change, details of the map may become ob
solete. Fignre 2.27.l
World map of !ime zones
58 EXPLANATORYSUPPLEI\4ENT 2 | rtME 59
DETERMINATIONS
2.3 PRACTICAL OF TIME is the ideal or reference frequency, Q the circular frequency in rad/s, and /(r) the
phase error of the clock. The clock error is then

Today, International Atomic Time (TAI) represents the most precise achievable
determination of time. This timescale is completely determined by applying the x(t)=0(t)/o, (.2.3r4)
methods of physics to the determinatioD of time. However, we live and make astro-
nomical observations on the surface of the rotating Earth; thus, it is necessary to and the instantaneous normalized frequency error is
relate TAI to the less consistent timescale deiermined from the Earth's rotation.
The most recent information on time standards and methods of disseminating y(t)=+=a 1dd:,t\ (2.3r-s
time and frequency is given in the volumes of Proceedings of the Precise Time and dt dt

Time Interval (PTTI) Applications and Planning Meeting and the documents of
The average normalized frequency error between tl and t2 is given by
CCIR Study Group 7. A review of frequency standards was given by Audoin (1984).

'v =
r(r2) r(rr) - | f' . .
I vtt)dt. (2.31-6)
t2 tt 12 tt J,
andClockPerformance
Standards
2.31 Frequency
All measures of clock performance will be related to y(/) or t(t), rvith age, euvi-
To produce uniform time intervals, a clock must generate a frequency that is accu- ronmental conditions, etc., as qualifying parameters. However, for various reasons,
rately known with respect to a standard. The standard frequency for atomic time frequency aspects y(r) have been emphasized over accumulated clock error x(t). The
was established by Markowitz ei al. (1958), who measured the resonance frequency basic problem in an1' measure of clock performance is the fact that .{(l) as well as
of Cesium in terms of the ephemeris second. The curreltly adopted standard is a y(t) mav be unbounded as / goes to infility.
hyperfinetransition of the Cesium-l33 atom (F = 3, Mp = Q to p = {, Mn = 0), which For details ol characterization of clock performance see Allarr (1985), Barnes
is assumedto be 9192631770 Hz in a magnetic field of H = 0. A description of the (1976), Altan (1983), Jenkins and Watts (1968), and Loeb (1972).
physics involved and a review of frequency standards are given by Mockler (1961),
Beehler el ol. (1965), and Vanier and Audoin (1989).
No clock is perfect. Since all clocks are subject to random-noisc rate variations 2.311 \pes of Clock Noise Various types ofrandom clock disturbances have been
and systematic effects, a perfect ciock is an ideai that can only be approached. recognized. Some of these are called rvhite phase noise, white rate noise, and random
but not reached. Improvements in clock performance require improvements in the walk in rate. Barnes and Allan (196,1)rccommended rneasuresof clock perlbrmance
signal-to-noise ratio of the carrier power to phase side-band power. Given these that do not depend critically on darta length or the cut-ofl frequencies in the per-
facts, it is necessaryto be able to specify clock performance. formance of drifts. One solution is to take the average of the r'ariances of only trvo
We first consider a clock as a generator of a nearly sinusoidal signal voltage: samples. This statistic is now gcncrally accepted and designated either pair vatiancc
or Allan variance. This is defirrcd bv

VU) = [Vo+ €(r)]sinO(r). ( 2 . 3 1r )

o1r;=^ frla 11 r)2, (2.31r-1
For our purposes, e(t) r'ill bc very smail or will vary slowly and we can ignore it for
most timing applications; r is the ideal time or the timc of a reference clock. We
can write for the instantaneous phase where 11 and y1 I represent successivesamples y computed as in Equation 2 31 6,
but with the conditions that /2 /l = i be the same for all samples and that therer
is no dead timc between samples.
@1r;= 621.. 61rr, (2.3r2)
Allan (1966) investigated the depencleuceof the general sample variance on the
parameters of dead timc, integration time, number of measurements in the sample,
where
and measurement-system bandwiclth. Further details are found in Barnes el al
F=Cll2tr (2.3r-3) (1971), Howe (1976), and CCIR Report No. 580.
60 EXPLANATORY
SUPPLEMENT 2 / ltME 61

2312 Clock Mod€ling The Allan variance (Equation 2.311-1) is a function of r. size and temperature (Peters and Washburn, 1984). Nerertheless, sorne hydrogen
Starting with sufficiently small values of r (a fraction of a second) and letting z masers equipped with automatic tuning of the cavity and kept in ternperature-
increase, we observe that, for all clocks, o(r) is first decreasing (frequency white controlled rooms have a long-term stability of the same order as the best cesium
noise), then reaches a smooth minimum (frequency flicker-noise floor), then in- standards. Ramsey (1968) gives principles and scientifrc background for masers that
creases(frequency random walk). The values of r corresponding to the flicker-noise make it clear why the hydrogen maser has gained such high interest among physi-
floor indicate the ability of the clock to ensure short-term or long-term stability. cists. Kleppner (1965) gives a discussion of the various physical efiects goverrring
The knowledge of a(r) versus r is essential for deciding which clock should be used the maser's behavior, along with detailed technical considerations.
in applications. It is also necessary to recognize abnormal behavior (Percival, 1976)
in order to predict the clock frequency and to maiatain synchronizations. Methods 2.32 Rubidium Vapor Cells The rubidium clock is an appropriate device rvhen
of optimal prediction have been discussed by Barnes (1976). Further referencesare a relatively low-cost clock is needed that has better stabilitv than a qrrartz-crystal
ll per day under the
Box and Jenkins (1970) and Anderson (1975). clock. The rubidium clock can reach a stability of 1 r l0
The methods of characterizing clock frequency can also be applied to computed best conditions, but it is subject to temperature- and pressure-induced frequency
timescales, such as TAI, under the assumption that such timescales are the output variations. Ringer el al. (1,975)describe the design and perforrrtance of a clock for
rr per day.
of fictitious clocks. the GPS satellites.This clock has achiev"d stability of 2 r l0

2.325 Mercury Ion Frequency Standard The mercury-ion frequency standard trses
of AtomicTime
2.32 Measurement
ions that are confined in a small region of space by an electromagnetic field trap.
2.321 Quartz-Crystal Oscillators Although they measure molecular rather than Thus the particles can be observed without having them collide t'ith the walls,
atomic resonances,quartz-crystal oscillators play a fundamental role in timekeeping. which would disturb the atontic tesonance. The mercrrry isotope Hg-199 has an
They are almost indispensable sources of stable signals in all systems, including extremely narrow microwave resonalce line at 40507 MHz. Although this type of
atomic clocks. They can be free-running or locked to a reference signal, with time frequency standard should be a large improvement over the cesium starrdard, it has
constants ranging from milliseconds to days. The performance characteristics may an unfortunately low signal-to-rtoise ratio, resulting in Iimited short-term stability.
range from 10-4 to 10 1l in frequency stability per day. Gerber and Sykes (1966) However, the long-term stability is very good, since integration can take placc over
give a review of quartz-crystal oscillators. a number of days (Winkler, 1987).

2,322 Cesium Beam Standards Cesium-bean frequencl' standards can be divided

2.33 EarthBotationMeasurement
into two categories. One is laboratory cesium-beam frequency standards. These re-
alize the second with the utmost accuracy (currently, 1.5 x 10-ta) and are, there- At the present time accurate determinations of UTI and polar motion are based on
fore, stable in the very long term. As of 1988, operating instruments are located VLBI, satellite laser-ranging, and lunar laser-ranging data. The difierent techniques
in Canada, the Federal Republic of Germany, Japan, the United States, and the are shown in Table 2.33.1.
USSR. Other instruments are under construction in several countries. The other
category includes commercially available frequency standards. These devices are 2.331 Optical methods Universal Time can be determined from transit instru-
Iess accurate (2 x l0 t'?), but may eqrral the laboratory stanclards for stability up ments, in particular the Danjon astrolabe and the photographic zcnith tube (PZT)
to about one year. The typical stability ofsuch high-performance clocks is 3 x l0-ra (Markowitz, 196C). The methods used are those of classical spherical astrolomy
for a tirne interval of one day (Percival, 1973). as presented in textbooks (Mueller, 1969; Woolard and Clemence, 1966; Danjon'
1959). Optical observations can determine UT to about 5 ms of time.
2.323 Hydrogen Masers The maser is the instrument of choice for applications The classical optical methods are restricted by three major problems: atmo-
where the utmost stability is needed for integration tirnes from ten seconds to ten spheric refraction, star-catalog errors, and spurious deflections of the vertical. Cur-
days, regardless of cost. The optimum stability reaches about 1 x 10-15 for integra- rent estimates of the capabilities of the various methods for Universal Tiure and
tion times of 1000 to 10000 seconds. The clock transition is the transitiol F = l. polar motion are given in Table 2.33.1. As there are generally correlations between
MF = 0 to F = 0, Mr = 0, with a transition frequency f h = 1420405751.768 Hz (Rein- successivemeasurement errors, increasing the averaging time by a factor J does not
hardt and Lavanceau, 1974). The frequency of transition is dependent on the cavity result in a decreaseof the uncertainties by f-1l2.
 62 EXPLANATORYSUPPLEIMENT 2 / T \rE 63

fable 2.33.1 lerotrclrl (VLBI) and Corrnected Elcrncnt Interferolnetrv (CEI) carr bc used to
Earth's Rotation
Techniquesfor l\.4easuring determine Unirersal Time arrd polar motion t'ith cxcellent resoiution. VLBI of-
lers higher precision. In addition, it provides synchronizatiorr betn'een clocks at the
Technique Polar Univelsal Averaging Measurement
Motion (" ) Time (ms) Time Conditions participating obscr-vatoriesto a precision of a fraction of a nanosecond. But since
the observations from participating observatories must be correlated, VLBI has the
Astrolabe 0.06 4 2 hours one inskumenl disadvantagc of a delay between the iime of observation and thc availabilitv of the
PhotographicZenith 0.04 4 2 hours one instrument
IUDE
results. CEI techniques have the advantagc of beillg less expensive and providing
Optical l\,4ethod 0.01 0.8 5 days networkof 80 instruments immcdiate results.VLBI techniqrresare describedby Cotrnselman(1976)and -\{eeks
Doppler 0.01 2 days 1 satellite,20 stations (1976). Short baselinetechniqucsare describedby Johnston (197,1).
0.01 '1 3 days one baseline
CEI
VLBI 0.001 0.05 24 hours networkof 5 or 6 stations
0.1 I nour 2 stations
SalelliteLaserRanging 0.001 0.1 3 days nelwork,- 20 stalions TimeDeterminations
2.34 Dynamical
(LAGEOS)
LunarLaser-Ranging 0.1 'I day l inskumentt
Thc IAU (1976) definition of d1-namicaltime (seeSection 2.22) relatesthe dynarrr
* LJndergood meleorclogicalcondilions ical time of solar-system ephelnerides to thc timescale of terrestrial atonic physics.
Horrcver, distinctiols lnltst be made betn'ccu an ideal dvnamical time, *'hich is
the indcpendent l'ariablc in the equations of rnotiou for the solar svstem, and both
TDT and TDB, rvhich are clrrrentl-y defined as ideal forms of atomic titrc.
2,332 The Doppler Method The Doppler satellite system lrses as a refefence a
In principle. thc tirue in the equationsof motion of the Sun. \l[oon, and planets
rrctwork of stations distributed or''er the globe. The system does )t have a lixcd
coukl <iivergefrom thc lirrre determined from obsetvations of phenorrxxla of terles-
relerqrcc in space and is dependent upon the dynarnics of thc satcllitcs and the
trial physics.At thc prcsent time. obserrational determinationsare not sufficienilv
knowledgeof the positions of the Earth-based stations. A rcporl on this method
accurate to indicate sucir a slrstematic diflerctrce. If a true diflcrcrrce q'ere detccted.
is given b1. Andelle (1976). The Doppler ruetlrod providcs length of day and polar
the scientific com[lurrity would have to dccide hou'to acconrrnodate that dillerence
motion r'alues. but cannot determine L'niversal Tinrc.
An ideal d-v-namicallirne can be determirrcd onll' b1' analysis of observal,ionsovcr
au extended pcriod of tirne.
2.333 Lunar Laser-Ranging Laser pulsc rarrging bett'een Earth-based telescopes Occultatiorr obser'\'ations.specifically observations of the time t'lten thc l'[oon
aucl rettoreflectors installed on the surfacc of the NIoon is currently done to a preci- passesin front of a star, provide thc rnost acculate currcut rnethod lor in'esl,igating
sion of approxinratelv 2 crn. A capability of determining Universal Tiure to a fraction dynamical time. The observationsrcquile a recording of tire timc lr'hen the occrrl-
of a millisecond w-asdemonstlatcd by Silr-erberg (197a) and is eflectivell' shorvrrbJ' tation took place and the location of the obserr-er.The predictecldYnar[ical time
recont results (Newhail el ol.. 1988;Veillet and Pharn Varr, 1988). of the occultation can lle cornpared rvith the recorded atomic time of 1,heobservcd
Because of the precisiorr of the measulemeuts. thcsc obsorvzltiols z[e vely serr- phenonrenon to provide inforrnation concerning tlrr: clifierencc betrccn dyrtami<:al
sitive to the dynamics of the Earth Nloon systcrrr. They provide very accurate tiure and Irrternational Atonic Time (TAI).
co[ections to the thcory of the Nloou's rnotion and librations and the theorv of A differcnce betrveerr<lyttamical time and TAI could also be revca]ed if difli:rerrt
gencral relativitv. Othcr significant pararncters are those that afiect the positions methods of detcr-rriniug lJniversal Timc give discordanl results. [n particular'. a
of the Nloon based rctroreflectors and the Earth-basecl telescope inclucling Earth disagrcement betrve<:nLutral Laser-Rarrgirrgand Very Loug Baseliue Intcrferometr-r"
ticlcs, contincntal drift, plate motion. and polar motion. The observations are also could indicate nonrnodelecl r''ariations of the \{oon's right asccnsion. Ihis n'ould
serlsitivc to the constant of preccssion,the theory of nutation, and the Earth's orltit. provide a <letermination of dynamical time.
All of these parameters are correlatcd irr complex lvavs. Rrture analysis of high-prccision planetary obserlations could reveal a cljfTer
ence in the timescales if differences betrvccrt predictcrl and observed positiorls are
2.334 Radio lnterferometry lhe r-otational posil,idr of the Earth is neasuled with systematically proportional to the r-nean rnotioDs of the planets. At the prescnt
respect to a plane paralhl u'ave frorn a cosrnic radio source that is so far awav thait time. the plarretary lcsiclualsdo not indicate such a systemati( deviation, anrl the
it can be safely assumed to be rvithout proper ntotion. Very Long Baselirrc Inter- possibility is only a conjecture.
64 EXPLANATORYSUPPLEMENT 2 I TIME 65
OF TIMETRANSFER
2-4 METHODS 2.41 RadioTimeSignals

Radio time signals continue to be a simple means of distributing time. The trans-
The determination and maintenance of a timescale at its source is a distinct prob-
mitted time is UTC, rvith a code or voice transmission of DUT1 so that UTl can
Iem from the dissemination of that timescale to many sites around the world. The
be obtained using the relation UTI = UTC + DUTI. A number of countries have
appropriate method of disseminating the timescale is Iargely dependent upon the
standard-frequency transmissions with a wide coverage and an extended period
accuracy required. Also, the appropriate method can be expected to change with
of continuous trarrsmission. The common frequencies for transmission are 2.5, 5,
time as new and improved technologies become available
10, 15, 20, and 25 Mc/s. However, a number of countries also have transmissions
Although this section will discuss the various techniques currently in use and
at special frequencies and specific times. Report 267 5 of CCIR Study Group 7
the accuraciesachievable, it can be expected that the contents of this section will be
and the Annual Report of the time section of BIPM give the characteristics of
one of the first to be out of date. The most up-to-date information on this subject
standard-frequency and time-signal emissions in allocated bands, and characteris-
can be found in the documents of the Consultative Comnittee on International
tics of stations emitting with regular schedules with stabilized frequencies outside
Radio (CCIR) Study Group 7: Standard Frequencies and Time Signals. A review
of allocated bands.
is given by Klepczynski (1984). A CCIR. Handbook on satellite time and frequency
The accuracy that can be obtained is severely restricted by unknown r'ariations
dissemination is in preparation.
in the travel time of the radio signal, particularly on the higher frequencies. The
There are two main techniques for transferring time: transportation of an op-
received signal may contain, in varying qrrantity, signals that have travcrsed differ-
erating clock and sending of time signals. The latter technique subdivides into
ent transmission paths. Reception conditions may difier widely between difierent
one-way and two-way methods.
reception sites and, at arly orle site, there may be a considerable diurnal variation.
In the one-way methods, knowledge of the propagation delay is required and is
The best result is generally achieved by measuring transmissions at different tirnes
often a limitation to the accuracy. For instance, the traditional high-frequency time throughout the day and by taking suitably weighted means. Anomalies are partic-
signals are subject to delay variations and an accuracy of only 1ms can be achieved. ularly troublesome when sunrise or sunset occurs on the transmission path; these
But in the use of the Global Positioning System (GPS) satellites, the propagation times should be avoidcd if possible. The most serious discordances may occur n'hen
can be well-modeled, and deternination of time from a specific satellite has an the reception site is within the skip area of the transmitte., since the predominating
uncertainty in the 10-20 ns range signal may be received bv backscatter from a distant point, cither on the grorlnd
One mode of operation of the one-way technique is to receive the same sig- or in the ionosphere.
nal at two or more sites. After the exchange of data, the difierence of readings of
tlte clocks is obtained at the emitter. Only the difierential delay of propagation
is needed. This method allows for accurate time comparisons using the signals of 2.42 Portable Clocks
commercial television, LORAN C, GPS and geosynchronorrssatellites, and natural
a^stronomicalsources such as VLBI. The use of portablc artomic clocks for thc corlparison of tinr€rscaleshas been the
Two-way techniques involve the nearly simrrltaneous exchange of time signals most reliable and accurate rncthod. It has also been one of the mosl expc'nsive
back and forth, along the same path. The uncertainties are reducc<l because the techniques on a cost-pel-comparison basis. Portable clocks have b<en ncc*ssarl for
propagation delay is eliminated in the data processing. Usually the method is ap- calibrating path delays for svnchronization by other means. Portablc atontic clocks
plied using a satellite to relay the signal. This requires complex equipment. Laser were first used by Redcr and \\iinklcr (1960) in experiments conductecl in 1959 and
pulses can be used either for a terrestrial link or for a link involving reflection on 1960. These early atomicrons were 7-feet high an<l weighed sevcral hundred portnds.
an artificial satellite. In the USSR, the two way technique is routinely uscd with re- They were portable in the sensethat they coulrl be transportctl u,hile mnning, but
flection on meteor trails, with accuracies reaching 20 25 ns. When the uncertainties they required cxtensive external power. In tlic carly sixties. the size of cesium-beam
of the time transfers are srnaller than 1,us, relativistic eflects must be considered, clocks was reduced enough that two nrerr could, u'ith diffiorlty, transport a clock
as explained in Section 2.47. This was demonstratedby Bagley and Cutler (196,1).
66 EXPLANATOFIY
SUPPLEMENT 2 | rtME 67
In 1965 the U.S. Naval Observatory establisheda portable clock service,which a LORAN-C time comparison is on the order of 50-100ns for one day, with the
has provided thousandsof clock synchronizations.The portable clock consistsof a uncertainty increasing to 300 1000ns for one hundred days. This increase is due
commercial cesium-beam oscillator, a counter or clock, and a pou'er supply ll'ith to annual q'eather-related effects over the LORAN C ground-wave path. Changes
rechargeable batteries, which are capable of approximately eight hours of continu- in ground conductivity as well as temperature will change the path delav of the
ous operation without connection to alternating current. The clock system weighs LORAN C ground-wave signal. Uncertainty of the path delay is the physical lim-
approximately 68kg and measuresapproximatcly 0.4nr by 0.,1rn by 0.53m. The itation of LORAN C timing networks, and is independent of the qualitl. of the
usual precautions of minimizing shock, vibration, hcat, cold, arrd magnetic flelds LORAN C system or the LORAN C receiver.
are mandatory to yield high accuracy on portable clock trips.
For a clock trip, the portable atomic clocks are adjusted and kept as close to
2.44 Television
ComparisonTechniques
the frequency of the master clock as possible. The tirne is set so that the diference
between the portable clock and the master clock passcs through zero at the mid- Another technique for time or frequency comparisons is television (TV) transmis-
point of the trip. On the day of departure. clock measurements are made against sions or high-porler, rvide-bandwidth communication links that broadca^stin VHF
the reference clock systems of the master clock with a resolution better than a and UHF frequency bands. (These bands have small propagation delay lariations.)
nanosecond,and with a small, portablc, ten-natrosecondcounter that is taken on Television receivers are incxpensive and readily available. There are two basic types
the trip. During the trip. clock data are recorded for each individual clock. Wherr of TV transfer links. The first is the differential transfer link. in u'hich both clocks
the clock is returned, it is zrgain compared rvith the rnaster clock. Once post-trip are in common view of a single TV transmitter. The second includes a microwave-
data are available, both the frequency offset of the portable clock and thc reductiorr network link as part of the differential path. The common-vie\\' transfer link is
and analysisof the trip data can be completed.For accurateresults, the changesin much rnore stable. since onlv the different propagation paths through the atmo-
the gravitational fields on the clock aud the clock r,'elocity relative to ground-based sphere affect the measurement. Long-tern] time-transfer stabilitl of 10 150ns can
or statiouary clocksmust bc considercd(seeScction 2.47). be expected. Such transfer links have been used for comparison for evervthing from
remote clocks to standard timing laboratories.
2.43 LORAN-C A clock-comparison lirrk that includes a network microriv'avesystern will have
greater time-transfer uncertainty. The differential path includes delays from uscr
The LORAN C navigation s1'stemhas been routinell'used as a time-transferstan- to transrnitter. delays from studio to transmitter link. delays lrom an internediatc
dard for time and frequencycoordination since 1969.The LORAN C system con- studio, delays through multiple microlvave repeaters to the net$'ork origination
sists of chains, each of which has a mastel transnlitter and two to four secondary, or point, and delays from thc transmitter to the other user. Except for the delay
slave. transmitters. Each of thc master transmitters is separated b.v several hundrecl between the user and the transmitter. the delays are subject to step changes lront
kilometers.The mastcr transrnitter emits a selics of coded 100-kHz bursts that arc a hundred nanoseconds to several-hundred microseconds.
repeated a,fter specificd and controlled delays by the secondary stertiols. Bursts of To ensure that the TV programs remaiu on the air, the links arc redundartt
100-kHz transmissions enable the receiver to scparatc t.hestablc ground wavc signal for reliability. Therefore. if there is a bad studio-to-transmitter link. a l.'ackup link
from the sky-rvave signal that arrives ,10ms or more later. A nnvigatiorr or timing rvill be sl'itchecl in its place. These changes to the backup c?rDcause dilTerenccs
receiver will track the third cycle of the received bursts. Each LORAN C station in the delay of as much as 500 nanoseconds. Due to this redundant chalacteristic
has a cesium standard for cleriving tirtre signals. The cotrrplete chain is ovorseen b1' of the TV networks, thc networks are largely unsuitable for high-accuracy tirrx:
one or nlore monitor stations that nteasureatrd control the emissiondelarls.Navi- coordination.
gation ac<:uracyu'ithin the grorrnd-mvc covcrageof tlx: chain is 0.10 0.25 nautictrl Another factor that reduces the use of TV-network transfer links is the use
Iniles. Ground-*'avc coverageof the LORAN C transrnitter is highly rarinble, and of digital frame synchronizers by the network stations. The purpose of the frarne
is a fuDction of thc radiated porver (275 1800 kilou'atts) and the conductir.itl. of synchronizer is to synchronize the netu'ork picture $'ith all locally generatcd signals,
the path. Over land, ground-wavecovcrageis typicallv 500 1000km; ovcr water thus preventing picture rolls when switching between a network program and a local
this coveragecan be over 2000km. Since each LORAN C chain is independent. program. A digital frame synchrolizer continuously digitizes and stores an incoming
tirne cornparisoDswith LORAN C errerestricted to laboratories that haye com- TV signal. The stored digital information is continuously read out and converted
lnon ground-wavcviews of a singlc chain. With a single chain, the monitor stations back to an analog video signal under control of the station's master synch generator.
maintain self-consistenttiDring to bettcr than a microsccond.The unccrtainty of The synchronizing pulscs for the nctwork picture have therefore been replaced by
6g EXPLANATORY
SUPPLET4ENT 2 r ltME 69

local station synchronizing pulses. Thus network s1'nchronizing pulses that are used With one-u'ar'l.ilne transfer, the accuracl of the method is basi<'alll linrited
for the time transfer have been lost. to the acculacl. of the ephemerisor the knouledge of the satellite s positiorr.Il a
two-&'a\.time tlansfer. a time-interral countcr is located at each Earth statiorr.The
counter is startcd by a pulse from a local clock and stopped b1'a pulsc rc<civeclvia
2.45 Useof Satellites
the satellite frorn the remote clock. Thus the time interlal bett'een pulses involves
Both the Tfansit satellite systern and the Global Positioning Svstem (GPS) can the transmitter delay, the propagation delay between the Earth station and the
be used for one-way time transfer. Each Trausit satellite contirruously transmits its satellitc, the delay in the satellite for the signal received, the propagation delay
current ephemeris encoded by a phasc modulation on two stable carrier frequencies between the satcllitc and the other Earth station, and the same list of clclays on
of approximately 150 MHz and 400 MHz. The navigational information is broadcast the return direction. In addition. the effect of the Earth's rotation nrust be taken
in two-minute intervals, which begin and end at the instant of each even minute. into account, bccause during the period of the time transfer. both Etuth stations
An encoded time marker is broadcast. with time uniquely marked at the instant of have moved and the satellite has rnoved. Thus each leg of thc coniperrison lnust
the even minute. be considcred individualll', and thel cannot be assumed to be equivalcnt to the
The broadcast frequencies are based on the highll'precise rraster oscillator. corresponding Ieg at a different time. The dificrential delay' eflects czrttscdbl the
The Transit satellite system is cxpected to be replaced by the NAVSTAR GPS Earth's atmosphere and the difierent frc<luertcicsof the difierent legs shoulcl not
navigation system. This system calls for a constellation of 18 satellites that are exceed a feu, nanoseconds, whether due to geonetric effects or to volocity effects
distributed in orbital planes inclined at 63'with respect to the Earth's equatorial caused bv the ionospherc and troposphere. At frequcncies of 20 30 GHz, lhc ef{ects
plane. Although the GPS system is designed for navigation capabilitics, it also of humidity mery cause significant inequtrlitics for tirne transfers at thc rtatrosc<on<l
coordinates most of the major timescales in the rvorld. The GPS system has an level (Oaks. 1985).
accuracy of about 10 ns for ten-day averaging. The most accruatc titne comparison
by GPS is basedon a common satellite procedure.GPS satelliteson a fl1'b1-mode
make two separatecomparisonstith respectto the GPS satellitc clock. A linear rate
for the satellite clock is assumed.This nrethod does not provide the same accuracl
as the cornmon-r.iew procedure, since the clock error and iorrospheric errors r.ill 2.46 Intercontinental by VLBI
ClockSynchronization
be present and more significant. The error budget for a comnrou-vieq' GPS time
transfer is eiven in Table 2.45.1.
With Very Long Baseline Interferometry (VLBI). radio signals frorn celcstial objects
are recorded at individual remote radio tolcscopes.The observational datl are then
Table2.45.1 brought togethel for correlation so that l)recise values for the differeltrr:s betu'cen
Error Budgetsfor Common View GPS Time Transfer the clocks at the individual antennas can bc determined. The fundarlentals of
VLBI have been dcscribedbv Klempcrcr (1972). rvho also presentsan extelrsivelist
Typeof Delay Besl Case Worsl Case
of references. Scvcral VLBI experimcrttal programs. and the funclanerltals of the
BMS RIVS
process,were describedby Clark (1972).FreqLrcncy compatisonsrvith an accur.rc]'oi
'lb
SatelliteEphemeris 3ns 10ns 10-13to 10-la and clock synchronizationof the order of 1 ns are possibk:. acrhieve
lonospheric 2ns 100ns this accuracy, 110-megahertzbandwidths arc required. These cart be aclticvcd by the
Tropospheric ns 20ns
User-Position ns lns bandwidth synthesis techniques described by Hinteregger et al. (1'972) and Rogers
Multipath ns 2ns (1970). The VLBI processrequires distinct measurementsbased on prcarranged
Receiver ns lns experime[ts involving difierent locations. The results of the time co:nparisous art'
Signal-To-Noise 7ns 1ns
TolalRl\rS 4.2ns 103ns not available until the trl,'odata sets are brought together and correlatcd. In theory,
(forsingle13-minute
kack) if sufficient data are available and a sufficiently large bandwidth can be obtained.
the VLBI svnchronizationmethod is mainly limited in accuracy by the difficulty
Note: Some ol ths €rrorsdep€ndon lhe distance(for exampl€lh€
salellileephemeris),Ths besl cas€ appliesto dislanceof
of determining thc overall system delay and the atmospheric or ionospheric delay
2000-3000kmi th€ worsl cas€ is lor distanceol 600H000 km. (Johnston el al.. 1983).
70 2 / T\4E 71
Effectsin TimeTransfer
2.47 Relativistic 2.472 Nonrotating, Local Inertial ReferenceFrame When time is transferred frorn
point P to point Q by means of a clock, the coordinate time elapsed during the mo-
When transferring time betrveen two points, the process can be viewed from one
tion of the clock is
of two reference frames-a geocentric, Earth-fixed, rotating frame or a geocentric,
nonrotating, local inertial frame. Discussions of time transfer vien'ed from these t, = a, ''':; '" . *) () l'71-1\

two frames are given in the following sections, The equations used in the discus-
l,n [,
sionsrepresentclock rates with ar accuracy better than 1 in lOra. In addition, the Here U(r) is the potertial at the location of the clock and v is the velocity of
equations are consistent with the proposal of the Consultative Committec for the the clock, both as vieu'ed-in contrast to Equation 2.471 | frorrl a geocentric
Definition of the Second (see Section 2.211), but they extend the proceduresto nonrotati[g reference frame. U, is the potential at the geoid, including thc effect
heights that include geostationary satellite orbits. on the potential of the Earth's rotational motion. Note that AU(r) / U(r') - Ue,
since U(r) does not include the effect of the Earth's rotation. This equation also
2,471 Clock Transport from a Rotating ReferenceFrame When time is transferred
point portable applies to clocks in geostationary orbits, but should not be used beyond a distance
from point P to Q by means of a clock, the coordinate time accu-
of about 50000km from the center of the Earth.
mulated during transport is

l,f t = If d,sl ,l l -A U, ( r )+. ^ r, 1, '+ .2u^

),{r. t ) L11_1\
2.473 Electrornagn€tic Signals TFansfer from a Rotating Reference Frame From
JP L c' zc'j the viewpoint of a geocentric, Earth-fixed, rotating frame, the coordinate time
elapsed between emission and reception of an electromagnetic signal is
where c is the speed of light; r,,,is the angular velocitv of the Earth's rotation; v is
the velocity of the clock with respect to the ground; r is a vector from the center of I la I AU{r)l 2,a
Ar= I doll--, l+ ,AE. tt 411 1\
the Earth to the clock, which is moving from P to Q; Ap is the equatorial projection cJP c'I
L
of the area swept out during the time transfer by the vector r as its terminus
moves fiom P fo Q; LU(r) is the potential difference beiween the location of the where do is the increment of standard length, or proper length, along the transmis-
clock at r and the geoid as viewed from an Earth-fixed coordinate system (with sion path; and AU(r) is the potential at the poirtt r on the transmission path less
the convention that AU(r) is positive when the clock is above the geoid); and ds is the potential at the geoid (see Equation 2.172 1), as viewed from an Earth-fixed
the increment of proper time accumulated on the portable clock. The increment of coordinate system. As is the area circumscribed by the equatorial projection of the
proper time is the time accumulated on the portable standard clock as measured triangle whose vertices are at the center of the Earth; at the point of transmission
in the rest frame of the clock; i.e.. the reference lrame traveling with the clock. AB of the signal, P; and at the point of reception of the signal, Q.
is measured in an Earth-fixed coordinate system. As the area As is swept out, it is The area As is positive when the signal path has an eastti'ard component. The
taken as positive when the projection of the clock's path on the equatorial plane second term amounts to about a nanosecond for a round-trip trajectory from the
2;
is eastward. When the height h of the clock is less than 24km above the geoid. Earth to a geostationarysatellite.In the third term, 2a/c2 = l-6227 x 10-6 ns knr
AU(r) may be approximated by gh, where g is the total acceleration due to gravity this term can contribute hundreds of nanoseconds for practical values of As. Thc
(including the rotational acceleration of the Earth) evaluated at the geoid. When i increment of proper length do can be taken as the Iength measured using standard
is greater than 2,1km, the potential difierence AU(r) must be calculated to greater rigid rods at rest in the rotating system. This is equivalent to measurement of length
accuracv as follows: by taking c/2 times the time (normalized to vacuum) of a two-way electrornagnetic
signal sent from P to Q and back along the transmission path.
ALttr')= l-l t^"r(/rin2e-ai1alt9u'
c M' \. rf ) ' l r * (rat r. ) ' r l . o , . 2 d -r r l .
at./ 2 2at Ll 2.474 Electromagnetic Signal Thansfer frorn a Nonrotating, Local Inertial Frame
(2.4',71-2) From the viewpoint of a geocentric nonrotating, local inertial frame, the coordinate
In this equation, a1 is the equatorial radius of the Earth; r is the magnitude of time elapsed between emission and reception of an electromagnetic signal is
the vector r; d is the colatitude; GM" is the product of the Earth's massand the

t,=! l,aa" - u,',rt''] ,

gravitational constant; and J2 is the quadrupolemoment coefficientof the Earth
( ' / z= + 1 . 0 8 3x l 0 3 ) . \2.4'74-t)
[r
 EXPLANATORY
SUPPLEMENT 2 | ltME 73
72

where U(r) and 4 are defined as in Equation 2.472 7, and do is the increment of where / is the latitude, ) is the longitude (the positive sense being eastward), and P is the
path over which the radio signal travels from A to B. If the two clocks are synchronized
standard length, or proper length, along the transmission path. The quantities do
by a portable clock, they will differ in coordinate time by:
appea.ring in Equations 2.473 | and 2.474 7 difrer slightly because the reference
frames in which they are measured are rotating with respect to each other.
a-o=[a,(^1 g) ' -'
-\ []"o""4as, (.2.41+-'1
JP \ /

Dxample I Due to relativistic effects, a clock at an elevated location will appear to be

where y is the portable clock's ground speed, and P is the portable clock's path from A
higher in frequeocy and will difier in normalized rate from TAI by
to B.

L.Lh I c2, (2.4'742) This difierence can be as much as several tenths of a microsecond. It is recommended
that Equations 2.174 6 and 2.474 7 be used as correction equations for long-distance
where AUr is the diference in the total potential (gravitational and centrifugal potentials), clock synchronization. Since these equations are path dependent, they must be taken into
and where c is the velocity of light. Near sea level this is given by account in any self-consistent cootdinate time system.

IfaclockistransportedfromapointAtoapointBandisbroughtbacktoAonadifferent
c(.d)ht c2, (2.4'74-3) path at infiniiely low speed at I = 0, its time will ditrer from that of a clock remaining in
Aby
where g(d) = (9.780+0.052sin267mfs2 is the total acceleration at sea level (gravita- ^.t=-2aAElc2, (2.474-8)
tional and centrifugal), / is the geographical latitude, and h is distance above sea level.
Equation 2.474-3 must be used in comparing primary sources of the SI second with TAI
where AE is the area defined by the projection of the round trip path onto the plane of the
and with each other. For example, at latitude 40', the rate of a clock will change by
Earth's equator. Ae is considered positive if the path is tra\€med in the clockt'ise sense
+1.091 x l0-r3 for each kilometer above sea level.
viewed from the South Pole.

Example 2 If a clock is moving relative to the Earth's surface with the speed v, which
Exanple 4 Since
may have the component v. g in the direction to the east, the normalized diference of the
6nskm-2, (2.4'7
4-9)
frequency of the moving clock from that of a clock at rest at sea level is 2 u I c 2 = 1 . 6 2 2 7x l 0

Ir2 ptd)fi the time of a clock carried eastward around the Earth at infinitely low speed at l' : 0 at
-I- ! r / r ( c o ds ) v E , (2.4744)
,F* f the eqnator will difier from a clock remaining at rest by -201.4ns.

where c..,is the angular rotational velocity of the Eaft} (a =1.992 x l0-5 rad/s), r is the
distance of the clock from the center of the Earth (r = 6378. l40km), c is the velocity of
light (c = 2.99'792458x 105km/s), and I is the geographical latitude.
2.5 HISTORICAL OFTIMEKEEPING
DEVELOPMENT

Dxample 3 A clock is moving 2?0 m/s eastward at 40" latitude at an altitude of I km. 2.51 Introduction
The normalized difference of frequency of the moving clock relative to that of a clock at
rest at sea level is The earliest attempts to measure time were probably based on observations of ce-
Iestial and meteorological phenomena: the seasonal motion of the Sun along the
- 4 . 0 6 x 1 0 - r r + 9 . 8 2 xl 0 1 3 l . 0 7 l x l 0 r 2 = 4 . 9 5 x 1 0 r 3 (2.414s) horizon, the progression of Moon phases, the shifting of winds, and so on. Time-
keeping, which we take to mean the measurement of subdivisions of the day, has
The choice of a coordinate frame is purely a discretionary one, but to define coordinate
time, a specific choice must be made. It is recommended that lor teDestrial use, a topocerl- long involved observations of both celestial and terrestrial phenomena. On the ce-
tric frame be chosen. In this frame, when a clock B is synchronized with a clock A by a lestial side is the measurement of shadows cast bv a vertical gnomon, which led
radio signal traveling from A to B (both clocks being stationary on the Earth), these two to the development of sundials. On the terrestrial side is the development of water
clocks differ in coordinate time bv clocks (clepsydrae). The origins of both approaches predate historical records. His-
torical references to water clocks are found in Babvlonian and Egyptian records of
u-o=3f1"o"'oas. (2.4144) the sixteenth century B.c. (Turner, 198,1).
 74 EXPLANATORY
SUPPLET\,ENT 2 trtMe /c
The system of hours developed first in Egypt, where day and night were each the only means of obtaining accurate time was by direct astronomical observation,
divided into 12 parts. This resulted in a system in which the lengths of daylight apparent solar time was in general use. Determinations of local apparent time were
and nighttime hours varied with the seasons.In summer, an hour of daylight was commonly made by observing altitudes of the Sun or stars, Thus apparent solar time
longer than an hour of darkness; in winter, the situation was reversed. Only at was the argument in The Naut'ical Almanac and other national ephemerides until
the equinoxes were hours of day and night of equal length. When this system of the early nineteenth century. Mean iime, when needed, was obtained by applying
temporal hours spread throughout Europe, it became a function of Iatitude as well the equation of time to the apparent time.
as season. It continued in civil use until the spread of mechanical clocks in the The equation of time, in the sense of the correction to be applied to apparent
fifteenth and sixteenth centuries encouraged the gradual acceptance of equal hours. time in order to obtain mean time, was tabulated in the national ephemerides
Our familiar system of 24 equal hours comes from Greek astronomv. At least from their earliest inception. It was used for regulating clocks and determining the
as early as Hipparchus, the Egyptian system of 24 temporal hours was merged with argument for entering astronomical tables. During the late eighteenth and early
the Babylonian division of the day into 360 equal units. The result was a system of nineteenth centuries, as clocks were improved and came into extensive use at sca,
24 equal hours, with each hour divided into 60 minutes (Neugebauer, 1975). This apparent time was gradually superseded in civil use by mean solar time.
established a tradition for astronomical tables that continues to this dav. In the mid-nineteenth century, when mean time was first introduced as the
Although their origin is a mystery, mechanical clocks were in use in Europe argument in the national ephemerides, the equation of time was supplemented by
near the beginning of the fourteenth century. They were driven by weights and an ephemeris of sidereal time at mean noon. This provided a means of calculating
controlled with the recently invented escapement. A regulating device allowed the mean solar time from sidereal time, without using the equation of time. Clocks were
clocks to be adjusted, morning and evening, to maintain temporal hours. They did not sufficiently perfected for the theoretical distinction between apparent sidereal
not provide adequate standards for scientific timekeeping, but were built for public time and mean sidereal time to be of any practical importance. This distinction was
use. therefore disregarded, and the very imperfect expressions then in use for nutation
In the mid-seventeenth century, the pendulum was successfullyapplied to clocks. were of no consequencefor the purpose of time determination.
By the late eighteenth century, clocks t'ere improved to provide accurate time for The Riefler clock, introduced about 1890, was the earliest timepiece with an
scientific purposes and for navigation at sea. accuracy comparable to determinations of time from observation. As the accuracy
Useful surveys of the history of timekeeping are provided by Andrewes and of clocks increased, the explicit recognition of mean sidereal time as distinguished
Atwood (1983), Landes (1983), and Whitrow (1988). from apparent sidereal tirne became necessary.The term uniform sidereal time was
often used at first, but this measure was noi strictly uniform, and the term was
2.52 AppareniSolarlime, MeanSolarTime,andthe Equation
of Time dropped from use.
After the introduction of the Short free-pendulumclock in 1921, the rpmoval
Apparent solar tirrre is the measrrre of time defined by the actual diurnal motion of of the short-period terms of nutation from the observed clock corrections was nec
the Sun. If the Sun's right ascension increased at a uniform rate, solar days would essary in order to check the clock satisfactorily. These terms were included in the
be equal in length throughout the year (assuming uniform rotation of the Earth). ephemerides of the sidereal time at 0h, beginning with 1933.
Apparent solar time would then provide a uniform measure of time. Hotvever. the The equation of time came to signify the opposite of the original concept'
Sun's motion in right ascensionis not uniform becausethe Sun moves in the ecliptic Apparent solar time was obtained by applying the equation of time to the mean
rather than the equator, and the Sun's motion in the ecliptic is not uniform. time kept by clocks, which were regulated by determinations of mean time from
Mean solar time was defined by the motion of an abstract fiducial point. which observations of sidereal time. The link between mean solar time and sidereal time
came to be known as the fictitious mean sun. This point was posited to move was founded on the relation
uniformly in the equatorial plane at a rate that is virtually equal to the mean rate
of the true Surr's motion in the ecliptic. The difference between mean solar time m o a n s o l a r l i m e a l C r o e n w i r - h= C H A Q - R 6 + l 2 ' , (2.s2-r)
and apparent solar time is called the equation of time. It is never larger than 16
minutes. where GHAQ is the Greenwich hour angle of the mean equinox of date (i.e., Green-
The concept of mean solar time, together with the principles for determining the wich mean sidereal time) and R6 is the right ascension of the fictitious mean Sun'
equation of time, extend back at least to Ptolemy. For Ptolemy and his successors, Underlying the concept of mean sola-rtime was the assumption that the rotation
mean solar time was useful for constructing tables of solar motion. But as long as of the Earth was uniform. In the first half cf the twentieth century, this assumption
 EXPLANATORYSUPPLETENT 2lrlvE II
76

was repudiated (seeSection 2.53), rvith the result that mean solar time was no longer rotation of the Earth. Consequenily, the observed hour angle of a celestial bodv
(which increases with the Earth's rotation) diffcrs from the position tabulated in
used in precise timekeepiug. It was replaced by two concepts: ephemeris time was
the ephemeris. An analysis of thc discrepancies between observed and computed
introduced to satisf;r the desire for a uniform measure of time, and Universal Time
(originally introduced to specify Greenwich mean time measrrred from midnight positions reveals the differcnce between a rneasure of the Earth's rotatiol and the

instead of nool see Section 2.521) came to designatea measure of the Earth's measure of a uniform gravitational time. The discrepancies are most evident in thc
rotation. mean motion of the Moon, due to the rapidity of its notion and the accrirercyryith
Over the centuries, the exact definition of mean solar time, and hence of the which the inequalities of motion can be observed. ilorvever, variations in thc Earth's
equation of time, evolved as net'tables for the motion of the Sun were inttodttced. rotation also prodrrce discrcpancies in the observed motions of the other bodics in
In analyzing historical observations that are expressedin mean time, care must be the solar system, in proportion to the magnitude of their respective nean motions.
taken to determine t'hat definition t'as used in the reductions. The flrst variation to be recognized was a secular retardation of the rate of ro
tation due to tidal interaction between the Earth and Moon. Adaurs (1853) shorved
2.521 Greenwich Mean Tim€ Prior to 1925, mean solar time was reckoned from that the observed secular acceleration of the rnean motion of the Nloon could not
noon in astronomical practice. The mean solar day beginning at noon, 12hafter the totally be produced by gravitational perturbations. However, the validity ol his re-
midnight at the beginning of the same civil date, q'as knorvn as the astronomical sults was hotly contested. The issue was resolved when Ferrel (186'1) and Delaunay
day. Mean solar time reckoned from mean noon on the meridian of Greenwich u'as (1865) showed from dynamical principles that the tides exert a retarding action on
designated Greenwich Mean Time (GNIT). Local mearr time (LNIT) was reckoned the rotation of the Earth, accompanied by a variation of the orbital velocity of the
from mean noon on a local meridian. Beginning rvith the national ephemerides for Moon in accordance with the conservation of momentunt.
1925, a discontinuity of 12hwas introduced in the tabular arguments, so the instant Newcomb (1878) consideredthe possibility of irregular variations of the Earth's
designated December 31.5 in the volumes for 1924 was designated January 1.0 in rotation as an explanation of lunar residuals. In the erxl, however, he could not
the volumes for 1925. find collaboration from planetary data (Ncu'comb, 1912). Some of the <lifilculties
In The Nautical Almanac, the designation Greenwich Nlean Time (GNIT) u'as faced by Newcomb, and prior astronomers, resulted from inadequacies in the lunar
still used for the new reckoning, together with local mean time where appropriate, theory. This situation was rcctified when Brown (1919) introduced his nerv tables. A
whereas in The American Ephemeris the designation Greenwich Civil Time (GCT) number of studies offered increasing evidcnce of irregular variations in the Earth's
was adopted, together with local civil timc (LCT). This confusion in telminology rotation. N{ostimportant &-crepapersby de Sitter (1927)and Spen,er Jones(1g:Jg).
was finally removed by dropping both dcsignatious and substituting Universal Time which correlatedirregularitiesin the Moon's motion with irregularitiesof the inner
(UT). Care is necessaryto avoid confusion. To distinguish between the tr,'o reckon- planets.
ings that have both been called Greenwich Mean Time, the designation Greenwich Meanwhile, the accuracY of crystal-controlled clocks was becoming compalable
Nlean Astronomical Tirne (GMAT) should be used for the reckoning fron noort. with that of the rotation of the Earth. By comparing thc observcd rates of the
The designation UT always relers to time reckorred from Greenwich midnight, evcu clocks of national time services, Sioyko (1937) found a peliodic seasonal variation
for epochsbefore 1925. in the rate of rotatiorr.
In addition, it should be ncted that in the United Kingdom, Greenwich Mean The Earth's rotation rate was therefore undcrstood to have seculat. irregular,
Time has meant the civil time or Coordinated Universal Time (UTC). For naviga and periodic variations. Since mean solar time was tletermined directly from sidcreal
tion, Greenwich Mean Time has meant UT1 (see Section 2.24). Thus GNIT has two observations. it could no loneer be considered unifbrm.
meanings that can differ by as much as 0i9. For these reasorrsGMT should not be
used for precise purposes.

2.54 Universal
Time
2.53 Rotation of the Earth
Newcomb (1898) gave the following expression for the right ascension of the ficti-
The continued failure of successivelunar theories to represent the observed motion
tious mean Sun:
of the Moon gradually led to the realization that thc rotation of the Earth is not
uniform. In ephemerides calculated from gravitational theories, the tabular times
r:) 54-ll
are the values of a uniform measure of tirne. quite independent of the variable Ro = l8'38"451836+ 8640184:542
r+ 0:0929f'7.
 78 EXPLANATORY
SUPPLEMENT
2 / ftME 79
where Z is measured in Julian cerrturies fronr 1900 Januarv 0, 0h Greenwich mean terminology was dropped when the distinction betu''een ephemeris time and Uni-
noon. This expression differs from that of the mean longitude of the Sun by onlv versal Time was introduced in national ephemerides in 1960. Equation 2.54 2, with
a slight, progressively increasing excess of 0:0203 f2. This difference is due to the ?.u as ihe independent variable, was used during the period 1960 1983 to calculatc
secular acceleration of the Sun and to the different rates of general precession on tabular ralues ol Grecnwich mean sidereal time at 0' UT. Beginning with ephemer-
the ecliptic and on the equator.
ides for 1964, Equation 2.24 t has served as the defining relation between Universal
By inserting Equation 2.54 1 into the formula for mean solar time, Equa-
Time and siderealtime (seeSection 2.24).
tio1' 2.52-I, and evaluating the result at 0h mean solar time on the Greenwich
meridian, we obtain the Greenwich mean sidereal tirne at 0h rnean solar time:
Time
2.55 Ephemeris

GHAQo= 6h38'45:836
+ 8640I84:542
T + 010929
72, (1 \t-)\ From evidencecompiled by de Sitter (1927) of fluctuations in the Earth's rotation,
Danjon (1929) recognized that mean solar time, which was observationally tied to
the rotation of the Earth, did not satisfy the need for a urriforrn timescale. He sug-
where T takes on successivevalues at a uniform interval of | /36525. But because
gested using the timescale of the Newtonian laws of planetary motion. His call for
the rotation of the Earth is not uniform, values calcrrlated frour this expression are
reform, appearing in a nontechnical journal, was perhaps unnoticed by specialists;
not consistent with observed hour angles of stars.
certainly it was unheeded. Two dccades later, Clemence (19'18) published a more
From the beginning of the century through 1959, these formulas li'ere used
specific proposal, apparently without being aware of Danjon's paper (Clemencc,
in the British Nautical Almanac and in the American Ephemeris to define mean
1971). From the results of SpencerJoncs (1939), Clenence derived the factor AT
solar time and to relate solar time and sidcreal time. As given by Newcomb, the
that would reduce Universal Time to the measure defined by Newcomb's tables. In
independent r''ariable in Equation 2.54 1 is consistent r-ith that of his solar tables.
addition, Clernence determined thc cr-rrrectionto Brown's lunar theory that would
As evidence accumuiated that the Earth did not rotate uniformly. Nen'comb's 7"
make the independent variable of the lunar theorv consistent wittr that of Nerv-
was irrterpreted as the independeot variable of the Earth's orbital revolution rather
comb's Sun. Thus a comparison of lunar observations with Brou'n's (corrected)
than its axial rotation. Since Newcornb's tables were the basis of the fundamental
theory could be used to determine Al.
ephemerides of the inner planets, the independent variable of his tables was called
In 1948 a proposal to establish a rnore uniform fundamental standard of time
ephemeris timc (ET). It becamc associated with the concept of a uniform measure
was referred by the Comit6 International des Poids et Nlesures to the Interna-
of time, based on the laws of orbital dynamics.
tional Astrononical Uniorr. This proposal was considered at the Conference on l,he
To clarify notation, Is (expressed in Julian centrrrics of 36525 ephemeris davs
Fundamental Constants of Astronomy held at Paris in 1950. At this confcrence,
from 1900 Jarruary0, l2n ET) can be used in Equation 2.54 | to signify ephemeris
Clemence proposcd usiirg the measure ol tirre defined by Ncwcomb's Tables oI the
time. Then R6(7s) = Re dcfines a uniformly mof ing fiducial point that is indepen-
Sun. The Confcrerrce adopted a resolution rccommending that this measure of time
dent of the Earth's rotation.
be adopted, be expressed in units of the sidereal vear at 1900.0, and be designated
To make Equation 2.54 2 consistent with observed hour angles of the stars.
by the name ephemeris lime. as suggcsted by Brou*'er. This recornmendation was
Universal Time (UT) was instituted as a measure of the Eartb's rotation. Then
adopted in 1952 by the International Astronomical Union at its General Assemblv
Equation 2.5,1 1 becomes a delinition of UT, with 7u (measured irr Juliarr ccnturics
in Rome.
of 36525 days of Universal Timc from 1900 January 0, 12h UT) scrving as the
Upon further consideration, the tropical year, rvhich was considercd more fun
independent variablo: R5(?.y) = Ru. The difierencc between Rs and Ru is given by
damental than the sidereal year, was chosen as the unit of tirne. Accordingly, the
Comit6 International des Poids et Mesrrres, at its session in Septcmber 1954 in
Paris, proposed to the 10th General Conferencc on Weights and N4easuresthat the
Rs Ru = 0.002738A7, a? 54_t\
fundamental unit of time be the second, redelined as 1/31556925.975 of the length
of the tropical year for 1900.0. The Conference authorized the Comit6 to adopt
where AZ is the differcnce between ephcmeris time and Universal Time (AZ =
this unit after formal action on the definition had beerr taken bv the International
ET - uT). Astronomical Union.
Prior to 1960, the designation "R.ight Ascension ol Mean Sun+ 12h" was some-
The IAU, at its General Assembly in Dublin in September 1955, approved the
times applied to tabulations of Greenwich sidereal time at 0h UT. This inexact definition proposed by the Comit6. However, the tropical year was understood to
 EXPLANATORY
SUPPLEI\,4ENT 2 | rtME 81
80
the theory of the Earth's motion around the Sun, but also in the heliocentric the-
be the mean tropical year defined by Newcomb's expression for the geometric mean
ories of the other planets. Ephemerides derived from Newcomb's tables of Lhc Su1
longitude of the Sun; the value ol the second required for exact agreement with
and planets could therefore be regarded as having ephemeris time as the inclepel-
Newcomb's tables is 1/31556925.97474 of the tropical year. The primary unit of
dent time argument. The mean longitude of any other planet, or even of thc N{ooq,
ephemeris time was the tropical year at the fundamental epoch of 1900 January
could have been used to define the origin and rate of a rrniform time system, and
0, 12h ET. The tropical year was defirred as the interval during which the Sun's
ephemerides of the Sun, Moon, and planets could have beerr constructed with that
mean longitude, referred to the mean equinox of the date, increased by 360'. The
time system as the independent argument.
adopted measure of this unit was determined by the coefiicient of 7, measured in
centuries of 36525 ephemeris days, in New'comb'sexpression for the geometric mean In principle, the ephemeris time at
2.551 The Detennination of Ephemeris fime
longitude of the Sun (referred to thc mcan equinox of date):
any moment can be determined by comparing the observed positions of the Sun,
Moon, and planets with their corresponding ephemerides. The tabular time for
L = 2'79'41'48104+ 1296027
68';13T + l':08972. (2.55-l) which the observations agreed n'ith the ephemcrides q'ould be the ephemeris time.
In practice, Universal Time, which can be detcrrnirred very accurately arrcl rvith
The tropical year at 1900January 0, 12nET accordinglycontained little delay from obscrvations of the diurnal motions of the stars. rvas use<las an
intermediary measure of time. The difference betn'een the tt'o measures of tine,
360t60 t60 Af = ET - UT, was thcn obtained by comparing observations *'ith ephemerides.
9747ephemerisseconds. (2.55-2)
x 36525x 86400= 31556925.
129602168.13 Observations of the N,Ioon,whose geocentric motion is much greater than those
of other bodies, provided thc nlost accurate determination of A7. Ilo*'ever, direct
Consequently, the Comitd at its sessionin Paris in October 1956, under the author-
comparison rvith the lunar ephcmcris calculated from Brot'n's lunar tables did not
ity given by the 10th Geleral Conference, adopted in place of the value formerly
give AT immediately, since Brot'rr's theory was not strictly gravitational and his
recommended a slightly rnore precise value of the tropical year at 1900 January 0,
tables were not in complete accord with Newcomb's solar tables. This was rectified
l2h ET: "La secondeest la fraction 1131556925.97 47 de l'ann6e tropique pour 1900
by eliminating the empirical term from the mean longitudc ol the N4oon and by
janvier 0 i, 12 heures de temps des ephernerides." (ProcEs Verbaux des S6ances,
applying the correction to thc tabular mean longitude:
1957).
At this session,a Comit6 Consultatif pour la Definition de la Secondewas estab-
L L = 8 l ' 1 2- 2 6 . ' 7 4 T 1 1 . 2 2 7 2 . (2.ssl-r
lished to coordinate the work of physicists on atomic standards and of astronomers
on the astronomical standard of ephemeris time. In 1960, the 1lth Conf6rence Gen-
Additional corlcctions u'ere needed Ibr souie of the periodicr terms in longitudc.
erale des Poids et Mesures formallv adooted the eohemeris second as a fraction of
latitude, and parallax. Beginning in 1960. the lunar ephencris q'as calculated from
the tropical year.
this amerrdcd theory directiy frorr thc lheoretical exprcssions for tlx: longilrtde'
At its 1958 General Assembly in Moscorv, the International Astronomical Union
latitude and parallax instead of from Brown's tables. Thc improved r:phernelis was
adopted the following definition for the epoch of ephemeris time: "Epheurclis time
made available for 1952 1959 in the Improued Lunar Eph,emeris(Narutical Alrnanac
is reckoned from the instant, near the beginning of the calendar year A.D. 1900,
Oflice, 195.1).
when the geometric mean longitudc ol the Sun was 279"41'48:04, at u4rich instant
A distilction llrd to be drawn betw'ccnUT'+ AZ and E'f' whcn AT was dcter-
the measure of ephemeris time was 1900 January 0dl2h precisely." mined using thc prcvious equationsand obscrvationsofthc NIoon U'f +A? difft:red
With its basic unit and initial epoch defined in this way, ephemeris time u'as from ET in tlvo main respccts:
equivalent to the system of time of Newcomb's solar tables. Ncwcomb originally
identified this tinescale with mean solar time. rvhich he considcred to be uniform. (1) by a quadratic exprcssiott in 7 of the forn a + bT + c1a, thc coeflicients of
When mean solar time, as determined from obserlations, was sho\\'n to be nonuni- which had been obscrvatioually determined to be zero, but n'hich alnrost
forn, Newcomb's timescale rvas identified directly with epheneris time. The first certainly difiered from zcro by signiiicant autounts (it should be noted that
two terms of the Sun's geometric mean lorrgitude rvere defined as absolute con- the term c72 is more of a fundarnental physical character t'bar' a + bT);
stants, and the corresponding values for the Moon and other planets were subject (2) by any deficiencies that miglrt have been present irt Brown's lunar theory,
to possible revision to bring them into accord with observation. The svmbol 7 in including revision of any constants involved; in particular, Brown used l/294
these expressions was therefore intended to represent ephemeris time, not only in for the flattening of the Earth.
 EXPLANATOBYSUPPLEMENT 2 / Tll\,,IE 83
82

In the period from 1960 to 1984, improvements were introduced in the lunar Moon was based on accepted values of various constants. Changes in the constants
theory, and the different versions of the lunar theory rl''ere designated by i = 0, resulted in different measures of ET, as described in Section 2.551
j = 1 and j = 2.'lhe ephemeris times determined from these lunar theories were
correspondingly labeled ET0, ETl and ET2.
2.553 AZ Formulas from Historical Observations The variations in the rotation
Thus the value of A 7 could difler systernatically depending upon rvhich method,
of the Earth can be determined from observations, though early observations are
or body, was used for its determination. This will continue to be true when AZ is
much less accurate than modern observations. The rate of rotation does not decrease
extrapolated back in timc prior to the availability of atomic time.
uniformly, as it rvould if tidal friction were the only mechanism afiecting the Earth's
rotation.
2.552 Difficulties in th€ D€finition ofEphemeris Time In hindsight, the definition
The formulas for AT must be divided into time periods. The equation for the
of ephemeris time (ET) was subject to a number of dificulties, both in concept and
time period from 1650 to the present is given by McCarthy and Babcock (1986):
in implementation. It was based on Newcomb's theory of the Sun and an associated
set of astronomical constants. Both the theory and the constants were replaced in
1984. At a deeper level, however, ET sufiered from its prerelativistic origins. Since A(TDT-UT1) = 5:1s6(10.404)+ 13.3O66e O.3264)(t- 0. 19(t 0.01)'?, (2.s53-l)
its definition ignored general relativity, ET cannot be categorized as a geocentric
or barycentric timescale, nor as a proper time or coordinate time. where t is given in centuries since 1800.0. Thc equivalent expression for the length-
Both the fundamental unit and thc epoch of ET contained technical problens. ening of the day is:
The tropical year was chosen as the unit of time under the assumption that it was
independent of astronomical constants, particularly the constant of precession. In L O D = 7 i 2 8 6 ( 10 . 1 7 0 ) x 1 0 - 6 ( y e a r 1 8 1 9 . 2 5 ( 11 0 4 ) . (2.553-2)
reality, determinations of both the tropical ycar and the sidereal year depend on
the adopted system of astronomical constants. Unfortunately, the possibility of a
Tabular values of AZ are given by McCarthy and Babcock (1986) for 1957 1984
time-dependent variation in the equinox, commonly called equinox motion, was not
and by Stepenson and Nlorrison (198'1) for 1630 1980. Current r''aluesare given in
recognized. As a result, ephemeris time \a'assubject to corrections introduccd in
lhe The Astronomical Almanac.
the system of constants in 1984.
The following expressionsfor historic periods are from Stepirensonand \'Iorrison
The fundamental epoch. dcsignated 1900 January 0, 12r' ET, *'as the instarrt
(1984). A parabolic represcntationof the data lrorn -q..1.948 to A.D. 1600 is given
at which the geometric mean lorgitude of the Sun, relerrcd to the mean equirr"x
by
of date, was 279" 41'48'.'04.Although this instant was definitive, its deterlninatiorr
depended on a comparison ol observations \a''ithan apparent ephemeris of the Surr. LT = 2515t2, (2.553-3)
The observatiorrsu,erethemseh'es definitive, but the appareni ephemeris as deducecl
from the geometric mean longitude depended on the value adoptecl for the constant where / is time in centuries from L.l. 1800. This result is equiralent to a rate ot
of aberration. All relevant observations and detcrminations had bccn made using lengthening of the day of 1.,1ms/century.
20'.'47fot the constant of aberration. A change in this vahre would lead to a changc For the period from 390 B.c. to A.D. 9'18the parabolic representationis
in the determination of the itrstaIrt of the fundamental epoch and, thus, to a corlc
sponding change in thc mcasurcs of ephemeris time assigned to all other instants.
AT = 1850"- 435r+ 44.3r2, r ) 55l-r1)
This particular difficulty could have bccn avoided by specifying the epoch as the
instant when the geometric mean longitude of the Srrn, reduced by the constalt of
where r is measured in cerrtrrriesfron ,q.1. 9'18.\\riih the origin at A D 1800, the
aberration and refcrred to the mean equinox of date, stas219'41'27'!57;bui there
corresponding expression is
were objections to the implied use of an "apparent mean longitude."
Ephcmeris timc was defined by thc lrtotion of a body that is difficult to observe
LT = 1360'+32Ot+44.3P. (2.553-5)
Determination of ET was better accomplished by observatiorrs of rapidly moving
bodies the \{oon, N4ercury,and Venus. However, there was no specified relation
ship between ET as defined and ET as determiled lrom observations of the Moon, The equivalent rate of lengthening of the day is 2.4 ms/century (Stephenson and
Mercurv. or Venus. In oractice. the determination of ET from the motion of the Morrison, 1984).
 84 EXPLANATOBYSUPPLEMENT
2 | ltME 85
The formulas above are based on the value of the lunar tidal acceleration n.
TAI took place irl 1977.In 1976.three laboratoriesnraintaining cesiumstandards
where n = 26100century-2. If another value of n is used, the ralrresof TDT - UTI the National Bureau of Standards,National R.esear< h Council, and the Phvsicalisc|-
should be corrected by the addition of TechnischeBundesanstalt agreedthat the frequcncyof TAI rvastoo high by l0-ll.
That same year, IAU recomrnended that the TAI frequency bc corrected by exactly
A?'(TDT-UTf)= -r".8zr(An
t2)(T- t9.ss), /) 5511\ -10 x 10 13on 1977.Ianuary 1. This adjustment was the first direct input of thc
laboratory standards ou TAI.
where T is measured in Julian centuries from 1900.0 (McCarthy and Babcock, In April 1977, the Consultative Committee for thc Definition of the Se<:oucl
1e86). recommended thtrt frequency steering by frequent small adjustments (of the saurc
order as the variations expected from random noisc) u,as better than noticeable
2.56 Historyof AtomicTime corrections at less fre<1uentinterrals. This practice rvas applied immediatch. to
TAI.
Until 1960,the unit of time, the second,was definedas a specifiedfraction (1/86400) In 1985 the IAU adopted a resolution agreeing to transfer the responsibilitl. for
of the mean solar da1..With the adoption of ephemeris time, thc second *.as definecl TAI from the BIH to the Bureau International des Poids et Nleasures. undcr th<r
to be a specifiedfraction of the year (see Section 2.55). This definition was more
responsibility of the International Committee on \{eights and N{easures (CIP\{)
precise than the definition based on the mean solar day, but it was difficult to
and the General Conference of Weishts and Measures,
measure from observational data and to implement with operational clocks.
The idea of using atomic resonances in an atomic clock is due to Rabi, who
suggested it in an American Phvsical Society lecture in January 1945. The first,
atomic clock in actual metrological use was built at the National Physical Labo
2.57 Historyof Coordinated
Universal
Time
ratory (NPL) by Esscn and Parry (1957). Details of the early developmentsarr:
describedby Ramsey (1972), Beehler (1967), and Forcuran (1985).
Beginning in 1962, an increasingnumber ol broad<.rsttime serlices cooperatql to
Since 1967,the SystEmeInternational (SI) second (seeSection 2.21) has beerr
provide a consistelrt tinre standard. \Iost broadcast signals $€re s1-nchronizedto
the standard unit of tiurc in all timescales. The length of the SI second $'as define(l
by Markowitz et aL. (1958) in tcrms of the observationally detcrnrined value of the the redefined UTC in 1972. Prior to 1972, broacl<rasttirne signals \r'ere kept $'ithil
0.1 secondof UT2 bv introducing step adjustments. norrnally of 0.1 second,arrcl
second of ephemeris timc. Thus, the SI second closely matches an cphcmeris second
occasionally by changirrg the duration of the second. Thc scale UT2 resulted lrout
and provides continuity with ephemeris time.
applying the following frrrrnula for the seasonal variation in the rate of the Earth's
Beginning in July 1955, atomic timescales were availablc from organizations in
rotation to UTl:
several countries. Thcse werc based on the cesium-beam frcquency standards at the
individual organizations. In 1971, the experimental atomic timescale established
by the Bureau International de I'Heure (BIH) r'as adopted arsthe worlds'ide time UT2= UTI + 01.O22
sin?nr 01012cos2rrr 01006sin4zrr + 0i0}i cos4r. . (2.5'l 1\
referencesystem, under the name International Atomic Time (TAI). As this decision
implied no discontinuity of the BIH timescale, it is convenient to designate it as where r represents the fraction of the Besselian vcar of thc obserlation.
TAI since its beginning in July 1955. The origin of TAI rva^sarbitrarily. chosen so The seasonalvariation in the rotation of the Earth s'a.sfirst detected bl mearts
that the TAI and UTI readingsat 1958 January 1,0h were the same. of excellent crystai clocks at the Physicalisch-Technische Bundesanstalt (Schcibe
Since 1971, TAI computation has undergone two major corrceptual changes. and Adelsberger, 1936). These variations \4'erecoDfirrned by Stoyko (1936, 1937).
In 1973, as a consequence of the 1972 meeting of the Consrrltative Committee Finch (1950), ancl Srnith and Tucker (1953). Lunar tidal variations were foun.l to
for the Definition of the Second (CCDS), a new TAI algorithm was irnplemented, be in reasonable agrcernent with the theory. More recent reports are available by
whereby each clock participates with a weight that is a firnction of its past and O'Hara (1975), Guinot (1970), and Markowitz (1976). Previously,the researchorr
present frequency. The mean frequency of each clock over a two-month interval long-term variations in the rotation of the Earth was correlated with the motion
rs computed with respect to TAI. The weight of a clock is proportional to the of the Moon, which was the only clock availablc rvith sufficient precision (Spencer
reciprocal of the variance of six mean frequencies. This takes into account the Jones, 1956; Munk and McDonald. 1960). Now UTC is based on UT1 as specified
changes of frequency or the short-term instability of the clock. The other change in in Section 2.24.

-$,
-,7
 Table 2.58.1,continued
86 EXPLANATORYSUPPLEMENT 'fime
and FrequencySteps by WWV and IVC (USNO)
2.58 Historyof Transmitted
TimeSignals
Step'
Prior to 1956, the transmittcd time signals were steppcd or steered to maintain '1958 J a n .1 5 36218 1900 20 WWV controlledat an oflset ol abo./i -100 x
a constant relationship to the lotation of the Earth. From 1956 to 1971, various Feb. 5 36239 1900 20 1 0 1 0d u n g 1 9 5 8 .
approaches were used by different organizations to maiutain the transmitted signals Feb. 19 36253 1900 20
close to the value of time detcrmined from the Earth's rotation. After 1971. all the Apr 9 36302 1900 20
J u n e1 1 36365 1900 20
transmissions ivere tied to UTC as deterrnined by the BIH. For the period 1956 July 2 36386 1900 20
1971 it is possible to obtain a correction to the transmitted time signal from thc July 16 36400 1900 20
O.r.22 36498 1900 20
available documerrtation of the history of the transmission during that time period. Nov.26 36533 1900 20
The Bureau International de I'Heure, located at the Paris Observatory, rvas Oec.24 36561 1900 20
founded to coordinate the practices followed by the national tirne services in ob- 1959 Jan.28 36596 1900 20 WwV controlledal an oltsel ot 100 x 10-10
Feb.25 36624 1900 20 duing 1959.
servations and calculations for the determination of time, and to establish precise Aug. 5 36663 1900 20
international standards; it carne rrnder the auspices of the International Astronom- Aug.26 36806 1900 20
ical Union in 1920. Sept. 30 36841 1900 20
Nov.4 36876 1900 20
In 1988, the BlH activities on astronomical time were taken over by a ncw Nov-18 36890 1900 20
service with a wider scope the International Earth Rotation Service (IERS), the D e c .1 6 36918 1900 2A LJSNO beoan conlrollino NBA at an oflset ol
- 1 7 0 x 1 0 1 od u r i n g1 9 ; 0 .
Central Bureau of which is locatetl at the Paris Obscrvatory. The responsibility of
No time sleps UTC began 1960 January 1.r The inilial partici
atomic tinescales has been transferred from the BIH to the Bureau Internatiorral patingobservaloriesand laboratorieswere USNO,
des Poids et Nlesures (BIPN4) in Sbvres, France. RGO, NBS, NRL and NPL. The o ginal offsetwas
1 5 0 x 1 0 1 0 .N o i n t e r n a t i o n U
a lT c o f f s e l w a s r n
Table 2.58.1 provides the documentatiol of the steps and frequency changes effectbetore 1960.
utilized by radio station WWV. In countlies r'"'herethe civil time system $'as not 1961 J a n .1 37300 0000 U T C o f i s e td u r i n g1 9 6 1w a s 1 5 0 x l O
10-
the same as WWV, a different table of corrcctions appropriate to the civil tirrrc 1962 A u g .1 37512 0000 50 advance
10.
system actuaily used should be introdrrced, if available. No time steps U T c o l l s e ld u r i n g1 9 6 2w a s 1 3 0 x 1 0
10.
1963 Nov. 1 38334 0000 100 U T Co t f s e d t u r i n g1 9 6 3w a s - 1 3 0 x 1 o
Table 2.58.1 '1964 Apr.1 38486 0000 100 UTC offsetduring 1964 was 150 x 1o
10.
Time and FrequencySteps by WWV and tVlC(USNO) Sepl.1 38639 0000 100
Oct.i 38669 0000 1 NrC(UsNo)advanced 1-6milliseconds10 Oclo6e(,
Step' 1964.WWV relarded 1.0 milliseconds10 Oclober,
1964.
10.
1956 Jan.4 35476 1900 60 advance 1965 J a n .1 0000 i00 U T Co f f s e ld u r i n g1 9 6 5w a s 1 5 0 ! 1 0
Mat.7 35539 1900 20 advance [,,lar 1 38820 0000 100
lvlar.28 35560 1900 20 J u l y1 34942 0000 100
July25 35679 i900 20 Sept. 1 39004 0000 100
Aug.22 35707 1900 20 1966 No lime sleps UTC ollsel during 1966 was -3OO x 10
10
sept.19 35735 1900 20
Oct.31 35777 1900 20 1967 Sept. 20 39753 0000 200 4 advance WwV only; MC(USNO) was not advanced 1967
Nov-14 35791 1900 20 September20.UTC offseiduring1967was -300 x
10'".
1957 J a n .2 3 35861 1900 20 On 1957 lvlarch13, the time step made at 1900 UT 10.
Mar 13 1968 Feb. 1 39887 0000 100 advance U T c o l l s e ld u r i n g1 9 6 8w a s 3 0 0 x 1 0
35910 2000 20 was obtainedfrom log books, bul the step made al
May 1 35959 1900 20 2000 UTwas fronr TS Bulletins. 1969
June 5 35994 2000 20 1S70 No time sieps UTC offsel dlring 1969, 1970, and 197'l was
June 19 36008 1900 20 -300 x 1O-'".
July 3 36022 1S00 20 r971
Juty 17 36036 1900 2Q
Aug. 14 36064 1900 20 *
All steps are retardationsin milliseconclsunlessolhefwise noled.
Ocl.16 36127 1900 20 I See Trans. tAU Beports,xta, 362-364.
Nov. 6 36148 1900 20
Nole: P or to 1956 January i, WwV did not make lime steps. Inst€adWwV sleered ils lrequencyto lollow
D e c .1 1 36183 1900 20 closelylhe Eanh's rolalion.Sieps were lhereJoreunnecessary fihe syslem followsd by WWV was the N2
system,used by ihe U.S. Naval Observaloryfrom 1953 April I until 1955 December31).
88 EXPLANATORYSUPPLEMENT 2 / ltME 89

2.6 REFERENCES BIH Circular D. (Bureau international de I'Heure, Paris, France).

Adams, J.C. (1853). "On the Secular Variation ofthe Moon's \Iean Motion" Pi.il. ?rons. Blair, B.E., cd. (197a). "Time and Frequency" (Red Book) Nat Bur. Stand (US) Nlonogr.
cxLIII. 397406. No. 140.
Allan, D.W. (1985). "Characterization Optimum Estimation and Time Prediction of Bodily, L.N. (1966). "Correlating Time frorn Europe to Asia and with Flying Clocks',
Clocks" Proceed,ingsoJ the Seuenteenth Annual Precise Time and Time Interua,l (PTTI) Hewlett- Pachanl, Journal 16, No. 8.
Applications anrl Planning Meeting, pp.45 66. Box, G.E.P., and Jenkins, G.\i. (1970). Time Series Analysis Forecastirtqaru) Control
Allan, D.W. (1983). "Clock Characterization Trto al" Proceedings oJ Fifteenth Annual (Holdcn-Day, San Francisco).
Precise Time and, Ti,me InteruaL (PTTI) Meeting, Washingtoz, pp.,159 476. Brown, E.W. (,1979). TabLesof the Motion oJ the Moon (Yale University Press, New Haven.
Allan, D.W., Blair, B.E., Davis, D.D., and Machlao, H.E. (April 1972). "Precision and cr).
Accuracy of Remote Synchronization via Network Television Broadcasts, LORAN C. Brunet, Ir{., Freon, G., and Parcelier, P. (1977). "Comparison of Dista t Atornic Clocks
and Portable Clocks" Metrologia E, 64 72. with the SYNIPHONIE Satellitc" Proc. Sgmposium on Telecontmunir:ation ltleasttre-
Anderle, R.J. (1976). "Polar Motion Determined by Doppler Satellite Observations" Re- ment, otgatized by URSI (Lannion, France, 3 7 Oclober 1977).
pod to Commission 19. lGeodes Bull. Grenoble). Buisson, J., N{cCaskill, T., Nlorgan, P., and Woodger, J. (1976). "Precise \lbrldrvide Sta-
Andrewes, W. and Atwood, S. (1983). The Time Museum: An Introduction (The Tirne tion Synchronization via the NAVSTAR GPS, Navigation Tedrnologv Satellite (NTS-
Museum. Rockford. IL ). 1)" Proceedings of the Eighth Annual Precise Time and Time Inter\al Meetitg.
Anderson, O.D. (1975). Time Series Analysis arul Forecasting; The Box-Jenkins Approach Capitaine, N- and Guinot, B. (1985). "Anomalies of Somc Tidal Waves in UT1" Geopi.gs
(Butterworth. London). J.R. Astr. Soc. El, 563.; C.C.D.S. (1970). Comite Consultatif pour la Definitio[ dc la
Seconde,5e Session(Bureau Interrrational des Poids et Nfestres,Sevres);CClR Dour-
Aoki, S., Guinot, B., Kaplan, G.H., Kinoshita, H., NlcCarthy, D.D., and Seidelmann,P.K.
(1982). "The New Definition of Universal Tirne" .4stron. AstrophAs.105, 359 361. ments. Study Gloup 7 (Standard Frequencies aDd Time Signals), Geneva, Sr,r'itzerland.
Chi, A.R. ard Byron. E. (1975). "A Two way Time Transfer Experinlcnt Usirrg a Syn-
Ashby, N. (1975). .An Earth-Based Coordinate Clock Network" NBS Tech. Note 659 (US 'l'une
chronous Satellite" Proceed,tngsof the Seuenth Anrnal Prec:ise and Time Interual
Govt. Printing Office, \\rashington, DC).
(PTTI) Applrcations and Planning Meeting, pp.357 37(j.
Ashby, N. and AIIan, D. (1979). "Practical Applications of Relativity for a Global Coor
Clark, T.A. (1972). ''Precision Timing and Very Long Baseline Intelferometry (VLBI)"
dinate Timescale" Rad,io Sc'i. 14. 649 669.
Proc. Fourth Annu,o,l NASA and DepartrnerLl oJ DeJense Precise Tinc (rnd Tirnc ltr
Audoin, C. (198,1). "Prospects for Atouric Frequency Standards" Proceedings of the Stx
terual Plannirry MeetirrgGSFC Rcport X 814 73 72. pp. 74 89 (Coddard SpaceFlight
teenth Annual Precise Tirne antl Tirne Interual (PTTI) Meeti,ng, pp. 1 18.
Center, Greenbelt, Ir'lD).
Bagley, A.S. and Cutler, L.S. (196,1)."A New Performance of the Flyirrg Clock Experi
Clemence,C.X,L (1948). "OD the Systerrrof Astronornical Constants" Astron.. .1.53. 169
ment" HewLett Packard, Journal15. 11.
779.
Barnes, J.A. (1976). "Models for the Interpretation of Frequency Stability N{easuremerrts"
Cooper, R.S. and Chi, A.R. (1979). "A Review of Satellite'lra:rsfer Tcchnolog-"-: Accc,rll-
NBS TechnicalNote 683 (U.S. Govt. Prirting Office, Washington, DC).
plishments and Future Applications" Radto Sci. 14, 4, 605 019.
Barnes, J.A., and AIIan, D.W. (1964). "Effects of Long-Term Stability on the Definitiolr
Costain, C.C., Daarrrs,H.. Boulanger. J-S., Hanson, D.\\r., and Klepr:zynski,\V.J. (1978).
and Measurement of Short-Term Stability" Proc. IEEE-NASA Symposiurn Short-
"Two-t'ay Time Transfers betlveen NRC/NBS and NRC/USNO via the Herncs (C'l'S)
Term FrequencyStability NASA Spec. Publ No. 80, 119 123.
Satellite" Proceedlngs ol the Tenth Awrual Precise Ttme and Tirne Inten:al Meeti.ng,
Becker, G. (1974). Time Scales Relativity: Report of Physikalisch Technische Burrdc- Washinqton NASA 'Ier:hnir:al Nlenorandrn 80250, pp. 585 600-
sanstalt, and acts of the secorrd Cagliari Meeting on Time Deternrination. Dissemi- Costain, C.C., Boulanger,.LS., Daarns, H., Hanson, D.W.. Beehler, R.E.. Clemerts. A..1.,
nation and Synchronizatiol, 197,1.Editor: Astron. Inst. of Cagliari, ltaly.
Davis, D.D., Klcpczynoki. W.J., Veenstra. L.. Kaiser, J.. Guinot. B., Azoubib, P,
Becker, G., Fischer, B., Kramer, G., and Nlullcr, E.K. (196?). "Die Definition der Sekunde Parcelier, P., Freon, G., and Brtnct, M. (1979). "Two n'a1'Time tansfer via Geosta-
und die allgemeine Relativitatstheorie" (The Definition of the Second and thc Theor\' tionary SatellitesNRC/NBS, NRC/USNO arrd NBS/USNO via Herrres anclNRC/LPF'l'
of General Relativity). PTB Mitt.77,111. (France) via Symphoiie" Proceed,ingsoJ the Eleuenth Arntual Precise I\ime antl Tirne
Beehler, R.E. (1967). "A Historical Review of Atomic Frequency Standards" Proc. IEEE InteruaL (PI'TI) ConferenceNASA Conferencc Publication 2129, pp.499 519.
55(6), 792 805. See also Blair (197,1),85 109. Counselman, C.C. (1976). "Raldio Astrometry" Annu. Reu. Astrorr. Astrophys. 14. 197
Beehler, R.E., Mockler, R.C., and Richardson, J.M. (1965). "Cesium RearrrAtomic l-re 211.
quency Standardsr!Metrologia lJ 114 131. Danjon, A. (1929). "Le temps, sa ddfinition pra"ctiquc,sa ntesure" L'Astronornte xliii,
Benavente, J., Besson, J., and Parcelier, P. (1979). "Clock Comparison by Laser in the 13 22.
NanosecondRarrge" Eadio Sci. L4, 1,70I 706. Danjon, A. (,1959). Astronomie Generole (Sennac, Paris).
Besson, J. (1970). "Comparison of National Time Standards by Sirnple Olefiight" IEEE Davis, D.)., Vv'eiss,
Nrl.,Clements,A., and Allan, D.W. (1981). "UnprecedentedSyntoniza-
Trans. Instr. and, Meas.,IM-19, 4, 227 232. tion and Synchronization Accuracy via Simultaneous Vieu'ing with GPS Receiversl
BIH (1973). Annual Report ol the B redTrInternationaL de I'Heure. Constnrction Characteristics ofan NBS/GPS Receiver" Proceedings oJ tlte Thirteenth
 EXPLANATORY
SUPPLEIVENT 2 r ltME 91
90

Annual Precise fime and Time Intenal (PTTI) Applications anil Planning Meettng Hurd, W.J. (1972). "An Analysis and Demonstration of Clock Slnchronization by VLBI"
NASA Conference Publication 2220, pp.527 541. Proceed,ingsoJ the Fourth Annual NASA and Department oJ Defense Precise Tzme
and, Time Interua,l (PTTI) Planning Meeting, pp. 100 122.
Delaunay, C.E. (1865). "Sur I'existence d'une cause nouvelle ayant une action sensible sur
la valeur de I'equation sdculaire de Ia Lune" Compte Rend,u des S4ances de I'Acad4mie IAU Fifteenth General Assembly (1973). "Report of Commission 31 Proceedings 1974
d,esSciences61. 1023 1032. Tmns. IAU XYB, 149.
IAU Sixteenth General Asscmbly (L976). Proceedings 1977 Tians l{a/ XVIB, 58.
de Sitter, W. (1927). "On the SecularAccelerationsaod the Fluctuations ofthe Longitudes
of the Moon, the Sun, X{ercury and Venus" Bull. oJ the Astron. Institues oJ the Ives, H.E. and Stillwell, G.R. (1938). "An Experimental Study of the Rate of a Moving
Atomic Clock" OpL Soc. Am.28,215.
Netherland,s iv,21 38.
Jenkins, G.M. and Watts, D.G. (1968). SpectralAnalEsisand its Applications (Holden-Day,
Essen, L. and Parry, J.V.L- (1957). "The Cesium Resonator as a Standard of Frequency
San Francisco).
and Time" Phil. Trans. R. Soc. London SeriesA 250,45-69.
Jespersen,J.L., Kamas, G., Gatterer, L.E., and McDoran, PF. (1968). "Satellite VHF
Farley et al. (196E). "Is the Special Theory Right or Wrong?" Nature 217, 17.
Transponder Time Synchronizatioi" Proc. IEEE, 56, 7, 1202 1206
Ferrell, W. (1864). "Note on the Influence of the Tides in Car.rsirrgan Apparent Secuiar
Johnston. K.J. et dl. (1983). "Precise Time Transfcr Using MK III VLBI Technology" Proc.
Acceleration of the Moon's N{ean Motion" Proc. oJ the American Acad.emE oJ Arts
oJ Fifieenth Annual Precise Tzrne aniL Time Interual (PTTI) Meeting' Washington,
and Sciences Yl, 379 383.
443,155.
Forman, P. (1985). "Atomichson: the Atomic Clock from Concept to Commercial Product"
Johnston, K.J. (1974). "Radio Astrometry" Proceedings of Sixth Annual Precise Time ancl
Proc. IEEE 73, 71a1.
Time InteraaL (PTTI) Conference, Washtngton, p.373.
Frisch and Smith (1963). "Measurement of Relativistic Time Dilation Using N{eason"
King-Hele, D.G. (1980). "The Gravity Field of the Eafth" Ph'il. TrcLns, A 294,3I7 328.
Amer. J. Phys. 31, 312.
Klemperer, W.K. (1972). "Long-baseline Radio Interferometr]' with Independent Fre-
Gerber, E.A. and Sykes,R.A. (1966). "State of Art: Quartz Crystal Units and Oscillators" quency Standards" Proc. IEEE 60.602 6Dg
Proc. IEEE 54(2), 103-116. See also in Blair (1974a), *'here the authors give a brief Klepczynski. W.J. (1984). "Time Transfer Techniques: Historicai Overview, Cuuent Prac-
update of the more recent developments in a Part B. tices and Future Capabilities" Proceed"ingsoJ the Sixteenth Annual Precise Time antj
Granveaud, D.N{. and Guinot, B. (1972). IEEE Trans. Insb. dnd Meas. IM-21, ,1,396. Time Interoal (PTTI) Meeting, Washington, pp. 385 402.
Granveaud, D.X'L (1979). "Evolution of the Internatioiral Atomic Time TAI Corrputa- Kleppner, D. et al. (1965). "Hydrogen Maser Principles and Techniques" Phys Rett. l3E'
tion" Proceedings oJ the Eleuenth Annual Precise Time and Time Interual (PTTI) 4A. A972 A983.
AppLications aru1,Planning Meeting NASA Conf. Pub. 2129, pp. 185 196. Landes, D. (1983). Remtution in ?irne (Harvard University Press, Cambridge. NIA).
Guinot, B. (1969). "Formation de I'echelle de temps coordonnec par le Bueau interna- Loeb. H.W. (1972). "Efficient Data Transfornatio.I in the Time Domain Spectrometry"
tional de l'heure" Actes d.u Colloque international d.e chronometrie A-20, Serie A, IEEE Trans. Instr. Meas.21(2), 166 168.
Paris. Markowitz, W. (1977). "Timckeeping and Its Applications" Adrances in ElcctT'onic and
Guinot, B. (1974). "Review of Time Scales" Proceed,ingsof the Sixth Annual Precise Tim.e Electron Pltysics 44, 33 87.
qnd Time InteruaL (PTTI) PLanning Meeting GSFC X-814-75-117 (Goddard Spacc Markowitz, W., Lidback, C.A., U1'eda,H., and Muramatsu, K. (1966). "Clock Synchro-
Flight Center, Greenbelt, N{D) nization via R€lay II Satellite" IEEE Trans. Inst. and Meas IM-15. 1, 177 189.
Hafele, J. and Keating. R. (1972). "Around-the-llbrld Atomic Clocks: Predicted Rela- Markowitz, W.. Hall, R.G., Essen, L., and Perry. J.V.L. (f958). "Frequeucy of Cesium ir
tivistic Tirne Gains" Science 177.166 17O. Terms of Ephemeris Tirne" Pi.gs. Reu. Lett. l, 105.
Halley, E. (1695). "Some Account of the Ancient State of the City of Palmyra with Shoit Martin, W. (1966). "Digital Communication and tacking: Time Svnchronization Exper-
Remarks upon the Inscdptions Found There" Phil. Trans. 19, 160 175. irnent" JPL-Space Programs Sumtnary Ill, 61. 37 12.
Hanson, D.W. and Harniltonj W.F. (1974). "satellite Broadcasting of W\\iV Signals" Meeks, M.L., ed. (1976). "Astrophysics, Part C: R.adio Observations" Methods oJ Etperi'
IEEE Ttans. on Ae-rospaceand, Electronica Sgslem AES-10, 5, 562 573. merLtal Phasics 12 (Acadcmic Press, New York).
Hellings, R.W. (1986). "Relativistic EIIects iD Astronomical Timing Nleasurements" ,4s- Misner, C., Thorne, K., and Wheeler, J. (1973). "Gravitation" (Freeman, San Francisco)
tron. J. 91. 650. Moritz, H. and Mueller, LI. (1987). Earth Rotation (Frederick Ungar, New York).
Higa, W.H. (1972). "Time Synchronization via Lunar Radar" Proc. IEEE 60,5,552 557 Moyer, T.D. (1981). "Transformation from Proper Time on Earth to Coordinate Time in
Hinteregger, H.F., Shapiro, I.L, Robertson, D.S., Knight, C.A., Ergas, R.A., Whitncl., Solar System Barycentric Spacc-Time Frame of Reference" Cele'stial Mechanics 23' 1,
A.R., R.ogers,A.E.E., X'Toran,J.N{., Clark, T.A.. ald Burke, B.F. (1972). "Precision 33 56 and 57 68.
Geodesy via Radio Interferometry" Science 17E, 396 398. Mueller, I.I. (1969). Sphericat and, Practical Astronomy (Frederick Ungar, New York).
Hudson, G.E. (1964). "spacetime Coordinate Systems" ,4cre.sd,u Congres international de Murray, J.A., Pritt, D.L., Blocker, L.W.. Leavitt, W.E., Hooten, P M , and Goring' W D'
(1971). "Tnne Transfer by Defense Communications Satellite" Proc oJ 25th Attnutl
chronometrie (Lausanne, Switzerland).
Frequencp Control Symposium, Atlantic City, New Jerseg, pp 186 193.
Hudson, G.E., Allan. D.W., Barrres, J.A., Hall, R.G., Lavanceau, J.D., and Winkler,
Nautical Almanac Office (1954). Improred Lunar Ephemeris 1952 1959 (U.S Governrnent
G.M.R- (1969). "A Coordinate Frequency and Time System" Proc. 23rtl Annual Fre-
quencA Control Sgmposizm USAE Com. Printing Office, Washington, DC).
92 EXPLANATORY
SUPPLEMENT 2 / rtME 93

Newhall, XX, Williams. J.G., and Dickey, J.O. (1988). "Earth Rotation from Lunar Laser Ringer, D.E., and Gandy, J., and Jechart, E. (1975). "Spacebornc Rubidium Frequerrcl
RarLgilg" BIH Annual Report Jor 1987, D 17 . Standard for Navstar GPS" Procee,TingsoJ the Setenth Annual Precise Time anrl.
Oaks, O.J. (1985). "Time Transfer by Satellite" Proc. of Setenteenth Annual Precise Time Time InteruaL (PTTI) ConJerence.pp.671 696.
and, Time Interual (PTTI) Applicati.ons and Planning Meeting, pp.77 90. Rogers, A.E.E. (1970). "Very Long Baseline lnterferometry with Large Effective Banrl
Neugebauer, O. (1975). A Historg of Ancient Mathernatical Astroaomy (Springer-Verlag, width for Phase Delay Nleasurements" Radio Sci.5, 1,239 124a.
New York). Rueger, L.J. and Bates A. (1978). "Nova Satellite Experiment" Radt'oSci. 14,707 714.
Newcomb, S. (1878). "Researches on the Motion of the \Ioon" Washington Obseruations Rutman, J. (1978). "Application ofSpace Techniquesto Time and FrequencyDissemintion
for 1875. and Synchro izatiorr" Proc. European Workshop on Space Oceanographa, Naxigation,
Newcomb, S. (1895). "Tables of the Motion of the Earth on its Axis and Around the Sun" and. GeoiLlnamics (SONG), SchLoss-ELrnau,Gerrnany ESAPS 137. Paris: European
Astronomical Papers of the American Ephemeris antl. Nautical ALmanac lY, L Space Agency SP-127, pp. 311 316.
Saburi, Y. (1976). "Observed Time Discontinuity of Clock Synchronization in Rotating
Newcomb, S. (1912). "Researches on the Motion of the Moon, Pa-rt ll" Astronornical
Frame of the Eaftln" J. RarLi,oRes. Labs.23, 11'2,255 265.
Papers of the American Ephemeris and Nautical Almanac lX, l.
Saburi, Y., Yamamoto, M., and Harada, K. (1976). "High-precision Time Comparison
Percival, D.B. (1973). Statistical properties of high performance cesium standards" Proc.
via Satellite and Observed Discrepancy of Synchronization" IEEE Trans. Instr. and
on the Fifth AnnuaL Precise Time and Time Interxal (PTTI) ConJ., pp. 239 -263.
Meas. IM-25, 1, 473.
Peters, H.E. and Washburn, P.J. (1981). "Atomic Hydrogen X4aserActive Oscillator Cav- Silverberg, E.C. (1974). "Operation and performarxie of a Lunar Laser Ranging Station"
ity and Bulb Design Optimization" Proc. oJ 16th Ann.al Preci.se Time and Tim.e Appl. Opt. 13, 565 574.
Intenal (PTTI) Meeting, Washington,pp.313 338. Spencer Jones, H. (1939). "The Rotation of the Earth and the Secular Acceleral,ionsof
Potts, C.E. and Wieder, B. 11972). "Precise Time and Frequency Dissetrination via the the Sun, l,Ioon and Planets" Month. Not. -R. '{S 99, 541.
LORAN C System" Proc. IEEE 60, 530 539. Steele, J. McA., N{arkowitz, W., and Lidback, C.A. (1964). "Tclstar Timc Synchroniza-
Pound, R. and Rebka, G. (1960). "Apparent \4reight ol Photons" Plrys. Rex. Lett. 4, 337. tlon" IEEE Trans. Instr. and. Meas.IM-13.4, 164 170.
Pound, R. and Snider, J. (1964). "Efiect of Gravity on Nuclear Resonanr:es" Phys. Re,t. Stoyko, N. (1937). .Sur la periodicit6 dans I'irregularitd de la rotation de la Terre"
Lett. 13, 539. Comptes-rend,us250, 79.
Comit6 International des Poids et Mesures (1957). Proces Verbaux tles Sdonces,deuxiimc Stoyko, M.N.(1937). "Sur la periodicit6 dans I'irregularit6 de la rotatior de la terra" C.fi.
s6rie 25. 75 79. Acad,. Sci 205. 79.
Ramsey, N.F. (1968). "The Atomic Hydrogen N{aser" ,4m. Sci.56(4\,42O 138. Taylor, R.J. (May 1974). "Satellite to Ground Timing Experiments" Proc. 28th Ann. Sgnt
Ramsey, N.F. (1972). "History of Atomic and Nlolecular Standards of Frequencl' and on Freq encu Cottrol, pp. 38,1 388.
Trans. IAU (1960).10, 72, Most!u,, 1958 (English translatidr, p. 500).
Tine" IEEE Trans. Instrum. and Meas.21(2),90 99.
Turner, A.J. (198,1). "Time Xleasuring lnstr:uments: Water-clocks, Sand glasses, Fire-
Rapp, R.H. (1974). "Current Estimates of [4ean Earth Ellipsoid Pararneters" Geophys.
clocks" Catalogue of the Collection, I, Part 3 (The Time XIuseum, Rockford).
fi€s. ,elt. 1(1), 35 .18.
Vanier, J. and Audoin, C. (1989). "The Quantum Physics of Atomic Frequency Standards"
Recommendations and Reports of the CCIR 1978. "Standard Frequencies and Time Sig- (Adams and Hilger, Bristoi).
nal* VIl. Feporr 363 4, 78 81. Veillet, C. and Pham Van, J. (1988). "Earth Rotation from Lunar Laser Ranging" 8,111
Reder, F.H. and Wiokler, G.X4.R. (1960). "Preliminary Flight Tests of an Atouric Clock Annual Report for 1987 D-45.
in Preparation of Long-Range Clock Synchronization Experiments" Natur€ 1E6,4725, Vessot, R.F.C. (197,1). "Lectures on Frequency Stability of Clocks in the Cravrtational
592 593. Redshift Experimenl" Expetirnental GraDitation B. Bertotti. ed. (School of Physics
Reder, F.H., and Winkler, c.M.R. (1960). "World Wide Clock Synchronization" ,IEE Enrico Fermi, Academic Press, New York), pp. 111 162.
Tlans. Mil. Electron. 4(213),366 376. Vessot, R.F.C. and Levile, I,'I.W. (1979). "A Test of the Equivalence Principle Using a
Reinhardt, V.S. (1974). "Relativistic Effects ofthe Rotation ofthe Earth on Renote Clock Space-Borne Clock" General Relatititg and, Grauitation 70, 3, 181 204.
Synchronization" Sixth Annual Precise Time and, Time Inten)at (PTTI) Planning Whitrow, G.J. Ogaa). Time in History (Oxford University Press, Oxford).
Meeting, Washington, pp. 395 124. Winkler, G.M.R. (1977). "Timekeeping and Its Applications" Ad,uanLesin Electronics and
Reinhardt, V.S. and Lavanceau, J. (1974). "A Comparison of the Cesiurn and Hydrogel ELectron Physics 44, 33- 97.
Hyperfine Frequencies by Means of Loran-C and Portable Clocks" Proceed.ingsof the Winkler, G.M.R. (198?). "High Performance Clocks" Ftrst Europeon Time and Frequency
Tuentl1-Eighth AnnTraI Symposi,um on FreqrrerlcgContro| Ft. Monmouth, New Jerse.g, Forum, 18 20 March 1987, Besancon, France,pp.7 1,9.
379. Winkler, G.M.R. and Van Flandern, T.C. (19??). "Ephemeris Tirne, Relativity and the
Reinhardt, V.S., Premo, D.A., Fitzmaurice, M.W., Wardrip, S.C., and Arvenka, P.O. Problem of Uniform Time in Astronomy" Astron. J.82, 1, 8,1 92.
(1978). "Nanosecond Time tansfer via Shuttle Laser Ranging Experimcnt" Proceed- Woolard, E.W. and Clemence, G.M. (1966). Spherical Astronomy (Acadcmic Press, New
ings of the Ninth Annual Precise Time and. Time Interual (PTTI) Applications and, York and London).
Planning Meeting, pp. 319 342.
*
\t

CHAPTER
3

CelestialReference
Systems
by C.Y Hohenkerk,
B.D.Yallop,
C.A.Smith,andA.T.Sinclair

REFERENCE
3.1 CELESTIAL SYSTEMS

In order to specify positions of astronomical objects, such as sta.rs,galaxies, quasars,

planets, and satellites, it is necessaryto have a referencesystem. The currently avail-
able reference system is based on the FK5 star catalog, which provides a reference
frame, the IAU 1976 System of Astronomical Constants, the IAU 1980 theory of
nutation, and the procedures that provide the methods for implementing the ref-
erence system. Chapter 4 describes in detail the distinction between a reference
systern and a reference frame.
An ideal celestial system would be an inertial system so that the equations
of motion could be written without any rotational terms. Unfortunately, the re-
alization of an ideal celestial reference system can only be approached, it cannot
be achieved. As this chapter was being prepared, the IAU appointed a working
group to study the many difierent aspects of reference systems and to formulate
improvements for the future. Therefore, although this chapter attempts to provide
the instructions and the methods for the current reference system, it also indicates
some of the difficulties with the present system and possible future changes.
The ideal referencesystem would provide an inertial frame that could be used at
all frequencies of the electro-optical spectrum, and at all magnitudes or intensities.
Unfortunately a number of problems interfere with realizing such a referencesystem.
The first problem is that ao known point sources radiate at all frequencies of the
spectrum such that the positions determined at one frequency can be accurately
related to those determined at another frequency. Thus we have different sets of
reference sources in the optical, radio, and X-ray frequencies.
We have the problem that the dynamicai range for all observing systems is
limited. Generally in the optical range only five magnitudes of intensity can be
observed astrometrically by one technique. Thus the method and the references
96 EXPLANATORY
SUPPLEMENT 3 / CELESTIALREFERENCESYSTEIIiIS 97

used for observing bright stars are difTerent from those for observing faint stars. We basic difference between the concepts remains, which is another reason for using
can define a reference frame based on the very distant extragalactic sources that the terminology quasi-inertial.
should be without proper motion and provide a pseudofixed reference frame, but Although we may desire a single ideal reference system, in practice we have to
it is impossible to measure the positions of brighter stars or solar system objects recognize not only the conceptual diflerences in reference frames used in the avail-
directly relative to those faint distant sources. Thus we have a problem realizing an able systems but also the differences in the realizations of these frames. Although
observable reference frame at all frequencies and magnitudes. the FK5 can be viewed as a kinetic frame based on the stars, its origin is tied to
There is also a problem delining the ideal referenceframe. In classical mechanics an equinox that is deflned by the solar system. But, in practice, the FKS does not
an inertial reference frame is either at rest or in a state of uniform rectilinear have a simple fiducial point. Its real origin is a meridian of zero right ascension.
motion with respect to absolute space. However, the theories of general and special The meridian does not exactly go through the equinox defined by a givcn solar svs-
relativity required some changes to this concept. In the general theory of relativity, tem ephemeris, nor is that meridian necessarill' a plane. It carr difler systematicallv
Einstein defined an inertial frame as a freely falling coordinate system in accordance in a declination dependent way. Therefore. the origin of the FK5 reference frame
with the local gravitational field which is due to all material matter of the universe is defined as a catalog equinox. This is contrasted with the eqrdnox delirred bv a
(Einstein, 1956). The special theory of relativity defines an inertial frame in a solar-system ephemeris and called a dynamical equinox. Each solar-svsterrrephcm-
space-time continuum in the absenceof gravitational fields, rather than in absolute eris defines its own origin or equinox. In the caseof the epherncrides DE200/LE200,
space (Moritz, 1967), and the transformation between inertial frames is given by an effort was made to make the origin agree with the FK5 equinox, but it was rec-
the Lorentz transfbrmation. ognized, and needs to be recognized, that there is some difference, *'hich will be
There are finite regions rvith respect to a chosen space of refercnce where ma determiued more accurately with time.
terial particles move freely without acceleration and in which the laws of special Therefore, although n'e may have a corrccptual ideal reference svstern, cach
relativity hold with remarkable accuracy, which can be regarded as quasi-inertial realization of a reference frame will be sornewhat different. and there r.ill be a need
frames. Thus the center of mass of the Earth-Nfoon system falling in an elliptic to determine the transformatiorls frorn orrc rclerence lrame to another. The FK5
orbit around the Sun in a relatively weak gravitational field is an example of suclt reference frame is verv close to the DE200/LE200 reference frame, u,hich in turn
a finite region. Therefore, we may assume quasi-ilertial frames of reference, arrd is very close to the radio source reference frarncs. The deviations are partly due
any violation of the principles when using classical mechanics can be taken into to the difierent acculacies lvith rvhich the various reference fraures arc dclcrmirred.
account with small corrections applied to the observations and by an appropri Similarly there can be time depenclent difiereuces between thc rcfcrcncc frames.
ate coordinate-time reference. The effects of special relativity for a system movirtg The FK4 reference frarne was forrnd to have a rnotion as a firnction of time. This
with the Earth around the Sun are of the order l0 8: Iikewise those rvith seneral could be vierved as a sysl,ernatic error in the proper motions of the FK,l systeui.
relativity are ofthe order 10 8. Although it is hoped that these clifferencesu'ill be srnall, it rnust be recoguized
It is necessary to distinguish between various types of quasi-inertial frames. when doing precise positional astrolomt- that these difierences do exist-
For example, the galactic reference frame. which is based on assumptions about the For precise reduction of observations it is necessaryto include all the displacc
proper motions of celestial bodies and their statistical properties. is a frame definr:d ments and motions involved in the reference systems. These include the differertces
kinetically. A reference frame can be constructed on extragalactic sources based oir in the origin of the frame, barycentric, geocentric. or topocentric, together lvith the
the assumption that the galaxies have no rotational components to their motions. appropriate timc coordinate as discussed in Chapter 2. Also the time-dcpenderrt
Alternatively, we can construct a quasi-inertial frame based on a dynamical systern. variations in the celestial coordinates must be included; thcsc are precession,nuta-
This reference frame is based on the theory of motions of the bodies of the solar tron, parallax, aberi-ation. and light deflection. The reduclion to topocqrtric coor-
system, and is constructed in such a way that there remains rlo rotational term itl dinates includes thc Earth-orientation parameters (UT1 UTC and polar motion),
the equations of motion. diurnal parallax, diurnal abcrration, and refraction. Thc basis for the exprcssions
In Newtonian mechanics, the various types should be equivalent, but this is not necessarylbr these conversions is the subject of thc rest ofthis chapter.
true in general relativity. In general relativity a dynamical system of coordinates
is a local reference frame that is locally tangent to thc general space-time mani- 3.11 Fundamental
Referencesystems
fold. In contrast the kinetic frame is defincd by assumptions about the apparent
motions of remote objects and is therefore a coordinate system that is subject to Stellar reference frames can be dcterrlined observationally. Bccause of the many
relativistic effects such as geodesic precession. Even if this is suitably corrected, a motions involved, it must be recognized that the observational determirration of the
98 EXPLANATORY
SUPPLEMENT REFERENCE
3 / CELESTTAL SYSTET',I|S gg

reference frame depends upon the dates of the observations. From the rotation of the 3.2 BASISOF REDUCTION
OF CELESTIAL
COORDINATES
Earth, the plane of the equator and hence its pole can be determined. A fiducial
direction in the equatorial plane must be determined either from the definitiorr
A necessary part of any reference system is the procedures required to relate a
of a given object, collectivelS' from a number of diflerent objects, or from solar-
reference frame for one time to one for another time. These procedures involve
system observations defining the ecliptic. The observational determination of the
precession, nutation, space motion, parallax, aberration, and light deflcctiorr. In
fundamental reference frame can then be converted to a conventional standard
this section, these different effects will be discLssed from a theoretical basis along
reference frame by application of certain specified procedures These procedures, in
with an explanation of the determination of these efiects. To complete the discus_
combination with the reference frame, constitute the reference system and will be
sion, formulas for polar motion and refraction, which are required in topoceutric
specified in this chaPter.
reductions, are given at the end of this section.

Reference
3.12 TheDynamical System
3.21 Precesslon
The motions in the solar system provide a number of possible bases for reference
frarnes, the standard being the equator of the Earth and the equinox defined by the Since the plane of the ecliptic and the plane of the cquator are usccl as planes of
reference, and their intersection, the equinox, is used as the fiducial point, it is
intersection of the mean planes of the equator and the ecliptic. Alternatively, the
ecliptic and the equinox can define a referenceftame. Some woulcl argue that the use necessary to inchtde the fact that these planes are in motion. Thc motion of the
of the invariable plane with an equinox provides a suitable reference frame, but this ecliptic is due to the gravitational action of thc planets on the Earth's orbit and
frame is not n'ell-determined. A referertceframe. horvel'er defined, can be determined makes a contribution to precession known as planetary precession. If the erluator
were fixed, this motion would produce a precession of the equinox of about I2" per
from the process of calculating the ephemerides of the solar-system bodies based
on their equations of motion in the chosen frame, fitting these computed positions. century and a decrease in the obliquitl. of thc ccliptic of about 47" pel centurt.
The motion of the equator is due to the torclue of the Sun, Nloon, and planets on
suitably reduced to the frame of observation, to observed positions, and hencc
the dynamical figure of the Earth. It cau be separatcd into t$'o parts, the lu,nisoLar
determining computed ephemerides that are consistent n'ith the observational datzr.
precession,which is the smooth, long-period motion of the mean pole of the eqrrator
Such a fitting process, with sufficiently accurate observation data, can also providc
about the pole of the ecliptic, r,i'ith a period of aborrt 26,000 years, and nutati,on,
the basis for determining the constants for the procedures ne(iessaryfor reducing
which is the short-peliod motion of the truc pole around the mcan polc with an
the observational data. Some observing techniques for solar-systern objects, such as
amplitude of about 9" and a varietv of periods of rrp to 18.6 years. The cornbination
range measurements and spacecraft obserlations, are independcnt of any external
of lunisolar and planctary precession is callcd genenrl precession,
frame, such as the stellar frame, but these observations have a strong dependence
The notation used here follows Lieske el ol. (1977),u'ho give a usefultable that
on the observingsites, i.e., the telrestrial coordinate systenl, the Earth-orientatiotr
contains cross-referencesto the notation used by other authors.
palameters, and an accurate basis for time. Other observatioms are made q'ith
A basic way of clescribing the motiorr of the ecliptic is to use angles r1 and
teference to stellar coordinate framcs. for which the dynamical reference frarne
II1, where zr,1is the angle betrveen the mean ecliptic at a fixed epodr e6 = lg a1d
must be determined with refercnce to the stellar refererrceftalttc for the consistent
the mean ecliptic of date €o = /, and I1,1 is the angular distance frorrr the equinox
use of these observations.
T6 at the fixed epoch to the a.scendingnode where the ecliptic of clate meets the
fixed ecliptic of epoch (Figrrre 3.21.1). Exprcssions for rq and JIl ar.e obtained
from Newcomb's thcory of the Earth's orbital motion about thc Sun, but with
3.13 The Conveniional Celestial Beference Svstem modifications to allow for irnproved determinations of the planctary rnasses.The
form of the expressions from the theorv are
Conventional reference systems catr be defined io provide a consistent basis for
comparison. Such conventional rcference systems must include specifrcation of the
sin zq sin /1a = s/ + sr I +sz/1,
reference frame at a given date, and all the necessary procedures and constants
required to transform this conventional frarne from one date to any other date. s i nr r Ac o sJ I , 4= c t + c t f +czt3, (3.2t-r)
 -
ft,
100 EXPLANATORY
SUPPLEMENT REFERENCE
3 / CELESTIAL SYSTEMS 101
The combined effects of planetary precession and lunisolar precessiou is called
generaL precession in longitude. Lieske el al. (1977) have adopted the definition
of the accumulated general precession pa as defined by Newcomb, shown in Fis-
ure 3.21.1o
,f
Eclipticat r

Eclipticat ,o
Pr=TR-ToR. 13.214)

Equatorat t
Equatorat to Newcomb actually adopted an expression for p1 that conputed the angb Tse. in
Figure 3.21.1,where Q is the projection of the equinox of date on the ecliptic at
the fixed epoch le. The difierence is small, being 0l'0005/ cy2; hou'ever, at epoch /s
both procedures give the same rate

p = q) - Xcos€0= P0cos€0- P, - scoteg, (3.21-s)

Figure3.21.1
The eclipticandequatorat epochanddate
and p may be resolved into general precession in right ascension, m. aud general
precession in declination, n, where
w h e r e f o r c o n t i n u i i yn 1 > 0 f o r t > 0 a n d r r , 4 < 0 f o r t < 0 . N u m e r i c a lv a l u e sf o r t h e
coefficientsJ, sl,.r2, c, ct, and c2 are given in Table 3.211.1(page 104), expressed m = qrcos €0- X = (P0 cos €o - Pa)cos eo s cosec€0,
conventionally in arc seconds. a = d sin es = (Pocos€o- Pr) sin eo;
The precessionaleffect due to the motion of the ecliptic is described by the arc
PT in Figure 3.21.1, and is referred to as the accumulated planetary precession. hence p = 2g9560+,1sin€0. ( 1 . 2l - 6 )
\ A . I t s r a t e o f c h a n g ea t e p o c hl g i s
The constants of precession in the past have been derived from the analysis
l=Jcosec€0, (3,21-2) of the proper motions of stars, In the solar neighborhood the stars are themselves
revolving about the center of the galaxy. The method was thoroughly reviewed
where e6 is the obliquity of the ecliptic at the epoch re. by Fricke (1977), who concluded that Newcomb's lunisolar precession should be
The basic parameter for describing lunisolar precession is Newcomb's preces- corrected by +l'JlO per Julian century with an uncertainty of 1 Oll5 per Julian
sional constant, which has the value P0 at epoch r0. It is not quite constant but has century.
a very slow rate of change Pl of approximately -0100369 per Julian century. Thc There are several ways of formulating precession. For example, the vector di-
constant occurs in the dynamical equations of motion for the equator becauseof the rections of the polar axis of the Earth and the polar axis of the ecliptic may be
torque produced by the Sun and Moonl so far it has beel impossible to calculate expressed in terms of polynomials involving time arguments (Fabri, 1980; Mur-
a sufficiently accurate value of Pe from its theoretical dependency on geophysical ray' 1983), which are then used to calculate the effects of precession. The method
parametels, and so the valuc used comes from the obsen'ed rate of precession. discussed below has been adopted by the IAU (Lieske el qt., lg77) for rigorous
Lunisolar precession at epoch ts, calculations. It has developed from the time of Newcomb, and is well-suited for
the practical application of corrections for precession to coordinates aud orbital
l=Pocos€o-P8, (3.2t-3) elements.
The accumulated precession angles (A, Za, ar\d.0e, which are used to calcrrlate
is the rate of change in longitude fa along the ecliptic due to the notion of the the efiect of precession on equatorial coordinates, are referred to a base epoch eg
equator at epoch t0, where P is called the geodesic precession,which is a relativistic and have time arguments that describe precession from an arbitrary fixed epoch
nonperiodic Coriolis effect amounting to 1l'92 per Julian century (Lieske et al., 6r to an epoch of date ep. The angles are shown on the surface of a sphere in
1977\. Figue 3.21.2, and polynomial expressions for them are given in Section 3,211.
102 EXPLANATORYSUPPLE[JENT
w 3 / CELESTIAL
REFERENCE
SYSTEIVS

The precessionmatrix P, made up ofthese rotations (seeSection 11,4), precesses

103

equatorial rectangular coordinates from an arbitrary fixed equinox and equator of

epoch €F to one of date eD and is given by

P[e;, e2] = f,.3(-s) R2(+0A)R3( (a). (3.211)

Rewriting Equation 3.21 7 in terms of spherical coordinates gives

cosZAcos dAcos (A - cos ZAcos dAsin (A - cos Z,{sin 0A

sin za sin G sin z^ cos (^

P= sin zAcos 0Acos CA - sin za cos da sin (a sin za sin 0e (3.21-8)

+ cos aAsin (,{ + cos tA cos (A

sin d,acos (A - sin d1 sin (,a cos d.q

Having calculated the precession angles for the matrix P, one can calculate the
inverse matrix P I in various waysl for example,
Figurr 3.21.2
Theprecession anglesCe,u, and0e
p I = R t r ( - ( 4 ) R l r r + d ar R , r r- z e r

= RJ(-CA)R;(+aA)Rl(-zA)
= R:(+&) Rz(-de) R:(+za)
In the figure the pole of the equator at e. is point P6 and the pole of the = Plep, e 71, ,/l tl-q)
ecliptic is point ly's. At the epoch ep, P6 has moved to P and Nn to N. Imagine a
right-handed set of three-dimensional Cartesian coordinate axes with origin at the 'where use
has been made of the property that the inverse of a rotation matrix is
center of the sphere, ,r-axis pointing to Ts, where the equator and ecliptic meet (the
its transpose(i.e., it is orthogonal).
equinox) at the fixed epoch ep. The y-a-xisis 90" away in an easterly direction along
the equator, and the z-axis points toward the pole Ps. Initially precession will move
P6 toward T6, but the movement of N6 as well will cause Po to move in a slightly
3'211 PrecessionAngles and Rates Adopted by IAU (1976) The new basis for pre-
different direction.
cession is taken directly from the discussion of Lieske et al. (1977) and modified
A rotation of -(,a about the z-axis makes P6^fs pass through P, This great circle
by the smatl amount discussed in Lieske (1979). The following values are adopted
meets the mean equator of epoch at right angles at point p and the mean equator
constants at epoch J2000.0
of date at right angles at R. The.r-axis is now in the direction Q; the y-axis points
toward the node S, where the two equators cross,
A rotation +da equal to the angular separation of P from P6 about the new p = 5029iO966 per Julian century
y-axis brings the mea,n equator of epoch to the mean equator of date. The z-axis
Pr = -01'00369 per Julian century
points to the pole of date P, the y-axis to the node S, and the x-axis now points
towa,rd R in the plane of the equator of date. Pe = l'!92 per Julian century
Finally, a rotation of -3a equal to the angle TPR about the z-axis brings R to
^1, eo= 23"26'211448 (3.21l-l)
so that the r-axis points toward ^f, the equinox of date, and still lies in the plane
of the mean equator of date.
 104 EXPLANATOBY
SUPPLEMENT 3 / CELESTIAL
FEFERENCE
SYSTEMS 105
Table3.211.1 which transforms the position vector rs referred to the fixed epoch e6 = /n, to the
Precession
Accumulated Angles position vector r referred to the epoch of date ep = /, and vice versa. The matrix
P (see Equation 3.21 8) is evaluated using the precession angles (,4, zA, and dA (see
sin,rAsin]1A=(4]1976-0:75250f+0l0oc/.31f2rt+(0:19447+0]000697DF-O:OOoi79f3
Table 3.211.1) and the appropriate time arguments.
sin?r,4cosn4=(-46:'8150-0:00117f+010054419f,)t+(olo5o59-O]OO3712D12+0:000344f
Equation 3,212 | can be rewritten in terms of right ascension and declination
r^ = (47:OO29 0106603I + 0:00059872)| + (-O]O33O2 + O:OOO598D P + 0]000060p
(o, 6) as
IIa = 174"52' !9a2+ 328914789 f+0:60622I + (-869:8089- 0:50491Dr+OtO3536f
{ A = F$Ai77U + O:49263T - O:0OO1 24 tz) f + (- 1t07259- 0:001106 n F _ O]OOI147 F
- 1:',88623
x^ = (10:'5526 r+01000096 rr) f+ (-2t38064- 0t000$3D f - O]OO1125 13 sin(o - zA)cos 6 = sin(oe + (a) cos ds
+ 2:22226T - OIOOOO42
p,a= (502910966 T2)r + (1t11113 - O]OOOO42D F - 0]000006F
(4 = (230612181
+ 1139656r - 0t00013912)t+ (0t30188- 0:'000344 cos(o - z,{)cosd= cos(o0+(A)cosd,acos60 sindasin6s (3.212-2)
D t2 + 01017998
t3
z1= (2306i2181+ 1:39656r - 01000139 12)r+ (1:09468 + 0]000066
I) t2+ 0101820313 sin d = cos(oo + (e) sin d,4cos 6e+ cos d,asin 66,
09 - 0t8s$0 r - 0100021
9^ = (2004131 7 12)| + (-o:4266s- otooo2l7 D F - 0:041833F
and the inverse is
sin(os + (1) cos 66 = + sin(a - za) cos 6

The equations in Table 3.211.1 give expressions for the accumulated precession cos(ae+ (1)cos6e= +cos((1- zA)cosd,{cos6 + sin da sin 6 (3.212-3)
angles as functions of time, where the base epoch of the equations is eo = J2000.0
sin d6 = - g6s1q - 21) sin d,acos 6 + cos ga sin d.
or JD(es) = 2451545.0, and the time arguments are in units ofa Julian century, i.e.,

I = (JD(6r)- JD(eo))/ 36525

3.213 Approximate Reduction for PrecessionApproxirnate formulas for the re-
,=(JD(€D) - JD(e)) 136525, (3.2tt-2) duction of coordinates and orbital elements referred to the mean equinox and equa-
tor or ecliptic of date (r) are
where as is a fixed epoch and ep is the epoch of date.
The expressions at epoch ep for the rates per Julian century of general preces- For reduction to J2000.0 For reduction from J2000.0
sion, in longitude, (p), right ascension (m) and declination (z), and n the rate of
rotation of the ecliptic, are ao=a-M Nsina-tand- o=a0+M+Nsinctmtan6m

5o=d-Ncoso. 6 = 6o+ Ncos cr.

.l
p= = 5029i0966 T - o:oooo42
+ 2i22226 T2 ,\o =.\ - a + &cos() + c')tanpo l = ) o + a - b c o s ( . \ o+ c ) t a n t
fiWett,=o
m=
d
+ zila = 4612:4362+ 2:793r2T- o:Un27872 l3o= p - bsin(^+ c') 3=Bo+Dsin(lo+c)
fi{ee
Qo= Q - a + b sin(Q + c') cot ie Q = Qo+a b s i n ( Q o+ c ) c o t i
)
n = I t9otlo' I09 - 0185330
- = 200413 r - 01000217
12
alI io=i-bcos(o+c') i=io+bcos(C)o+c)
ll
r = = 4'7:0029 .0:'06603r+0:'0005S872. c.ro= (., - Dsin(f,)+ c') cosecie o = rro + , sin(Oo + c) cosecL
fitratlr=o

3.212 Rigorous Reduction for Precession The most convenient method of rigor- The subscript zero refers to epoch J2000.0, and cr. and d. refer to the mean epoch;
ously precessing mean equatorial rectangular coordinates is to use the precession with sumcient accuracy
matrix P as follows:
om = o - j{iu + lu sino tano)
r=Pro
$
and 16 = P-r1, (3.2r2-r) 6*=6-jru"oro.

-,
-i
-
 --&
106 EXPLANATORY
SUPPLEMENT
3 / CELESTIALREFEBENCESYSTEMS 107

o r o m= o o + j , t * t r ' no o , - O o , Table3.214.1
Precession
Angles1984Januarldoh
6' = do+ coso. (3.21
3-I )
llv Woolardand Clemence Klnoshita Adopted

and the precessionangles,obtained to suf8cientaccuracy,from expressionsin Ta-

ble 3.211.1for the fixedepochof J2000.0,I = 0, are (o 783.70925 783.70938 7A3.7079A 7A3.7092
z 783.80093 783.80106 783.79942 783.8009
rt sinIIe = 4i 1976t + 0:19447? - 0:000179t
681.38830 681.38849 681.38732 681.3883
trrtcosIId = -4618150r + 01050591+ 010003441
M = (Ce+ ziT t + 0:0003879I + 0:0000101y'
= 0, t) = l? 2812323
- 0?000118s
N = 0t(T = 0,t) = 095567530r I - 0:0000116
f discussion by Woolard and Clemence (1966) was drawn, and the formulation given
a = pA(T= 0,t) = l:396971t + 0:0003086
I in the introduction to the SAOC (1966), as well as a more recent discussion by Ki-
noshita (1975) and repeated by Aoki el al. (1983), where the discussion is carried
= 0, r) = 0:013056,- 0:00000921
b = 1tA(T to one more significant figure than that given by Andoyer.
I
Among the various represerrtations, differences in the accumulated precession
c = I 8 0 " - I I ^ + l a = 5 ?1 2 3 6+2 O ? 2 4 1 6 1+4O
t ?OOO|\22P
angles of the order of I mas are found in as short a time as 30 years. For example,
I a comparison of results from Woolard and Clemence (1966), Kinoshita (1975), and
c' = 180' - II^-;a=s?12362 l:155358t-0:00019641 G.2r3-2)
lhe Enplanatory Supplement (ES) for the accumulated precession angles (s, z, and
d between equinox and equator of 81950.0 and the equinox and equator of 1984
where 'r1. = II|(T = | t, t = O1is evaluated at the mid-epoch. Jan 1d0his shown in Table 3.214.1.
Formulas, ralid over short periods, for the reduction from the mean equinox Results from the Kinoshita formulation agree with that of Woolard and Clem-
and equator or ecliptic of epoch t1 (e.g., the middle of the year) to date (r = r; + r). ence at the 0.1 mas level. Disagreement of the results from the ES at the 1 to 2 mas
where z is a fraction of a year and lrl < 1, are level is brought about by the neglect of higher-order terms in the ES (p. 30). The
Andoyer (Woolard and Clemence) expressions for the precession angles are
o = .tl + r(m +nsin ol tan6l) 6 = 6 1 + r l 1c o so l

) = . \ r + r ( p - 7 rc o s ( . \ t+ 6 " ) t a n B ) B= Pl +irsin()r +6') (0 = (230351545+ l3g:'120tt + 010604) r + (301240- Oi2'1OtJ 12 + t7!995 13

(3.2t3-3)
Q = A+r(p - ? rs i n ( O l+ 6 ' ) c o t i ) i = i r + z r r c o s ( Q r+ 6 0 ) z = (23O35'!545+ 139:720^ + 01060f ) r + ( I 09:480 + 0'1390tr ) 72 + I 8:325 13

.r = Jrt + r7rsin(Ot + 6") coseci 0 = (2o05r:12 - 85:29tt - 037 &)r + (-42:65 - o:3'1t)r2 - 4l]80 13

(3.214-t)
where p, m, n, and r are the annual rates, evaluated at ,l .
where tr and t2 are the intervals in units of 1000 tropical years ol 365242,19a782
3'214 Newcomb's Precession A section on Newcomb's precessionangles has been days between the initial fixed epoch e6 and 81850.0, and between the final epoch
included since it is relevant for the transformation of old catalogs. Since Newcomb's of date e2 and B1850.0, respectively; thus
value of the precession constant (Newcomb, 1895) came into general user several
representations have been used for the practical realization of the accumulated ty = (ep - 81850.0) / (i000 tropical years),
precession angles denoted by (s, z, and 0, which distinguishes them from Lieske's
precession angles (a, 2a, and 0a, Onc may cite the formulations given in the Er- t2 = (ep - B1850.0) / (1000 tropical yea.rs),
planatory Supplement (1961), the development by Andoyer (1911), from which the (3.214-2)
and T=t2-h.
 108 EXpLANAToBv
suppLEt\4ENT 3 / CELESTIAR
L EFERENCE
SYSTEIVS 109
The rates of general precession in riglrt ascctrsion m0 and declination n0 are correction to right ascrnsiou= e tan6 Ao - 1sec26 46,
given by
correction to declirration= fAo,

,n = zl),':= 46Oit':Og|
+27g';44Ott where e= c o s a ( z r + s i n eA q l ) sinoA'!,
+O')t2O4,
fKo+
/ = + sin.r (r1+ sin € Air!) coso A€,
no= - 8s':29
= 20ost':12 tt - o:l 4 (3.2r4-3)
#G);n e = 23?44, ( 3 . 2l 6 - | )

3.215 Approximate Reduction for Prec€ssionand Nutation The folloning fornru- t is the time from the standard epoch to the time of observatior, al(l n tlrc r.rte
Ias are based on the day number method (see Section 3.3.11). and transforrn the of precession in declination. If I is in years then tt is in radians per vear, an<l Ar:'
right ascension and declination (o6, 6s) from the standard epoch and equinox of and Ae are the nutations in Iongitrrde and obliquity at the tinto of obst:r'r'ation.
J2000.0to the true equinox and equator of date, expressedin radians.
8t2sec?6
The errors in arc rrnits caused by rrsing these formulas are of orclcr l0
(r = ac + I + g sin(G + (lo) tan 60, multiplied by the displacement in arc from the companion star.

d = 60 + I cos(G + (1o), ( 3 . 2 1 5I-)

3.22 Nutation
where J, g, and G are given by
The long-period motion of the Earth's rotation axis lvith respc(rl t,o tltr: axis of
'['he'
the ecliptic caused bv lunisolar torque is call lunisolar precessiorr. sltort-1>eriod
J=M+A(n/n)+E, motion of the Earth's rotation axis $'ith respect to a space-flxed coordiuatc s1's-
g=132+1A+1v)2;r/2, tem is called nutation. It is intirnatell' connectedu'ith polar motiou. rvlticlt is thc
movement of the Earth's rot.ttioll axis rlith respect to an Earth-fixed ((Drdinate
G=t;In-t(Bt(A+t )+:M. (3.2t5-2) system. This short-period rot?rtionalmotion of the Earth includes effectsof both
forced motion and free motiott. The forced motion is due to the tor'<1teproduct:cl b1'
The expressions for /, g, and G are an extcnsion of the use of the day numbers. the gravitational attraction of the Nloon and to a lesser extent that of the Srrn arld
which apply precessionand nutation from thc epoch I (e.g., the middle of the nearest planets on the equatorial bulge of thc Earth and any deforrnations of thc Earth'
year) only to thc epoch of date t+r, where r is the fraction of a year. Thus,4, B, ancl For the free motion, the external forces are set to zero in the eqllatidrs of ntotion,
E are the Besselian day numbers, and m and rl are the annual rates of prccessiorr which then vield particular intcgrals. The free motion can be deternrirteclonly frorn
(see Section 3.342).The additional terms M and N (Eqrration B.2tB 2), q.,hicharc observations and mav be excitecl by internal processes.If the llarrth wcrc a rigid
the accumulated precession arrgles, are rcquiredto apply precession frorn J2000.0 body, it would be possible to conlpute the coefficients of the forced periorlic tt:rnts'
to epoch l. The observed values of sonte of the terms are differcnt frorn the calculated valucs,
The values of l. g, and G are tabulated at 10-di*, dates in Z[e Ast.ronomictLl becauseof the difficulty of utodeling the nonrigidity of the Earth. Tlrn prinlipal
AlmantLc. difierence occuts in the coefficicnt of the 18.6-year nutation terrns, wltich is knou'n
as t}ae constant of nutation.
The reference pole for nrrtation and polar motion is called the CelestiaLEphem-
3.216 Differential Precessionand Nutation fhc corrections for difierential preces- eris Pole.It is chosen to be along the axis of figure for the mean surface of a model
sion and nutation are given below. These are to be addecj to the obsened diflercnce of Earth in which the free motion has zero amplitude. This pole has the advan-
in the riglrt ascensionand declination, Aa and A6, of arr object relative to a com- tage that it has no nearly diurnal components of motion with respect to either
parison star to obtain the difierences in the urezr[ a space-fixed or an Earth-fixed coordinate system. Previously, the ephemeris pole
lllace for a standard epoch (e.g.,
J2000.0 or the beginning of the year). The diffct:nc.es Arr and .46 are in the sense was chosen to be the instantancous axis of rotation, but it was pointed out by
"object - comparison star," and the corrections lre in the same units as Ao and Atkinson (1973, 1975) that this ted to quasi-diurnal motions and was not the best
46. In the correction to right ascensionthe same units must be used for Arr and Ad. reference pole.

"1,
 -I
110 EXPLANATORYSUPPLEMENT 3 / CELESTIAL SYSTEMS
REFERENCE 11

The Celestial Ephemeris Pole is the axis about which the diurnal rotation of terrestrial pole was defined by a largcr number of observatories operating opticerl
the Earth is applied in thc transformatiorr betwccn celestial and terrestrial frames. equipment (e.g., PZT's). This system atternpted to maintain continuity $'ith thc
and so it is often referrcd to as the axis of rotation in general discussions. as it was CIO system. It rvascoordinatedby the Bureau Ilternational de l'Heure (BIH), and
in the first palagraph of this section. It mav also be leferred to as the true celestial so it is called the BIH polc. IvIore recently SLR and VLBI have become the most
pole of date (IAU, 1976). accurate observational techniques, and the BIH pole bccame strongly dependcnt
on these data, In 1988 January the international activities of timekecping and of
3.221 Celestial Ephemeris Pole The principal axcs ol a body are the orthogonal monitoring Earth orientation were separated, and the mainienance of the terrestrial
directions that define a cartesian coordinate system in rvhich the morncnt of inertia reference i'rame was taken ovcr by the International Earth Rotation Servicc (IERS).
tensor is a diagonal matrix, i.e,, the products of inertia are zero. For a rigid ellipsoid again attempting to naintain the same pole as the CIO and BIH, brrt bccause of the
of revolution, two of the principal moments of inertia are equal, and the principal higher accuracies now obtained it is necessary to include a model of plate motions
moment is larger about the third. This latter axis is the axis of figure. The Earth in the reductiott of observations. This reference frame is called the IERS l'erres
imperfectly resembles an ellipsoid of revolution, and, in addition, is elastic. Thus trial Reference Frame (ITRF). International timekeeping is non'coordinated by the
its inertial tensor, and the directions of the principal axcs, are functions of time. Bureau International des Poids et \Iesures (BIPNI)
Nevertheless, the instantaneous axis of figure F can be defined as the line passing The terms nutati,on ar'd polar motion ltave been defined as the forced and frce
through the center of mass ol the Earth, r'hich is parallel to the prirnary eigenvector motions, respectively,of the adopted referenccpole (the CclestialEphemerisPole),
of the instantaneous inertia tensor of the Earth. The intersectiorr of F *.ith the but in sone casesthe separation of these effects is not clear. The forced solution is
surface of the Earth is the instantaneous pole of flgure. the solution of the Earth's equations of rotational motion that accounts for all the
The instantaneous axis of figure of a deformable Earth is subject to substantial external forces (specifically the gravitational forces due to the Sun' N'Ioon,and pla.rt
motions due to distortions of the Earth, such as those caused by the bodv tides. ets). The free solution is that solution which rcsults from setting the extcrnal forces
This effect makes it diflicult to interpret the motions of axis F, above, in terms of to zero (ihe particular solution ol the difierential equations). For a rigid Earth this
intuitive, rigid-body concepts. It then becomes useful to define the mean surface results in the Euleriarr free motion. a componcnt of polar motion, rvhich rvould have
geographic axis, B. as an axis attached in a least-squaressenseto the Earth,s outer annual periods in a terrestrial frane. For a nonrigid Earth the period is lengthened
surface. Consider a network of observatories on a rigid Earth and an axis fixed to about 1.1months. and is called the Chandler compo ent of polar rnotiol Hencc
with respect to the positions of the observatories. The only possible motion of the the forced annular polar motiorr due to the climate along with rnotions resultitLg
observatory network is a possibly time-dependent, rigid rotation: that portion of from geophysical strains and meteorological effects should logically be includcd in
the rigid rotation vrhich is not parallel to the axis will cause motior of the axis as nutatiolr. but in practice they are included in polar motion because these effects
the latter follows motions of the observatories definirg it. On a deformablc planet cannot be calculated as accurately as the external forcing eflects
we must allow lor the possibility of other motions of the obsenatorics. and r.e: For an Earth q''ith a fluid core thcre exist a second lree solution, thc so-called
generalize the definition of the axis so that it is defined in a least-squarcs senseb1- nearly diurnal free polar motion (also called the nearly diurnal free rvobble or the
the position of the observatories. The axis B is exactlv this axis in the lirnit of arr free core nutation), t'hich has not yct been clearly obserred. Like polar motiou rt
infinite number of unilormly distributed observatories. Thus if we decompose th<: is part of thc trarsfornation from space-fixed to body-fixed coordinates (Yatskiv'
motion of the Earth's surface into a mean rigid rotation plus a residual dcformatiorr^ 1980).
B moves with the sense prescribed by the mean rotation. The axis B does not
respond to body tides. Note that for a rigid Earth B coincides with F at all times. 3.222 The 1980IAU Theory of Nutation and Reduction The 1980 IAU Th'-ory of
There is no nontidal periodic motion of observatories on the Earth with respect to Nutation was computed for the Celcstial EphernerisPole by deternriniltg the tru-
B. There are motions of observatories rvith resDect to B caused bv crustal motiorrs tations in longitude and obliquity of a rigid Earth (Kinoshita, 1977), and uraking
and, of course, b1' tidal forces. modifications for thc lonrigid Earth model 1066 A (Gilbert and Dzieirvonski, 1975)
The Conventional International Origin, the CIO, was defined by adopted l.ti- by Wahr (1981) in such a fashion that therc are no nearly diurnal motions of this
tudes of the frve International Latitude Service (ILS) observatories contributing to celestial pole with respect to either space-fixed or body-fixed (crust-fixed) coordi-
the International Polar Nfotion Service (IP\{S) (Nlarkowitz and Guinot, 1968). It nates. Tables 3.222.1 and 3.222.2 gfuethe 1980 IAU Nutation seriesaud the set of
is not knorvn whether the CIO is fixed *,ith respcct to B, but it is believed that the fundamental arguments (Van Flandern, 1981) based on the orbital rnotion of the
CIO roughly coincides with the mean pole F over the period 1900 to 1905. Later the Earth and Moon needed for evaluating the series.
fabb 9.D,'1 FEFERENCE
3 / CELESTIAL SYSTEMS 113
Nutationin Longitudeand ObliquityReferredto the Mean Eclipticof Date
Table3.222.1,continued
Argunent Co€ltlclent to. andObliquityReferredto the MeanEclipticof Date
Nulalionin Longitude
Multlple ol Longltud. Obllquity
No. D9 Si Ci A.gument Co€fflclenttor
l,x to-a Pedod Longltud. Obllqulty
No. I D9 Ci
1 0 0 0 0 i 6798.4 -171996 -1742T 9ZOZ5 +a.gI 1'x 1o-a
2 O 0 0 O 2 3399.2 2062 +0.27 -895 +0.5r
3-202011305.5 48 ,24
56202-2212.40 -3
420-2001095.2 tl 0
570002114.8-6 3
5-202021615.7-3i
s8 0 0 2 2 1 7.1 -7 3
61-1 0-l 03232.9 -3 0 123.9 0 -3
59 10 2-2
7 0-2 2 216786.3 -2 1 14.7 -5 3
60 o 0 0-2 1
a20-2 01943.2 10
9 0 0 2 -2 2 1p.6 -13187 _1.6 r _3.1 r or 1-1 0 00 29.85 0
5rc6
1o o 1 0 0 0 365.3 14a6 -3.47 62202016.9-5 3
54 o.t r
63010-2015.44 0
1r 0 1 2 -2 2 121.7 -sl7 +1.2T 224 -0.6r
64 10-2 00 2e.94 0
12 O -1 2 -2 2 365'2 217 -0.5 r -95 +03 r 65000i029.5-4 0
13 0 0 2 2 1 1n.E 129 +0.1r _70
142 00-2 0 205.9 66 11000 25.6 3 0
48 1
15002-20173.3220 67 102 00 9.1 3 0
'17 08 1-1 2 02 9.4 -3 1
1602000182.6 0-17 0 -1 -1 2 22 9.8-3 1
69
170100138e0_15 9 1
70-2000113.72
14022-2291.3_16 +.O.1T 7
190-i00r346.6-126 71 3 0 2 02 5.5-3 1
20-20021.t99.8_63 72 0-1 2 22 7.2 -3 1
73 11202 4.92 1
21 0-1 2-2 1346.6 -5 3 -1 2 1
71 0 2-2 132.6
42 00-2 1212.3 4 _2
752000113.82
23012-2 1119.6 4 _2
24 100104i1_8 _4 76 10 0 02 27.8-2 1
o
25 210-2 013t.7 77300009.22 0
1o
7A002129.32 1
26 0 0-2 2116S.0 10 -1 02 27.3 1
79 0 0 I
2701-220329.8 -r 0 010.1 -1 0
a0 10 0-4
2801002409.2 10
?9-1 0 0113aa.3 81 2022214.6 1
10
30012-20117.5 82 -1 0 2 42 5.8-2 1
10
a32004015.9-1 0
s1 0 0 2 0 2 13.7 _274 0.2f gn o.5r 22.5 1 1
81 112-22
32 1 0 0 O O t.6 712 rO.1r _7 1 1
85 102215.6
33 0 0 2 o 1 13.6 _3a6 _0.4I mO
34 10202 -0.1 r 86-202427.3-l 1
9,1 3ot 129
35i00-2031.8_158 -1 87 -1 0 4 02 9.1 r 0
88 11020 29.3 1 0
36 -1 O 2 O2 27.1 -l
89 2 0 2-2 1 12.4 I
370002014.8 03 -2 4.7 1 0
90 2 0222
38 I000127.7 63 +o1 r -33
39 -1 o o -58 9i100219.6-l 0
0127.4 0.1r 32
40 *1 O 2 22 --59 92 0 0 4-2212.7 I 0
9.6 26
93 3 0222 4.7 I 0
41 102 019.1 _51 27 0
94 10 2 20 23.4-1
4200222 7.1 _38 l6 1 0
95012 0113.1
432000013.a29 -1
44 1O 2-22 96-t-r 0 2135.0 1 0
23.9 29 -12
45202028,9_31 9700-20113.6-l 0
98 0 0 2-1 2 254-1 0
460020013_626 1 0
990102014.2-1
47 -1 02 0127.O 21 -t0 0
-1 1oot0-2209.5-l
48 002132.0 16 a
49 100_2 _13 101 0-l 2 0114.2-l 0
19.1.7 7 '102
50 - 1 t l o -2 1 u.7 -l 0
O 2 21 9.5 -t0 5
109r0-22032.8-l 0
5ttro_2034.8 70 7.1 1 0
104 20020
520120213,27 4.4 1 0
105 0 0 242
53 0-1 2 02142
s4 10222 _8 106 01010 27.3 1 0
5.6 I
5,5 10 O 20
114 EXPLANATORYSUPPLEMENT REFERENCE
3 / CELESTIAL SYSTEMS 115
Table3.222.2
FundamentalArguments Meaneclipticof date
I = 134"57'46:733
+11325r + 198"52'021633)r
+ 311310
72+ 01064 fg
// = 357'31'39:'804
+ (99r -
+ 359003'011224)T 0:577f2 0:U2f3
F = 93o16'18:877+ (1342r - 13!257
+ 82'01'031137)T 12+0:011Ij Mean equatorof date
41i328)f - 6189172 + 0101
D = 297"51'Ol !3O7+ (1236r+ 307006' 9 73
J?= 135'02'401280 - (5r+ 134'08'10i$9) r + 71455I + 0i00873 True equatorof date

Not6: 1' = 3600 and 7 = (JO - 2,151545.0)i 36525 is the numberol Julianc€nturieso136525days
of 86400" ol dynamicaliimgfrom lh€ lundamenlalepoch olJ2000.0lo date-The lundanental Figrre 3,222,1
argum€ntsarel Themeanandtlue equators
of date
/ = the mean longiludeof the Moon minuslhe mean longitudeol the Moon'sp€rig€e,
// = lhe m6an longiludeol the Sun minusthe mean longitudeot the Sun'sperigee,
F = lhe mean longitudeollhe Moon minusths mean longatud€ of lhe Moon'snode,
D = the mean longiludeof lhe Moon minus th€ m€an longitudeof lhs Sun, i.e-, lhe mean which is rewritten in full as
elongationoflhe Moon lrom tho Sun,
- ,'no* t,n.:

t;*,:.
f2 = the longiludeollhe m€an ascendingnodo of th€ lunar orbit on the eclipticmeasuredlrom - sinAry'cos€o
lhe mean equinoxol dat€,

N= cosAd cos€cos€o+ sinesines

. ,
C O SL \ L , C O S€ S l n € 0 -
I
S l n ( C O S€ 0 |

cosAf sinecose6- cosesine6 cosAtl sinesineo+ cosecoseqJ

(3.2224)
Nutation may be rigorously applied to equatorial rectangular coordinates thus:

Nutation describes the motion of the true pole relative to the mean pole and r = Nro, (3.222-s)
may be resolved into the components Al in longitude and Ae in obliquity, which
are shown in Figure 3.222.L The nutation matrix N is a sequenceof three rotations where the position vector r0 refeued to the mean equinox of date is transformed to
which uses these angles and the mean obliquity of the ecliptic, €0, to transforrn the position vector r referred to the true equinox of date.
equatorial coordinates referred to the mean equinox and equator of date to the true The nutation matrix is taken from Emerson (1973), and it is the complete
equinox and equator of date. rotation matrix without any approximations. The nutation angles Ary', Ae ma1'be
The rotations are a lotation of eo about the ,r-axis, which transforms the eraluated from
ry-plane from the mean equator of date to the mean ecliptic of date (see Fig-
ure 3.222.1),followedby a rotation of -Al about the z-axis,which appliesnutation and At =
s.-^
Ab= J,sinAi
in longitude and rotates the x-axis, pointing toward the mean equinox of date '1 , L LLtcosAi
i=l Fl
to the true equinox of date TT, and lastly a rotation of -e, the true obliquity of
date, about the -r-axis, which rotates the ry-plane to the true equator of date. The -vhere Ai = ail + bil' + ciF + diD + eil) (3.222-6)
expressions for the mean and true obliquity of date are given by
and the multipliets ai) bi, ci, d;, e1, and, the coefficients & and C are given in
Table 3.222.7, which defines the 106 terms of the IAU (1980) nutation series The
eo= 23'26'211448 4618150I- 0l'00059
12+010018r3?3, (3.222-1) terms of the series are grouped according to their periods, and arranged roughly
€=€o+A€, (3.2222) in order of magnitude of the coefficient of the nutation in longitude within each
group. The fundamental arguments /, //, F, D, auJ' O, in the FK5 reference system,
where 7'= (JD - 2451545.0)/ 36525.The nutation matrix is a.regiven in Tab\e 3,222,2, Values for these nutation angles and e, calculated using
N = Rr(-e) R:(-Alr) Rl(+€o), (3.2221) these series, are tabulated daily in ?he Astronomical Almanac
 116 EXPLANATORYSUPPLEIJENT

3.223 The Equation of the Equinoxes and the Uniform Equinox The equation of
the equinoxes, which in editions prior to 1960 u'as callcd "nutation in right ascen-
sion," is the right ascension of the mean equinox rcferred to the true equaror ara
equinox and equalsAr/ cose. In Figure 3.222.1it is representedby the arc T1Q and
T
3 / CELESTIALREFEBENCESYSTEI.4S

Al. and Ae" to be added to tho IAU 1980 nutations in longitudc artrl oltlirlrritv.
A{' arrd Ae, respectively arr:
117

it represents the difference betv'een the true zrnd mean right ascensionsof a bodl- A U . = ) ( L J ,s, r n A ,+, L c , ,c o s A , , ) ,
on the true equator; it is thus the difference betlvcen apparent and mean sidereal
time; i.e., :|
Ar. = ) ( U ( ; , c o sA t , + U ' n s t n A n ) , (\ ))t_t r
GAST = GNIST+ Ap cose; (3.223-t\

where GAST and GNIST are the Greenrvich apparerrt and mean sidereal times. where A,, the sum of thr: re<1tirc<l arglrlnentsis gir-enin Equation 3.222 6. Thesc
respectively(Section2.23), which, togcther rvith thc equation of the equinoxesare corrections are made up of trvo tcrms: thc first terms, lS, for longitrrrl<rarrcl OCn
tabulated daily to 0?0001in The Astronomi,caLAlrnanac. for obliquity are called in-phasc, or dircct: the other t$'o terms LC,, arr<l OS,'. arr:
It is sometimes more convenient to refer right ascensiorrsto the uniform equinox called out-of-phase, or indirect.
(Q inFrgrte 3.222.1)(Atkinson and Sacller,1951),which is the true equinox ofdate There are other effects cluc to the planets that are callcd tho plu.rlctl.ly terms
rninus the equation of the equinoxes. The adlantages zrre that in the determirrtr- in nutation. Vondrak (1983) has produced a table (Table 3.224.2)ol85 tcrtrts lltet
tion of horrr angle, right ascensionsof objccts refc'rrcd to the uniform equinox are gives the combined direct and irxlirect planetary efiects that lttt,,c a rttitxtlttllrt
subtracted from mcan sidereal time. These right ascensionsare smooiher functions amplitude of severalparts itr l0 o arcsecondsand should be uscclif rrtilliiucscconcl
ol tinc rvhich reducc the number of ternrs rcquirc'd u,hen they are represented lrl accuracy is required.
polynonial or Fourier-type expression over a fixed time interral. It also sinplifies The planetarl' nutation iu lolrgitrrde.Ar,7. and in obliquitl'Ar,, IItal bc calctt-
interpolation if tablcs gir'e mean siderealtirre ald right ascensionr.eferrcdto the lated from
uniforrn equinox.
85
A ' p = (LJ,, \rn 44 + L( ,, cos A/, )
3.224 More Accurate Nutation Recent analvsis of highly accurate obser\.ations L

(VLBI) irxlicates the need for some systerratic rnrrcctions to the IAU 1980 Nuta-
85
tion scries. Presumably these corrections are duc to the difference of the real Earth
A,.p= L {OC,co\A,,+ OS"sinA,,) | 7 ))4-) |
obtaincd from the model of Wahr. The latest dillcrences Herring (1987) publishcrl
in BIH Annual R.eportfor 1987 (1988) are giverr in Table 3.22,1.1.The correctiors
and, An = anl I s,F + c,,D+ dnQ + enQ+ f ,v + g,,E+ h.M + i,J + i.S

whereLS,,LC,,OC,,,OS,,,zrrrdth<:rnultiplielsan,b,.c,t,C,,e,,.f,.gi.h,t.i,t',
are given in Table 3.224.2. Tlrc fuDdamental arguments /. F, D, and r? ar('given ilr
Table3.224.1
Correctionsto IAU 1980 NutationSeries Table 3.222.2.Q. V, E. M. J, and S, the rnciu heliocentriclongitudcs of tltt' lrlattets
Mercury. Venus. Earth, Nlars. .Iupiter, and Saturn, respectivel!',are gir',tr rrs:
Argument Coeflicienlfor
Multipleot Period Longitude Obliquity T M = 353?3+ t9I4O.3T
Q= 252?3+ 149472.7
No. llF Dn Y = 1 7 9 9+95 8 5 1 7 .r8 J=32?3+3034.97 r 1 t)4-1)
LS, LCN OCN OSN
o 2n bn cn dn en days 1"x 1O-s E=98?4+35999.4T S = 4 8 9 0+ 1 2 2 2 . l T
100001 6798.4 725 417 224
20r000 365.3 523 61 20a -24 where 7 is measuredin.lulian ccrrtrrriesfrom 24515.15.0.
3002-22 182.6 102 -118 ,47 The IAU is at present studying rarious proposals prior to adoptillg aII intploved
400202 13.7 -81 0 32 0 nutation series.In the future l,hcre will be a need lor an improved nutiltion transfor-
mation with improvements to both the rigid body and nonrigid body comporrents'
 3 / CELESTIALREFERENCESYSTEMS 119
Table3.224.2
planetaryTermsin Nutation,CombinedDirectand IndirectEffects
Table3.224.2.continued
Termsin Nutation,CombinedDirectand IndirectEflects
Planetarv
Argumonl Coetticienl fot
Multlpleol Perlod Longllude Obllqulty
Argumenl Coetficient tor
No, IFD!)QVEMJS LS, LCn OCn OSn Pedod Longllude Obllqulty
an bn cn dn en fn 9n hn in k days 1'x10-5 ,-^-.-\,-^-
No. !FDAQVEMJS LSn LCn OCn OSn
'| o 000 002 0 0 0 183 10 00 anbncndnenlngnhDhrn days 1'x 10-5
1 0 -2 -1 301 0 0 o 12924 01 00
3 0 000 012 0 0 0 975 01 00 46 1 1 o 20 -21 0 0 0 14610 0 I I 0
000 0 1 -1 0 0 0 584 16 0 00 00 0 001100780 -1 0 0 0
0 470
0 000 011 0 0 0 139 -40 0l 00 0 oo 12 00177 -l 1 0 0
5 480
490 00 0 001200576/'-53 0 0
0 000 013 0 0 0 266 -1 0 00 -6
500 00 1 o 0 -1 2 0 0 378a3 6 3
7 0 000 o24 0 0 0 448 40 03
0 000 o2-3 0 0 0 1455 o7 00 0 00 0 00 02 00 344 -1 0 00
51
0 000 02-5 0 0 0 209 01 10 520 00 0 002200390-40 00
10 0 000 o2-2 0 0 0 292 -6 1 00 530 00 0 0 0 3 2 0 0 189 I 1 00
00 0 0042 00124 10 0 -1
-1 s40
tl 0 000 o21 0 0 0 162 0,2 0 00 0 00 23 00 902 -1 0 00
12 0 000 o20 0 0 o 1t2 40 0-2
13 0 000 037 0 0 o t72 10 00 00 0 o o 2 -4 0 0 28€,2 0 1 00
'14 0 000 036 0 0 0 325 0 -1 00 00 0 003-400414 11 00
't5 0 000 035 0 0 0 2959 204 17 580 00 0 002-a 30183 14 20
590 00 1 00-48306870 14
0 001 035 0 0 0 5240 0 1 00 00 1 oo 4a 306720 -1 3 21
600
17 0 000 03-4 0 0 0 417 0 4 00
0 0 0 195 2 0 00 0 0 6 I 3 0 183 -4 -2 1
18 0 000 03-3 61 0 000
19 0 033 0 0 0 2061 I 0 01 620 000 0 0 1 0-3 0 489 000
20 2 o20 033 0 0 0 3562 1 0 00 6:t 0 000 0 o2030209 000
u2 0-2-1 0 o20304524 001
21 0 0 0 0 032 0 00127 0 -1 -1 0 0 0 -1
652 0-2'l 0 o-203013630
22 0 0 0 0 o47 0 00734 0l 00
23 0 0 0 0 046 0 00724 51 -1 -2 662 o,20 0 0-203013563 4000
24 0 0 0 0 044 0 0 0 146 10 00 000 0 01o-20439 ,2300
670
25 0 0 0 0 05-8 0 0 0 2864 0-3 20 680 000 0 o 2 0 2 0 199 4000
o-2 1 0 o 2 0 2 0100130 -500-3
692
26 0 000 o5-7 0 00324 20 01 702 0-2 1 0 o20203288 1000
0 0-56 0 0 0 4948 o-4 20
2a 0 2-20 0-56 0 0 0 18186 o2 00 71 0 0 0202016/,2 1000
29 0 000 05-5 0 00117 -i 0 00 722 o-20 0 o20206366 4000
30 0 000 06-9 0 00485 01 00 730 000 0 03020129 1-1 00
740 000 0 o402095 -2001
31 0 000 068 0 00204 10 01 750 000 0 01010399 -16 0 0 0
32 2 o20 0-68 0 0 0 17496 't0 00
33 0 154 '1 0 00 7600 00 0 0 2 0-1 0191 -2001
000 07-9 0 0 0
34 0 000 0 8 10 0 00122 10 00 7700 00 0 03010125 510-2
-1 -1 -1 0
35 0 000 0 8 15 0 0 0 183 00 7800 00 0 000104333
7900 01 0 0001011945 0100
36 0 000 0 I -13 0 0 0 87265 3-4 0 0 0 010'1 0 337 -1 0 0 1
8000 00
37 0 001 0 I -13 0 o o 7373 1l 0 0
38 0 001 0-8 13 0 0 0 6307 0 0 al 00 00 0 o2010175 2001
-12214
39 0 000 0 8 -11 0 00142 0 0 8200 00 0 000202166
40 1 0 0 -1 0 -10 3 0 0 0 6733 0 -1 0 0 8300 00 0 0io20313 o -1 0 0
8400 00 0 000301444 -1 o 0 0
4l 1001 0 -10 3 0 0 0 6865 00 00 0 0 1 0 0-1 378 -1 0 0 0
8500
42 r 0 0 -1 0 -18 16 0 0 o 7296 0 00
,[3 1001 018 16 0 0 0 6364 0 00
1-20-2 0 -18 16 0 o o14 01
,t5 1202 0 18 16 0 0 o14 -1 0 -1

E ph@toty supdmd b nE A5r@.'rl..r ^rfuea -6tnh p..ot, Jury !o, rct2

 120 ExpLANAroRy
suppLEMENT 3 / CELESNAL
AEFERENCE
SYSTEMS 121
3.225 ApproximateNutationand Reduction The rutation matrix given in Equa_ 3.23 SDaceMotion
tion 3.222-4can be simplified to
The motion in space of stellar objects as a function of time must be considered since
I -Aty'cose -Alsine star catalogs give the position of a star at an epoch referred to a mean equator and
I I
N= l+A?y'cos€ I -Ae L (3.22s-r) equinox at a chosen epoch.
taditionally, stellar motion has been divided into proper motion, which occurs
| +Aty'sine +A€ I l
perpendicular to the line of sight and is given as components in right ascension(po)
where Ary' and Ae are expressed in radians and second-order terms have been and declination (p6), and radial velocity i, which occurs along the line of sight. In
neglected, since they can reach only one unit in the eighth decimal place. f-undamental catalogs the epoch of the star positions is identical with that of the
Equatorial rectangular coordinates referred to the mea,n equinox can be con_ equinox to which the catalog refers, but in observational catalogs the epoch may
verted to the true equinox by application of the corrections differ from that of the equinox to which the positions are referred. In all cases,
the correction to position consists of th€ product of the star's space motion vector
with the interval between the required epoch and the epoch of the star's position
Ar= -(ycos e+ 2sine)Ary', in the catalog. In some catalogs the secular variations of the proper motions are
Ay = +rcos eAry'- zAe, also included, and in such cases the mean value of the proper motion during the
interval must be used. For double stars the orbital motion of the comDonents with
Az = +r sin€A?y'+yA€. (3.22s,2) respect to each other may also have to be included.
The position vector r of a star at the catalog epoch r is given by
Similarly the first-order corrections (Aa, Ad) to right ascensionand declination
may be calculated directly from Ircos6cosa]
r=lrcosdsinol, (3.23-r)
I rsin6 l
Ao = (cose + sinesino tanf)Aty'- cosotan64€,
where o and 6 are the catalog right ascensionand declination, and r is the barycen-
.4'6= sinecosoAry'+ sinoAe. (3.22s-3
) tric distance to the star in au, which can be computed from

These can be combined with the reduction for precession from the mean equinox
r = | /sinp, (3.23-2)
of the middle of the year by means of day nurnbers (see Sections 3.21b and 3.34).
It is also possible to calculate the nutations in longitude and obliquity to about where p is the parallax of the star. If p is unknown, unavailable, or zero to within
l" by considering only the dominant terms of the series (see Table 3.222.1), thus the accuracy of measurement, set it to some small but finite positive number in
order to avoid mathematical indeterminacy; a choice of l" x 10 7 will efiectively
A/ = -0?0048 sin (t25?0 * O?05295d)- 030004sin (200?9+ l?9712gd) place objects of unknown parallax at a radius of l0 Mpc.
The space motion vector i can be obtained by differentiating Equation 3.23 1,
Ae = +030026cos( LZS?O- O?OS295d)+030002cos(20019 + t?g7tTgd), giving

(3.2254)
cos6sina - sin, coso cosdcosrll t5sra.'l
. f- f
r= cos0 coso - s i n 6 s i n a cos6sina (3.23-3)
wheie d is the number of days ftom 245L548.O. I I srutI ,
L cos6 sin6 I L fti l
where po and p6 are the proper motions in right ascensionand declination, respec-
tively, and i is the radial velocity. The factors 15s, s, and /< convert pd, p6, ar'd i
3.226 Differential Nutation For objects within a small area of sky difierential nu- into the required units for the space motion vector, assuming that po is in units
tation is always combined with difierential precessionl see Section 3.216. of time. If i is required in AU per day, and po and p6 are in seconds of time and
 3 / CELESTIALREFERENCESYSTEMS | <\t
122 EXPLANATORYSUPPLEMENT

A linear approximation to a star's motion in space, which ignores the changes

seconds of arc per Julian century, respectively, aud i is in kilometers per second,
then in proper motion due to the changesin right ascensionand declination, and ignores
foreshortening and radial relocity, may be made when the highest accuracy is not
needed, or if the time interval (r) is small such as in the day-numbers method
s =2r l(36Ox3600 x 36525) convertsfrom arcseconds/cyto radians/day,
(Section 3.34) when r is a fraction of a year. Corrections are made directly to the
k = 86400/ (1.49597870x l0*") convertsfrorn km/s to ,lu/day. r I t'l--d r right ascensionand declination. via

AII catalogs do not necessarily give the proper motions as rates per century; some- ot=cr+Tpa, =6+rpo, (3.234)
times they are per year.
It is worth considering the casesin which ihe available data are incomplete. The where r = 11- / is expressed in the appropriate units.
radial velocity is required only for relatively nearby stars for which foreshortening
effects (second-order changes in the apparent motion of the star due to the shifting
aspect of its motion) are significant. However, the radial velocity is uselessin this
regard (and should be set to zero), if the parallax (distance) is unknown. If the star's
3.24 Parallax
radial velocity is zero or unknorvn, or has been set to zero because the parallax is
not known, then the above space motion is tangent to the celestial sphcre at the
Pamllax is the term used to describe the difference ir] the apparent direction of
star's catalog position r. Convelsely, if the proper motion components are zero.
an object as seen from two different locations. For example, the Sun, Nloon, and
then the star has no known tangential velocity. Also, it may be that the propcr
planets are observed from the surface of the Earth, and most almanacs publish their
motion cornponents are known but not the parallax. In such a case, if a "reasonablt:
geocentric positions. The term is usually used in connection with a shlft of origin
guess" parallax lalue is not used, then the computed velocity components corrlcl bc
from the center of the Earth to the surface (diurnal parallax), or a shift from the
greater than seemsphysically plausible. However, this is a computational curiositr
barycenter or heliocenter to the geocenter (annual parallax).
with no physical meaning or practical effect on the results of the calculation.
The following is a simple and rigorous lncthod of making correctious for both
The position rl of the star at epoch /l is calculated from
diurnal and annual parallax, and it is recommcrrded that it be used in prefcrence to
approximate methods. The position vector of a body rs, with respect to the origin
rr =r+(tr -t)i, (3.23s) E, is given by

re=uB-Ea, r'l )1-l \

where /l - / is the interval between the required epoch and the epoch of the catalog.
If i is in eu/day then clearly (r1- t) must be in days. Use of the space motion vcclor
(Equation 3.23 3) includes both proper motion and foreshortcningefiectsirr 11. where Es and ug are the position vectors of the nerv origin and the bocly,respec-
The errors that are being made in this method of calculating space motiorr har.e tively, with respcct to the old origin and refcrrcd to the same referencc fratne.
been investigated by Stumpff (1985). Basically, proper motion ancl radial velocitl' Approximate methods arc given for calculating diurnal and annual parallax
are observablesthat should be converted to inertial quantities. Thus catalog proper that are useful when the appropriate data are not available; e.g., the distance of
motions include thc effect of aberration. whereas radial velocities include a rcla- the bodv is not known.
tivistic Doppler effcct. Fortunately, these effects arc I'ery srnall and are importtrnt
for only a few nearby stars whose apparent rnotions are charrging rapidly across
the line of sight. Siumpff has shown that the crrors in neglecting these elfects can
3.Zl Diurnal Parallax Diurnal parallax is a shift of origin from the center of
rncrease progressively with time, so that after several decades they may bccome
the Earth to the surface of the Earth. Coordinates with an origin at the surface
significant at the mas level.
of the Earth are called topocentric coordinates, and depend on the position of the
There is no explicit correction for light-time in stellar apparent place compu-
observer. Using Equation 3.2,1 1, the topocentric rectangular equatorial coordinates
tation; it is assumed that the position and space motion vectors implicitly include
of a body r/ are given by
the light-time and its time derivative.
 '124
EXPLANATORY
SUPPLEMENT 3 / CELESTIAR
L EFERENCE
SYSTEMS t<c
r'=r-g(/);
where a and A (the unit distance)are dcfinetl b1.the IAU (1976)sl.stemof constanrs.
Le. x' = r' cos6'cos a' = rcos 6 coso - p cos cos(d
@' + )), The equatorial horizontal pnrallax (horizontal paralla-x) of any objcct lrav bc
y' = r'cos 6'sin a' = rcos 6 sin a - pcos calculated from
4tlsin (d + )),
z' = r' sin6' = rsin 6
(3.242-3)
- psin (ft,

(3.241-r) where 4 is the geocentric distance of the object in AU. In preliminan- rvork orr
where r and g(r) are the geocentric position vectors of the comets and minor pla'ets, where the geocentric distance is u.known, it is convcniett
body and the observer,
respectively, referred to the same reference frarne. and to calculate parallax factors po and p5 for each observation; these nrav bc usetl. oncc
the geocentric distances are deternlined, to give the parallax corrections i'thc fo.u
/2 = x'2 +y'2 + z'2, at =tan-t1/ / xt), 5' = sirr-t(2,
/ r,). (1)ll-)\
661=p,"lA, trf=p6tA, (3.2424)
The geocentric position of the observer is given here in
terms of p, the distance
of the observer from the center of the Earth. the geoce[tric where Aa = o o/ anri 46 = 6 - 6/. The parallax factors are calculatecl frou
latitude //. the Iongitude
A^(east-longitudes are positive), and the Greenwich sidereal
tirne d. If the position
of the body is referred to the true equinox of date, p. = 8'.'7941
48 p cos // sin i sec6 = 015862765 p cos // sin I sec6,
then apparent sidereal tirne
must be used. Section 3.244 gives the formulas for the positiorr
of the observer in p6 = 8'.'794148
p (sin/'cos 6 - cos/'cos & sin 6),
an Earth-fixed reference frame. e.242_5)
The topocentric hour angle (lt,) may be calculatcd
from l/ = g +,\ - a,, where
apparent or mean sidereal time is used as appropriate, where p is expressed in -tu.
and g and a/ are expressed
in the same units. At upper meridian transit The parallax correctionfor altitude. callcd parallax in altitude, is gir cn approxi_
l/ = 0.
mately by rcosa'. rvhcrea/ is the topocentric altitude: it is added to the lopocentric
altitude to form the geocentricaltitude to low precision(about 0ll ); therc is no cor-
3'242 Approximate Diurnar paralax Approxirnations rection for azimnth (Z). This approxirnation does [ot allorv for the oblatent,ss of
to Equation 3.241 1 nar-
be made in various ways. For objccts relatively the Earth, which, for the N,Ioon,can produtx, an crror of up to 012. For an obsr:rver
close to thc E.;;f,,-"; ;"';;;
Moon, whose parallax is significant, it can at latitude d the oblatcnessof the Earth can bc corrected for b1. atlding thc next
be assumed that the Earth is spherir:al
by setting p = I Earth radii. For the more t e r m o f t h e e x p a , n s i o nw. h i c h i s r / s i n 2 / c o s a / + r r l s i n 2 / c o s z s i n a t , t h e r c
distant bodics, su<:h as the Sun. tlrt: f is
planets, or comets, ll'hose parallax t h e f l a l l e r i r g { s r . cS e , . r i o n3 . 2 1 1 . ) .
amou.ts to onry a fc'i, secr'ds of arc, it carrr
be assumed that the geocentric and
topocentric distance are equal; j,e., / = r.
O:Oiesthe ropocentric right ascensionand
}::j:l :r* declination (o,, 6,) mav 3.243 Approximate Annual Parallax The reduction for anrrual pa,rallax frorrr the
oe approxrmated using the geocentricposition (o. 6) to sufficierrt
accuracv bv trrc bary-centric place (o, 6) to [he geocentricplace (n1, 61) is given b1.
nlsl-ordcl approXimatiOn

o r = a + ? r ( X s i n a- / c o s o ) / ( l 5 c o s 6 ) ,
o' = o - d/sinisecd,
" :cos 6 r = 6 + r ( X c o s n s i n 6 + l s i n o s i n 6 - Z c o s6 ) , (3.243-l)
6 ' = 6 - r e $ i n O t c o s 6- c o s @ / c o s i s i nd ) , (\)L) t\
where ?r is the parallax in arcsec, X, Y. Z arc thc barycentric coordirrates of thc
where ,r is the equatorial horizontal parallax, Earth in eu, and the right .Lscensionis expressed in units of time.
a l,he equatorial radius of the Earth.
p the geocentric distance of the These expressions may be simplified by using the star constants c, d, c', d, (sce
obseryer, and l the gr:ocentric hour arrgle.
The solar paralla-x is given by Section 3.342) incorporated into the day-number urethod (Sectiorr 3.341) in order
to correct for annual parallax. Thrrs

sinTa = a /A and ?ro = B':j94149, (3.242-2) at - d= L(}=T(dX cY),

 126 EXPLANATOBYSUPPLEIVENT 3 / CELESTIALBEFERENCESYSTEIJS 127
61 6= L6 =r(d'X c'1), t 7 )L1_) \
j;r)*"oo- *:r3"o'6d,
. (*tt
v/here Ao and 46 are in the same units as rr. Also, they may be combined with
the correction for annual aberration (see The Astronomical Almanoc,822). c=t +)r+*r. *r - ()t.)r .frr) *,'o
The times of reception of periodic phenomena, such as pulsar signals, may be
reduced to a common origin at the barycenter by adding the light-time correspond- * (*1r'*]f) *"oo- fir' coseo,
ing to the component of the Earth's position vector along the direction of the object,
that is, by adding to the observed times (X cos crcos 6+ y sin o cos 6+Z sin 6)/c, where
c = 173.14Au/day, and the light-time for 1 AU is | /c =U0057755.
o=r- )r*trr. *r. ()t- ftr)*",r
. + cos6d,
3.244 Terrestrial Coordinates There are three commonly used ways of express-
(*, *r)cos+d fi /3
ing terrestrial coordinates (i.e., Earth-fixed): (i) geocentric equatorial rectangular
d 6 ' = (, . - + * srn od. (3.2444)
coordinates, ,x, y, z; (ii) geocentric longitude, Iatitude, and distauce, ), $', p; (iii) )r) sn\2,f(:r'z f,f) "i,+o l/3
geodetic longitude, latitude, and height; l, /, h. Geodetic coordinates a.rereferred
to a reference spheroid (an ellipse of revolution), which is normally geocentric, and The expressions for p and $ - $' are for points on the spheroid (l = 0), and the
is defined by its equatorial radius (a) and flattening (f). For example, those of latter quantity is sometimes known as the "reduction in latitude" or "the angle of
MERIT (1983) are the vertical." and it is of order 10'in midlatitudes. To a first approximation when
ft is small, the geocentric radius is increased by h I a and'the angle of the vertical is
a = 6378.137
k:n and f =11298.25'7. (3.2441) unchanged. The height ft refers to a height above the reference spheroid and difiers
'\rndrrlation of
from the height above mean sea level (i.e., above the geoid) by the
The geodetic and geocentric longitudes of a point are the same. The following the geoid" at the point.
relationships hold between the geocentric and geodetic (geographic) coordinates:

.ir= p cos d/ cos ^ = (aC + &) cos @cos,\, 3.25 Aberration

y = pcos @/sin.\ = (4C + lr) cos srn^, (1)L.l
d )\ The velocity of light is finite, and so the apparent direction of a moving celestial
2= psrn6' = (aS + /l) sin f, object from a moving observer is not the same as the geometric direction of the
object from the observer at the same instant. This displacement of the apparent
where a is the equatorial radius of the spheroid, and C and S are auxiliary functions position from the geometric position may be attributed in part to the motion of
that depend on the geodetic latitude and the flattening / of the reference spheroid. the object, and in part to the motion of the observer, these motions being referred
It follows from the properties of the ellipse that to an inertial frame of reference. The former part, independent of the motion of the
observer, may be considered to be a correction for light-time; the latter part' inde-
pendent of the motion or distance of the object' is referred to as stellar aberration'
C = ( c o s 2 d + ( l - 1 1 2 s i n621 - 1 1 2 , s = ( l - f l r c . | 1 )44-1\
since for the stars the normal practice is to ignore the correction for light-time The
sum of the two parts is called.planetary aberration, since it is applicable to planets
Geocentric coordinates may be calculated directly from geodetic coordinates
and other members of the solar system.
and conversely. It is easier to use an iterative procedure for the inverse calculation
(see Section 4.2.2). Series expansions, which contain terms up to
f3, for S, C, p, and
Q-6'arc 3.251 Light-Time In Figure 3.251 | E is a stationary observer at time t, and P
is the position of a celestial object also at time /. The dotted curve represents the
s=r - 3ur
+*r. *r - (it - )r - )r) *"ro orbit of P. The light which is received at E at time , was emitted by the celestial
obiect when it was at P/ at time (t - 7), where 7 is the light-timei i'e', the time
 128 EXPLANATORY
SUPPLEMENT 3 / CELESI1AL SYSTEMS
REFERENCE 129
The classical Newtoniau expression for the direction of the source as seen by
the moving observer is obtained by vector addition of veiocities a-sfollows:

p+Y lc
P t= (3.2s2-t)
lp+v/cl'
F('-:) ---r_t(,)
Takiug the scalar pa-rt of the vector cross product of p with Equation 3'252-1, then

sinAd =
,rc+rgfd+9i7fr
=I"^'-i
(V lclsin0
(:)' s i n2 0+ . . .,

s i n c el p n p l l = 5 i 1 7 \ pl ,P a p =
l 0, and lt"Il= !sinl. (3.2s2-'2

The term of order V / c is about 0. 0001 or 20"; the term (V / c)2 has a maximum
value of about 01001.
In special relativity, the velocity of Iight is constant in the moving and station-
axy frame, and the Lorentz formula for the addition of velocities applies. Hence

E(t) r p + ( V i c ) + ( p ' V I c l ( VI c \ | ( l + a - 1|
B.
p r= T (3.2s?-3)
Figure3.251.1
Light-time aberration
where 6-r = \fr=O /4, again taking the modulus of the vector crossproducts
of Equation 3.252 3 with P, then
taken for the light to travel from P/ to E. The direction EP' is called the geometric tl
( V / c ) s i n d ++ ( v/ c ) 2s i n 2 l0\ l + B
direction of the object allowing for light-time. stn^o=ffi
Light-time is calculated iteratively, with a first approimation I calculated
from the geometric distance between the object and the observer at time r, and
the next approximation 12 calculated from the distance between Pf at (t - rl) and
=i"^'-i(I)' +...,
s\n20 (3.2s24)
the observer at time t, No allowance has been made here for the relativistic delav
caused by the Sun's gravitational field. However, the apparent-place algorithm given which shows that special-relativistic aberration and classical Newtonian aberration
in Section 3.31 (3.315) includes this efiect, which is of the order of 1 microarcsecond. agree to order V/c (mas precision). However, it is recommended that special-
relativistic aberration (Equation 3.252-3) be used, particularly when high precision
3.252 Stellar Aberration In general, the observer at E will be moving with a ve- is requied (see also apparent-place algorithm, Sections 3.315 and 3 317)'
locity V. The appa.rent change in the geometric direction of the celestial object at The motion of an observer on the Earth is the resultant of diurnal rotation of
I/ due to the orbital motion of the Ea.rth about the barycenter is called stellar the Ea.rth, the orbital motion of the Earth about the center of mass of the solar
aberration. In Figure 3.252-f, p is a unit vector in the geometric direction EP', system, and the motion of this center of mass in space. The stellar aberration is
i.e., in the direction of the body at time t allowing for light-time (but ignoring the therefore made up of three components, which are referred'to as d'iumal aberration,
effect of light deflection). The observer is moving with a velocity V relative to the annual abermti,on, and. secular abenation. The stars and the center of mass of the
stationhxy frame, and at time t will observe the body at P// in the direction p1, solar system may ea,chbe considered to be in uniform rectilinear motion; in this
where FfB = 0 is the angle between the direction of motion and p in the stationaxy caae the correction for light-time and the secular aberration are indistinguishable,
frame, and FEP" = Ad is the displacement due to aberration in the moving frame, a,nd the aberrational displacement due to the relative motion is merely equal to the
which is always toward the direction of motion. proper motion of the star multiplied by the light-time; it is constant for each sta'r'
 130 EXPLANATORY
SUPPLEMENT REFERENCE
3 / CELESTIAL SYSTEMS 131
and in general, is not known, and is therefore ignored (see Section 3.23 on space the Earth with respect to the Sun, obtained as a couection to be applied to the
motion for further discussion). The term "stellar aberration" is sometimes loosely geometric place a, 6 in order to obtain the apparent place o1 , d1:
used in contexts where "annual aberration" should strictly be used.
a; - a = -(n sin O + re sin 11)sin o sec6

3,253 Classical Annual Aberration In accordance with recommendations of the - (/r cos O cos € + Kecos ll cos e) cos o sec6
(3.253-3)
International Astronomical Union (1952) the annual aberration is calculated as
6r -6= - ( K s i nO + r e s i nZ ) c o sa s i n6
from 1960 from the actual motion of the Earth, referred to an inertial frame of
reference and to the center of mass of the solar system. The resulting aberrational - (/i cos O cos 6 + Kecos JI cos €)(tan€ cos 6 _
sin o sin 6).
displacement Ad may be resolved into corrections to the directional coordinates by
standard methods. If, for example *, i, and 2 are the components of the Earth's where O is the true geometric longitude of the Sun, e and II are the eccentricitv
velocity parallel to equatorial rectangular axes, the corrections to right ascension and longitude of perigee of the solar orbit, e is the mean obliquity of the ecliptic.
and declination, referred to the same equator and equinox, in the sense "apparent and rcis the constant ofaberration (see Equation 3.253 4). The second term in each
place minus mean place" are, to second order in V / c, factor in Equation 3 253-3 depends explicitly on the eccentricity and represents the
components of the displacement due to the departure of the elliptic orbital motion
YY from a circle. The component of the aberration that depends on e is known as
cosdAa = -a sina + l cosa
cc elliptic aberration.
I The constant of aberration r is the ratio of the mean orbital speed of the Earth
+ ,(Xsino y c o s o ) ( X c o s o+ y s i n o ) \ e c d+ . ,
c. to the speed of light, where perturbatiorrs and the motion of the Sun relative to the
barycenter are neglected. It is derived from
AO = --
x vC OzS OS l n O - - S l n A S r n O+ - C O S O
ccc
n=rot\rtI-"r), /1 t 51--dI
* ,in o - i'coso)2tan6
$<*
where c is the speed of light, a is the mean distance of the Earth from the Sun, n
I
+ ;(-tcos 6cosa + i'cos6 sin ri + 2sin6) is the mean motion, and e is the eccentricity of the orbit.
The annual aberration due to the barycentric motion of the Sun or to thc action
x ( - l s i n dc o sa + i ' s i n 6 s i no - Z c o s6 ) + . . . . (3.253l) of any particular planet may be obtained, when the ecliptic latitude of the planet
is neglected, from
These equations are usually used to first order in (V/c), and ignoring the Z term ar'
expressed in terms of the Besselian day numbers C and D (Section 3.342), u'hiclt o, - o = -!!!6in o sin 1+ cose coscrcos/) sec6,
c
represent the classical annual aberration terms. An account of how C, D, and the
E-terms of aberration have been calculated in the past is given in Section 3.53. 6 1- 6 = - ! ! ! 6 o s a s i n f s i n l + c o s / ( s i n ec o s 6 c o s € s i n a s i n d ) ) ,( 3 . 2 5 35 )
c
Atkinson (1972) gives a useful algorithm using a truncated series for calculating C
and D to 0'i001. where I is the heliocentric ecliptic longitude of the planet, tn is the ratio of mass of
To a lower precision it is possible to use the expressions the planet to Sun, a is the meal distance, and n is the mean motion. The cocfficients
to be used are given in Table 3.253.1. The Sun's aberration in longiiude, a-ssuming
X=+0.0172sin.\, i= -0.0158cos.\, Z= -0.0068cos) /1t<1-)\ unperturbed elliptical motion, can be givcn as

for the barycentric velocity of the Earth with respect to the mean equator and A , \ = - n s e c0 ( 1 * e c o s y ) = r i c ( l e2'11R, (3.2s34)
equinox of J2000.0, where ) is the apparent longitude of the Sun.
Also to first order in V/ c, we obtain for the aberration in right ascension and where tc is the constant of aberration; a, e, and. y are the semi-major a-,<is,eccen-
declination due to the unperturbed elliptic component of the orbital motion of tricity, and true anomaly of the Earth's orbit; and
B and R are the Sun,s latitude
132 EXPLANATORY
SUPPLEMENT 3 / CELESTIALREFEFENCESYSTEMS I J\t

Table3.253.1 where the plus and minus signs are used for upper and lower transits. respectivelr':
(nma/ c) for the
Coetficients
this may be regarded as a correction to the time of transit.
MajorPlanets
Alternatively, the effect may be computed in rectangular coordinates using
venus 010001 the following cxpression for the geocentric velocity (i) vector of the observer wilh
Eadh 010001 respect to the celcstial equatorial reference frame of date:
Jupiter 0t0086
Saturn 010019
uranus 0t0002
{/sin(d+ )) l
I c,'pcos
Neptune 0t0002 t= d'cos(o+ I) ,
L.,ncos (\ )\/--ll

| I

where d is the Greenwich sidereal time (mean or apparent as appropriate). and ,\

is the longitude (east longitudes are positive).
and true radius vector. The e2 term afiects only the fifth significant figure; so until The geocentric velocity vector of the observer is added to the barycentric veloc-
the constant of aberration was specified io five significant figures and the Iongitude ity of the Earth's center, to obtain the corresponding barycentric velocity vectors
of the Sun was given to three decimal figures, this term was ignored, of the observer(seeSection 3.353)
Measurements of radial velocity may be reduced to a common origin at the
3.255 Planetary Aberration Planetary aberration is the apparent displacement of
barycenter by adding the component of the Earth's velocity in the direction of the
the observed position of a celestialbody produced by both the motion of the body
object, that is, by adding
and the motion of thc Earth. It is often calculated by adding the couection for

(3.2s3-7) stellar aberration (Section3.252) to the geometric position correctedfor light-time

icos o cos6 + l'sin o cosd + Zsind.
(Section 3.251). On the other hand it may be calculated clirectll'.either b1-using
-
the barycentric positions of thc body and the Earth at tinrc (t 7), or by using the
3.254 Diurnal Aberration The rotation of the Earth on its axis carries the ob- barycentric positions and velocities of the body and the Earth at time r as described
server toward the east with a velocity apcos$'. where u.,is the equatorial angular below. Horvever, although these latter tt'o methods are sirnple and rvidely used. thc
velocity of the Earth (the standard value of c,:is given in Section 3.353; if a is thc first method may bc preferrcd, since it brings out the principle that aberration
equatorial radius, then eu -- 0.4@ km/s is the equatorial rotational velocity of thc depends upon the relative vclocity of observer and obje<:t.
surface of the Earth), and p and $' ate the geocentric distance and latitude of the If we denote the barvcentric position of the Earth at tinc t by EB(/) and thc
observer, respectively. The corresponding constant of diurnal aberration is barycentric position of the planet at time / by uB(/)r thcn the geometric geocentrlc
qa position of the planct at timc t allowing for light-time is given by
2 cosd'=o':32ooe cosd/.
cosd/=0:021334 / ? )54-l )
ca a a
P(r)=us(r-r)-Es(t), (3 . 2 5 5l-)
The aberrational displacenrent may be resolved into couections (apparent mean)
in risht ascensionand declinatior whete z = P/c is the liglrt-time and P = Pl Il P1 is the geocentricposition of
the planet allorving for planetary aberration, then using classical stellar aberration
A a = O i O 2 l 343 c o s d ' c o s l , s e c 6 , (Equation 3.252 1)
a (3.25s-2)
Pr=P+rEs.
! cos/' sinft sind,
A6 = Oi32OO t 1, )</-)\
a
If we assume that thc velocity of the Earth EB and the velocity of the planet Ds
where i is the hour angle. The efiect is small but is of importance in meridian axe constant during the light-time, then Equation 3.255 1 may be written as
observations.For a star at transit. i = 0" or 180', so Ad is zero. but
Pr = us(r- t) - EB(r- t), (3.2s5-3)
Aa = t0i021334 cosd'sec6, / I t54-1)
o
 134 EXPLANATOF]Y
SUPPLEMENT 3 / CELESTIALBEFERENCESYSTEMS roc

Pr = us(r)- Ee(r) r(us En). f l 755-4r In a transit of a plartct acrossthe disk of the Sun, c.g., thc cxternal contacts
occur when the observcr is ou the colrical surface that circu[]scribes 1'hc Sun and
Since Newtoniau arbcrrationand special-relativityabcrration agrce to order y/c, it the planet and has its vcrtex bctween thc Sun and tlte planet. Thc intcrlal cotrtact
is only when mas precision is required that it bccomes necessary Lo use a rlgorous occur when the observer is on thc cone circumscribing the pl.ulct lr,uclthc Sun
formula (seeSection 3.252). having its vertex betlvcen the planet and the Earth. The obsen'ed contacts arc at
the instants whcrr the apparent positions of a point on the limb of tlle plarret and a
3.256 Differential Annual Aberration The diffcrcrrtial <.oorclinatesof a rnor.ing ob- point on the limb of the Stut are tlte sarne;i.e., the ray of light froln thc Sun that
ject with respcct to a fixcd star rvill be affectcd bl.diffcrential aberration: if Aa. reachesthe geometric positiott of thc observer at the instant 7. of contact has grazed
Ad arc the observeddillerencesof the coordinatesin the senscmovinq obiect minus the planet on the rval'. This ral left the Sun at a previous tinrc ? 12 and reachcd
star. the! thc corrcctions for differential annual abcrration tu.e the planet at time (f - 12)+ 71. The circumscribing cones are forrn(l br t he grazing
rays; hence. the points on the Earth and the planet that lie in thc strnre straight
A(a1 - o) = (Dcos o - C sin a) sec6A(r + (D sin o + Ccos o) sec6tan 6l\6, Iine on one of the cones at the instani of a contact are the geotlletric position of
A ( 6 r - 6 ) = - ( D s i n o + C c o s a ) s i n d A a + ( D c o s r-i C s i n o ) c o s d A d the observer at the timc I, and the Seometric position of the point olr the planet at
time T - 12+ 11.Therefore, in the formulas of the theory of trartsits, for anv value
Ctan essin dA6, (3.2s6-
l) of the time T, all quantitics dcpcnding on the time must be derivtrl frorn the values
-
of the geometric coordinatcs (r, 1, b) of the planet at T 12+ 11 artd the geometric
wherc C ancl D are the aberration dav numbers ck'firx,d in Section 3.342, ancl thc
coordinates(/,1', b') of the Earth at /.
units of cach tcrrtr are consistent.Thc corrections shotrlclbe applied q,ith thoso Similarly, in cornparing obscrvedpositions of objects iIr the solar systcrn \\'ith
for diiTcrcntialprecessionand nutation to give nrciur positions referredto the s.tme
one another or $'ith reference stars, in order to determine the coorclinatcs of a body.
equator and equirroxas those of the stars. great care is required iD corrocting the observations for aberratiolt. zrccorclingto the
Alternatir.elv the.orrections for differential anrlral aberration to be added to
means of observatiorr used, attd thc method of comparison.
the obselved cliflereuc.es(in the sensemor.ing oltject rninus star) cf right asccnsioll
In eclipsing binarJ-sJstclns. an apparent variation of the perio'l tnal-he pro-
and declination to aive true differelces are
duced by the rariation in light-time with changing distance froln thc obscrver due
to an orbital motion of the eclipsingpair t'ith respectto a distant third cornpone[t
in right ascension aAo + bAd iu units of 01001 (Irwin. 1959).
(3.256-2)
in dr:r:lination cAn + dA6 irr urrits of 0l0l

rvhere Ao, 46 are thc observeddifierenccsin uni1,sof l. arrcl I' respectir.ely,ancl 3.26 Gravitationallightdeflection
where a. b, c. d are coeflicieuts defined by
Gravitational light dcflection rvas predicted by Einstein and first t:oltfirrncd photo-
graphically at the eclipse of 1919 NIay by expeditions frorn Grecnwich and Cam-
a = -5. 701cos(f/ + (r) sec6,
bridge, England (Dyson et ol., 1920). It has been measuredmany tirrres'and mr'st
b = -0. 380sin(I1+ o) secd tan6, recent high-precision rDeasuremcrttshavo been made q'ith radio illterferolnel'els'
c = + 8 . 5 5 2s i n ( H+ o ) s i n6 , which can observesourcesvcty closeto the Sun. Fomalont and Sranlek (1975) lrave
confirmed the deflection predicted by Einstein's theory rvith an accuracv of 1%'
d= 0 . 5 7 0 c o s ( H+ a ) c o s 6 , which is an order of magtritttde lnore accurate than that achieved opticalll from
eclipse observations. The deflection incrcasesthe closer the light path is to the Sun;
Il =23.4 ( d a vo f l e a r / 1 5 . 2 ) . ( 3 . 2 5 63 )
at a heliocentric elongation of 90o, however, the deflection has decrr:asedto 01004
3.257 Differential Planetary Aberration Abcrral iorr, bccausc of its dcpendence orr The algorithm for the deflcction of light is ihat of Yallop (1984) as givcn in ?'€
the relative rnotions and distances, sometirnes has t:orlplcx effects q4rere trvo or Astronomical Almanac 1981, whlch is an a<laptation of Murray's (1981) formulas'
rnorc bodies arc involvd, as, e.g., in eclipscs, lrarrsits, lrrd the phenomena of satel- The isotropic metric has been assumed. Only the Sun's gravitational ficld has been
lite systerns; a,rrd orr some pilst occasions, tho detcrtniniltiotr of lhese effects has included; each of the planets causes a similar effect ihat is smallcr by a factor
The
presentcd an iutricate problem. equal to the ratio of thc planct's mass to that of the Sun (1/1047 for Jupiter)
 136 EXPLANATOBY
SUPPLEMENT SYSTEMS
REFERENCE
3 / CELESTIAL 137
since (q ^ e) ^ p / sin ry' is a unit vector in the plane PES, at right angles to p and
pointing away from the Sun. This equation may be written in the form that is more
useful for comPutation'

pr = p+ * t(p.q)e- (e.p)ql. (3.26-3)

The dimensionless scalar quantities 8l and 82 are

2u (3.2tu)
gt= b and 82=l+q e,

where p = 1.32'112438 x 1020m3s-2, c = 864OO/ ra = l73.1M633 xu/day, and

'rA = 499.W4782 is the light-time for 1eu in seconds, based on the IAU (1976)

Figure3.26.1 system of constants. The value of 8l is always close to 2 x -10-8, but 82 varies
Gravitational
light deflecrion between 0 and +2. The vector pl is a unit vector lo order p I c'.
Stars represent the asymptotic case when p = q, and Equation 3.26 3 becomes

e-(p e)p
P r= P + 8 r (3.26,5)
l-p*
gravitational field of the Earth, also ignored here, can deflect light by a few tenths
of a milliarcsecond for ground-based observers. It may also have to be allowed for When one is applying the light deflection correction in the algorithm for appar-
in precise astrometry with Earth satellites. ent places (see Section 3.316), using the vector Es in the relativistic deflection com-
In Figure 3.26.1,S is the Sun, P the body that is being observed,and E is the putation introduces a minor approximation resulting from the use of the barycentric
Earth. The unit vectors e and q represent the heliocentric directions of the Earth position of the Sun at the epoch of observation. The resulting error cannot exceed
and the body, respectively. The heliocentric elongation of the Earth from p is r,l 0. l mas in the worst case (object observed at the limb of the Sun with barycen-
where cosry' = q. e. The geocentric direction to the body p when the light left it tric motion of the Sun orthogonal to the line of sight) and is generally much less
is given by the unit vector p. The dotted arc AEB represerrts the light-path as it Furthermore, the deflection algorithm itself results from a first-order development
passesthe Earth. The tangent to the light-path at EisXEy. As the light, which that assumessmall deviations of the photon track from a straight line in Euclidian
was emitted at P, travels along the path AB, it is always deflectecl tov,ard the Sun. space; the error in neglecting second-order effects can reach about 0 5 mas for an
At E the direction between p and the tangent to the light-path is Ad (as shorvn irr object observed at the Sun's limb (Kammeyer, 1988, private communication)'
the figure). Einstein's general relativity theory predicts that Figure 3.26.2 shows the magnitude of the deflection of light, as viewed frotn the
Earth, for planets and stars as a function of the geocentric angular separation of the
2p sin dr observed body from the center of the Sun. Maximum deflection occurs for bodies
AQ = --;=: ------- , (3.26-t) that are about to be occulted by the Sun. Minimum deflection for Mercury and
c't I +cosi,
Venus occurs when thev are about to transit the Sun. For the other bodies' minimum
where E is the distance of the Earth from the Sun, p is the heliocentric gravitational deflection occurs at 180o elongation from the Sun. The figure was produced fronl
constant, and c is the speed of light. Equation 3.26 1 assuming circular orbits.
The apparent direction of P is along the tangent to the light path p1, and by Table 3.26.1 tabulates the deflection angles A{ = g1 tan(l/2) (Equation 3'26- 1)
vector addition for various values of D = 180" - ty' (for stars, D approximates the geocentric elonga-
tion). The body disappears behind the Sun when D is less than the limiting grazing
p ,= p + 9 1 j T I a d , (3.26-2) value of about 0925.
slnp
 138 EXPLANATOBYSUPPLEI\,,IENI
3 / CELESTIALREFERENCESYSTEIUS 139
o?27 3.261 Approximat€ Light Deflection for Stars The increments to be added to the
45' 90" 180' calculated right ascension and declination of the star may be evaluated app1e11i-
matelY from
10100
cosD = sin 6 sin 6. + cos 6 cos d. cos (o - ar),
l:'00
cos 6. sin (a - a.)
Aa = 0i000271 '
0:'1 (l cosD) cosd
sin 5 cos5. cost o - o.) - cosd sin d,
46 = 01'00407 , (3.26r-t)
I - cosD
0:'001
where c!, d, os, 6s are the geocentric right ascensionsand declinations of the star
l" x 10-a and the Sun, respectively. For stars, D, the geocentric elongation. approximates
l80o - {, where I is the heliocentric elongation. These corrections mav be included
l"xl0 5 with the day-number reduction for stars given in Section 3.341.

0:l t' 10" 100'

Geocentric
elongation 3.27 PolarMotion
The rotation of the Earth is represented bv a diurnal rotation around a reference
Figure3.26.2 axis whose rnotion with respcct to the inertial reference frame is represented by the
Light from theplanetsandstarsdeflectedby thesun theories of precession aird nutation. The reference axis docs not coincide *'ith the
axis of figure (maxirnum nroment of inertia) of the Earth, but uroves slorl.ly (in a
The table u'as produced assurning that thc Earth Sun distance E = I, aucl terrestrial reference frarne) in a quasi-circular path around it. The rcfercncc axis
thus 91 = 0l'00407. The variatiol of the Earth Sun distance modulates the va]ue is the Celestial Ephemeris Pole (normal to the true equator), and its motion &'ith
of 91 by less t\an 27e,.This sitnplc form of the gravitational deflection forrnula respect to the terrestrial referencc frane is known as polo,r motion. The Celestial
has been noted previously by Shapiro (1967) and by Fukushima (1982, private Ephemeris Pole is the axis about, which the diurnal rotation of the Earth is applir:d
communication). In this forur. tlie deflection is not explicitly dependcnt on thc in the transformation between celestial and terrestrial frames. and so must not be
distance of the emitting body lrom the Sun or the Earth. Thereforc. to an obsen'er confused with the instantaneous a,ris of rotation. The maxinum amplitude of the
polar motion is typically about 013 (corresponding to a displacemcrrt of about 9 m
on the Earth, the apparent gravitational deflection is the samc for all objccts that
lie anywhere on a given Iine that extends radially outivard from the Sun. This result on the surface of the Earth) and the principal periods arc about 365 and 428 days.
holds regardlessof the orientation of the line n'ilh respect to the Earth. The motion is afiected by unpredictable geophvsical forces, and is determined from
observationsof stars, radio sources,the Moon, and appropriate Earth stilellites,
Thc equations for Iight deflection have arr indeterminacy for light paths startirrg
using relevant techniqucs including VLBI and laser ranging.
beyond the Sun on thc extension of the Sun Earth linel bodies there ar.ehidden Lr1
the Suu's disk and urrobservable in any cvent. For these bodies or the Suu itsclf, The pole ancl zero (Greenwich) rneridian of the terrestrial referencc frame are
defined implicitlv by the adoption of a set of coordinates for the instruments that are
thc deflectiorr can bc considered to bc zero.
used to determine UT and polar motion from astronomical observations. (The pole
of this systcm is known as the Conventional International Origin.) The position of
Table3.26.1 this terrestrial reference fraurc r-ith respect to the true cquator and equinox of date
ApparentDeflection
Angles is defined by strccessiverotations through tn'o small angles x, y, and the Greenwi<:h
apparent sidereal tirnc d. The angles r and I correspond to the coordinates of the
D 0925 095 2.5"'10"20"50.90" Celestial Ephemeris Pole with respect to thc terrestlial pole measured along the
ad 11866 0t'933 0:466 0:233 0:',093 0:047 01023 0t009 01004
meridians at lorrgitude 0' and 270' (90' west). Current values are published by the
International Earth Rotation Service; values from 1970 January 1 onward are given
140 EXPLANATORYSUPPLEMENT 3 / CELESTIAB
L E F E B E N C ES Y S T E I \ , 4 S 141

in section K oI The Astronomical Almanac. An 80-yearJong series of values on a temperature, pressure, and humidity and even temperature lapse rate, gives a,well-
consistent basis has been published by the International Polar Motion Service. The defined reference vahre. Thus, for example, residuals of the observed values from
coordinates x and y are usually measured in seconds of arc. the reference values could be used to model the local variation of refraction.
Polar motion causes variations in the zenith distance and azimuth of the Ce-
Iestial Ephemeris Pole and hence in the values of the terrestrial latitude (/) and
longitude ()) that are determined from direct astronomical observations of latitude
and time.
The rigorous transformation of a vector r0 with respect to the frame of the 3.281 Refraction-Numerical Int€gration In this section we give a precise algo,
rithm for calculating refraction using numerical quadrature. The real accuracy is
true equator and equinox of date to the corresponding vector r with respect to the
very dependent on the conditions, particularly for large zenith distances, and on
terrestrial frame is given by the formula
the atmospheric model.
The atmosphere is assumed to be spherically symmetric and in hydrostatic
r = Rz(--r)Rr(-y) Rs(d)ro; (3.27-r)
equilibriun.r, and to obey the perfect gas law for the combined mixtrrre of dry air
conversely, a,nd water vapor, and also for the dry air and watcr vapor separately. The two
layers of the atmosphere are the troposphere and the stratosphere. The troposphere
ro = R:( d) Rr0) Rz(x)r, (3.27-2)
extends from the surface of the Earth to the tropopause, which is assumed to be
11 km (ft,) from the Earth's surface. In this region the temperature decreasesat a
where Equation 3.27 1 represents a rotation of 0 about the z-axis, followed by a constant rate (a = 0.0065" km
r) and the relative humidity is constant and equal
rotation of -y about the x-axis, and finally a rotation of -l about the y-axis (see to its value at the observer. In the stratosphere the temperature remains constant
Section 11.4). The vector r could represent, for example, the coordinates of a point and equal to the temperature at the tropopause (7,) and there is no pressure due
on the Earth's surface or of a satellite in orbit around the Earth. to water vapor. The upper height (r,) of the stratosphere is taken as 80km, sirrce
Alternatively the transformation can be expressed as a variation of the longi- above this height refractiort becomes negligible.
tude a"ndlatitude of a point, and to first order, the departures from the mean values The total bending of a ray is given by
,\-, S. (corresponding to vector r0) are given by

- l'^ rdn ldr

Ad=-xcos.\.-ysin.\. and A . \ = ( . xs i n) . + y c o s ) . ) t a n d m . (.327-3) t=
' t --------------:- atz. (3.281-l)
Jo n+rdnldr

The variation in longitude must be taken into account in the determination of

This integral is a transformation of the usual refraction integral. It has been rec-
Greenwich mean sidereal time, and hence of UT, from observations.
ommended by Auer and Standish (1979) becausc it is more suitable for numerical
quadrature; it is a more slowly varying function over the whole range of ; and re-
moves the problem at z = 90'. Howevcr, there is a discontinuity in the function
3.28 Refraction dn I dr aL the tropopause, and so the iutegral is evaluated in two parts, from ; = ;"
to ;," in the stratosphere, and from z= e t,o zo in the troposphere.
Atmospheric refraction is included in lhe The Astronomical Alm,onac in only a few
The following procedurc gives the steps required to calculate the total refrac-
topocentric phenomena, such as the times of rising and setting of the Sun and tion, (, i.e., the sum of the refraction caused by the troposphere ((1) and the strato-
Moon, and in theory in the predictions of local circumstances of eclipses. How- sphere ({"), from
ever, for observational reductions the effect of refraction must be included. In this
section we give a low-precisiou formula, valid at the horizon, that is particularly
useful for navigation; a formula valid for zenith distances less than 70o, which takes
=l'' ru,*1",',,'
(=€,+€, ro, (3.28r,2)
into account the amount of water vapor in the atmosphere; and an algorithm for
evaluating the integral in the classical formulation using quadrature. This latter
algorithm, which allows for the principal change of refraction due to variations of to a specified tolerance e, where / is the integrand given in Equation 3.281 1.
 142 EXPLANATORY
SUPPLEMENT 3 / CELESTIAL
REFERENCE
SYSTEMS 143
Step 1: The parameters of the Earth aad the atmospheric model, wh.ich a.re as- In the troposphere at distance r 1! 4 from the center of the Earth, the temper_
sumedto be constantare ature f, the refractive index a. and dn / dr ate calculated from

R = 8 3 1 4 . 3 6 M a= 2 8 . 9 6 6 M * = 1 8 . 0 1 6 6 = 1 8 . 3 6 T=To-a(r-rr),

re = 6378120 ir=!1000 ft" = 80000 a=0.0065 (3.231_3) -,(+)'-'l+,

,=,.f..(^r)"-
Step 2: The parametersof observation*the initial conditions.The observedzenith
dista.nce,zs, of an object with wavelength of ) pm is recorded by an observer at
latitude {, height lro meters abovethe geoid i.e., rn = r" I 114
meters from the cen_
#=-,(#)'-'.",(;)'-' (3.2814)

ter of the Earth. The meteorologicalconditionsat the time of observationare the

In the stratosphere at a distance r where 4 ( r ( r, and the temperature is givel
temperature S K, the pressureP0mb a.ndthe relative hurnidity R;.
by T1= T6 - a(rt - ro) the refra.ctive index n and, dn I dr are calculated from

Step 3: Calculate the para.rnetersdependent on the initial conditions and the I r-L\
n = | + ( n t_ 1 l e x p (_ r r i
model of the atmosphere. Those denoted by C; a.re needed at each step of the ),
integration.

P,a = Rn(To
I 247.1)6,
#=-7''- Itexp(-
"+)
(3.281-:7)

E = 9.78a0 - 0. 0026cos2/ - 0. 00000028to), The integrand is

dnf rdnldr
/ -| ' :4',r88
.1 0.0t36\ 273.15 I" ll r , n ,q=r ll = - - . (3.281-8)
A = l 2 8 7 . 6 0 4+ L n + rantdr
\ ^2-* x )tblJfjxr0-"'
Ct=a, C2=lMa/R, C3=C2/Ct="/, Step 5: Calculate the following para.meterswhich are required for the limits ofthe
Ct=6, Cs=P,o(l - MatM*).tt6-i, two integrals. At the observer, the observed zenith dista.nce is z6; set r0 = r" +r, and
cala:Jate n6, dnnldrn /6. At the tropopause, in the troposphere calculate rt = re*hrt
Cs=A(po+C)/To, C7 = (AC5+ 11.2694x 10-6 p,{-)/ To, th, dnt / dh, fl, and using Snell's law
Ca = a(t - l)Cs / To, Cs = e(6 - l)Ct I To. e.2814) norosrnzo
Zr= sin-l (3.28r-9)
lltft

Step 4: Use the following expressions to calculate r, n, and. At the tropopause, in the stratosphere ns = rt, and ca.lculate n6, dn6 | dr6, f6, and
dn / dr, which depend
on the atmosphere, and thus evaluate the integrand ?a. At the limit of the stratosphere calculate rs = r. + hs, n", dn" I dr", f", utd 2".
f at each step in z along the
path of the ray.
Step 6: Integrate the function over the required interral usirg numerical quadra-
The value of r corresponding to the curent step in zenith distance ture, e.g., Simpsorr's rule, by forming the suumations over equal steps of z. In the
_ z is found
by solving Snell's law zrsinz = norosin zo using Newton-Raphson troposphere S = (zt - zd | 16 is a convenient step length to start with. Repeat the
iteration thus:
integration, halving the step size each time, until there is no significa.nt difierence
niri -
,,*r=rr-f
nitridnildri
=l ;
',o/o sin z0 / sin , 1
for i = 1,2,..., (3.281-5)
between two consecutive ralues and the tolerance permitted (i.e., 0.5e). Repeat this
step for the stratosphere. Thus the total amount of refrarction will be = +
{ $ $,

where r1 is the value of r calculated at the previous step of the 3.282 Saastsmoinen'sReftaction Formula
integration. Con_ This formula, rralid for zenith dista.nce.s
vergence is rapid, so the initial estimate rl is trot critical. Four iterations to 70o, was devised by Saastamoinen (1972), a.nd it is equivalent to the re.
should be
su.ftcient.
tables in ?he Star Alrnanac. Given the observed zenith distance, zs, the
144 EXPLANATORYSUPPLEMENT
T 3 / CELESTIALREFERENCESYSTEMS
'145

parallactic angle -i'e ' t'he angle at thc object

temperature, Pressure' and partial pressure of water lapor' (Zo,Po,P'o)' the refrac- where R is the refraction and C is the
Thus when the object is on the meridian
tion for a wavelength of 0.574pm. ie., visible light, and for an observer at sea in the spherical triangle pole-object-zenith
R, is a direct correction to declination onlv'
level is f= 0",^una the amount of refraction,
but the right ascension is unafiected'
and pressure are printed in
t = t6'2't t Q tanzo(l + 0. 00003940 tan2zo) Refraction tables based on a st,andard tempelatule
(1985)' Pulkova (1956)' together with
r"arious publications, for example, Abalakin
- 0'i0000749Po(tanzo+ tan3uo), and pressures The standard conditions
correcti;n tables for other temperatures
andP= 1010mb; in Tlrc Star Almartac
w h e r e Q = ( P e - 0 . l 5 6 P ' 4 )/ ? 0 . (3.282-1) usedin ?he Nautical AlmanacareI= lO'C
=
f =7'C and,the Pressure P 1005mb'
A formula to calculate P,x in tenns of the relative humidity is given at the start of
Step 3 in Section 3.281. Saastamoinen gives correction tables for other wavelengths PLACEALGORITHMS
ANDTOPOCENTRIC
3.3 APPARENT
and heights above sea level.
calculating apparent and topocentric
This section presents a set of algorithms for
3,283 Low-precision Refraction Topocentric phenomena such as the times of ris- given in a steprvise forrn using matrix
places of planets and stars. The methods are
ing and setting, and navigational reductions require tltc amount of rcfraction ac programmcd usillg a set of subroutines'
.nd u""to, operations which carr then be
curate to 0i5. The following formula is dependent on the apparent altitude. IJ: of the reduction Frrll details on the
each of which handles one particular aspect
i.e., the observed altitude corrected for thc index crror of the sextant and dip (see part of Section 3 2 The algorithm
various topics are discussed in the appropriate
Section 9.331), and on the temperature and pressure.and gives the amottrtt of solar-system bodies is given first'
for calculating apparent places of planets and
procedures lol calctllating apparent
refraction, R, in degrees; because it is the nore comprehensivc The
places (Scctiort3 35)' and differential
places (Sections3.31 ancl 3.32), topoccntric
/ 0 _ 2 8 P\| 0?0167 (1989)'
p- | _ (3.283-l) astrometry (Section 3-4) are bascd on Kaplan et ol
" - \ r + z t t ) t a n ( H+ ' 1 . 3 1l ( H + 4 . 4 ) ) ' srrbscripl' B refers to the
The following notation is uscd in the algorithms The
root of the sum of tlte
solar-system barlcenter. a = lul rneans calculate the squarc
where I,P are the temperature in <legreesCelsius and the pressure in millibars. product' i e ' the sum of
squares of the coDrponents, and (u r) iDdicates t'he scalar
If 7 and P are unknown, assume thc terrn in the first bracket is unity The elTect
the products of their corresponding componerlts
of refraction is removed frorn an observation by subtracting R from the obselved
altitude. When only the truc altitude is known, it is possibleto iterate using thc
particttlarll
formula above,iu order to find the observedaltitude. This is necessary', 3.31 Aoparent-Place for Planets
Algorithm
at lo\,,,'altitudes, when the refraction correction is large. a Planel or otlter solar-
The algorithm used to compute thc apparc[t place of
Alternatively, at low altitudes, when H < 15', the followirrgmay be used: rvith origin at the
system"body at an epoch of observation t/' given its cphenreris
to thc Earth's rneatr equalor-
soltr-"y"t"m barYcerrter zrnd its coordittates rcferred
0:1594+ 0.0196H+ 0. 00002H2 as
^=(;*) (3.2832) ur1d.qritt.r* of a Ieferertce epoch, t0' can be succinctly teptesenterJ
I + 0. 505H+ 0. 0845F

For rising and sctting phenomenaa constant of 34' is used fbr flc ltorizr'utaL
= N(/)P(t)fts[uB(r- ') - Es(t)]1,
u4(r')
( 3 . 31 - l )

refraction; thus for the center of the Sun or N'Ioon to be coincidcnt rvith the horizon
the true altitude will be 34'. where
The observed right ascensionand declinatiort, (o, 6), of an objcct mav be // is the cpoch of observation'in thc TDT timescalc:
corrected, to first order, to givc the true position al , 61' usirrg the following forDrulas: t is the cpoch of observation, in the TDB tirnescale;
referred' e'g ' J2000 0i
/s is the reference epodl to rvhich thc epherneris is
al =o in tbe TDB timescale'
Rsec6sinC T is the light travel tirne from the planet to thc Earth'
dr=6-RcosC, (3.283-3) for light arriving at the epoch of observation t;
 146 EXPLANATOBY
SUPPLEMENT REFERENCE
3 / CELESTIAL SYSTEMS 147
uB(r - r) is the barycentric position of the planet at epoch / - r, referred to the Step e: Compute T, the number of Julian centuries in the TDB timescale elapsed
mean equator and equinox of t6; since J2000.0 TDB (JD 2451545.0 TDB)
Er(t) is the barycentric position of the Earth at the epoch of observation r.
referred to the mean equator and equinox of t6;
8t...1 is the function representing the gravitational deflection of light; T = (t - 2451545.0\ / 36525. t3.31r-4)
/t...1 is the function representing the aberration of light;
P(t) is the precession matrix, a rotation from the mean equinox and equator
of to to the epoch of observation !; In the expression for s, the lunar and planetary terms of order l0-5 secondshave
N0) is the nutation matrix, a rotation from the mean equinox and equator been ignored. Furthermore, the expression for m (Step c) strictly requires a time
of date to the true equinox and equator of date r; argument in the TDB, not the TDT, timescale. See Moyer (1981) for a complete
u4(/) is the apparent geocentric place ofthe planet at the epoch of observation discussion.
/' represented as a three-dimensional position vector (.r,y, z) and referred However, the algorithm given above is much more precise than is required for
to the true equinox and equator of observation. the computation of apparent places of stars and most solar-system bodies. For
stellarappaxentplaces,sets=0,t=/,andT=T'withnegligibleerror.Forsolar-
This expression is schematic; the full functional forms of / and g, the elements
system bodies, the same approximation can be used for all bodies except the Moon
of the P and N matrices, and other auxilia.ry calculations are not indicated.
and close-approachlng comets and asteroids, where the error in using the t = tl
approximation may approach 0'1001in very unfavorable circumstances.

3.311 Relevant Time Arguments

Step a: Express the epoch of observation / as a TDT Julian date. 3312 Ephemeris Data for the Earth and Sun

Step br Compute 7, the number of Julian centuries in the TDT timescale from Step f: Extract from the ephemeris, for the time t, the barycentric position and
J2000.0TDT (JD 2451s45.0)TDT. velocity of the Earth, Eg(t) and lis14, and the barycentric position of the Sun,
Ss(r), referred to the Earth's mean equator and equinox of the reference epoch /6.
T'=(/ - 24srs4s.0)
/ 36525 ( 3 . 3 1 ll ) Also form the heliocentric position of the Earth Es(t) = Eu11)- Ss(/). Note that the
barycentric position of the Earth is that of the center of mass of the Earth, and
not that of the Eaxth Moon barycenter. All position vectors are in AU and velocity
Step c: Compute the mean anomaly rn of the Earth in its orbit, in radians, at the vectors in eu/day.
epoch of observation
The standard ephemeris of the major bodies in the solar system is the Jet
Propulsion Laboratory ephemeris designated DE200 (Standish 1982a). The posi-
m = (357.528 + 35999.O5O1)x 2r I 36O. (3.3rr-2) tions of the Sun, Moon, and planets given in The Astronomical Almanac and other
international almanacs are now obtained from this ephemeris. Values of the com-
ponents of the vectors E3, Es, and -EH at 1-day intervals are tabulated in Tft.e
Step d: Compute s, the diflerence, in seconds,between the clock reading in the two Astronomical Almanac, A set of analytical planetary theories fitted to DE200 has
timescales (in the sense TDB - TDT), and r, the TDB Julian date coresponding been developed by Bretagnon (1982).
to the epoch of observation The barycentric position of the Earth is used to form the geocentric position
of the body (see Sections 3.314 and 3.315); the barycentric velocity of the Earth is
s = 0.001658
sin(rn+ 0. 01671sinz), used in the aberration computation (Section 3.317); and the heliocentric position
of the Earth is used in the computation of the relativistic gravitational deflection
t=t'+s1864N. (3.31l-3) of light (Section 3.316).
 148 '|
EXPLANATORY
SUPPLEMENT 3 / CELESTIALREFERENCESYSTEMS 49
3.313 EphemerisDats for the Planet

Step g: Extract from the ephemeris, for the time r, the barycentric position of the
planet, uB(t), referred to the Earth,s mean equator and equinox of the reference cr'= Ll+l2tttc2\1"
,E-ffi.
(3.31s-3)
epoch J2000.0.

3.314 G€ometric Distance Between Earth and planet where U = lfll. E = lEHl. Q = lQ, p is the heliocentricgravitational constaDt.rDd
c is the speedof light in eu/day (Equation 3.31'l 2). Using the IAU 1976 svstem.
Step h: Compute d, the geometric distance between the positions of the center
of u = 1.32712438x l02omls-2.
mass of the planet and the Earth at time /, in .ru, from
Compare r' t'ith z: if they are identical tithin some small tolerance,corttinue
d=lus(r)-EsOl. (3.3l4- I ) to Step l. If thel are not. then replacethe value of r *'ith the value of r' and repeat
Steps i through k until the light-time convergesto u'ithin the tolerancepermitted
The geometric distance is the quantity tabulated in The Astronomico,lAlmawlc Since the speed of bodies in the solar system is small compared to the specd of
as
the "true distance" of solar-system bodies. light, this process converges rapidly.
The tolerance permitted depends on the precisiondesired in the final coordi-
Using d compute r, a first approximation to the light-travel time between nates and the apparent angular speed of the bodl' as vieu'ed lron Earth. The most
the
planet and the Earth, as rapidly moving objects in the sky are the Nloon (angular rate = 0'15per second).
Mercury (angular rate at inferior conjunction 0105per second), and the SuIr (an-
/ I 11.1-l r gular rate 0104per second). Hou'ever, occasionalll' an Earth-crossing asteroid or
comet may exceed these rates for short periods of time. For a conlputational preci-
where c is the speed of light expressedin eu/day; its precise value sion of one milliarcsecond. therefore. the light-time convergeDcetolerance must be
may be computecl
from 86400/r,a,where 2,4is the light-time for unit distance (1eu)
in seconds.In thc f t 0 0 2 = 2 x l 0 8 d a j s u r l e s . :u - es t l g B c sl rx l 0 - 8 d a l s .
IAU (1976) system c= 86400/499.004782= 173.144633 Kaplan el al. (1989)ignorethe secondterm in Equation 3.315 3, (order 1 nicro-
^\r /day.
arcsecond).which is due to the relativistic delal causedb1'the Sun's gravitational
field.
3.315 Geocentric Position of planet, Accounting for
Light-Time

S.tepit Extract from the epherneris, for the time - Step l: Once r has converged,set U = us(, r)-Es(r) and Q = utlt 7)-SB(/-7),
/ r, the barycentric position of
the planet and the Sun, uB(r - 7) and Sg(r _ r), respectively. the geocentric and heliocentric positions of the body at time t - 7

Step j: Calculate U and

Q, approximations to the geocentric and heliocentric po_ 3.316 Relativistic D€flection of Light
sition of the center of mass of the planet, respective\r,
at the epoch of observation
t, from
Step m: Form the following unit vectors and dimensionless scalar: quantities

U=uB(r-r)-Es(r), (3.31s,1)
u = U/ lul, q=Q/lQl, e=EH/lEHl,
2! ( 3 . 3 1 6 -)l
Q=us(r-r)-Ss(r- r). (3.315-2) e).
s,=;;. 82=l+(q

Step k: Next compute r/, a better approximation to the light-travel

time between where the constants p and c are given in Steps k and h. A detailed explanation on
the planet and the Earth, as
light deflection is given in Section 3.26.
 \r
15O EXPLANATORY
SUPPLEM€NT 3 / CELESTIAL
REFERENCE
SYSTEMS 151
Step n: The deflected position ofthe body relative to the geocentric inertial frame
Step r: Transform the coordinates from the fixed reference epoch 16= J2999. g 1o
that is instantaneously stationaxy in the space-time reference frame of the solar
the epoch of observation r, by applying the precession matrix P to the vector u2:
system (the natural frame) is theu

u ' = l u l ( o * 4 t ( u .4 e - ( e .u ) q l ) . u: = Puz. (3.318-l)

(3.316-2)

The vector u1/ lUl is a unit vector to order u t *.

3319 Nutation
For the Sun or bodieslying behind the sun on the sun Earth line, the light deflec-
tion ca.nbe consideredzero, so that ur = U. Step s: Eraluate the mea.nobliquity of the ecliptic, €0, the two fundamental nuta-
tion angles, Ary', Ae and the true obliquity of the ecliptic € = €0+ .4'6,for the epoch
3317 Aberration of Light of observation /, and calculate the associated nutation matrix N. Full details a"re
given in Section 3.222.
Step o: Form the following quantities:
Step t: Transform the coordinate system to that defined by the Earth's true equa-
Y=EsOlc, tor a.nd equinox at the epoch of observation r, by applying the nutation matrix N
P=ur/lurl,
lO VeCiOr U3:

B-t=1/t-1t2, .fr=(p.V), fz=t+ft/0+p-t), (3.317-l) u += N u : , (3.319-l)

where the velocity (V) is expressed in units of the velocity
of light a.rrd is equal to
the Earth's velocity in the barycentric frame to order V2. where u4 is the apparent geocentricposition vector of the planet at the epoch of
observation t/ .
step p: The aberrated position ofthe body in the geocentric
inertial frame that is
moving with instantaneous velocity (v) of the Earth relative The combined precession-nutation matrices given in The Astronomical Almanac
to the natural frame
is then given by the vector (R = NP) may be used in place of Steps q through t above; thus u+ = Ruz, where
the ralues of the elements of R must be interpolated to the epoch of observation,
tz = (B-tur+/2|u,lv) / (l + /r). (3.317-2)

The above algorithm includes relativistic terms (Murray, 33110 The Apparent Position in Spherical Coordinates
lggl) (see Section 3.252),
which a,re of the order 1 milliarcsecond. Therefore,
fo. -urry applications one may
use the much simpler classical formula Step u: Compute the object's apparent geocentric right ascension, o, and decli-
nation, d, at the epoch of observation t/, using the three components of the vector
u2= ur + lullV. U4, i.e., .r, y, and z:
(3.317-3)

o=tan-r1 and 6 = tan-' (3.311O-l)

3318.Prccession

Step q: Elaluate the three fundamental precession angles (,a,

zA, and gA using Note: ua is not a unit vector. Most computers have a double argument function
the equations given in Section 9.211 lvith epoch e6 = J2000.0
and ep = r and the for inverse tangent (e.g., ATAN2) which will provide the correct quadrant if the
associated matrix P (Equation 3.21_3).
and denominator are entered as separated arguments.
152 EXPLANATORY
SUPPLEMENT
T 3 / CELESTIALREFERENCESYSTEMS 153

Algorithm
3.32 Apparent-Place for Stars Julian centuries in the TDB timescale elapsed from the reference epoch J2000.0
TDB. For starsJskip Step c and simply sel t = l/
The algorithm used to compute the apparent placc of a star at an epoch of ob-
servation t/, given its mean place, proper motion, and other data (as available) at 3.322 Ephemeris Data for the Earth and Sun
reference epoch to, can be succinctly represented as
Step 2: Follow Step f in Section 3.312 and extract from an ephenreris, for time l,
+ (, - /0)uB(t0)
ur(r') = N(t)P(r)"ftsluB(r0) - EB(r)ll, (3.32-
I) EB(t), [lB(t), and Se(r), the barycentric position and velocity of the Earth' and the
-
barycentric position of the Sun. Also, form EH(l) = En(t) Sg(t)' the heliocentric
where position of the Earth. All position vectors are in eu and velocity vectols in eu/day.
r/ is the epoch of observation, in the TDT timescale;
/ is the epoch of observation, in the TDB timescale; For apparent places of stars accurate to a few milliarcsecond it is necessary
ro is the referenceepoch and equinox, e.g., J2000.0,of the sta,rcatalog, irr only to obtain the Earth and Sun's position components to three significant dig-
the TDB timescalei its and the Earth's velocity components to five significant digits lt is therefore
uB(r0) is the catalog mean place of the star at the referenceepoch 10,represented feasible to construct relatively compact closed-form algorithms which provide the
as a three-dimensional position vector in AU, with origin, solar-systcm required data (Stumpff, 1980b) The U.S. Naval Observatory has developed com-
barycenter, and coordinates referred to thc rnean equator and equinox puter subroutines, based on a truncated and modified form of Newcourb's theory'
ol 16; which evaluates the barycentric position and velocity of the Earth, without using
Og(to) is thc spacemotion r€ctor (in ,lu/day) of the star at the referenceepoch external files. Howevet, for the highest precision, or s'hen the apparent places of
/0, obtained from the catalog proper motions. parallax, and radial veloc- planets are being computed more complex algorithms or external files arc re<luired
itv;
Es is the barycentric position of thc Earth at the epoch of obser\ation 1.
referred to the mean equator and cquinox of t9; 3.323 Star's Position and Space Motion V€ctors at the Catalog Epoch
gt...l is the furction representingthe gravitatioual deflection of light;
ft.. .1 is thc function representingthe abeu'ation of light; Step 3: Given are o and 6, the catalog mean barycentric right ascensionand decli-
P(/) is the precession matrix, a rotation from thc nea.n equinox and erluator nation of the star al to = 2451545,O.the reference epoch J2000 0, and p,' and p6 the
of te to the epoch of observation li corresponding proper motion componcrrts in seconds of time arrd alc, rcspcctivcly,
N(1) is the nutation matrix, a rotatiolr fron) thc mean equinox and cquator- per Julian century (of TDB). Also given are p, the parallax of the star in scc'rnds
of date to the true equinox arrd cquator of date t; of arc, and i, its radial velocity in km/s.
u.r(t') is the apparent geocentric place of thc st.u at the epoch of observatiorL
// represented as a three-dimensional position vector (x, y,.z) and refcrretl Form the barvcentric positiorr vector, uB(t0) (in AU), and spacelnotidr vector,
to the true equinox and equator of obserlatiorr. ns(rs) (eu/day), of the star at the catalog epoch t0, refeued to the Eartlt's nteart
equator and eqrrinoxof the catalog epoch using Equations 3.23 1 antl 3 23 3 in
This expressionis similar to Equation 3.31 1 for pliurets. \,Iost of the algorithnr
Section 3.23.
is identical; the difference is due onl1. to the mole cornplex rnotion of a plauet
compaled to that of a star. In SectioD3.31 r,i'ehacl to obtain the position of the
planet from an ephernerisand deal rvith the light-tirnc ploblern explicitll': in this
3.32 Star's Position Vector at the Epoch of Obs€rvation
scction r'"'eu'ill rxsurne uniform rectilinear rnotion for the star and neglect lariatious
in light-tinre as the star moves. Step 4: Compute the barycentric position of the star. at the cpoch of olrscrvatiolt'
by adding the distancemoved during the intcrval of time betweentlre cal alog ep'Jch
3.321 Relevant Time Argum€nts
and the eooch of observation. as

Step l: Follow Steps a through e in Section 3.311. This gives the value of t, the
uB(/) = uB(to)+ (/ - r0)uB(t0). (3.324-t)
TDB .Iulian date corresponding to the epot:h of observation, and ? the number of
 154 EXPLANATORYSUPPLEMENT 3 / CELESTIALREFERENCESYSTEMS tcc

The assumption is made that there is no explicit correction for light-time in stellar for the Earth, and the precession and nutation matrices can be saved and re-usecl.
apparent-place computations (see Section 3.23). For example, when there are a large number of apparent places to be computed,
it is more efficient to compute the places of all bodies at a given observing epoch
3.325 Geocentric Position ofth€ Star before moving to a new observing epoch. The major computational burden in the
algorithm involves the retrieval of data from the planetary ephemeris in Steps f
Step 5: Form the vector U, which represents the geocentric position of the star at through I and the evaluation of the two nutation angles Al and Ae in Step s.
the epoch of observation, l. If computational time becomes critical (such as within telescope-coltrol sys-
U = us(r) - Es(t). (3.325-l) tems or in microcomputer implementations), the method of obtaining the epheur-
eris and nutation data must be considered. Self-contained algorithms can often be
This step introduces annual parallax, and is a shift of the origin from the solar simplified by truncating small terms from the series, if high precision in the final
system barycenter to the Earth's center of mass. apparent place is not required. Consideration should also be given to precomputing
the required data for fixed intervals and storing the data in an external file that can
3,326 R€latiyistic Deflection of Light be efficiently accessedand interpolated. Planetary ephemeris data are frequentlv
distributed in this form anyway. Precomputing and storing the elements of the com-
Step 6; Follow Steps m through n in Section 3.316, with Q = us(t) Ss(r) th<: bined precession-nutation matrix (seeend of Step t) is also feasible; seeSection B of
heliocentric position of the star, to obtain the geocentricposition of the star ut, The Astronomico,l Almanac. The convenience of self contained aigorithrns must be
corrected for relativistic light deflection. weighed against the number of calculations rcquired. the accuracy of the algorithm,
and the accuracy needed for the final result.
3.327 Aberration, Precessionand Nutation
3.34 ApparenfPlaces-Day-NumberTechnique
Step 7: Follow Steps o and p in Section 3.317.
The 1976 IAU resohrtions on the astronomical reference frame included two state-
Step 8: Follow Steps q and r in Section 3.318. ments about the procedures for computing apparent places and the reduction of
observations. First, reductions to an apparent place shail be computed rigorously
Step 9: Follow Steps s and t in Section 3.319. and directlv without the intermediary of the mean place for the beginning of the
yea,r whenever high precision is required. Second, stellar aberration shall be com-
3.328 Star's Apparent Position in Spherical Coordinates puted for the total r''elocity of the Earth referrcd to the barycenter of the solar
system and the mean places shall not contain E-tcrurs, Implicit in that recommcl-
Step 10: Follow Step u in Section 3.3110. dation is the fact that, when accurate results are desired, dav mrmbcrs should not
be used. Thus, this section on day numbers is inchrded as a means to help unravcl
what may have been done to observaliorrs in the past and to understand the proce-
3.33 TheComputer
lmplementation
of Apparent-Place
Algorithms dures that were followed in the past for historical purposesl and for thosc who can
The development of the apparent-place algorithms using matrix/vector notation is be satisfied with a reduced lcvel of accuracy, e.g., for making rapid calculations ol
rigorous, and allows the planet and star algorithms to use those parts u'hich are star positions il real time for poitrting telescopes.
common to both. It also allows users to tailor their program to their own necds, as Before we explain thc rnethods of day numbers, an additional warning concern
well as the testing of alternative, or simplified, algorithms for special purposes. ing the past practice for computing day numbers seemsadvisable. Specilically there
In this unified approach for planets and stars, position vectors have bcen uscd have been a number of changes in the method used for computing the aberration
throughout. N{any other algorithms use unit vectors (see-The Astronomical AIma- day numbers, C and D. A detailed discussion is given in Section 3.53.
noc). There are penalties either way, and iI the procedures are carefully implemented The day nurnber technique may also be used for objects within the solar system,
the results will be identical. but the geocentric position of the body is required at the time t - At, where z11is
The method given here involves a series of steps. However, if the observation the tight-time. Thus the proper motion, and annual parallax corrections must be
time remains the same, quantities such as the time arguments, the ephemeris data omitted.
1EA EXPLANATORYSUPPLEI\,IENT REFERENCE
3 / CELESTIAL SYSTEI\,S 157

The calculation of the apparent place of a star for a date t+t, where / represents Besselianday numbersA. B. C, D, E, and star constants (Section 3.3,12),fronr
a fixed epoch such as the beginning of the Besselian year or, as is current practice,
the middle of the Julian year, and r represents a fraction of the year, first requires a t = a + A a + B b + C c + D d + E + J t a n 26 + r p . + n ( d X c y ) ,
the calculation of the mean place lor the mean equinox and epoch of r. The reduction
5t = 6 + Aa' + Bb' + Cct + Ddt + Jt tan6 + r p6 + r(dt X - ct la, (3.341_t)
then involves the application of corrections for precession, from the epoch I for the
interval r, nutation, stellar aberration, annual parallax, proper motion, and orbital
or using the inclependent day rrumbers f, S, G, h, H (Section 3.343), fronr
motlon.
Proper motion, orbital motion, and stellar aberration do not affect the frame
( r r = a + l + g s i n ( G + o ) t a n6 + h s i n ( H + a ) s e c d+ - / t a n 26
of reference,but cause changes in the actual direction in which the star is observed;
the corresponding corrections must therefore be calculated with respect to a par- +rlto+lt(d.X cn,
ticular reference system and applied to the position of the star in the same system.
6r = 6 +8cos(G+ a) + ftcos(I1+ (:Y)
sin 6 + icos6 + "/' tan6
Precession and nutation, however, are changes in the frame of reference and do not
affect the actual direction in which the star is observed, These two corrections, and + r1t5 + r(.d'X - c'l), 11 1z1l )r
that for annual aberration, are sufficiently large to make their order of applicatir.rn
of significance if cross-product terms are neglected. The corrections for parallax, where a, d are the right ascension ald declination of the star for the mean equinox
proper motion, and orbital motion are generally very small and can be applied at and epoch of 1, and po ancl p6 arc the annual proper motions in riglrt ascension
any convenrent stage. and declination, respectively, and rr is the parallax. X and Y arc the barycerrtric
Since nutation is calculated from the longitudes of the Sun and X{oon referred to coordinates of the Earth. Thc dir,ynumbers are usually tabulatcd in seconds of arc;
the mean equinox of date, it is (theoretically) necessaryto apply precession before when used for calculating the star's right ascension, measured iu timp. either ther'
nutation. There are then two methods for calculating the reduction to apparerlt or the star constants bv ri,hich they are multiplied should be divided by 15.
place; if precession and nutation are applied first (Method 1), then the aberration For stars, the correctio s for the deflection of light, which has not bcur included,
correction should strictly be applied to a fixed star whose coordinates (refcrred to are given in Section 3.261.
the moving frame of reference) are continuously changing. If aberration is applied
first (Method 2) then the corrections for precession and nutation should strictil'be
3.342 BesselianDay Numbers and the Star Constants To the first order, the Besselian
applied to the changing position of the star.
day numbers A, B, an<l E correct for precession and mrtation: C and D correct for
The two methods give identical results, of course, but for systematic calcula-
annual abcrration, and are gir,-cnby
tion oqe must apply corrections for precession, nutation, and aberration to a fixed
star, any residual corrections (if appreciable) being applied separately. As night be
Aq.
expected, the largest correcting term is the same for both rnethods, but the other A = nr * n---i = flr + srn€oAu,
terms difier. An analysis of the magnitude of the residual terms, taken in conjunc-
tion wii,h the second order terrns of precession,nutation, and aberration thenmelves, B= Ae,
show conclusively that Method 2 leads to smaller residual errors (Porter and Sadler,
1953; see also Section 3.344; also, it is more logical to apply aberration with respect
E=^'+, ( 3 . 3 4 2l )

to a fixed frame of reference.)

where X = XA(T=O,t+r), tlt = dA(.T=O,t+.r), n, and m (for the in<lependentday
From 1984 onward, the day numbers printed in The Astronomi,caLAlnt'anac att,
calculated according to Method 2, and are designed for use rqith star places referred numbers) are the precessionalparameters(Section 3.211), and the nutation angles
to the mean epoch and equinox of the middle of thc Julian year. A/, Ae, and E0thc mean obliquity of thc ecliptic, arc givcn irr Section 3.222. They
are all evaluated for the epoch / + T.
3.341 M€thods Using Day Numbers The apparent right ascensionand declination Since 1960 (see Section 3.53 for previous methods) the aberrational day num-
(o1 , 61), of a star at an epoch t+ r-, which includes the effects of precession,nutation, bers are a function of the Earth's barycentric velocity i, and are given by
annual aberration, proper motion, and annual paralla><,aud includes the second-
order day numbers ("/ and "/' see Section 3.344), may be calculated either using the D= -k* C=+kY t=+k2, (3.342-2)
 158 EXPLANATORYSUPPLEMENT
3 / CELESTIALREFERENCESYSTEIUS 159
*Lt"rs it = (X,Y,4 are the velocity components, in AU per day, referred to the mean Table 3.344.1
equinox of I (since aberration is applied first), and f = (za/86400) x R x 3600, where Second-OrderTerms
re = 499.004782 is the light-time for unit distance in seconds, and R converts from
Adcos 6
radians to degrees (i.e., k converts from lu/per day to seconds of arc).
If, however, the barycentric velocity of the Earth (fe) is referred to the standard 1 +fgcos (G + a)sin 5 -/gsin (G + a)
equinox of J2000.0, (e.g., as in DE200), then use the precessionmatrix P (see -
l.fg cos G cos a sin 6 +;/9cos Gsin c!
Section 3.21) to precess these coordinates from the fixed equinox of J2000.0 to the
3 +lgf sin2(G + a)tan 5 sin6 j d s i n ' ? ( G +o ; t a nr
eouinox of date /, i.e.,
'l 4 +]ef cososin (2G + a)cos 6
i = P1J2000.
0, tl io r 1z1r-1\
5 +rltF sin2(H + a)sec 5 ;,f sin2(H + a)tan 6
The quantities A,{, Ae, A, B, C, D, E, in seconds of arc, and 7 are tabulated 6 +;,f cos2(H + a)sin 26
d,aily in The Astronomical Almanac. +hl cos(H + (x)cos26
7
The star constants, which are constant only for the mean equinox of the flxed
epoch /, are defined by 8 tF sin26
I +9hsin(G+H+2d)tand -ghsin (G + a)sin (H + a)sec 6

a=mln+sirrotan6, c ' = c o sc r . 10 +9lsin(G + (})

b=cosotal6, b' = sina,

/1 t1)_/\
c=cososec6, c' = taneocos6 sina sin6,
d=sinasec6, d/ = cosa sin6, Terms 1 to 4 are neglected in the approximations for precessionand nutation, 5
to 8 are neglected at the annual aberration stage, and 9 and 10 result from Method
where cr and 6 are the mean place of the star for epoch /. The precession rates 2. Method 1, which is not considered here, gives rise to more terms.
(Section 3.211) m and z, and e6, also should be evaluated for epoch t. Thus the second-order corrections J and "// are such that the corrections to right
ascension and declination are Jtan26 and "// tan 6, respectively. They are derived
from the rnost significant terms, i.e., terms 3, 5, and 9, given in Table 3.344.1, and
3.343 Independent Day Numbers The independent day numbers /, g, and G cor
replacing secd by t tan d, witir an error that vanishes at the poles. This gives
rect for precessionand nutation, andh, H, and i correct for annual aberration. They
are given by
J = +[g sin (G + a) t ft sin (I1 + rr)] [g cos (G + a) t,rr cos (H + o)]

f = ( m l n ) A + E = m r + c o s eA l , h s i n H= C , = +[(A1D)sino+(Bl O c o sa ] [ ( A t D ) c o so - ( B 1 C ) s i no ] ,
l
gsinG=8, hcosH = D, (.3.343-t) - : [ Rs i n( C + o ) t l ? s i n( H + o ) J 2
"/'= I
gcosG=A, i = Ctan€.
= - jt{a t o) rino + (BI Qcosal2, (3.344-t)
These day numbers a"redefined in terms of the Besselian day numbers, which are
the upper sign being taken for positive declinations, and the lower sigtr for negative
given in Section 3.342.
declinations. The expressionsfor second-order terms and J and.// assume that /, g,
h a,remeasured in radians, and,/ must be divided by 15 to be converted to seconds
3.344 Second-Order Day Numbers In the equations of Section 3.341 the terms of time.
"/ and,// are called the second-order day numbers, and approximate the second- These second-order dav numbers J and J/ are tabulated it The AstronomicaL
order terms that have been ignored. Table 3.344.1 lists all the second-order terms Almano,c as simple functions of north or south declinations, right ascension' and
neglected from approximations made at the various stages. date, so that the complete reduction may be made in one operation.
 REFERENCE
3 / CELESTIAL SYSTEMS 161
160 EXPLANATORYSUPPLEMENT

Table 3.344.2 as it is usually considered a correction to observations rather than an effect to be

Errors Oue to NeglectingSecond-OrderTerms taken into account when computing a topocentric place, it is not considered here.
The simplest way of computing a topocentric place is to compute an apparent
0:007 0:008 0:009 0:010 0:012 0:014 0:016 0:018 0:020
pla.ceusing the position and velocity vectors of the observer rather than the center
5o 25" 350 490 540 57' 60' 63" of mass of the Earth. That is, modify the barycentric vectors Es(r) and ts(r) by
-1.,..i 574 62' 670 70' 760 7A" 810 adding the position and velocity of the observer relative to the center of the Earth,
see Step h. This procedure is equally applicable to stars and solar-system objects.
The development that follows requires quantities related to precessionand nutation
that, in the computation of geocentric apparent places, are not needed before Step
q. These quantities should be computed, used here, and saved for later use: they
are specifically noted as they arise.
The method is of advantage in the routine calculation of a number of star places.
In particular, Porter and Sadler (1953) clearly show the advantages of restricting
the range of r to i 0.5; this has the effect of reducing all the second-order terms
3.351 Location and Universal Time ofthe Obs€rvation
in / and g, which are functions of time. A more detailed analysis of the magnitudc
of the neglected terms in different methods has confirmed the conclusion that if
Step a: Determine the universal time of observation, specifically, the epoch of
r is allowed to reach +1, and no secorrd-order corrections arc applied, there arc
observation in the UT1 timescale. UT1 is affected by unpredictable irregularities
unavoidableerrors of 0'.,010,even at declinationsof 45o. If r is restricted to I 0.5.
in the Earth's rotation, but is always within 0i9 of UTC, the lattor defining civil
and the J and J/ terms are included, then the maximum error reduces to 0l'003, arrd
the range of declinations over which second-order corrections rnav be neglected is time and broadcast worldwide according to international convention. The difference
AUT = UTI - UTC is determined and distributed by the International Earth
correspondingly increased.
Rotation Service. The predicted r'alue of AUT to rvithin 0il (denotcd DUT) is also
Table 3.344.2 gives the upper limit of declination for a given error '!r'hen no
coded into UTC broadcasts.(SeeSection 2.24.)
second-order terms are applied.

Step b: Obtain r (in meters), the position vector of the observer in an Earth-
3.35 Topocentric-Place fixed, geocentric, right-handed coordinate system, with the ,ry-plane the Earth's
Algorithm
equator, the xz-plane the Greenwich meridian, and the z-axis pointed toward the
The topocentric place of a star or planct rcfcrs to its apparent direction as it would north terrestrial pole. In terms of the observer's geodetic latitude e, longitude )
actually be observed from some placc on Earth, neglecting atnrospher.icrefi.action. (east longitudes positive), and height i above the Earth's reference cllipsoi<l [frrr
The apparent place, developedin Sections3.31 and 3.32. can be thought of as the most purposes the height abol'e mean sea level (the regional geoid) can bc used]. r
apparent place of an object for a fictitious observerlocated at the center of a trarrs- is given by
parent nonrefracting Earth. The differencebetween the apparent aDd topocent.ic + li)cos ocosIl
place is due to the slightly diffclcnt position and velocity of an obserr.er on the
f(aC
r= (3.351-r)
Earth's surfacecomparedwith those of the fictitious obscrvcrat l.hc Earth's center.
l ( a c + n ) s i n d s i n Il ,
L ( a S + , ) s i n q tl
The change itt direction of the observcrl body due to the tliffcrcucc of position is
referred to as geocentric parallaz, and is significant only fbr objc<rtsin thr: solar sys- where a is the eqrratorial radius of thc Ea,rth,
tem. It is typically a few arcsccondsfor most solar-systcm trodics, bul, reachesabout
lo for the Moon. The changc in dircction due to the difierencein vclocity (due to
C = ( c o s @ 2 + ( l - f ) 2 s i n 2E ) - 1 1 2 , S=(l -J)'?C, (3.3s
1-2)
the rotation of the Earth) is referred Lo a^sd,iurnal aben'atiott, and is independeDt
of the distance of the observedbodv, and is ahval.slessthan 0i32.
and / is the adopted flattening of the Ealt|'s rcfcrenceellipsoid. The IAU (1976)
Atmospheric refraction also afiects the apparent direction of celestial objects.
c o n s t a n t sa r e a = 6 3 7 8 1 4 0 m a n d f = I 1 2 9 8 . 2 5 7 .
In fact, refraction at all wavelengths is orders of magnitude larger than either geo-
centric parallax or diurnal aberration. Refraction is discussed in Section 3.28, and
 162 EXPLANATOBY
SUPPLET\,|ENT g / CELESTIALREFERENCESYSTEMS 16s
3.352 Apparent Sidereal Time at the Epoch of Observation The preceding expressions take into account polar motion, which affects the
components of the observer's geocentric position vector at the 10-meter (013) level,
Step c: Use Step s of Section 3.319 to obtain the two fundamental nutation angles, and may be neglected if desired by replacing R1(yr) and Rz(xp) by unit matrices. Ne-
Ary' and Ae, and the mean (e6) and true (e) obliquity of the ecliptic. Save all these glecting polar motion affects the computed topocentric place of the Moon by several
quantities for later use. arilliarcseconds, with a much smaller effect, inversely proportional to distance, for
other bodies, If effects at this level are important, corrections are also required to
Step d: Using the UT1 epoch of observation as the argument, compute the Green- refer the regional geoid (the coordinate system for the observer's geodetic latitude,
wich mean sidereal time 9. longitude, and height) to the Earth's reference ellipsoid. In forming the velocity of
the observer a standard value for the rotation rate of the Earth has been used, and
0n = 67310".54841
+ (876600h
+ 8640184a812866)7, (3.352-l) the small effects due to the variation in rate frorn this value and due to the change
+ 0:0931041 - 6".2x tO 6Tl of the polar motion components have been ignored.
More information on the computation of an observer's geocentric coordinates
where 7u is the number of centuries of 36525 days of universal time from 2000 are given in Section K ol Th.eAstronomical Almanac, Mueller (1969), Taff (198r),
January 1, 12' UT1 (JD 2451545.0 UTl). The Greenwich apparent sidereal time is as well as ChaPter 4.
then
0 = 0^ + A{ cose, (3.3522) Step f; Converi the geocentric position and velocity vectors to units of AU and
Au/day, respectively, by multiplying g(l) by I I A ar'd E(t) by 86400 /A, where A =
where d- and Al must be in the same units. Greenwich mean and apparent 1.495978'70x 10ll is the number of meters in 1,q.u(IAU (1976)).
sidereal time are the angles between the Greenwich meridian and the mean and
true equinoxes of date, respectiveln and with the equation of the equinoxes (Sec-
3.354 Position and Velocity ofthe Obs€rver in the Space-Fixed Frame
tion 3.223) are tabulated daily at 0hUT1 in Lhe The Astronom.ical Almanac.

Step g: Transform the vector g(/) and g(/) to the coordinate system defined by the
3,353 Geocentric Position and Yelocitv Vectors ofth€ Obs€ryer
Earth's mean equator and equinox of the reference epoch t6, which is the space-fixed
coordinate system in which the position and velocity of the Earth are expressed:
Step e: Compute the geocentric position and velocity vectors of the observer. rvith
respect to the true equator and equinox of date, in meters, and meters/second,
respectively, from G t r l = P t P l B t l )= P r P r * , r , = R r g l t .

c(/) = P-'N-1 g(/)= PrNrE(r)= Rr E(r), (3.35,1.-

I)
g(r) = R:(-d) Rr0p) R2(,rp)r, ( 3 . 3 5 3l )

A(r)=wR,rg(r) where P and N are the precession and nutaiion matrices, developed in Steps q and
cosd 0l s in Sections3.318 and 3.319.Here the inversematrix is simply its transpose.The
='" [ - s i n d _;"' R'(yo)R2(ro)r, (3.3s3-2) most efficient procedure would be to evaluate the elements of these matrices at
";" 3] this point and save them for later use in Steps r and t. Steps (q) and (s) could be
L
skipped.
whete w = 7.2921751467x l0 5 radians/second(Aoki et ol., 1982) is the standard
value of the rotational angular velocity of the Earth, d is the Greenwich apparent Strictly, the precession and nutation matrices define a transformation between
sidereal time at the time of observation, calculated in Step d, and R is a unit a space-fixed system and a slowly rotating system. The slow rotation is the chang-
vector pointing toward the north Celestial Ephemeris Pole of date. The angles .r2, ing orientation of the Earth's axis due to external torques that the precession and
yp correspond to the coordinates of the Celestial Ephemeris Pole with respect to nutation theories describe. Therefore the conversion of the observer's velocity given
the terrestrial pole mcasured along the meridians at longitude 0' and 270" (90" in Equation 3.354 1 is missing a Coriolis term. However, the equivalent linear ve-
west) (seeSection 3.27). locity of this rotation is of order 10-5 m/second for an observer on the surlace of
 '164 EXPLANATORYSUPPLEMENT

the Earth, comparable to the tracking velocity of large telcscopes and con4rlcttll'
negligible.
r 3 / CELESTIALBEFEBENCE
SYSTEMS

3.4 DIFFERENTIALASTROMETRY
t03

the algorithms in Sections 3 3i and 3 32

For differential astrornetric measurement,
it is necessarYto consider onlv effects that
can be simplified. In differential work
of the observed bodies, i e , arc
can alter the angles between the position vectors
sphere. The orientation of the coordinate system is not
3.355 Barycentric Position and Velocity ofthe Observer lengths on the celestial
of fundamental irnportance, since in most casesthe celestial and instrrr-
"onrid.r"d event, in dificrential observirrg. the
Step h: Calculate the barycentric position and velocitl'of the obserr:er Os(t) :uld mental coordinate Systems ale coupled. In any
the fact, during the reduction of the
Os(t), by adding vectors G(t) and G(l). obtajned above,representingthe gcocclt]-ic coordinate systern is not established until after
position and velocity of the observcr. to EB(t) aDd Es14, obtained in Stcp f of observatrons.
place'
Section 3.31, which representsthe barycentric position and velocity of the Ealth. There are three types of difierential positions: the virtual place the local
differential observa
thus and the astrometric place. For the reduction of high-precision
power' it is straight-
tions, the local place should bc uscd. With available computing
place of all objects within a field and rrsel'lte ensemble
forward to compute the local
os=Es(t)+G(t), o f l o c a l p l a . e sa s t h c s t a r l i n g p o i n t f o r l h e r c d u c t i o n p r o c c d r t r e '

Os=Ee(r)+G(r). ( -1.-155-l )
3.41 VirtualPlace
in thr: coordinate
Then reclcfinethe vectors Es(t) and Es(/) to be identical to Os(t) and Oelr;. The virtual place can be thought of as an apparent place expressed
plaDet as it
system of the reference epoch I0 It represents the position of the star oI
Step i: Continuing at Step g in Section 3.31. the other stcps follos- as lrcfitt'. would be seen frorn the centcr of mass of the Earth at some date in thc coordinate
except that the elementsof the precessionand urrtation matrices ueed not lrc t' system defined by the Earih's rnean equator and equinox of the refcrcltce epoch'
computed in Steps q and s. assuming that the Earth and its atmosphere werc transparcut and nonrefracting'
therr'-
For the reduction of high-precision difierential astrornetric obs"rvatious'
The right ascensiono and declination 6 obtained at Step u represcrlt tl)e fore, the final precession and nutation rotations need not bc performed' and Eqrra-
topocentric place of the object at tlrc epoch of obscrration. The topocentri( hou1' tions 3,31 1 and 3.32 1 reduce to. respectivr:ly.
angle of thc object is given by LHA = d+,\ - a, whcre objects w'estof the nr('ri(liaD
(setting) have positive hour angles. u1(/')= f[sluB(t r) - Es(r)]1, (3.41-1)
Irr rnany cases,the previous pro<cdure may be simplified. but most ,rLle is
- Es(t)]1,
aud ua(') = /[g[uB(r0)+ (1- r0)uB(ro) (3.4r-2)
neededfor objects in the inner solar system s-hen the highest precisionis rt'<1uitctl.
For objects beyond thc inner solar system. to rnilliarcsecoud precision. nrrtation
can be ignored throughout. That is, rnean sidereal time can be used inslctrcl of where all the sytnbols have been define<l in Sections 3 31 and 3 32 Thc resulting
apparelt sidereal time (the equation of the equinoxes can be considered zero) a,nd position is called the virtual plzrceand is computed by following, in Section 3 31'
the nutation rotation matrix in Equations 3,35,1-1can be neglccted(i.e.. N can be Steps a through p, then setting u4 = u2 and skipping to Stcp u, or equivalcntly ftrr
consideredthe unit matrix). Additionalll'. the differcncebetvct:n UT1 and LITC stars, in Section3.32, Steps 1 through 7, and Step 10 where ua = uz' The ornissionof
timescales can be ignored. However, these simplifications may not result in aul real the final precession and nutation rotations rloes not introduce any approximations
computational saving. The nutation parameters would have to lle calcuiatetl a,rryri/tl] s rc omitted'
o r d i s l o r t i o n . ,5 i p c eo n l y o r t h o g o n a tl r a n s f u r m a l i o n a
in a later step. Furthernrore, these sirnplifications zrffectthe computed sidereal tilne
at the l" level, and will therefore cause errors in the computed topocentri( llotlr 3.42 Local Place
angle of this magnitude. Therefore. irr rnany cases,carrying out the full proccdulc.
and saving the r'alues of the relevant nutation and time variables for later use, n1a) The Iocal place is essentially the topocentric place expressed in the coordinate
be the most Drudent course. system of the reference epoch ts. Local place is related to topocentric place in the
 too EXPLANATOBYSUPPLEMENT
3 / CELESTIALREFEBENCESYSTEMS toI
same way that the virtual place is relatcd to apparent place. Specifically, the local
For observations of solar-system bodies, it is frequently useful to cornpute a
place represents the position of zr star or planct as it rvould be seen from a specifi<
location on Earth at some date and time, in the coordinate svstem deflned bv topocentric astronetric place. Only the correction for geocentric parallax is ap-
trrt, p1ed. To compute a topocentric astrometric place, simply add the procedure given
Earth's mean equator and equinox of the referenceepoch,assumingthe atmospheri,
were nonrefracting. To computc it, simpll,add the procedurc for calculatirre in Section 3.35 to the procedure given above for computing an astrometric place.
th(, (Again, the precessionand nutation rotations in Section 3.354 should not be omit_
topocentric place (Section 3.35) to the procedure for computing tlre virtual pJa.,,
(section 3.41). (Note: The precessionand nutation rotations in sectio'3.354 shouLr ted). In this case, however, the observer's velocity vector need not be corr4)uted,
not be omitted.) The local place has utility beyond its use in rerarrvc astrorretr\,. since aberration is ignored. Topocentric astrometric places of stars are I1cver re-
quired, since the topocentric correction is vanishingly small.
Before 1984, the effects of the E-terms of aberration were omittcd: a detailed
discussionon them is given in Section 3.53. Also, in the past the term ,,astrographic,'
3.43 AstrometricPlace place was used instead of astrometric place.

In differential rr,'ork it has also becn customarl,. if not stricth, correct.

to llegle(.t
both the gravitational deflectionof light (function g[... and thc aberratiou 3.5 TRANSFOBMATION
TO FKsSYSTEMANDEPOCHJ2OOO.O
]) of light
(function /[...]). The assumptionis that for sufficientlysmall ficlds thcse
diflerential
effectsare so small that the relative positionsofobjects remain unafier:ted.
Sirnilarh. Expressions are given for converting both observational and conpiled catalogs from
difierential refraction is iguored, although its efiect coulcl be much
larger becaust, the old FK4 reference system of 81950.0 to the new FK5 refercnce svstem of
of the variation in color over the samc ficlds. An1. rcsidual
distortion of tho fiekl J2000.0. A detailed discussion of the formal background is also given.
resulting from the neglect of thcse effects is assumed to be
absorbcd into platr.
consta[ts or similar paranleters solved for in the data_rcduction
Drocess.
The astrometric place of the planet or star js obtailled fro[r
Equations i].31 1 3.51 FK4Zerc-PointCorrectionin RightAscension
and 3.32-1, which reduce to
An early discussion of the problem of the position and motion of Nervcomb,sequinox
and the FK4 equinox can be found in Black*ell (1977). Later, it was deternined by
u+(1')=us(r-r)-Es(/) (3.43-r) Fricke (1982) that the FK4 right-ascension system requires a correction of +0:035
at 81950.0. This has been referred to as an equinox correction. It implies that the
and ua(ri)= uB(r0)
+ (/ - r0)[B(r0)
- EB(r), (3.132) right ascension of every star in any catalog whose right ascensionsare referred to
the system of FK4 at epoch and equinox B1950.0 must be increased by +01035.The
and is obtained by following. in Soction3.31,
Steps a through l, or equivalenth.fbr motion of the equinox of +0i085 / cy. Fricke (1982) and the correction to lunisolar
stars, Steps 1 through 5, then setting ua =
U and skipping to the rasr srep. precessionof +1110/ cy Fricke (1977) implies that the FK4 system has a rotation
However. it should be recognized that the gravitational
deflection of lisht shoujd ln right ascension relative to our present best realization of a,n inertial system of
not really be ignored in this way, since it cannot
irr princriple bc absorbedinto l,late (G085 - l:10) per century. This has been corrected in the FKS by altering the
constants or similar reduction parameters:
the deflection is a firnction of positioD proper motions by 0085/cy and precessionby l'll0/cy.
and distance. Although in any part of the
sky the direr:tionof the deflectionis thc The equinox couection at any epoch, 7, given as a function of its value at
same for all bodies, its magnitudc is less for
solar-s5rstembodies than tbr stars (see 81950.0 and the difierence in epoch (I
Figure 3.26.2). Generally,this detail is of little practical ?n6),has been given by Fricke (1982) as
importance, since orrll,il follows:
a few special casescan it cause errors exceedirrs
0:,01.
Astrometric placesare simple to compute! and thprefore
have beenwidell. used. E(O = 0:035+ 01085(T Td / 36524.2198'182, (3.51-l)
Another attractive fcature is that they can be directly plotted
on an ordinary star
map with negligible error in the resulting field configuration.
Astrometric places wnere ? is the Julian date at any epoch and 76 is thc Julian date corresponding
are-therefore used for the ephemerides of faint or fast-moving
solar_system bodies, to the beginning of the Besselian year 81950.0, i.e., Is = 2433282.42345905.We
such as minor planets and Pluto, in Z[e Astronomi.cal Almanac.
suppose that the divisor of the epoch difference should be the number of Julian
 168 EXPLANATORY
SUPPLEIVENT 3 / CELESTIALREFERENCESYSTEMS
169

days per tropical century, but no significant error is made if the number of Julian been available at thc time of observation. In the case of individual obserlational
days per Julian century (36525) is adopted instead. catalogs, such as the series of firndamentally observed Washington Six-inch Transit
Aoki et al. (1983) have made a convincing argument that, for the sake ofconsis- Circle catalogs, the reductiorr to the FK5 equinox is best done by directly calculat-
tency with the new definition of UT introduced at 0h UT on 1984 January 1 (Aoki ing the difierence FK5-Catalog from the stars in common to both catalogs.
et aI., 7982), it is important that the FK4 equinox correction should be applied to
right ascensionsof the FK4 catalog referred to the same equinox and epoch (198,1
Termsin Aberration
3.53 Elliptic
January 1) as the date on which the new definition went into effect. The equinox
correction corresponding to that epoch is accordingly, 3.531 Effect on Position RecommendatioD 4 of IAU 1976 Resolution No. 1 indi-
cates that from 1984 onward, stellar aberration is to be computed from the total
5) = +0i06:qO.
E(2445700. (3.51-2) velocity of the Earth referred to the barycenier of the solar systeul, and mean
catalog places are not to contain elliptic terms of aberration.
The reduction to the IAU 1976 system of constants at J2000.0 of the right ascension The problem of reducing observational catalogs to a unilorm system is compli-
system of an observational catalog (as distinct from a compiled catalog) referred cated by the change irl the conventional value of the constani of aberration. n'hich
to the FK4 system at an arbitrary equinox and at the mean epoch of observation since 1911 had been taken as 20'i47 (Paris conference,1911), and from 1968 has bccn
should be done in a manner consistent with the way in which the catalog rvas taken as 20'1496(IAU 1967). Before 1911, the value of StrLtveand Peters. 20i4551,
referred to the system of FK4 at the time of obscrvation. On the other hand. if an was in common use, and from 1984 the value of 2O'.'19552has been introduced as
observational catalog has been constmcted from a fundamental treatment of thc the conventional value.
observations of solar-system objects and thus referred to the observed dynaniical Reduction to a uniform system is furthcr complicated by changes in the methocl
equinox,the reduction to the IAU 1976system of constantsat J2000.0I'ould ignore of computing the aberrational day numbers, C and D, given in the national ephem-
the zero-point correction normally required by the FK4 system in right ascerrsion. erides and almanacs. Before 1960. the aberrational dav numbers C and D *'ere
calculated lrom a circular approximation io the motion of the Earth in rvhich small
periodic terms due to the action of the Nloon and planets and the elliptic terms in
3.52 TheCorrection
to the FK4ProperMotjonSystemin RightAscension aberration q'ere neglected (see Explanatory Supplement, p. 48fr.).
Beginning u'ith 1960, lhe aberrational da1'numbers C and D rvere derived
The existenceof a secular term of the lorrn 0i085(2, Ti /36524.2198782in tbe
from the true velocity of the Earth rcferred to the center of rnass ol the solar
equilox correction has generally been interpreted as a constant, nonprecession-
system and to a dynamically determincd frarne of rcference. The elliptic terms in
dependent conectiotr applicable to the FK4 right-ascension proper-motioD svstern
aberration due to the eccentricity of the Earth's orbit rvere then removed from an
(Fricke 1982). This implies that at any epoch, and regardless of the equinox. all
otherwise completcly correct expression of the annual aberratiolt. Furthermorc, as ot
right-ascension proper motions of the FK,l catalog (or any catalog r.ith the propcr.
1960, epheuteris values of the aberlationa] clay nurnbers properly reflect thr: slou'ly
motions referred to the FK4 system) must be increased by +0:085 per tropical
changing eccentricity and longitudc of perihelion of the Earth's orblt (E:rplanatory
centurv.
Supplement,p. 158tr.).
The nonlinear character of precessiol vie*.ed as a mathenatical operation sug-
When one is working u'itir jndividual obscrvational catalogs, reductions to a
gests that subtle differences can be introduced ilto a cataloe transformation frorrr
uniform system of annual abcrration should take into account the changes in the
one epoch arrd eqrrinoxto alothcr. dependjlg on th..p.,ch at rvhich correctious
value of the constalt of abcrration and the method of calculation of the published
arc introduced. A number of possibilitios come to mind. Onc could, for exanpl<:,
apply the corrections to positions and pr.oper motions rcferred to: values of the aberrational dav numbers, C and D This will adjust the published
values of the observed posil,ions in a manner consistcnt with conveutions in use at
(1) the epoch and equinox of the beginning of 1984; or the time the observatious were made arrd reduced from apparent to mean place'
(2) the epoch and equinox of B1950.0;or If this procedure is not followed, then corlupted data will be produced, and the
(3) the epoch and equinox of the mean epoch of observation. potential for svstematic improvement will be diminished.
The FK4 catalog, and other fundancntal catalogs such as the N30 (N'Iorgan'
Among these three options, the third one most closely approximates the op- 1952) and thc Gencral Catalog (Boss 1937), were compiled from observations nrade
eration that would have been carried out if the improved IAU 1976 system had prior to 1960, whe[ a circular approximation was nade to the rnotion of the Earth
 Y
170 EXPLANATOBYSUPPLEI.4ENT SYSTEMS
BEFEBENCE
g / CELESTIAL 171

in its orbit, and the elliptic terurs in aberration were ignored, except for stars $,ithit v/here t = 20149552is the constant of abenation at J2000 0'
l0' of the poles in the caseof the FK4 catalog,
Due to the rapid, daill' change in the value of the circular component of an- e = Eccentricity of the Earth's orbit

nual aberration, it is not possible to correct mean positions for changes that havq (T - Tdz,
= 0. 01673011 0. 00004193(I- T0) - 0. 000000126
occurred in the adopted lalue of the constant of aberration eiiher in the conpilccl
f = N'Ieanlongitude of perigee of the solar orbit
fundamental catalogs or in observational catalogs of mean positions. Observations
of solar-system objects are normally given on a daily basis and may be corrccLed = 282"04' 4g:g5l + 6tg}]61(T - I0) + I165(7 Z0)2+ 01012(T r0)3,
with a high degree of rigor. However, becausethe elliptic terms in aberration change
6 = Obliquity of the ecliptic
so very slowly, their influence can be removed from the FK4 catalog positions bv
subtracting the elliptic terms in aberration from the catalog right ascension anrl = 23a26'41':836 46:8495(T- T 0100319(7- T0)2+ 0100181(7 r0)3,
declination (o*,, 6"") to obtain the corrected right ascension and declination (o, 6)
as follorvs: where (I Te) is in urrits of Julian centuries of 36525 days aud I0 is the epoch
B1950.0 (2433282.123459os). Equation 3 531 2 has bcen developedfrom ES, p 98'
and may be used to refer Equation 3.531 1 to any epoch of obserr'ationrelative to
d = ocat (ACcosa.u, + ADsinn""t) / (15cos6""r),
t h e m o v i n g( i . e . o f t l a t e ) e q u i n o x .

t
6=6,u, (AD cosa.,, AC sin a"o,)sin 6"", ACtan€cosAat, (3.531 l) For recently observed \\rashington catalogs, srrch as the W5(50) Six-inch.Tran
sit Circle Catalog (1963 1971) (Hughesand Scott 1982)and the WL(50) Ser'ciiirtch
where AC = 01065838,AD = +01'335299. and ACtan e = 0:'028553at epoch Tfansit Circle Catalog (1967 1973) (Hughes, Snith, and Branham, in press) ob-
81950.0. using the J2000.0value of the constant of aberration. served from El Leoncito in San Juan, Argcritina. thc assumed constant of aberration
These equations represent the classical elliptic aberration with firsl-order ac used in the apparent place conputations rvas201496.The W6(50) Six inch Transit
curacy iu the ratio of the Earth's velocity to thc r.elocity of light (\lbolard :rrrd Circle Catalog (1973 1982) (currently under discussion)will be re-rcdtrcedin strict
Clemence 1966, p. 113tr). N{ore than the necessary number of significant figures is accordance with the IAU 1976 rcsolution, atrd \r ill be renamed to rcliect J2000 0 as
carried in order to ensure correct roundinp;. the equinox to $'hich it t'ill be referred
For applications requiring accuracies better than 1 milliarcsecond (mas), tlic
discussionsof Stumpfi (1979, 1980a) regarding the second-orderrclativistic terms 3,532 Effect on Proper Motion Aoki el nl. (1983), introduce a (ronection to the
in elliptic aberration should be consulted. proper motions to compcrlsate lbr a secular change of thc elliptic aberration asso
The nurnerical values for AC. AD, anti ACtane given here arc no!. generalll ciatgd witrh the \ariatiorr of the cccentricity of thc Earth's orbit. of thc longitrrcle of
applicable to every case in which positiorrs have bcon referred to the equinox of solar pcrigec, and of thc obliquity of the ecliptic u'ith time. Ledclle and Schwan re-
B1950.0. For an observational catalog referred to an arbitrarv equinox, each positiorL mark that this is unneccssary bccause of the prir,cticefbllowed irt the t:ornpilation ol'
(where right ascension and declination must be treated separately if their rneal FK4 proper rnotions by which differencr:sin right ascerrsionand declinatiorr bctwecI
epochs of obserr'ation are not thc same) may be precessedto thc equinox of the rnezur observed positions and positions comprrted on the btuis of the FKll were usecl ln
epoch of observation and then corrected for the elliptic tcrms of aberration rrsirrg fact, the problen is complicatecl by the practice of having climinatecl the ittflutnctr
values referrcd to the rnoving equinox (see,e.g., Scott 1964). \Ianv obserrational of elliptic aberratiott for circurlpolar stars bcfore prccessing to il lel ctluttiox antt
catalogs wiil require an epoch-dependent correction of this type. Compiled calalogs epoch, ancl thcn re introducing thc effect of elliptic aberration appropriate for the
-80"
may or mav not requirc a correctiorr, deperrding on how the compilation rvas dotrc new equinox ancl epoch. Tltis rvas not done for the stars between +80o atd
(Lederle and Scht an. 1984). declinatiorr(l'l{3, 1937;FK4, 1963),*'hich means that uo single practice will u'ork
The time-depcndelt expressions lor the coefiicients of Equation 3.531 I ma1, equallv well or consistently for all stars ol the FK'l, nor for any catalogs refcrred to
be written as follows (\\roolard and Clernence, 1966, p. l1,lI{): the FK4 svstem. Corrcctions for the secrrlarchange of elliptic terms in aberrirtion
are not necessary lor proper motions of cilculnpolar FK4 stars. brrt a corrcctlolr ls
required for FK4 stars bctween +80o ancl 80'
AC = -,te cos I cos e, (3.s3r-2)
A very good casc can be rnade that the highest-accuracy transformation of
AD = -/<esin I, the FK4 catalog to the equinox of J2000.0 would result if the procedure outliled
 172 EXPLANATORYSUPPLEMENT g / CELESTIALREFERENCESYSTEMS 173
above were folloq'ed paying spccial attention to the treatment of the stars as fegards
Hence rvith
the elliptic terms in aberration. Stars within lO. of the poles necd have only the
elliptic terms in aberration at 81950.0 removed from their positions. Their proper
motions need no modification. For stars between +80o and _g0. in declination, the e= pertropical
0.00004193 cy
ft"bn= mo = 4609:90 per tropical cy
positions and proper motions should be precessedto the equinox of the mean epoch
of observation and the elliptic terms iI] aberration (referred to the moving equinox f = +6190154 cY
PertroPical no = 2004!26 per tropical cy
#,bn=
of the mean epoch of observation given by Equation 3.531 2) should be removed po = 5O26!75per tropical cy
from the positions. Equation 3.831 1 may be differentiated with respecr ro nme to i= -4618485
pertropical
c1'
ft4rtt=
give the equations for the correctcd proper motions (p,p/) in right ascension and
(3.s324)
declination in terms of the catalog proper motions in right ascensionand declination
the numerical values for AC, AD, and ACtan e in the fixed frame at 81950.0 are
1t*r, pL^), and the catalog positions ((r*,, d.u,),

AC = -01001580per tropicalcentury,
Lb = -0'!OOl245per tropical centurl',
p = pca,- (Accos ocar+ AD sin clcat)
/ ( 15 cos6""1) (3.532_l)
ACtan € = -01000677per tropical certur]'. (3.s32-5)
- p"", sin I "(_AC sin a"u,+ AD cos o"u,)/ cos
d"u1
- plu, sin l "(ACcos a"", + AD sin n"n,)tan 6""r ( The equations for ms and n0 are given in Section 3.214, and p6 = m0cos r + trosin €.
/ 15 cos 6c"t),

pt = p|^, - (Lb coso.,, - ACsin a"u,)sin 6.u, (3.532_2)

- l5p"u,sin l"(-ADsina",, _ ACcos o"",) sin 6"o,
3.54 Precession
- pl,, sin I "(AD cos o"", A C sin a"u,)cos d"u,
- (ACtan € + ACa, sec2e) cos 6"", A detailed discussionon precrxsion is given in Section 3.21. The rrew IAU 1976
precession angles are defincd in Section 3.211, and the old Newcornb angles are
+ l5 p""tsin l"ACtan e sin 6"u,, discussedin Section 3.214.

rvhere the factor 15 converts arc to time or vice versa. and

3.541 Newcornb's Precession The precessionalmotion of the referencc frame from
the equinox and equator of B1950.0to 1984 Jan ldOhis calculated using Andol'er's
AC = +tei, cosesin| - f(A cos€ - e€,sine)cosf, (3.s323) Equation 3.214 1. The adopted values are tabulated in the last column of the Ta-
ble 3.214,1.Thesevalueswere obtained with the following time argutnents;/t = 0. I'
AD = -te sinf - &ef, cosf,
tz = Q.133999566814,and r = 0.033999566814. The rcsulting precessionrnatrix is
ACtane = +kei, sinesinf - i(?cose - et,sine)cosf tane,
-0. 003303433841
| +0. 999965667560 0. 007599409538 I
P r = | + 0 . 0 0 7 5 9 9 4 0 9 5 3 5+ 0 . 9 9 9 9 7 1 1 2 3 9 9 2 0 . 0 0 0 0 1 2 5 5 J 0 2 1
- l. (3.541-l)
-0.000012ss1s54
f +0.003303433s46 +0.ggsss4543s6s
)
where i" and i, are the r.ates of change of I ancl e irr radians per
tr.opical century.
The quantities AC, AD, AC tan e have to bc evaluated in a fixc<l
frtrme at
Bi950.0. To convert from thc rotating frame of clate to this
fixed fiame. the in- 3.542 The IAU 1976 Precession The precessional motion of the reference frame
slantaneous rates of change mo, noj and po at B 1950.0 have to
be removed frorn the from the equinox and equator of 1984 Jan 1d0hto J2000.0 is calculated using the IAU
displacements in right ascension(I cos €), declination (i sin e), and longitude (f)
due (1976) precession angles defined by Lieske (1979) and and given in Section 3 211'
to precession of the mean longitude of perigee.
The values obtained by evaluating these expressions with ? = -0. 1600136893and
 174 EXPLANATORY
SUPPLEIVENT
3 / CELESTIALREFERENCESYSTEMS 175
y = ,7, and the accumulatedanglesliom 1984January 1d0hto J2000.0are (1 =
where
36819985,zA= 369:0188,and 0A= 320'7279,and the resulting precessionmatrix is
lt, p' are the proper motions in right ascension and declination, respec-
tively, referred to the IAU 1976 basis and precessedto 1984 January
+0. 999992390029 -0. 003577999042 -0. 001554929623
| I 1"0';
p, = | +0.003577999042+0.9999935989J7-0.00000278 t855 | . {3.5a211
fr, ft are the centennial general precession in right ascension and decli-
L+o.Oorss4s2s624 0.000002781702+o.sgsg9879r0s2 ) nation, respectively, referred to the IAU 1976 basis at 1984 January
1"0';
a0, 60 are the righi ascension and declination for the FK4 catalog pre-
cessedto 1984 Jan 1;
a = oo + 0:06390;
3.55 The Proper-Motion
System 6 = 6oi
E = 01085per tropical century;
lr'o, Plo are the proper motions in right ascension and declination, respec-
The discontinuity in the right-ascension proper motion system of the FK4 catalog tively, from the FK4 catalog, for example, and precessedto 1984
is to be imposed at 1984 January 1o0n.Proper motions referred to that equinox and January 1o0n;
epoch should: mo, no are the centennial general precession in right ascension and decli-
nation, respectively, referred to Newcomb's precession at 1984 Jan-
(1) be corrected for the zero point of the right-ascension proper-motion svstern uary 1d0n;
as discussedin Section 3.52,
F is the factor for converting from tropical centuries to Julian
(2) be changed from the Newcomb to the IAU 1976 (Lieske et al., I9T7) pre-
centurres.
cession basis, and
(3) be expressedin units of Julian rather than tropical centurres.
The quantities m and n are calculated using the IAU 1976 precession (Sec-
tion 3.211) and at 1984 January 1d0nwith 7 = -0. 1600136893
Only points 2 and 3 apply to the proper motions in declination.
The fundamental condition to be satisfied at 1984 Jaruary 1d0his that the
n = 4611i98926 / Julian century = 307:465950 / Julian centurv, and
centennial variation on the new basis shall equal the centennial variation on thc old
basis plus a constant correction in the case of the right-ascerrsion proper motions: n = 2O04i44743/ Julian century = 133".629829/ Julian century (3'554)

Similarly ms and n6 may be derived from the equations for Newcomb's preces-
<rn.*=tirora+8, sion in Section 3.214, and at 1984 January 1d0hwith tr = 0. 13399956681,

6** = dora. (3.55 l)

/ Julian century, and
mo = 4610'95218 / Julian century = 30'7".396812

no = 200410"126/ Julian century = 133:600750/ Julian century. (3 55-5)

Let quantities without subscripts be taken as the new values referred to the IAU
1976 basis, and those with subscripts 'o' be ta,ken as referring to the old (Newcomb)
The values ol m, n, ms, and ne will be used later in Section 3.566.
basis, then in risht ascension

of Catalogsfrom81950.0to J2000'0
3.56 Equationsfor the Transformation
p + m + n sirLo'tan6 = (ps + rne+ ne sinae tan 66+ i F, (3.ss-2)
This transformation is conversion from a mean place to mean place. That is, it trans-
and in declination forms catalog mean places and proper motions from one refetence epoch (81950 0)
p ' + n c o s o = 1 p l + 1 1 6 c oass ) f , (3.ss-3) to another (J2000.0). The complexity of the transformation results from the changes
 176 EXPLAMTOBY
SUPPLEMENT CEIESTIALREFERENCESYSTEMS 177
in constants, timesceles, a.nd procedures mandatedby the IAU for epoch J2000.0 frcm 81950.0to l984Jan ld(I'
Pr€cession Apply the precession
ftom 81950.0
catalog data.
lg84 Jan ld0li to the vectors u2(tl) and Ul, using the precession matrix Pr calcu-
ftom the angles(0, z, and 0 given in Section3.541.Then

3561 Units and the System of Positions and proper Motions It is assumedthat u3(tr)= Pl u2(tl) (3.s6,t1)
the positionsand proper motions of the catalogto be transformedto J2000.0are
referred to the epoch and equinox of 81950.0. The positions are assumedto be
in units of secondsof time in right ascensior and secondsof arc in declination. riz = Pt tlr. (3.s64-2)
However, when these quantities are used as arguments of trigonometric fuactions,
it is left to the user to expressthem in degreesor radians as necessary.similarly, the .[!om the compouents.r3' ]3' 23, arrd'i2, y2, 22,of the vectors u3 and U2' respectively'
inverseof a trigonometric f'nction is assumedto be degrees.The proper motrons calculate the right ascension and declination and their proper motions at 1984 Jan
are aasumedto be in units of secondsof time per tropical century in right ascensiou 1d0:
and secoadsof arc per tropical century in declination.Radial velocitiesare taken
to be in r.rnitsof km/s and the pa,rallaxesare in secondsof arc. It is also assumed at = tan-l,b I h), (3s64-3)
that the catalog iu question is referred to the system of the FK4. If uot, steps must
be ta.keneither to refer the catalog to the FK4 system or to alter the appropriate d r= t a n -(rr r , y l t G r ) , G56+a)
stepsin the discussionwhich follows.
(xzyz- yz*z)
(3.564-5)
15s(fi +f)'
3,562 EllipticAberration Correct the catalog right ascensionard declination (o"",,
fi22- zt\xtiz+ rtiz + ztzzl (3.sffi)
d""1)for the elliptic terms in aberration using Equations 3.531 1 and j.SZl_2. If t. ,
necessar1,correct the cata"logproper motioDs (p""t, p!,) for the elliptic terms in " \ln-zi
sG
aberration using Equations3.832_J,and 8.532-2.
where
fi=fi+y\+fi.
The equations for p2 and p!, lollow directly from the total time derivative of the
3.563 Position and Velocity Vectors(81950.0to l9g4 equations for a1 and 61. The units of or and 6l are degreesl p2 and pl2 a.re in seconds
Jan ldl)n) At the equinox and
epoch of 81950,0form the position vector ul, and the velocitv vector of time per tropical ceutury, and secouds of arc per tropical century, respectively,
ur from the
correctedpositionso and 6. the proper motions p and p/ (corrected with s defined in Section 3.563.
if necessary).
the parallaxp, and the radial velocity i, usingEquations3.23 1 and 3.23_2givenin The parallax, ht at L984 Jan 1o0n,consistent with a model assuming linear
Section3.23.Note that to obtain the requiredunits of AU/ tropical centurv spane motion, is given by
s is the
conversionfactor from secondsof arc to radians a.nd* = g6a00x
36524-2rggig2x
1.495n87Ox l0-8 is the conversionfrom km/s to AU per (3.564-7)
tropical century. Thus P3 = sin-'(1 //3)
the unit of u1 is astronomicalunits (lu), arld ri1 is in .lu per
tropical century. and the new radial velocity in AU per tropical century, i3, is
The position vector at epoch111984Jan 1d0h(244b200.b),b;t still referredto
-
the equinoxof B19b0.0,is h=@3i2+yi2+4b)t4. (3.s@8)

u2(r1)= ur(ro) + dr(rr - r0), 9.563-1) 3565 Right AscensionZero-Point Correction Apply the zero-point correction to
fhe right ascension:
where (11 - f6) is the number of tropical centuries betweeu epoch
,0 (B19S0.0) aud
epoch t1. 02=cl+0106390, (3.s6s-l)
6z=6r
 178 EXPLANATORYSUPPLEI\4ENT 3 / CELESTIALBEFERENCESYSTEMS
179
v/here the right ascension and declination (a1, d1) are expressed in seconds of tiure the conversion factor to
The radial velocity will be in eu per Julian century, and
and seconds of arc, respectivell.. km/s is 1.49597870x 108/ (86400x 36525)
at the
If neither the radial velocity nor the parallax of a star were known
3.566 Tiansformation of the Proper Motion System Appll,ing the couectiors and the fictitious values produced by the procedurc
beginning of this procedure, then
transformations discussed in section 3. 55 to the proper motions: shouldbeignored.Ifonlytheparallaxisknown,thentheoriginalBl9500rahx:
should be brought forward without modification
p, = ip2 + 0. 085)F (m , md - (lr sin a2 - rs sin a1) tan 62
= (p2 + 0.085)F (307.465950- 307.396812) Catalogs
Observational
3.57 Transformationof
- (133.629829sina2 _ 133.600750sina1)tan catalog in which no proper
62, If the transformation is applied to an observational
the mean epoch of obser
p\ = plzP (n cos Q2 - n6 cos o1) motions are given, and the mean positions are referred to
the mean positions should bc precessedto
vation and the equinox of 81950.0, then
= t1i F cosa2 - 2OO4.Ol
(2OO4.14743 l26 cosa1) general' rvill bc different
the equinox of the mean epoch of obserlation, rl'hich' in
be cottected lor terms in
w h e r eF = 1 . 0 0 0 0 2 1 3 5 9 0 3 . (3.566-l) for each star. The right ascension and declination should
to the nean epoch of observation' provided they
elliptic aberration correspondirrg
the catalog place Special care must
From Equation 3.566 1 it becomes evidcnt that a star precisely at or rvithin ir hai not already been removed rvhen deriving
arr: differerrt'
ferv seconds of arc of a pole musl be treated as a special case. Historicalll.. proper be taken if tlte epochs of observation in riglrt ascensionand declination
motions have been giverr in right ascension and declination without problems. If w h i c h h a p p c n sf r e q u e n t l yi n m o d e l n o b s e r r a t i o n acl a l a l o g s '
should be cor
a star is very close to the pole, the singularity is avoided by consideriug the real fhe right ascensions,if they n'ere referred to thc FK4 system
at the mean epoch of observation in riglrt
displacement p3 cos 62. rected for the zero-point error elaluated
0 usirrg the equafions for
ascension.The positions may be brought forward to J2000
eraluated for
3.567 Position and Yelocity Vectors (1984 Jan ld0h to J2000.0) Form rne \.€ctrors the IAU 1976 prccessionangles(1,;a, and d; given in Table 3 211.l
oI J2000 0'
u4 and u4 frorn o2, d2,/{:, A!, using Equations 3.23 1 and 3.23 3 ensuring lhat the the time interval from the mean equinox of obscrr''ation to the cqrrinox
unit of ua is astronomical units (lu). and ua is in AU per Julian cenrur1,. In this case. the proper motion part of the complttation may bc ignored
positions
Forn the position vector u5 at epoch.12000.0: The same is true of catalogs of radio-interferometrically deterrninetl
dist'ances'
of quasars and other obiects presumed to be located at extragalactic
(VLBI) and
u5(/2)= u4(4) + [4(r2 - rr), However. the transformation of very Long Baseline Interferometry
(3.567-1) of
connected-element interierometrv (CEI) catalog positions from the eqrrinox
than the transformation of the
where (12- tt) is the number of Julian centuries bct$,eenepoch 81950.0 to J2000.0 is slightly more complicated
r: (J2000.0) and
epoctr11,1984 Jan 1d0h. FK4 catalog because of the way the zero point irr right ascension is establishe<l'
adjust-
This usually involves a systematic adjustrnent to all right ascensions \Iost
givcn by
3.56E Precession from 19g4 Jan 1d0hto J2000.0 Applv the precessiorr ments refer to the averaged optical and raclio right asccllsiorl of 3C 2738
fi.orn 198.1 to the
Jan 1d0hto J2000.0to the vectors u5 and u4, using the prccessiolr Hazard el at. (1971\. To the extent that this right ascension was referred
mahix pz calcrl, a correction to right
lated from the angles(,a,;a, and da qiven in Section 3.542 FK4 system at its mean epoch ol obsen'ation, it also requires
ascensionbv an amount that the FK4 equinox requiled at thc same epoch'
lrF
Another factor that should be taken into consideration when transforming
u6 = P2 u5(/2); (3.568-I )
terferometrically observed positions from the 81950 0 basis to the J2000 0 is the
ri5 = P2 ua. (3.568-2) intrinsically high accuracy of such a catalog ln or<ler to preserve the high accuracl-
in
of such catalogs, the cffects that have been discussed earlier the elliptic terrns
From the compolents of the \€ctors u6 and u5, the riglit asccnsion aberration, the equinox corection, and the use of Newcourb's precession constant
and cieclirration
as
and their proper motions at J2000.0 along u,ith the rnodified parallax
and raciial to refer observations to a fixed equinox should all be removed in a manncr
velocity rnay be computcd from a reapplication of Equations 3.5tj4
3 to 3.564 8. closely analogous to the operations rvhich I'ould have been employed in reducing
 180 EXPLANATORYSUPPLEMENT '181
3 / CELESTIALREFERENCESYSTEMS
the original obsenations at the epoch of observation had the irnproved constalts Table3'58.1
been available at that time, In general. this means that catalog positions rnust be SelectedStarPositionson FK4and FK5Systems
precessed back to the equinox of the mean epoch of obscrvation using the same
precession constants as the catalog compiler used, and if necessarv corrected for FK4 tt tt'

the elliptic terms in abeuation and equinox error of the FK.l svstem. The modific{ hms s/cy "tcy Km/s
position is then precessed to the equinox of J2000.0 using the Lieske el al. pre_ -65 10
1 0 00 17 24.774 06.70 +27.141 +116.74 0.134 + 8.70
cession. This procedure does not recognize 1984 as an epoch at which any spccial 't0 00 20 04.3100 -64 52 29.332 +26.8649 +116.285 0.1340 + 8.74
condition is to be imposed and best preserves without distortion the high pr.ecisigl 1 1 9 0 3 17 55.447 -43 15 35.74 +27.827 + 74.76 0.156 +86.80
of interferometricallv detcrmined positions. 1 1 9 0 3 19 55.6785 43 04 10.830 +27.7694 + 73.050 0.1559 +86.87
239 06 11 43.975 74 44 12.46 +3.105 - 21.12 0.115 +35.00
239 06 1 0 1 4 . 5 1 9 6 ,74 45 11.036 + 3 . 1 3 1 0 - 2 1 . 3 0 4
0.1150 +35.00
3.58 NumericalExamples 538 14 36 11.250 -60 37 4a.85 -49.042 0.751 -22.20
+ 71.20
538 14 39 36.1869 50 07.393 49.5060 + 69.934
0.7516 22lA
Numerical examples are given in Table 3.58.1 for selected FK4 stars. ,lhc sclcctiol
793 21 o4 39.935 +38 2 9 5 9 . 1 0 +35.227 +318.47 0.292 -64.00
li'as made on the basis of proximity to the north and south poles, the magrritude:
793 21 06 54.5901 +38 44 44.965 +35.3528 +320.206 0.2923 63.89
of the proper motion, parallax and radial r.clocit1..arrd one star n,hich crosses thc
907 01 48 44.744 +89 01 43.74 +18.107 0.43 0.000 + 0.00
equator bet$€en 81950.0 and J2000.0.For each star. the first line givesthe 8t950.0 907 02 49.8131 +89 15 50.661 +21.7272 - 1.571 0.0000 + 0.00
position on the FK4 system and the second line gives tlu.tlarrsformecl rltrta for -89 08 18.48
923 20 03.004 +11.702 0.09 0.000 + 0.00
J2000.0. The transfor.med FK4 data given in Table 3.58.1 a,re not the saure as the 923 21 08 46.0652 -88 57 23.667 +8.4469 + 0.171 0.0000 + 0.00
FK5 data resulting fron the incorporation of ncw observational results prorlrrcecl -580.57
1307 1 1 50 06.172 +38 0 4 3 9 . 1 5 +33.873 0.116 98.30
since the publication of the FK4 catalos. 1307 1 1 52 58.7461 +37 43 07.456 +33.7'156 581.216 0.1161 -97.81

1393 1 4 54 59.224 +00 01 58.08 +0.411 - 2.73 0.000 + 0.00

'f393 14 33.2650 00 10 03.240 +O.4273 2.402 0.0000 + 0.00
3.59 MatrixMethod

There has been much controvcrsy in thc ljterature over the correct procc(hu.e fclr
transforming catalog positions and proper motiorls to the FK5 svsteul dis<.rrssecl Step 2: Remove the E-terms of aberration frou] the B1950.0 catalog ulczrll place.
in the precedingsections (Smith el o/.. 1989). The matrix nrethod describcd herc There is also a question rvhether thc E-terms should be rernor.ed frour thc propel
(Yallop el ol., 1989) is baseclon Standish (1982b) but modified motions or not. Thc problem is dis<:usscd urore fulll in Section 3.53 and thc corre
to agree \\ith thc
proceduregivcn in Section 3.56. The transfbrmation is catcgorized sponding vector devcloprnentis given irr Yallop el al. (1989). If the FK.l catalog is
iuto nirru steps;
six of the steps of the transformation are represented by a six-bv six llral,l.ix (le- used, they certainlv do not have to be rcrnoved from stars within l0' of the poles,
scribed below. Sections3.591 and 3.592give thc completealgorithms for conr,e,r.iirrg becausethey have not been included (Lcderle, 1984).
star placesfrom 81950.0 to J2000.0.and from J2000.0to 81950,0 bl.rcr.crsiug thc
Step 3: Apply spa(c motion to the positior.rlectol to the epoch 198.1.lartrtart'1 0,
transfcrruration. It should be noted that whcrr transferring ildir.idual obser.\ations.
which is the epoch at rvhich the sidereal time expression iu terms of flT is changed
as opposed to a catalog rneaa place, thc safest method is to transform the.bser
(IAU, 1e76).
vation back to the epoch of the observation, on the FK4 systern
(or in the systcnr
that was used to produce the observed mean place). convert Step 4: Precess, usiug the FK4 preccssion constants, the position, aucl velocity
to the !.K5 svstiem,
and transform to the epoch and equinox of J2000.0. f r o m 8 1 9 5 0 . 0t o l 9 8 l J a n u a n t .
The transforrnation lor a fundamental catalog position is as follos,s:
Step 5: Apply the equinoxcorrection FK.l to FK5 at 198,1Januar)' 1.0 to the riglrt
Step 1: Form the position and velocity vector from the FK4 catalog ascension and proper motion in riglrt ascension.
dtrtai i.e.,
form the position and velocitv vector flom the right asccnsion, declination, proper Step 6: Convert the proper motions frorn seconds of arc per tropical century to
motions, paralla;i, and radial velocitv. seconds of arc per Julian century
 182 EXpLANAToRy
suppLEMENT SYSTEMS
REFERENCE
/ CELESTIAL 183
Step 7: Precess,using the FK5 precessionconstants,the position and velocitv a noninertial ftame (i.e., rotating hame of date)' The expressionfor P, usingthe
from 19E4January 1.0 l,o J2000.0. accumulated precession angle uotation, becomes

Step t: Apply space motion to the position vector from 19E4 Ja.nuary 1.0 to
J2000.0.
P[ep, e2, s] = lr
L"P
rrl
rl
+s0e)Qr(-G, -sG).
I | = Q:(-2,r,-saq)Qz(+O,r, (3.59-4)

In order to calculate Po and P. the equatorial precession a.ngles on the FK4

Step 9: Convert the position and velocity vector back to right ascensionand dec-
(Section3.214)and FKS (Section3.211)systemsare required,respectively,together
lination and proper motions in right ascensionand declination, and extract the
parallax and ra.dial velocity in the FK5 system. with their rates of changewith respectto time. The precessionrates on the FK4
and FK5 sYstemsare given bY
The Steps 3-8 may be representedby successivemultiplication by the matrices M1,
Mz, M:, l&, Ms, and ltlr on the position and velocityvector at Bl9b0.0 to produce
z=f"zlr,r)e=fie6',t'
a=fi a<n.'t (3.5e-5)

the position and velocity vector at J2000.0 where

ze=
a=fta<r,o e^=
ftu<t,tt !eo1r,t1' (3.se-6)

=' =Mo
Msr\&
', on,', (3.59-1)
il] [ll] l;i] The six matrices are defined as follows:
M1 Adds space motion between the standard epoch 81950.0 and e to the
and the position and velocity vectors r and i a,re expressed in au and seconds of
Dosition vector at 81950.0:
axc per century, tropical or Julian, as appropriate. The conversion of the units of

",=[; "l],
the velocity vector is doue by matrix M1. The followiag is a brief expla.nation of
the notation used: (3.s9-7)

Cs = 36524.21987817305 days, the length of the tropical centurSr;

where !s = c(JD(e) - JD(B1950.0)) / CB and c = url(180 x 3600) is the factor that
G = 36525 days, the length of the Julian centuryl
converts the proper motions to radians per century.
e = Epoch, e.g., 1984 January 1.0;
JD(date) = Julian date,JD(81950.0)= 2433282.4234sso5,JD(J2ooo.o)=,24.s1545.o, M2 Applies FK4 precession from 81950.0 to 6, to the position and velocity
a.ndJD(1984Ja.nuary1.0) = 2445700.5i iu 6-space:
I and O = Unit and null 3 x3 matrices.
The rotation matricesM2, l\[3, and M5 are formedfrom the e;, i = 2,3 matrices, M2 = PolB1e50.0,s,tl = Q3(-2, -i) Qz,le,+i)\Q3Geo, -cd. (3.59-8)

which are given by

M3 Adds the equiuox correction to the right ascension and proper motion in
right ascension at epoch e:
o,@,ir=lr#,fft,
.%,], (3.se-2)
M: = Q:(-8", -E), (3.5e-9)
where R1's are the standa.rd 3 x 3 rotation matrices, a.nd is a function of time.
/
The derivatives of the required rotation matrices are whereE = E50+ E(JD(€)- JD(B1950.0))/ CB, Eso= 0!525 and E = li275.

-cos{l l[4 Convertsthe proper motions from tropical centuriesto Julian centuries:
A J-sind 0 r f -sind cosd 0l
;-Rr(d)=l 0 0 0 l, *nrtOl=l-*rO -rinj ol. (3.s9-3)
I cosC0 -sindl oq L o o 0J ^"-[I o I (3.5e-lo)
"*= [o (crtcB ]'
In &space the notation for the precession matrix (Section 3.21) has to be mod-
ifed to include a further parameter s, where s = 0 when precessing from one inertial IttI5 Applies FK5 precession from e to J2000.0, to the position and velocity
frame to a.nother inertial frame, a,nd s = I when precessiag from an inertial fra.rne Gspace:
 184 EXPLANATOBYSUPPLEMENT 3 / CELESTIALREFEBENCESYSTEI\.,IS
185

Ms= P,-'Uzooo.
o,e,1l Step c: Remove the efiects of the E-terms of aberration to form r1 and i1 from

+&)Q2(-0e,-0/JQzGa,+2i.
= Q3(+e4, (3.59-l1.1 rr=ro-A+(ro A)ro,

Adds space motion between e and J2000.0 to the position vector at €: ir=io-A+(ro A)ro,
M6

',=l; -'i],
| -r.625s't1
whereA=| -o.rtltl I x l0 6 radians,
(3.59-12)
L-0. 138431

where tr = c(JD(e) - JD(J2000.0)) / Cr and c = z / (180 x 3600).

| +r:usl
A = I -11580| x to-3 p". tropicalcy, (3.s91-2)
L-or65el
The following two sections give complete algorithms for the conversion to and
a,rrd(ro A) is the scalar Product.
from J2000.0 using the matrix M and its inverse M-r. It should be noted that here.
and in the method given on pages B42 and 843 of The Astronomical Almanac,
Kinoshita's development of Andoyer's precession angles (Aoki el ol., 1983) have step d; Form the vector .' = and calculate the vector * = *.-
[:] ] [i]
been used. However, the numerical values for the vector components for correcting
the Bterms of aberration in Equation 3.591 2 have been taken from Yallop el al. (3.s91-3)
R=MRr
r1989).
Planetary positions may also be transformed using N{p, where l\4 is a 3 x 3
where M is a constant 6 x 6 matrrx:
matrix formed from partitioning M = MzM:MqMs (Ma has no effect on position)
into four 3 x 3 matrices and extracting the leading matrix. Alternatively, use the -0.01 I 182061
nseeszseiaz l -0.00000002710663,0.00000001177656
0.0048s79477+0.0000{1242395018
appropriate 3 x 3 rotational matrices from M2, M3, and M5, I - 0.00000m0006587
+0.9999374784-0.0000271765+0.00000002710663 +0.00000242397878
l + o . 0 1 1 1 8 2 0 -6 1 0
+o.0048s7e47e 177656 -0 00000000006582+0.00000242,1101
1997 +0.00000001
o.tno)27| 414 +0.999988 73
3.591 Conyersion of Stellar Positions and Proper Motions from the FK4 System |
+0.99994704 -0 01I 18251 o.00/85167
0.000ssr 0-238565 +0.415739
at 81950.0 to FK5 System at J2000.0 A matrix method for calculating the mean | +0.99995883 0.00002718
+o2385r4 0.002667 0.008541 +0.01118251
place of a star at J2000.0 on the FK5 system from the mean place at 81950.0 or | 0.00002714 + 1.00000956
+O.O12254 +0.002117 +0.00,185767
L-04r5623
the FK4 system, ignoring the systematic corrections FK5 FK4 and individual star (3.591-4)
corrections to the FK5, is as follows:
and set (x,y,2,i,j,,i) = R'.
Step a: From a star catalog obtain the FK4 position (a6, 66), in degrees, proper
motions (poo,1.u6o)
in seconds of arc per tropical century, parallax (16) in secondsof Step e: Calculate the FK5 mean positior (a1.61), proper motions (po,,p6,) in
arc, and radial velocity (vs) in km s-r for 81950.0. If n6 or v6 are unspecified, set secondsof arc per Julian century, parallax (r1) in secondsof arc, and radial velocity
them both equal to zero. (vl) in kms-' for J2000.0from

Step b: Calculate the rectangula,r components of the position vector rs and veloc- sindl =z/r,
c o so t c o s6 t = , r / r , s i n ( I lc o s 6 l= y / r ,
ity vector i6 from
x y --yj 1x. a(f +)4 - z(xi + yi)
f cosoecos6ol
= l1b,=
trol )
rt\/x'+Y'
r o = sino6cos6s
I I
I sin66 j
( 3 . 5 9 11-) v1 = @i + yy + zi) / (21.095r0r), |tt=1to/r, (3.s91-s)
-loo sin a9 cos 6s - cos c!0sin 60I
f /-.160
ro= +4."cosoocosdX* ttn u. + 21.O95vororo,
"o'. where r = tF;f;V If rrp is zero then v1 = vp.
| f;r .1
 186
3.592 Conversion of Stellar positions and proper Motions from the FKS
at J2000.0 to FK4 System at 81950.0 A matrix method for calcutating
place of a star at 81950.0 on the FK4 s;lstem from the mean place
Syst€m
the mearr
EXPLANATORYSUPPLEMENT
r 3 / CELESTIAL
F E F E B E N C ES Y S T E M S

Step 5: Calculate the FK4 mean (o1, 61), proper motions (po,,p6,) in secondsof
arc per tropical century, parallax (21) in seconds of arc, and radial velocity (u1) in
187

at J2000.0orr km s-r for B1950.0,as given in Section3.591Step e, by setting (r, ), z) = r/, (i, i, a) =
the FKS system, ignoring the systematic corrections FK,l FKS and
individual star i/and r= '6i7iV
corrections to the FK4, is as follows:

I n S t e p4 s e t ( r r , ) r , z r )= r i , ( i r , i r . 2 r , = i | , a n d , 1 = , @ + f i Z r , thel
Step l: From a star catalog obtain the FKS position (o6, 6e), in
degrecs,propol
motions (pa., p6o) in seconds of arc per Julian century, parallax (&l)
rn seconds of vr =(rrit +yrir + zrh) / (2l.o95nort), 1tt=1fo/rt. (3.s924)
arc, and radial velocity vs in kms-l for J2000.0.If rs or (u6) are
unspecified,set
them both equal to zero.
If z'o is zero then v1 = Ys.
Step 2: Calculate the rectangular componentsof the position vector
16 iurd veloc_
ity vectol ig from Equatiol 3.591 1.
3.6 REFERENCES
rrnl '1,..
S t e-p3 : F o r m r h e v p c t o r-n"" = l l a n d c a j c u l a t c t h c
v c c t o r = f r r _
f R, Those publications referenced in Chapter 3 are indicated by a f.
1,u., li, |t'"-'
t Abalakin, V., ed. (1985). Refraction TabLesof Pulkota Obsenatory,5th edition, Cen-
Rr= M-'Ro, tral Astronomical Observatory, Academy of Scienceof the USSR. Nauka Publishing
( 3 . 5 9 2 -)l House, Leningrad Section, Leningrad, USSR.
tAndoyer, M.H. (1911). "Les Formules de la Pr6cessiond'aprbs S. Newcomb" BrlL. Astron.
where M-l is a constant 6 x 6 matrix: 28. 67.
j Aoki, S., Guinot, t}., Kaplan, G.H., Kinoshita. H., \'IcCarth5 D.D., and Seidelmann.P.K.
+0.9999256795r{1.0t|I8t482s +0.otx8590039-0.00000242389840 (1982). "The New Defrnition of Universal Time" -,lsfron. Astrophys. 105,359.
| 0.m000002? I0544 -0.0000000t 1777.12
00111814828 +0.999e1718:19u.oo(N211i71 {.000000027105:t4 Aoki, S. and Kinoshita, H. (1983). "Note on the Relation Between the Equinox and
l..-0@185q0O40 O.o0t!,)o2423g2.702
+0.00uxr000006585
Lr.oorj.l:?t5s7+)ooo9881946 +O.0000{J0O Guinot's \on-rotating Origin" Celes. Meclt. 29. 335.
| I | 777.12 {.0qX10000006585 0.uJ({X12.12404995
+0 2tx50,r -0,rts6t,t +0.99ggt!j2 tAoki, S., Soma, trI., Kinoshita, H., and lnoue, K. (1983). "Conversion trIatrix of Epoch
l...000o5:l +0.0l El:15 +0.tM85852
-00112667 -0.ollt8l.t5
B1950.0 FK4-based Positions of Stars to Epoch J2000.0Positions in Accordancewith
+o.r)t2254 +0.99991613
l-0238500 0.L,LXJ(J?7|7 thc New IAU Resolutions" Astron. Astrophys. 128,263.
L +0.435730 -0.008541 +0.(rj2l17 0.ixx85852 _0.0{1002716 +0.9999668.1 Aoki, S. (1988). "Relation Bet*'een the Celestial Reference Systcm and the Terrestrial
(3.s922) ReferenceSystem of a R.igid Earth" Celes. Mech. 42,3O9.
Arias, E. F., Lestrade, L.-F., and Feissel,M. (1988). "Relative OrientatioD of VLBI Celestial
Step4: Include the efiectsofthe E-terms Reference Frames" in The Earth's Rotation and R.eJerenceFrarnes Jor C,odesg and
of aberratiol trs follorvs.lbrn s1 = 11/7.1
and Sr = il / 11 whcrc ,t = GeodynamicsA.K. Babcock and G.A. \\rilkins, eds. IAU Symposium 128 (Klurvcr,
Jiiii?t. Dordrecht, Holland), p. 61.
Ashby, N. and Allan, D.W. (1979). "Practical Implications of Relativity for a Global
Scts=st,andcaiculaterfronr=sr
+ A - ( s . A ) s , s , h e r cA i s g i v e l i n S t c p c o l . Coordinate Time Scale" Rad.ioSct. 14.619.
Section3.591.
tAtkinson, R.d'E. (1972). "\,lodern Aberration for Observations of any Date Astron. J.
77, 512.
set s= r/r and iterate thc expressiql
for r on<r..r tq.ic. u'til a co'srstcnt \.alue01. t Atkinson, R.d'E. (1973). "On the 'Dynamical Variations' of Latitude and Time" ,4slron
r is obtained, then calculate J. 78. 117.
fAtkinson, R.d'E. (1975). "On the Earth's Axes of Rotation and Figure" Mon. Not. R.
Astr. Soc. l7l, 381.
i=3r + A _ (s.A)s, (3.s92-3) Atkinson, R.d'E. (1976). "On the Earth's Axes of Rotation and Figure" Presentedat XVI
Gen. Assembly of IUGG, Grenoble.
whereA is givenin Stepc oI SectronJ.591. t Atkinson, R.d'E. and Sadler, D.H. (1951). "On the Use of Mean SiderealTime" Mon. Not
R. .4srr. Soc. 111. 619.

$
 V
'188
EXPLANATORYSUPPLEMENT
3 / CELESTIALREFEBENCESYSTEI,IS 189
fAuer, L.H. and Standish, E.l\{. Jr.. (1929). ,,Astronomical Refractron; Coinputational
Method for all Zenith Angles,' yale Universit1. Astronorny Department. Carter, W.E. (1988). "Corrections to thc IAU 1980 Nutation Series" BIH Antnal Rep. Jor
1987 D-105. Paris.
Backer, D.C., Fomalont, E.B.. Goss, \\r.M., TayJor, J.H., and Weisberg.J.M. (19S5). ,,Ac_
curate Timitg and Interferoneter positions for the l,{iliisecond pulsar Chapple, W.N,l. and Tullis, T.E. (1977). "Evaluatiort of the Forces that Drive the Plates"
1937+21 and J. Geophys.rRes.82, No. 1'1, 1967.
the Binary Pulsar 1913+16,' Astrorr. J. g0.2275.
Chauvenet, W., (sth ed., revised and corrected 1908), A Manual of Spherical and Practi-
Barbieri, C. and Bernacca, p.L., eds. (1979). European Satellite Astronetrg
(lnst. di As_ cal Astrornmy 2 vol. (Lippincott, Philadelphia), rcprinted (Dovcr Publications. Nerv
tronomia, Univ. di Padova, Italv)_
York, 1960).
Bender, P.L. (1974). "Reference Coordinate System Requirements for Geophysics,,
in On Corbin, T.E. (1978). "The Proper Nlotions of the AGK3 R and SRS Stars" in ,{4odern
Reference coordinate systems for Ear*t Dynamics B. Koraczek and G.
\\reiffenbach. Astrometrg F.V. Prochazka and R..H. Tucker. eds IAU Colloquium 48, (University
eds. IAU Colloquium 26. (Smithsonian Astrophys. Obs., Canbridge, NIA), p.85.
Observatorv,Vienna), p. 505.
Bender, P.L. and Coad, C. (1979). The ftse oJ Artificial Satellites
for Geod,esyantl Geo_ de Sitter, W. (1910). "On the Formula for the Con-rparisonof Observed Phenomcna of
dynamics Yol,. II (National Technical University, Athens).
',Equinox Jupiter's Satellites with Theory" Mon. Not fl. ,Astr' Soc. 71, 85
t Blackwell, K.C. (1977). position and motion duri.g 250 years, Mon. Not. R.. de Sitter, W. (1911). "On the Harvard Eclipses of Jupiter's Satellite IY" tr'Ion Not. R-
Astr. Soc. 180. Short Communication. 65p.
Astr. Soc.71. 596 and 602.
t Boss, B. (1937). Generar catarog oJ s3slp stars Jor the Epoch /95, publication No. 468 Dicke, R.H. (t9OO). "Averagc Acceleration of the Earth's Rotation and thc Visr:osit-"- of
(Carnegie Inst. of Washington).
the Deep N{antle" J. Geoplrys -Res.74, 25, 5895.
Boncher, C. (1985). "Tcrrestrial and Inertial Relativistic Refercnce Dickey, J.O. (1989). "Intercomparisons Between Kinernatic and Dynamical S]'stems" iD
Frarnes,, Grottpe dc
Reche,rceen Geodesie Spatiale (GRGS) Rep. No. 3 (Institur de Geograptrie ReJerence Frarnes in Astronomy and Ge-ophysicsI Kovalevskv. I.I \'Iuellcr. and B
National.
rarts). Kolaczek. eds. (Kluwer, Dordrecht, Holland), p.305.
Boucher, C. (1987). ',Defilition
and Realization of Terrestrial R.eferenceSystcrns ibr NIon Duncombe, R.L., Seidelmarrl, P.K , and Van Flandern, T.C. (1974). "Celestial Referctcc
itoring Earth Rotation', Institut de Geographic Natiotai/Service Systems Derir-able from Solar System Dynamics" it On ReJercnce CoodirLate Sys
de la Geoclesie ct
du Nivellement (IGN/SGN), paris No. 27, .1b9. tems Jor Earth Dynamics B. Kolaczek and G \!'eiffenbach, eds lr\U Colloquium 26
Boucher, C. and Altaminmi, Z. (Ig87). ,,Intercornparison of (Srnithsonian Astrophys. Obs., Cambrirlge. N{,4.).p 223
VLBI, LLR, SLR, and GpS
Derived Baselines on a Global Basis" Institut de Geographic
National/Service clc la Duncombe. R.L.. Fricke, W., Seidelmann,P.K.. and \\iilkins, G A. (1976). "Joint RePort of
Geodesieet du Nivellement (IGN/SGN). par;s No. Z,1SO. the \\rorkirrg Groups of thc IAU Commission '1 or Precession. Planetary Ephemerides,
t Bretagnon, P. (1982). ',Th6oriedu nrouvementde l,ensembledes planites. Units and Time-Scalcs" Trans lAU XYIB,56.
Sol.tion VSOp82. 'A DcterDlination of the
Astrort Astrophys. ll4. 278. tDysor, Sir F.W., Eddington. A.S., and Davidson, C (1920)
Brouwer, D. aDd de sitter, w. (193s). Deflection of Light by the Sun's Gravitational Field from (Jbsen'ations ntade at the
"oD the systenr of Astronomicar constants,, nzl1.
Astron. Inst. ol the Netherlandsg.213. Total Solar Eclipsc of \Iay 29. 1919" Plil Trans R. Soc LowJon 4220,291'
1981 (198s). (Bureat Intcrnationai de l,Heure, paris). Eichhorn, H.K. and Leacock. R.J. (19S6). Astrometric TechniquesIAU SymJrosirul 109
:I:.1nnua! !"e fttr
Jor 1g8S 1gaq. (Bureau Ilternational de l,Heure. paris). (Reidel, Dordrecht, Holland).
Y.!,1""*r,l"o
Rep. for 1986 \1987). (Burcau International de l,Heure. paris). t Einstein, A. (1956). Th.e Meantng oJ Relatittity (Princeton University Press. Princeton,
,o.,,!,1**t,
o"r:.:,_!"f NJ).
,:!n /or 196? (1988). (Bureau Intcrnarioual de l.IIeure, l,aris).
LanDon, W.H. (1978). "The ClassicaiArralysis t Emerson, B. (1973). "A Nltethotlof Calculating App:rrent Places of Stars" fi' Obs lluLl
of the Responseof a Lon6; l3aselineRadio
Interferometer', Geophgs.J. R. Astr. Soc.5J. 178.299.
bO3.
capitaine, N , w liams, J.G., and seid"rmanr. p.K. Eubanks. T.N'I., Steppe, J.A., and Speith, X'{A. (19s5). "Earth Orient:rtion Parameters
(1985). ..clarincations cronccrrirlB
the Definition and Detcrmination of the Celestial by Vcry Long Base Interferometr-v" BIH Annual Rep Jor 1981D-l9- Paris'
Eihemelis puiil e**r.
146,381.
Astraphus. ! Explanatoryl Srryplement to the Astronomi(:al Ephemeris and tlLeAmeTican Ephc tnerts nnd
capitaine, N. (19s6). "The Earth Rotation parameter: Nautical Almanat: (1961). (Her Nlajestv's Stationery Ofiice' LondorL,r"printed 19?l)
corrceptual and conventi.nal Del_ tFabri, E. (1980). "Advocating the Use of Vector-N{atrix Notatirxr in PrecessionTheorv"
initions" Astron. Astrophys. 162, 323.
Capitaine, N., Guinot, B.. antl Souchay, Astron. Astrophys. 82, L23.
J. (1986). "A Non Rotating Origin on the lnsral- Fanselow,J.L., Sovers,O.J., Thomas, J.B., Purcell, G H. Jr., Cohen. E..I , Rogstad, D H '
Equator: Definitions. properties ani Use,, Ccles.
^ .,.lrn"u"
Lapr,taifle.N. and Cuinot. B. (1988). ,,A
U"j. N. zaz. Skjerve, L.J., and SpitzmesseriD J. (1984). "Radio Interferometric Detcfmination o{
Non_Rotating Origin on the InstantaneousEqua_ Source Positions Utilizirg Deep Space Network Altcnnas 1971 to 1980" Astrc'x J'
tor" in TAe Earth's Rotati.on. and Reference
froln""jo, Geodesg and Geotlgnamit:s E9,987.
A.K. Babcock and G.A. Wilkins, eds. IAU
Symposium 128 (Klu;r, Dordreclrt, Hol_ Fedorov, E.P., Smith, NLL., and Beirder, P.L., eds. (1980) Nutation and the Earth's llo-
land), p. 33.
tation lA\l Symposium 78 (Reidel, Dordrecht, Holland).
Carter,.W.E., Robertson, D.S., and Fallon,
F.W. (1987). ,,polar Motion and UT1 Tirre Feissel, M. (rgsOj. "b"t..rlrin.tion of the Earth Rotation Parameters by the Bure'ru Irt-
Series Derived fronr VLBI Observations" BIH Annual Rep.
for 1986D_Ig. paris. ternational de I'Heure. 1962 1979" BulL. Geotlesique 54,81

tr
t"
 191
SYSTEMS
190 EXPLANATORYSUPPLEI\,IENT 3 / CELESTIALREFERENCE
the Earth"
in the Kinematics of the Rotation of
tFK3(1937). "Dritter Rrndamcntalkatalogdes Berliner AstrourischenJahrbuchs" I/ero.fent- Guinot, B. (1979)' "Basic Problems
Iichungen des Astronomischen Rechen Instituts zu Be,r'lin-Dahlem \r. 51. |ttTimeandt|rcEar|'h.sRotationD.D.Ir'lcCarthyandJ'D.H.Pilkington.eds'IAU
Holland)' p 7'
tFK4 (1963). "Fourth Fundantental Catalogue" Ver6ffentlichungen des Astronomischen Sy-po"iottt E2 (Reidel, Dordrecht'
S. tfgZ9t "lrregulalities of t'he Polar Nlotion" in Time and the Earth's Rotatnrt
Rechen-Instituts Eeid,elbergNo. 1O. Cti"*, 82 (Reidel' Dordrecht'
"""o.b. anJ J D H Pilkington' eds lAU Svmposium
t Fomalont, E.B. and Sramek, R.A. (1975). "A Confirmation of Einstein's Gencral Theory il.C-,tt,
of Relativity by Nleasuring the Bending of N{icrorval-eRadiatiol irt the Gravitational Holland), P.,279' Origin ofthe Lon-
onthe Terrestrial Pole of Reference'the
Field of the Sun" Astrophys. J. 199, 749. Guinot, B. (1981) "Comments Coorilinate S1stertLsfor EttTtlt
"-'*ti"a"., .nann ttt" Defirrition of UT1" in ReJetence
Fricke, W. (1967). "Precession and Galactic Rotation Derivod frorn !'undamcntal Proper.
IAU Colloquium 56 (Rcidel' Dor-
Motions of Distant Stars" ,4sfron. J.72, 1368. ?ir".-i* n nf Gaposchkin ancl B Kolaczek' eds
Fricke, W. (i971). "A Rediscussionof Neu'courb'sDetermination of Precession" ,4stron. drecht, Holland), P 125' H K Eich-
Astrophgs. 13, 298. Cdn"i g.'(tg8e) "Concepts of Reference Systems" i\ Astrolneh'ic ?eci'nigues
""';;;""";;R.: 109 (Renlel' Dordrecht' Holl,nd)' p l'
Fricke, W. (1974). "Definition ofthe Celestial ReferenceCoordinate System in Fundamen Lcacoci' cds IAU Svmposirrr'
Tirne Scales" in ReJc'renre Frarnes in As'
tal Catalogues" it On Reference Coordinate Systems for Earth DynamicsB. Kolaczek G"t;,;. (1989) "Astrolonical and Atomic
""';;^;;."; I I \l'eller a.d B Kolaczek' eds (Kluu'er'
and G. \\'eiffenbach, eds. IAU Colloquium 26 (Smithsonian Astrophys. Obs., Cam- GeophgsicsJ Kovalevskv'
b r i d g e ,M A ) , p . 2 0 1 . Dordrecht' Holland), P 351
-f Fricke, W. (1977). "Basic Material for the Determination of Precession and il der Atmosph:ire der
of Galactic H..;;';. (1924). "Berechnung der Ablenkung der Lichtstrahlen
"";;" Grundlage" Publikat;on der Sternuurtt
Rotation and a Review of }lethods aod Results" Vertiffentlichungen Astron. Rechen- ."f;"i" \leteorologisch-Phvsikalischer
Inst. Heidelbergno.28, Verlag G. Braun, Karlsruhe. in Kiei No. 13'
Fricke, W. and Gliese, W. (1978). "Progress in the Compilation of the FKs" in Modern zur Berechrrungder Ablenkurrge der Lit:lrtstrahlen
Harzer, P. (1924) "Gcbrauchstabellen
Astrometry F.V. Prochazka and R.H. Tucker, eds. IAU Colloquirrm 48, (l.lnir-ersitv fiit ii" B"obachtungen arn Grossen Kider \Ieri<lirnkreise"
in der Atnosphile d"t E'd"
Observatory Vienna). p. .121. \o 1'1'
Pubtikation cler Sternu'artc rn -Kiel
Fricke, W. (1979). "Progress Report on Preparatiorr ofFK5 for Commission 4" Prot:eedings J Argrrc, A N ' Kcnworthy' C N{'' tr{orrison''-L V'' and \{rrrray C' A
i H"r*J, C., Sutton, , sr:i'233'8e'
of the Seventeenth General Assembly" N{ontreal 1979, Trans. IAU XVIIB (Reidei. "-"ildtijl i.c..,,.^* n"ii' of3c27llB'
opticalPo'iti''rr" .*?:',':'n'"
"'rd Effects in Astronomi(i'l Timirr6 \leasurcnrents" ;1s-
Dordrecht, Holland), p. 65. (1986) "Relativistic
-Hellings, R\{
Fricke, W. (1979). "Nlethods of Compiling a Fundarnental ReferenceSystem" in Collo tion. J 91, 651 (see also 92 1'146for erratum)
qtium on European Satellite Astrornetru C. Barbieri ald P.L. Bernacca, eds. (Ist. di (1986) "Geodesy by Radio Intcrferometry:
H".ring,
""--ii f.,A.., Gu'inl, C R, aud Shapiro' I I
Res 91 l7l5'
Astromomia, Univ. di Padova, Italy), p. 175. *fi". Forced Nutations of the Earth" J Geophvs
t Fricke, W. (1982). "Determination of the Equinox and Equator of the FK5" ,4stron. ,,ls- "i,ft" - tht lAU 1980 \utation Series' BIH ArnuaL
c.)'*'iti-t
trophys. lO7 , L13. t Herring, T.A. (1987) "HerriDes
RePort for 1987 D-1O6'
Froeschl6, M. and Kovalevsky J. (1982). "The Conrrection of a Cataloguc of Stars with an Comptrtation of Angular Attnosphcric
Hohenkerk, C.Y, ancl Sinclnir, A T (1985) "The 63'
Extragalactic RefereoceFrame",4stroa. Astroph!,Js. 116. 89. ail Nuut"ol itr,'ono.- oflict: TechnittL Note
Refraction at Large Zenith e"gri"::
Gaposchkin, E.M. and Kolaczek, B. (1981). Reference Coordirate Sgstens for Earth Dy-
Royal Creenu i, h Obs'rvarorl
namics IAU Colloquium 56 (Reidel, Dordrecht, Holland). of Observ:rtions- Maclo with the Six-In(lr
Garfinkel, B. (1944). "An Investigation in the Theory of Astronornical Refraction" Astrun- I Hughes, J.A. and Scott, D K (1982) "Results Naral ODserrraforvSecond
y'iil"t;t^ oJ tl*: tlnited Stotes
Transit Circle 1963 1971"
J. 50. 169.
Series,XXIII, Pt. tll. 16i''
Garfinkel B. (1967). "Astronomical Refraction in a Polytropic Atmosphere" Astron. J.72. or 1992) Pttbl of the Lirrited Stat's
t Hughes, .1.A , Srnith, C.A ' antl l3ranh:rru' R L (1991
235.
Na.ual ObsenatorY (irl Prcss)'
t Gilbert, F. and Dziewonski,A.M. (1975). "An Application of Norrnal l\4odeTheory to the o! the Se'uetnh C;enerrrlAssenrblg
IAU (1948). "Comnission l Ephernelicles" Trartsactions
Retrieval of Structural Pararneters and Sour(:e llechanisms fronr Seismic Spectra' Carnbridge'
Trans. IAU VII,75, (Ca'nbriige University Press' .England) 98' (catrr
Phil. Tfans. R. Soc. Lond,on A278. I87. Assernbl'/'rians IAU vllr' 67'
Goad, C. (1979). "Gravimetric Tidal Loadirrg Conrputed frorl Integrattd Green's Func- IIAU (1952). \I'at$actions t7 tn" nll'nti i"*oui
bridgc University Press, Cambriclgc' England)
tions" 1fO.4-4 Tech. Memorand.um NOS NGS 22 (NOS/NOAA, Rockville. J\'ID). ol the Thirtc"tlth General AsstttL'
Green, R.M. (1985). Spherical Astronomy (Cambridge Universiry Press, Carnbridge, Eng- t IAU (1967) "comDission 4' llphernerid's" Froce"dings
land). 6fg tans. IAU XIIIB (Iteidel, Dordrecht' Holland)' .f
4' 8'-and 31 on the New Svstcnr
Greenwich Observations for 1898 (1900). "Refraction Tables Arranged for Usc at the t IAU (1976). "Report of Joint u"*i'i* "i-"tttissions
Astronomical constants" ''i'ilio"i"-"ll'" s'irteenth General Assemblu 1976Trans
Royal Observatory, Greenwich" Appendix I, by P.H. Cowell (Her Majesty's Stationerr-
IAU XVIB IReiJel. Dorrlr"cht llulland)
Office, London). Holland)' p 8'
t IAU (1980). Reports on AsLroTLo-i'f""" IAU XVIIIA (Reid.e\
Cuinot, B. (1978). Dept. of Geod,.Sci. r?ep. 280, LI. Mueller, ed. (Ohio State Uliversity, P:r.1recht'
'4stron' J 64' 149'
Columbus). p. 13. tt,*in, J g (lgsg) "standarcl Light-Time Curves"
 193
192 EXPLANATORYSUPPLEMENT 3 / CELESTIALREFEBENCESYSTEMS
'PrccessionNlatrix Based on IAU (1976) System of Astronomical
Jefireys, W.H. (1980). "Astrornetry with the Space Telescope" Celes. Mech.22. 175. i Lieske, J.H (1979)
Kaplan, G.H., ed. (1981). "The IAU Resolutionson Astronomical Constants, Time Scales, C o n s ra n r s " A ' t r o t t A " t r o p h u s . 7 3 . 2 8 2
Applied to Polar Motion' Rt:latiuity'
and the Fundamental Reference Rame" United, States Naual Obsen)atorA Circular ]^6J Ma, C. (1979). Very Long Baseline InterJerometry
(U.S. Naval Observatory, Washington). Geoiesy NASA Technical N{emorandum 79582 (Goddard Space Fliglrt Center'
and,
Kaplan, G.H., Josties, F.J., Angerhofer. P.E., Johnston, K.J., and Spencer, J.H. (i982). Greenbelt, IVID).
A Precise Celestial Refercnce Frama'
"Precise Radio Source Positions from Interferometric Observations', Astron. J. 87 Ma, C. (1983). "Geodesy bY Radio Interferometry:
570. EOS Tmnsact;on oJ tlrc American Geophysical Union 64' 45' 671'
T.A., Ryan, J W, Herring, T A , Shapiro' I I ' Core1" B E ' Hilrleregger'
t Kaplan, G.H., Hughes, J.A., Seidelmann, P.K., Smith, C.A., and Yallop, B.D. (1989). . F .Clark,
Ma,' HC., ,
A . E E , W h i t n e v , A R , K n i g h t , C A , L u n d q u i s t G L ' S h a f f r : rD
' 'B'
"Mean and Apparent Place Computations in the New IAU system. IIL Apparent, Rogers,
Bo' (l98ti) "Ra{io
Topocentric, and Astrometric Places of Planets and Sta6" Astron. J.97.llg7. Vu,ri"rrUi.g, NR., Pigg, J.C. Schupler. BR and Riinnii'ng
Kaula, W.M. (1975). "Absolute Plate l,{otious by Boundry Velocit5r Nlinimizations" -I. source Positions from VLBI' Astron J 92' 1O2O' 'EartL
(1987) Orientati{)n fnrIrr VLRI'
Geophgs.Res.80, No. 2, 244 Ma, C.. Himrvich, W., Ilallarra. A., and Kao. \L
BIH Annual Rep. Jor 1986 D-L1, Paris
t Kinoshita, H. (1975). "Formulas for Precession" Smithsonian Astrophysrcal Obseruatorg
Fr-ameof the NASA Crusta] Dvnamics
SpeciaLReport 364, (Smithsonian Astrophys. Obs., Carnbridge, It,IA). Ma, C. (1983). "The VLBI Cclestial Rclcrerrce
Geodesll un'd Ccodynam-
JKinoshita, H. (1977). "Theory ofthe Rotation ofthe Rigid Earth" Celes. Mech. 18,277. Proj'ect" in The EartlL's Rottttion and ReJerenceFratnes Jor
Wilkins, cds IAU Syrnposium 128 (Rcirlcl l)orclrccht'
Kinoshita, H., Nakajima, K., Kubo, Y., Nakagawa,L, Sasao,T., and Yokoyama, K. (1979). ics i.K. Babcock alrd C.A
"Note on Nutation in Ephernerides" Publ. Int. Lat. Obs. oJ trlizusawa XIl, 71. Holland), P. 73
Frarnestn Aslronotnu and
Kolaczek, B. and Weiffenbach, G., eds. (197{). "On Reference Coordinate Systens for. Ma,' C. (1989). "Extragalacti(: ReferenccFranes in ReJerence
eds (Klurvcr' Dordrecht'
Earth Dynamics" IAU Colloquium 26, (SmithsoniaD Astrophys_ Obs.. Carubridge, Giphyri"" J. Koval.,'sk1'' IL \luellcl and B' Kolaczek
MA). Holland), P. '13
Se.cularl[otion of Llrc PoLe
Kovalevsky, J. (1979). "The Refcrence Systems" in Time and, the Earth's Rotation D.D. Markowitz, \'V. and Guin,,i,, B , eds. (1968) C)ttnttnc:ntalDrilt
t'"-'
oJ the Earth IAU Symposiurn 32 (Reidel Dordrecht' Hollnrrd)
McCarthy and J.D.H, Pilkington, eds. 82 (Reidel, Dordreclrt, Holland), p. 151. and, Rototion
Polal hlotions rvith
Kovalevsky, J. (1980). "Global Astrometry by Space Techniques" Celes. Mech.22.753. Markowitz, W. (1976). "Conparison of ILS IPNIS BIH and Doppler
Grenoblc'
Kovalevsk5,',J. and Nlueller, I.L (1981). "Cornrnents on Con.!entional Terrestrial and Quasi- Theoreticali Rcp to L{U Conlm lg and 31 lAU Gen' Assembll
'lndepcndent Polar \lotions Opti<al ald Doppler: C h;rlcllel Un
Inertial Reference Systems" in rRefereaceCoord.inate Sgstem for Earth DgnamicsE.\1. Markowitz. \\'. (197d).
to IAU Comm. 19 alrd 31, IAU GeD Assembl)" llontro?rl
Gaposchkin and B. Kolaczck, eds. IAU Colloquium 56 (Rcidcl, Dordrecht, Holland). cettainties" Rep 'The Role of Non tidal
Mather, R.S., Colcrnan, R., Rizos, C , and Hirsch, B (1977)
Kovalevsky J. (1985). "Systbmes de r6f6rencc tenestres et c6lestes,, Bzll. Astronontiqut: Gravity Variations it tile Nrlaintenancc of Rcfereltce Systerrrs lbr Sccular Ceotlvnant-
d,e Obseruatoire Rogal de Belgique 10, 87. ics" UNISURV G 26, I 25.
ofCEOS-3
Kovalevsky, J. (1989). "Stellar Reference Frames" in ReJerence Frames in Astronong and Mathcr, R.S.. \'Iasters. E.G . and Colenrarr.R (1977) "A prclirtrinarl Anah-sis
Altimeter Data in the Tasman and Coral Seas' U\ISL]RV C 26' 27 )6
Geophysics J. Kovalevsky, I-I. tr{ueller ald B. Kolaczek, eds. (Klun-er. Dordrecht.
'On the Recovcr-r'of Gcodlnami<s 1rrJ"t6;rtiorr
Holland), p. 15. Mather. R.S. and Larden. D.R. (1978).
Kozai,, \. (1974). "Hybrid Systems for Use in thc Dynamics of Artificial Satellites', il On from Secular Gravitv Change" UNISURV C 29,11 22'
Reference Coord,inate Systems Jor Earth Dynamics B. Kolaczek and G. Weificnbarl, McCarthv, D.D. and l'ilkiugton, .I D.ll , uls (1979) Tint: ontl the'Earllr's nolotionlAU
eds. IAU Colloquium 26 (Smithsonian Asl,rophys.Obs.. Cambridge, \,lA), p.235. Synrpositm s2 (Reiclcl,Dordreclrt, tlollatrd)'
Space
Larden, D. (1980). "Some Gcophysical Efiects on Geodetic Levelling Netrorks,, Pror:ccd- McClure. P. (1973\. Dtunnl I'oLarllcttiott GSFC Report X 592-73-259(Clodrlald
lngs of the Second.Inte,rnational Sgmposium on Problems Rclated to the Retk:Jittition Flight Center. Grecnbelt, \lD).
B '
oJ NoTth Amencan Vertical GeodetxcNeturor,ts (Canadian Inst. of Sur.!€).iDg, Otfa\\.a)- Nlelbourne.$.I., Anclerlc. It.. Feissi. \1.. King. R.IIcCalthr" D D Smith D l'apler"
and Vicente. R. (11183)."Project \IEIllT standartls" Ilnited States Natnl Obstrtatory
t Lederle, T. and Schwan, H. (1984). "Procedurc for Computing the Apparerrt places of
Fundamental Stars (APFS) from 1984 Onwards,, Astron. Astrophys. ll4, 1. Circular 167 (U.S. Naval Observatolv, Wa-shington)
(Reiclel'
Leick, A. and Mueller, I.I. (1979). ,,Definingthe Celestial pole,, Monus georl,aetica4, 749. Melchior, P. and Yunri, S., ods. (1972) Rolntion oJ the Earttt l A'IJSvrnposiulrr48
Lestrade, J,-F., Requiime, Y., Rapaport, M., and Preston, R.A. (1988). ,,Rclatiolship Dordrecht, Holland).
Between the JPL Radio Celestial Referencc Frame and a PreliDrinary FKS Frarue" Melchior, P. (1978). Tidts oJ thc Pkutr,t Eatth (PergatnorrPrcss Oxfor'<l linglarrd)'
83'
j,n The Earth's Rotation and Reference
Fratnes t'or Geode.syand Ceoclynarnics A.I{. "n{, T.H. (1978). "Piesellt-dal'Plat{'lloti{)ns J' Gr:oplrusRr:s'
Minster, J.B. and Jorda|.
Babcock and G.A. Wilkins, eds. IAU Symposium 128 (Rcidel, Dordrccht, I{olland). B11,5331.
()uatriime
p.67. Molodenskij, M.S. (1961). "The Theory of Nrrtation and Dirrrnal Earth Tides"
Clom-
fLieske, J.H., Lederle, T., Fricke, W., and Morando, B. (1977). "Expressionsfor thc Pre- Symposium International sur les mar'6es lerresfres' Brrrxellcs' P Nlelchior'' ed
municatiors de l'Obscrvatorre Roval de Belgique, No ltl8' and S6rie G6oplrystque'
cession Quantities Bascd upon the IAU (1976) System of Astronornical Constauts"
Astron. Astrophys. 58, I. No.58.25.
 REFERENCE
3 / CELESTIAL SYSTEI\'{S 195
194 EXPLANATORYSUPPLEMENT
(1987). "Earth Rotation (UTOR) frorn
Morabito, D.D., Preston, R.A., Linfield, R.P., Slade, N{.A., and Jaunccy, D.L. (1986). Newhall, XX, trVilliams, J G., and Dickcy, J O.
"Arcsecond Positions for Milliarcsecond VLBI Nuclei of Extragalactic Radio Sources: Lunar Laser Rangirg" B1I1 ,,lnnual Rep. for 1986 D-29, Paris.
IV, SeventeenSources" ,4slron. J.92,546. Newton, I. (1968). "Experimental Eviderrce for a Secular Decrease in the Gravitatiolal
Moran, J.M. (1974). "Geodetic and Astrometric Results of Very Long-Baseline Interfero- Constant G" J. Geophys. Res 73. 12,3765.
metdc Measurements ofNatural Radio Sources" in On Reference Coonl,inate SEstems Nes/ton, R.R. (1974). "Coordinatcs Used in Range or R.ange-Rate Systems and Their Ex
for Earth Dynamics B. Kolaczek and G. Weillenbach, eds. IAU Colloquium 26 (Smith- tension to a Dynamic Earth" in On Refereru:e Coordinatc Sgstems for Earth Dynamit:s
sonian Astrophys. Obs., Cambridge, MA), pp. 269 292. B. Kolaczek and G. Weifenbach, eds. IAU Colloquium 26 (Smithsonian Asliroph)'s.
Obs., Cambridge, N{A), PP. 181 200.
t Morgan, H.R. (1952). "Results of Observations with the Nine-Inch tansit Circle 1913-
(1911). Contrcss International des EpherneridesAstrononriques tenu a
1926" Publication of the United States Naual Obseruatory SecondSeries, XIII, Pt. IIL tparis Conference
I O b l p r \ a l o i r p d o P a r i sd u 2 3 a u 2 6 O . l o b p r l 9 l l .
J Moritz, H. (1967). Dept. oJ Geod,.Sci. -Rep.92, (Ohio State University, Columbus).
Moritz, H. (1979). Dept. oJ Geotl. Sci. Rep.294, (.Ohro State University, Columbus). Perley, R.A. (1982). "The Positions, Structures and Polarizations of 40'1 CornPacl,Radio
Sources" ,4slroz J. 87, 859.
t Moyer, T.D. (1981). "Transformation from Proper Time on Earth to Coordinated Time
in Solar System Barycentric Space-Time Frame of Reference (parts 1 and 2)" Celes. tPorter, J.G. and Sadler, D.H. (1953). "The Accurate Cak:ulation of Apparent Places of
Mech. 23, 33 and, 57. Slars" Mon. Not. R Astr' Soc. 113,455.
tMueller, I.L (1969). Spherical and, Pract'ical Astronomy as Applied to Geodesg (Ungar, Prochazka, F.V. and Tucker, R.H., eds. (1978). Modern Asbometrll IAU Colloquium, 48
New York). (University Observatory. Vienna).
Mueller, I.L (1975). "Tracking Station Positioning from Artilicial Satellite Observations" t Pulkova (1956). Refraction tables oJ Pulkoua Obseruatory'lth edition, (Acadcrnv of Sci-
Geophgs Su|L)eAs2 243 276 ences Pressr NIoscow, Lcningrad).
Mueller, I.I., ed. (1975). Proceedings of the Geodesy/Soiid Earth and Ocean Phvsics Purcell, G.H. Jr., Cohen, E..1.,Fansclow. J.L., Rogstad, D.H., Skjerve, L..1.,Spitznesscr,
(GEOP) Research Conferences 1972 1971. Dept. oJ Geod. Sci. Rep.23l (Ohio State D.J-, and Thomas, J.B. (1978). "Current Results and Developnents in Astromctric
University, Columbrrs). VLBI at the Jet Propulsiorr Laboratory" in Llodern Astrornc:lrAF.V. Prochazka and
Mueller, I.L, ed. (1978). "Applications of Geodesy and Geodynamics. An Internationa.l R.H. Tucker, eds. IAU Colloquiun 48 (University Obserlatory, Vienna). p. 185.
Symposium" Proceerl.inlls oJ the Geod,esg/Sol'idEarth and Ocean Phgsics (GEOP) Purcell, G.H., Jr.. Fanselow,J.L., Thomas, J.8., Cohen, E.J., Rogstad, D.lI.. Sovers.O.J.,
Research ConJerenceOctober 2 5, 1978, Columbus, OH Dept. of Geod. Sci. Rep.280 Skjerve, L.J.. and Spitzmesser,D..J. (1980). "VLBI f4ea^suremerrts of Radio Source
(Ohio State Universitl', Columbus). Positions at the Jet Propulsion Laboratory" Radio Interferometry Ter:htnquesJor
Mueller, LI. (1981). "Reference Coordinate System for Earth Dynamics: A Prcview" in GeodesgNASA Conference Publication 2115 (NASA Scientific and Tech. Inlbrmation
Reference Coordinate System Jor Earth Dynamics E.M. Gaposchkin and B. Kolaczek, Office, Washington, DC), p. 165.
eds. IAU Colloquiurn 56 (Reidel, Dordrecht, Holland), p. 1. Reasenberg,R.D. (1986). "NllicrosccondAstrometric IniierferoDctry" in Astrotnetric TeclL-
Mueller, I.L (1985). "ReferenceCoordinate Systems and Frames: Corrceptsantl R.ealiza niqzes H.I{. Eichhorn, and R.J. Le:r.t:ock. eds. IAU Symposium 109 (Rcidei, Dordrccht,
tion" Bull. c€od. 59. 181. H o l l a n d ) ,p . 3 2 1 .
Reigber, Ch.. Sclrlr.intzer.P.. Nhreller. H.. and \{assmaur, F.H. (1987). "Earth Rotation
Mueller, LL, ed. (1985). Proceedings ol the InterrLational Conference on Earth Rotatxon
from Laser Ranging to LAGEOS: Station Cooldinates fron Lascr Ranging to LA
and the Terrestrial Reference Frame (Dept. of Geod. Sci. and Surveying, Ohio Stale
GEOS" t111,4nnuol Rep. for 1986 D 39. Paris.
University).
Robertson, D.S., Fallon, F.W.. and Cartcr, W.E. (1986). "Cclcstial Refcrcnce Coor.di-
Mueller, I.L and Wilkins, G.A. (1986). "Earth Rotation and the Tenestrial Reference
nate Systens: SubDrilliarcsecondPrecisio Dcmonstrated with VLIII Observations'
Frame" -lldz'.Space Res.6,9, 5.
Astron. .1.9L. 1456.
Muhleman, D.O., Berge, G.L., Rudv, D.J., Nieli, A.E., Linfield, R.P., and Standish, E.\4.
t SaastarnoinenJ. (f972). "lntrodu<1io to Practical Comgrl,rltion of ,,\sfrolromirrnlRefrac-
Jr. (1985). "Precise Position Measurements of Jupiter, Saturn and Uranus Systenrs
tion" Bull. Geod. 106. 383.
with Very Large Array" CeLes.Mech,.37.329.
ISAOC (1966). Smithsctniarr AstroplLgsrcalObseruatorg Star Cdaloq. Positions and Propcr
Murray, C.A. (1979). "The Ephemeris Reference Framc for Astrometry" tn Tirne antl the
Motions oJ 25t997 Stars Jor the Epoch and Equinor. o.f 1950.t7Smitllsonian Publicatrrltr
Earth's Rotation D.D. McCarthy and J.D.H. Pilkington, eds. 82 (Reiclel,Dordrecht,
4652 (Smithsonian Institution, Washington. DC).
Flolland), p. 165.
Schutz, B.8., Taplen B.D., ancl Earcs, R.J. (19S7). "Earth Ilotation fiom LAGEOS Lascr
t Murray, C.A. (1981). "Relativistic Astrometry" Mon. Nctt.r?. ,4sfr'.Soc. 195, 639.
Ranging" 8111,4zrnnl t?.ep.lor 1986 D 33, Paris.
t Murray, C.A. (1983). Vectorial Astrornetj"g (Hilger, Bristol, England). Schwan,H. (1986). .A New TechDiqrrefor the Analvtical Detcrmination of a Furdam€Dtal
l Newcomb, S. (1895). Elements oJ the Four Inner Planets and the Fund,am,ental Clon-
System of Positions and Proper Nlotiors" in AstrometTic TechniquesH-K. Eichholn
stozfs supplement"ne
to the Americarr Ephemeris and ltrautical Alnranac (USGPO. W-ash-
and R.J. Leacock, eds. IAU S1-nposium 109 (Reidcl. Dordrccht, Holland), p.63.
ington, DC). Schwan, H. (19E8). 'rconstruction of the Systen of Positiorrsand Proper NlotioDs in thc
Newhall, XX, Prcston, R.A.. and Esposito,P.B. (t98e). "Relating the JPL VLBI Referencc
FK1" ti Mapping fAe 5*y S. Ddbarbat. .J.A. Eddy, Il.K. Eichhorrr, aDd A.R. Upgrcn,
Frame and Planetary Ephemerides" in Astrometric Techniqu,esH.K. Eichhorn and
eds. IAU Symposium 133 (Kluwcr, Dordrccht, I{ollald), p. 151.
R.J. Leacock, eds. IAU Symposium 109 (Reidcl, Dordrccht, Holland), p. 789.
 196 EXPLANATOBYSUPPLEMENT SYSTEIVS
197
3 / CELESTIALREFERENCE
Schwiderski, E. (1978). "Global Ocean Tides, part l; A Detailed Hydrodynarrical Improvcd IAU SYsicm of AstroDonl-
h- +Van Flalr(lcrD.T C. (1961) "The Introducti(tr ofthe 'l'he Astronornial Ahnurtttl'
terpolation [,{odel, US Naval Surface \{'eapons Center TR-3S66 (U.S. Naval Surfa<:e t "-"i"ut and Referencc Frame i to
Con.t.oas Time S' ale'
Weapons Center. Dahlgren, Virginia). 1981(U S Govcrnnent Printing Oftice'
The Supplement to the AsttonomicaL Alnwnac
Scott. F.P. and Hughes, J.A. (1964). ',Cornputation of Apparent places for the Southern p 526'
Washirrgton.and HNI Stationer] Officc' London)'
ReferenceStar Program" ,4sfron. J.6g.368. ''The \\iater BalaDce of the Earth" Drexel lrtstitulc oJ
V-
'- ffyf.f."-", T E A (1956)
t Scott, F.P. (1961). "A tr{ethod for Evaluating the Elliptic E Terms of the Aberration,. T"chnolnggP , r t b l i e n l t o nisn ' l i n o t o l o t t y 9 ' 5 !
|17
Astron. J. 69. 372. p.
1i6ao1. "fia.f Corrections to GrNxlctic Quantities" NoAA Tech Rcp NoS
Vuni""f,,
Seidelmann, P.K. (1982). "1980 IAU Theory of Nutation: Thc Final Report of the IALr 11 (NOS/NOAA, Rockvillc, NID)'
83 NGS
Working Group on Nutation,, Celes. Mech.27,79. lnfluen<:cofthe Planets oll Nutation l Flfft'tts of
'"^'-pf"""".tl. (fsSi). "On thc IDdirect
VonJ*t
I.L (1967). "New Method for the Detection of Light Deflection by Solar cravit\. Czechosl S4' 184
t Shapiro, Perturbati,,n" in Lunar Orbit" Bull Astrort lnst
Srr.nce157.806. Influcncr: of the Planets olr \trtation. II Eflt'cts of
'""ir"r"o.u.l. 1t9aSl "On thc Indirect
t' V"r,illL
Smart, W.l,I. (79U). Tcxt-book in Spherical Astronomg.{th editiou (Cambridge Universirl. perturbations of the Earth,s orbit.. Bull. Astron. Irtst. czechosl.34.311.
Press, Cambridge, England). olRadio Sorrrccs V Positions
ma", C.Vf. and Johnston, I{ J (19?7) "Prccise Positions
Smith, C.A. (1986). "On Desiderata for Star Catalogs for thc Remaindcr of the Twentieth "--iiiu J.82,791. Err?rl.um,4stro?. ,,I84,1932.
sou.."" oD a Basclire of35knf' , stron.
Century: A Report on Catalog Work Now in Progress at the U.S. Naval Observatorv.. j.rl,r. (1929). "The Ticlal Nttotionsof a Rotating, Elliptical, Elastic and ot:r'anless
warrr,
";";;l;
in Astrometric TeclniquesH.K. Eir'hhorn and R.J. Lcacock, eds. IAU Synrposirrnr109 il.D.'th."i" (Dept of Phvsics Univclsitv of Colorado' Boulder)
(Reidel, Dordrecht, Holland), p. 669. j.lvf. (1981). -The Forced \utations of nn Ellipitical' Rotatilg Elastic and occan-
+W.fr.,
Smith, C.A., Kaplan. G.H., Hughes, J.A., Seidelmann,P.K., yallop, B.D.. and Hohenkerk. 705'
f less Earth" Geophgs l R' ' str' Soc 64' 'Reduction
C.Y. (1988). "l\leal and Apparent Place Computations in the New IAU S1.stcrn.l. (190a). T;blcs XXVI XXXIV of Tables for Transit Circle Ob
The TransforrnatiouofAstrometric C.italog Systems to the Equinox.l2000.0,,,4stron " *""ri"ti,"*"
W*iitgt"n
Norral Obscruatory' Second Scrics 4'
Publicotiorls oJ the Llnited slates
J. 57. 265.
Appendix II.
Solomon, S.C. and Slecp, N.H. (1974). "Sor)reSimple Physical \4odels for Absolutc platt: Report of theJAU/IUGG Wrrkitrg
'' '-A;;" G.A. and Mueller'.I l' (1986) "Joitt Sumrnarl'
Wilkins,
IVlotions" -/. Ceopl41s.Res. 79, 17. 2557. Svstem" in Hiohhghts
." the Rotatiort of the Earth arrd Terrestrial Referctrcc
Sovers, O.J., Edwards, C., Jacobs. C., Lan1.i, G., Lierver. K.. and Treuhaft. R. (1989). SlrnJrositrm ? (Reidel Dordrrrtrt Holland) p' 771'
oj ilrrono^n J.P Stings IAU
"Astrometric Results of 1978 1985 Deep Space )ietnork Radio Interfer-orrretl.:The 'Courrnelltson tho EITectof Adopting \erv
Williams, J.G. and \Ielboume. \\: G (1;82)
JPL 1987 1 Extlagalactic Source Catalog" Astron. J. gS.1617. E$th Rotntion anri Eartlt
Precession and Equinox Correctiorrs" it High-precision
t Standish, E.NL Jr. (1982a). "OrientatioD of the.IPL Ephenrr.rides,DE200/LE200, to the 63 (9a)' (Rcidel' Dordre<:hl Hol-
Moon Dynamics O. Cnlame, ed IAIJ Colloquium
Dynamical Equirrox of J2000.0" Astron_ Astrophys. 114,297.
land),p.293.
t Standish, E.M. Jr. (l9S2b). "Conversion of Positions and proper X{otions froD 81950.0 j.G., Dickey, J O. N{elbourne' WG, an<l Stanclish' E \'l Jr' (1983) "Utlifi(rur
Willlams,
to IAU System at J2000.0" Astron. Astrophys. 118,20. of lAc syr posia'
tion of celestial and Tcrrestrial coorrlilrttte Slsterns" Proccrrlings
Stolz. A. and Larden. D.R. (1979). ,,seasorralDisplacementand Deformation of tle E:rrth (Dept of Geod Sci and Survelilg' Ohicr
IUGG XVIII Gcneml Assembl)" Hambtrlg 2
by the Atmosphere" J. Geophys..l?es. E4. E}11,6185. State Univc6ity), p. {93
t Stunrpff, P. (1979). "The Rigorous Trcatmcrrt of Stellar Aberration and Doppler Shift. Fra'rnesin l hc Plan-
Williams, J.G. ,.na iio,tai"l, E.I'I Jr. (1989) "D1'namical Rcfercrrce
and the BarvccDtri(: I{otion of the Earth,' ,4stron. Astropl41s. 7E,22g. and Gutpltllstcs
etary and Earth N{oorr Svstems" in Re'Jt]^cnceFrames in Astronomy
t stuntpff, P. (1980a). "o' the Relatio^ship bctween classical and Relativistic:Theorv of J. Kovalevsky, I L Mucuer and B K"la<'zck, cds (Kluwcr' Dorclrccht'
l:th]td)':::l;llitta
PhYsical
Stellar Aberration', Astron. Astrophys. 84, 257. Willis, J.E (194i) "A Dct'ernination of Astrotlomical Refrat:tion fronL
t Stumpff, P. (1980b). "Two Self-Consiste't Fortran Subrorrtinesfor the Comprrtation of Transactions of the Atnerit:an Geophgsia Lrrrion part II' 132'1'
tlre Earth's \Iotion" ,4stron. AstroplLys.Supp. Ser. 41, l. ..Ephcnr.ri" Tinre' Relatir'itr,. ;urd the
Winkler. G.\I,R' and Van Flandern. T.cl (1977)-
f StuInpff, P. (1985). "Rigorous Treat'rerrt of thc Heliocentri. \Iotion of Star:i-- .,lst?.o,. Problems of Uniform Tinre in Astrononry" Astron' I E2' 8'l'
Astrophgs. 144, 232. (Acadernic Press' New
fWoolard. E.W. and Cle[ronce, G.NI (1966). Splrcrical Astronorng
Stumpff, P. (1986). "R.('lativistic and Pcrspective Efects in prooer Motions ;urcl Ratli,rl York and Londorr).
Velocities of Stars" i.t\ Relatil)itg nt Ctic"tio,l Mechanics tttrl Astronomy .1. Kovalesky Prirltitrg Olfi(:c'
tYallop, B.D. (1984). in flre ,.lstro rnmicdl Ahnltuac l96l (U S Governrnent
and V.A. Brumberg, eds. IAU Symposiurn 1l Washington, and HM Stationery Office, Lorrdon), pp L116 B'11'
, (Re el, Dordrecht, Hollarxl), p. 193.
and Seitleltnann'
tTatr, L.G (1981). CoDLputational Spherical Astronomg (Wilcy, New york). tYallop, B.D; Holenkerk, C.Y., srniti, c.,q...Kaplan, G H, Hughes'J A '
I The Astronomical Ahnanac (19a4). (U.S. Government printing Office, li'o-rhingtol. and Pla(:c ComPutatiols iu the New IAU Svstem'
PK. (1989). "\'leaD and Apparent
HM Stationery Office, London). grgso o to FKs J2000'0 Using
ll. Transformation of lt".,t !iu, Placcs frorl Ff+
Van Altena, W. (1978). ,,Spacc Telescope Astrometry,, itt Moilcrn Astrometry F.V . proc- Matrices in 6 space" ,4slron. J.97,274
Astronomical Lat-
hazka and R.H. Trrcker, eds. IAU Colloquiurn 48, (University Observatory. Vienna), tvtt"ti"'v i iintoi-;;" tt Diurnal Free Polar Motion Derived from
E'P' Fedoto't '
D. 561. tiude and Time observationsl in Nutation and the Earth's Rotation
198 EXPLANATORY
SUPPLEMENT
M.L. Smith a.ndP.L. BeDder,eds. IAU Syrnposium7E (Reidel, Dordrecht, Holland).
P.59.
Zhu, S.Y. and Mueller, I.i, (1983). ,.Efiectsof Adopting New precession,Nutatior and
Equinox Correctionson the Terrestrial ReferenceFlame,, BtlL Geodes.i7,29.
CHAPTER
Zvetev, M.2., Polozhentsev,D.D., Stepanora,E.E., Khrutskaya, E.V., yagudin, L.I., and
Polozhentsev,A.D. (1986). ,,On the SRS Catalogue',in Astrornetric Techniques H.K.
Eichhorn and R.J. Leacock,eds. IAU Symposium 109 (Reidel, Dordrecht, Holland).
o.691. Coordinates
Terrestrial
of the Earth
andthe Rotation
by BrentA. Archinal

INTRODUCTION

The task of establishing or defining the terrestrial coordinates of a point is inextrica-

lly linked to establishing the rotation of the Ea.rth over time. The definition of any
fiprrestrial reference coordinate system (TRS) is given by establishing a celestial
reference coordinate sysiern (CRS) and a suitable tra.nsformation between them'
This transformation is what is generally cafled Earth rotation (ot Earth orienta-
tion), including the various changes in the orientation of the Ea.rth's axis in space
and internally, and the rotation (a.nd rotational variations) about that axis' (This
latter rotation is also often defined as Eat'th rotation-both uses of this term will
be made here with the meaning being given from context.) One could connect the
celestial and terrestrial coordiuate system at any given instant using three Eulerian
angles to deflle Eaxth rotation However,
"g.ngles, atrd then form a time series of such
fbr historical reasons (mostly in order to associate the motions of the Earth with
physical processes a,rrd to allo$/ for successive approximations of the motion)' in
practice these angles a,re expressed iu terms of precession, nutation, polar motiorr'
and rotation about the Earth's axrs.
Before we proceed into any explanation of terrestrial coordinates, it is also nec-
eBsaxy to distinguish between a (celestial or terrestrial) "reference system" and a
coor-
fteference frame." A reJerence systern is the conceptual definition of an "ideal"
$lnate system, based on some abstract principle(s). For example, an ideal terrestrial
would be one in which the Ea,rth's crust shows only deformation with no
or tradrslations. A conuentional reJerencesYsternis one where the model
to define coordinates is given in detail; that is, speciflc model choices have
ma.de. The realization of such a converrtional reference system, using actual
199
 tF-
200 EXPLANATORYSUPPLEI\,4ENT
4 / T E B B E S T R I AC
OF THEEARTH
SN D T H E R O T A T I O N
L O O R D I N A T EA
201
observations, adopted station coordinates, etc., provides a convcntional re.ference
(Kovalevskyand N{ueller.1981;Ntlueller,1988;Koralevskv et cl., 1989).
/rame Topography
This chapter initially will provide the basic concepts of various adopted tcrrcs_ Geop(throughp)
trial reference frames (mostly based ol the use of rcference ellipsoids), and of ther Geoid
Conventional Terrcstrial Reference Coordinate System (CTRS) as it is currentlv Ellipsoid
realized by the International Earth Rotation Service (IERS). Also included is irr-
formation on the actual transforrnation of coordinatcs from onc frame to another.
This will be followed by information, rnostly extracted from the IEI|S Standard::
Figure 4.21.1
(McCarthy, 1990), regarding the Earth's gravity field (to a limited extent), ti.les. surfaces
TheEaith'ssurfaceandreference
and plate tectonics. Methods for monitoring the Earth's rotation (past, present, and
future) are described and information on the IERS is provided. Finally, a summary
of determinatiotrs of historical variations in the rotalion of the Earth's axis is giverr.
An overview of Earth rotation and its carrscs in general has not been gil-en a1rdyet is still a fairly simPle rnathematical surface. (A rnore complicated surfact:.
becausespace is lacking for such a cornplete treatment, and the arnple coveragefrorn such as a triaxial ellipsoid, l'ill not greatly inprove thc rr:preseutation, but u'ill
this vie*'point by Teisscyre(1989). Nloritz and Mueller (1982), Lambeck (1930). greatly complicatr: the rrrathematics involved in using it ) Such a reference srrrface
and NIunk and \,IacDonald (1960). The rnotion of the Earth's celestial pole due lo has proven useful for relcrencing the position of points and mcasllrements on or ncar
precession and nutation has already been covcred in Chapter 3. (llcaders scckilg the Earth's surfacc. Refercnce ellipsoids ma1' be eithcr local ot Ien'estrial. depcn<lilg
more detailed information on thc subjects of gravity, gcodesy, geooeuc asrronom\-. olr whether they are an approxintrtion for a given area or tln] ettire E:r'rtll'
and terrestrial refereuce systems should consult standard texts such as N{uellcr anrl Another typc of hlrpothetical surface that proves useful for refercnciug obser-
Rockie, 1966;Heiskanenarrd \,Ioritz, 1967; NIueller. 1969; Bomford, 1971; NIoritz. vations is a geopotential srlrface, or geop, a surface of equai gravitl- potential (In
1980; Torge, 1980; \'Ioritz and NIueiler. 1987; Kovalevsky et r:1.,1989.) NlcCarthr. geodesy, grat,ity includes both the effects of gravitational potential and potential
(1990) also provides further informatiorr on accepled models of ef{ects on various due to centrifugal force.) The gravity vector is perpendicular to the geop at e\€rv
space geodetrc sysl,ems. point. Such a definition is quitc natural when nost instruDlents "peratc'l on tlre
sudace of the Earth alc sensitivc to local gravity (through the use of a pltrnb bob,
bubble level, pool of Xlercury. ctc.). It is also critical for anv hydrologic purpose'
4,2 TERRESTRIAL
COORDINATE
SYSTEMS where anv water surface closely follows a geop, and &'hcre waier florvs "dorvnhill"
from a geop ol higher potential to orre of lower potcntial A parti'rtJrr geop of
In this scction the lundamentals of various tcrrestrial coordinate syslrcmsarc pre- great irnpotance is the geoid, a geop \vhose potential is definr:d to l)e th'lrt oI thc
sented, covering both the various tvorldwide and common local svstems used to Earth at sea level. N4any dillerent dcflnitions have bcen givt:l in the past fi)r the
deflrre the coordirrates of an objcct on or rrear the Earth's surface. It includcs th<: geoid, perhaps the best conceptualll' being an imaginary "strrfact:coirrcidilLgtith
neccssary explanations of the common abstract representations of the Earth's srrr mean sea-level in the occans, and lying un<ler thc land at the lr:vel 1o lr4rich the
face, such as with the geoid or various cllipsoids. It encls rvith irrformation on thr: sea would reach il a<lmitted by small frictionless channels" (Bonfor'<l' 1971) Vari-
current Conventional Terrestrial Refercnce Coordinate Systcur. as it is realized b1: ous other dcfinitioris are given in Heiskanenand N'loritz (1967, pp 48 50)l N{oritz
the IER.S Terrestrial Coordinatc Svstem. (1980, p. 6); and Torge (1980, pp. 2 3). I practice, the avera.gegravity poterrtral
of the Earth at sea level is deterrnined ernpirically (norvadays usuallv via a'ltirrtet-
ric measurements of the world's oceans by satcllite), ancl tlte sur-faceof the samo
4.21 The Figure of the Earth and the Reference Ellipsoid
potential is by definition the geoid. Thc geoid srrrface is then elTectively the same
Most tcrrestrial coordinate systems for rucasuring positions near the Earth's surface surface usuallv referred to a rnean sed leu.'1.
use as their referencc an abstract surface near that of the Earth. A best-fitting The overall situation is shorvn in Figure 4.21 1. I'hc obscrver is at point p
rotational ellipsoid is rnost commonly assuIrredas an approxirnation to the surface on the surfaceof the Earth (topography). A rcferenceellipsoid (parficularll a local
of the Earth. A rotational ellipsoicl is used, sincc it represents the shape of the one) approximates the gcoid and lies generally below the topography A geop passes
Earth far better than a sphere, because of the flattcning of the Earth at the poles. throughp, but thc geoiJ itself lies close to the refcrence ellipsoid. Note that although
 202 EXPLANATORY
SUPPLEMENT 203
Refercnceellipsoid
af,isof rotation

d Defle.ctionof vertical

Tangentto
ellipseat p'
/"
equatorialradius

Figure 4.22,1
Geocentricandgeodeticlatitude
Figare 4.D.2
Astsonomiclatitude atld lorgitude

the geop and geoid surfaces are not the sarre as that of the topography, there are
some similarities since tbe mass of the topography influences the local gravity. Also
geodetic
shown in the ffgure arc lhe geoid und,ulation (N), which is the height of the geoid fhe ellipsoidal equator and line fron p to the center of the ellipsoid. The
above the ellipsoid, the ellipsoidal height (h), which is the height ofp above the latitude 4 is the angle between the ellipsoidal equator and the normal from p to
ellipsoid, and the mean sea leael height ot or.thometric height (H), which is the the ellipsoid (at point p/); a and b are !,he serni-maior and semi-minor axes of lhe
height ofp above the geoid. Note that: elipsoid, respectively;f,the ffattening,or I //, the inuerseflattening, is often used'
with
, a-b (4.22-r)
N=h-H. (4.21-t) a

This relationship is approximate since Il is actually measured along the curved Fhe geocentric longitude ()) is defined by the angle between lhe relerence (or zero)
v€rtical or plumb line through p, whereas i and N a.re measured rmeridian a.ndthe meridian ofpointp (andp/), measured eastwa.rdaround the Earth
along a straight
line through p perpendicula.r to the ellipsoid. ' hom 0' to 360' (IAU, f9$, p. Z). The geodetic longitude will be the same as the
'geocentric
longitude if the reference ellipsoid has been chosen with the same axes
4.22 Geocentric, and reference meridian as the Earth this is assumed to be the case in further
Geodetic,and Astronomical
Coordinates here. The astronomical coordinates of point p a,re determined relative to
Va.rious coordinate systems exist for measuring longitude and local vertical direclion (or zenith) at point P (seeFigtre 4.22.2, where z indicates
la;';ilude, Geocen-
tric coordirates a,re determined from the center of the Earth; geod,etic direction of the zenith and z/ the direction of the ellipsoidal normal)' A vertical
coordinates
are determined relative to a reference ellipsoid; and astronomical (or geogmphic) plumb)
rb) line
line passing through p will
passing through the Earth's
iutersect the
will iutersect equatorial plane
Earth's equatorial plane with
with I
coordinates a,re determined relative to the observer,s local vertical. anglee O,
O, the
the astronomical latitude ofp.
astronomical latitude ofp. Astronomical longitude A is measured
Astronomical longitude measured I
Figwe 4.22.!
illustrates the diflerence between geocentric and geodetic latitude, once rgle between
angle between the meridian plane
reference meridian
the reference plane and plane including
the plane
and the including the local
the local I
aga.in for
a point p on the Earth's surface. The geocentric latitude /, is the ansle between thmugh p and perpendicula.r to the equatorial plane. Note ,n* ,0" t"t,

J
 204 EXPLANATOBY
SUPPLEMENT OFTHE EARTH
ANDTHEROTATION
TEBBESTRIALCOORDINATES 205

Figure 4.22.3 Ftgrc 4.22.4

Deflectionof the vertical on a unit sDhere Geocentriccartesiancoordinates

vertical does not necessarilSrpass through the Earth's axis. The angle between the
local vertical direction and ellipsoidal normal through p is known as lhe d,eflection
of the aertical0. The deflection of the vertical is often broken into two orthogonal
components (see Figure 4.22.3, where z indicates the direction of the zenith, and z/
the direction of the ellipsoidal normat), one { being in the meridional direction, the
level or geoidal height 11 have already been mentioned. The geocentric radius r,
other 4 in the orthogonal ptrime verii.cal direction. Relations between the geodetic
is also shown in Figure 4.22.4. "fhese heights are associated with the geodetic,
and astronomical longitude and latitude and these components of the deflection of
astronomical, and geocentric systems, respectively.
the vertical are
In addition to these s;rstems of latitude, longitude, and height, it is common
to e:rpress the position of a point on or near the Earth in a right-handed geocen-
sin@= cos4311q- 6',' (4.22-2) tric cartesian coordinate system (Figure 4.22.4)- Tbe coordinates of a point p are
sin4 = css4'in1'1- ^'' (4.22-3) expressed as a triplet including X, I, and Z coordinates. Here the direction of the
Za)ds is that of the rotational reference ellipsoid, and the X-axis is perpendicula"r
Using small-angle formulas, these can also be given as ',\o
that through the reference meridian. The l-axis is of course perpendicular to
1tf,e X- and Y-axes in a right-ha.nded sense. As with the other systems, the origin is
€=a-d, (4.224) {ideally) the center of the Earth. Conversion from geocentric to geocentric cartesian

**
loordinates is given by
? = (A - l)cos d. (4.22-s)

As to the meaaurement of the ,,height" of point p, several types of ,,height,' lxl ft l cosd'cosI I
measurements are possible, referred either to a nearby surface, such as the geoid
or ellipsoid, or to the center of the Ea.rth. The ellipsoidal height ft, a.nd mean se&
l"l
tz )
=rl
I
c o s d ' s i n )l .
sind' I
 206 EXPLANATORY
SUPPLEMENT OF THE EARTH
ANDTHEROTATION
COORDTNATES
TERRESTRIAL 207
Conversion from geodetic to geocentric ca,rtesian coordinates is given by
v = ( P i- O + - @ i + Q ) + , (4.22-17)
lt - 1
a=)fre+v+el. (4.22-18)
f xl f rN6+tr)cos4cos.\'l
=rl (Md+r)cosdsin.\ (4.22-7)
lY | I,
Lz ), | 1 1-re 2 y a l r + l y s iJn 6 (4.22-19)
le.r-:s]-c
where e is the eccentricity of the ellipsoid
, the latitude and ellipsoidal height are computed from:
'/F=6
,trr=, (4.22-8) , I r-r I
@= a r c m nl a , r L , l , (4.22-20)

and N4 is the ellipsoidal mdius of cutature in the meridian, and is given by h=(r - at)cos6+k- b)sinf. (4.22:21)

/;--------;---:- ' (4.22-9) If D < 0, e.g., if less than about 45km ftom the Earth's center, the following
vt -e-s'n Q eouation should be used for v in order to avoid the use of complex numbers:

Conversion from geocentric ca.rtesian to geocentdc coordinates is given by

"=znz:F*,
{} -""" [3t-a-11
(4.22--22)
1
6'j
)l=l 1w-'ffi
t^-,,
I
I (4.22-10) To obtain the proper sign (and solution of the fourth-degree polynomial since up
, )o L^,@iF;V ) to four solutions actually exist), the sign of Zr should be set to that of z before
beginning. Borkowski also notes that, of course, this solution is singular for points
Note that d'= sin-r Z may also be used to determine d,, and that these functions at the z-axis (1 = 0) or on the Xl-plane (Z = 0). Additionally for points close to
a.reevaluated such that -90. < d (o. d,) < +90., and 0. < ) < 360.. those conditions, some roundoff error may be avoided and the accuracy improved
Conversion ftom geocentric cartesian to geodetic coordinates is not as simple slightly by replacing the value of v with
a matter, however. This problem has long been an interesting one for geodesy.
However, Borkowski (1989) has recently reviewed the various techniques used to v 3+ 2 Q (4.22-23)
solve this problem, and has proposed two solutions, one of which is exact and is 3P
reproduced here. It is based on using an expression for the reduced latitude in a
Finally
solvable fourth-degree polynomial.
) = arctan;. (4.22-24)
Several intermediate values a,re computed in order to perform a solution:

r=vx.+I., (4.22-rr) 4.23 LocalCoordinateSystems

-.)
F_Loz-@--p')l (4.22-12)
at Ma.ny possibilities exist for establishing local coordinate systems, either systems
- lbz + @2- b2\l designed for a specific observer or systems designed to cover a given local area' Most
lr = _-=__-, (4.22-t3)
ar systems specific to a.n observer are horizon (or altitude and azimuth) coordinate
systems and a "standa,rd" one is described below. Coordinate systems designed to
P = ;1EF+ l), (4.22-14)
J cover a localized area are usually based on a "map projection" coordinate system,
such that the coordinates of a point relative to the reference ellipsoid are projected
Q = z ( E- P l , (4.22-r5)
Qnto a pla.ne upon which computations can be done with relative simpliciiy (as
o=tr +A, (4.22-16) to working on the ellipsoid itself). Some examples of the more commonly
 208 EXPLANATORYSUPPLEMENT
V d / TEFRESTRIALcOORDINATESAND THE RoTATIONOF THE
EARTH 209
and the azimuth will have
used of these types of systems are given, such as for Universal Transverse Mercator the object is below the horizon and positive il above),
Instead of this spherical coordinate systern,
coordinates and (in the U.S.) state plane coordinates. an angular value of from 0' to 360".
system is also used, with the
a left handed three-dimensional cartesian coordinate
positive north, and the y-axis positive east (through
4.231 Horizon System Depending on the author, slight variations exist in the def_ z-axis positive upl the x-axis
inition of any Iocal horizon coordinate system. However, a more or less ,,standard,, 90" azimuth).
system is described as follows. A further variation of this system is to use the zenith distance z instead of
'fhe zenith d'istance is simply the complement of the altitudc, i e ,
First, any such system must have at least two reference planes. The primarv the altitude.
the advantage of always beiug
reference plane here is the local hori.zon, and the secondary is the locol meridian the angular distance of p2 from the zenith z. It has
(seeFigure 4.231.1).The horizon is the plane perpendicularto the observer,slocal positive, wiih angular lalues of from 0o to 180'. It is computed from altitude as
vertical, while the meridian is the plane perpendicular to the horizon that passes
through the celestial poles. The direction of any point pz (from the observer at z=90' - a, rzl)ll-l\

pr ) in this system is then given by its altitude and azimuth. If we imagine a plaue
passing through p1 and the zenith z, the angle between the horizon and the point where the z and 4 are in degrees.
p2 in this plane is the altitude (alno sometimes known as t]rre eleuati,on anglc) a,
and the a,nglebetween the meridian and this plane measured clockwise from north. 4.232GridCoordinatesystemsForcomputationalcorrvenience'iliisd)nrmonto
as viewed from above, is the azimuth A. (Some users in the past, pa,rticularly for represent sections of the curved surface of the Earth lvith a plane. various systems
geodetic purposes, have defined azimuth as being measured clockrvise from south. the
have been devised to do this, all based on given transformation equations fronr
This practice is no longer much in use but users of old observations should be arvarc sphere or ellipsoid to the plane. These transformations are, in fact' map projcctions,
of it.) The altitude wilt have an angular value of from -90" to +90" (neeative if in manv casesthe same as those used for mapping even large segrnelts of, or all of
the Earth. Conformal projections are usrrally used so that directions on the surface
being mapped arc prescrved on the grid. The coordinate systems based on such
map projections are usually termed grid coordimtte syslems The rnost l'idely used
''f
are the Universal Transverse Mercator, and the state plane coordinate syslems
the United States. Obviously many other systems exist for various countrjes and
to celestial pole
Direction of the areas, but the basic principles (and many times the projection formulas, rvith the
plumb line
exception of constants) are thc same
(to zenith)
For any grid coordinate system, several firndamental transforrnations and quan-
tities are of importance. The firndamental translormations are the direct converslon
fiom geodetic coordinates (S, )) to grid coordinates (X, i')' and thc inrerse tratrsfor-
Meridian plane
mation lrom grid coordinatcs to geodetic coordinates. Other quantities of interest
include: (a) the convergence of the mericlians, i.e., the angle betwt'en grid north
and true north on the grid (1); (b) the angle on the grid between the geodesic(the
shortest distance between two points on the ellipsoid) and the straight line between
) (north)
them (known in some of the referencesas "t ?"'); (c) the scale at a givcn poirrt (k):
-{ (east)
and (d) the scale factor alolg a line. Equations for the transformatiorts atrd scale
are given belon, (based on Snyder (1983)); equations for the other quantities are
available from the other references,particularly Thomas (1952)'
Fundamental quantities and expressions rrsed are summarized as follows:

y'r= geodetic latitude of Poittt;

Figure4.231.1 -\ = geodetic longitude of Point;
Altitudeandazimuth d'= foot point latitude (see definition below);
 SUPPLEMENT
EXPLANATORY
210
,\o = longitude of the origin (and the central meridian for the transverse Mer-
cator grid);
conformal conic grid;
dr, dz latitude of the standard parallels for Lambert
=
- A6) (in
A) =difference of longitudc from the central meridian (A) = )
radians);
a semi-major axis of the ellipsoid;
=
b = semi-minor axis of the ellipsoid;
"
e = c c c e n t r i c i l yo f t h e e l l i p s o i d1 " 2= o - - ! f

e' = secondeccerttricity of t he ellipsoid ,"u = {--!- ),

No = ellipsoidalradius of curvature in the meridian tr, =

l, !Hh)

Fortran programs for these transfornlations are available from the U' S Na-
tional Geodetic Survey (NGS, 1985).For rnore information write National Oceanic
and Atmospheric Administration, Natiorral GeodeticSurvey (N/CG174), Rockville,
MD 20852.

4.233 Universal TransYerseMercator and Universal Polar Stereographic The Uni-

versal Transvetsc trlercator (UTfl) [along q'ith the Universal Polar Stereographit
(UPS)I coordinate system is a rvorldwide grid system of scparatc zones lt rvas dr:
veloped in 19,17by the U. S. Army as a worldo'ide military grid s1-stcrn Originalll
designed to be useful for artillery operations, it has since been used cornmercialh'
as well. Thc UTIU system consists of sixty 6o wide zones of longitudc, with zonc 1
going from -180" to -174o longitude, zone 2 from -174" to -168o loDgitude,r:tc
Lettcr desiguatiolts are sometimes trpplic<l to zones rr.rnning fLottt sorrth to nolth'
The various zoDesare as shou'n in Figulc 4.233.1.The central ncridiarn is the ccu
tral meridian of cach zone, t'ith a sctrlc factor of &s = 0.9996. This results iIr a
maximum scalc crror of 1/1000 anywltcrc iD a zone, ot 1f25OOfor points on the
central meridiart or above 45" or belon' 45" latitude. (Of cortrsc.as n'ith all glict
systems, there is no scale error if thc zrppropriate corrcctions frrr scalc are ma'I" )
The use of UTNI is ruually limited to bctwccn +84o and 80' latilrrde Grid coor-
dinates arc dcfined with X = 500000 mcters for all points on tltc ccntral meridian'
and Y = 0 rnctcrs fot all points on the oquator when points arc bciltg rcferenceditL
the Northcrn Hemisphere, and Y = 10000000 lneters whcn points are being ref"r
enced in the Southern Hernisphere.Although the latitude and longitude of a point
referenced to any cllipsoid may be coltverted to UT\l coordinates' for certain arcas
of the world, speciflc ellipsoids arc associated u'ith UT\I coordinates. For examplt"
the specified ellipsoids are Clarke 1866 for North America, Clarkc 1880 for Africa'
and the Intcrnational for most of the rest of the world, except for the far east and
Australia. [See Dept. of the Army (1973, plate 1) for these ellipsoid designations'
Fisure 4.233.1
Section 4.24 explains the subject of ellipsoids in more detail ]
UiM andUPSgrid zonedesignations
 212 EXPLANATORY
SUPPLEMENT OFTHE EARTH
ANDTHEROTATION
TEFRESTRIALCOORDINATES 213
The Universa.l Polar Stereographic Grid System (UpS) consists of two pola.r scale factor at any given point is given by
Etereographic projections, one for the north polar area and one for the south polar
axea. The north zone covers all areas qorth of 84" latitude, a.rrd the south zone all
areas south of -80' latitude. The scale factor at the origin (the poles) is &6= 9.994. t =tofr +<r+Af; +<s- 4r + 42c+ BC - 28e'\#
In both zones, a 2000000 meter X-coordinate line coincides with the 0. ard 1g0.
meridians. The pole will have coordinates of X = 2000000 meters, a,nd y = 2000000 16r\#l
+(61-148?+ (4.233-7)
meters. The international ellipsoid (a = 6y83SS meters, f = 1/297) is associated
with UPS coordinates. Because of the a"reasit covers, the UpS is little used.
[For The inverse formulas, for converting (X, I to (@'I) a're
more information see Dept. of the Army (1958).]
Although UTM and UPS were originally developed for use with tables and - - - gen)D4/ 24
graphs, with modern computers it is easier to use the actual transformation equa- 6 = h - @0.,tal 6t / R)lD2 I 2 (5 + 3 4 + LOQ 44
tions. The transformation procedures a.regiven below [from Snyder (19g3, pp. 6Z_ +(61+904 +2g8q+4s4 - 2s2en- 3q)ff t72Ol, (4.233-a)
69, 160 164)]; both UTM and UPS a.realso discussed in Parker and Bartholomew ,\ = )o + [D - (1+ 2Tr + C)D3/ 6
(1e8e).
For UTM, sta"rting with our previously given expressions,we continue by com_
+ ( 5- z c t + 2 8 7 1- 3 Q + 8 e ' 2+ 2 4 4 ) f f / l 2 o l l c o s h , (4.233-9)
puting 56, the length of the meridional arc, the true meridional distance on the
ellipsoid from the equator to latitude /: where/lislinefootpointl&titud'eotthelatitudeatthecentralmeridianwhichhas
the 6ame Y coordinate as that of the point (d,I) It is given by

so = a[(l - e2/ 4 - 3e4/ 64 - 5e6| 256*.)d -

S , = 1 r +( 3 e 1 1 2 2 7e 3 r1 3 2 +" ' ) s i n z p +
( 2 1 e ?I 1 6 - 5 5 e Il 3 2 - " ' ) s i n 4 p
- ( 3 e zt 8 + 3 e 4/ 3 2 + 4 5 e 6t t o z 4 + . . . ) s i n 2 4 (4.233-rO)
+ ( 1 5t e i / 9 6+ . . . )s i n6 p + . . .,
+ (15e4I 256+ 45e6| fi24 + ...)sin44
- (35e6t3072+ ...\sin6d + ...1, (4.233-t
)
r-\n=v (4.233-11)
where / is in radians.Also neededare other auxilia,ryquantities: l+vr-e'
p = s l l a ( l - e 2 l 4 - 3 " a / 6 4 - 5 e 6 1 2 5 6 -" ) 1 , (4.233-12)
t = tan' Q, (4.233-2) S=Soo+//ko. (4.233-13)
L=e-cos-@, (4.233-3)
.For the other terms, in addition to needing C, T, and N6 for the footpoint latitude
4 = cos d(A)). t L ) 17-4\
f1 we also have

We also compute S4o as given by S computed at the latitude

/s at the origin of the
X and I coordinates. Then the direct transformation from (d, )) to (X, I/) is given
h = a(1- ez)I (l - ezsir?4)3t2, (4.233-14)

by o=X/(No,kd. (4.233-rs)

For UPS coordinates, the direct equations are

x = & o N o l A + ( 1- T + q A 3 / 6
+(5 - t8T +72 +72C - 58et2)A5 / 1201, (4.233_5) X=2000000+psin), (4.233-16)
y = [o{Sd - S4o+ N|tzn glA2t 2 + (5 - T + 9C + 4C)A4 (4233-17)
t 24 Y.on = 2000000- Pcos)'
+(61 - 587+72 +6cf,,C- noet2)A6t72OlI. (.233-6) I"outr,= 2000000+ Pcos). (4.233-18)
 OF THE EARTH
ANOTHEROTATION
215
COORDINATES
/ TERRESTRIAL
214 SUPPLEMENT
EXPLANATORY
in surveying urbanized a'reasis
off in the future since: (a) a 1 in 10000 error pur-
The first Y is for the north zone, the second for the south zone. The auxilia.ry i"".-i"t and (b) the NAD1927has beenreplaced.formost
ouantities are obtained from ""**ptable, 1983, and the state plane systems will now
bv the iorth American Datum of 'ure since it involves changing laws in several
zaLot to be uPdated-a drthcult Proceq
^_ curently use
P- - e1t-"
(4.233-19) . H"*eL, many governmental and commercial organizations
lF * plro" Lavilv' [For more information
"rr''-11 "',"]";rttr
t7r d\ [t -esindl'2 "oo.dintt""Mitchell uoi Si*^oo" (1945)a,rrdTSnvder (1983':tlit:flt""r'
pp' 56-63)'
t=tanl--:rr-l (4.233-20) J;;;;" see
containinforma-
\4 2 ) l l + e s r n $l IJ'"*u"rins **ts, suchas MofEt and Bouchard(1975)'also
tion on state Planecoordinates'l
We also have for the scale ]'"*fi" for the transverse Mercator projection are *:tU{ u*"1
"lL":
(4.233--21) "oo"ii.". UTM above' The only changeswill be for constants such as the
am trc Jo"ripiion for
(&o), and the edd
LI,j _""4-, ()o), the scate factor ou the central meridian
where cpordinates of the origin
*"-ii"
cos/ axe taken ftom Sny-
(4.233-22) fo. th" L"'rnb"rt conformal conic projection
1/t-dsinz6 "qo",i"tt" equationsare
der {1983, p. 107-109).The direct
The inverse solution for { and ) is given by (4.234-1)
X = psine,
( -esindl"'?'l (4.234-2)
ll Y=po-pcose,
Q =-" - a r c t a n - r r , n d l (4.233-23)
, Pon (4.234-3)
lt L, J' am
,\ = arctan
+. (4.233-24)

4.234 StrtePlane Coordinates In the United States, a state plane coordinate sys- (4.2344)
p = aF{,
tem has been established so that each state has its own grid coordinate system.
(4.234-s)
These systems were originally designed by the U.S. Geological Survey in the 1930s d=n()-,\o),
and 1940s, and have since been legally adopted by the state legislatures of most (4.23M)
po = aFl6'
states. For ea,chstate, they consist of one or more zones of transverse Mercator for
bt my - lnm2 (4.234-7)
states or areas of states with a mostly north-south extent or Lambert conformal n=
htl, -ln;'
conic projections for states or areas with mostly east-west extent (and one oblique
cos @ (4.234-4)
transverse Mercator projection for pa"rt of Alaska). There are 136 sepaxate zones n=
covering all of the 50 states (the U.S. Virgin Islands, American Samoa, etc.' all
$:7srFW'
/tr d\ I (4.234-9)
also have their owlr projections, some of which a.re noi conformal). Snyder (f983, '=h(a-r) h;;in'
pp. 58-62, Table 8) lists all the zones, by projection and state, with the central '
t dt"h-Atl
meridian, its scale error for the La,rnbert conformal conic zones, and the standa.rd ,nl (4.234-10)
pa,rallels and origin for the transverse Mercator zones, The coordinates of the ori- nti
gins are also given in a footnote to that table. Note that surtey leet (where I m :
th: eq:ation
39.37 in.) are used rather than standard U.S. feet (where 2.54cm:1 in.). These with the sa.mesubscripts 1, 2, or none applied to z and { in Y lr'l:n:
systems were established under the assumption that any calculations done assum-
0, 1, 2, or ,ton"aooti"aio t tlna4 i".'l:.*:?:tt"
the subscripts l:: i^l^1ffrt"tfl
ing plane coordinates would not have errors worse than 1in 10000. Their defitrition
ffi ;illffi ;';;;;;;;;' tu"ipouti't'"l
Not" Jt :::"::.:3ii"{ with
itr many states is also legally tied to the North American Datum of 1927 (see the
;;il.;";;l"" t"ln.iulJain"'""""s, asthetableswerecomputed
next section), and la^nd pa,rcels may be defined or even rnlst be defined in terms of 1983, p' f08)'
,'lee+precise formulas (Snyder,
state plane coordiaates. It seems likely that the use of state plane coordinates may
,1
 216
The inverse formulas are
EXPLANATORYSUPPLEMENT
r d / TERRESTRIALCOORDINATESAND THE ROTATIONOF THE EARTH

4,241 Datums, Geodetic Ref€rence Systems' and Terrestrial Reference Systems A

geodetic datutn is a coordinate system defined relative to a refetence ellipsoid (as
217

a=|_z**^{,i;##1"}, (4.234-rr) defined in Section .1.21)and an initial point (or origin) near the surface of the ellip-
soid. Therefore the following arc needed: (a) the size and shape of the (rotationzr,l)
ellipsoid (e.g., a. I / /); (b) the a^stronomicalcoordinatesof the initial point of thc
where
ellipsoid (Q,A), and an azirnuth (o) determincd fr<xr that point; (c) dcflections
of the vertical and a geoid undulation at the initial point Obviouslv an infinite
(4.234-12)
lumber of choices are possible for the above palarneters. but scveral trlethods are
p=1fyz-1ru-ry, (4.234-13) usually applied in order to provide the most uscful possible referencc sl'stern. As
e previously dcscribed. the ellipsoid is chosen so as to represent the gcoirl as rvell as
(4.234-t4) possible over the desired arca (e.g., a country) or lltc entire Earth. Fol a local da-
Y tum, the deflections and unclulation are sometimes chosen for sirnpli<itv to bc zero
d = arctan --- ;. (4.234-15)
Por at the initial point but tnore usefully ovcr larger areas or for a tcrrcstrial <latum,
these values ale choserr so that thc rotational axis of the ellipsoicl u'ill bc parallel
The sign of p is taken frour l. If r is negati\.e, the signs of X, I/. and p0 in to or identical rvith that of some refetencerotational axis of the Ealtlr (arrd if the
the
equation for d must be rel.ersed, and the quaclrant of d determincd accordingll.. deflectioncomponentsare absolute,then thc daturn is also geocentrit).
Also note that thc equation for / requires iteratiorr on {, and will corrvt:rgerarpidll, As an historical notc, it should be addecl that classical referencc cllipsoids and
with the initial assumDtion geodetic dat,ums were usuir,llYcstablished in support of thc surveyiltg of grcat hor-
o = ! - 2 * " t u nt . 14.234-16) izontal geodetic networks. usitrg triangulation networks over largc arcas as rvell as
auxiliary baseline. astronoutical. and gravity llleasllrelnents. Thc rl:rt;r from these
4.235 Other Grid Systems Obr,ioush.. the equations that ha1.ebcerr qir.err
abor.e surveyscoulclbe acijustcdntathematically in ordcr to specifv thc bcst-fitting ellip-
for the Transverse llercator, Polar Stereographic. ancl Lambert r,-onfornrai
Conic soid to the area being surr-cterl. as rvell as thc ol ltcr tlesiredparantclcr-sfol a datum'
projections can be used with nrrury other grid systerns with just
a chir,ngcol the An initial point for the (latllln lvas usually cllosclr neaL the ceulcr of thc net'"r'ork
appropriate constants. in order to nrinimize thc ettors in rncasuremctlts rolative to the ilititrl point as one
moved outward in the uetwolk.
4.24 GeodeticDatumsand ReferenceSystems Howevcr.the spaccage iut<lits int roductiolt of artificial satellitesgrr';rtlt t lranged
the methods of daturn detcrtnination. Non' the absolute positions rclativc to the
A geodetic daturn or referenco sl.stem provides a fraurelr.ork to u,hit.h center of the Earth coulcl bc detcrrninedfot variottspoints in the gcotlctic netu'ork
1hc coorcli,
nates ofarry point rnay be refercnr:r:d.A geodetic tlatrrur has (the techniquesused for this are <lescribedin Section'1.J) These p"sitiorr Jetermi-
a r.ery spcc.ifi<r
rlcfinition
in terms of classical geodesy; ltower.er. some rccent ,,(latums', nations cortld be combineclrvitlt the conventiotttrlsrtrvel-ingobsr:tvrrtiolrsitr order
and mosl tcrrestrial
reference systems do not fall strictlv under this ck:finition. to obtain Imtch more precisc datums lor muclt ltrtger areas Frtttirtl. co"rdilat'
A geodetic datum may also be either a horizon.tal or rerlicol datuur. systems(usuallv with sets of slation coordinatcs)c:rtrbe cleterrnint'rlwithout tef"r-
clifler.cuti-
ating rvhethcr the datum is to scrvc primarily for clotcrrnining ence to any ground-basedsrttvcvingobservatioltsor e\€n to an ellipsoi<l(although
(horizorrl,al)location
on or near the surface of the Earth. or for detorrnining an ellipsoid is often uscclto (xpress station <:<xrrrlitratesin the Ilr,,rt' colrvelltional
hcight (actuallv gr.alit1.
potential) of :r gir.enpoint. In thc discussion geodetic latitude and longitrtcleinstcad of gr:occutriccartesian coor<linates)'Thc
that follolr,s.it is a horizorrttrlclaturn
that is bcing referred to. orientation and scale of thcsc spacc based systctttsis usualll' obtaincd fiom some
In this sectiou, we brielly rlcscribehow thcsc dlipsoicls, combination of holding the coordinates of ont'or nrore stations fixc<I, trrlcl using or
daturns, gcodetic ref_
erence systems, and terrestrial rcferonce systems iur determining an Earth rotatiou scriesthat fixes thc cltangesin the svstttn rrvertimc
created, and therr list nrany
of the orresirr use. Irrformation orr how transforrnations due to the Earth's rotation. Itt practice thesc assumptions arc urzrck'so that the
may be pcrfonnt:rl bctween
these various datums and rcfercrrce systems is also provided. new system will have the saure oricntation artcl sctrlc as some previorts "standard"
Finallv, a short cxpla-
nation is given of horv to red.cc cor'mon geodetic a.d astronomicar system, such as the BIH or IERS systems (clcscribed belorv) l'hc ctnter of the
observatious
to the refercnce ellipsoid. Earth is determined dynarnically by the satellitc obserrations.
 218 EXPLANATORYSUPPLEMENT

One of the first consequencesof the availability of the new techniques *as
the establishment of geodetic referc.ncesyste.msthat have much in conrmon rvitli
a geodetic datum but have slight differenccs in their definition. Such a rcferent' 9
6
system does not rely on the definition of an initial point. The orientation of sucir
a system is assumedto bc "perfect," i.e., with rotational axis, referenccmeridian. =
-
and center, being that of the Earth. The rcfcrence ellipsoid in this case is ahvavs
a terrestrial reference ellipsoid, and the best possible lepresentation of t,he rnean
Earth elLipsoid,or that ellipsoid that best approximates the geoid. To accomplish
the latter, the gravity potential at the surface of the cllipsoid is specified (or al- o
o
ternatively, "/2, the dynamical forn factor of the Earth) as well as the rotational
o
velocity of the Earth. The constants a,rrdparameters of such a system arc of coursc
always undergoing refinement. but the specification of such sl'stcms allows for their' :
use in referencing geodetic observations.
A fintrl refinement beyond a geodetic refcrence systeln is a terrestriaL referert<t'
sgslem,using a rcfetencefrarne consistingof a set of cartesianstation coordinates
and their velocities to provide the uuderlying definition rathel than a referencc
'
ellipsoid. Many other pararneters ancl even methods of dnttr reduction ("Standards )
are also given as part of t]rc referencc s1'stern generally in order to dcfine ho*'
observatiorrs will bc used in order to dcfine the dynamic connection betiveeu such a
terrestrial system and a celestialone. By interrrationalagreemclrt,one such svstcnr ill
is defined as fundamental and it is thc IERS Terrestrinl St'stenr.describcd furthcr
in Sectiorl1.2.5.

.? "s
4,242 Specific Ellipsoids, Datums, and Referenc€ Systems Tablc -1.2-12.1 lists \ar- @
ious ellipsoidsthat have brtn used as the basis of geo<lcticdatrrms and rcfelelcc
systems. Note that as previ(msly meutioned, only a an<l I /.f aro necessarv for tht'
definition of an ellipsoid. Table 4.212.2lists geocleticdaturns (also giving 1hc origirr
station if any and its latiturlc aud longitude) aDd Dotesthat are associatedrvith spc-
cific terrestrial reference systcrns. Sirrrilar, but abbreviaLed versions of thosc tabl<'s =@

zrrcgiven in the crrrrent Astronorniarl Almantu:. pagc K13. Firrallr'.Figrrrt'.1.2.12.1

(after Rapp. 1980,p. 82) shorvson a rvorld map the co\,erageof tlrr: rnajor geodetit
<iatums. 3

-L Im
4.243 Datum Conversions The subjcct ofcorrvertirrgcoordinir,lcsfi orr oilc datunl LJ

to arrother is a cornplexonr:, lvith a vcr.yexl(]lsive literaltrrc. Inrleecl,the ptoblenr

alrnost alu,avs arisos arnd is rrsually cliscusse<l.!vhenev('r?r nelv dlturn or rcferencc
sYstemis formed. sirrcethcrc t'ill usrrallybe a rreedto transfornr the coor<liuateso1
i !.:
points in other syskrns to <oordinatesin the rrcw systcrrr.
3NT-J
Two general rrrcthodshave been used in 1ho past firr datum transfornlations.
The first is a difiere.ntialmcthod. s hcre differential (in tLe partial-derivativcscnsc)
equations are forrncrl that Bive thc ch:rrrgeilt statiort t:oordinates(Ad,A)) as a
Fi9we 4'242,1
furrction of changcs in datrrrn/ellipsoid pararnetcrs (such as for c:hangesin a. /, Majot geodeticdatumlocations
 220 EXPLANATORYSUPPLE[,4ENT Table4.242.2
Table 4.242.1 GeodeticDatums
EarthEllipsoids Latltude Longitude
Oalum Ellipsoid Orlgin

Ellipsold Semi-maior Inverse

C arke 1880 + 2 2 1 0 4 7 .| 1 0 0 3 1 2 9 2 1 . 6 0 8
(nameandyearcompuled) axis (a) in flattening - 1 42 00 8 . 3 4 1 1 8 9 1 7 0 77 5 0
C ark€ 1866 Betry 13 ECC
Arne @n Samoa 1962
meters ( 1/ f l
Arc-CaPe(SoulhAiica) C a|ke 1880 33 59 32 000 0253A44 622
3 55 8 1 7 297 494A
-07 57 ua37
Airy1830 6377563.396 299.324964 1 3 3 1 2 3 00 8
JohnslonMemonalCal.n 25 56 54 55
Everesl1830 6377276.345 300.8017 Ausldlian Geodelic 1966
Clarke 1866 Fl GeorqeB 1937 +322244 360 2 9 5 1 9 0 18 9 0
Everest1830(Bonialternate) 6377301 .243 300.8017 Eemuda 1957
+465708660 0072622 335
B€me 1898
Bessel1841 6377397.155 299.152813 + 0 r2 1 4 2 0 3 1 7 2 5 5 4 79 0
Clarke1866 6378206.4 294.978698 Camp Area Asl.o 1961-1962USGS
77 5052.521 1 6 6 4 0 1 37 5 3
1966 CanlonSecorAslo 024624.99 r 8 8 1 64 3 4 7
Clarke1880 6378249.145 293.465 Clarke 1866 +242932 364 2792521 230
Clarke1880(modified) 6374249.145 293.4663 sal T .sra.059 RM3 +0200359l 2023521 A2
ch.ishas lslandAslro 1967
Helmert1906 6378200 298.3 Chua Aslro (B€zil_Geodelic)
1 9 4 5 4 11 6 3t r 5352 44
1 95 0 1 5 1 4 0 31 | 02 t7 25A
lnternational
1909 6378388 297 Cori€goAlegE (Brazil_[,laPPins)
-271039 95 250341681
Krassovsky 6374245 298.3 Ea$er lsland 1967Aslo
1 6 8 2 0 3 32 5 0
Mercury1960 6378166 298.3 +522251 45 0 1 3 0 35 8 7 4
+390354934 3 3 1 5 7 3 61 1 8
wGs60- 6378165 298.3 Gacosa lsland(Azores)
GUX1 092705 272 1 5 9 5 8 3 17 5 2
IAU1964 6378160 294.25 Clark€ 1866 +13223849 1 4 4 4 55 1 5 6
Australian
National'1965 6378160 298.25 5 3 0 11 1 6 8 a 1 3 2 3 2 26 4
WGS66' 6378145 294.25 tbsn Asrrc, Navy 1947 (truk) + 0 7 2 9r 3 0 5 1 5 1 4 9 4 44 2
+ 2 4 0 7 1 12 6 0773917 57
ModifiedMercury1968 6378150 298.3 C alke 1866 Slalion 03a +18434493 2 4 9 0 2 3 92 8
SouthAmerican1969 6378160 298.25 +t64449 729 19029 04 781
GeodeticReferenceSystem1967' 6378160 254.2471674273 Kourcu (FrenchGuiana) 05r553 699 524S09149
WGS72' 6378135 294.26 Kusai€,Asl.o 1962.1965 +05214880 1 6 2 5 8 0 32 3
+r33341 000 1215203.000
tAG1975 6378140 298.256 Luzon 1911 (Philippines) Clafte la66
1A23624 2A
+ 2 4 1 1 3 45 0
tAu 1976 6378140 298.257 4 1 1 9 0 49 0 0 r 7 50 25 r . 0 0 0
Geodetac
Reference
System1980' 6374137 294.257222101 Cla.ke 1866 + 3 9 1 3 2 66 8 6 2 6 1 2 72 9 4 9 4
MERIT1983- 6378137 298.257 GFS 80
+ 4 8 0 8 2 0 0 0 0 0 1 1 3 4 2 64 8 3
wGs84" 6378137 298.257223563 + 2 1 1 a 1 38 9 2A2a904 21
C arke I866
tERS1989' 6378136 294.257 OrdnancoSurueyG.B 1936 A ry 1830 + 5 0 5 1 5 5 . 2 7 1 0002045 882
(sN)
osca 1970 +505155.271 0002045.882
1 Ellipsoid tsTs050 -644635.71 2955639.53
is definedor recommended
as partof a geodelicor l€rrestrial
relerence Paher Aslro 1969 (Antarctica)
syslem, Pico de las Nieves(cana.ies) + 2 75 7 4 t . 2 7 3
,250406.97 229s312 t7
+522253 954 0 1 3 0 4 0 11 5 3
Sourc€siRapp(1980,p. 83r,Astrcnonicat
Atmanac
(1990,p. K130),D[]tA{1987, +a83417 1T 2960A25 12
parl1,p.3-41),Mcoanhy ProvisionalS Anercan 1966
i1989). PbvisionalS. Chile 1963 5357 a7 76 251232A 76
+59461855 0 3 0 1 9 4 20 9

,194541 653 3 1 1 5 3 5 59 3 6
Sourhea$ tsland(Mahe) 044039 460 0553200 166
lsTS 061 Aslro Poinr1968 5 4 1 63 8 9 3 3 2 3 3 0 4 39 7
Swallowlsrands{sotomons) 1 0 1 8 2 14 2 1661756.79
scale, etc.). This tlpe of rlethod has been most rrscfitl in the past rvhere stnail TananariveObseryatory 1 8 5 5 0 21 0 047 33 06 75
differenccsexist bet&'eeuthe two datums. or a minor <.harrgr: TokyoObseflalory (old) +35 39 17 51 1 3 9 4 4 4 05 0
irr an existing datuur ,37 0326 79 3474053.21
Inrsalrig06s FIMNo. 2
was being madc. Horvever, the derivation of thc differentitrl equations for. a specific vili Lovu 1916 (Fij) M6naval! (la1ii!d€onlY) ,17 5324 245
C afk€ 1840
case and their appli<:ation arnd accuracy is a cornplicatcd subject and will not be 1 7 8 2 5 3 58 3 5
Wake lsland,Astrcnomc 1952 + 1 9 1 7 1 9 . 9 9 1 16638 46 294
discussed firrthcr here. +323027.Ole 2 5 3 3 1 0 1 . 3 0 6
flnformation on this type of ructlrod is available in Rapp Clark€ 1466
3423052.98
Yol Asro 1967 (Oaka4
(1980, pp.42 74); Bonforcl (1971, pp. t99 2O7);HoisL.anon and Moritz (1967, pp.
Clark€ 1gso

204-21o)i Molodenskii et al. (1962, pp. 13 t7); Tc,rge(1980, pp. 150 152); DN{A -
Localdatumsol soecialouroose,based on NAD 1927 values lor th€ originslalions
Sources:napp ( t sjgo),lviuolierei ar (1973.p. 1 t), AstrcnomicalAtmanact 1990 p K13)
(1987, part 1, pp, 7-1 to 8-8).]
 tt2 RoTATIONOF THE EABTH 223
EXPLANATOBY
SUPPLEMENT r' / TEFRESTBIALCOoRDINATESANOTHE
purpose arc those of the Navy Navigation Satellite Svstenl
A second mcthod is to do a rigorous transformation bctwecn the svstens of usually observed for this
on these sYsiems
each datum. This is often required if high accuracy is desired, or there are several (NNSS) or Global Positioning System (GPS) More information
'1
or large diflerences between the datums. For any pair of datums, trarrsformatiorr is given in Section 4'
parameters may have already been deterrnined, at least for oflsets and sometimes
to the Ellipsoid In manl cascswhele astlonomi-
for rotation and scale. A three step procedure is required to convert (/, ), h) lrorl 4.244 Reduction of observations
made' it is desirable to "reduce" the bbservations so
system I to system 2 cal o. g"od"ti" .rbservations are
observations made on the ellipsoid itself using
ihr,,h"y would correspo[d to such
vertical This is usually for when one is referring
( 4 ,^ , h ) t = ( X , Y , 4 t , (4.243-t) the ellipsoidal normal as the local
or for that nattcr creatilg a ncw
observaiions to or densifying a current network,
(X,Y,4t ) (X,Y,Z)2, (4.243-2)
network itself.
depending
( x , Y , 4 2 1 ( d ,^ , h ) 2 . (4.243-3) Various formulas of various accuraciesexist for this type of reduction'
(chord or arc distances' azinruths angles etc )
on the type of observation involved
will contairr variolrs velsions of
The first tn'o transformations are already described in Section 4.22. The equatiorr and the accuracy desired. Any good surveying text
91' 120 122; Heiskanenand
for the third transformation is these formulas. (Also see Bomford. 1971' pp 45, 89
Moritz, 1967,PP 184 197')
txl f xl Irr I I D - R . ' . R . zI l x l
l'l=l'l lr, l l^, D -R,ll'1. (4.2434) System
Reference
Terrestrial
4.25 fhe Conventional
LZ)t LZ)t LrrJ L_R: R1 D )LZJ, as part of tlte
The conventional Terrestrial Reference Systent is currently defirted
where 71, 22, and 73 are the componentsof ertranslation vector; R1, R2. and Rt irre IERS Reference System The IERS Reference S-v-stem itself cottsists of the IERS
the rotation angles (usiug small-angle forrnulas here): and D is thc cxcess 1 scalc Reference Frames and the IERS Standards'
rl'ith
factor (BIH, 1988. p. A-4). Various forms of this seven-parallctcr transformation The IERS Reference Frames include both a terrestrial and a celestial one'
exist under vatious names (some not as rigorous as others). (Ser: Kumar. 1g72: the terrestrial frame (ITRF) being defined u'ith the following characteristics: The
b1'
Rapp, 1980,p. 48; DMA, 1987,part 1, pp. 7-6, 7-I2 to 7-74.) It is also possiblethat origin, the reference directions, and the scale of the ITRF are implicitly defined
some of the parameters may not be known or can bc assumed to be zcro. In modern the cartesian coordinates and velocities aclopted for various "primary" observing
relerence systens, D is usually close to zcro, as is ft and ?2, and the rotations ar-(l stations of the IERS. If geodetic coordinates are needed' the GRS 80 ellipsoid is
-- 298 257222101)
quite small. recommendecl for use in the convcrsiou (a = 6318137m, I l f
In some cases, the transformation paranletcrs are not rvell-derterrninedor not The origin of the ITRF is located at the center of mass of tlxr Earth with an
knorvn at all. The user must then deternine thc parametersdirectll,frorn obserr.a- uncertainty of less than 1 l0cm (as deterrnined bl la.ser-ranging'rbservations)'
tions that tie stations together betrveenthc tlvo systems.If cnough points have their. The standard unit of length is the SI rnetcr- The IERS Referclr<e Pole (IRP) and
coordinatesdeterrninedin both sl.sterns,E<lutrtion,1.2434 or sirnilar equationscarl Reference Meridian are consistent with the corresponding directions in the BIH
be solved (usuallv in a least-squares solution) for the transforrnation parameters. Terrestrial System (BTS) within i 0.005". The BIH referencc pole was adjusted
Of course, this may not take into account arry deformations of eithcr. of the datums to the Conventional International Origin (CIO) in 1967. It was thelr kt'pt stablc
from their idealizcd reference systems (due to surveying errors). Any deformations independently until 1987. Recent re-reduction of the available data has shown (BIH'
will be absorbed into the transformatiorr paraueters when they are solved for. 1988, p. A-3) that the tie of the BTS (and therefore the IRP) with the CIO is
Of course, nowadays an alternative exists to doiug transforrnatiorrs at all. One accurate to t 0.03". The ITRF shoulcl show no global net rotation or translation
could just measurethe coordinatesof all the clesiredpoints with a satcllite receiver n'ith time (due to the motions of the stations or the tectonic plates they lie on) A
thus deterrnining the coordinates of the points in some current relercncc system. more complete description of the IERS Terrestrial Frame is given by Boucher and
bypassing any transformations entirely. This climinates the need to use the old Altamimi (1989).
datums or old obscrvations. The coordinates can be measured in an absolute sense. The Celestial Frame (ICRF) is defined similarly, using the coordinates of 23
relative to the center of the Earth at the 1- to b-meter accuracy level, or relative to "primaryu radio sources (a.it" ,r1,,1988). The origin is at the barycenter of the
"t the IAU
other points nearby (within tens of kilometers say) to the few-cm levcl. The satellites .o1., The direction of th" poit, u*i" is given for epoch J2000 0 by
"yrt"-.
 ANDTHE RoTATIoNOF THE EARTH
COOBDINATES
225
EXPLANATORYSUPPLEMENT r' / TERRESTRIAL
224
field can be considered a potential field
The Earth's external gravitatiortal
1976 Precessionand the IAU 1980 Theory of Nutation (see Chapter 3). Comparison fulfills the Laplace equatiort
of optical and radio observations shows that the origin of right ascension is il whose strength V
agreement with that ofthe FK5 star catalog to within t 0.04". The parameters AV=0, (4.31-l)
that describe the rotation of the ITRF relative to the ICRF (in conjunction with
the precession and nutation model) are the Earth rotation parameters (ERP). For oPerator
where A is the LaPlacian
further information on celestial refetence ftames, see Chapter 3.
'Ihe IERS Standards (l\'lcCarthy, 1990) describes tlte models to be used u'hen Az
a2 a2 rA 1l-)r
-*
data are being reduced for use by the IERS (and asks that the effects of deviations ax, ay" a?
from the standards be determined). The next section contains material substantialll
systen]'
drawn from those standards. in a cartesian X, Y' and Z coordinate
Equation 4 31 1 (Heiskanen atrcl \{oritz'
The IERS ReferenceSystem evolved from a series of previous systens. As Iroted It is possible to devclop a solution to
pp 26 29) in terms of a seriesof LeqcndrepolgnorniaLs:
above. the ITRF is itself a direct continuation of the BIH Terrestrial System (BTS). fSOZ,pp.'fS 35; Torge' 1980,
first established in 1985 and continued through 1987. The BTS had in turn evolvecl
from the BIH 1968 System, which was established in 1969 by the BIH on the basis ^i\'r-tt
e"ut=71,11 (4.31-3)
of a comparisort of data collected by the BIH and the International Polar Motion
Service(IPMS) during the rnid-1960s.However,it should be emphasizedthat this with
- ))' (4.3t4)
was nol a complete terrestrial refelence system; it q'as a referettce system for di- / = c o s p = c o s 0 c o s d '+ s i n d s i n d ' c o s ( I '
rection only, beilg based on the observed astronornical coordinates of the invoh'ed
observing stations. The courparison of the BIH collected data rvith ttre IPNIS data an attractcd point (at spherical
ry'is the spherical distance on a unit sphere betu'eeD
allowed the BIH 1968 System to be refercnced to the IPMS refcrence pole. This (at / ' 0" ^')'
coordinatesr, 9, ,\) and an attracting point
coorclinatesd an<l '\' it
pole is thc Conventional International Origin (CIO) thich was the approxitrtatt: In order to expressP,,(t)as a furtction of the spherical
of the lirst Airrrj of dcgre<:
average position of the true celestial pole of the Earth from 1900 0 to 1905 0, but may be decornposodusing ossocioletl Legen'dre Furtctions
was actuallv defined by the adopted latitudes of 5 International Latitude Selvice l 5 , l i f f e r e ni t; r t i n gP , , t l t m t i t t t " s
n a n d o r d e rm , o l ) t a i n F ,b
(ILS) stations. (For further information on all of the above,seeIERS, 1989,or the
BIH Annual Reports for the appropriate year(s), or see \'Iueller, 1969 for firrther 1= 11 l t"' ft, e""'
P,,.,rt (4.31-5)
information oD these systemsfor the pre-1969period.)

It can then be exprcssed bv the fitnctiolts

4.3 GRAVITY. OF THECRUST

THETIDES.ANDMOTIONS P,,n(cos0) coszr'\'
(4.31-6)
(4.3t-'7)
0) sin ml,
P,,,,,(cos
In this section, various phenornena will be covercd that nflect thc Ea.rth's gravitl
field and coordinates of a given station. A very short sunlmary of orto method of
modeling the Earth's gravity field (as a spherical harmonic series)is given. al"ng known as surface spherical harmonic functions
with the IERS tecommendations for a "standard" gravity field. This is follou'txl by. A general equation for thc gravitational potential then becortrcs
alnrost verbatim. the IERS Stand,ardsrecotnmendations on Earth tides, ocean tides.
ocean and atmospheric loading, and plate motions (MCarthy, 1989). Finally, the ,' ,- _ 6 1 a { - o t n - z t !
effect of tides on the Earth's rotation itself is described, as given by thc Stantlards' , Z-7L'ln+mll
r -( p , , . t . o , o r . o t , . ^ t''P,,"\cor|'rcosm^dm
* [ [ [
4.31 r' \ J J Jr.tnh
Modeling the Earth's Gravity Field

The Earth's gravity can be considercd as a force that is the sum of the Earth's +pn,,(cos
d)sinm^ n); (4.31-8)
III^^^r'r,^r""s0')sinn'\r
gravitational force and the centrifugal force resultirrg from the Earth's rotatiort. idr,
-
 226 EXPLANATORYSUPPLEMENT
, / TEFFESTRIALCOORDINATESAND
THE BOTATIONOF THE EARTH 227
(l lorm=o
*, = where c,r"is the angular velocity of the Earth, and X and
I are the coordinates of
\z form1o-
ihe point in question (P(X'Y'D).
It should be noted that many other representations of the Earth's gravity field
If the spherical component of the potential of the Earth's mass M is separated application Most commonly, l7 may
exist and are widely used, depending on the
out (with n = 0), the ellipsoidal semi-major axis a is used to provide a scale for the Earth itself' using measured gravity-related
be determined over a wide area or the
unit sphere, and the integrals are replaced by harmoni.c cofficients, C,. and Sn., previously mentioned on geodesy and geodetic
data. Any of the standard references
we then have of computations-
astronomy contain detailed information on these types
^.. I
v = "-y lr + ')- (1 g)l| .
nr)+ S,-sinz.\rp,-(cos
- - - -)- \ r l )'' 16,.cos
(4.3t-9)
r | I
L n=t m=u I

In many cases,an alternative notation is of the Earth'sGravityField

4.32 A Representation

(4.31-10)
The current IERS recommended geopotential field is the GEM-T1 model given in
( 4 . 3 1 - l) Table 4.32.1. The GM and a values reported wiih GEM-TI (398600 436 kmr/s2 and
tz-< (4 11-1)\ 6378137m) should be used as scale parameters with the geopotential coefficients.
The number of terms actually needed to provide sufficient accuracy for a given
It may be convenient to :usefully normalized surface spheri,calharmonics: application will, of course, vary. For example, when used in computing the orbit ol
the LAGEOS satellite, only terms through degree and order 20 are required.
Values for the Czr and S2l coefficients are not included in the GEM-T1 model
(43r-r3)
(h

r={;2:EZ
-
(they were constrained to be zero in the solution), and so they shorrld be handled
h\l
P,,,(cos0) = k12n+ 11-:--:: . P,.1coseY
\n+m)! separately.
The C21 and S2l coefficients describe the position of the Earth's figure axis
This results in fully normalized. coelJicients with respect to the ITRF pole. When averaged over many years, the figure axis
should closely coincide with the observed position of the rotation axis (or "mean
pole" ) averaged over the same time period. Any diflerences between the mean figure
Ie" \_ (n+ m)l
I C^ \. r={;2:E?, e3,-,4) and mean rotation axes averaged would be due to long-period fluid motions in the
I S,.J- k(2n + 1)(n - m)l I s,. /' atrnosphere, oceans, or Earth's fluid core (Wahr, 198?, 1990). At present, there is
no independent evidence that such motions are important; so it is recommended
It should be noted that the expansions for V converge rigorously only outside a
that the mean values used for Gr and Szr give a mean figure axis that corresponds
sphere of radius a that just enclosesthe Earth, so that the Laplace equation may be:
to the mean pole position.
satisfied, In practice, however, since the coefficients are normally dctermined ern-
The BIH Circular D mean pole positions from 1982 through 1988 are consistent
pirically from measurement (e.g., from satellite observations), the series can alu'avs
to within t 0.0005" corresponding to an uncertainty of t 0. 0l x l0 e i" c21 (IERS)
be considered to convelge. Once a set of coefficients are determined (see the nexl
and sI (IERS).
subsection) the expression for V becomes ralid and useful for anywher.ein spacc.
If the mean pole during this period coincided with the ITRF pole, then C21=
The same expression may be used for determining the gravitational potential
Szr = 0 could indeed be used. However. the poles are ofiset by the angular displace-
on the Earth's surface. However, as noted prcviously, it is the grotrrilypotential, here
ments t and t. so that
denoted l4l (inctuding centrifugal force) that is of primary interest on or near (i.e.,
moving with) the Earth's surface. We have

Czt = iCzo, (4.32-r)

t-^
W=v+tultX'+Y'\, (4.3r-15) Szt = YCzo
(4.32-2)
 ,r-
Table4'32.1,continued
Table4.32.1 GEM-T1 Normalized (x 106)
Coefticients
Normalized
GEI\,4-Tl (x 106)
Coefficients
ln0ex

Zonals
lndex lndex 243 0.0069i48- 0.0105925 25 3 0.00326420.0031087 26 3 0 0002621 0.0037215
273 0.00510350.0018585 2A 3 0.00026460.0011662 29 3 0.00438640 0017609
303 -0.0016075 0.0014770 31 3 00018668 0.0040769 32 3 0.0006663 0.0029992
93 3 -0.4017072 0.4024776 3 4 3 -0.0005145 0.0022065 35 3 0.0006733 0.0006861
'20 ,4841649906 30 o.9572357 4 0 o.5387322 50 0.0687802 6 0 0I48t004
36 3 -0.0008129 0.0013852 4 4 0.1900348 0.3084595 5 4 0.2948236 0.0524087
70 0 0905337 80 0.0459023 9 0 0.0283764 10 0 o 45t2211 1 1 0 0.0512619
120 0 0320806 13 0 o.04223
t9 1 4 0 o.o197327 15 0 0.0018731 1 6 0 -0 00S3772 6 4 0.0927975 0.4733069 7 4 0.2737404 0.1224207 A 4 0 2460639 0.0677453
170 0.0203968 18 0 0.4112912 1 9 0 -0.0046084 204 0.0153150 2 1 0 0 0097754 I 4 0 0i28303 0.0232637 1 0 4 0.0973123 0.0693825 11 4 0.03321080.0700036
220 0.0049440 230 0.0241260 2 4 0 -0.0009556 250 0.0068867 2 6 0 0.001&91 12 4 0.06530200.0030125 1 3 4 0.0088182 0.0003732 14 4 0.00883290.0018783
270 0.0041234 280 0.0058541 2 9 0 -0.0039091 300 o.0042749 31 0 0.0051154 15 4 0.04437600.0126416 1 6 4 0.0365123 0.0438559 17 4 o0125878 0.A312242
320 0 0000819 330 0.0022286 3 4 0 -0.0024803 350 0.0012731 3 6 0 0 0007396 1A 4 0.0434167 0.0060924 19 4 0.0025699 0.0076747 20 4 0.0017s510.0008281
21 4 -0.0002550 0.0069894 22 4 0 0094462 0.0167100 2 3 4 0.0100336 00016696
24 4 0.00605800.0181145 25 4 0.00635030.0015335 26 4 0.0053222 0 0048968
27 4 0.0029164 -0 0006401 2A 4 0.0028786 0.0024855 2 9 4 0 0063641 00007289
30 4 -0.0021102 0 0039250 31 4 0.00495090.0016392 3 2 4 0.0020179 0.0033825
33 4 0.0002362 0.0003938 34 4 0.0028502 0.0018773 35 4 4.0027442 0AA14A64
36 4 00001181 0.0058408 5 5 0.17775630.6660281 6 5 0.2657650 0.5377472
SeclorialsandTesserals 7 5 0.00347500.0196519 8 5 0.024933500853003 9 5 0 4141122 A 060A627
Index Value 10 5 0.0504401 0.0438269 11 5 0 0459086 00552848 1 2 5 0 0306040 00014745
13 5 0.05961760 0574460 1 4 5 0.0227952 0A116078 15 5 00160742 00108864
16 5 0.4077242 40016773 1 7 5 0.0111472 0.0056031 18 5 0.OO1t426 0.0211276
-3 1 2.0297737 0.2495946 4 1 0.5334272 0.4751189 5 1 -0.0589503 0.0955435 19 5 10.00244570.0173789 20 5 0 0104182 0.0003034 2 1 5 0 0177593 0.0158911
6 r 0.0813751 0.0238300 7 1 0.2770971 0.0974177 8 1 0.0288561 0 0547223 22 5 -0 0046335 0.0001257 23 5 00019341 0.0079407 2 4 5 0.0140838 0 0079805
9 1 0.1440477 0.0245251 1 0 1 0 . 0 7 6 9 6 5 50 . 1 3 8 1 1 1 0 11 I 0.0095019 0.027811 25 5 -0.0024585 0.0023514 26 5 0 0043007 0.0106472 2 7 5 0.0015884 00037103
12 1 -0.0492610 0.0496520 13 1 0.0540617 0.0434555 14 r -0.0147462 0.0232244 28 5 0.0029222 0.4402361 29 5 0.0034464 0.0435422 3 0 5 0.0033916 0.0006s83
15 1 0.0082868 0 0142124 16 1 0.0317099 0.0173493 17 1 00309381 0.0268459 31 5 0.0009096 0.0014649 32 5 0.0000480 0.0027949 3 3 5 0 0005132 00032353
18 1 0.0002253,0.0456055 19 1 0.0115942 0.0053764 20 | 0 . 0 1 4 51i 9 0.0212711 34 5 0.00129520 0000401 35 5 -0.0002351 0.0007374 3 6 5 0.0011798 0.0003403
21 1 0.0153942 0.0417459 22 1 0.0083946 0.0147250 23 1 0.0008657 0.0145970 6 6 0.0090593 0.2363344 7 6 0.35785270.1509175 8 6 0.0664178 03128323
24 1 0 . 0 0 8 1 1 7 8- 0 . 0 2 3 1 9 8 7 25 1 0.0037145 0.0043498 26 1 0.0049741 0.0172518 9 6 0.0705263 0.2166285 10 6 0.03473660.0777189 11 6 A.OOa4723 0.0242910
27 1 0.0005230 0.0066113 2A 1 0.0065294 00100251 29 1 0.0034699 0.0024176 12 6 0.00138810.0458322 13 6 0.02238690.0118360 1 4 6 0 0031868 0.0065119
30 1 - 0 . 0 0 16 i 7 1 - 0 . 0 0 9 0 8 8 6 31 1 0.0051782 0.0023829 32 1 0.009i529 0.0092451 15 6 A0272318...0.051707716 6 0.01799490.0267835 1 7 6 0.0002920 0.0204166
33 1 0 0012672 0.0021258 34 1 0.0015608 0.0091011 3 5 1 0 . 0 0 1 9 0 6 90 0 0 2 0 1 1 0 18 6 0.03119910.0085569 19 6 0.00625450.0039169 2 0 6 0 0127607 00009662
36 1 0.0028774 -0.0058408 2 2 2 4389280 1.3998397 3 2 0.9035491 0 6204198 21 6 0.00421460.0083501 22 6 0.014626100024348 2 3 6 00099988 0.0049568
4 2 0.3470021 0.6640304 5 2 0.6557302 0.3234056 6 2 0.0516096 0.3749956 24 6 0.0003055 0.0006641 25 6 0.0059358 0.0067430 2 6 6 0.0085388 0.0031968
7 2 0.3177108 0.0916083 I 2 0.0703801 0.0684494 I2 0 . 0 3 1 1 3 6 50 . 0 3 2 3 8 8 2 27 6 0.0017746 A.0021342 28 6 0.0083296 0.0020029 29 6 0.0002960 0.0024663
10 2 0.0805212 00513356 11 2 0 0090541 0 . 0 9 9 2 4 1 4 12 2 0 0076400 0.0349183 30 6 -0 0032177 0.0c41679 31 6 0.0005622 0.0007667 3 2 6 0.0038310 0 0002190
13 2 0.0534361 -0.0575844 14 2 0.0348122 0.0060681 15 2 -0.0216258 0.0364425 33 6 a.0013a24 0.0012542 34 6 0.0005808 0.0003220 3 5 6 0.0008304 0.0014752
16 2 0.0156437 0.0245431 17 2 0.0057800 00171247 18 2 0.0084083 0.0168428 36 6 0 0005757 0.0008988 7 7 0.00159760.0220013 a 7 0.070424A 0.074A626
19 2 0.0084369 0.0104744 20 2 0.0198772 0 0032259 21 2 0.0009874 -0.0026067 9 7 -0.1005510 10
0.1186233 7 0.00974680.0042901 1 1 7 0.0096093 0.0918891
22 2 ,0.0142925 0.OO2095A 23 2 -0.0005313 0.0017780 24 2 0.0058515 0.0052022 12 7 0.01269750.0348291 13 7 0.00357360.0066171 14 7 0.0374843 0.0043588
25 2 0.0037220 0.0052068 26 2 -0.4052847 0.0A02522 27 2 0.A102174 0.OA2A223 15 7 0.06671300.0114545 16 7 0.00305110.0090737 1 1 7 0.0229520 0.0119809

28 2 -0.0084276 0.0115524 29 2 0.0094632 0 . 0 0 4 3 3 1 1 30 2 0.0040515 0.0053645 18 7 -0.0007957 0.0067159 19 7 0.00514430.0016620 20 7 0.0077913 0 0048958
31 2 0.0066008 0.0008362 3 2 2 0.0018649 00043749 33 2 0.0010231 0.0009548 21 7 0.01222790.0014058 22 7 0.01275380.0013042 23 7 0.OO232A2 0.0026452
34 2 0.0035817 0.0051783 35 2 -0.0025529 0 . 0 0 1 0 2 7 7 36 2 a.O001790 0.au2922 24 7 o.oo251130.0050662 25 7 0.0002632 0.0034596 26 7 0.0054524 0.0025831
3 3 0.7209866 1.4131694 4 3 0.9909779 0.2006215 5 3 0.4482036 0.2151363 27 7 0.0069279 0.0028217 2a 7 0.0046459 0.0015444 2 9 t 0.0011830 0.0072940
6 3 0.0619709 0.0046430 7 3 0 2507429 02091639 8 3 - 0 . 0 1 9 9 6 6 4- 0 . 0 8 6 9 3 6 7 30 7 -0.0001720 0.0001291 31 7 0.OOt48170.0017479 3 2 7 0.0030832 0 0018611
9 3 0 . 1 5 5 3 7 4 2_ 0 . 0 8 4 0 1 5 8 10 3 0.001319 0.1614824 11 3 0.0288895 0.1324963 33 7 -0 0000405 00017756 34 t 0.00231480.0001028 35 7 0.0000748 0.0015282
12 3 0.0324198 0.0179438 13 3 0.0140259 0.0836615 14 3 0.0365311 0.0224222 36 7 -0.0004199 8
0.0001662 8 0.i88827 0.1223320 9 8 0.1844954 0.0018494
15 3 0.0446271 0.0265447 16 3 0.a320A41 0.0450272 17 3 0.0101214 0.0099939 1o I 0.0437468 0.0924808 11 a 0.00635300.0225827 P a 0.0212177 0.0169046
18 3 0 . 0 0 1 0 0 2 0- 0 . 0 0 7 0 4 8 3 19 3 0.001439r 0.0141955 20 3 0.0082691 0.0137151 13 I -0 0122964 0.0110925 14 8 --0.03294r60.0131814 15 8 0.0406660 0.0247325
21 3 0.0019941 0.0226923 22 3 0.0067253 0.0080913 23 3 0.0045642 0.0119458 16 8 0.01343760.0022804 1t I 0.03115640.0087750 18 8 0.0457191 0.0004305
 7-

Table4.32.1,continued Table4.32.1'continued
(,
(x 106)
NormalizedCoetficients
GEI\.4-T1 G?NI-T1NormalizedCoefficients 10o)

lndex

19 8 0.0148626 0.0113273 20 8 0.0020109 0.0012912 21 I 0.01810080.0025208 16 14 -0.01912260.0382895 17 14 0.01333700.0117613 1 8 14 0 . 0 0 9 2 8 2 80 . 0 1 0 9 4 0 0

22 I -0.0098173 0.0068267 23 8 0.00422090.0067954 24 8 0.00243300.0075574 $ -0.0126448 20
14 -0.0051227 14 0.01032280.0117620 21 14 0.0187760 0.0086994
25 a 0.00142560.0041241 26 8 0.00308890.0021407 27 8 0.0041 794,0.00440s9 14 0.ooa728oo.o1a2407 23 14 0.00461080.0032737 2 4 14 ,0.0186436 0.0014570
22
28 8 0.00056550.0031778 2S 8 -0.0064245 0.0025406 30 8 0.00295360.0005720 25 14 -0.02194180.0132058 26 14 0.00392900.0056350 2 7 14 0 . 0 1 1 9 7 0 20 . 0 0 6 6 3 7 9
31 8 0.0000872- 0.0011845 32 I 0.00088180.0033226 33 I ,-0.00008230.0015750 28 14 0.00210640.0065025 29 14 0.00515250.001940S 30 14 0.0000327 0.0025522

34 I 0.00070830.0007860 35 8 0.00026790.0001772 36 I 0.00104870.0005378 31 14 0.oo72a2a0.0012441 32 14 0.00465690 0069216 33 14 0.0092319 0.0025099

't1 - 0.0001
14 - 0.0004836 238 36
I 9 -0.0555457 0.0975889 10 I 0.1281797,0.0481860 I 0.03877740.4402849 34 14 -0.0010431 0.0002878 35 14 0.0048384-0.0040665
12 9 0.04693800.0132223 13 9 0.02038270.0457820 14 9 0.0371609 0.0179332 15 15 -0.01809480.0080854 16 15 -0.01253210.0322958 17 15 0.0043435 0.0057493
15 I 0.0134441 0.0410187 16 I 0.01657500.0509825 17 I 0.0032015-0.0343246 18 15 -0.03776190.0198247 19 15 -0.01831640.0127675 20 15 -0.0227306 0.0004135
18 I 0.01352i60.0192446 19 I 0.00175660.0086593 20 I 0.0228121 0.0072350 21 15 0.0166205 0.0i49837 22 15 0.02793730.0031033 2 3 15 0 . 0 1 7 7 3 1 80 . 0 0 2 2 8 1 3

21 I 0.0173205 0.0093913 22 I 0.0i251060.0094816 23 I 0.00402740.0103916 24 15 0.0098097-0.013528625 15 0.0019899 0.0022710 2 6 15 0 . 0 1 1 3 7 9 70 0 0 4 7 0 1 0

24 I 0.00389400.0014302 2s I 0.00602970.0098442 26 I 0.00251910.0006592 27 15 -0.0043373 0.0001002 28 15 -0.0082106 0.005375i 2 9 15 0 . 0 0 1 2 7 1 80 . 0 0 2 4 9 1 8
27 I 0.00039900.002i553 28 I 0.00294140.00305s0 29 I 0.00163940.0024790 30 15 0.0028146-0.00927803i 15 0.00045410.0043732 32 15 0.0039107 0.0049375
30 9 665- 0.0039351 31
0.0001 S -0.00389410.0017016 32 I 0.00194610.0007681 33 15 -0.0030055 0.0021726 34 i5 0.00076430.0030005 35 15 0.0002609 0.0028304
33 I 0.00029390.0018691 34 S 0.0012660 0.0015093 35 I 0.00108260.0020731 36 15 -0.00182230.0018634 16 i6 0.0324114 0.0043686 1 7 16 0 . 0 2 9 0 6 8 30 . 0 0 1 8 8 4 a

36 I 0.00031700.0005484 10 10 0.0945596 0.0201041 11 10 0.05205820.0176126 18 16 0.00978800.0050024 19 16 0.01990470 0119326 2 0 16 0 . 0 1 0 6 6 8 50 . 0 0 1 6 9 1 9

12 10 0.00912730.03i6782 13 10 0.04330280.0380383 14 10 0.03699530.0027966 21 16 0.0087331 0 0051553 22 16 0.00008920.0049265 2 3 r6 0 . 0 0 4 9 0 2 90 . 0 1 1 7 6 7 1
15 10 0.00959280.0160812 16 10 -0.01041710.0066056 17 10 0.00210500.0201193 24 16 0.00049080.0062766 25 16 0.00304190.0127972 2 6 16 0 . 0 0 5 8 2 4 10 . 0 0 4 1 5 1 0
18 10 0.0090063-0.010861919 10 0.03535380.0026556 20 10 0 02242010.0080s26 27 16 0.00658490.0041072 28 16 0.00830730.0076860 29 16 0.0021980 0.0055327
21 10 0.00365430.0018356 22 10 0.00500620.0203830 23 10 0.0i997580.0037585 30 16 0.00061890.0056597 31 16 0.0045i410.0048013 3 2 t6 0 . 0 0 2 9 1 8 70 . 0 0 4 1 0 1 0

24 10 0.01735350.0092954 25 10 0.00569070.0044669 26 10 0.00483700.0016645 33 16 -0.0003879 0.0019724 34 16 0.00113370.0026578 3 s 16 0 . 0 0 0 1 3 1 40 . 0 0 1 3 3 6 9

27 10 0.0083100 12
0.0060i 2A tO -0.0072826 0.0012446 29 10 0.00006020.0060144 36 16 0.00134050.0020055 17 17 -0.03831060.0206234 1 8 i7 0 . 0 0 6 1 1 4 20 0 0 8 7 6 6 3
30 10 0.0012265- 0.001
0402 31 t0 0.00251740.0037447 32 10 0.00083080.0019507 19 17 0.027S459 0.0108837 20 17 0.00429340.0449776 21 i7 0.0067459 0.0008396
33 10 0.00023410.0009449 34 10 ,0.0014879 0.0000093 35 10 0.00143130.0008354 22 17 0.0138079 0.0111258 23 17 0.007212s-0.0066031 24 17 0 . 0 0 8 4 6 2 50 . 0 0 1 8 8 1 7
36 10 -0.0003570 0.0004671 11 11 00543322 0.0547288 12 11 0.00541430.0095228 25 17 ,0.0083083 0.0005500 26 17 0.00488900.0082820 2 7 17 0 . 0 0 5 5 5 0 50 . 0 0 1 5 9 1 4

13 i'l 0.04019060.0055015 't4 11 0.0080835 0.0413614 11 0.OOt71710.0249322 28 17 0.00452010.0042606 29 17 0.00457080 0027849 3 0 17 0 0010057 0.0015438
16 11 0.01401560.0064368 17 11 -0.01711080.0175019 1 8 11 0.0127S890.0005971 31 17 0.00593350.0025686 32 17 0.00366020.0018909 33 17 0.0021426 0.0030432
19 11 0.01648040.0134748 20 11 0.0113787 0.0239305 11 0.00928060.0367834 34 17 0.00036260.0025764 3s 17 0.00338630.0024391 36 17 0.0021679- 0.0008490
22 11 0.00937400.0183775 23 11 0.00384900.0136794 24 11 0.01273960.0121180 18 18,0.0044492-0.005064719 18 0.02164670.0031i31 2 0 18 0 . 0 1 0 5 7 7 10 . 0 0 1 3 0 2 4
25 11 0.00557930.0012766 26 11 0.00320450.0050467 11 0.0011883-0.0030951 21 18 0.01683040.0065631 22 18 0.00703110.0102955 2 3 18 -0.0019056 0.0063023

28 11 -0.00062140.0008290 29 11 -0.0093226 0.0004548 30 11 0.00162070.0050752 24 1a O.OO430720 0450725 25 18 -0 0013004-00i06697 2 6 18 0 0090148 0.0075516

31 11 -0.00166310.0058521 32 11 -0.0024010 0.0006463 33 11 0.00555680.0004803 27 18 0.00518980 0059394 28 18 0.00082s5 2 9
0.0003645- 18 0 0 0 2 0 0 9 5 - 00 0 0 1 3 7 8
34 11 0.00124060.0038325 35 11 0.00067850.0036583 36 11 0.0005880-0.0007308 30 18 0.00030350.0008530 31 18 0.00256450.0008491 3 2 18 0 . 0 0 2 2 1 0 50 . 0 0 1 4 7 9 5
12 12 0.003s2800.0117964 13 12 -0.02800590.0864102 12 0.0089681-0.0320668 33 18 - 0.0012393
0.0008702 34 18 0.00272320.0000869 3 5 18 0 . 0 0 1 s 0 0 30 . 0 0 0 5 9 3 7
'15 12 0.a2a33170.O124a72 16 12 0.02088030.0057370 1 7 12 0.03427340.0172570 36 18 0 00006930.0005306 19 19 0.00646380.0104244 20 19 -0.0070980 0.0084586

18 12 0.0261819 0.0165262 19 12 0.00320370.0043292 20 12 0.00405810.0172980 21 19 0 02095150.0158790 22 19 0.00662100.0046952 2 3 19 0 0086827 0.0074916

21 12 0.002A236 0 0127066 22 t2 0.0074377-0.4074426 23 12 0.0215777-O.a166812 24 19 0.00052740.0150179 25 19 0.00918020.00213i7 2 6 19 0 0016260 0.0007199
24 12 0.0123406 0.0095152 25 12 0.0055425 0.0110100 26 12 -0.0196457 0.00s4338 27 19 0.00092540.0062305 28 19 0.00441820.0138003 2 9 19 0 . 0 0 2 1 9 2 00 . 0 0 1 5 0 5 9
27 12 -O.OO171A22A
0.OOO422A 12 0.OOO4024 0.0024269 29 12 -0.0008339 0.0049254 30 19 0 00562660.0026384 31 1S 0.00201720.0028162 3 2 19 0 . 0 0 3 0 8 8 10 . 0 0 1 5 1 4 6
30 12 5- 0.0034240 31
0.003791 12 0.0001491 0.0046159 32 12 0.0017118 0.0041940 33 19 0.00167960.0002369 34 19 0.00059530.0010032 35 19 0.0029031 0.0005s32

33 12 0.00522080.0041268 34 12 0.00045500.0024876 35 23 0.00148410.0020619 36 19 0.0002562-0.000069520 20 0.0017085 0.0135051 2 1 20 -0.0190411 0.018536i

36 12 0.0002182 0.0016068 13 13 0.06154830.0682661 14 13 0.031s3330.0446234 22 20 -0.0133i52 0.0147789 23 20 0A17224aO.OO9A47524 20 0.0060619 0.0003298
15 13 -0.02610510.0049829 16 0.01690750.O20t122 25 20 -0.00373150.0066217 26 20 0.00948300.0109488 27 20 0.0029525 0.0030106
13 0.01307540.0006134 1/ 13
28 20 0.00098050.0011271 29 20 0.00479960.0031625 30 20 0.0000276 0.0035666
18 13 0 00658150.035155i 19 13 -0.00608940.0291709 20 13 0.02664910.0048913
21 13 0 01816940.01.1596922 13 0.01694550.0174453 23 13 0.0104578-0.007s112 31 20 0 0018947 0.0005595 32 20 -0.00158810.0014199 33 20 0.0020822-0.0007525

24 13 -0.0036235-0.0003824 25 13 0.0073795 0.0151883 26 13 0.00272300.0014151 34 20 o.ooog917-0.000397535 20 -0.0007745-0.0008763 36 20 0.0009128 0.0008661

27 13 -0.00598130.0041287 28 21 21 0.00247750.0068510 22 21 0.01322440.0075983 2 3 21 0 . 0 1 0 8 1 9 50 . 0 0 7 6 4 2 8
13 0.00009830 0035308 29 13 0.00114580.0019753
30 24 21 0.0105744 o.oo1ii97 25 21 0.00539680.0031404 26 21 0 0003907 0.0024147
13 0.01467420.0000192 3i 13 0.00568710.0013250 32 13 0.00725560.0002229
33 13 0.00366840.0067767 34 13 0.00808030.0012768 35 13 0.00118490.0044812 27 21 o.oo2o712--o.oo4s4g2 2A 21 0.00249000.0002697 29 21 -0.0093986-0.0059494
36 13 0.00076850.0037948 14 l4 0.05056570.0063741 15 14 0.00617070.0256132 30 21 -0.0074518 o.oo3o982 31 21 0.00233580.0035767 g2 21 0 0011690 C.0058848
 V-
t?2 EXPLANATORYSUPPLEI\,IENT ANDTl.lEBOTATION
COORDINATES OFTHEEARTH 233
4 / TERRESTRTAL
geopotentialrnoclcl(Larrrbeck,1971).This givesnormaliz14
Table4.32.1,continued should be added to thc
(x 106)
GEM-T'1NormalizedCoefficients coefficients of

lndex C : r ( I E R S )= 0 . 1 7x l 0 8 , (4.32-3)
e. (4.324)
S 2 r 0 E R S )= 1 . 1 9 x 1 0
33 21 0 00073190.0008394 34 2t 0.00137430.0006590 35 21 0.00138430.0027357
36 21 0.00074110.002i175 22 22 0.OOt46230.0047142 23 22 0.00090340.002144s
24 22 0.0017322-0.401303325 22 -O.OO14741
-O.OO17574 26 22 0.0109119 0.0091610 For consistelcy rvith the IERS Terrestrial Reference Flalrc, thc C21(IERS) and
27 22 0 00014300.0029341 28 22 0.00484650.0005118 29 22 0 00965620.0044395
22 0.00324580.0055410 31 22 0.00626640.0057325 32 22 -0.0046595 0.0007455 Srr(IERS) should be used in placc of Or(GEN{-T1) and Sr(GENI-T1).
30
33 22 0.00409920 0010974 34 22 A0008294 0.0004221 3s 22 0.00001590.0033003
36 22 0 00057320.0006864 23 23 0.00084460.0002030 24 23 0 00214350.0090055
25 23 0.0045743 0.0024633 26 23 0.00235370.0089516 27 23 0.A0538814.0027372 4.33 Solid Earth Tides
28 23 0.00264030.0063891 29 23 0.0050i18 0.0000780 30 23 0.00158330.0053377
31 23 0.00950470.0056544 32 23 0.00385730.0004759 33 23 0.00062770 0043975 The solid Earth ticle model is basccl on an abbreviatecl lbrm of the \\iahr model
34 23 0.00093000.0021403 35 23 0.00237830.0015551 36 23 0.0012168 0 0005s18 (Wahr, 1981) using the Earth model 1{)66Aof Gilbert and Dzieu'onski(1975). Thc
24 24 0.0023434-O.OO12129 25 24 0.00360650.0038584 26 24 0.0013736 0.0121837
Love numbers for thc iudrrced free spacc pote[tial, k, and for thc vertical and
27 24 0.00193610.0026238 28 24 0 00687620 0150841 29 24 0.0425342 0.0031024
30 24 0.0025120-0.00003773i 24 0.00382980.0019i01 32 24 0.00659050.0053251 horizontal displacencnts, /r and l, havc bectt taken from \\iahr's thesis. Tables 13
33 24 0.0039903-0.000482134 24 0.00672670.0008511 35 24 0.40252770.0021953 and 16. The long period, tliurnal, and semi diulnal ternls are incluclecll'hirtl <lt:gret:
36 24 0.00065890.0014201 25 25 0.00494550.004014i 26 25 0.00398760.0082488 terms are neglectecl.
27 25 0.0118014 0.0031453 28 25 0.00112950.0048096 29 25 0.00830570.0036286
30 25 0.0088475-0.005606231 25 0.40774490.4002757 32 25 0.01317300 0077043
33 25 ,0.4012628-O.OO44224 34 25 0.00625160.0082080 35 25 0 00338920.0015231 4.331 Calculations of the Potential Coefficients The solid-tide-induced flcc sl)ace
36 25 0.00002700.0086477 26 26 0.00342810.0042690 27 26 0.00500800.0040035
potential is most casily rnodeled as lariations irt the standard geopotcnti:rJr:oei-
28 26 0.0034421 0.0016811 29 26 0.00625840.0036878 30 26 0 00320870.0081399
3'1 26 0.00469230.0003639 32 26 0.00105110.0015337 33 26 0.00813430.0055386 ficienis C,. arrd Sn,, (Eanes el al., 1983). The Wahr model (or auv other having
34 26 0.0010816 0.0090061 35 26 0.01434380.0001412 36 26 0.0084463 0.0110849 frequencv-dependent Love numbers) is rlost efliciently cornpulcd in two steps- The
27 27 0.00689450.0034538 28 27 0.00992470.0013336 29 27 4.0074725-.
0.0421507
30 27 -0.0019206 0 0078152 31 27
first step uses a frequencv-independent Love llulnber t2 and tn e\aluatiou oI the
0.00699600.0r22525 32 27 0.00309490.0030221
tidal potcntial in the time domairr frorl a lunal and soliu ephemelis. The second
33 27 -0.0103323 0.0021297 34 27 0.00689740.0005442 35 27 - 0.0191294
0.00277A7
36 27 0.0101912 0.0042305 28 28 0.00676890.0019493 29 28 0.0103226 0 0019625 step corrects those arguments of i1 harrnonic expansionof the tide-gr:lerating 1;o-
30 28 0.00896s80 0051589 31 28 0.00034280.0017261 32 28 0.0015717 0.0023530 tential for which tlic cllol fionr usiug the k2 of step 1 is above sor r cutofl.
33 28 0.0i08851 0.0018952 34 28 0.00455020.0081979 35 28 ,0.01089360.0233542
36 28 0 00694800.0056619 29 29 0.00863360.0031601 3A 29 0.AO4a227 0.0441240 The changes iu rrornralizrxl secotd-depilee geopotcrtial coeffi(rients tirr step 1
31 29 0.00542150.0059502 32 29 0.00334280.0025024 33 29 0.0213209 0 0001932 are
34 29 0 00381130.0044842 35 29 0.00397910.0005664 36 29 0 00134i5 0.0024474
30
33
36
30 -0.0015075
-0.0004221 31
30 0.00252i7-0.0134935 34
30 -0.0015551 0.0020261 31
30
30
31
0.00245270.0084230 32
0.00611610.0000268 35
0.00023190.0000737 32
30
30
31
0.0082893 0.0016713
0.00371280.0028852
0.0007616 0 0027125
s^,=ftr.fi i ff ,*,,,",,,,, (4.33r-r)
33 31 0.00017450.0011s66 34 31 0.00235340.0022983 35 31 0.0013305 0 0010339
36
34
31
32
0.00355160.0013604 32
0 0008166000i1610 3s
32
32
0.00072770.0005068 33
0 0039791 0.0005664 36
32 0.0024100 0 0000219
32 -0.00088040 0003000 Ac:r- 'As2,
=:l'rr,#t?fi ,^,,
PuGi,,ii)" (4.33
r 2)
33 33 0.00026760.0003603 34 33 0.0010234 0.0013813 35 33 -0.00028810.0016655
36 33 0 00239720.0026170 34 34 0.00020930 0006039 35 34 0.0005168 0.0001237
36
36
34
36
0.00083240.0017560 35
0.00014840.0004197
35 0.00009380.0001450 36 35 0.00021520.0007347 - 'As2r
Ac?? =irr:r.*fffe,x,,,oit" ,,^" (4.33
r-3)
' C2o
does nol ncludethe zero-trequencyte rmi see EqLiaton 4 332-5 lor ihe adjustedvalue
" - 2 1 a n d 3 2 1 s h o u l db e t h e I E B Sv a l u e s si e e E q u a t o n s4 . 3 2 - 3a n d 4 . 3 2 - 4 l o rr e c o m m e n d evda r u e s . where

kz = norninal second degree Lrx e number;

R" = equatolial radirrs of the Earth;
 2g EXPLANATORY
SUPPLEMENT COOROINATES
ANDTHEROTATION
OF THE EARTH 4t:tc

GM = gravitational parameter for the Earth; Table4.3i11.1

GM1= gralitalional parameter for the Moon (l = 2) and Sun (7 = 3); Step2 SolidTideCorrec'tionsWhenk2= 0.30 in Step'l (Usinga
r/ = distance from geocenter to Moon or Sunl CutoftAmplitudeot 9 x l0-12for A,i6/cHs)
di = body fixed geocentric latitude of Moon or Sunl Long-PorfodTld€s (, = 2, n = Ol
)t = body fixed east longitude (from Greenwich) of Sun or Moon. None6xc€ptzero-frequency tide,

The changes in normalized coefficients from step 2 are: DlumalTldor (r = 2, m= 1)

t, argum€ntmultipli€E
I n+rn even Number r
Doodson s h p /4m6/GHsx1012'
^.c, -ttS, =t^ L 6 k E H s \ -i)
\ ',i6. ' (4.3314) A
n+n odd
{r,,r) 145.555 (q) 000 0 -16.4
16:|.555 (Pi -200 0 -49.6
where 165.545 0 0 -1 0 -9.4
. (1 itn=0 (K1) 000 0 507.4
o o=
' i (4.33r-s) '165.565 001 0
to ifn/o 166.554 btn 100 -15.2

6[s = difierence between Wahr model for t at frequency s and the nominal Scmi-DlumalTldee(n = 2, m = 2l
value &2 in the sense ks - &2; n, argument
multipliers
IIs = amplitude (m) of term at frequency s from the Cartwright a.nd Taylor Doodson
Number r s h p A 4n6lc4xt0tz-
(1971) and Cartwright and Edden (1973) harmonic expansion of the
tide-generating potential; 255.555 (M21 2 0 00 0 0 39.5
273.555(Sr)22-2000 18.4
o" =fi.p =D:Etnpt;
t = six vector of multipliers of the Doodson (1921) variables;
0 = the Doodson variables;
6520= 0.
a nomiual &2of 0.3 and an amplitude cutofi of 9 x 10-12 change in normalized
The Doodson variables are related to the fundamental arpruments of the nuta- coeftcients, the summation S(n, m) requires six terms for the diurnal
tion series (see Chapter 3) by (n = 2, n = l) modifying C21 and 521 and two semidiurnal terms (z = 2,
= 2) modifyinge2zand Sz. With the exceptionofthe zero-frequency
tide, no long-
s = F + Q = h (Moon's mean longitude); terms are necessaxy,Table 4.331.1 gives required quantities for correcting
h = s - D = h (Sun's mean longitude); (2'l) and (2,2) coefficients. The correction to ?zo is discussed in more detail
p = s - I = B4 (longitude of Moon's mean perigee);
Nt = -Q = 0s (negative longitude of Moon's mean node);
h = s - D - l' = 0s (longitude of Sun's mea.nperigee); The step 2 correction due to the Kl constituent is
r = 0c + T - s = 0t (time angle in lunar days reckoned from lower transit);
dg = mean sidereal time of the conventional zero meridian. (A-C21
x 10r2)y,= 507.4sin(r + s)
The normalized geopotential coefficients (Q.,3,-) = 507.4 sin(9r + zr)
are related to the unnormal-
ized coefrcients (C, ,5" ) by = _507.4sin(Oe), (4.331-9)
(AS21x l012)6,= -507.4cos(es). (4.331-10)
cnn = N^enn, (4.33r4)
s,^=N^5", (4.331-1) The total rariation in geopotential coefficients due to the solid tide is obtained
t'2 the results of step 2 (Equation 4.331-4) to those of step 1 (Equations
., (n + n1l(2n+
| --------------- t)(2 - 6o)f
/Yan = |
__i:=
| (4.331-8) 4.331-2,and 4.33r-3).
L \n+mtt I
236 EXPLANATORY
SUPPLEMENT COORDINATES
ANDTHE BOTATION
OF THE EARTH 237
4332 Theatment of the Pennanent fide The mean value of Aezo ftom equation coefrcient that includes the zero-frequencyterm; then the GEM-T1 values of
4.331-1 is not zero, and this permanent tide deserves special attention. Ideallg the with the permanent tide restored are
mean value of the correction should be included in the adopted value of Qs and
hence not included in the Aezo. The practical situation is not so clear, because eioQau -Tr)= -484.16499 x 10-6- (1.39119
x t0-8)x0.30,(4.3324)
satellite-derived values of ?2s, as in the GEM geopotentials, have been obtained
using a mixture of methods, some applying the corrections and others not applying
eio9nu - Tt) = -484 169025x 10-6' (4.332-s)
it, There is no way to ensure consistency in this regard short of re-estimating C26
valuesfor el6 are recommendedfor use with the respectivegravity field and
with a consistent technique. If this is done, the inclusion of the zero frequency term
be addedto the periodictidal correctiongiven as Ae;O in Equation 4.332-1
in Equation 4.331-1 should be avoided, because /<2 is not the appropriate Love
get the total time.dependent value of C26.
number to use for such a term.
The zero frequeucy change in e1 can be removed by computing Ae2s as
Solid Tide Elfect on Station Coordinates The variations of station coordi-
caused by solid Ea.rth tides predicted using Wahr's theory a"realso most
Adio = 46ro1Bo"ation4.331-l)- (adro) , (4.332-r) implemented using a twostep procedure. Only the second degree tides
necessaryto retain 0.01m precision.Also terms proportional to y, h+, h-, z, l+,
where ; and le- axe ignored. The first step uses frequency-independent Love and Shida
dhmbers and a computation ofthe tidal potential in the time domain. A convenient
(te 2o)= eoPolr, of the displacement is given in the documentation for the GEODYN
= (4.4228 (Ma.rtin el al.,1976, p.5.5-1) . The vector displacementofthe station due
x 10-8X-0.31455)t2
i tidal deformation for step 1 can be computed from
= -1.39119 x l0-8t2.

Using t2 = 0.30 then ^=Dlffil

{o,,r*r.n,*r*
-
l, (+ r),*,n'-?]"),
(4.333-r)

(aero)= -4.1736x ro-e, (4.332-2)

GM; = gravitational parameter for the Moon U =2) or the Sun f = 3);
@1,1= = gravitational parameter for the Earth;
(a.lz)= - (aero)t/5 =9.3324xt}-e rd 11r-3\
&, R, = ,rott vector from the geocenter to Moon or Sun and the magnitude of
that vector;
The decision to remove or not to remove the mean from the corrections depends r' r = unit vector from the geocenter to the station and the magnitude of that
on whether the adopted Qs does or does not already contain it and on whether &2is
vector;
a potential "solve for" parameter. If k2 is to be estimated then it must not multiply lrz = nominal second-degree Love number;
the zero ftequency term in the correction. In the most recent data reductions leading lz = nominal Shida number.
to GEM-T1, the total tide correction was applied. If we assume the more recent
data has most of the weight in the determination of ?29 then we conclude that the If nominal values for i2 and lz of 0.6090 and 0.0852, respectively, are used with
permarent deformation is not included in the GEM-TI value of d26. Hence, if t2 is of 0.005 m of radial displacement, ouly one term needs to be corrected in
to be estimated, first (ACr6) must be a.ddedto ezo and then AEls should be used 2. This is the r(1 frequencywhere I from Wahr's theory is 0.5203 OnIy the
in ptace of d2s of Equation 4.331-1. The [2 used for restoring the permaneut tide displacement needs to be corrected and to sufficient accuracy this can be
should match what was used in deriving the adopted value of d26. as a periodic change in station height given by
The GEM-T1 value of d26 is -484.16499 x 10-6 and does not include the
permanent deformation. The tidal corrections employed in the computations leading 5 (4.333--2)
6hs71= 66*,P*, 3 sin d cosd sin(gx, + .l),
to GEM-Tf were equiralent to Equatiou 4.331-1 with tz = 0.30. Let dio denote ( U"
238 EXPLANATORY
SUPPLEMENT OFTHE EARTH
ANDTHEROTATION
COORDINATES 239
where dl = tt(mLr+mzi+ (1 +n3)2). O is the meana.ngularvelocity of rotation of the
: m1a.resmall dimensiouless parameters' ml, rn2 descdbing polar motion, and
6hK,= hK, (Wahr) -ft2 (Nominal) = -0.0887;
raniation in the rotation rate; r is the radial distance to the station.
Ilr, = amplitude of Kl term (165.555)in the harmonic expa.nsionof the tide
generatingpotential = 0. 36878m; eglecting the variations in m3 t'hat induce displacements that a.re below the
the mr and m2 terms give a first-order pertubation in the potertial y
4t= geocentric latitude of station;
) = east longitude of station; , re85)
Or, =Kr tide argument= 7+J= 08 +7',
022
or, simplifying' AV(r, O, .\.1= -"f sin2o(q cos) + m2sin )), (4.334-2)
disTa= -6.9253.;tt @cos@sin(@,+ )). (4.333-3)
O is the colatitude, a,rrd I is the eastward longitude.
The effect is maximum at 0 = 45o where the a.rnplitudeis 0.013m.
liet us define the ra.dial displa.cement S,, the horizontal displacements 56 and
There is also a zero-frequency station displacement that may or may not be positive upwa,rd, south, a,rrdeast, respectively, in a horizon system at the station
included in the nominal station coordinates. The mean correction could be removed to AV using the formulation of tidal Love numbers (Munk and MacDonald,
analogously to the discussion above. When baselines or coordinates are compa.red at pp. 24-25).
the few-cm level, care must be taken that the correction vras handled consistently.
It is essential that published station coordinates identify how the zero-frequency
contribution was included.
^ .av
Jt = n-, (4.334-3)
I
If nomina.l Love and Shida numbers of 0.6090 and 0.0852, respectively, a.reused
with Equation 4.333-1, the permatrent deformation introduced is Se= __:aeA% (4.3344)
I
^ 11- (4.334-5)
sl=_.-_a ray,
Ar.r= /4 11?--4) {J
I Sln
(3 l
i r \ r s r n' - @ _ \r / m
= -o.tzot' (4.333-5) g is the gravitational acceleration at the Earth's surface, and lr, I are the
body tide displacement Love numbers'
in the radial direction, a.nd In general, these computed displacements have a nonzero average over any
timespan because m1andrn2, used to find AV, have a nonzero average. Con-
, the use of these results will lead to a change in the estimated mean
t---

A r . e' , = 1 / i 1 0 . 0 8 5 2 ) ( - 0 .3 1 4 5 5 ; 3 c o
d s i n@ coordinates. When mean coordinates produced by difierent users a're com-
y +71
at the centimeter level, it is importart to ensure that this effect has been
= -0.05071cosd sind m (4.3334)
consistently. It is recommended lhal m1 and m2 used in Equation 4'334-2
replaced by paxameters defined to be zero for the terrestrial reference fiame
in the uorth direction.
discussedin the previous section.
4.334 Rotational Deformation Due to Polar Motion The variation of station coor-
Thus, define
dinates caused by the pola.r tide should be taken into account. Let us choose l, 9,
a.nd 2 as a terrestrial system of reference. The 2-axis is oriented along the Ea,rth's xp=ml-x' (4.3344)
mean rotation axis, the fi-axis is in the direction of the a.dopted origin of longitude,
(4.334-7)
and the i-a>ris is oriented along the 90o E meridian.
The ceutrifugal potential caused by the Earth's rotation is
t and t axe the values of nr1 arLd.-ru2 for the ITRF. Then, using ft = 0'6,
v=rtzlllar -(r.a)21, (4.334-1) 085,andr=a=6.4x10Pm,

$r
 2& EXPLANATORY
SUPPLEMENT COORDINATES
ANDTHEROTATION
OF THEEARTH 241

,-r - =
4rGp, I (n + m)l 1"' t +k1 (4.34-2)
S, = -32 sin29(x, cos) - yo sin )) mn, (4.334-8) - -
g I (n n)l (2n+ 1)(2 6"^)) z r +r '
Se = -9 cos2€(tp cos,\ - yosin )) mm, (4.334-9)
Sr = 9cosO(.rpsin) + yocos)) mm. (4.33+-10) g -- GE l az = 9.798261ms-z;
G = the universalgravitational constant= 6.6?3 x 10-11m3kg ls-2.
for xo and.yo in seconds of arc. -3;
pr. = density of seawater= | 025kg m
Taking into account that x, and yo vary, at most, 0.8 axcsec, the maximum radial
( = load deformationcoefficients(14= -0.3015,14 = -O.1954, k'q= -O.tZZ,
displacement is approxirrately 25 mm, and the maxirnum horizontal displacement
l/, = -o. t032,ti = -0.0892);
is about 7 mm.
Cr-, .9,,. = o"eao-tide coefrcients in meters for the tide constituent s (see Ta-
If X, Y, ald Z are cartesiar coordinates of a station in a right-hand equatorial
ble 4.34.7);
coordinate system, we have the displacements of coordinates
9, = argumeut of the tide constituent s as defined in the solid tide model (see
r, orevioussubsection).
ldX,dY,d4r = Plr1So,J.r,S,l (4.33+-rr)
The summation, !*, implies addition of the expression using the top signs
' (the prograde waves
where 'l Qn, and .$-) to that using the bottom signs (the retrograde
f c o s9 c o s) cosOsin^ -sinO mw G. and.S"-,). The ocean-tidecoefficientsCt, *a $,' as used here a.re
R=l -sin^ cos^ 0 (4.334-r2) related to the Schwiderski(1983)ocean-tideamplitude and phaseby
l.
I sinOcos) sinOsin,\ cosO I
C" - it, = -it|^ei("**x't, (4.34-3)
Equation 4.334 10 can be used for determination of the corrections to station
coordinates due to pola,r tide. where
The deformation caused by the pola.r tide also leads to time-dependent pertur-
bations in the C21 and S21geopotential coefficients. The change in the external po- . 4- = o"""tt-tide amplitude for constituent s using the Schwiderski notation;
e*- = ocean-tide phase for constituent s;
tential caused by this deformation is tAV, where AV is given by Equation 4.334-2,
f 0 ) Semidiurnal, and long period
and & is the degree-2 potential Love number. Using /r = 0. 30 gives
=
X" { r/2 f1; (or in general for constituents with 11"> 0),
}
-rl2 Or, P:,Qti (or in generalfor constituentswith FI" < 0).
| )
en=-1.3x10-e(xo), (4.334-13)
For clarity, Equation 4.34-1 is rewritten in two forms below:
3zr= -1.3 x lO-e(-yp), (4334-14)

where xo and y, are in seconds of a,rc and are used instead of z1 and -mz so lhat dC,^= F, ! tt4,, * c;,l"os o, + (E,n+ s;,,) sinoJ, (4.344)
no mean is introduced into Q; a.nd 321when making this correction. {u,m,

4.34 .OceanTideModel A-c, = p^^! t4,, sinto"+ .i,,, + x")+ c;n sin(os+ e;n+ x)l' (4.34-5)
s\nfi)
The dynamicat effect of ocean tides is most easily implemented a€ periodic variations
in the normalized geopotential coefficients, The variations can be writteu as follows
aS,,' = F,- | t{{,. * s;) or - (qn,,,+ c;-) sinerl, (4.344)
{r,rr)
"os
(Eaneset ot, 1983):
aS^^= P^ | t4,.co(o" + ei, +x")
^e* - i^S^ =F^ > Dtd.*+,.9--)l6., (4.34-r)
s\n,mt

+ C;,t cos(Or + r; + Xr)]. (4.34-7)

242 EXPLANATOBYSUPPLE[.IENT Table4.34.1,continued
Oiean Tide Coetficientsfrom the Schwiderskilvlodel
Table 4.34.1
Ocean Tide Coetficientslrom the Schwiderskil\4odel nm €i,.
Argumenl
Number (cm) (des) (cm) (cm)
Argumenl €!nn
Number (cm) (deg)
135.655 1 .5373 313.735 ,.3715 .3882
',t07.346 .0935 .2994
135.655 Qt 1 .3t36
057. 555 &,20 .6215 221.672 .8264 .9284 1 .2930 2a8.992 -.0953 -.2770
135.655
ss30 .0311 1.735 .0019 .0621 135.655 1 .2209 112.383 .0841 .2042
ss40 .1624 92.674 .3244 ,.0152 1 .0396 2a7 .824 -.012 t -.0377
135.655
.2624 251 .737 -.4991 -.1647
1 2.4186 3 1 3 . 7 1 6 -1.6715 -1.7481
ss60 .4363 145.744 .4512 -.7213 145.555
145.555 O1 1 1.3161 83.599 .1467 1.3079
065.455 Mn20 .5313 258.900 1.O42a .2046 O1 1 1.4301 276.242 .1565 -1.4215
145.555
065.455 Mn30 .0317 94.298 .0632 .0047 145.555 Ot 1 .9505 109.128 .3115 .8980
065.455 Mm40 .0998 69.054 .1863 .0713 Ot 1 .1870 242.623 ,.0409 .1825
145.555
065.455 Mm50 .2279 292.291 -.4214 .1729 -.6256 .6498
163.555 R 1 .9020 313.912
065.455 Mn60 .0660 39.882 .0447 .1014 P1 1 .2976 39.958 -.2281 .1911
163.555
Pl 1 .6346 258.311 .1286 -.6215
075.555 ry20 .8525 251.956 1.6211 .5281 163.555
075.555 Mf30 .0951 148.236 .1001 .1617 163.555 B 1 .4130 104.438 .1030 .4000
MJ40 .2944 102.723 .5422 -.1315 163.555 A 1 .0583 276.591 .0067 .0579
075.555
075.555 MJ50 .2960 223.167 -.4050 -.4318 165.555 K1 1 2.8158 315.113 1.9950 1.9472
075.555 Mf60 .0880 107.916 .1675 .0542 165.555 K1 1 .8925 33.752 .7421 -.4959
165.555 Kl 1 1.9121 254.229 .5197 1.8401
165.555 K1 1 1.2111 104.672 .3068 1.1716
165.555 K1 1 .1645 281.867 .0338 .1610
245.655 N2 2 .6516 321.744 .4030 .5120
245.655 N2 2 .10a4 171.923 .0152 .1074
245.655 N2 2 .2137 141.779 .1322 -.1679
245.655 N2 2 .0836 5.034 .0073 .0432
The summation over .r(2.n ) should include all constituents for which Schr-iderski 245.655 2 .0674 346.544 .0157 .0656
has computed a model. Except for cases of near resonance! the retrograde ternts W 2 2.9551 310.553 -2.2453 1.9213
do not produce long-period (> 1 day) orbit perturbations for the diurnal and semi- M2 2 .3610 168.623 .0712 .3539
M2 2 1.0066 124.755 .a270 -.5738
diurnal tides. The root mean square of the along-track perturbations on LAGEOS -.0165
M2 2 .2751 356.561 .2746
due to the combination of all of the retrograde waves is less than t 5 cm. M2 2 .4130 329.056 .2124 .3542
For computing inclination and node perturbations. only the even-degrce terns 314.011 -.6682 .6456
52 2 .9291
are required, but for the eccentricity and periapsis thc odd degree terms are rot 52 2 .2633 201.968 -.0985 .2442
negligible. Long-period perturbations are only produced when the degrce (n) is 52 2 .3716 103.027 .3621 .0838
52 2 .1365 3.772 .0090 .1362
greater than 1 and the order (m) is 0 for long-period tides, l for diurnal tidcs, and 2
for semidiurnal tides. Finally, the ocean-tide amplitudes and their cffect on satcllite
c 2 .1726 2ao.38l .1698 .0311
K2 2 .2593 315.069 .1a32 .1836
orbits decreasewith increasing degree, so truncation above degree 6 is justified fol K2 2 .0943 195.007 -.0244 -.0911
LAGEOS. K2 2 .1059 103.521 .1029 .0247
Thus, for the diurnal tides (Q1, Ot, Pt, Kj) only the n = 2, 3, 4, 5, 6 and m = I 275.555 K2 2 .0382 .411 .0003 .0382
K2 2 .0467 241357 -.0458 .0092
terms rreed be computed. For the semidiurnal tides (N2, M2, 52, K?) only n = 2. 3,
4, 5, 6 and m = 2 terms need be computed. For the long-period tides (S",, M,,, M1) (1) The Ooodsonvariablemullipliers(t) are coded inlo the argumenlnumber(A) atter Doodson(1921)as

only n = 2,3, 4, 5, 6 and m = 0 terms need be computed. Table 4.34.1 gives the A = \ ( n 2 + 5 ) 1 r y + 5 ) \ n a + 5 ) ( n 5+ 5 ) ( r b + 5 )
values required for each of the constituents for which Schwiderski has computed (2) For the long-periodtides (rn = o), lhe value ot Oj,m used to compute 4nm and Enm was iwice that
a model. Note that the units in Table 4.34.1 are centimeters and hence must be shown to accountfor the combinedelfectoJthe retrcgrad€and progradewaves.
scaled to meters for use with the constants given with Equation 4.34 1. (3) The sphericalharmonicdecomposilionoJ Schwiderski'smodels was computed by C Goad ol ohio
State University.
lliillii;li'ii,ilii+lrEgiriii
# iiii iiiii
i* lri:ii:ir il*iii+
i
ii.,ilil;iiuii;1il*'ii
ot
M2 sz\01 Nz Pt Kz Qt MI Mh Ssa
AMP PHAS AMP PHAS AIVP PHAS AMP PHAS AI\,IP PHAS AMP PHAS AI\4P PHAS AIJP PHAS AIVP PHAS A[]IP PHAS AMP PHAS 6'o
o i.
Chlboltn 1.37 46.1 .47 i 1 . 1 3t -58.5 .10 127.5 .31 63.7 10 58.9 .12 5.4 . 0 11 0 6 . 4 .09 4.2 .03 26.8 .04 27.5 a9
Madid64 1.37 -46.4 .46 6 t . 4 .23 70.0 .03 151.3 .29 106.3 .07-69.4 .12 63.8 .02 48.1 .04 25.5 .02 -98.r .01 102.4
Robleds2 1.37 -86.4 .46 6 1 . 5 .23 70.0 .03-151.0 .29 106.3 .07 69.4 .12 63.8 .02 48.0 .04 25.5 .02 98.1 0 1 1 0 25 o
Cebrer26 1.37 -87.0 .45 62.0 .23 70.4 .O3 149.7 .28 106.6 07 69.5 .12 64.6 .42 44.1 .o4 252 .O2 97 8 .01 102.8
Haystack .96 176.8 .26 154.4 .39 -5.4 .26 4.8 .21 165.2 12 3.2 .07 161.6 .05 1.7 04 10.5 02 5S.2 04 96.0 o
Westford .96 -176.9 .26 154.8 .39 54 .26 4.A 2 t 165.2 12 3.2 .07 161.6 .05 1.7 .o4 10.5 .02 59.2 .04 96.1
l\,4arpoint .95 158.7 .24 t704 .33 1.9 .22 1.6 1 6 14t.7 11 .7 .06 175.0 .05 2.3 .02 17.7 .41 74.3 .05 117.3 o
A)
NRAO'140 .68 152.1 .2Q 173.2 .29 6 .20 .9 . 1 1 133.7 .10 3.6 05 177.1 04 2.9 .01 20.5 .00 66.4 .04 121.8 :J
Richmond .82 165.9 .24 162.3 .t8 15.2 .12 25.1 1 4 146.6 .06 13.1 .06 169.7 .03 35.0 .05 154.8 .02 160.5 .1A 122.2 r
H B A S0 8 5 . 0 8 - 1 6 8 . 8 .15 128.1 47 32.5
o
32 17.1 .05 69.3 .15 30.8 .04 r09 1 07 10.0 .o2 1677 .Ot 157.3 01 1373
q
Platlvil .22 111.6 .i3 168.3 .47 39.5 31 25.2 .02 , . 1 12 .15 387 .03 151.3 .06 178 .01 6.5 01 49 2 .02 92.7
vLA .04 169.6 .14 131.9 .57 38.0 37 22.8 .06 67.0 .18 36.8 .04 108.8 .08 146 . 0 1 1 3 8 . 3 . O i 1 2 35 .02 90.2
Vemal .22 96.5 13 170.8 .58 44.5 37 29.9 .03 18.3 .18 43.6 .o2 14A.3 .At 2t 5 .01 1.9 .Oi 49 3 .03 830
Flagstaf .O1 2.A 14 129.2 .69 41.6 44 26.3 o7 63.9 22 4A5 .04 103.0 .09 176 .01 88.3 .0r r05.0 .03 79.9
3
j'
Yuma .26 57.2 21 98.3 .90 40.1 .57 24.4 13 70.8 28 38.9 .07 837 r1 159 .01 117.4 01 1452 .05 13.7
Ely 24 70.2 .12 164.2 .76 47.4 4a 32.4 06 27.1 24 46.6 .02 132.5 .09 23I .01 16.7 .01 55.0 .04 77.6 3
BlkBune 2A 42.4 .19 -99.9 .96 415 61 26.3 .13 65.6 30 40.6 .o7 82.1 12 17.4 .01 100.7 .01 1 3 77 05 72.9 9
Ocotillo 42 49.5 .26 86.6 106 396 67 24.3 .17 67.8 .$ 3A4 . 0 9 7 6 . 2 . 1 3 1 56 .01 110.5 .01 1s48 .06 71 6
DeadLlanl 28 34 2 .18 99.2 100 421 63 27.O .13 62.5 .31 412 .07 81.2 .12 181 .01 89.6 .01 130.4 06 734
l\,lonPeak .51 45.7 .27 a3 1 1.12 39.7 .71 24_5 .19 66.2 35 3a7 10 73.6 13 15.7 .01 106.4 01 1s7.0 .06 71.4
PinFlals .38 39.5 .22 89.7 107 41.0 68 25.9 16 63.5 33 40.1 o a 762 t3 17.1 .ol 96.2 .01 143.t .06 726 o
Godvenu .22 13.4 .ls 108.7 .98 43.6 .62 28.6 12 o
56.7 31 42.A 05 840 .12 19.7 .01 698 .01 109.4 06 73.8 o
l\Iojave12 .22 -11.3 .ls 1075 436 (o
.99 .62 2al 12 55.3 .31 42.9 .0s 825 12 19.8 .01 67.2 .O1 107.7 .06 73 I
GoldEcho .22 111 15 108.6 .98 43.6 .62 281 .12 56 5 .31 42.8 .05 o
83.7 12 19.8 .01 68.8 .01 108.5 .06 73I o
Otay .70 45 0 35 74.0 129 38.2 .81 23.1 .24 647 .40 37.1 12 64.2 15 14.6 02 104.4 .41 167.0 .O7 71.1 -
GoldPion .22 -1O.3 .14 108.6 .99 43.7 .62 28.8 .12 55.O .31 429 .05 82a .12 i9.9 .01 66.1 01 106.3 06 741
Goldl\,4ars .23 -8.5 14 109.7 99 43.8 62 289 .r2 54.3 .31 43 1 .05 83.1 12 2AO .01 64.4 01 104.9 06 742 A)
Lalolla .84 41.1 .36 718 136 38.4 .85 23.3 .25 632 43 37 3 .13 66.3 16 148 .02 994 .Or 1676 .08 71.3 o
PBlossom .40 23.8 .19 -89.1 1 15 423 72 27.3 .17 57.4 36 41.5 .o7 73.4 .14 1A.s .01 7 7 . 2 A 1 1 3 13 . 0 7 72.6
J P LM V 3 . 5 1 2 7 . 0 .21 44.7 1.21 419 76 26.9 .18 57.9 .38 41 2 .08 7i.1 .14 18.1 01 8 1 . 2 . 0 1 1 4 13 07 71.9
JPL t\.4V1 .51 27.0 21 84.7 t _21 41.9 .76 26.9 18 57I 38 41.2 .08 71.1 14 1 8 . 1 .01 8 1 . 2 . 0 r 1 4 13 07 71 9
JPL t!,1V2 .51 27 A .21 44.7 1.21 41.9 .16 26 I .18 57.9 .38 41.2 .04 71.1 14 181 0l 81.2 01 141.3 .O7 719
ovRo 130 .28 32.6 .10 140.6 1.01 46.8 .63 321 .10 37.6 .31 46.3 .o3 914 t2 23.2 .01 38.4 01 73.4 .06 74.4
ovRo 90 28 32.7 .10 140.6 1.01 46.8 63 321 .T0 37.7 .31 463 03 915 t2 2 3 . 2 .01 38 5 01 73.5 06 74.4
 ;o
M2 s2 K1 o1 N2 K2 01 Mt ssa €s
6'o
AI\,,IPPHAS AMP PHAS AMP PHAS AMP PHAS A[.'P PHAS A[.IP PHAS AIJP PHAS AMP PHAS AMP PHAS AI\,IP PHAS AMP PHAS
i0i
Pverdes .80 31.7 .32 -69.9 1.42 39.9 .90 24.9 .25 58.3 .45 39.0 .12 -63.4 .17 16.2 .o2 44.2 .01 165.3 . 0 8 71.9
Mafibu .U 27.1 .27 -72.8 1.37 40.8 .86 25.4 .23 56.5 .43 40.1 .10 64.2 .16 17.O .o2 -4o.4 .01 158.0 .o8 71 7
SaddfePk .64 -26.4 .27 -735 1.36 40.9
9;
.86 25.9 .23 56.5 .43 40.2 .10 64.5 .16 t7.1 .02 -80.3 .01 157.1 . o 8 71.7
Gorman .45 -'13.3 .18 846 1.25 43.0 .78 24.1 .18 -51.4 .39 42.6 .07 68.2 .14 19.3 .01 -67.9 .ot 128.7 .o7 72.3 ;=,
Mammothl .35 39.7 .09 155.2 1.O2 47.A .64 33.2 .10 30.s .32 47.3 .02 -95.6 .12 24.3 .o2 33.5 .01 65.9 .06 74.2
SanPaula .59 -20.6 .24 73.3 1.38 41.6 .a6 26.7 .22 -53.4 .43 41.1 09 63.2 . 1 6 1 7 . 5 .o2 -74.2 .01 150.7 .08 72.1 oo
oo.
Vandenbg .92 - ' l 1 . 2 -27 59.1 1.65 42.O 1.O3 27.2 . 2 4 - 4 5 . 7 .51 42.1 .11 52.4 . 1 9 1 8 . 3 .02 -64.6 .01 162.9 . 1 0 71.8
Vndnberg .97 -11.3 .26 58.7 1.67 42.O 1.O4 2 7 . 3 .28 45.0 .52 42.3 .11 5 1 . 6 .19 14.2 .02 -64.3 .01 164.9 . 1 0 7i.6 -
Quincy .62 56.2 . 1 3 1 5 0 . 31 . 1 1 51.5 .69 3 7 4 .12 2.5 .34 51.1 .01 138.9 . 1 3 2 8 . 6 .03 20l .02 49.4 .o7 74.2
Hatcreek .74 59.8 . 1 6 1 3 8 . 8 1 . 1 5 52.7 .71 38.7 14 5.5 .35 52.1 .02 115.6 . 1 3 2 9 . 9 .03 17.3 .02 45.8 . o 7 74.1
FonOrd .89 20.3 .04 53.4 1.55 46.4 .96 31.9 .23 25.1 .47 46.4 .05 34 1 . 1 8 2 3 . 1 .o3 -41.2 .01
='
(o
92.8 0 9 73.1
Vacavill .74 42O .07 133.3 1.37 49.2 .85 35.1 . 1 8 - 1 0 . 8 .42 49.4 .o2 1.4 . 1 6 2 6 . 3 .03 -30.1 .01 62.6 .08 73.9
SanFranc 1.04 33.8 .05 105.5 1.53 48.4 .95 34.2 . 2 2 - 1 2 . 9 .47 48.8 03 4.1 . 1 4 2 5 . 4 .03 -33.6 .01 7 1 . 1 .09 73.8 3
Presidio 1.04 33.8 .05 105.5 1.53 48.4 .95 34.2 . 2 2 - 1 2 . 9 .47 48.8 .03 4.1 .14 25.4 .03 -33.6 .01 71.1 .09 73.8
FlReyes 1.19 36.4 .10 81.4 1 69 44.6 1.05 34.5 .27 8.7 51 49.1 .04 16.7 . 1 9 2 5 . 6 .o4 -32.5 .o2 66.3 . 1 0 73.9
Yakalaga 2.49 100.7 .89 137.4 1.32 A7.4 .a3 75.O .43 74.1 .43 85 7 .23 133.1 . 1 5 6 6 . 9 10 4 . 7 .09 32.1 . 1 3 59.6
Giloreek .78 100.7 .32 140.1 51 96.3 .35 88 1 .10 87.5 .16 95.3 .08 134.7 .06 79.4 .09 16.7 .06 23.1 . 1 0 71.7
e
Kodiak 2.82 1131 1.O7 147.3 1.61 98.0 1.05 85.7 .51 94.5 .52 96.9 .27 144.7 . 1 9 7 9 . 7 .12 . 8 .10 38.8 . 1 5 56.6
o
Kauai .95 111.3 .38 130.0 1.15 61.9 .66 56.2 .21 132.3 .33 61.9 .12 1347 . 1 0 5 8 . 4 .05-140.1 .03 170.8 . 0 6 112.5 0)
SndPoint 1.94 127.6 .80 150.9 1.45 112.9 1.06 104.1 .33 122.2 .49 115.6 .21 151.1.19 102.8 . 1 3 4 . 7 .0s 56.9 . ' t 5 63.6
Kwaial26 3.03 46.7 1.60 -2A.1 .95 -125.9 .66 151 2 .44 45.2 .31 127.2 .39 -30.3 . 1 3- 1 6 5 . 3 't0 162.3 .o4 179.4 . 1 1 161.9
Idbin6a .91 125.1 .12 175.5 .26 116.5 .27 652 .19 95.2 .10 108.8 .03 160.7 .08 50.5 o
(o
01 18.0 .00 26.6 . 0 3 126.8
Kashima .87 5 1 . 5 . 4 6 7 5 . 4 1 . 1 4- 1 3 8 . 3 8 9 1576 .13 64.7 .35 138.5 .13 74.7 . 1 8 1 6 3 . 0 .o2 12.2 .o5 42.9 . 1 0 105_6 o
Johani26 1.63 127.4 .74 -102.4 .09 128.3 14 116.8 .33 135.6 .03 127.6 .20 101.8 . 0 3 1 0 1 . 6 't12.6 . 0 7 o
.o2 171_6 .03 146.0
Wettzell .52 -64.4 .14 -36.6 .18 57.9 .08 97.7 .11 88.3 .05 -54.1 .o4 32.7 .01 40.5 .05 6.1 . 0 3 119 .05 75.4
='
Onsala60 .39 62.3 .1O -25 .2 .22 -5O.4 .11 103.3 .0a a4.2 .o7 47.7 02 - 1 2 . 6 .00-130.0 J
.08 13.2 .05 23.1 .06 51.5
OnsalaSs .39 62.3 .1O -29.5 .22 ,50.2 . 1 1 1 0 3 . 2 .08 84.3 07 -47.6 .o2 ,12.7 . 0 0 1 2 9 . 6 .08 13.2 .05 23.0 .06 51.5 o
Werlhovn .66 59.7 .19 -28.6 .20 ,58.0 .09 925 . 1 4 81.1 .06 55.4 .04 -25.5 .01 35 4 .06 3.3 .04 6.5 .03 52.O
Etlsberg .64 607 .18 -31.5 .21 -58.2 0 9 9 3 0 14 42.2 06 55.8 .04 -30 2 . 0 1 3 6 I .06 2.7 0 4 59 03 50.6
Ofioral 91 1256 .13 174.3 28 116.5 .28 66 1 20 97.0 .11 109.1 .03 1600 . 0 8 5 1 . 6 .01 - 16.3 .00 24.0 .03 -r20.6
Yaragade . 3 7 1 6 8 . 3 1 3 1 0 54 .81 1 35 64 6.5 o7 947 . 2 6 1 34 .02 95.0 1 3 2.6 .01 129.5 .o4 131.7 08 63.8
Graz .53 65.1 .14 -37.6 .16 60.3 07 99 0 11 8 9 . 1 05 55.6 .44 32.9 0 1 4 2 7 . o 4 6 1 .03 13.5 .05 a2.o

,{

]4 ct
6'o
Kt Ot Pt t'lzKz Qt MJ i'ii S" RI
M, Sz
PHAS AMP PHA' AMP PHAS AMP PFIAS AMP PHAS f(,l
TMF_FFI';ilF FHFAMP PHAS AMP PI1AS AMP PI]AS AMP PHAS AMP

.03 34.3 . 0 1 5 1 . 7 .03 7.A .03 20.2 .06 91.0 9;

Plana .42 -63.9 .11 -39.9 .10 71.8 _06 104.4 .09 -90.3 .03 -64.2
.03 -32.8 .01 46.0 .05 6.8 .03 13.2 .05 76.7
.14 -37.5 .17 -58.1 .08 98.3 . 1 1 8 9 . 0 .05 -54.1
.o4 -27.A .01 55.5 .05 8.0 . 0 3 13.6 .o5 75.7 7=.
Ondreiov .50 65.1 .14 ,53.6 .17 57.8 .08 98.9 . 1 1 4 7 . 2
93.4 . o 2 - 1 7 4 . 4 .m -44.1 .oo 124.4 .00 -42.3 .o2 30.7 .07 99.6 OE
Hefwan .37 67.0 .11 52.3 .08 -170.9 . 0 5 - 1 5 0 . 5 .09
.o4 -32.7 _0't 39.9 . 0 5 6 . 1 .03 12.O .o5 75.4 oo
wenzell .53 63.8 .14 36.6 .18 -57.8 .08 97.5 . 1 1 8 8 . 2 .05 -54.0 oo.
-59.8 48.4 .o2 1.4 . 0 1 1 1 0 . 3 . 0 7 1 9 . 8 . 0 5 1 6 . 8 .07 74.4
Metsahov .28 .07 -23.3 .17
52.2 . 1 1 1 0 2 . 9 .06 -89.2 .05
58.3 .05 44.8 .01 5.5 .04 -4.5 .02 2.O .03 83.2 t-
Grasse .73 71.4 .21 ,45.7 .14
62.0 .06 97.9 . 1 5 9 3 . 3 05 o
. 1 0 8 2 . 2 .10 122.5 . 0 1 1 3 2 . 1 .o7 174.6 .03 99.4 . 0 8 - 1 5 8 . 2 . 0 4 - 1 5 1 . 5 .05 141.4
Huahine .46 107 5 .23 167.4 142.9
.O4
-63.3 98.4 . 1 0 - 4 4 . 2 .06 52.3 .o3 24.4 .01 62.3 .06 9.0 . M 14.9 .05 67.4
Potsdam .51 13 -31.3 .19
55.5 .09 5'
-65.3 76.9 .03 39.7 . 0 1 - 4 6 . 6 .02 3.6 .o2 22.4 .06 95.1
Oionysos .42 .11 -44.0 .07
8s.5 . 0 5 1 0 91 . 0 9 - 9 1 . 3 .o2
.07 98.5 . 1 1 9 0 . 2 .04 -55.2 03 -34.8 01 42.3 .04 6.6 .03 14.7 .05 83.5
Malera .49 65.9 .13 39.4 .15 -60.4 f
.76 -154.4 . 1 2 8 1 . 8 . 3 0 1 3 5 . 4 .11 88.3 . 1 5 1 5 9 . 9 0 1 1 0 . 1 .04 46.0 .09 110.9
Dodair .72 65.1 .37 U.4 .97 -135.2
.86 145.5 .25 95.7 .34 126-6 .'t7 108.9 . 1 7- 1 5 3 . 0 .o1 32.2 .o5 43.7 . 1 1 1 1 30
Simosato 1.25 87.6 .57 107.0 1.11 126.5 q)
5 . 2 .22 -41.6 . 2 5 1 5 . 5 .16 85.0 . 1 2 - 2 . 6 . 0 5 - 1 5 8 . 7 . 0 3 1 7 1 . 5 .06 48.5
Mazatlan .74 -89.2 .53 -88.2 .81 20.0 .56
62.3 .04 1.6 .o2 38.0 .0s 1.3 . 0 5 1 5 . 0 .o4 12.O
Kootwijk .60 -36.6 .19 -7.1 .20 63.1 . 1 0 6 9 . 5 . 1 5 - 6 3 . 1 .06
. 5 1- 1 1 3 . 2 . 1 7 1 6 3 . 6 1 9 - 9 2 . 8 .04 172.2 . 1 0 1 1 5 . 7 .01 105.3 .03 95.7 .09 146.8 c
Shanghai .58 169.3 .14 178.2 .62 -97.4
.26 -160.2 . 1 3 9 6 . 5 . 1 5- 1 4 1 . 7 .04 99.4 .o4 177.2 .04 176.8 . 0 1 1 2 6 . 8 .03 105.0 o
A@qtlipa .37 124.2 .10 95.2 .50 - 140.1
85.2 . 0 5 - 5 2 . 0 . 0 3 - 1 9 7 . 0 1 7 3 . 1 .05 11.2 . M 15.5 .06 74.9 !)
Borowiec .r|4 61.4 .12 2a.A .17 56.1 .09 99.9 .09 f
. 0 7 1 4 1 . 5 .4a 124.4 .07 -90.9 .21 86.0 .04 53.0 .o3 -47-7 . 0 3 - 1 2 4 . 5 03 108.1
San Fem 2-29 105.7 .77 U.4 .24 -49.5
.07 99.9 . 1 5 9 0 . 7 . 0 6 - 5 6 . 3 .05 -40.0 . 0 1 1 . 9 .05 -.2 .02 2.5 .03 71.9
.21 41.4 .20 -59.2 o
-4.4 .03 132.9 .01 141.7 .09 1.0 .05 18.7 .03 -25.3 (o
Herstmon .rK -88.0 .11 -74.6 .30 60.2 .10 111.6 .06 .08 61.8
-70.3 .04 108.3 .o7 10.0 .02- 167.9 . 0 1 157.8 .01 135.4 o
Fl. Davi .08 164.8 .15,127 .2 .4A 32.5 .32 17.2 .05 .15 30.8
-1O.7 24.7 .12 ,55.2 .31 42.9 .05 82.5 12 19.9 .01 -67.1 . 0 1 107.7 .06 73.9
Mojave .22 . 1 5- 1 0 7 . 6 .99 4:].6 .62
.23 1.9 .17 143.6 .11 .1 .o7-174.4 .05 2.O .o2 19.5 .01 68.9 .05 115.5
Gr€enbel .94 159.1 .25 169.9 34 2.3
.73 49.2 .2A 134.7 .37 57.5 1 5- 1 3 5 1 .11 47.2 06 143.7 . 0 3 169.1 .05 109.5 D
M a u i ,H l 1.18 117.1 .4a 134.4 1.28 57.7
-45.7 24.5 . 1 8 -66 2 .35 38.7 .10 73.8 .13 15.8 .01 106.4 . 0 1 156.7 .06 71.4 o
Monument .51 .27 835 1 12 39.8 .71
.31 252 .o2 11.7 .15 38.7 0 3 - 1 5 10 .06 17I .01 6.3 .01 49.2 .o2 92.7
Plattevi .23 111.4 .13 1681 .47 39.5
2.6 .34 51.0 .01 1408 .13 28.6 .03 -20.1 .02 49.3 .o7 74.2
Quincy .62 56.3 . 1 3 1 s 0 . 81 . 1 1 5 1 . 5 .69 37.4 . 1 2
-55.1 . o 7 - 1 1 2 . 4 .08 87.0 .o2 -77.O .o2 19.4 .01 85.5 .03 16.1 .03 22.2 .07 91.0
Simeiz .34 .09 27.6 .07 -86.3
-53.1 87.6 .04 -57.9 .o2 2.1 .01 94.4 .05 21.4 .05 19.7 .08 85.5
Zvenigor .28 .o7 16.7 .12 -64.1 . 0 9 - 1 0 5 . 2 .06
 248 EXPLANATOBYSUPPLEIVENT COORDINATESAND THE RoTATIoN oF THE EARTH 249
I / TERRESIRIAL

The hciglrt tlisplacement.in ( r'ntinteters.carr lrc forrlcl for a giv.rr corrstituelt i a: giving
follows: P=Ao+(A:+As)Rr/4, (1.3524)
&(i)= n|npl;.1 /) - phase(i)),
x cos(arg(1, (4.35t _2)
p2 = (.r2+ y2).
,'1"ru
to pt:rforrtt a quadratic {il to the avail
where arg(i,l) is lhe proper angrrlar. argument to be uscd with tho Schu,iderski It remains thc task of the rlata alalyst
to detcrrnirrc the coellicients Ao s. I'uture aclvattccs in under-
phases (a Fortran srrbroutirrc frrl this exists il the IERS Starulauls (\lcCartlrr.. able weather data
elarsticinterarctiolrcan probabll lrt' a<tomntodateri
1989. pp. .10 -12)). Thc total ti(lal displacerncrrt is lorrncl bl.suurrrrirrg ovcr a]l nine standing the atmosphere-crltst
irr Equation '1 352 l Furthermore' expartsiouof Equa
constitucnls. A rrcgative lr(i) lrr('arlsthat the surfacc:rt thc predicterl lxrirrt lras beerr bv adjusting the coefficicttts
Iolvered. t,ion 4.352 I rnight be reqrtilctl for sfaliionsclosc to thc coast'

4.352 Atmospheric Loading l hc procedure clost libecl bclol' is takcl fr.orrr tho pub 4.36 PlateMotions
lication of Sovers and l'anscloq' (1987). A tiurr.r':rrlitrg atmosph<:li< irrossulc di-s
of t he tectonic
tributiorr catr induce crustal rk'forrnation. Rablxl arrrl Schuh (1986) ('sliniarc rhe one of the factor.sthat can afli:tt Earth rotatiott t crsltltsis the nrol it)Ir
plates that make up the Earl['s surface As thc plates mo\-e fix{\l coorclinates
efiects of atrrrosphcric loacling orr VLBI basclirrc dct crrninat ions. iin(l (!)l(.lll(lc tlLat
fite rates of
they mav iuu()rult to manv nrillirneters of scasorral varitrtiou. Ilr contrast to ()ceall- for the observing stations rvill beconreinconsistcut rvith each ollrcr.
l)c 5 ccntinrcturs
tidal effects, turalysis of the sil llatior in the atniospiiolic casc does rrot lrctrefil frorn relative motions lor somr:rcgular obscn'ing sitt.s ax' believed to
per year or larger. The obsr.rvat,ionsof plate Droliorrs so far by satr:llil <rl{scr ranging
the prescrr<rrof a well-urrdt:rsl ood periodic drivirrg folce. Otherq'isc. cstirnirtion ol'
appeiu to lre roughly <rotrsisl|rlt q'ith tht:
atmospheli( loading via (,lr<r'rr'sfrrrrclion techni<1trcsis aua.logous to rlctllr<ls usecl and very Long Baselinc Irrtcr.fcrorrrctr-v-
average rates over the last ft:tr, million 1-earsrlt'r.ivtrl fron the gcological Iccord
to <alculatc occan-loading t:f[c<ts. Iiabbel and S<lrrrh Lecorrrurenda sinrplifictl for.rl
of the dcperrdcnce of the lt'r'lical crustal clisplar.crrr<rltou pressure tlistr.ibrrtion. lt and other geophl,sicalinforrrration.Thus. in or(lcr l() reducc iucoDsist('I(irrsin thc
invoh es onlt lllc instantluc()lls In-r.ssureat thc silc in cluestion. arrl ari il eLage station coordinatesand to urtrkethe results frorrrdifl:rent tccilDi(lucslllol-.'dir.e(th
prcssure o\tr a circular rcgiou C t,ith a 2000-klr r.adius surrounclilg tlrc sitc. Tlrc comparable,a nodel for plitlc ntotions basul olt 1he lelative plai{'rr)otioll nrodel
exprcssiorr for' 1,heverticir,l disl>l:rr:r:rncnt(nrn) is RM-2 of N,Iirrstcrancl Jorclaut(1978) is reconrrrx'Ir(lf(I.
FloIIltheRN'{-2model,NIittslerandJorcl:rrrtlt't'ivclfirr.rr.diflcr-t'rrlirllsollrt
arr: ANI0-2. \1,hic}rhas
motion mociels.The t1\,oth:rt hiu-t:been discrrsstrl [rost rvidel-v-
Ar = 0.35p- 0.55tj. ( . 1 . 3 5 2l ) zero net rotation of the E:rltlt s sttrface.ancl A\11-2. $ ltich mirtirlizcs l ltt' mol ion of
a set of hot spots. Kaula (197ir) and others ha\'('(liscrrssedalternali\1'geophl-sital
rvherc p is I lte local pressrrlr':rrrorrralr'.andf the l)r(,isure alonrall sitlrir 1ll. !01:J(l- constraintsthat can be rrsedin ordcr to investigalc the plate nloti(tls Nillr resPecl
k m c i r c u l a r r c g i o n t n e n t i o l r . < la l r o v r : ( b o t h q u a l t l i l i c s a r . c i r r r n b a r ) . N o l ( ' t l t a t t l l c to the brrlk of the mantle. Firr cortveuience.u'lx'1l fllttlrc chang{rsilrc ll)ad(rin lhe
refereltce poirl, for this disyrlacctrrcrrlis the site lot.aliorr at stitrrclturl pttssru.c (101ii relative-rnoliorr model, ancl 1o zrvoi<ldependr:n(uoll thc choice of tlro llot spots to
rnbar). be held fixed, thc (ANI0-2) lrx)(lcl is recomlnerxlt'tl Tlte cartcsiarrrolilti(nl vector
Arr a<Llil,ionalurechanisrrrlor clrtrracterizirrgf rrrav be appliccl. Th<' I rvo-rlirncrlsi{)lll for each of the major platcs is giwrn in Tabb .1.:l(i.1.
stttfart prr:ssrrrcdistributiorr sut.rounding a site is <k.scribcd br- Future iurprovemcuts ar(' I)lanned lbl fittilrt' lEllS StanclartlsirLthrcling:(1)
replacementof R\I-2 b1 a lrc\\' rnodel.ncrv vekrcitics.ancl ne\r piat.s lirt cxample
p(t, -.r')= A0 + A I,r'+ A:-v l- A:-rl + A1.r.r'+ A5-r'1, (1.352-2) with thc NUVEL-1 model of Dc\lets el al. (199{)):and (2) adoptiorrol a re|tical-
motiort rnoclel (post-glacial r-cbound).In any casc, thc A\I0-2 trtork:l sltould bc
where r all(l ) arc the local t'asl arrd north dislarrcs of tlre poirrt irr <luosl,iorrfionr used as a default, for statiorts tltat appear to lollorv reasonably il's values For
the VLBI sil,c. T'hc pressrrre .llolraly some statiols, particularly irr the vicinity of plltc boundaricsj us(rrslnay benefit
t rnav be e\,;lluatcd bl the sirrrplc irt(.gra,tiorl
by estimating velocitiesor rrsirtgspecilic valrroslrot dt:rived frotrr ANl0-2 This is
also a way to take into accourrtr()\\'some nonncgligiblevertical rnotiotis Publishc'cl
n=| f, a'a,oa.e d,a,. (,1.35
2-3) station coordinatesshould irx lttck:the cpoch associatcdrvith thr: coortlirtates'
f ||
 250 EXPLANATORY
SUPPLE[4EN-r COORDINATESANO THE BOTATIONOF THE EARTH 251
4 / TERRESTBIAL
Table4.36,1 for the fortnightly and monthly terms The influence of the ocean
CartesianRotationVectorfor EachPlateUsing ef J 0.08 msec
KinematicPlateModelAM0-2(NoNetRotation) periods less than or eqttal to 24h is at the level of t 0 1 msec'
iides with

!7v Qy Q7
Plale Name deg/My degr'My deg/My OFTHEEARTH
OFTHEROTATION
4.4 THEMONITORING
Pacific -o.12276 0.31163 -0.65537
-0.63726 - l.33142 0.72556 rotation began after Kiistner's discovery of
Cocos
-0.09086 -0.53281 continuous monitoring of the Earth's
Nazca 0.63061 Chandler's explanation of it as polar mo-
Caribbean 0.02787 -0.05661 0.10780 the variation of latitude in 1884' and
wobble). Optical a-strometric methods then in usc have continrred
South America 0.05604 -0.10672 0.08642 fion (Chandler
-0.09492 continued to be tlte only source of Earth-rotation infor-
Antarctica 0.05286 0.21570 to this day, and in fact
lndia/Australia 0.48372 0.25011 0.43132 techniques then grzrdually took over
0.05660 nration up until the mid 1960s. More rnodern
Africa 0.19249 0.24016 of the Earth. $'ith successivell'
Arabia 0.27845 0.1674d 0.37359 the bulk of the activity of monitoring the rotation
Eurasia -0.03071 -0.15865 0.19605 Doppler tracking of satellites. Connected Element Radio Interferometrl' (CERI),
North America 0.03299 -0.22428 -O.O1427 Satellite Laser-Ranging (SLR)' and Ver.v Long Base-
Lunar Laser-Ranging (LLR),
line Interferometry (vLBI) assisting in such monitoring. At prcsetrt, the techniques
of SLR and VLBI provide most of the current high-accuracv observaiions for this
purpose, with LLR providing such data sporadically'
4.37 TidalEffectson UT1 Before we describe the observational systerns and services' the basic sensitivity
of the systems should be briefly explained. All the current observ'rtioDal systen)s
To properly cornpare UTI results obtairred at diflerent epochs, the high-frequencr-
depend on the existence of observing astrolromical targets u'ltose position can be
tidal effects must be accounted for. Thus, a starrdard expressiorr frr tlrr: tidal rar.j
inferred in some celestial reference fralne. This includes stars and solar-s1'stemob-
ations in UT1 is required. The periodic !ariations in UT1 duc to ti<lal deformation jects for optical astrolnetry,the Nloon and artiflcial satellitesfor ltrserranging, an']
of the polar moment of inertia have beeu reclelived(Yoder el ol.. 1981) includilg
(usually distant) radio sources for radio iuterferometry. Thc obsorvatiorts of these
the tidal deforrnation of the Earth with a decoupled core.
systems are then used to infcr the utotiorr of the observing instrtrments or, Inore
This model leads to effective Lovc rlurrrbers that differ from thc bulk value of specifically, the Earth to which they.rre altached. After thc predicted motion frorn
0.301 becauseof the oceans and the fluid <ort:, which give risc to dillorcnt theoretical
all other "defined" sources(as explairloclpreviously.includirrgprt'cession,nutati.n'
values of the ratio t/C for the fortnightly iur<l rnonthly terrns. Ho\,(rver, Yoder el al..
the Earth's siderealrotation, tides. tertoltic motions, etc,) has be'tl retrroved'thc
recomtnendthe value of 0.94 for t/C for both cases.Periodic terns irr fiTl are eiven
motion that rernainsis attributed to polar motion and UTl-UTC variations Horv-
in Table 4.37.1and are discussedit Erplanatory Supplementto thc IE!?S Bttlletins ever,as we shall scc, some proposcclfitture s1'stenrs of Earth rotatiorr lni)nitoring Jo
A and B, March 1989 as rvell a^stlrc Annual Reports of the IER.S.'Ihble 1.37.1 depend instead ou the sensingof a local "inertial" coordinatc s\-s1enr- Iather tharr
includes terms due to zonal tides. .rl,ithpcriods up to 35 days. willr k/C = 0.9-1. an astronomicallv defined one.
Table 4.37.2 lisls the remainder of the tidal terrns (u'ith pcriods grr:ater than 35 It is also inportant to note that a. siuglc instrumcnt (or sillglo baselinefor in-
davs). terferometricobserving),being at a givcn point on the Earth's sttrfl' e. ' an be used
UT1R, 4R. and trR representthe rcgrrlarizedforms of UT1 (in<:luclingperiods to determine only two comporrents of thc Earth's rolation. 'fhc colnponents com-
up to 35 days), of the duration of the clal', 21, and ol the angular vr:locit1. of thc monly determined using singlc-statiou (or single-baseline) data inr:ludc "variatiort
Earth, ,"'. Thc urrits are 10-4 s for UT, 10 5 s for 21, and 10-la ra<l/s for r.,. in latitude" (AO) due to any rorth south motion of thc station, and UT0-UTC
It should bc rnentionedthat thc occzrnictides rvith short perio<lscausc \-aria- (or just UT0) when there is any east \\'cst motion of the statiorl, attributable to
tions in UT1 that are only partially rcprescntedby the rnodcl givcn belorv.Accord- Earth rotation variations. To obtairr all three components of Ezrrth rotation' it is
ing to Broscheel al, (1989) the contribution of the oceanictides should bc split iuto necessary and sufficient to use data from two stations (or baselint's) rvith different
a part that is in phaseu,'iththe solid Earth tides and into an out-of-phascpart. Thc longitudes. If multiple stations (or baselincs) with difierent longitudcs are involved'
discrepancy betwr:en the model of Yoder. el ol., (rvhich already contairs a general the problern is overdetcrmined arrd gcnerally a least-sqrrares solution is done to
influence of the oceans) and the model of Broschc ei ol., anounts to a maxirrum determine the Earth-rotation pararncters (ERP).
 OFTHE EARTH
ANDTHE ROTATION
COORDINATES
TERRESTRIAL zcrJ
Table4.37.1
ZonalTideTsrmswithPeriodsUpto 35 Days 4.97.2
TideTermswithPeriodsGreaterthan35 Days
A19ulranf P . t f o d U T i- U T 1 F A-AR o-qB
Co€tficientof Cooffcientof Atgunenr Pcrlod UT1-UT1F A-AB a-aR
Dayg Eln(aryument) cls (argumonD Coeftcionlol Co€tlicientol
F D f, Days sln (argumont) cos (argument)
1 0 2 5.9 -o.o2 0.3 -0.2
0 20 r 6.95 -0.04 0.4 -0.3 IO 2 91.31 -0.06 0.0 0.0
-0.10 -0.8 22-2
0 20 0,9 I 119.61 0,03 0.o 0.0
o 1 7.09 -0.05 0,4 -o.4 l. 0 't2-2 _ 1.88 r.0 -0.8
0 2 121.75
0 0 22 2 z.to -o.12 1.1 -O.9 173.31 0.25 -0.1 o.1
0 o2-2 0
'| 1Tl.U 1.17 -0.4 0.3
I 0 20 0 g.fi -0.04 03 -o.2 0 o2-2
I o 20 1 Lt2 -0.41 2,8 -2.4 2 18P.62 -4€.25 16.6 -14.0
-0.99 -5.8 0 02-2
I 0 20 2 9.13 6.8 200 0 182.63 -0.19 0.1 -0.1
-0.02 -0.1 0
0 00 0 9.18 0.1 00-2 -l 199.84 0.05 0.0 0.0
'| -0.08 -0.5 2
0 9.54 0.5 00-2 0 205.89 0.2 -0.1
2
-0.20 2 00-2 1 212.32 0.04 0.0 0.0
0 22 1.3 1.1
o o2 0 9.61 -0.08 0.5 -0.4 I 346.60 -0.05 0.0 0.0
-0.1 0-1
o 2 12.41 o.u2 0.1 00 -t 346.64 0.09 o.0 0.0
'| -0.1 01
0 20 2 13.17 0.cKl 0.1 2 &5.2 0.€|l, -0.1 0.1
-o.30 -1.2 0-1
0 o 20 0 t3.61 1.1 01 00 0 365.26 -15.36 2.6 -2-2
't -12.5 01 00 1 386.m -o.14 0.0 0.0
0 0 0 13.63 -3.21 14.8
0 0 0 2 13.66 -7.76 35.7 -30.1 411.78 0.03 0.0 0.0
I 00-l 0
0 0 0 1 13.75 0.02 -0.1 0.1 1095.17 -0.14 0.0 0.0
-0.34 -1.3 o-20 0
0 0 0 0 13.78 1.5 020 1 1305.47 0.42 o.0 0.0
0 0 0 1 13.81 0.02 -0.1 0.1 3232.85 o.(x o.0 0.0
l0'l 0
o 000 2 3!)S9.r8 7.90 0.1 -o.1
0-l o 2 14.19 -0.@ 0.1 0.1
0 000 1 6790.36 -1617.27 -14.9 12.6
00 0 2 -1 14.73 o.05 -o.2 0.2
oo 2 0 14.V -o.73 3.t 2.6
00 0 2 1 t4.80 -0.o5 0.2 -o.2 ' t= 134.96" + 13.0649930(MJD-515,14.5)
Moananomalyol the Moon
0 -l 0 2 0 15.39 -0.o5 0.2 -0.2 / = 357.53"+ 0.9S56000(MJD-51$4.5)M€ananomalyof lhe Sun
F= 93.270+ 13.229350o(MJD-519.5)L - n: L: M6anlongitudeol lhe Moon
10 -2 1 23.86 0.05 -0.1 0.1 O= 297.85o+ 12.1907496 (MJD-515,14.5) ol th€Moontromlhe Sun
M€anelongation
10 -2 2 23.94 0.10 -0.3 0.2 A= 125.O4o- o.052954o (MJD-515!44.5) of theascending
M€anlongltudo nodoof the Moon
11 0 0 0 2s.62 o.u -0.t 0.1
-t0 0 0 26.88 0.05 -0.t 0,1
-1 0 2 0 1 26.98 0.18 -0.4 0.3
-1 0 o 2 27.@ o.44 -1.0 0.9
10 0 0 -1 27.44 0.53 -1.2 1.0 The actual equations for expressing A{ and UTO-UTC in terms of xp, Yp, and
10 0 0 0 27.56 -8.26 18.8 ,15.9
10 o -1.2 UT1-UTC are
0 1 n.67 0.54 1.0
00 o 1 0 29.53 -0.1 0.1 Ad=rpcosl-'psin), (4.4-1)
1-1 0 0 0 29.80 -0.06 0.1 -0.1
-1 0 0 2 -1 31.66 o.12 -0.2 0.2
-1 0 0 2 0 31.81 -1.82 3.6 -3-0 UTO - UTC = tand(rp sin^ + ypcosl) + (UTl - UTc)' (4.4-2)
0 2 1 31.S6 0.13 -0.3 0.2
10 2 -t 32.61 0.02 0.0 0.0
and .\ are the latitude and longitude of the station (or baseline), with
0 2 0 34.85 -0.09 o.2 -0.1 "Phere {
longitude again measured toward the east. In cases where data ftom more than
' etations (baselines) exist, a least-squares solution is generally performed with
!= 134.960+ 13.0649910 (ruD-51544.5) Meananomaty ot the Moon
| = 35r.53o + 0.9456000(ruD-51544.5) Meananomalyol lh€ Sun abwe equations (for most multistation optical a.strometry artd LLR quick-look
F= 93.27o + 13.22$35 o(ruD-515445} L - O: L: Meanlongltudool th6 Moon for the three
D= 297.85o+ 12. | 90719o(iruD-5t544.5)Moanelong€tionof tho Moonlrom th6 Sun ), or else a general least-squa,res solution is performed, solviug
A -- 125.0,1o- 0. O529elo(MrD-51544.5)M€anlongitudeol th€ aecendlnonodeot th6 Moon nts of Earth rotation (arnong many other parameters) directly ftom the
I
ons. I
I
I
 v
254 EXpLANAToRv
suppLEMENT COORDINATES OFTHEEARTH
ANDTHEBOTATIoN 255
r' / TEFIFIESTRIAL
Follou'ing a,re discussions of the irrdividrral methods for Earth-rotation detel- gravity field Thc coordinates of the stations courputed by this
nents of the Eari,lt's
thought to be coirrcident v!'ith the center of urass
mination. First. the current high accuracy methods are describcd including laser technique have an origin u,hich is
ranging (both lunar and satellite) and VLBL Second, historical methods ar.o dis- to within t l0 cm. Earth-rotation results are computed at 3 day intcr-
of the Earth
cusscd. Third. some proposed luture or alterrtate techniques are described. For all old' u'ith polar motion arrd UTI-UTC accurate
\,als using data less than a week
tenths of a millisecond respectivcll'.
these methods, only basic descriptions of the observing systefirs are presented. along to a few milliseconds of arc (mas) and a ferl"
with summaries of the actual opcrational svstems. Firrally, the currenlt internationill (Further information on tlte computational techniques ilvolved is given in Kauit1,
service for Earth rotation is described. (A similar rcview also exists in \'loritz :rnd 1 9 6 6 p, p . 7 8 8 1 . a n d G S F C . 1 9 7 6 p
. p 7 - l l l o 7 1 5 )
Muellcr^ 1987,as u"ell as in IERS, 1988.)
4.412 Lrnar Laser-Ranging Lunar Laser-Ranging (LLR) is idcntical in conccpt
with SLR, except that thc targets are arrays of cube corner reflectors left on the
4.41 LaserBanging Moon by the U.S. Apollo astronauts and USSR unmanncd
probes. Hou'ever,the
distance to the N4oon makes it more clifficult to point a laser reliably at the re
Laser ranging of celestial targets, lirst proposed (for the X,loon) in the late 1950sand
flectors, and when there are reiurn signals, they are orders of magrritudes weaker
early 1960s(Alley, 1983), is conceptuallya verv sirnpleobservationalprocedrue.A
than those from the LAGEOS satellite Although LLR has bccrr <lone since Au-
Iaser beam is directed to a celestial target carrying retrorcflectors, such as the NIoon
gust of 1969 (initiallv using the a.ray left by tire Apollo 11 astronauts (Arrnstrong,
or an artificial satellite. Half the round-trip travel time of the light multiplied by thc
1977)), these technical difficulties have resulted in few stations reaching continued
velocity of light gives the distance to the object (subject to r,.ariorrscorrections clrrc,
operational status. Curreltly data is obtained regularly by the Centrr: d'Etrrdes et
to relative motion, refraction, etc.). These rangcs can then be compared to their.
des Recherches Geodynamics et Astronomiques (CERGA) station near'Ioulousc'
predicted ralues, and in an adjustment (usuall)' q,.cightedleast squares), corrections
France: the Universitv of Hawaii LURE Observatory ou Maui' Hawaiil and thc N{c-
to the parameters involved may be estimated. such as the or.bital parameters. the
Donald Observatory. Texas. Nluch of the samc information obtairrr:d with SLR can
station positions, and the orientation of the Earth.
be obtained from LLR. Hot'ever, in practice, with only a limited rlumber of stationsi
the primary use of the LLR systemshas been for the determination of IJTO-UTC
4.411 Satellite Laser-Ranging Currcntly there are several satellites arailable for. (and thereby UTI UTC), parameters of thc Nloon's orbit and the \'Ioon's rotation
laser ranging, but observation$of the Laser GeodynamicsSatellite (LAGEOS) cur (librations), special- and general-rclativity parameter tests, ancl inforrnation on tlte
rently provide the highest-accuracy vaiues for the above parameters. (However, tlrc tidal paranteters of the Earth N{oon svstem. LLR also shot's promise in provitling
USSR has recently launched trl,'osimilar satellites. Etalon 1 and 2, u'hich shorv grc:rt quick determinationsof UT0-UTC, sinceits computation is a simple matter as soon
prolnise as targets lor providing such information (Arrodina and Prilepin, 1989).) as the day's observations are completed at a given station. Culreutly UT0-IJ IC is
LAGEOS itself is a composite (aluminum shell and brass core) sphere, 60cn in determined irregularly (at each of the three operating stations) with an irccuracy
drameter.rveighing407kg. nith 4263.8cm cubc corner r.etroreflectors embeddcdirr of t 0. 3 to t 0.7 ms (\IcCarthy and Luzurn, 1990).Lardcn (1982.p. 250) indicates
its surlbce. It is irr a nearly circular orbit at a height of about 5900km. and arL that with a four or five station net*'ork, and ranging regularly with t 3 cnt ran11e
inclination of 109.8'. Its rclatively largc rnass,smali surface area. and high orbit accuracy, that t 5 cm (t l. 6 mas for polar motion, t 0. I ms for lITl UTC) trvo tlav
result in lot' atmospheric drag and inserrsitivity to short,war€length componerrts of ERP determination should bc possible. Further information or the 'nmputational
the Earth's gravity fielcl. NIore inforuration on LAGEOS aucl otlicr such Aeocletic techniquesilvolved is given by Arnold (1974,pp. 249 255), Mulhollarrd (1977) and
satellitesis availablein Cohen ef al. (1985). Larden (1982,pp. 69 90).
There are approximately 25 SLR stations irr rcgular operation. located on ciglrt
tectonic platcs. regularly obtainilg observational datil- The precisiou of rangcs ob- 4,42 VeryLongBaseline
Interferometry
tained from these instruments are on the order of I I to t l5 cm. (For more infor-
mation on the instrumeuts thernselves,sec Shaweald Adclman, 1g85 and Dcgncn, In geodeticVery Long BaselineIntcrferometry (VLBI), the primary targets being
1985.) These stations are operated by various governmerrtal and university research observed are a sct of extragalactic radio sources. These objects are presumably so
organizations, in ordcr to obtain varioqs geophysir:aldata relating not onlv to Earth distant that none of their motions are detectable, and they therefore form a fixed
rotation, but station coordinates and baselirre lengths, changes in those values ovr:r celestial reference frame. Currently, sorne 228 sources have positions established
time (thereby giving tectonic,plate nlotion velocities), and long-wavelength compo- (published) within such a systerrr, but only about 74 objects have been observed
 256 EXPLANATORY
SUPPLEN4ENT ANDTHERoTATIoNoF THE EARTH
COORDINATES zct
r' / TERFESTRIAL
from rnore tharr one netrvork, and only about 23 repcatecliy frorn more thau two 2nd Connected-Element Radio Interferometry (CEllI) having beerr
etellite tracking
networks (Arias, el nl., 19E8). purposesin the 1960sand 1970s
used for operational
The VLBI observing networks consist of two to generally four or fir'e radio tele-
scopes that simultaneously track extra"galactic radio sources. Each station rccords An explanatiorr ofthe motion ofthe EarLh's
X and S band radio signals ir digital form on nragnetic tape, alottg with precist' 4.431 Optical Astrometric Techniques
the early 1890s (Chandler. 1891, 1892). rrsing
time sigrrals provided bv hydrogen-maser frequeucy starrdards. Typically, 10 to 20 oole was first made by Chandler in
trunsit .i.cle observations by Kiistner dating frorn the late 1880s. Charrdler. ancl
sources are tracked for periods of 3 to 6 rnilutcs several times or,'erthe coursc ol a astromctric observatiorrs from transit circlcs. rgad
2,l-hour observing session.Thc tapcs are sent to a correlator center. il'herc lhey nre later others, also used variolrs
back to the eallv 170ils.Thc
replayed. Thc diffcrcnces in arlival tines of the radio sourcc sigrrals ("delays"), arrcl rants, and other visual optical iDstruments dating
1890s saw the establishment of the Internatiorral Latitude Servicc (ILS), which op
their differences in time ("delay rates'') are deterrrriuccl betr.een each pair of sttr- latitude until the 1970s. The use of
erated several stations at approxirrrately 39'
tions. Similar to the laser rangilrg solutions, these delavs and delay rates arc thclr photographic zenith tubes (in 1913 bv
visual zenith tubes and the invention of
rrsed in a least-sqtrares adjustrncnt irr orcler to estimate a r'ariety of geophysicaL
Ross) brought thc measurcntent of polar motion to accutacy lcvels pretty rtruch
parameters ancl the coordinates of the radio sources.
maintained into the 1960s. The use of precise pendulum clocks in lhe 1930sallorvcd
Currently some 20 radio telescopesin the u''orld have beerr outfittcd lbr making
the detection of seasoral l-ariations in the lcngth of da1" Spccial \'Ioon cameras were
high accuracy geodetic VLBI observations (using -\'lark III cquipment, see Clark e/
also used frorn 1952 until the early 1960s in order to measure changes in the length
ol., 1985). Although many of these are radio tclcscopcsshared betu.eengeorlctir:
of day against ephemeris time clcrived from thc observed position of the \Ioon. Thc
and astrophysical obserl'ations, seveLalare dcdicatcd plimalily to determinatiorr of
same equipment was also used to (ompare ephcrneris time with atomic tirrrc and
ERP. These inclrrclepriurarily trvo rrctu,orks:Projects Polaris/IRIS (Carter el ol..
to establishthe defirrition of the sccond (\'Iarkon'itz el ol . 1958;\'Iar-kowitz,1960).
1984;Carter and R.obertsotr.198,1).opcratcd b1'the U.S. National Geodetic Survo.
The International Polar Nlotion Service (IP\'IS), lbllou'ing on the ILS. tlordinated
consisti[g of foul or fivc stations operating in the United States and Europr:: arrcl
and reduced the polar motion obscrvations. Thc Bureau International tle l'Hcure
the Navy Netrvork, opcratcd by the U.S. Naval Obserr'atorv. corlsisting of forrr' 1o and also eventuallv be-
fBIH) did the same for UT0-UTC lelated obser-vations,
six stations operating in \orth America and Harvaii. The IR.IS statiorrs obscrr-r:
gan to handle polar rrlotion data irt orclerto provide rapid servi<eUTl-UTC. Tlic
fbr 2,1 hours. ever1. 7 days. using 14 difierent radio sources. Current ac:cru-acies
optical techniqucs continued as the onlv soulce of ERP rncasttrements rrrtil tlLe
of the data obtairred incluclepoial motion at t I 2 rnas, UT1 UTC at t 0.05
mid-1960s,whut Doppler satellitc tracking begari to providc information on polar
0.10ms, nutation and plecession (ceiestialpole) cdrectious at t I 2rnas, radio
motion. N{ore than 100 optical instruments all over the world lver{r at oue timc
source coordinatcs at less than I I mas. and \rLBI statilrr coordinatcs a1,it fe\\'
in use for such measLlrements.Currentll'. a srllall numbcr of such ilstrumelts are
centimeters. Daily observations of four sorrrcesfor t hour, frorn trvo statiols. also
still in operation, primarily-as a backup to and as a p()ssibleclx)ck on the spacc
provides daily detcrminations of UTO U'l'Cl. The Na.v1-Nctnork obscrvr:sfor' 21
geodetictcchniques,and to corrtjnlretheir usefirllong rccrordof colltinuotlsobserva
hotus every 7 days. using 26 different ra.dio sources. Obserrations everl-Icrr' da\.s tions. The primary disaclvarrtages (for inproved ERP determirration)'rf lh. uptir al
for three hours provide daily U'tIl UTC r-alucs (u.ith tr.o stations) lvith acdrra( ies astrometric lcchniques arc lhcir clepenclenceon the krcal r''ertical as their prirnary
similar to the IRIS resrrlts. Obscrvations are also rnade by stations opcratecl al referencc and thc accura(ry of the star ca.talogsthey rrse to define a cek:stial relcr-
occasionalintervals bv the U.S. NASA Crustal D.vnamicsProject (CDP) for thc ence system. Thc gcodctic space-basedtccbniquesthat havc mosth'replacsl theln
morritoring of plate tectonics and Earth rotation (Coatcs el ol.. 1985). are lit e aflected by variations in the lor:al vertical arid pror-icleinsteacl for a c rst
Furthcr information o1lthc computational techrriclucsirrvolveclis gir-en bl CiSFCl fixed three-dirnensional reference svstcm. Inforrnation olt thesc illsirum(rnts and
(1976,pp. 7-47 t,o7-42) Robcrtson (197;), Nl.r (1978),\Iclintock (1!180),aD(lBock the various international services involvcti is givr:u in revicrvs (\'Ioritz and NIucller.
(1e80). 1987,pp. 324 337, 124 430,456 465: N{ueller,1969,pp. 80 84; I-ambeck,1980'pp'
62-66; X{unk and NlacDonald, 1960,pp. 58 61; I}omforcl, 1971,pp. 312, 335 33E.)
4.43 HistoricalMethods

Various rnethods of lesser accuracy for EllP chl,errnination have exisl.crl ir the past. 4.432 Doppler Satellite Tracking Durirrg the 1960s, Doppier tracking of the U S'
rvith some still in opclational use today 1br special pruposes. Thcse techniqut:s in- Navy Navigational Satellite System (NNSS) began to provide polar-rrrotiol iufor-
clude primarily optical astrornctric methocls, u'ith spacc techniques such as Doppler mation on the same levcl of accuracy or better than the optical techniques The

[,
258 EXPLANATORY
SUPPLEMENT COORDINATES ANDTHEROTATIONOFTHEEARTH 259
! / IERFESTBIAL
ERP values obtained via this system rvere used by the BIH in combination with thc alternatirl- ol)servationalslstems. of coursc, it should also be realized that
"ible in thc tulrtlttly used observationalsystelns (prcviously mentionerl)
optically obtained values. However, this techniquc did suffer from the dependeucl imp.ou"m"ttt"
of the ERP values on the particular satellites in use and from long-term drifts, abo show great Prornise'
and as SLR became available for polar motion monitoring in the late 1970s, begal
to be phased out for active ERP use. An excellent review of the NNSS Doppl<rr
satellite-tracking technique is given by Kouba (1983), artd a review of ERP-relatcrl
use of such tracking is given by Moritz and Mueller (1987, pp. 338''3'14' 467 '170). (GPS) The Global Positiorting System was
4.441 The Global Positioning Syst€m
and global satellite navigation system. In
designed as an all-weather, continuous'
it u'as to bc used for navigation, but is also quite usable for
its basic configuration
4.433 Connected-Element Radio Interferometry One other method that enjolrrl gcodctic positioning of points on the ground. It rvas designed, is
time transfer and
some use in the late 1970s and 1980s was the use of Connected-Element Raclio of Defense.
under development bv, and will be operated bv the U.S Department
Interferometry (CERI). CERI operates in principle in the same fashion as VLBI The "space segurent" of the system till consist of 24 satellites in 12-horrr,
(and, in fact, provided the initial tests of the VLBI principle), except that the radio 20000-km-altitudc orbits, r,i'ith the satellites equally spaced in three orbital plancs,
telescopes are physically connected by a wideband commu[ications link, so thal inclined at 55" to the equator' Each satellite will broadcast ou tu'o L band frequen-
the observations (delays and delay rates see Sectiorr 1.42) arc made in real time. cies,Lt (1,575.4NIHz) aud L2 (1227.6 MHz), with the two ficqrtcncies allowing the
A least-squaressolution to obtain ERP, station positions, and source positiolls call elimination of first-orcler ionospheric refraction effccts. Modulatcd onto these fre-
then be made as soon as observing ends for the day, Although the speed at u'hich quencies will be thrce codcs, including (1) a "D" code, corrttrining inlbrmatiorr (at
results could be made available was a strong point of CERI, the small size of tht' 50 bits/second)on the prcdicted position of the satellite and thc satellite itself, (2)
arrays usually only allorved two ERP components to be determined simultaneouslr'. a coarse acquisition "C/A" code, a pseudorandom noise (PRN) code at a frequency
such as Ad and UT0-UTC. The U.S. Naval Observatory operated such an inter- ol L.O22MHz. which repezrtseverY one millisecond. and (3) a precise "P" code.
ferometer at Green Bank, West Virginia, from October 1978 until the end of 1987 another PRN code at a frequencl'of 10.23\'{Hz. rvhich repeatseverl'38 u'eeks.but
Polar motion and UTO UTC could be determined to an accuracy of about 50mas will be reset in practice evcrl'u'eek Each satellite also llas a vcrt'precise cesium or
and 4.6ms respectively (N{cCarthy et ol., 1985; I\'Iatsakisel ol.' 1986). Anotircr' rubidium clock to provide prccise time and stable freqrtcncv information- Currentlv
CERI was also operated for a short time (August-September,1980) at Cambridgo (1991 September),there arc six operating prototype (Block l) satellitesand eleven
U.K. (Klepczynski. 1982). The eventual improveutcnt il the turnaround time ancl operational (Block II) satellites in orbit that allow the svstem to be used inter-
much higher accuracy of VLBI results led to the evcntual demise of this techrriqut' rnittently. The lautrch of morc Block II GPS satellitcs is pltruled to '-ontinue using
for ERP determination. Delta-2 boosters. This proccss will contimre until thc svstcm is frrlll' operatiorral
(with 21 satellitesand 3 sparcs)in 1993.
The "user segrnent" u'ill generalll' consist of uscrs lvith navigation receivers
4.44 AlternativeTechniques These receivers will be ca,pableof "locking on" to fottr satellitcs at a tirne, gcner'-
ating C/A- (or for morc pro<:ision. P-) codes.shifting those codesin time to nratclL
Several obserr,ational rnethods are under developrnent t]tat shot' promise irl tlr(' those of the satellites. arrd decoding the D-codc inforltrzrtiolrfrom the satellites
field of Earth-rotation lneasuretnent. The current level of dcvelopment of thcs{' The position of the satcllitr:srvill then be knos'n ftonr thc D-code informatiol, arLd
techniques of course varies greatl)', and the future levtls of doployment aDd ac( lt- it is only necessaly to takt'the four obserr.ations (oue to ea<:ltsatdlite) and solve
racies achievableby these techniquesis still not clear. The primarl' conteuder Jot four equations for the forrr unknor.ns of the recciver's positiou (si,1 X' Y. Z) and
providing high-accuracy Earth-rotation information is the Global Positioning Sl s- time (this process bcirrg known as "pseudorangirrg"). A "stnrtdard Positioning Ser'-
tem, a global network of 24 satellites intended for navigtr,tion, which is currcntll' vice" will be availa,blo to all usels. This servicc allows tts<:of the C/A code (but
only partially deploycd. Thc use of laser ring gyrosrcpcs or superfluid helium inter- with a degraded signal for the Block II satellitcs in a process known as "Selectivc
ference gyroscopes rnay also allow such changcs in the Earth's rotational velocitt Availability" (SA)), for dctcrrrriningthe receiver'sposition itnywhere on Earth lo
to be monitored. These last techniques may be of particular interest, since thcl approxirnately 100 200 mctcrs. Using the P code, whit:h will be available only to
qualified users as thc "Precise Positioning Service," positiolral accuracies ol 15 30
have no dependencc on astro[omical observations and rnay also be used as a check
on cuuent techniques. What follows are more detailed accounts of these three pos- [reters are possible. Once "locked on" to the satellites (a process that takes at most
260 EXPLANATORY
SUPPLEIVENT COORDINATES ANDTHEROTATION oF THEEARTH 261
4 / TEFFESTFIAL
a fe*'minutes), real-time updates $'ill be ar''ailable every 4 5 seconds. Differcntitll Applied ResearchLaboratories (1986).Institute of Navigatior (1980,
Goad (1985),
techniques are under development by the private sector that will eventually iil- Physical SciencesLaboratory (1989) )
L984), and'
low instantaneous cm-lcvcl relative positioning. In these techniques it is envisionecl
A similar system,known as the Global NavigatiorrSatelliteSystem (GLONASS),
that permanent ground stations will continuously track GPS satellites and suppll' by the former republics ofthe USSR IT consisted of 10 to
is also under development
their observations (in some form) over a cornmunications link to mobile users. Tlre end of 1990.and 24 satellitessometimein the periocl1991 1995.
12 satellitesby the
observations may then be "diflerenced" or combined in some way to get very high- purpose will be for I'orldwide aircraft navigation, but it is also planued
Its primary
accuracy station positions. Since any change in the satellite signals will aflect both system is in reality very closely
to be used by naval and lishing fleets as well. The
receivers equally, any degradation or coding of the signals will have little elTect61
patterned after the GPS, with the primary exception that each satellite I'ill trans-
the accuracy of the relative station positions. just using dilTerent codes as the GPS
mit difierent l-band frequencies (rather than
For geodetic use of these satellites, observations can simply be made for lolgqr The receiver hardware, therefore, will have sourc differencesfronr that
satellites do).
periods, and with more satellites, in order to increase the positional accurac3r.It is current reccivers are, and it is likely many future
of GPS receivers; however, sonte
estimated that absolute accuracies on the order of 1 I rneter or lcss are possiblc.
ones will be, able to receive and process signals from either system. Sorne additional
Alternativelv. tivo or more receivers may also be used in a differential or interfer'-
details on this system are presentcd in Anodina and Prilepin (1989).
ometric mode, analogous to the VLBI obserr.ations discussed previously, and the
relative positions of the two receiversestablished to the millimeter level on baselines
of, say, up to 50 kilometers in length. Centimeter-level accuracy is availablc on ba,se
lines of much longer length, but uncertainties introduced by the Earth's atrnosphcre 4,442 Laser-Ring Gyroscopes Another rnore exotic method lor determining vari-
and the satellite positions cause increasing errors with increasing baseline length. ations in the Earth's rotation has also been proposed lf a laser beam is split arrd
As has already been describedin Chapter 2, in addition lo the positiolal inlbrna sent in opposite directions around a given path on the Earth's surfacc, the light can
tion. the rrser'sclock can be set relative to the GPS satellite clocks to accuraciesas be made to interfere rvith itsclf ripon meeting again Changes in this interference
good as t l0 nanosecondsor better, thrrs providing a time transfer rnechanisrool. pattern will occur because of changes in the Doppler shift of ttre light travcrsi[el
extremely high accuracy. the two different paths, due to variations in thc Earth's rotatiou. Such a dcvice is
For the purposes of Earth rotation. one must assumethat three ol more satellitc known as a "laser ring" or "sagnac laser" gyroscope (Soffel et nl., 1988). Although
tracking stations are in use at known locations. It is then possible to deterurine the theoretically it might be able to provide usable Earth rotation data, in practice the
position of the satellites with respect to the grourrd stations, as well rLsthe oriertto,- stability of the platform on rvhich the laser light is directed seems to be a limitittg
ti,on of the Eorlh with respect to the "inertial" coordinate system cstablished by tht: factor, because of changes in light paths due to vibrations, microearthquakes. etc.
satellites. In practice, such tracking rvill be dorre by the "control segment." Dept. Such a device (covering a 58 m2 area) is currentlv undergoing testiug and upgracl
of Defense tracking stations responsible for opelating thc satellite, other govcrrr ing, with reported accuracies of 10-6 ovcr averaging tintes on thc order of a dar-.
and a predicted theoretical accriracy of a few parts in 10 r0 u'ith avelaging tirnes
ment networks (e.g.,that operated by the li.S. National Geodetic Survey (NGS) at
of a few secorrds(Rotge et al., 1985).
U.S. IRIS VLBI siations), or private gloups. Any of these data are norv partialll'
usable for Earth rotation deiermination, and should be capable of providiripl ltigh
accuracy ERP values once the entire GPS satellite system is opcrational. Abbol. el
ol. (1988) have already used data obtained in 1985 to shorv that errors could be: 4.443 Superfluid Helium Interference Gyroscopes Thc trse of srrperfluid helium
kept to J:40masin,ro,t 70masin yo,and t 0.9ms in UT1-UTC over5 dars .,[ interference gyroscopes (SHIG) has just recently becn proposcd for the detcrmi-
measurement.Zelensky el al. (1990), in a corarianceanalysisof observationsof 18 nation of Earth rotation and polar motion (Packard, 1988). It invoh,'esthe use of
Block II satellites,show that accuraciesin the range of 0.2 to 1.0mas for all three a small (less than 1-meter diameter) horizontal torus of superfluid Leliuln llHe or
ERP componentsu'ill be possible.Paquet and Louis (1990) in a simulatiorL(u ith 'He),
in which a blockage with a pinhole is positioned. When this torus undetgoes
18 satellites and 10 tracking stations making I l0 cm accuracy rrreasuremeDts)sholi' rotation (for example, bccauserof the rotation of the Earth), it should be possi-
possible accuracies in the range of less than t 3 mas for 2,1irours and + 3 5 rnas tor to -"*u." the phase tliflerence or phasr: slips across thc blockage (Avenel and
-tr
6 hours of observationfor pole position, and t 0. 15 0.30ms for UTl-UTC. varoguaux, 1988). This in turn is proportional to the rotational vebcity of the
(For another general overview of the GPS, see Archinal and \'Iueller (1986). torus and therefore
the Earth, and could be rneasured to 1 part in 109 over a few
\{ore detailed inforrrration is also available in Chapter 3 and in Remondi (19t'1), seconds.Of course technical problems still exist in building such a device, such as
 tr
262 EXpLANAToRv
supplEt!1EN-r COORDINATES ANDTHEROTATION OFTHEEARTH 263
4 / TEFBESTBIAL
Survey of the Ilnited S1,atts.Tlrc NEOS acts as a sub-b[rcau
(a) buiiding a pinhole diaphragm of vcr1,small aperture (0.1p holes arc neeclerl). lqational Geodctic
IERS BuL[etin A is issued bv 0 hours UT' Friclay. of each week,
and (b) stabilizing thc irrstrument or determining its tilt r{'ith rospect to thc locql of the IERS. The
vertical (i.e.. its connection to the Earth) \fith the accuracy needed (changcsof Earth rotation paranretersrlctcrmined from the combination of other
snd contains
10-9 meters in tilt would be significant eicrossa l-meter instrumcrrt). If they can recently determined ERP series, predictions of Earth rotation pararneters daily for
successfully be built, horl'ever. trvo or more of this type of instrument. as with llrc up to 90 days in the future, predictions of Earth rotation parametels monthly qr
laser ring gyroscopesJwould provide high-speed dcterminations of LOD and polar to on" y"ar in the future, and equations lor predictions even firrther into the futrrre.
motion independent from astrononical obscrvations. Other miscellaneous data atrd information are also availablc, such as the ERP de
termiled from observations being used in the combined solution. plots of current
Dolar motion and UTl UTC lalues, values of TAI IITC, DUT1, announcerrents
of l"up r""ond., etc. For the combined ERP serics, data given incluclcs for each
4.45 lnternationalServices day: the calendar date and trodified Jrrlian date (NIJD): thc ralues olxn, yo. and
their expected accuracies in arcsccorrds: al)d the values of UTl IJTCI and their
On 1988 January 1. the Intelnational Earth Rotation Scrvice (IERS) licgan o1> expected accuracies in seconds of time. Valucs are ustrally given for approximatelv
a week's worth of data. When available, nerv values from the 1El?5 BulLetin.B are
eration, under the arrspicesof the IAU and the IUGG. It replaced the Bureau
also published. Predictiorrs are given for each day lbr 90 days past the end of thc
International de l'Herrre (BIH) and the InternationerlPolar N{otion Service(IPNIS)
combined data, consisting of the calendar date and N{JD, values of t, and ), in
as the organization rcsponsible for determining ancl predicting the oricrrtation ol thc
arcseconds,UT1 UTC in secorrds,and a tabie giving estirrratcd ir,ccuraciesof these
Earth. The IERS consistsof a boarcl of directors that overseesa "central brrrcau'
values. Equations are also providc<l in older to extend these predictions even fur-
and "coordiuating centers" for each obscrr.ational technique atrcl "sub-bureaus" fol
ther. Similar predictions are givcu for the lirst dav of each rnorrth lbr ()te year in
speciai purposes. Further, each tcchnique may have several "computational ceulcrs'
the future. Predicted celestial pole oflscts (corrections to Arl and Ae) ate also Eiiven
actually responsible for processirrg data and ERP series tir:r.t are then collecrl,crl.
for each dav for 90 days in the futurc.
used, and <listributed br- the Ccntral Bureau.
The combined series being publishcd is actually forned at lhe U.S. Naval Ob
The IERS publishes two different prirnary bullctins for the dissemiuation of
servatory, using ERP series based on LLR., SLR to LAGEOS. and VLBI cla.ta.The
Earth-rotation data as detcrmined from the combination of data frorn diflerelt
individual seriesin use are providecl by the Department of Asl,rononlv at tire Univer-
obserrational techniques. The 1E-RS Bulletin A provides information u'eeklr- ol
sity of Texas (UTX) using LLR data; the Center for Spacc li.esearchat UTX and the
the rapid determination of the Earth's rotation. emwell as predictions for future
Technical Urrivelsity of Delft for SLll to LAGEOS data; a.rrdthe Natiolal Geodetic
rotation. The IERS B ulletin B provicles "lirral" smoothed data rrrorrl,hly on the
Survey and the U.S. Nilval Obscrvatoly for VLBI data. Each series has syslematir:
rotation of thc Earth. Of course,other orgernizationsboth withiu alxl outsiclethc
errors remol'ed fiorn cerchof its available componcnts (,r;, yr,. arxl UT1 UTC). A
IERS also provide othcr ERP series, usually based on obser.valionzrldata of a giveu
spline curve fitted ttrrough cnch of the components is used to detcrmine tlte onc ciay
type.
values for each conpontlt. Accrrracy estimates lbr cach input s(r'ies are estirnatql
The IERS is also responsiblc for deciding wherr a step in UTC is neccssary in periodicallv via intercorrrparisols of the series. Thc iortg-term biases artd trctrds are
order to keep it within 0.7s of lJTI. These steps, conrrnonly knorvn as leap secor,,ds.
adjusted on an annual levei to those of the serics published i the IERS BuLletin
rnaY be positive or negative. are of l second duration, ale announced at least 8
B. The prcdictions for.rn aud )p ale clone from a k:ast-squalcs solution ttsing the
weeks ahead of tiure, and nay be inserted at the end of any month (prcferabh. last two years ol data. l'he predictionsfor UTl IITC are cornputedrtsirtga rrtodcl
and ir practice, at the end of June or Dccember) (BIH, 1972. pp. A-11 to A-14). based on the last 365 days of data. A complete dcscription of the cornbinatiorr ancl
Sirrce lheir introdrrction in 1972, these steps have allvays bccn positir.e, and ha\-(l prediction tcchnique is giver in NEOS (1989) ard NlcCartllv and Luzutr (1990)
been required from evcry 6 months to 2 ycars. A table of these steps is gi\,cn in thc
AII ol the IERS Butletin,4 informatiou is also available on the "Autornated
t:lrrrcnt Astronomiarl Almanar:, p. K9, as diflerences of UTC fron-r 'IAL Data Service" of the U.S. Naval Observatolv, and from the Arpanet, Bll NET, Iu
ternet, G.E. Mark III. ancl SPAN electronic netrvorks. Fbl further inforrnation, or to
be placed on the mailing list for the |ERS Butlctin,4, contact: National Eartlt Ori
entation Service, Earth Orientation Parameters Division, U.S. Naval Observal,ory'
4.451 IERS Bulletn A The IERS But.letin ,4 is published by the National Earth
Washington,D.C. 20392.
Orientation Service(NEOS), a joiDt activity of the U.S. Naval Observatoryand thcr
 264 EXPLANATOF]Y
SUPPLEMENT COOROINATESAND THE ROTATIONOF THE EARTH zoJ
4 / TERFESTRIAL
4.452 IERS Bulktin B The IERS Bulletin B is published by the Central Bureau
For further inforntation, contact the Central Burcau of the IERS at tht: arldress
ofthe IERS, located at thc Paris Observatory in Paris, France. It is issued during tlie
given in Section 4 452'
first week of each month, and contains final ERP values from a combination solutiol
of a month's or more ERP series data, urore recent ERP values and predictions as
available from the 1485 Bulletin A, along with information on the length of dar
(LOD) and ERP series data used in the cornbined solution. The final values, givel OF PASTVARIATIONS
4.5 DETERMTNATION lN LENGTHOF DAY
for every 5 days, include the calendar date and MJD, smoothed values of x, and ANDTHEPOSITION
OFTHEPOLE
yp in arcseconds, UTIR-UTC and UTIR-TAI in seconds, and raw values of .rn
and lp in arcseconds, and UT1 UTC in seconds. Following this are sonc of t|e The determination of historical variatiorts in the rotational rate of the E:rrth has
rapid service and prediction values from IERS Bulletin,4 in the same fornrat. These been a topic of discussion since the latter part of the nineteenth centur'\'. rvhen such
values are also interpolated daily and given with the calendar date and IIJD. _r, variations were first suspected. \Iost of the research in this area has cerrtt'reclon the
and y, in arcseconds,UTi-UTC, UT1-TAI, UTl-UTIR and the excesslength use of ancient astronomical observations of eclipses and tclescopic obscrvatiorrs of
of day ("D") in seconds.Smoothedvalues of "D" and the rotational vclocitv oI the lunar occultations, iIr order to determine At, the differerr<c betq'een Eurrtlt rotatiolal
Earth in microradians/second are also given for every 5 days. time (UTl) and dynarnical time (TDT ol formerly ET). The time dcrivzrl,ive of this
The final values ("raw" above) are computed for every 5 da3rs,using tlata fi.on lue is the length of day (LOD). A sulnmary of recent work in this arca rvill be
various obserr,-ationalERP series as input. The input series are tire sarrrc as rnosc present€d along with sone current restrlts.
listed for the IERS BulLetin A, \Mith. in addition, series obtained from .lPL \,'LBI In addition, availirbleresults on historical variations in polar rrr,'tion arl prc-
observations also being used. The values are then smoothed using a ,,Vondrak', sented. However, since these variatious do not result irr a curnulativ<: cffcct as thev
filter (Vondrak, 1969), The values of UTl-UTC are also corrected for v.rriations do with UTl, older rneasurements of these variations zrre based entirely orr precise
due to zonal Earth tides rvith periods up to 35 da1's.giving UTIR-UTC. Thc telescopic observations and so can only be extended bar:k to the uri<l rrirrctecnth
values are referred to the IERS Terrestrial System, u'hich is identical to 1he earlier century with any dcgree of continuity.
BIH Terrestrial System, with the BIH origin having been adiusted to the CIO in
1967and maintained independentlysirrcethen. This scriesis also a direct cxtension
4.51 Historical
Variations
in UT1and Lengthof Day
of lhe BIH Circular D serles,which ended on Deceurber il, 1987.
All of this inforrnation is also available from the G.E. Mark III and thc EAIiN/ Sincestudies by Newcomb (1882,p. a65) indicated th.rt rneasurernents of thc tran-
BITNET electronicnetworks. For further information, or to be placcd orr the rnail sits of Mercury, the N{oon. and Jupiter's satellites nll s}rou'edsimilar variations
ing list for the 1Z?5 Bulletin B, contact: Central Bureau of the IERS (IERS/CB). possibly related to the Earth's rotatioll. man]'stulics and papers hart been un
Observatoirede Paris, 61, Avenue de I'Obsenatoire. 750l4 Paris. Franrr:. dertaken to determine prore accruatcly these variations over time. Houeter, these
studies have been complicated by thcir dependenceorr the position of thc Nloon
(during eclipsesor occultations), a1d hcnce on wheth.r there has cvcl l)een any
4.453 Other Publicationsofthe IERS change in the acceleration of the l\'Ioon's mean motioll (a) (and cflrrtivolv, lhen.
Other publicatiorrsofthe IEIIS il<lrrde:
any changein the rate of ET deterninations basedol 1hc \Ioon's nrr.rtion).\Ian1'
studies of this problern that hare been undertaken up until the 1970sattcrnptcd
Special BulLetin C Issued irregularly to announcc upconling steps (lcap scc- to solve simultaneously for n and Ar. An example of this type of approach is that
onds) in UTC.
by Newton (1970). However, rr can be assumed constant, an assurnption thilt seeuts
Specio,LBulletin D Issued irregularly to announce upcoming chaugcs 1,othe
to be supported by nxrdern astrononrical observations (including radar alcl LLR
value of DUT1 to be transmitted u,ith time sisuals.
Annual Reportof the IERS Issuedannually fionr lg89 June on, r('lx)rting on measurements) of the Moon. \IaDy studies during thc 1970s and 1980s have been
varioussolutionsfor Earth orientation parameters?uldterrestrial aud t.eltrtiai uudertaken with this assumption as their basis. The rrxrst complcte crrrrent ref-
refetence systems. erences on historical UT1 and LOD varriations (*'ith n assurned constarrt) are by
Technical Notes of the 18R.9 Issucd irregularly as reports and corrrplemen- Stephensonand Morrison
(198a), and McCarthy and Babcock (1986) from u,hich
tary information of relevanceto the work of the IERS on Earth rotntion and much of the material prcscntedhere is derived.The resultspresentedin both of these
related reference systems. reterencesagree well, with
difierpnces duc only to slightly different data sources and
 7-
266 EXPLANAToRY
SUPPLEMENT AND THE BoTATIoN oF THE EABTH
, / TERFESTRIALCOORDINATES 267
different degrees of smoothing before 1861 (McCarthy, personal communication).
Some newer results are also availablein Stephensouand Lieske (1987). Leneth of dav-86400 seconds
The results can conveniently bc considered by examining various historical pc-
riods separately, based upon the type of observations being considered. The specific
periods to be mentioned here ivill include (working backward) 1980 present, 1955
1980. 1620 195,1,700 1600, and 700 B.c. A.D. 200. Observationsalso exist for the
period ,r.n. 200 700, but have not yet becomc available (in published form). For
example Hilton el al. (1989) rcport Al values determined from 58 occultations of
planets by the \'Ioon observedby the Chinesein the period 68 B.c. to A.D 575.
Extended details of the methods for tlte determinations of UTl and LOD rvill
o^
o{,
^
h'tu\,'l''\
not be presented here. For further information on the rnethods used when consid
ering lunar occultation data, Morrison (1979b) may be consulted. Descriptiolrs of [,t'\'
i
-4
how eclipse data were used are presented in Stephenson and Nlorrison (198'1) along
with many additional teferences on data sources. Other chapters here may also
be consulted. Current rrethods for observing UT1 (since 1980) are firlly described
,8
earlier fin Section4.5). 1800
1700
Observations of UTI from 1955 to 1980 are rrnique in that they can be refer
enced to TDT. due to the advcnt of the first atornic time standards in the 1950s Year
However, since mostly optical tcchniques and thcn developing space based tech Figure4.51.1
niques were in use. the accuracy levcls were an order of magnitude lowcr than at Lengthof dayfrom 1656to 1988
the preseni time. Determinations by the BIH providc the best results from 1963
1980 (originally in the form of TAI-UTI, rvith TAI diflcrent from TDT onlv b1-
an additive constant to TAI of 32.184").N{cCarthy and Babcock (198'1)present
the most accurate arailablc resrilts during the period 1956 1962. based on USNO
clock star. The untimcd observations were made largely from Europe, usrrally in
Photographic Zenith Tube (PZ I) observatiols of UT1, as refcrencedto the USNO
nonastronomical works.
atomic timescale A.1 (rvhich is dilTerent from TDT only by an additive constant ltr
A.1 of 32.150"). "Ancient" obsenations during the period circa 700 B.c. to A.D. 200 arc of a
rather limited number, but in some casesare quite useful. Stephenson and Nl[olrison
The period of 1620 to 1955 covers substantia,llythe period where telcscopic
(1984) have been able to use (a) 5 Chinese sightings of total and rrear total eclipses
observations may be used in deiermirrations of UT1. Lunar occultatiorrs alonc pro-
betweeu198 B.c. and A.D. 120); (b) a Babylonian obserr''ation ofa total solar cr:lipse
vide lairly precisevalues from 1861 to 1954 as describedby Nlorrison (1979a).Th(,
in 136 B.c.; (c) "timed" (possibly with rvater clocks) Babylonian observationsof
sparsenessof srrch observations before that timc results in some difiiculties in re
lunal eclipses; and (d) 20 Babylonian records indicatilg that the N{oon rose or set
covering UT1. R.esultsfrorn occliltations and now also eclipses are fitted lvith cubic
while eclipsed.
splines in order to obtain a continuous series of values. Beyond Stephct''orr arr'L
As a sumntary of the preceding results, graphs of Al arc presented in Figures
N{orrison's1984 treatnent, part of this period (1672 1806) has bcen rec('usi.lered
4.51.1 4.51.3.Figure,1.51.1.presentsa curve oI A/ vs time for the recent historical
with nervlv ar,ailable solar cclipsc timings by Stephruson ancl Lieskc (1987).
period of 1656 to the present, after NlcCarthy and Babcock (1985) with recent
Thc "medieval period" ol 700 1600 refers to the gencral peric,d frorn linred
values from the IERS Rapid Service Sub-Bureau. The determinations of Ar by
Arabian eclipse obserr'ations aucl othcr untint:d obsr:rvations until the inrentit rr .'I
Stephenson and N,lorrison (198,1) from isolated observations are presented lor the
the telescope. During this perio<l, a sirrglc "curve" of values for Al rnay rto iongr:r
period,l.o. 700 to 1600in Figurc 4.51.2,and for thc period 700 B.c- to A.D. 200 in
be estimated. One mav only set lirnits on 1,hevalu:s at sporadic epochs dur-irrg
t'l8ure 4.51.3.Finally, their logarithmic plot of Ar is presentedfor 700 B.c. to A.D
the period. The timed Arabian obserl'ations actuallv rart from approximately A.D'
1980 in Figure 4.51.4. The current Ailrononrical Alm.anat also contains a tablc of
800 to 1000, with the timings made by measuring the altitude of the ![oon or a
At valuesfor the period 1620.0to the present.
 COORDINATES OF THEEARTH
ANDTHEROTATION <oY

3000

.:l lr..
-' I
2000
' trl f",l ttlrtlr

ye^r 3 Babyloniantimed eclipse

+ Arabian timed eclipse
III untimed total or annuler solar eclipae
Figure 451.2
I untimed partial solar eclipse
At from A.D.700to 1600
!r Babylonianuntimedlunaror solar
Ir ecliisein which moonse!or roseeclipsed

io r ye^r
o otp
!
o o^i
a
700 B.c. to A.D. 1980(30 secondshavebeenaddedto eachvalue to makethemall positive.
'b !'l line showsthe comDutedtidal de{€lerationof the Earth.)
!
;: ?,r
_t:

lil h"F "l

\ ol o;r
or HistoricalVariationsin PolarMotion

of the International Latitude Service in the 1890s brought about mom-

of polar motion that has continued to this day. However, except for spora'dic
duriag the nineteenth century little data on polax motion eists before that

high-precision polar-motion data, as with UT1 data' exists only ftom

with the introduction of the space-based geodetic techniques discussed in
-600 4.4. The recent motion of the pole is presented in Figure 4.52.1, using llalues
- 400 -200 200 by the IERS Ra.pid Service Sub-Bureau.
yeat
pola,r-motion data, based on la.rge numbers of optical astrometric
Figure 4.513 became available just prior to the formation of the International Polar
At ftom 700B.c.to A.D.200 Sen'ice in 1962. Doppler satellite tracking data also began providing polar-
in the late 1960s, aud was iq use by the BIH by the early 1970s'
 270 E X P L A N A T O RSYU P P L E M E N T

Pola.rmotion ILS polar motionin y

u,o
J- | meter

-.2
5)
a o.2
i5
c
0.0

c
-0.2

,4
-0.4
',1919
7< 1899 1909 1929 19J9 1949 1959 1969 1979 1989

Y (scr.of arc) yeilr

Figure4.52.1 Figure 4.52.2

Polarmotion, I980 Scptember28 1()1990July 27 (lrianglesnrarkthe beginningol' x and y components
of polarmotion tiom the ILS. 1899.9to I979 0
eacn year,

dating from 1863 to 11185from rarious obserlatolirs, h<r <lemonsl,ratt:<l<lecisiveh'

the existcncc of the rrrof,ior now kno$'l as L[e Chatrrllct wobble. Chnnclltrl actuallv
This type of data stiil continuesto bt: ilvailable uutil 1.hcpresentdal, brrl ltas bectr considered data over tlrt: r'ltire perirxl 1726 (from Br-adlt'r') to 1890. rvhich iIt<licated
srrpcrsecled iD accuLacr_ lrr. the ner\r:r ltchniques. large changes were o(( ul'rin€! in the lcrrglh of thc Chatr<ilel pcriod. Tllcsc chatlges
Tlte stations of tlro ILS providr:rl urouitoring of polar nlotiolr frorl Scplem]rer' are usuall]- considererl lo l)e ar.tila(ts oI the sparse 1611'precisiolrdata itr ltse.
of i899 until 1987. A cdnplete arr<lligor.ousrerttlrr.tiou of ail of rlrr, Il,S clata
fronr 1899.9until 1979.0is describtrl rur(l rcported by Yrrrui and Yok<xrr,rrra (1980).
Figurc 4.52.2,shou'irrgthe X atrd f valucs vr:rsustinrrrrlrrring that poli(xl. is l2rkel
IroIn Figtrrc 8 of thll 1>ublication.Forlorov et aL.(\972) also present Jrolar-ruotion
results from 1890.0to 19ti9.0.
Br.forc 1900..1?rlathat cau bc uscrl for polar-notion <k:tr:rrrilatiorr :rre q]litc
sporarlic.All such rlal,a are based on shod observation<arnpaigns, sllalh- at olle IM
statiolt, and althorrglrsome observcrswrrrc in fact lookirrg for politr trn)l,iol. niost
of thc observationsu,'('rctransit-cir(:loor visual zenith irrslnrtnent oltscrvati()llsb{r-
lng used to detcrmiDt's1aI positiolls allrl/or-timc. Polar-rrrotiondata arc a\ailabl:
trorrt the pleliouslv rrrorrtiorrcd rvork lrr.Fecloror.el a1.fi.om 1890.0to 1900.0.Co1-
ering a ntore lengthl' por.iodduring tlrt. nineteenth <.crr1.rrrl.,specificalll,rlrrringthe
periotl 1846 1891.5,is 1hcstudy by R.vkhkrva(1969).!'igufe 4.b2.3showsthr: polar-
motiorr resrrlts durirrg that period. Surprisinglr,.,one oI the ol,her few disotssions of
Figure4.52.3
polar-motion data bef<xc1900is th:rt of Chandler hirns<tf(1891, 1892).Using data ofpolarmotion,1846.0to1891.5
x andfcomponcnts
 EXPLANATORYSUPPLEMENT
w
COORDINATES
4 / TEHBESTBIAL OFTHEEARTH
ANDTHEROTATION 273
4,6 REFERENCES (1971). "New Computations of the Tide-Generating
Cartwright, D.E. and Tayler, R.J.
Potential" Geophgs.J. Roy. Astron. Soc,23, 15 71.
of T,atitude" The Asttonomical Jounurl ll,83.
Abbot, R.I., King, R.W., Bock, Y., and Counselman III, C.C. (1988). "Earth Rotatiorr Chaodler, S.C. (1891). "On the Variation
from Radio Interferometric tacking of CPS Satcllites," in Babcock and Wilkirrs. Chandler, S.C. (1892). "on the Variation of Latitrtde" The Astronomical lournal L2, 17.
1988, 209 213. Cbapman, S. and Lindzen. R. (1970). Atmospheric ?ides (D. Reidel, Dordrecht, Hollald).
Agnew, D.C. (1983). "Conservation of Mass in Tidal Loading Computations" Geophys. J. Clark, T.A., Corey, B E., Davis, J.L., Elgered, G., Herring' T.A., Flinteregger, H.F.,
Roy. Astrcn. Soc. 72, 321 325. Knight, C.A., Levine, J.L, Lurtdqvist, G., Ma, C., Nesman, E.F., Phiilips, R.B.,
Alley, C.O. (1983). "Laser Ranging to Retro Reflectors on the Moon as a Test of Theo- Rogers, A.E.E., Ronrrang, B.O.. Ryan, J.W., Schupler, B.R., Shafler. D.B.. Shapiro,
ries of Gravity" in Quantum Ctptics, Experimento,l Gralritg, aniL Measurernent Theorq 1.1.,Vandenberg, N.R., Webber, J.C., and Whitney, A.R. (1985). "Precision Geodesy
'.frans.
( P l e n u m ,N c w Y n r k , N Y ) , 4 3 1 . Using the Mark III Very Long Baseline Intelferometer Svstem" IEfl on
Applied Research Laboratories (1986). Proceedings of the Fourth International Symposium Geosci. and Rem. Sens. GE-23, 4' 438 449
' 'Space Age Geodesy:
on Satellite Positi.oning (ARL, University of Texas, Austin). Coates,R.J., Ftey, H.. N{ead,G.D., and Bosl"'rth, J.XL (1985) The
Anodina, T.G. and Prilepin, M.T. (1989). "The GLONASS System" in Physical Science NASA Crustal Dynamics Prcject)' IEEE Trans. on Geosci. and, RerrL. Sens. G8 23,
Laboratory, 1989, 1, 13.18. 4, 360 368.
Archinal, B. and Mueller, I.L (1986). "GPS An Overview" Professional Surueyor 6j no. 3. Cohen, S.C., King, R.W., Kolenkiewicz, R., Rosen, R.D, and Schutz, B.E., eds. (1985).
7 11. "LACEOS Scientific Resuits" speciai issue ol Journal of GeophgsicaLResearch 9O,
Arias, E.F., Feissel,M., aud J.-F. Lestrade (1988). "An Extragalactic Celestial Refererr:e 811, 9215 9438.
Frame Consistent with the BIH Terrestrial System (1987)," in 81ff 19E8, D-113 t,:, Degnan, J. (1985). "Satellite Laser R.arrging: Cunent Status and Futrtre Prospects" 1E-E-E
D-121. Trans. on Geosci. arLdRem. Sens.GE 23, 4, 398 413.
Armstrcng, N. (1977). Foreword to Mulholland 1977, xiii xiv. DeMets, C., Gordon, R.G., Argus, D.F., and Stein, S. (1990). "Current Plate Nlotions"
Geophys. J. Int.7O7, 125 474.
Arnold, K. (1974). "Ceodetic Aspects of Laser Distance N{easurementsto the Moorr arrcl Department of the Army (1958). UnLxersaLPolar Stereographtc Grtd TIl 5 2a1 I (De-
Radio-Interference trlleasurements to Quasars" Cerland.s Beitr. Ce,ophgszk83, rLo. 1. partment of the Army, \lhshirrgton, DC).
219 269. Department of the Army (1973). Ilni\ersal Trnnsuerse Mercator Gnd TNI 5 211 8 (De
The Astronomical Almanac (,1990). (U.S. Govcrnment Printing Olice, \\'ashington, DC). partment of the Army, Washington, DC).
Avenel, O., and Varoquaux, E. (1988). "Josephson EIIect and Qualtum Phase Slippagc DMA (1987). Suppletnent to Departrnent oJ DeJense |lorld Geodetic Sastcnl 1984 Tech-
in Superfluids" Phllstcal Reoiew Letterc 60, no. 5, '116 419. nical Report D}{A TR 8350.2 A (Defense \{apping Agency, U.S. Naval (lbscrvatory,
Babcock, A.K. and Wilkins, G., eds. (1988). The Earth's Rotation and ReferenceFtnnes Washington, DC).
Jor Geodesg and Geodgnamics IAU Syrnposium 128 (Kluwer. Dordrecht, Holland). Doodson (1921). Proc. R. Soc.,4. 100, 305 329.
BIH (1973). BIH Annual Report Jor 1972 (BIH, Paris Observatory, Paris, France). Eanes,R.J., Schutz. B., and Taplel'. B. (1983). "Earth and Ocean Tirle Eflects ort LAGF-:OS
BIH (1989). BIH Annual Report lor 1988 (BIH, Paris Observatory, Paris, France). and Starlette" ProceerT'ingsoJ the NirLth International Sgntposiunt ort Earth Tides (E
Bock, Y. (1980). "A VLBI Variance Covariance Arralysis Interactive Cornputer Prograrri' Schweizerbartsche Verlag Buchhandlung, Sluttgart).
Dept. oJ Geotl,etrcScience Report 298 (Ohio Statc Univcrsity, Columbus, OH). Farrell, W.E. (1972). "Deformation of the Earth by Surface Loads" Beu. GcoPhlls Space
Bomford, G. (1971). Ceodesy,3rd ed. (Oxford University Press, London, UK). Phys. 10,761 797.
Borkowski, K.M. (1989)- "Accurate Algorithms to Transform Geocentric to Geodetic CL) Fedorov, E.P., Korsun, A.A., [4a1'or.S.P., Panchenko. N.I., Taradii. V.K., ard Yatskiv.
ordinates" BuLletin GictdEsique63, no. 1,50 56. Y.S. (19?2). The Motion oJ the Pole of thc Earth lrom 1890.0 to 1969 0 (Ukr:rinialt
Boucher, C. and Altarnimi, Z. (1989). "The Initial IERS Terrestrial Referencc Frame' Academy of Sciences,Naukol'a Dumka, Kiev, USSR).
IERS Technical Note 7 (IERS, Paris. France). Gilbert, F. and DziewoDskiJA.N4. (1975). "An Application of Nornal Nlode Theory to
Brosche, P., Seiler, U., Suendermann, J., and Wuensch, J. (1989). "Perbdic Changes in the Retricval of Structural Paranreters and Space trllechanisnrsfrom Seismic Spe<:tra"
Earth's Rotation due to Oceanic Tidcs" Astronomu and Astrophys;cs 220. 318 320. Phil. Trans. R. Soc. of Lond. A27E, 187 269.
Carter, W.E. and Robertson, D.S. (198,1)."IRIS Earth Rotation and Polar N{otiorr llc- Goad, C.C. (1980). "Gravimetric Tidal Loading Computed from lntcgrated Greerr's Func-
sults" in Proc. Int. SAmposium Space Techniques Jor Geodynamics Vol. 1 ,1. Sornogli liors" J. Ceophgs. fies. 85, 2679 2683.
and C. Reigber, eds., (Hutgarian Acaderrry of Sciences,Sopron, Hungary), pp. 21'1 Goad, Clyde (ecl.) (L9a5\. Proceedingsof thc First InteT"nati.onalSymposium on Precis(-Po'
222. sitioning u;th the Global Positzoning Sgrlem (Natio al Geodctic Inforlratiorr Centcr,
Carter, W.E.. Robertson, D.S., Pettey,.LE., Taplcv. B.D., Shutz,8.E., Eanes, R.J., and NOAA. Rockville. MD).
Lufeng, i\4. (198'1)."Va ations in the Rotation ofthe Earth" ScierLce GSFC (1976). Mathcmatical Theorg of the God.tl,ard,Traje(:torl! Determ.ination Svstem X
224,957 961.
582 76-77 (GSFC, Greenbelt, NID).
Cartwright, D.E. and Edden, A.C. (1973). "Correctcd Tables of Tidal Harmonics" Geo-
Heiskanen, W. and Moritz, H. (1967). Physical Geod.esg(W.H. Freeman, San Francisco,
phys. .1. Roy. Astron. 5oc.33,253 264.
CA).
 THE BOTATIONOF THE EARTH 275
274 EXPLANATORYSUPPLEMENT 4 / TERRESTRIALCOORDINATESANO

Hilton, J.L.. Seidellnanrr,P.K., and L. Ciyuan (19E9). ,,ChinescRecords of Lunar Otr:n1_ I,{,rsakis.D.N..Josties,F.J.'Angerhofer.P'E'.Florkowski'D,R..}'IcCarthy'D.D'.Ji-
tations of the Planets" Bull. of Am. Ast. Soc. 21. no. ,1. 1153. Abstract only. "'""1"^", X, and Yunl"u P (1986)' "The Greel Bank lnterferorncter as a Tool for the
gl' no 6'
IAU (1983). "Proceedings of the Eightee{th General Assenbly patras 1982,, Tians ol tlre fi"^",rr"*""t of Earth Orientatiorr ParalnetemJJ The Astronomical Journal'
1,4tI 18B (Reidel, Dordrecht, Holland). 1.163 1473.
(1985) "Connected Fllement'
IERS (1988). IERS Internat,tonal Earth Rokttion Sertrice (U.S. Na\,al Obserratort, \\rash- r,r-n""ihv. D.D.. Angerhofer, P.E , and Florkowski, D R
irrgton, DC). "" "r]ifii"ronr" ter R"sults ir Project NIERIT" in \'Iucller, 1985' 1' 341
350
(1986) "The Length of Day since \656" Plrysxcs of
IERS (1989). An Etplanatory Supplement to the IERS Bulletins A and B (IER.S, Lr.S McCarthy, D.D. an<i Babcock, A K
Naval Observatory, \Vashington, DC). t h eE o r t h o n d l l o t t t o r q I n t ' t ' o " ' 4 4 . 2 a 1 2 ! 2
Institute of Navigation (1980). Gfobal Positioning ^9ysfem(Thc Insijitute of Navigatitu. McCarthy,DD.,ed.(1989)"IERSStandards(1989)"ltnsTechnicalNote3(Ceniral
Paris' Paris' France)
Washington, DC). Bu."l., of the IERS. Observatoire de
Institute of Navigation (1981). GLobal Pu;tioning SgstenL Volume 11 (The lnstitute of Frequency Variations In Earth Rotation" in trIccart6Y €t
McCarthy, D D. (1990) "High
Navigation, Washington, DC) al., 1990
(1990) "Combination of Precise Obsen'ations of ihe
Kaula, W.M. 11966). Theory of Satellite Geodesg (Blaisdell, \\raltham, MA). M"Cr,.ihy, D.D and Luzurn, B
Kaula W.trI. (1975). "Absolute Plate N{otions by Bourrdary Vclocity N,Iinimizations,,,/. Orientation of the Earth" Submitted to Bulletin Gdodistque'
Geophys. Res. EO, 244 248. D., Paquet, P, and Carter, W ' eds' (1990) Proceedings ol llLe IUGG S1l7n-
McCarthy, D
*"-;;;;;*
(American Geophysica) Union' \\in'shington'
Klepczyrrski,W.J. (1982). "Coordirator's Rcport: Connectcd Elements Radio lnterferotr- "Variations it Earth Rotatictn"
etr']'JJin ProJec, I\IERIT: Report on the Short Campaign G.A. \\rilkins and NI. Feissel. DC).
Applit:ations"
eds. (Royal Greenrvich Observatory, Herstmonceux, UK), 34. 77. Mclintock,
",'-ph.l. D.N- (1980) "Ver1'Long Baseline lnterferometr) and.Geodeti(r
Kouba. J. (1983). "A Review of Geodetic and Geodynamic Satellite Doppler Positioning (Univ of Nottingham, Dept of Civil Engineering' IJK)
ag"rt.tion
Nltotiors" -/ Geophgs Res SS'
Reu. Geoph.ys.Space Phgs.2l,27 10. l{i""t".,.l.S. and Jordan, T H. (1978) "Present-Dav Plate
Kovalevsky,J. and NIueller, L (1981). "Comrlertts on Conventional Tcrr€strial and euasi 811, 5331 5354.
Sgste'ms(A Manual
Inertial Reference Svstens" irr ReJerence Coordi.nate SystenLsJor Earth Dynamics E. Ivlitchell,H.C. and Simmons. L G. (1945) The StatePlane Coortlinate
Publication i:s Gcodetic Surl'e-Y Rockville' NID)-
Gaposchkin and B. Kolar:zek,eds. (Reidel. Dordreclrt. Holland), pp. 325 :18,1. for Suruegors) Special lNut;ntltl
Publishers.
Koralcvsky, J., Nltellcr, L,:rrrd Kolaczek. B.. eds. (1989). ReJerenceFrarnc:s.tlrA ronc)n1g l,toit, f.. .rrj Bouchartl, H. (1925). Surueying,6th ed. (Intext Educalional
ancl Ceophgsics(Klulr'er, Dorclrecht,Flolland). New York, NY).
oJ the EatetnaL
Kurnar. \'I. (1972). "Coordinatc Transformation by N,Iinimizing Correlations bctwcen Pa- Molodenskii, tr{.S.,Ercmeev, V F . Yurkina' NI I. (1962) Mcthods fa1'Studll
Figurt: of tlte Eorth (Israel Prograrn for Sciettifir: Translations'
ramcters" ,€pl. of Geodetic Science Repctrt 18/, (Ohio Statc Unil,ersity, ColurnbrLs. Grauitational Fieltl ttnrl
oH). Jerusalem).
UK)
Lambcck. K. (1971). '*Detcrmiration of thc Enr.th's Polc of R.otation frorn Laser liatge Moritz, H. (1980). Aduance,l Phgsical Geodesg(Al''acus Press, Tunbrirlge \Vells' KeIrt'
Thettrv and Obserualion (Lingar/Clon-
Obscrvations to Satellites" Ru,lletitLCeadesique101. 2fj3 280. Moritz, H. and llluellcr, LI. (1987). Earth R.otatian:
Lanrbeck. K. (1980). Tire Earth's Var.ia.bl.e1?otation (Carnbridge Universitl prcss. C:[r- tinuum, Nen' York, NY).
of
bridge, Englancl). Morrison, L.V. (19?9a). "Re-clcterlninationof the Decade FluctuaLions in the Rotrtion
the Earth in the period 1861 1978- Gcophvs JR Astr' Soc' 58':l'19 360
Larden. D.R. (1982). ",\'Ionitorilg the Earth's Rotation bv Lurrar Laser Ranging rRc2r.,r/s tor
Morrison, L.V. (1979;). .An Alalysis of Lunar (lccultations in the Years 1943 1974
front the SchctoloJ Srrueging UNISIJRV S 20 (Univ. of,\eu,South \\iales, Kcnsington. Tlreory, tllc RiSht Ascensjon Systcm of the
corrections to thc Constants in Brorvn,s
Australia).
FK4, and \\iatts'Lurrar profile l)atum" ,L|oz. Not ll Astr' Soc' 187' 41 t2
\4a, C. (1978). Verg Lon.g Dastline InterfcronrctrlJ Applictl ttt Polttr Mottotr, IlcktLtlitu ntl o1
Mueller, i.I., and Rockie. .f.fi. lrooe;. Grauimetric arrl Celestial Gcodcsv:A Glossarg
GeodesyTM 79582 (NASA GSFC. Greenbclt. \ID).
?errls (Ungar, Ncrv York, NY).
Nlarko$,'itz,W.. Hall, R.c., Essen,L., and Parry, .l.V.L. (1958). ,,Frcquerrcl.of Cesiun irr
Mneller, LII 16ss1. s1,n"r;r"i anri Prar:tical Astrotornv as ApplierL to Gr:odcs37(llDg:rr'
Ternrs of EplrerrrerisTimc" PAgsicalr?euiettLetters.l, ul 3.
New York, NY).
Nlarkorvitz,\V. (1960). "1'he Photographic Zenirh Tubc alld t]re Dual Rate Nloon Camera' Mueller, I.I., fi,r-"lr, U , Rcillv, J.P., Saxena, N, arxl Soler. 1' (1973) "Global Satellite
Stars artd Stello.r Systents G-P. Kt\rer and B.\,1. X,liddlehurst.cds., I (Univcr.sitl of (Solu
Trianguiation and Trilateration for the \ational Gcodetic Satellite Prograrn
Chicago Press. Chicago, IL). trons WN 12, 1.1,nnd 16)" DepartmerLtoJ Gcodetic St:ienrcl?e'portNo' 799 (Thc Ohitr
N{arsh, J.G., Lerc}r. F..1., Putney.8.H.. (lhristodoulirlis. D.C., Smit}r, D.E., Fclsterrtrcger'. State University. Colurnbus, OII).
T.L., Sanchez,B.V.. Klosko, S.\'L. Pavlis, E.C., \lartin, T.V.. Robbins..l.lV.. \Villiar:r Mueller, I.L, ed. (1;85). Proce",t;ngs' uf the. Intc:rnational Cont'erttce on Earth RoLation
"Fro)r"
son, R..G., Colombo, O.L.. Ro*.lands, D.D., Hddy. \\'.F.. CtLandier. N.L., RacllirL, ond the Terrestria,l R"f.r"nr. (Dcpt. of Geocl sci and Sttrveying The Ohio
K.E., Patel. G.B., Bhati. S., and Chinrr. D.S. (1gfls). ..A N€l,i,cravitational N{odel State University, Columbts,'Co.,rai"at.
OII).
for the Earth frorn Satellite liacking Datn: C]EN,{'l I" J. GeoplLys.l?es. 93.6169 6215. Mueller, Ll. (rsssj. "n"f"ror.. Systerns: An Upclate irr Tht:orv oJ Sotellite
\ilartin, T.V., Oh, LH.. Eddy, W.F., and Kogut. J.A. (1976). Volume I GEODYN Sust.em. Ceodeqt ant),'Grauity Field Determxnation R Rumrnel :rnd F Sanso' eds (Sprirger
D esuiption (EG&G/\\'ASC. Riverdale, NID). Verlag, Heidelbcrg).
276 EXPLANATOBYSUPPLEMENT
x ! / TEFFESTRIAL
ANDTHEROTATION
COORDINATES OFTHEEARTH 277

Mulholland, J.D., ed. (1977). ScientiJicApplications of Lunar Laser Ran4ing (Reidel. Dol- q"vder, J.P. (1983). "\Iap Projections Used by the U.S. Geological SuI\ey" Geological
"-"
drecht, Holland). Suru"y Bulletin 1532 2nd ed. lU.S. Government Printing Olfice, 1'Vashington.DC).
H. and Schneider, N'I- (1988). "ReferenceFrarnes in Rela-
Munk, W.H. and MacDonald, G.J.F. (1960). The Rotation oJ the Earth: A Geophgsical Sofiel,M., Herold, H, Rnder.
Discussion (Cambridge Univ. Press, Cambridge, UK). rivisric S p a c e T i m " i n B a b r o c ka n d W i l k i n s . 1 9 E 83
. 9 I03
Newton, R.R. (1970). Anr:ient Astronomical Obseruations an!1,the Accelerations of the O.J. and Fanselow, J.L. (1987). "Observation trIodel and Parameter Partials for the
Sovers, 'MASTERFIT'-1987" JPL Publicatlon E3 39, Rev.3.
Earth and Mooa (The Johns Hopkins Press, Baltimore, MD). JpL VLBI Parameter Software
(1988). "Changes in the Earth's R.ate of Rotatior
Newcomb, S. (1882). "I)iscussion and Results of Observations on tansits of N{erctr\-. Stephenson,F.R. and Lieske, J.H
from 1677 to 1881" ,4slr. Pap. Am. Eph. 1. part 6, 363 487. Betweel A.D. 1672 and 1806 as Deduced from Solar Eclipse Timings" Aslronorng azl
NGS (1985). "Geodetic Software" Information flyer 85 10 (NGS, Rockville, MD 20852). Astrophgsics 200. 218 224.
Rep nted January 1987. Stepheoson,F.R. and N{orrison,L.V (1981). "Long term Changes irr the Rotation of the
Packard, R.E. (1988). "Principles of a Superfluid Helium Interference Gyroscope: A Neu' Earth: 700 B.c. to A.D. 1980" Pntl. Trans. R Soc I'ond.) A 313, 17 70.
Tool to Study thc Earth's Rotation" EOS 69, no. 4.1,1153. Abstract only. Teisseyre, R., ed. (1989). Graxity ond Low-Frequenca Geodynamics (Elsevier Science Pub-
Pagiatakis, S.D., Langley, R.B., and Vanicek, P. (tOS+). "Ocean Tide Loading: A Global lishels, Amsterdam.)
Model for the Alalysis of VLBI Observations" in Proc. oJ the 3rd International Syrn Thomas, P.D. (1952). "Conformal Projectiols in Geodesy and Cartographv" U.S. Coast
posium on the Use of Artificial Satellites for Geod,esyaru1,Geodynamics Errr'ioni. and Geodetic Survey Special Publication No- 251 (U.S. Govcrrme t Printing Office.
Greece, Sept. 1982, G. Veis, ed. (National Techn. Unir,., Athens), 328 340. Washington, DC). Fifth printing, 1968.
Paquet, P. and Louis, L. (1990). "Simulations to RecoverEarth Rotation Parameters with Torge, W. (79a0). Geodesy, C. Jekeli, Trans (Walter de Gruyter, New Yolk, NY)
GPS System" in McCarthy et al., 1990.
Parker, W. and Bartholomew, R. (1989). "Issuesin Autonating the Nlilitary Gritl Refer Vondrak. J. (1969). "A Contribttion to the Problern of Smoothing Observational Data"
ence System" in Physical ScienceLaboratory, 1989, 1, 303 913. BulL Astron. Inst. Czechoslotnkia 20, 349 355
Wahr. J.M. (1981). "The Forced Nutations of an Elliptical. Rotating, Elastic, and Ocean-
Physical Science Laboratory (1989). Proceedings of tlte Fifth International Georletic Sryn
less Earth" Geophys. J. Rolj. Astrorr. 9oc. 64, 7O5 727.
posiurn on Satellite Positioning, 2 vols. (Neu'Mexico State University. Las Cruccs.
Wahr, J.M. (1985). "Deformation Induced b1-Polar N{otion" J. Geophys.,Res.90. 9363
NI,I),
9368.
Rabbel, W. and Schuh, H. (1986). "The Influence of Atmospheric Loading on VLBI Ex
Wahr, J. (i98?). "The Earth's C21 and S21 gravity coellicients and thc rotation of the
periments" J. Geophysics 59, 164 170.
corc" GeophlJs.J.R. Astr. 9oc.88,265 276.
Rapp, R.H. (1980). Geometric Geot)esy Volume II (Arhanced, Techniques) (Department of Wahr, J. (1990). Correction to the above paper, Geophys. ./. 1zl. In press.
Geodetic Science,The Ohio State University, Columbus, Ohio). Yoder, C.F., Wiliiams,.LG.. and Parke, X'I.E.(1981). "Tidal Variatiorrs of Earth li.otation"
Remondi, B.\V. (1984). "Using the Global Positionirrg System *(GPS) Phase (Jbservablc J. Geophys.,Res.E6, 881 891.
for Relative Geodesy: Modeling, Processing, and Results" Ph.D. dissertation ar-ailablc Yumi, S. and Yokoyama, K. (1980). "Results of thc International Latitude Scrvice in a
as CSR 84-2 (Center for Space Research, The University of Texas at Austin, Austin, HomogeneousSystem" (International Polar lr'lotion Service, Mizusaua, lapan).
Texas). Zelensky, N., Ray, J.. and Liebrecht, P. (1990). "Error Analysis for Earth Oricntation
Robertson, D.S. (1975). Geodetic and,Astrometric Measurements uith Very Lory1 Base:hrLc Recovery from GPS Data" in N{cCarthy el ol- (1990).
InterJerometrg X 922 77 228 (NASA CSFC, Greenbelt, MD).
Rotge, J.R., Shaw, G.L., and Emrick, H.\'V. (1985). "\{easuring Earth Rate Perturbatio s
with a Large Passive Ring Laser Gyro" in Mueller, 1985, 2, 719 729.
Rykhlova, L.V. (1969). "Motion of the Earth's Pole, 1846.0 1891.5,from Obserr':rtiorrsai
Pulkovo, Greenwich, and Washington" Soliet AstronomA 12, no. 5, E98 899.
Sato, T. and Hanada, H. (1985). "A Program for the Computation ofOcean Tidal Loadtng
Effects 'GOTIC"' in Mueller, 1985 2,712 748.
Scherneck, H.G. (1983). Crustal Loadi.ng AJJecting VLBLSites Department of Gcociesy.
Report No. 20 (University of Uppsala, Institute of Geophysics, Uppsala, Swedcn).
Schwiderski, E.W. (1978). Global Ocean Tides, Part 1, A Detailed, Hydrotlyna,mical ModeL
Report TR-3866 (U.S. Nav. Surface Weapons Center, Dahlgren, VA).
Schwiderski, E.W. (1980). "On Chartirrg Global Ocean Tides" Reo. Ceophys. Spo.cePh1ls
18, 243 268.
Schwiderski, E. (1983). "Atlas of Ocean Tidal Charts atd Maps, Part I: The Semidiurnal
Principal Lunar Ticle Mz" Marine Geotlesy 6,219 256. (See also Chapter 8.)
Shawe, M.E. and Adelman, A.G. (1985). "Precision Laser Tracking for Global and Polar
Motion" .LE -E Trans. on Geosci. anaLR.em. Sen3. GE-23, 4. 391 397.
 CHAPTER

OrbitalEphemeride
of theSun,Moon,andPlanets
XX Newhall,
by E.M.Standish, J.G.Williams,
andD.K.Yeomans

EPHEMERIDES
FUNDAMENTAL

fundamental ephemerides are the basic computation, based on the equations

and fit to the observational data. of the positions and velocities of so-
slrsten bodies. Funda.mental ephemerides a.re the bases for computing apparent
representational ephemerides, phenomena, orbital elements, and sta-
characteristics. The fundamental planetary and lunar ephemerides of ?lre
Almanac are D8200 /LE200. constructed at the Jet Propulsion Labe
in L980. They result from a least-squa.res fit to a rariety of obserrational data
by a numerical integration of the dynamical equations of motion governing
major bodies of the solar system.DE200/LE200 are the basesfor computing
apparent ephemeridesand phenomenathat a"relisted in the almanac.
' 'This chapter describes the equations of motion used for the major bodies in the
hr system, the numerical integration program used to compute the motions, the
data to which the ephemerides were adiusted. the reference frame in
the ephemerides were adjusted, the reference ftame in which the coordinates
orpressed, and the estimated accuracies of the resultant positions,
Numerical integration of the equations of motion is currently the most accurate
of computing fundamental ephemerides. The accuracy is limited by the
of the observational data, by the completeness of the model of the sola.r
ueed, and, if integrating for a very long time period, by the round-ofi and
errors of the numerical intecrator
analytical or numerical theories ca,u be developed as fundamental
the number of terms required for current observational accuracy is
Rather, such theories are currently used for limited-accuracy ephemer-
the sake of convenience.

1l{ril'
 w
280 EXPLANAToRY
suPPLEr\rENT 5 / ORBITAL
EPHEMER,DESOF THE SUN, lrOON, AND PLANETS 281

Representatil,e ephemerides to specified accuracies can be developed based ol OF EPHEMERIDES

5.2 coMPUTATION
Chebyshev polynomials alone and in combination ll'ith reference orbits. N{ean orbit
elements can be used to compute approxirnate positions and velocities. S.21 MathematicalModel
Historical information about former methods of computing ephemcrides is also
Newhall et al. (1983) present the equations of motion integrated for the creatiorr
provided.
of the planetary and lunar ephemerides, DE200/LE200. That descriptior is t-epro-
duced here. The mathcmatical model includes contributions froln: (a) poirrt-mass
5.11 GravitationalModel interactions among the N{oon, planets' and Sun; (b) general relativity (isotropic,
parametrized post-Newtonian); (c) Ncwtonian perturbations of a few selected astcr-
The gravitational model used in DE200 includes all the knorvn relevant forces acting oids; (d) Moon and Sun action on the figure of the Earth; (e) Earth and Sun action
upon and within the solar system. A relevant force is defined as any force that on the figure of the Nloon; (f) Earth tide action on the Nloon: and (g) phvsical
produces an observable or measurable effect. libration of the Moon
Included in the equations of motion were the following: (1) point-mass irr-
teractions amo[g the N{oon, planets, and Sun; (2) general relativity (isotropic, 5.211 Point-Mass Interactions Tlte principal gravitational lbrce on thc nine plan-
parametrized post-Newtonian), complete to order I / c2; (3) Newtonian perturb?r- ets, the Sun, and the Nloon is trodeled by considering those bodies to be point
tions of a few selected asteroids; (4) lloon and Sun action upon thc figure of the massesin the isotropic, parametrized post Newtonian (PPN) n-bodv rrrctric (Will,
Earth; (5) Earth and Sun action upon the figure of the \'loon; (6) Earth tide action 1974). Also included are Nervtonian gravitational perturbations frorn the ast,eroids
'Iirese
upon the Moon: and (7) physical libration of the Moon. The equations for these Ceres, Pallas, Vesta, Iris, and Bamberga. five asteloids were found to have
efects are described in detail in Section 5.21. The relativistic efiects on the tirrre the most pronounce<leffect on the Earth \Iars range over the time span covered by
arguments are discussed in Chapter 2. the Mariner 9 and Viking spacecraft range mcasurentents. The n-body equations
were derived from thc variation of a time-indcpcndent Lagrangian actiorl irltegral
formulated in a nonrotating solar-systerl baryccrttric cartesian coordiuate liame
5.12 TheEphemeris
Reference
Frame The referenceplane is the Earth's mean equator of B1950.0. Thc r axis coincides
approximately u'ith thc 1950 equinox of the FK4 catalog.
The positions and velocities of the ephemerides are referenced t'ith respect to incr
For each body I, tlte point-rnassacceleratiouis given bv Equation 5.211 1
tial space, since the equations of motion are defincd with respect to inertial space.
Alry systematic error present in the ephemeris positions ol motions should be re- - r,rf 2l J+.r\- pt l,\_ f,.,.l!)'
garded as ephemeris error! not as a rotation of the coordinate refelence fiame. i' ,. p o i n \ n ^=. .\ /- . r / , { r , t _\2t
l- c- L r,* c' (_ r,, \1t
tlt
I' \ lt (11
The orientatiol of the referenceframe of DE200/LE200 is described bv Standish
(1982).Sirrcethe dynamical equinox is a quantity inhererrt irr the ephemerides,it . tv,t) - 2(l+i) .r ltr,- r,r i, It I
+r+.)(;) i1 ii r2-{r'-r'i.riJ -l
is possible to derive its location frorl the ephemeris itself. The alignnrcnt of the ,, 2rj | .
epiremerisreference frame with thc J2000 equinox of the ephemerides then b<:t'oures
relatively straightforward. Thus, the dynamical equinox, i[helent in an ephcmeris.
is not equivalent to the equinox of aly specific stellar or radio-source catalog.
(5.211-
5.13 TheAstronomical
Constants
Usedin the Ephemerides
where r;, i;, and i; are the solar-system barJrcentric position, r'elocitl', and atrcelcr-
Associated with the ephemerides is the set of ast,ronornical constants used in the
ation vectors of bod1, i;
creation of the ephemeri<les.Many of thesc values do nol, agree cxactly with the 1t1= Gmi, ll,.hcr-cG is the gravitationa.l consiant arrd ,rj is
the mass of body = rj - ri ; B is the PPN parameter rneasuring the nonlirrear-
standard IAU 1976 set of constants. I'he dificrencc has treen necessary in order to i; rij
Ity in superpositiou of gravilv;
provide a best fit of the epherneridesto the obselvational data. Thcse constants are 1 is the PPN parameter rneasurirrg space curvatule
produced by unit rest rnass (in tliis intcgration, as in general relativit]-. ij = 1 = 1.1;
associateddirectly with the ephemerides and are considcred to be an integral part
vi = li;]; and c is thc vd<xrity of liglrt.
of them. They are presented in Section 5..12.
 282 EXPLANATOBYSUPPLEMENT E P H E [ ' 4 E R I D EOSF T H E S U N ,M O O N .A N D P L A N E T S 283
5 / ORBITAL
In the last term on the right-hand side of Equation 5.211 1. quartities em- ze
ploying the index rz refer to the asteroids. The positions of thc asteroids are lot
integrated but are obtained from polynornials representing heliocentric Keplcrian
ellipses. The polynomials give good representations for perturbations on the plalets
at the present tirne. At times in the distant past the polynomials drifted frorn the
real asteroid orbits, but the perturbations were snaller (Williams, 1984) for Nlars
than any ancient optical-measurement accuracy.
The quantity it appearing in two terms on the right-hand side ofEquation 5.211 1
denotes the barycentric acceleration of each body I due to Newtonian effecis of the
remaining bodies and the astcroids.

5.212 Figure Effects Long-term accuracy of the integrated lunar orbit requircs
the inclusion of the figures of the Earth and Moon in the mathematical model. In
D E 2 U 0l h e g r a v i t a t i o n a e
l f f e r - rtsl u c t o f i g u r e si n c l u d e :
Figure5.212.1
accelerations
system,in whichfigure-induced
The€4( coordinate arecalculated
( 1) The force of attraction between the zonal harmonics (through fourth degree)
of the Earth and the point-mass Moon and Sun;
(2) The force of attraction between the zonal harmonics (through fourth degree)
and the second- and third-degree tesseral harrnonics of the Nloon and thc
point-mass Earth and Sun. which the harmonics are expressed; and I is the east longitude of thr: point mass
in the same bodv-fixed coordinate systcln.
The primes denote differentiation l'ith respect to the argument sin {'. Thc ac-
The mutual interaction between the figures of the Earth and Moon is ignorerl.
celerations are transformed into the solar system barycentric cartesian system by
The contribution to the inertial acceleration of an extenderi body arising frorn
application of appropriatc rotation matrices: first by a rotation frorn the {4( systcn
the interaction of its own figure with an external point mass is expressed in the
to the selenographic systerl, followed by the application of the inversc libration-
{4( coordinate system. where the {-axis is directed outward fiom t}re extenclc<l
angle matrix.
body to the point mass; the 4-axis is directed east (lying in the selenographic ry-
The interaction bett-ccn the figurc of an extended body and a poilt rnass also
plane, perpendicularto the {-axis); and the ( axis is directed north, completing lhe
induces an inertial accclcration of thc point mass. If iip11;clenotes the accelerafion
right-hand system (seeFigure 5.212.1).Iu that system (NIoyer, 1971),
given in Equation 5.212 1 when exprcssedin solar-systembarycentric cr'ordilr;rtes.

)&{sin@) then the correspondiug acceleration ipxl of the point mass is

= ' l ;i,, (.1'
fL;la l +{ ' ' ' 1'l''+ l (s.212-1)
.. lrFic. (5.212-2)
rtNI = - rFIG'
l-co.'orj,rsinorl PPNI
[ - t r + | t P f , r s ionr l G , ,c o sr r \ + S,. sinrz)JI q
. t (" ) f I n s e c p P l l s i n p l [C , , , s i n r a+S,,,cosm^l
\ where pp16 aDd ppv are the gravitation constant 6 times the massesof the exterrcled
I l,
;7''' ; I c o . ' o e ; i r s i n , a r l C , , , c o sS,,, + n.\l I .,
r r \sin body and point mass, respectively.

where p is the gravitational coustant G times the mass of the poiut bodyi r is th(l
5.213 Earth Tides The tidr:s raised by the Nloort on the llarth appear as a bulge
center-of-urass separation betwecn the two bodies; l?tJ and n2 arc the maxinull
leading the Earth NIoon line by a phnsc angle 6. l'he rcsulting geocentric at:celer-
degreesof the zonal and tesseral expansions, respectively; P,(sin /) is the Lcgerrdrc
a t i o no l t h e N l o n r ir. g i r l r r h . vr h p p x p : r r r : i o n
polyrromial of degree n; Pi'(sin d) is the associated Legendre function of degrce n and
order m; J, is the zonal harmonic for the extendccl bodv; C,., Sn,, are thc tesseral
harmonics for thc extended body; a is the equatorial radius of the extended bod-vi
@is the latitude of the poirrt rnass relative to the body-fixed coordinate system in
.. 3k214, .
',,u,==..- l'-t/
(, p^\
(*)'l;:*1 , (5.213-1)
 284 EXPLANATORYSUPPLEMENT EPHEMERIOESOF THE SUN I\''IOONAND PLANETS 285
5 / OBBTTAL
where 12 is the potential Love number of the Earth; a" the radius of the Earth: Table5.214.1
ren1is the geocentric lunar distancel and xr ), and z are the geocentric cartesian LunarLibrationAnglesand Rates
coordinates of the Moon expressed in the true-of-date system. (The rotation to
Epoch Epoch
mean equator and equinox of J2000.0 is performed by the application of an inverse JD 2440400.5 JD 2,144400.5
nutation matrix followed by an inverse precession matrix. The nutation matrix is
evaluated using only the leading (18.6-yr) term.) In addition, p. is the gravitation d 5. 13326253997944214O03 -3.88899107819826403 002
4i 1.00533929643038618 oo4 - 1.27558140671
247413 004
constant times the mass of the Moon; p" is the gravitation constant times the mass 001 4.30912863340948250 001
0 3.82365872861401043
of the Earth. e - 005 5.09287249525085733
1.44167407096036579 005
The inertial accelerations follow frorn the conservation of the celler or mass. 1, + 000 1. 33180551
1.294226a0A46346332 + 000
146558290
$ - oo4 9.3330009498581
1.19916273482678420 5842 005
For further discussion, see Williams et al. (1,978).

5.214 Lunar Librations It is necessary to form a matrix transforming between

the inertial coordinate system of the integration and the selenographic system. The
Euler angle definitions and their differential equations were taken from Goldstein
(1950). d is the angle along the Earth's fixed equator from the fixed equinox to The difierential equations for the angular velocity come fron Errler's equations'
the line of nodes with the Moon's true equator; d is the inclination of the Nloon,s
tt-Bt N,
true equator to the Earth's fixed equator; and I is the angle along the Nloon,s ;, =; + :
equator from the line of nodes to the reference meridian of the selenographic system. hr:1,,",:
Following customary procedures, we define ar= pLe.e\+; (5.214-5)

N.
C_A J.=. trr\r!+
tatll-1\ C.

and Finally, the difierential equations lor the three Euler angles are
B_A
t\ )14-)\
L, ./ dr sind + dr)cos 4, + 0(4,- 4)cos0)
o=
where A, B, and C are the three principal moments of inertia of the Nloon and
*rt
C > B > A. The relationship between p1, .tt, Jz, Czz and C I me2 (where ra is the 6i=;acosl - orsinry'- 4t sin9 (5.214-6)
mass of the Moon and a is the lunar radius) is described by Ferrari el ol. (1980). li = it, - 4cos0+ d0 sine.
Let Fp16 be the force on the Moon due to the gravitational interaction of
the lunar figure and an external point-mass Earth or Sun. Fprc is derived from The second-order difierential equations for the Euler angles of a rigid body Moon
Equation 5.212-1 . Then the torque N on the Moon is given by a,reintegrated numericallyr considering torques induced by ihe Earth and Sun
The libration states at the epoch 2440400.5 axe given in Table 5 214 1 From
N=rxFp16, 15 ? I4-l) the nature of its definition the angle ty' (denoting the angular displacement of the
selenographicmeridian from the Earth's equator) grows rapidly with tirne' changirtg
where r is the vector from the lunar centel of mass to the point mass. In the by 2r radians per lunar revolution. To limit the magnitude of the numbers carried
selenographic principal-axis system the equations for the angular velocity vector o by the numerical integrator, a linear polynomial was removed from the initial con-
are related to the Euler angles through ditions for rl presented to the integrator. In all calculations the actual values of iy'
and tl used are:
c,.,,= /sin d sin{ + dcos ry'
ar=Qsin0cost!-0sint! (s.2144) {(t) = Vi Q)+ 4)o+ tn,
u,= $cos0 + rj-,. 4,(t)=,l't"t@+tr, \5.214-7)
286 SUPPLEI\,IENT
EXPLANATORY OF THESUN,MOON,ANOPLANETS
EPHEMERIDES 287
where ?y'i"r(t), ry';"6(r) denote the angle and rate provided by the integrator: Forre Model Eyaluation The calculation and arrangement of accelerations
integration step is as follows:
lto = o'
(1) The integrator subroutineprovidesnew states(positionsand velocities)for
h = 0.2299715021898189
rad d-r (5.21,1-8)
the nine pla.nets,the Moon, and the libration angles.
(2) The asteroidstates are er.aluatedfrom fixed polynomials.
Discussions of numerically integrated physical librations are given by Williams
iS) Equationsfor the relativistic massesa,reevaluatedfor the planets,Moon,
et al. (7973), Cappallo (1980), and Cappallo el al. (198r). Nonrigid-body effects asteroids,aud Sun, usingcurrent statesfor all bodiesexceptthe Sur. (The
a.re described by Yoder (1979) and Cappallo (f980). The numerically integrated barycentric state of the Sun calculated at the end of the previous step is
physical librations used to fit the lunar laser data came hom the program described retained for this evaluatiou.)
by Williams et al. (1973) rather than from that described here, since higher arccuracy j (4) The present approximate state of the Sun is obtained from the constraint
is needed for the data reduction than for the integration of the lunal orbit (Equation5.215-1).
(5) Equations5.215-2are er,aluatedagain,using this new estimateof the solar
slaf,e.
5.215 Solar-SystemBarycenter In the z-body metric, all dynamical quantities are
(6) Equation 5.215-1 is evaluated a second time to provide the current state of
expressed with respect to a center of mass whose definition is modified from the
the Sun.
usual Newtonian formulation. The solar-system barycenter is given by Estabrook (7) Equations 5.211-1 are eraluated to obtain the accelerations of the nine
(1971, private communication) as planets and the Moon.
(8) It has proved numerically more suitable to integrate the luna.r ephemeris
t i
Piri = 0, (s.21s
r) . relative to the Earth rather than to the solar-system barycenter, The solar-
system barycentric Earth and Moon states a.re replaced by the quantities
r.,' and r3, given by
where
fen= fn - fe (s.22t-r)
ui=u,{r**"?-#n?;} (5.215-2)

pef e + pnrn
In Equation 5.215-2 pi is defined as before a.nd vi is the barycentric speed of body fB = -'---- ' (s.221-2)
lte + ltn
t:

rij=fi-ril (s.215-3) where the subscripts e and m denote the Ea.rth and Moou, respectively.
Note that r"- is the difierence of solar-system barycentric vectors and is
During the process of numerical integration only the equations of motion for the distinguished from a geocentric vector by the relativistic transformation
Moon and planets were actually evaluated and integrated. The barycentric position from the ba,rycenter to geocenter. (The vector ra can be interpreted as
and velocity of the Sun were obtained from Equation 5.215-1. It should be noted representing the coordinates of the Newtonian Earth-Moon barycenter rel-
that each of Equations 5.215-1 and 5.215 2 depends on the other, requiring a"n ative to the solar-system barycenter. It has no physical significance and
does not appea,r in force calculations; it is solely a vehicle for improving
iteration during the evaluation of the solar position and velocity.
the numerical behavior of l,he differential equations.)
(9) The equations for the libration angle accelerations are evaluated.
5.22 NumericalIntegration
Estimated Integration Error The method of error control used in the in-
The numerical integration of Equations 5.211-1 and 5.214-6 was carried out using puts a limit on the absolute value of the estimated error in velocity of
a variable-stepsize, va.riable-order Adams method (Krogh, 1972). The maximum equation at the end of every iutegration step. Step size and integration orders
allowable order of auy of the 33 equations is 14; the actual order at any instant is adjusted on the basis of estimated error. The limits selected for DE200 were
determined by a specified error bound ard by the behavior of backward difierences r? lud-l in each compoqent of the equations of motion for the planets a.nd
of accelerations. and 2 x 10-15 ra.d d-l for each cemponent of the libration equations.
288 EXpLANAToRy
suppLEMENr 5 / ORBITAL
OF THE SUN, [4OON,AND PLANETS
EPHEIVIERIDES 289
5.223 Adopted Constants The integration requires inputting the numerical val- (5) DE119/ was adjusted onto its own dynamical equinox of J2000, thereby
ues of a number of parameters. Some of these parameters, such as the initial po- producing DE200:
sitions and velocities of the planets and Moon, result from the least-squares fits
= R.(0100073)ri'r.
1266 (5.2314)
and are different in each fit. Other parameters, such as some of the massesand the
Earth's zonal harmonics, come from outside sources and are only rarely changecl
for these present purposes. Some parameters, such as the mass of the Earth Nloorr 5.232 Astronomical Constants The preceding analyses and comparisons rer''eala
system, can be derived from the data, but for convenience are changed only when number of features that merit further discussion.
statistically significant improvements can be made over the standard values. The origin of the reference system of DE118 should be approximately that of
The lunar secular acceleration is an important parameter that has been subject the FK4, since the ephemeris has been fit ta data that included transit observa-
to uncertainty and confusion. Lunar Iaser data now provide an estirnate of the lunal tions of the U.S. Naval Observatory that have been refererrcedto the FK4. As such,
secular acceleration (Calame and Nlulholland, 1978; Williams et al., 1928; Dickey the determinatior of Ei,s(1950) may be interpreted as a deterDrination of the FK4
et al., 1982; and Dickey and Williams, 1982). Input can be a Love numbcr and a equinox. The value of 015316 lound in Step 1 of Section 5.231 agrees remarkablv
phase shift that are converted to an acceleration (Chapront-Touze and Chapront, closely with that of Fricke (1982) for the FK4: (8ru'4(1950)= 0.]525). The smallness
1988). The value used in DE118 (DE200) was h = 23i9 t ti3 / cy2. This may be of the difference of 010066must be foftuiious, however, for the expected uncertain-
compared with the more recently estimated value of -2419 ! 1'.'O I cy2 (Newhall ct ties of the two determinations are nearly an order of magnitude greater.
ol.,1989). One may calculate the mean obliquity, €, at a given epoch dircctiy from the
ephemerides,using an analysis similar to the one used for computing J2. This deter-
mination should be quite accurate (t 0l'01) since the data set used in the adjustment
5.23 Orientation
of Ephemerides ofthe ephemeridesincluded 10 years ofLunar Laser-Ranging As such, the obliquity'
especially at the nlean epoch of the laser-ranging data (1975), is well represented
5.231 Procedure Surnmary The J2000-basedephemerides,DE200/LE200, u,et: crc- by the ephemerides. The analysis gires the result
ated directly from the 81950-basedephernerides,DEl18/LE62, by a simple rcori-
entation of the reference frame. The orientation is presentcd by Standish (19g2a):
the essential features are reproduced here. This procedure rvas done in the follon,ilg = 23' 26'2l';4 | 19,
ei66("/2000) (s.232-r)
nve sleps:

where the rotating (Standish. 1981) senseol the definition has been rrsed.
(1) The 1950 dynamical equinox of DEl18 rvasdetermined to bc
This latter number is then to be comDared with IAU (1976) value of

E i , 3 t1 9 5 0=r Q ' { 1 9 5 0=) + 0 : ' 5tJe , , r)

(5.231
et,L'= 23"26'21')448, ( 5 . 2 3 22 )
where the superscript signifies that the ,,rotating', dcfinition of the equirrox
(Standish, 1981) was used.
giving
(2) DE118 was rotated onto its own dynamical equinox of 1950, procluciDg
DE119:
Er,ru €ior= 010361. (s.232,3)
r r r s= R . ( - 0 1 ' 5 3 1 6 ) r r r s . (s.231-2)

(3) DEl19 was precessedto the epoch J2000, using the 3 x 3 matrix p givcn Bretagnonand Chapront (1982)havernadean indcpendenta alysisolDE200|LE2O\.
by Lieske (1979): using their analytical planetary ard lunar theories(Bretagnon,1980,1981;Chapront-
Touz6, 1980; Chapront-Touz6 and Chapront, 1980). The analysis, coveriug 100
rlle = Pr11e. (.s.231-3)
years, produces a value for the dynamical equinox (rotating scnse) of 0:'00068.This
(4) The dvnamical equinox of DE119, at J2000 was determined to be E!,n = independent check verifics that DE200/ LE200 is ildeed referred to its o$'n dynam-
-0100073. ical equinox of J2000.0to r.ithin an accuracy of t 01001.
 29O

5,3 OBSERVATIONAL
DATAFITBYTHEPLANETARY
ANDLUNAR
EXpLANAToRy
suppLEMENr w OFTHESUN'[/OON'ANDPLANETS
EPHEMERIDES
5 I oRBTTAL
T a b l e5 . 3 1 1 . 1
291

EPHEMERIDES fromthe U.S.NavalObservatory

TransitCircleObservations
thathavebeenusedin theJPLEphemerides, DEl lS
The planetary and lunar ephemerides represent the results of numerical integra-
Catalog Timespan Telescope vovPa Numberot
tion6 of the equations of motion that describe the gravitational physics of the solar Observalions
system. It is assumed that the equations of motion accurately describe the physics,
at least to the presently observable accuracy. Also, the cornputer program that in- w(10) 1911-19'18 6', XI 2436
w(20) 19'r3-r925 9', x l 3381
tegrates these equations has been demonstrated to be sufficientlv accurate. It is w(25) 1925-193s 6', XVI/I 6911
the least-squaresadjustment process, of course, that produccs the initial conditions w(40) 1935-1944 9', 4547
and the set of astronomical constants that are used by the numerical integration w(50) 1935-1941 6', XVYI 3777
w2(s0) 1941-1949 6' XVVIII 3444
computer program. xtvl
w3(50) |949-19s6 6" 3678
The most critical feature of the construction of modern-day lunar and plan- w4(50) | 9s6-1962 6' xlxtl 4051
etary ephemerides is the accuracy and variety of the observational data to r,i'hiclr w5(s0) 1963-1971 xx t/!tl 5 8 t1
(Cir.) 1975-1977 6" 1543
the ephemerides are fit. This section (taken from Standish, 1990) describes thr:
observational data used in the creation of DE200/LE200. totats 39579
The observational data include nlcridian transits, satellite astrometry, radar
Not€:The Vol/Partretersto the second Seriesol the L,sNo Publicalions
ranging to a planet's surface, ranging to various spacecraft, and laser ranging t<r
the lunar reflectors. The reduction formulas are given for each data type. Tht'
accuracies of each type are given alorrg with the numbers of observations and thr:
time spans covered by the difierent sets of data.

Data since the end of the \\15(50) iu 1971 have been provicled in urachine-
5.31 ODtical
Data readable format by the USNO. These must be considered as provisional. since the
There have been three types of optical data used in the Jet Propulsion Laboratorl. transit circle was refurbished durins the interval 1972 197.1.
(JPL) ephemeridcsup to and including DEl18. These are the transit observations
from the U.S. Naval Observatory (USNO), some astrometric platc data of Saturt
5.3111 Basic reduction ofthe transit obs€ryations Transit observations are differ-
taken at the Universitl' of Virginia's Leander N{cCormick Observatorv, and a set of
ential in nature the planetary observatious undergo the sante prcrcessitrgas fhose
normal points of Neptune and Pluto provided bv the USNO.
of the observed stars, both being relatcd to the standard catalog of thc epoch The
observations are prrblished as geocentric apparcnt right ascerlsionsiuxl declinations'
5.311 tansit Circle Data The onlv transit data used in thc JPL ephemeridesup
taken at the tirne of meridian passage. For cornparisoni then, ottt: ol)tains a com-
to artd including DEl18 were the observationsfrom the U.S. Naval Observatorv puted position from the ephemerides by iteratirrg to find the tinrc at rvhich the local
(USNO) taken with thc six-inch and uirre-inchmeridian circles. Frrrtherrnore.onll
apparent hour anglc of the planet is zelo. The formulation for conlputing apparent
those sincc 1911have been used, thc date signifying the introdrrction of the impcr- places has been esseutially identical to that described in Chapter ;.
sonal micromcter. (SeoTable 5,311.1.)
The observations are recordcd in thc publications of the USNO. sccond scrics.
The data listcd in thesc publications havc bccn reduced to thc \\rtrshington catalog of 5.3112 Phase corrections A phase-eflect colrection is determirted bv the USNO
the concurrent epoch- Table 5.311.1 prcsents rcferencesfor thc \!'ashington catalogs and applied to the observations before publication. This correctiort is llased on the
in which the observationsare givetr. theoretical cflect that should be expected due to the geomctric situation. Here,
All the trarrsit observations for catalogs 1\r(25) through W4(50) were trans- empirical formulas have been used to fit and remove further phase effects fiom the
formed to the refcrence systern of thc FK,l using the formulas of Schlvan (1977). transit-circle observations. It seems that changes result nainly frorn "irradiation
For catalog W(10) the table published irr thc secondseriesof the U.S. Naval Ob- efiects" whereby an observer tends to mcasure a bright illuminated Iirnb differently
servatory publications was used. rrom a darker terminator.
292 EXPLANAToRY
suPPLEt\rENT l* 5 / OFBITAL
EPHEMERIDESOF THE SUN, |\.{OON.AND PLANETS 293
For the center of light measureme[ts. for both \{ercury and Venus, the USNO The mean motions of the inner planets are determined pri-
SSll4 C^t^logdrift
couection was temoved from the observations, and, instead, the following forrnula the least-squares fit. Therefore, anv
marily by the strength of the ranging data in
was applied: between these mean motions and those determined from the optical
inconsistency
A(t s sinO data will appear as secular trends in plots of the optical residuals. These drifts are
l c u + C t I + C 2 P+ Q i 3 l , (5.3112-l)
A6 coso due pa,rtly to an incorrect value of precession and to an equinox motion in the FK4
referencesystem, which have been accounted for in two ways.
where 1is the phase angle expressedin units of 90 degrees(1 = i/90); O is the position
angle of the midpoint of the illuminated edge; and I is the phase angle between the 5J12 Corrections to prec€ssion and equinox drift The secular-like drifts in the
Sun and the Earth, subtended at the observed body. The coefficients, Ce through optical residuals have been modeled by the standard formulas,
C3, are solved for in the ephemeris solutions. The resulting cubic polynomial agrces
quite well with that found by Lindegren (1977). Ao = (A/< + Az sin a tan 6)156
For an illuminated limb measurcmcnt, for both Mercury and Venus, thc USNO {f = (An cosa)25e, (5.312-t)
correction was retained, and the following formula was applied in addition:
where ?59is the time in centuries past 1950. These corrections rvere subtracted frorn
Aa sin @
[Lo+hl + Lzi2+ LtI3l. (5.3rr2-2) the observed values of all the transit obscrvations of the planets and the Sun.
A6 cos@ If one assumes that these parameters comc fiom precession error aud equinox
drift exclusively, then the following relations apply:
Here. the coefrcients are in units of seconds of arc. For measurements of four linbs.
the USNO correction was retained and the following formula was applied in addi- (.s.3r22)
Aft= -E+Apcose - A/ and An=Apsine,
tion:
sin @
Bp sin2i, (5,3r 12-3) where ii is the equinox rnotion; Ap is the correction to lunisolar precession; A/ is
Ad cosO
the correction to planetary precession; and e is the value of the Earth's obliquity
For DE118, the value of An was set to +01'438/cy,corresponding to Lp =
where,t = 4, 5,..., 8 for Mars, Earth,..., Neptune. This empirical formula rnal'br:
+lll0/cy, Fricke's(1971) determination of the correctionto precession.At was lixed
compared with that of Chollet (1984), the form of which has been derived frorn
to -0i266/cy, corresponding to Fricke's (1982) determination of the FK4 equinox
physical considerations.
motion, E = +li275lcy. The value of A/< should have been -0'2371cy, since the
The forms of these preceding "phase correction" formulas were all choselr
cortection for planetary preccssion of Ll = -0:029lcy (see Lieske et al., L977) was
s t r i c tl y f r o m e m p i r i c a lc o n s i d e r a r i o n s .
inadvertentlv omitted.
With A/< and An constrained to the values determined by Fricke, the corrections
5.3113 Day corrections For each \\rashington catalog, corrections are applied tcr to the computed residuals of the transit data are identical to those that \i'ould
the obsenations obtained during daylight hours; rramel5r.observations of thc Srur. have been found if the whole procedure (observations and reductions) had been
N{ercury, and Vemrs. Typically, t}rcsc are given in tables as ftrnctions of the object s performed in the J2000.0 (lAU 1976) referencc systern.
declination and of the tirnc of day. Day corrections have not been applied. howeler',
5.3121 Catalog offsets With A& flxed for the solution to DE118, there was the
to the prcliminary observations given in the USNO circul:r,rs. Therefore. lbl thc
possibility of a residual secular trend in the optical lesiduals. It was dccided to
1975 1977 data the following day corrections were solved for and removeclfrorn thc'
introduce constant ofisets in both right ascensionand in declination for each of the
observations of the Sun, N{crcury, and Venus:
Washington catalogs. The values for these are included in Table 5.42.1.

A a = A r + A z s i n6 (s.3113-1) 5.313 Astrometric Plate Data for Saturn Over the years 1973 1979, the Univer-
A6=Dr+Dusin6, (5.3| l3-2) sity of Virginia, under contract from JPL, provided measurements of photographic
plates taken of the planet Saturn and its satellites (Ianna, 1974, 1980). The mea-
where 6 is the declination of the body. A1. A2. D1, arld D2 are ernpiricalquantiti('s. surements of the satellites and the background stars were provided to JPL arrd
 OFTHESUN,MOON,ANDPLANETS
EPHEMERIDES 295
294 EXPLANATORY
SUPPLEMENT

were reducd by Sta.ndish, using the satellite ephemerides of Null (1978) a.nd the
Planetatr-rd
Zodiacal Catalog of the USNO (Schmidt et al., 1980). The "observed" positions
of the satellites from the reductions were then compared to "computed" positions
derived ftom the satellite and planetary ephemerides. Since the errors of the satel-
lite ephemerides were expected to be small in comparison to those of the planeta.ry
ephemerides, and since the satellites errors enter in a quasi random way when many
observations are used, it was assumed that the differences (O - C's) could be at-
tributed to the errors in the planetary ephemerides. Accordingly, the difierences
were used to adjust only the ephemeris of Saturn.

Ear$ at t Ewthatt-ra-1
5314 Normal Points for Neptune and Pluto As mentioned before, the main op-
tical data used by JPL a.re the transit observations from the USNO' beginning in
1911. Clearly, for the outermost pla.netsthis is not enough data to determine a full
orbit. Consequently the USNO transmitted a set of normal points that gave the
differencesbetween one of their experimental ephemerides and an intermedia.ry JPL rAO) (t - rd)
rAQ-rd-r)
ephemeris (DE114). The optically based USNO ephemeris gave what was believed
to be reasonable fits to the observations of Neptune and Pluto. These data were
then used to adjust the solution for DE118. The data were transmitted in the form
of corrections to longitude and latitude, one pair of points every 400 days. The time
span for Neptune was 1846-1974; for Pluto, 7974-1974. Solar-systembarycenter

Flgutt 5.321.1
5.32 Radar-RangingData Diagramof geometryfor planetaryradarranging

5321 Computation ofRador-Ranging Measurements Ra.dar-ranging measurements

are actual timings of the round-trip light-time of the electromagnetic signal from
respectively. The two formulas a.re solved iteratively, first for z then
the time that it leaves the transmitter until the time that it arrives at the receiver
(Figure 5.321.1). The timing is done by an atomic clock at the observing station.
For an observation received at the time t (expressed in TDB time units), the location of the antenna is computed in a straightforward manner using a
rould-trip light-time is given by the difereuce UTC(I) - UTC(I - rd - ru), wherc ephemeris and the proper formulas of precession, nutation, timiug, and
motion with which one orients the Ea.rth into the refereqce fta'ue of the
The location of the point of reflection was computed by assuming the
r, = lrp(r - Td) - r t - ra - r")l l c + Ar,[rel] + At'[cor] + A?],ltropol (5.321-l)
of Mercury and Venus to be spherical. Radar ranging to Mars v/as not
in the solution for DEl18.
and

q = lr sQ) - rn(r - za)l / c + Atrlrell + Atdlcorl + Azzftropol, (5-321-2) Time delaysfor relativlty' the solar corona, and the troposphere The time-
due to relativity, given by Shapiro (1964), is obtained by integrating along
Tu a!.d 16 are the light-times (in TDB units) of the upleg and the downleg, respec- path over the value of the potential. For each leg of the signal path, the
tively; ra is the solar-system barycentric position of the antenna on the Eaxth's sur- is given by the formulas
face; rp is solar-system barycentric position of the reflection point on the planet's ( l + 1 \ G M , l e + p + o- l
surfacel c is the velocity of light; and the three Az's are the correctioDs to the light- Aztrell = :-----::- llr | L (5.32rr-r)
c' lc+p-ql
times due to relativity, the electrou content of the solar corona, and the Ea^rth's

$Sd+","
 I'|il[
296 EXPLANATOBY
SUPPLE[,,IENT S F T H E S U N ,M O O N ,A N D P L A N E T S
E P H E M E R I D EO
5 / ORBITAL
297
where I is the PPN parameter of general relativity and where e, p, and 4 are the Table5.322.1
heliocentric distance of the Earth, the heliocentric distance of the planet, and the Radar-BangingObservations in DE1l8
UsedDirecily
geocentric distance of the planet, respectively. These distances are evaluated at
TimesDan Anlenna Number
t-rdfot the planet, at t -r,1 - zu for the Earth during the upleg, and at lforthe ol Observatlons
Earth during the downleg.
The delay from the solar corona (see Muhleman and Anderson, 1981) is ob- Mercury 1966-1971 Arecibo 106
1966-1971 Haystack 217
tained by integrating along the signal path from point P1 to point P2 over the 1971 Gold 13/14 I
density of ionized electrons.N. 1cm l.;. 1972-1974 Gold 14 30

MercuryTotals 362
Li \ fP,
Arfcor]= r.o'. (s.321r-2)
Arecibo 248
# Jr, 1966-1970
1966-1971 Haystack 219
1964-1967 Millstone 101
where c is the speed of light (cm/sec), / is the frequency (Hz), and s is the linerr 1964-1970 Gold 13 294
distance (cm). The density is given by 1970-1971 Gold 13/14 14
1971-1973 Gold 14 44
1973-1977 Gold 14 25
A oUt yF s'rn2A + b'zcos')
B
(s.3211-3) VenusTotals 945
f

where r is the heliocentric distance expressedin units of the solar radius and d is thc
solar latitude. The values for the constants, A, c, and b are included in Table 5.42 1
The delay from the Earth's troposphere is discussed by Chao (1970). For cadr
IeE.it is

Arltropo] = 7 nsec/ (cosz + 0. 0014/ (0.045+ cotz)), rS 1tI lJ\ 5.33 SpacecraftRangePoints

There were three sets of spacecraft-ranging data used in thc ephemeris solutio[s
wherez is the zenith distanceat the antenna.
for DE118. They are (1) normal points from the Pioneer Missions to Jupiter, (2)
normal points to the Mariner 9 orbiter of Mars, and (3) actual range measurements
5.322 Radar-Ranging Data The strength of a radar echo from a planet's surface to the Viking Landers on the surface of Mars.
varies as the inverse fourth power of the distance. Up to the present time. higlr Normal points represent modified distance measuremelts The original round-
quality planetary radar-ranging observations have been conlined to times whcn the trip range and Doppler measurements have been reduced using tbe JPL Orbit
distances are less than about 1AU. At these distances, the precision of the ranging Determination Program (Nloyer, 1971). This reduction is an adjustrnent for various
paxameters, including the spacecraft orbit, the planct's mass and gravity field, and
observations has been seen to be about 100 meters. However, for ephemeris mea-
so forth. As such, the resultant range residuals represent derived corrt'ctions to the
surements, variations in the topography of the planet's surface introduce variatiorrs
nominal planetary ephemeris used in the reduction. These residuals are then added
into the observations which tend to dominate the uncertainties. For Nlercury and
to the geometric (instantaneous Earth planet) ralge in order to give a pseudo
Venus these amount to about 1 km.
"observed', range point.
The radar-ranging measurements used in the JPL ephemerides have conre
from five antennas. These are located in Arecibo, Puerto Rico; Tyngsboro, N'IA
(Haystack); Westford, NIA (Millstone); and Goldstone, CA (DSS 13 and DSS 14)'
5'331 Pioneer Normal Range Points to Jupiter There was one normal range point
In addition, some of the measurements at Goldstone were taken in the bistatic
from each of the two Pioneer encounters of Jupiter. The rnajor uncertainty for each
mode: transmitting with DSS 14 and receiving with DSS 13. Table 5 322 1 presents
point comes from the uncertainties in the determination of the spacecraft's orbit
a list of the measurements which have been used directly in DE118.
 298 EpLANAToRY SUPPLEMENT OFTHESUN.MOON.ANDPLANETS
EPHEMERIDES 299
Tablo5.332.1 of the solar corona, the calibration of the inherent time-delays ir_
Marinor9 OrbiterNormalRangoPointsto Mars station antennas (done before and after each pass), and the calibration

Delea Numberof transpondersof the landersthemselves(done beforelaunch).

(,lsec) Ob8ervatlong range data were reduced using the formulas given in Section 5.321 with
point on the planet's surfacebeing representedby the location of
244'1272-2441361 0.25 77 on Mars. For this, one needs the Martian coordinates of the landers as
2441359-2441540 0.29 81
2441541-2441555 o.7a 47 a set of angles used to express the orientation of Ma.rs within the ephemeris
2441577-2441W2 fta.me, The position of the lander, expressed in the frame of the ephemeris,
Total by

NoleiTheiours€tsot pointsaregrouped
accodlngto th€
poximiiyintim€torhesolarconjunction
of Mars(JD2,141568). r = &(-€)&(-o)&(-4&(-o)R (-rr&(-y')ro, (5.333-l)

e is the obliquity of the ecliptic; O and 1 are the longitude and latitude of the
with respect to the center of mass of the planet. The observations, transmitted by
Martian orbit upon the ecliptic; Qn and Io a.re the mean node and inclination
Null (1976), are as follows:
Ma.rtian equator upon the mean orbit; V is the Iongitude of the Martian
meridian measured along the equator from the intersection of the orbit; and
Pioneer 10: 1973 Dec 04 0r'ET (2442020.50) / 2754747323!40 p,sec.
the Ma.rs-fixed coordinates of the lander, and where
Pioneer 11: 1974 Dec 03 0h ET (2442384.50) / 2439811990110psec.

These values represent one-way geometric (instantaneous) distances from the center a'o= Qo - Arlt, / = In+ Ae and V' = V + N! cosI'0, (5.333-2)
of mass of the Earth to the barycenter of the Jovian planetaxy system.
Arl and Ae exDress the nutation of Mars. comDuted ftom the fornulation
5.332 Mariner 9 Normal Range Points to Mars During its lifetime, the onboard et oI. (7979). The Ma.rs-fixed coordinates of the la.nder a.re computed
range transponder of the Mariner 9 Orbiter allowed accurate range measuements the cylindrical coordinates,
to the planet Mars. These data exist in three sets according to their proximity to
the Martian solar conjunction (JD 2441568) when the 2300-MHz ranging signal r[ = [z cos), z sin), v]r. /5 ???-?\
passed within 4 solar radii of the Sun at the heliographic latitude of +79". When
near conju.nction the major uncertainties of these points are due to uncertainties in i,A summary for the Viking Lander ra.nging data is as follows:
the densities of the ionized electrons in the solar corona through which the signal
passedl away from conjunction, the orbital uucertainties dominate. As shown in Lander #1 2442980-2444054 Points: 683
Table 5.332.1, the three sets of data are weighed accordingly. The data have been Land.er#2 2443026 2443417 Points: 78.
couected for the solar corona using the formula given in Section 5.32.
values for the pararneters used in the reductions a.re included Table 5.42.1'
5333 Yiking Lander Range Data The most accurate of all planetary position data values for €, O, a.rrdI were adopted: those for Qq, I* and V, as well as the
are the two.way ranging measurements taken of the Viking Landers on the surface of of the landers, were estimated in the least-squares adjustments.
Mars. Though only a single frequency (2300 MHz) was used, the solar corona could
be calibrated by using the nearly simultaneous dual-frequency measurements of the
Viking Orbiters. Observations were made on the averageof once per week, typically LunarLaserRangeData
with about six range points per day. The residuals are seen to have a scatter of only
2 to 3 meters about the mean of the day; the means for ea.ch day, however, show tunar LaserRange(LLR) data cousistof time.of-flight measurements from
a scatter of about 6 meterc among themselves. Without orbital uncertainties being Observatory to a,rry one of four retroreflectors on the Moon and back
present, the dominating contributions to these residuals come from three sources: The retroreflectors are at the Apollo 11, 14, a.nd 15 landing sites and on
300 EXPLANATORYSUPPLE[,4ENT
5 / ORBITAL
EPHEI4ERIDESOF THE SUN, MOON, AND PLANETS 301
the Lunakhod 2 vehicle. These 2954 range points are distributed from August 1969 Table5.41.1
to June 1980. The normal points through 1973 have been published by Abbot ez The ObservationalData Used for the Adjustmentof DE 118/LE62
al. (1973), Mulholland et al. (1975), and Shelus et al. (7975). The LLR data are
Typeol Observation Number of
deposited in the National Space Science Data Center. During the least-squares fi1 Observations StandardDeviation
the ranges have been weighted according to the instrumental errors that accompany
each point. The general trend is toward improving accuracy with time. The sirnple transits
Washington 1911-1977 39579 110(Sun, l\,,ler, Ven)
0 . 5( M a r s , . . N
.,ep)
post-fit rms residual is 31 cm in orre-way range. The q,'eightedrms rcsidual is 27 cm
Saturnastrometry 1973-1979 4790 013
(Dickey et al., 1982).
Neptune,PlutonormalPoints 1825-1974 386 015
'1964-1977 10 km (belore'1967)
For the Moon, the mean orbital elements that affect its geocentric distant:e Radar-ranging 1307
will be strongly determined by the laser-ranging data. The eccentricity is implicitlr. 1.5km (after1967)
Pioneerrangesat Jupiter 1973-1974 2 3 km, 12km
determined to a few parts in 10v and the mean anomaly to a ferv milliarcseconds I rangesat N4ars
lvlariner 1971-1972 645 4G-100m
There is a strong determination of differential geocentric lunar and solar eciiptir: VikingLanderranges 1976-1980 761 7m
longitudes resulting from strong solar perturbations on the lunar distance (arnpli- Lunarlaserranges 1969-1980 2954 18cm
tudes of about 3000 km and 4000 km for the two leading terms). During the span of Total 50424
the observations the differential Iongitudes are knou''n to a few milliarcseconds. Orrt-
side the data sparr the error in thc lunar longitude is dominated by the unccrtaill.l
in the lunar tidal acceleratiorr. (Sce Williarns and Standish, 1989.)

5.41 TheObservational
Equations

Table 5.41.1 gives the number of observatio[a] equations for each differelt set of
5.4 LEAST-SQUARES ADJUSTMENT OF THE EPHEMERIDES data. Each equation was nornalized by multiplying it by the factor, I / d0, where
d0 is the a priori standard deviation of a single observation of a particular sct of
data. Previous experience in working with the rarious sets has Ied to a knou"ledge
The least-squares adjustmeut for DE118/LE62 involved 175 explicit unknor.n pa-
of the individual accuracies of each tvpe of obsenation. The values of the a priori
rameters and 50424 observational equatious. The adjustment was done in trvo slep-s.
sta.ndard deviations, listed in Table 5.41.1, $.ere chosen to be approximately cqual
Using a previousepheneris (DE111/LE55) as thc rrorninalephemeris,a preliminalr'
to the rms post-fit residuals.
solution was made involving all solution pararnctcrs that were judged to be rek:rant.
Then, the values for certain parameters were rounded off and forced into thc filal
solution for DE118/LE62 which readjusted, slightly, the rcrraining parameters. 5.42 TheSolution
Parameters
There are a number of reasons for forcing certain paramctcrs to have spec:ifJc
values: the adoption of someoneelse's, presumably more accuratc, determinatiol ol Table 5.42.1 givcs the values and the formal standard deviations of the solution
the constant (e.g.. the mass of Pluto); the rounding of values for cosmetic reasorls parametersof the least-squaresadjustment for DE11S/LE118. Th" pararnetersfor
(e.g., a truncated value for the inverse mass of Jupiter); or the need for consistr:ncr which values were adopted from outside sources are enclosed in square brackets.
n'ith outside sorrrces(e.g., the IAU values of A/r and At). Parameters that were determined from the preliminary solution, rounded, and then
The creation of DE118/LE62 *'as done partially with the intent that tllis forced into the final solution are cnclosed in parentheses.The adjustment of the or-
ephemeris (when rotated onto the J2000 equator and eqrrinox) *'ould be used for t he bital parameters is inherent in the valrresof tLe initial conditions of the ephcnrerides
foundation of future national almanacs. Ideally, the ephemeris should have iucorpcr that are listed in Table 5.42.2.
rated all of the astronomicalconstants of the IAU (1976) J2000 referencesystenl. In addition to the parameters listed in the table, there were a number of pa-
However, it walsrecognized that more rnodern valucs for some constants l\'erc lec- rameters pertainilg to the platc solutions and catalog adjustment of the Saturn
essary in order to fit the more accurate obsen'ational data propcrly. In palticular' astrometrical data. These entered into the solutions implicitly; they were present in
it was not possible to use the IAU sc1,of corrstants and still produce ac<:eptable the solution just as if they had been carried along explicitly; however, their presence
solutions for thc Viking Lander ranging data or the Lunar laser-ranging clata- rs not seen directlv.
 EPHEI/|EBIDESOFTHE SUN, MOON, AND PLANETS 303
Table5.42.1 Table5.42,1,continued
ot the Constants
Valuesand FormalStandardDeviations ol the ConstantsUsedin
Valuesand FormalStandardDeviations
Usedin DE118/1E118(DE200/1E200) DE118/LEl 1I (DE200/LE2o0)
scale FactorlkntAu] 049597870.661o.0o2) = 1 . 2 2i m x 1 0 8
solarCoronaIcm-3] A
EaivMoonMassRatio [81.300584 a = 0.,14f. 02 x 106
Sun/PlanetlllassRatios Sun/Mercury = [6023600] b = 0.,14t. 03 x loo
Sun/Venus = [4O8s23.5]
SUdEM-bary = [328900.5s] MarsOdentation Y [deg] 132A.707423251
SurvMars = y [deg/dayl 350.89199047 i 0.00000024
[3098710]
SurVJupiter = (1047.350 r 0.000r) /q [d6g] 25.180841510.0013
Survsatum = (3498.0 r 0. 014) iq ldeglcy) 0.03010.006
SurVUranus = (22960r 6) oq [deS] 35 3371s55t 0.0026
Sun/N€ptune = Aq ld,eglcy) -0.118 + 0.010
[19314]
Sur/Plulo = [130000000] / idegl t1.851
, Idegr'cyl [-0.00820]
Ast€roidGMvalue Gli(Ceres) = I0.1746x lO-r2l o ldesl [4s.17193]
tAu3/daf l GM(Pallas) = lo 3200x 1o-t3l A ldeg/cyl I-0.294701
GM(V€sta) = [O.4O8Ox 10 13] 6 [deg] [23.,145789]
GM(kis) = IO.1600x 10-141 e ldegr'cy] [-0.01301]
GM(Bamberga) = t0.2600x 1o-r4l
u1 3136.515t 0.001
VikingLander [kmJ
Radii 2,{39.990 1 0. 093 z1 [km] 12U.587l:O.O24
6051.81310.054 )1 [deg] 311.8027t 0.0033
e lkml P77.374t0.002
t3397.sls] OO.IUlO O22
v2 lk.nl
Day Conections At = 01268t 01048 )2 [deg] 134.03.43
10.0034
4 = 0:378r 01160
Dr = 0:t53 r 01048 Note:Numbersin squarebracketswe€ not solvedfor but adoptedlrom other
= 1:131t 01160 sources.Numbersin palenthgs€sare roundodvaluesd€terminedlrom a
4 pr€liminary
so[]lionairdthenforcedunchanged inlothepresentfinalsolution.
Phas6 Conections C9 = -0.O3 t 0.O17
Ct = 0.39 t 0.087
Cz = 0.471 0.11
Co = -0.26 t 0.039 The StandardDeviations
h = -ol.ls r 01067
Lr = 2181t 0130
= _2:4s t 0:37 starrdard deviations listed in Table 5.42.7 ate the formal values, straight from
4
Lg = 0:73t0:13 'least-squares solutions. They do not represent realistic uncertainties, for it is
=
Bl
=
0:40 t 0:022 l'l<nown that formal uncertainties tend to be overly optimistic, sometimes by an
Bs 0:97 i 0:077
Bo = O:74tO:14 of masnitude. This is a,direct result of incorrect or incomplete rnodeling, either
Bz = 0120t 0130 the equations of motion or through the data reductions. In the present case'
Ba = 0193t 0149
of the sources of the incomDleteness a,re known. For discussions of what are
catalogDrift ak = I_0t266/cyl to be realistic uncertainties associated with the pla,netary ephemerides, see
A/r = [0:438/ cy]
et al. (1983), Standish (1985), and williams and Standish (1989)'
CalalogOffsots w10 0136s 0:303
w20 01009 ,0:033
w25 01081 0:084
w,() -01027 -0:015 .NUMERICAL OF THEEPHEMERIDES
REPRESENTATION
w150 -0:02a 0:090
w250 -0102 0:0s4
w350 -0:018 -01021 produced by the Jet Propulsion Laboratory axe obtained by nu-
w450 -0506 -01001
-0:047 -01032 integration of the equations of motion of the Moon and planets. The imme-
w550
wcfR -0:231 -01020 Ltput from the integrating program is not suited to subsequent rapid eval-
of ephemeris quantities. To provide ephemerides convenient for distribution

r$t il
 fable 5.42.2 5 / oFB|TAL
E P H E M E R I D EO
S F T H E S U N ,M O O N ,A N D P L A N E T S 305
The InltialConditionsof the Ephemeridesat JED 2440400.5in AUand Au/oay
drd use throughout the astronomical communitl', an efficient procedurc n.as ds-
generation of coefficients representing an interpolating polynomial.
veloped for the
Coordinales Velocltles
This section (taken from Newhall, 1989) describes the generation and accrtracy of
Mercury x.l 3. 57260212546963715-001 i1 3.36784520455775328-003
those coefficients'
The numerical intcgration process carries the states of the Sun, \{oon. and
y1 9. 15490552856159762002 h 2.44493428375858480-002
-8. 59810041345356578-002 4 1.29440715971588809- 002 olanets as their solar-svstem barycentric Cartesian coordinates in the inertial J2000
referenc" frame. The integration method used is a variable-step-size, variable-order
Venus x2 6. 08249 437766441072 - 00 1 XP -002
1. 09524199354744185
-3. 49132444047697970 - OO1 j,e
Adams Procedure.
v2 1.56 125069115477042 002
Z2 ,1. 95544325580217404-001 6. 32887643692262960-003

Earth-Moon xB 1. 16014917044544758-001 *6 't.

68116200395885947-002
5.51 ChebYshevPolynomials
barycenter yB -9. 26605558053098135 -001 j,a 1. 74313126183694599 003
-4. 0180626511782,1489-001
Chebyshev polynomials are the functious ofchoice for ephemr:ris interpolatiol. They
ZB 7. 55975079765192612 -004
are stable during eraluation, the,r'gir'. a near minimax rcprescntatiotr.and thel
- 1. 14688565462040833
_0014 1.44820048365775564-002 provide a readily appalent estimate of the effect of neglected ternrs on iltcrpolation
Y4 1. 32836653338579221
+000 iq 2. 37285174568730153-004 error (For a thorough treatment of these polyrromials see Rivlin. 1974).
4 -6. 06155187469280320-001 -2. 437 4475686161 1a22 004 The nth Chebyshev poll'nomial 7,,(l) is defined by the recrrrsion formula
Juprter \ -5. 38420864140637830+000 i5 1.09236745067075960 003
y5 -001
8. 31249997353602621 j,s -6. 523293903169766S9-003 T , ( t )= 2 t T ^ t ( t )- T n _ z Q ) , n=2,3,... ( 5 . 5 1 - l)
2. 25098029260032085-001 -2. 82301211072311896-003
with 70(t)= I and Ir(t) = t. The applicablerange oll for interpolation is -l < t < 1.
Saturn 4 7. 8498494267322747a+000 & -3. 21720514 122007756-O03
Any function /(l) has arr approxiurate Nth-degree expausion in Chebvsher-poll
Y6 4. 59570992672261122+O0O j,a 4. 33063208949070216- 003
1.55842916634453457+000 nomials
1. 92641681926973271
003
-_ s- _
uranus -1. 82698911379855169+001 \1 2. 2 l 544461295a71596 - O04 J t t t = ) r a n t a l t) , ( 5 . 5 12 )
Y7 1. 16273304991353263+000 h -3. 76765491663647351 003 n=0
-2.50376504345852463 001 1.65324389089726956-003
and, when differentiated,
Neptune Xg - 1. 60595043341729160+001 is 2. 64312595263412502 003
Ya -2. 39429413060150989+001j,a - 1. 50348686458462071
-003 J(tl = LanIr(t), (5.5r-3)
ZB -9. 40042772957514666+000 -6. 81268556592018307-004

Pluto X9 3. 04879969725404637+001 x, 3. 2254 176a79a 400992- O04 where the c, are chosenin a marurrt appropri;rtc for /1r1 and /1r). ln the present
t9 -8.73216536230233241001 j,e 3. 14875996554t 92878 003 caae, where.f(r) and /(tj) clcnote a coordinate and its derivative comprttr:d at dis-
Zg 8. 91 |35208725031935+000 - 1. 080t 8551253387161003
crete times
1 by the iutegrating program, the an serve to define the function f(r)
Sun xS 4. 50479585567460182-003 is -3. 52445744568339381 007 as a polynonial. The task becornesthe deterrnination of a set of a, that pro-
/S 7.732544746a9074171 -004 Ys 5. 177 6377806722212A 006 vide interpolated values suitably approximating those availablc from the original
ZS 2.68503998557327098 004 ZS 2. 22911325240040085-006 backward-differencerepresentations carried by the integrator. The following section
Moon -8.08177235835125058004 details the generation of the rz, for cphemeris botiy coordinates.
i1a 6. 01084831482911873 004
YM - 1. 99463003744199594 003 - 1. 67445469150060619
-004
YM
ZM - 1. 047262721 62086794 -O03 ZM 8. 55620810990486240-005 5.52 ChebyshevCoefficientGeneration
Nol€: Coordinat€sgven are hetiocsntnccoordinaleslor ptanets,g€oc€nlriccoordinalosJorlhe Moon, and
solar-syslembaryc€ntriccoordtnaleslor the Surl An algorithm was developed that creates Chebyshev polynornial coemcients repre-
senting the Cartesian coordinates of the epherneris bodies. The full span of, say,
306 EXPLANATORYSUPPLEI\r1ENT S F T H E S U N ,M O O N ,A N D P L A N E T S
E P H E M E R I D EO
5 7 OFBITAL 307
60 years of an epherneris file is segmented into contiguous intenals. or granules. rl1 Table5.53.1
fixed length. (The actual length of a granule dependson the body; seeTable 5.53.I Granule Length and PolynomialDegree for the 11 EphemerisBodies
for details.) For each coordinate of an ephemeris body the algorithm produces ther
Granule Polynomial
Chebyshev coefficients a, that define the interpolating polynomial valid over a giverr length(days) degree(N)
Body
granule. There are as many sets of coefficients generatcd for each coordinate as therc
a r e g r a n u l e sc o v e r i n gt h c c p h e m e r i ss p a n . Mercury 8
In the procedure, nine pairs of position and velocity lalues are obtained lcr- Venus 9
barycenter
Earth-Moon 12
each granule at equally spaced times one pair at each end point and seven in th(, Mars 32 10
interior. The output is the set of an for the polynornial that is an exact fit to t|c Jupiter 32 7
end points of the set of positions and a least-squares fit to the interior positious. Saturn 32
Uranus 32 5
The differentiated polynomial is an exact fit to the end points of the velocitl. set Neptune 32 5
and a least-squaresflt to the interior velocities. (The derivation of this proccdrrro Pluto 32 5
is given in Newhall, 1989.) Moon 4
Sun 16 10
This approach has thc advantage that interpolated positiou altd velocitl ir,rt'
continuous at the comnton end point of adjacent granulcsl it also minimizes lhc
efiect of noise that would otherwise degrade the interpolatcd velocity obtained froli
differentiation of a polyrromial based on position valucs alorre.

Investigation has sho*,n that the granule length and the poll'nomial degree N cart
5.53 Interpolation
ErrorandPolynomial
Degree be chosenso that 1a,.1/a^l = 9.1 or less for r ) N. This implies that thc rnaxi-
mum expectedintcrpolation error is about one tenth thc uragnitudeof the highest
It is essentialto have a quantitatil.e estimate of the llaximun error expectedfrorrL retained coeflicient a,,.
the interpolation pror:css rvhen the polynomials clescribcclpreviously are usctl to The accuracy critcrion for standard JPL ephemcridcsis that tLe interpola-
extract coordinate ralues at arbitrary times. (It shorrkl be noted that tho tenrr tion error for all <rrorrlinatcvalucs rnust be lesstharr 0.5 millimcters. (The DE102
"error" hetc refers to the difference betweel interltolatccl and integrator-supplicrl ephemeris covers 44(X)r-cars; in thc interests of providing n significantlv corlpressed
vahtes; it does not irrdicate the degree of accuracy to which the original intogratctl file, the interpolatiorr-<:rrol critcrion was relaxed.) Thc ruinimtrm degrce N of the
ephemerisrepresentsthc dynamical state of thc solar svstern.) interpolating polynorrrial is 3. as the requirement that thc cnd-point positiort ancl
The Chebyshcv polyuornialsprovide a convcniclt arrd reliable estimatc of irL- velocity values be rnatchc<lcxacl.ly yields four constlainl,s: t hr: 1E conrbined positiot
terpolation error. Arr arbitriuy function has the exact rcl)rcsclrtation as an infinitc and velocity values pcrrrrit a maxirnum degreeof 17. Tablc 5.53.1 lists the gralultr
Chebyshevexpansiorr length and polynoruial rl<'gr-cc for each bodl'on tho.IPL cplretnerisfilcs.

Itt|= )_a t,tttt. (5.53-l)

,=0
5.6 COMPUTATION OF OBSERVATIONAL EPHEMERIDES
The maxirnurn valtc of In(l) is unity on the intenal I l,ll, lhc domain of validitl
for interpolation. T'hcrefore, u'hcn a function is ap1>roxirnatcd by an Nth-degrc(' Computing the fundarur:rrtalcphemeridesas des<ribrrl in Sections5.2 to 5.4 pro-
polyrromial, as in Etlration 5.51 2, the maximum olror.6 itrising from the onittrrl
duces ephemerides irr a space-fixed coorclin.rtc system at a givon epoch in terms of
remainder of thc scrios h:rs the upper bouncl the barycenter of thc solar system and in barycentri< tirnt:. Frorn these ephemcr
ides, observatiolal rlrhcmcridcs can be comlnrted that errecalled apparent geocen-
tric ephemerides, with respect to a fixed epoch and.r rcfcr(rr(:e star catalog Transit
(5.s3-2)
times of the objr:t ts and .rstronomical pherromena can be computed as well
308
W, OFTHESUN'[,iOON' ANDPLANETS
EXPLANATORYSUPPLEMENT EPHEMEBIDES 309
5 / oRB|TAL
5.61 AooarentPositions Ephemeris transit is the time rvhen the right ascensionof the object is equal to
substituting the TDT time into the equation for sidereal
the le,lue determined by
The apparent position of a planet or the Moon represents the position as it would The times ol transit ate expressed in TDT and refer to the transits over the
dme,
be seen from the center of mass of the Earth at some date in the coordinate sys- meridian. Because the longitude of the ephemerisrneridian (expresscdin
ephemeris
tem defined by the Earth's true equator and equinox of date, if the Earth and its AT east of the Greenwichmeridian. the TDT of transit
time measure)is 1.002718
atmosphere wele massless,transparent, and nonrefracting. The apparent position
over the Greenwich meridian is later than that of ephemeris transit by
could then be related to observablesfor a particular instrument at some location on
the surface of the Ea,rth that is rotating and contains a refracting atmosphere. In
LT I (l - 0.99'127x rate of changeof R.A.) ( 5 . 6 3 -)l
addition to the transformation of the coordinate systems. the tine argument musl
be changed from barycentric time to Terrestrial Time. The formula and methods
for performing the transformation from barycenter to geocenter are provided i4 or aPProximatelY,
Section 3.32. Reduction to topocentric coordinates, with the effects of refractiou AT(1 + rate of changeof R.A.) (s.63-2)
included, is given in Section 3.7. The relationships between barycentric time ancl
Terrestrial Time are given in Section 2,22,
if the right ascension is expressed in the same unit as the time irttcrr-al. The UT
of Greenwich transit is therefore later than the tabulated TDT of the ephemeris
5.62 AstrometricPositions transit by the small quantitY:

Astrometric positions are useful for faint objects such as Pluto, minor planets, and
AI z rate of changeof R.A., ( 5 . 6 33 )
comets where observations are made directly with respect to the star positiorls on
some fixed epoch in the same field. Thus, if a photographic plate or a CCD image is
taken with respect to star positions that are known on a catalog for a fixed epoch. which may, at present, reach about 2" for the Moon, but is less than 013 for the
it is desirable to relate the solar-system object to those positions. An astrometric planets.
position is therefore easier to compute than an apparent place since the differential For the Sun the TDT of ephemeris transit is simply 12n minus the equation
position is not concerned with the orientation of the coordinate system. The details of time interpolated to the time ol transit; the UT of Greenwich transit differs by
concetning computation of astrometric positions are giverr in Section 3.62. 0.002738AT.
For the Moon, the TDT of upper or lower transit of the Nloon oler a local
meridian may be obtained by interpolation of the tabulated time of upper or lower
5.63 TransitEphemerides transit over the ephemeris meridian, where the first dilTerencesare about 25 hours.
The interpolation factor p is given by
For observational planning and for certain types of observations, it is desirable
to know when the body will be on the meridian. Because Terrestrial Dynamicai (5.634)
p=^+l.OO2738LT,
Time (TDT) is independent of the rotation of the Earth and because the rotation
of the Earth is not precisely predictable, it is not possible to calculate or publisL
where ) is the east longitude and the right-hand side is expressed in days. (Divide
in advance the times of transits over a fixed geographical meridian. Therefore, arr
longitude in degreesby 360 and AI in seconds by 86400). In general' second-order
auxiliary reference has been introduced. The position of the ephemeris meridian in
differencesare suflEcientto give times to a few seconds,but higher order differences
space is conceived as being where the Greenwich meridian would be, if the Earth
must be taken into account if a precision of better than 1 second is required. The
rotated uniformly with the rate implicit in the definition of Terrestrial Dynamical
UT of transit is obtained by subtracting A? from the TDT of transit, which is
Time. Thus, the ephemerismeridian is 1.002738A?'east of the actual meridian of
obtained by interpolation.
Greenwich, where A?'is the difference TDT - UT. When referred to the ephemeris
For planets the UT of transit over a local meridian may be obtained from:
meridian, phenomena depending on the rotation of the Earth may be calculated in
terms of TDT by the same methods as those by which calculations referred to the
Greenwich meridian are made in terms of Universal Time. time of ephemeris transit (,\ / 24) x first difierence. (s.63-s)
L.,1.
 310 EXPLANATORYSUPPLEIVENT
EPHEMEFIDES OFTHESUN, liooN,ANDPLANETS 31 1
5 / OFBITAL
This produces an elror that is usually less than 1second. ll'here ) is the ea51 comcts, the generation of accurate ephemerides often requires that,
longitude in hours and the first diflerence betiveen transit times is about 24 hours. short-period
the planetary perturbations, the comet's rocketlike outgassing forces
in addition to
be taken into accorrnt.The mathematical models trsedto describethese so-called
by \Iarsden ct aL. (1973) and Yeomans a1d
nongravitational effects are described
5.7 ORBITANDEPHEMEBIDES
OF OTHERBODIES Chodas (1989)'
WITHINTHESOLARSYSTEM The magnitude parameters (11,G) given for the minor planets can bc used to
compute the object's apparent magnitude using thc follorving relationships:
5.71 MinorPlanetsandComets
apparent n'ragnitude= i1+ 5log(Dr) - 2.5log[(1 - G)Pt + GPzl 6.11 l)
Orbital information is available for several thousand minor planets (often called
asteroids) and scveral hundred comets. Orbital data for all numbercd minor plauets.
and ephemerides lbr those objects that reach opposition drrring the rolume year. are
Pr = exp[-3. 33(tanaI 2)o631
published annually by the Institute for Thcoretical Astronomy in St. Petersburg.
Russia. These annual \,'olumesJerLtitled Ephemerides of the Minor Pla,nets.can be Pz = exp[ 1.87(tana l2)]?21, (s.7t-2)
purchased from the White Nights Trading Company, 520 N.E.83rd St., Scattle.
w A 9 8 1 1 5 .U S A . where
The N{inor Planet Center (60 Garden St.. Cambridge, X.lA 02138,USA) olTers
a variety of serviccs lbr obtaining information on minor planets and colncts. Cal.r | = geocentric distance in Ali
Loguesof Cornetary Orbi.ts are published every few years. Approximately ten tines r = heliocentricdistance in ,Lu
each )'ear, astrometric observations, ephemerides, and updalcd orbits are published a = Sun-asteroid Earth angle.
for both comets and minor planets in tlte Minor Pla,ne.tCirculars. These circulars
are available in printed form, bv electronic mail, ard as machine-readable files orr In lhe Astronomical ALrrr.ano,c:,the ephemeris positions and orbital elements
personal courputer disks. Nlagnetic tapes, contailing large files of astrometric data are given with respect to a coordinate systern defined by the equinox of the FK5
for thousands of minor planets atrd corncts, can also be purchased from The \Iirlrr fundamental referencestar catalog and the J2000.0epoch (FK5/J2000.0 systern),
Planet Center. whereas the vast majority of solutions for orbital elements, prior to 1992, were
The accuracy of published orbital parameters for minor planets and cornets computed u"ith respect to the equinox of the FK4 reference star catalog and the
varies fiom object to object. In gcrrcral, objects rvhose orbits are based upon op- B1950.0 epoch (FK4/81950.0 system). In Scctions 5.711 5.71,1,the expressions
tical observational data sprcad over lengthy time intenals *'ill have nrore orbit:rl "1950.01'and "2000.0" refer to the FK4/B1950.0 and FK5/J2000.0 systemsrespec
acculac). than those objects whose orbits are based upon shorter data intclvals. tively. Comission 20 of the International Astrolomical Union has rc(rommolded the
The inclusiott of radar datn in the orbit determination process *'ill dramaticalll following conversiol procedures betq'een the two svstems.
irnprove the orbital accuracy, particularly if the existing optical data are sprcad
overonly a -lrorr inrertll ol rime.
5.711 Conversion of Comet and Asteroid Astrom€tric Positions from 81950.0 to
Tlre annual Astronomical ,4lmanac volurncs publish cphcrnerides for thc first
J2000.0 Let d0=object's right ascensionrcfcrred to 81950.0 s)rstem.
four numbered minor planets. These volumes also incluclc orbital elements, nag-
do= object's declination referred to B1950.0 svstem.
nitude parameters,opposition dates, aud opposition apparent mitgnitudes lor the
li = object's right ascension rcferred to J2000 0 system'
larger minor planets that reach opposition during the volume ycar. In addition, the
d = object's declination referrcd to J2000.0system.
osculating orbital elements are given for the pcriodic comcts that are scheduleclto
pass perihclion within the volu[rc year.
1. Calculate the rectangular components of the objcct's position vector ro.
The orbital parautcters given for the minor planets and comets can be used to
generate approximate ephemeris information using two body formulations. These
same elements catr be employed to gcncrate more accurate ephemcrides $.hclr used I cos(ao)cos(do)l
(5.71l-l)
to initialize numerical integrations that include planetary perturbations. F-oractive
I sin(ao)cos(do)| .
L sin(do) l
3'12 EXPLAMTORYSUPPLEMENl OFIHE SUN.MOON,ANDPLANETS
EPHEMERIDES 313
2. Remove the efrects og eltiptical aberration (Bterms) to form astrographic of Comet snd Astemid Astmnetric Posidons ftom J20ffi.0 to
position vector (r1). as = object's right ascensionreferredto J2000.0syst€m.
Lel
Let rj be the vector transpose of r9. Then form 11.
do = object's declinationreferredto J2000.0system.
= object's right ascensionreferred to 81950.0 system.
| -1.62557 x 10-61 d = object's declinationreferredto 81950.0system.
r r = r 0 - A + ( r d . A ) r 0 w h e r e Al=- o . f r e t e xr o - c| 6311+)
L-0. rcs+g x to-ol
Calculate the recta.ngula.rcomponents of the object's position vector r0.
and (rj ' A) is a scalar product.
3. Form the J2000.0vector r from the vector 11using the matrix M; fcos(ao)cos(do)l
1 6 = l s i n ( a 6 ) c o ( d 6l .) (s.7r2-r)
L sin(do) I
r = Mrr, (5.71l-3)
C a l c u l a t e t h e 3 x l vectorrl(r1,)1,zl).
wnere
rr =M-'ro where16isa3x l vector. (5.7r2--2)
M=x(0)+r*(0);
The M matrix is given in Section5.711.
I= Julian centuriesfrom I}1980.0epochto the obserlation time (t) Include the effectsof elliptic aberration (E-terms).
= (t - u33282.423) | 3652s.
r = rr +A - (ri.A)rr, (5.712:3)
The equation_ r = Mr1 is a modificationto Equation 30 in Murray (1989),
where the following matrices are those defiaed as his equations ZA dnd Z6]
, where the vector A was defined above in Section 5.711 and (r,t.A) is a scalar
_0.01118148322s662 Product.
| 0.9999256i%s56a77 _0.0048s900381535921
x(0)= J0.01I181483239t717O.9W314848933r35_0.W27162ss47142 Calculate the new right ascensionand declinations (a,d) wing the following
|
_O-U]fp.27170293744O
10.0048590037723143 0.999988t94ft23742 expre€srons,
J
. f -0.n264ss262-1.1s399t86892.t t346l90l Let r=
f cPX(o) = r. 1540628161 -o.or2su2wt o.v23g2t47a. (r.r,,_o)
| | r = r(x,y,z)
L-2. |t2n9M8 _0.ms6r.24448 0.0t0258?134
)
cos(4= I
cos(a)
The matrix expression (fi((0)) removes the error introduced when the v
81950,0 reference staxs were advanced to the epoch of obsenation using sin(a)cos(d)= I
r
their proper motions (i.e., the error results from the incorect constantof z-.
precessionand the equinox drift between the FK4 and FKS systems). The sln(d, = (s.7t24)
r
matrix X(0) conects for the half arcsecond equinox ofiset of the FK4 at
the 81950.0 epoch and precesses the vector 11 from the epochof B19S0.O Conversion oflnertial Equatorial Coordinat€s fmm 81950.0 to J20fi).0 Let
to that of J2000.0using the 1976IAU constantof precessi,on.
4. Calculatethe new J2000.0right ascensionand declination (a,d) fromr(x,y,z) = equatorial rectangular position and velocity coordinates in J2000.0 system.
using the followine exDressious.
= equatorial rectangulax position and velocity coordinates in 81950,0 system.
Lerr=u@lpQ'
X(O)rr,
cos(4) cos(d) = : X(0)vr, (s.713-1)
r
sin(a) cos(d) = 1
is the 3 x 3 matrix definedin Section5.711.
inversetra.nsformationfrom J2000.0to 81950.0 iB performed using the
slo(.1,= -. (5.71l-5)
r traospoee) of the X(0) matrix.
314 EXPLANATORY
SUPPLEMENT / OBEITAL oFTHE SUN,MOoN,ANDPLANETS
EPHEMERIDES 315
5.714 J2m0.0 Keplerian Orbital Elements from 81950.0 Elements Given the six the known angles, Z/, L,and,J, are given by
Keplerian orbital elements a, e, Ts, u, Q, arld.i, in the I}1950.0 reference system,
one may compute the corresponding orbital elements for the J2000.0-basedsystem / = 4?50001688
(denoted by primes):
L = 5?198s62o9
a, a'= Semi-rnajoraxes in B1950.0,J2000.0system "/ = 0300651966.
e, e' = Eccentricity in 81950.0, J2000.0 system
Zo, 4 = Perihelion passagetime in 81950 0, J2000.0 system compute the corresponding angles a,,, Q, ar.d I (81950.0) given the a.ngles r,.r/,
'9,, undf (J2000.0),use the following expressions;
r..r,ru' = Argument of perihelion in I}1950.0, J2000.0 system
J2, J?/ = Longitude of the ascending node in 81950.0, J2000.0 system
sin(c.,r/- c..r)sin(I) = sin(-Q sin(Z' + J?')
i, i' = Inclination in 81950.0, J2000.0 system.
cos(url- cr) sin(I) = sin(/) cos(, - cos(I') sin("I)cos(Z' + O')
Then at=a, e=e',To=T'o and'w', Q', and i'are computed ftom the following cos(I) = cos(/) cos(, + sin(1')sin(, cos(L' + Qt) (5.714-2)
expresslons: sin(Z + l'l) sin(I) = sin(I') sin(/ + o')
cos(Z + O) sin(4 = - cos(/) sin(, + sin(I') cos("/)cos(l' + O').
sin(cu'- c..,)sin(I') = sin("Dsin(Z + l-?)
cos(o/- ir) sin(I') = sin(/)cos("D+ cos(| sin(, cos(Z+ f,2)
The values for the constants L', L, and,"/ have been determined using the fol-
cos(1/)= cos(I)cos("4- sin(4sin(Ocos(Z+ O) (5.714-l)
lowing values for the obliquity of the ecliptic;
sin(/ + g') sin(/) = sin(I)sin(Z+ O)
cos(Z/+ J'l/) sin(1')= cos(4 sin("D+ sin(1)cos(, cos(Z+ r/), 23?44578787for (81950.0)
23?43929111for (J2000.0)

5.8 KEPLERIAN OF THEMAJOR

FORTHEPOSITIONS
ELEMENTS
PLANETS

Lower accuracy formulas for planetary positions have a number of important appli-
pations when one doesn't need the full accuracy of an integrated ephemeris. They
a,reoften used in observation scheduling, telescope pointing, and prediction of cer-
tain phenomena as well as in the planning and design of spacecraft missions. Also,
they are more easily incorporated into a computer program.
I' Classical Keplerian orbital elements a.regiven in Table 5.8.1. The approximate
errors over the oeriod 1800-2050 are shown in Table 5.8.2. The errors
when extrapolating outside this interval can be substantially greater. The
of the approximate formulas may be compared to those of the integrated
Figure 5.714.1
The relationshipbetweenthe 81950.0 andJ2000.0 es. DE200. which are less than 1 arcsecond throushout 1800-2050' often
referenceframesandthe orbital plane smaller.
 5 / oFB|TAL
E P H E M E R I D EO . NOPLANETS
S F T H E S U N ' l \ r O O NA 317
Table5.8.1
ClassicalKeplerianelementsat the epochJ2000 (JED2451545.0),givenwith respectto EPHEMERIDES
5.9 BASIsFoR PRE-1984
the meaneclipticand equinoxof J2000.The anglesare in degrees,and theircentennjal
rates are in arcsecondsper century.The six elementsare the semi-majoraxis, prior to 1984, the ephemerides for the Sun, N{ercurl', Venus, arrd N{ars were based
eccentricity, longitudeof ascendingnode,longitudeof perihelion,
inclination, and mean
(1898). Computerized evaluations of the
longitude,respectively on the theories and tables of Newcomb
tables were used from 1960 through 1980. From 1981 to 1983' the ephemeridcs
were based on the evaluations of the theories themselves. The ephemerides of the
Sun were derived from the algorithm given by S. Newcomb 1n TabLesof the Sul
Me.cury
0.38709893 0.205630697.00487 48.33167 77.45645 252.25084 (Newcomb, 1898). Nct'comb's theories of the inner planets (1895 1898) served as
0.00000066 0.00002527 - 2 3 . 5 1 -446.30 573.57 261628.29 + 415 rev the basis for the heliocentric ephemerides of Nlercury, Venus, and Mars. In the case
VENUS
of Mars, the corrections derived by F.E. Ross (1917) were applied
0.723331990.00677323 3.39471 76. 68069 '|31.53298 1 8 1 . 9 7 9 7 3
0.00000092 0.00004938 2 . 8 6 9 9 6 . 8 9 - 1 0 8 . 8 0 7 1 2 1 3 6 . 0 6+ 1 6 2r e v Ephemeridesof the outer planets,Jupiter, Saturn, Uranus, Neptune. and Pluto,
E-M barycenter were computed from the heliocentric rectangular coordinates obtained by'numeri-
1 . 0 0 0 0 0 0 1 10 . 0 1 6 7 1 0 2 0.00005
2 - 1 1 . 2 6 0 6 4102.94719 100.46435
cal integration (Eckert el al., 1951). Although perturbations bv the irrner plancts
0.00000005 -0.00003804 -46.94 18228.25 1 1 9 8 . 2 81293740.63+ 99 rev
(Clemence, 1954) were included in the printed geocentric ephemericlcsof the outer
lrars
1.523662310.093412331 . 8 5 0 6 1 49.57454 336.04084 355.45332 planets, they were omitted from the printed heliocentric ephemerid<:sand orbitai
-0. 00007221 0 . 0 0 0 1 1 s 0 2-25.47 - 1 0 2 0 .1 9 1 5 6 0 . 7 8 2'l7103.78 + 53 rev elements.
Jupiter The lunar ephemeris, designated by the serial number j = 2. was calculated
5.203363010.048392661 . 3 0 5 3 0 1 0 0 . 5 5 6 1 5 1 4 . 7 5 3 8 5 34.40434
0.00060737 -0.00012880 4.15 1 2 1 7 . 1 7 8 3 9 . 9 3 557078.35 +8rev directlv from E.W. Brou'n's algorithm instead of from his ?ables of Motion of the
Saturn Moon (1979). To obtain a strictly gravitational ephemeris expressedin the measure
9.53707032 0.054150602.44446 113.71504 92.43194 49.94432 of time defined by Newcomb's Tables of the Sun, Ihe fundamental argutrellts of
-0.00301530-0.00036762 6.11 1 5 9 1 . 0 5 1948.89 5 1 3 0 5 2 . 9 5+ 3 r e v
uranus Brorvn's tables were amended by removing the empirical term and by applying to
1 9 . 1 9 1 2 6 3 9 0.04716771
3 0. 76986 74.22944 170.96424 313.23218 the Moon's mean lonsitude the correction
0.00152025 -0.00019150 -2.09 1 6 8 1 . 4 0 1 3 1 2 . 5 6 246547.79 +l tev
Neptune , 8':',72
- 26':'74 (5.9I )
T t1':2274,
30.068963480.008585871.76917 131.72169 44.97135 304.88003
- 0 . 0 0 1 2 5 1 9 60.00002514 - 3 . 6 4 -151.25 844.43 7A6449.21
Pluto where Z is measuredin Julian centuriesfrom 1900January 0.5 ET = JED 24150200.
39.481686770.24880766 1 7 . 1 4 1 7 51 1 0 . 3 0 3 4 7224. 06676 238.92881 In addition, this ephemeris was based on the IAU (1964) System of Astronornical
-0.00076912 0.00006465 1 1 . 0 7 - 3 7 . 3 3 - 1 3 2 . 2 5 522747.90
Constants, and was further improved in its precision by transformation corrections
(Eckert el ol., 1966,1969).The expressionsfor the mean longitude of the Nloon and
Table 5.8.2 its perigee were adjusted to remove the implicit partial correction for aberration
Approximatemaximumerrors of the Keplerianformulasover the
interval1800-2050,givenin heliocentric (Clemence et aI., 1952).
rightascension,
declination,and distance.

RightAscension Declination Distance 5.91 Introductionof New Constanrs

I" l [1000km] (') [1000km] l" l [1000km]
In 1964 the International Astronornical Union adopted the 1968 IAU System of
lvlercLrry 20 1 Astronomical Constants (IA|J T,rans.,1966).The complcte list of these constantsr
Venus 20 10 5 220 10 with detailed explanation, referenceand formulas were published in the Supplement
Earth 20 420 15 to the Astronctmical Ephemeri,d.es1968, which was reprinted in the 197'l edition of
Mars 25 25 30 30 40 40
Jupiter 300 1000 100 350 200 600 NheEx:planatory Supplement.
Saturn 600 4000 200 1400 600 4000 By 1970 it was recognized that the ephemerides being used for the national
utanus 60 800 25 400 125 2000 almanac publications required improvements that the fundamental caialog of star
Neptune 40 800 20 400 100 2000
250 70 positions (FK4) should be replaced by a new fundamental catalog of star positions'
Pluto 40 1100 t0 2000
 318 EXPLANATORYSUPPLE[4ENT
319
SF T H E S U N ,M O O N ,A N D P L A N E T S
< / O F B I T A LE P H E [ 4 E R I D EO
that better values t'ere knorvn for some astronomical constants, that the definitiou
it was necessary to adjlrst some of the constants in order to produce a
and practical realization of ephcureris time rvas inadequate ancl inconsistent, th41 data, that
the least-squares sense
atomic time was available, and that a new epoch miglrt be introduced to roplace best-fit in
B1950.0.
Since the necessarv changcs would affect many aspects of astronomv. it was
decided to introduce as many changesas possiblc at one time in a consistent sysLe11. 5,10 REFERENCES
rather than to introduce them at many dillerent times $'ith resulting inconsistencies. J D.. and Silberberg' E C (1973) "Lascr Observa-
Abbot, R.L, Shelus, P J., Nlulholland.
Agreement was reached that all changes should be introduced into the national
tions of the tr4oon: Identification and Construction of Normal Points for 1969 1971"
publications of the ephernerides for the 198,1editiotrs, and that the changes shoukl Astron J. 78,784 793
(196'1) "Results ofthe Six Inch Transit Circle.
be eflective on 1984 January 1. Adams, A.N., Bestul, S.M., and Scott, D K
The changes were designed to be the best that can be achieved on the basis 1949 1956" PubL. U S. Naool Obs. XIX, I.
Adams, A.N. and Scott, D.K. (1968) "Results of Observations Nlade with the Six-Inch
of current kno*'ledge and should reduce the systematic errors between theorl and
Transit Circle, 1956 1962" Publ. U.S Noual Obs XIX, II.
observations by several ordcrs of magnitude. Holvever, it was recognized that per-
Aoki, S., Guinot, B., Kaplan, G H , Kinoshita, H.H., NlcCarthy, D D , and Seidelmann,
lection cannot be achicved and the improvenlents introduced should perrrrit thc P.K. (1982). "The New Definition of Universal Time" Asiron ,4sli"ophys.105, 359
deterrnination of further improvements. after the systematic deviations are redrrcerl 361.
by the prcsent improvements. Bretagnon, P. (1980). "Th6orie au deuxiimc ordre des plandtes inf6rieures" A stron. As
The ncq' resolutions forrn a self-colsistent and interrelated system of changcs trophgs. 84, 329 341.
necessary to improve the astronornical rclerence svstem. The resolutious shoirld Bretagnon, P. (1981). "Constructron d'uDe th6orie des grossespla ites par unc rn6thode
it6rative" Astron. Astrophlls. 101, 342 349.
not be introduced individuallv or selcctivell.. They must be implemented simulta
Bretagnon, P. ard Chapront, J. (1981). "Note sur les formules pour le calcul de la
neously.It is intended that whencverthe nerl,standard epoch (J2000.0)is used, all pr6cession" AstrorL. Astrophys. 103' 103 107-
the courputations and reductions will be based on the new astrorrotnical constants Bretagnon, P. (1982). "Th6orie du mouvement de I' cnsemble des planites. Solution
and resolutions. Furtherurore, proper utilization of srrch data requires adherence to VSOP82" ,4stron. Astrophgs. 114, 278 288
the resolutions in any calculations or procedures. Bretagnon, P. and Chapront, J. (1982). Private communication in Standish (1982a)
Bretagnon, P. and Francou, G. (1988). "Planetary Theories in Rectangular and Spherical
The majoritv of the resolutions were prepared atrd adoptcd by the General [s-
Variables VSOP 87 Solution" Astron. AstrophAs.202' 309 315.
senrbly of the IAU at the 1976 and 1979 meetings (lAU Tr.ans.,1977,1980). fhr: Brown, E.W. (1919). ?ables oJ the Motion oJ the Moon (Yale Uriversitv Press New l{ar'-en.
1979 IAU Theory of Nutation was adopted at thc 1979 IAU General Assetrblv to CT)
provide an improved nutation theory for astronorrrical cornputations. Thc IUGCI Calame, O. and Nlulhollanrl. J.D. (1978a). "Lunar Tidal Acceleration Detcrmitred from
requested that the IAU reconsider its recoDrmendatiorrs.not because of thc ina{le Laser Range \lleasurements" Scicnce 199, 977 978.
Calame, O. and Nlulholland, J.D. (l978b). rn Tidal Frictzon and. Llrc Earth's Rotattort P
quacy of the mrtation theory. but becausethc Earth moclel underlving the nutatiol
Broche and J. Sunderman, eds. (Springcr, Nerv York) p. '13
theory rvas rrot corrsidered an adequate rnodcl. Although differences in thc Earth's Cappallo, R.J. (19E0). "The Rotation of the N4oon" Ph D. Thesis (NIIT, Cambridge)
nodels could not be demonstrated lo significantly afiect the nutation. the r-elerartt Cappallo, R.J., King, R.W., Couselman, C.C. III, and Shapiro, l.L (1981) ")'lurncrir:al
commissionsof the IAU have adoptcd thc 1980IAU Theory of Nutation. The Firal Model of the [4oon's Rotation" Moon antl Planets 24,281 289.
Report of the \Vorking Group orr Nutation rvas published bv Seidelmann(1982). Chao, C.C. (1970). .A Prelinirrary Estimation of Tropospheric Influencc on the Rangc
The origin of the FKs with respr:ct to the FK.4 u'as specified by Frickc (1.982). and the Range Ratc Data During thc Closet Approach of thc l"1X{71 N{ars \lission"
Jet Prcp. Lab Tech. Memo #391-129 (JPL, Pasadena,CA).
The nevr definition of the formula rclating Greenrvich rnean sidereal timc to U[i-
Chapront, J. (1970). "Construction d'une th6oric litt6rale plandtairejusqrr'au secondordre
versal Tinre rvaspublished bv Aoki et ol. (1982). des masscs" Astron. Astrophgs. T,175 2O:1.
The fortrrulation using the new constir,nt of precession u,as publishecl by Liesk: Chapront, J. and Simon. J.L. Og72\. "Variations s6.culaires au premicr ordre des 6l6ments
et at. (7977) and Lieske (1979). des quatre grosses planates. Comparaison avec Le Verrier et Gaillot Astron Astro-
The 1976 IAU System of Astronomical Constants is giverr iu Chapter 15. 'fhe phgs. 79,231 234.
Chapront-Touz6, lvl (1980). "La solutioD ELP tiu problime central de la Lune" '4slron'
constantsusedin thc creation ofthe pianetary and hrrrarephernerides, DE200/LE2r)tl.
Astrophus.83. 86 94.
are also given in Chapter 15. These are not all idcntit:al to those of thc 1976 IAII Chapront-To"uz6, M. and Chapr.xt, J. (1980). "Les perturbatiors plan6taircs de la Lune"
System. for it had become apparent in fitting the ephcmerides to the observational Astron. AstrophlJs. SL, 233 216.
 EXPLANAToBY
SUPPLEMENT EPHETERIDESOFTHESUN'I,OON'ANDPLANETS 321
320 5 / oFB|TAL

Chapront-Touz6, M. and Chapront, J. (1983). "The Lunar Ephemeris ELP 2000" , stron (rogh, F.T. (1972). Lecture Notes in Mathematics, 362 (Springer-Verlag, New York) p
Astrophgs. 124, 50 62. 22.
Terms of Classical Planetary Theories Using the Results of
Chapront-Touz6, M. and Chapront, J. (1988). "ELP 2000-85: a Semi-analytical Lunar LasW, J.(1986). "Secular
Ephemeris Adequate for Historical Times" Astron. Astrophgs. 19O,342 352. General Th"ory Aslron. Aslrophls 157. 59 70
Chollet, F. (1984). "Evaluation des corections de phase dans I'observation des positiols J.H. (1979). "Precession Matrix Based on IAU (1976) System of Astronomical
Lieske,
des planbtes" Astron. Astrophgs. 139, 215 219. Constants" Astron A strophl|s. 73, 242 281.
Morando, B. (1977). "Exprcssions for the Pre-
Clemence, G.M. (1954). "Perturbations of the Five (Juter Planets by the Four Inner Lieske, J.H., Lederle, T., Fricke, W., and
Ones" in ,4slronomicaL Papers American Ephemeris (U.S. Government Printing Of cession Quantities Based upon the IAU (1976) System of Astronomical Constants"
fice, Washington, DC). Astron. AstroPhYs 73,282 284
Clernence,G.M., Porter, J.G., and Sadler, D.H. (1952). "Aberration in the Lunar Ephem- Lindegren, L. (1977). "Meridian Observations of Planets with a Photoelectric N'|ritislit
e rs" Astron. J. 57, 16 47. Micrometer" Astron. Astrophgs. 57, 55 72.
Davies, M.E., Abalakin, V.K., Lieske, J.H., Seidelmann,P.K., Sinclair, A.T., Sinzi, A.trl Lyttleton, R.A., Cain, D.L., and Liu, A.S. (1979). JPL Publication 79 85 (JPL. Pasadena,
Smith, F}.A., and Tjuflin, Y.S. (1983). "Report of the IAU Working Group on Car- cA).
tographic Coordinates and Rotational Elements of the Planets and Satellites: 1982" Marsden, B.G., Sekanina, 2., and Yeomans, D.K. (1973). "Comets and Nongravitational
CeL Mech. 29, 3Og 321. ForcesV." Astron. J.78, 211 115.
('Results
Dickey, J.O., Williams, J.G., and Yoder, C.F. (1982). "High-PrecisionEarth Rotation and Morgan, H.R. (1933). of Observations with the Nirre-Inch Trausit Circle. 1913
Earth-Moon Dynamics: Lunar Distances and Related Observations" in Proc. IAtj 7926" Publ. U.S Naual Obs, XIII
Co|\.63,0. Calame, Ed. (Reidel, Dordrecht, Holland) p. 209. Morgan, H.R. and Scott, F.P. (1948). "Results of Observationsnade lviih the Nine-T .h
Dickey, J.O. and Williams, J.G. (1982). "Geophysical Applications of Lunar Laser Rang- Transit Circle. 1935 19,15' Publ. U.S. Naual Obs., XV. part V
tng" EOS 63,3O1. Moyer, T.D. (1971). "Mathematical Formulation of the Dolble-Precision Orbit Determi-
Duncombe, R.L. (1969). "Heliocentdc Coordinates of Ceres, Pallas Juno, Vesta 1928 nation Program (DPODP)" Jet Prctp. Lab. Tech Report #32-1527 ()PL. Pasadena,
2OOO"Astronomical Papers American Ephemeris XX part II (U.S. Governrnent Print- cA).
ing Office, Washington, DC). Moyer, T.D. (1981). "Transformation from Proper-Time on Earth to Coordinatc Time in
Ecked, W., Brouwer, D., and Clemence, G.M. (1951). "Coordinates of the Five Outer Solar System Barycenter Spacetime Frame of Refererce I, If' Cel. Mech.23.33 56,
Planets 1953 2060" in Astronomical Papers Americart Ephemerid,es XII (U.S. Gor- 57 68.
ernment Printing Office, Washington, DC). Muhleman, D.O. and Anderson. J.A. (1981). "Solar WiDd Electron Densities fIon Viking
Eckert, W.J., Walker, M.J., and Eckert, D.D. (1966). "Transformation of the Lunar Co Dual Frequency Radio Nleasurements" Astrophys. J.247,7093 1101
ordinates and Orbital Parameters" Astron. J. 71.311 332. Mulholland, J.D., Shelus. P.,L, arrd Silverberg, E.C. (1975). "Laser Observations of the
Eckert, W.J., Van Flandern, T.C., and Wilkins, G.A. (1969). .A Note on the Evaluaticrn Moon: Normai Points for 1973" Astron. J. E0. 1087 1093.
of the Latitude of the Moon" Mon Not. R.A.S. 146. 173 17a. Murray, C.A. (1989). "The TransforBation of Coordilatcs Betweetrthe Systens Lrf81950.0
Estabrook, F.B. (1971). "Derivations of Relativistic Lagrangian for n-Body EquatioDs and J2000.0J and the Principal Galacti<: Axes Referred to J2000.0" Astron AstrotrhlJs
Containing Relativity Parameters D and c" JPL Internal Communications 218, 325 329.
Ezplanatorg Supplement to the Astronomica,l Ephemeris dnd the American Ephemeris anLI Newcomb, S. (1898). "Tables of \,Iotion of the Earth on its Axis and Arourrcl the Sun''
Nautical ALmanac (H.M. Stationery Ofnce, London), 1961. 1n Astronomical Papers American Ephcmerts VI, part I (U S. Covernurent Printing
Ferrari, A.J., Sinclair, W.S., Sjogren, W.J., Williams, J.G., and Yoder, C.F. (1980). "Gco Oflice, \I'ashirgton, DC).
physical Parameters of the Earth lvloon System" J. Geophys. -Res.85, 3939 3951. Newcomb, S. (1898). "Tables of the Heliocenrric N{otion of N4ercurl'' n AshartornicaLPa'
Fricke, W. (1971). "A Rediscussion of Ncwcomb's Determination of Precession" .,lstrorr. pers Americart EphemertsYl, part II (U.S. (lovernment PrintilrEiOflice. \Vashington,
AstrophAs. 13, 298 308. Dc).
Fricke. W. (1982). "Determination of the Equinox and Equator of the FK5" .'lsfrun.. '1s- Newcomb, S. (1898). "Tables of Helioccntric Nlotion of Verus" in Astrcnttm.iutl Ptpers
trophAs. I07 , L13 L16. Amerzcan Epherneris VI, part III ([J.S. Governmert Prirrting Olfice. \[ashilrgton,
Goldstein, H. (1950). ClassicalMechanics (Addison-Wesley,Reading, MA). Dc).
Newcomb, S. (1898). "Tables of the Heliocentric Nlotions of Nlars" in AstTonotnicaLPapers
Hammond, J.C. and Watts, C.B. (1927). "Results of Observationswith the Six-lrrch 'Ilan-
sit Circle. 1909 1918" Publ. U.S. Natal Obs.Xl. American Epherne:risVI, part IV (U.S. Government Prjnting Oflice, \4iashington. DC)'
Newhall, XX, Standish, E.\'L. and \Villianrs, J.G. (1983). "DEr02r a Nrrrnerit:all1'IItte
Hellings, R.W. (1986). "Relativistic ElTects ir Astronomical Timing Measurements" ,,1s
tron. J. 92, 65O 659. grated Ephemeris of the l,Ioon and Plancts Spanning Forty-Four Ccnlurics" Astr"7r"
Hughes, J.A. and Scott, D.K. (1982). "Results of ObselvatioDs Madc with the Six-hrch AstropllUs.125, 150 1tj7.
Newhall, XX, Williams, J.G., and Dickcy, J.O. (1989). "Earth RotalioIr from Lulal Laser
Transit Circle, 1963-1971' Publ. U.S. Na,xal()bs. XXIII, part III.
Ianna, P.A. (1974. 1980). Private communication. Ranging" IAU Symposiun 128 (Coolforrt, West \rilginia).
 322 EXPLANAToBY
SUPPLEMENT OFTHESUN,IVOON,
EPHEMERIDES ANDPLANETS 323
S / ORBITAL
Newhall, XX (1989). "Numerical Representation of Planetary Ephr:merides',Cct. Mcclt. (1981). "Two Differing DefiIritions of the Dynamical Equinox and thc
S@ndish, E.l,{.
45. 305 310. Mean Obliquity" Astron. Astrophgs. 101, L17-L18.
Null, G.W. (1976). Prirate communication. Standish, E.[4. (1982a). "orientation of the JPL Ephemerides, DE200/LE200. to thc
Null, G.\V. (1978). "Programmer Information for the Saturn Satellite Ephemeris 'pARSAT. Dynamical Eqrrinox of J2000" Astron. Astrophys. L14,297 302.
(1982b) "The.IPL Planetary Ephemerides" Cel. Mech. J. 26. 181 186.
1108 Software Package" Jet Prop. Lab Interoffice [,Iem o JI1.7-2b2 (using date elcment Srandish,E.M.
#360) (JPL, Pasadena,CA). Standish, E.M. (1985). "Numerical Planetary and Lunar Ephemerides: Presc[t Status,
Reasenberg,R.D. and King, R.W. (1979). ,,The Rotation of Mars,' ./. Geophys.R.es.8!. Precision and Accuracics" h Relatiuitg in Celestial Mechanics and Astrometru J.
n o . 8 1 1 .6 2 3 1 6 2 4 0 . Kovalevsky and V.A. Brumberg, eds. (Reidel, Dordrecht, Holland) pp. 7l 83.
Richter, G.W. and Nlatzncr, R.A. (1983). "Secoid-Order Contributions to Relati\-isri. Standish, E.M. (1990). "The Observational Basis for JPL's DE200, the Planct ar1'Epherl-
Time Delay in thc Parameterized Post-Nettoniar Forrnalisnf' PhlJ. Reu D. 28,3007 erides of the Astronomical Almanac" Astron. AstrophlJs.233,252 271.
3072. USNO (1984). "The Astronomical Almanac for the Year f984" (U.S. GovernrnentPinting
Rivlin, T.J. (197a). The ChebEshet Polynomials (.Iohrr Wiley & Sons, New York). Office, Washington, DC, H.M. Stationery Office, London).
Ross, F.E. (1917). "Nerv Elements of \lars" Astronomical Papers American Ephetneris Watts, C.B. and Adams, A.N. (1949). "Results of Observations N4adewith the Six-Inch
IX, part lI (U.S. Government Printing Oflice. \\ashington, DC). Tlansit Circle, 1925 19,11" Publ. U.S. Naual Obs. XVI. part L
Schmidt, R.E., Corbin, T.E.. and Van Flandern, T.C. (1980). ,'New US\O Zodiacirl Srar Watts, C.B., Scott, F.P., and Adams. A.N. (1952). "Results of Obsenaiions \lade rvith
Catalog" Bull. Anw. Astron. Soc. 12,740. the Six-Inch Transit Circle, 1941 1949" Pu,bl. U.S. Nanal Obs. XVI, Part III.
Will, C.M. (7974). EryerimerLtLrl Gralritation B. Bertotti, ed. (Academic Press, New York).
Schwan, H. (1977). "Develqrment and Testing of a Method to Derive an Instrumeltal Svs
Williams, J.G., Sindair, \V.S., and Yoder, C.F. (1978). "Tidal Acceleration of the N{oon"
tem of Positions and Proper N{otions ol Stars" Veroffentlichurtgen 27. Astrononrischcs
Ceophys. Res. Letters 5,913 946.
Rechen-Institut, Heidelberg.
Williams, J.G., Dicke, R.H.. Bender. P.L., Allo, C.O., Carter \\'.E.. Currie. D.G.. Eck
Seidelmann, P.K. (1982). "1980 IAU Theory of Nutation: The Final Report of thc IAU
hardt, D.H.. Fluer, J.E., Kaula, W.\L, tv'lulholland,J.D., Plotkin. ILH., Pcultner'.
Working Group on Nutation" CeL.Mech.27,79 106.
S.K., Shelus, P.J., Silvcrbcrg, E.C., Sinclair, W.S.. Slade, NLA., and Wilkinson, D.T.
Shapiro, LL (1964). "Fourth Test of CeDeral Relarivity" Phgs. Reu. refters 13, 289 791. (1976). "New Tcst of the Equivalencc Principle from Lunar Laser R.anging" Phys.
Shapiro, LI., Counselmann,C.C.,III, and KiDg. R.W. (1926). ,,Verificationofthc principle Reu. Letters 36. 551 554.
of Equivalence for f,lassite Bodies" PAus. Reu. Letters 36. 555 558. 'Lunar Phvsical
Williams, J.G., Slade. \1.A.. Eckhardt, D.FL, and Kaula. W.ll. (1973).
Shelus, P.J.. \'lulholland, J.D., arrd Silvcrberg, E.C. (1975). ',Lascr Obserrations of the Librations and Lunar Laser Ranging" lvloon 8, 169 1a3.
N,Ioon,Normal Points for 1972" Astron. J. 80. 15.1 161. Williams, J.G. (198.1). "Determinirrg Asteroid Nlassesfrom Pctturbations olt N{ars" lcarus
Simon, J.L. and Chapront, J. (1974). "Perturbations du secondordre des planatcs,lupitcr 57, 1-13.
et Saturne: Comparaison avec Lc Verrier" Astron. AstrophUs.32. 51 64. Williams, J.G. ard Standish, E.1\'1.(1989). "Dl narnical Rcfcrr:nceFrames in the PlanetarY
Simon, J.L. and Bretagnol. P. (1975). 'Perturbations du premier ordre des quatre grosses and Earth Nloon Systems" tn ReJerence Frames J. Kovalevskl'. I trlrreller. and B
plandtes: Va ations litt6rales" ,4stron. Astrophgs.42,259 26:1. Kolaczek, eds. (Klurver, Dordrecht, Holland) 67 90.
Simon, J.L. and BretagnonJP. (1975). ',R6sultats des perturbations du premier orth.c dcs Yeomans,D.K. and Chodas, P.W. (1989). ",{n Asymmetric OutgassingN{odolfor Cometaly
quatre grosses planltcs: Variatiorrs LittArales" Aslron. Astrophys. Stppl.22,1O7 l(,0. Nongravitational Accelerations" Astron.,J. 98, 1083 1093.
Simon, J.L. and Britagr)on. P. (1978). ,,Perturbatiorrs du deuxilrDc ordrc des quatre Yoder, C.F. (1979). "Etrects of the Spin-Spirr lnteraction and the Ine]astic 'Iidal Defor-
grosses plalites: Variations sdcuiaircs du delrri graDd a_xe' ,,lsrlron_Astrophus.69. mation on the Lunar Phlsical Librations" iIr Natural and Artifcial S t.eltite^fotnl)
369 372. P. Nacozy and S. Feucz \lcllo. cds. (Universitl' of Texas Press. Austilr) p 211
Sirnon, J.L. and Bretagnon, P. (1978). "R6sultats des perturbations du deuxiinrr,ordre
des quatre grosses planitcs'r AstT.on. Astrophgs. Suppl. 34, 183 194
Simon, J.L,. and Francou, C. (1982). "Arn6lioration dcs thdories de Jupiter et Satu re par
analyse harnroniquc" Astron. Astrophys. 114, 125 130.
Simon, J.L. (1983). "Th6orie du mouvement des quatrc grossesplanitcs. Solution TOP82'
Astrotr. AstrophAs. 120, 197 202
Simon, J.L. and Bretagnon, P. (1984). ,,Th6orie du nouvement <lc Jupiter et S:rturne
sur un intervalle de tornps de 6000 ans. SolutioD .IASON84' Astrcn. Astroph.tls. 138.
169 178.
Standish, E.NI. (1973). "The Figure of Nlars and its Eficct on Radar-ranging, !rh?n.
AstrophAs. 26, ,163 {66.
Standish, E.M., Keesey,M.S.W., and Newhall, X X (1976). "JPL Development Ephcrneris
Numbcr 96" Jet Prop Lob Tech. Report 32-160.9(JPL, Pa*sadena, CA).
 CHAPTER

rbitalEphemerides
and Ringsof Satellites
by JamesR. RohdewithAndrewSinclair

INTRODUCTION

qrhenerides of most satellites in The Astronomical Almanac a,re calculated

a set of analytical expressions, termed a "theory," for the motion of the satellite
orbit that is basically an ellipse, but subject to various perturbing forces, such
oblateness of the pla.net and the attractions of other satellites and the Sun.
orbits of the outer satellites of Jupiter are highly perturbed and representations
are not suficientlv accurate. In these cases numerical intecrations of the
of motion of the satellites a.re used.
objective of this chapter is to describe the theory of the motion of each of
principal satellites in euough detail for search and identification purposes. In
cases, a description of the complete theory that can be used for the precise
of observations is given, but in cases where the theories contain large
of periodic terms, only a.n outline of the theory is given, together with
to where the complete theory can be found.
tables for all of the known satellites can be found in Chapter 15. Table
Pres€nts orbital data; Table 15.10 gives physical a.nd photometric data; a.nd
15.11 presents north poles of rotation and prime meridia.n data.
An alternative to Section F in The Astronomical Almanac, called The Satellite
is available on diskette. The Sotellite Almanac is a progra.rn for MS-
based nicrocomputers that ruas on the IBM PC computers and compatible
The time span of the Satellite Alrnanac is currently 1990 December 28
January 5.

OtbitalElements
 7.,
326 EXPLANATORYSUPPLEMENT
6 / ORBITALEPHEMERIDESAND RINGSOF SATELLITES 327

In addition, the size and shape of thc orbit are described by the serlr-rna;or a,115
, and eccentricity e. The position of the satellite along the orbit. measrrrcd frqrl
the periceDter, is described by the mean anornaly M or the true anomaly f. The
difference / - M is known as the equation of center. The equation of centcr can be
approximated to sumcient accuracy for the satellites by

S (Positionof satellite)
* = (r" , js) ,t", * soe2
sin2M
+
t f;", "insu.
( 6 .r1- r )

Referenceplane (Pericenteroforbit)
For many satellites the pcricenter is ill-defined si[ce e is small, or the nocle is ill-
defined since i is small. In the former case it is usual to use the quantity ), rathcr
than M, and in the latter case the quantity o, rather than r.u.q,here

(Ascending ),= M +u + Q (the mean longitude),

node oi
orDrr)
(the longitude of pericenter). (6.1r-2)
Figure 6.11.1
The orbital elementsused to describethe orbital plane relative 6.12 SecularPerturbations
of theOrbit
to a referenceplane and an origin of longitude O.
If the orbit q,'ereunperturbed, then all the elements would reurain constant except
the mearr anomaly M, which would inclease at a constant rate n. the rnean motiorr
of the satellite. For a perturbed orbit the same elliptical model can be used. but the
orbital elements, now called osculating elements, arc functiorrs of time, typically of
the form of a constant plus periodic terms. In addition, the angular elements carr
have a lilear (or secular) variation rvith time. The form of the expressions for the
osculating elements depcnds on the rcference plane chosen. The oblateness of thc
planet causcsthe pole of the orbital plane to precessaround the pole of the planet.
ly' (Ascending node of orbit)
and solar perturbations r:ause it to prccess around the pole of thc planct's orbit.
Perturbations bl' other satellites cause precession around the polcs of their orbital
planes, but usually this effect is smaller than the other two effects. Because thr:
satellites' orbital poles are usually close to the pole of the planet, their cffect is
Figure 6.11.2 primarily an addition to the oblateness cflect. The Laplacian plane is the reference
An equivalentform of the orbital elements.The end-on view of a great circle on plane about whosc axis thc satellite's orbit precessesas a result of thcse two major
the celestialsphereis representedas a straight line. Trianglesform;d by such
lines are sphericaltriangles,and the formulas of sphericaitrigonometryapply.
effects. The axis of the Laplacian planc is coplanar with and betl'een the polar
and orbital axes of the planet. Hence thc node of the orbit legressesaround a planc
that lies between the equatorial and orbital planes of the planet, and has a common
relerence plane and origin are dcscribed by tltree eletnents. which line of intersection with the planes. The orbit maintains a constant inclination to
are shou,n irl the Laplacian plane (apart from other, usually smaller perturbing efiects). The
F i g u r c s6 . I l I a n d 0 . 1I . 2 . T h e s nc l o m p n r sa r e
Laplacian plane is also known a-slhe inuariable plane lhrougb the planet.

u, = argumcnt of the pericenter, The location of the Laplacian plane is shown in Figure 6.12.1. It divides the
angle between the equatorial and orbital planes into parts it and 4, where
O = longitude of the ascendingnode,
i = inclination. 2n2J2fisin2i2= oznn(l - e'2)-)'2 sin2it, (6.r2-t)
328 APLANATORYSUPPLEMENT ANDRINGSOF SAIELLITES
EPHEMERIDES 329
plane, then the osculating inclination and longitude of the node will
Equatorialplane a.nglesi and ,f?.The formulas of spherical trigonometry give

sinisin(O- Or)= sintsin0,

(6.12-3)
sinicos(O- f,2r)= cos7 sini + sinTcosir cosd,

ir, Or a.re the inclination and longitude of the node of the Laplacian plane
to the adopted reference plane; and 7 and 0 are defined above. In most
,y is small, and so (O - O1) is small as well. Thus the following approximate
are obtained:

i = i 1+ 7 c o s 0 ,
O=Or+'Ysind/sinir. (6.124)

the use of a refereuce plane other than the Laplacian plane introduces a
oscillation into the osculating inclinatiou and longitude of the node' Note
FtSur€6.12.1 it and JZr are quantities whose values are determined by the physical prop
TheLaplacialrplane
of the system. The arbitra,ry constants associated with the inclination and
axe the quantities 7 and the phase of 0- This choice of reference pla'ne also
- -
a periodic variation, equal to the augula"r difference $ + Q Q1 0, irtto
longitude of the pericenter and the mean longitude. This periodic variation is
approximately by
where J2 is the oblateness coefficient of the planet; 16 is the equatorial radius of the
A@, Al = ? tan sin d. (6.r2-s)
planet; z' is the orbital mea.n motion of the planet; e' is the orbital eccentricity of ;,r
the pla.net; and n is the mean motion of the satellite.
Apa.rt from small short-period perturbations, the motion of the satellite's orbit
most satellites the oblateness efiect on the orbital plane is much larger than
relative to the Laplacian plane is such that
solar perturbations, and so the Laplacian plane is very close to the equatodal
For a more distant satellite, such as Iapetus' the effects are comparable, so
7 = COrXlta , Laplacian pla.ne lies about midway between the equatorial and orbital planes'
(6.r2-2)
0=eo-K(t-t , the Moon. the solar perturbations dominate, and the Laplacian plane virtually
with the Earth's mean orbital plane the ecliptic'
where 7 is the inclination of the satellite's orbit to the Laplacian plane, and d is The oblateness. solar. and other perlurbations also cause secular variations of
the a.rc reckoned from the intersection of the planet's equatorial and orbital planes of pericenter (in most cases the pericenter advances a'round the orbit)
along the Laplacian pla.ne to the node of the satellite's orbit on the Laplacian plane of the mea[longitude of the satellite, so that the mean motion in longitude is
However, it is usually more couvenient to ta.ke a plane other than the Laplacian from (generally greater than) the value derived from Kepler's law for ao
plane as the reference plane, as it eaaes the calculation of pertubations by other orbit. It is this overall rate of motion that is quoted as the satellite's
objects, and also eases the transformation of the calculated position to a reference motion, and it is thus an observed mean motion. Hence, when using the mean
frame based on the Earth's equator (which is needed for comparison with observa- n to determine the mass of the planet from Kepler's law, one musi allow for
tions). If, for exarnple, the orbital plane of the pla.net in Figure 6.12.1 is taken as of it due to secular perturbations.
 SN D R I N G SO F S A T E L L I T E S
E P H E M E R I D EA o.t I
330 EXPLANATOBYSUPPIE[,ENT 5 i OBBITAL
the resonance itself As a result the angular
It is useful to note itr this context that the semi-major axis of the orbit is planet, other pedurbations, and bv
affected in a similar way. The various perturbing forces cause perturbations of the argumenI
eccentricity and the longitude of the pcricenter of the form 0=2\'-\-a (6.13-3)

then the motion is

Ae=AcosM, varies very slowly. If 0 either increases or decreasescontinually'
(6.12-6) circu[,ation, because the position of conjunctions of the satellites circulates
eA.ot= AsinM. l.1111ed
around their orbits, and in this case the mutual perturbations are not particularly
of I is zero, and the value
The effect of these terms on the true longitude is zero to first-order. The perturba- Iarge. In many casesj however, the average rate of change
of d oscillates about either 0' or 180", depending on the type of commcnsurability.
tion of the distance r arisins from Ae and Ao is obtained from
This type of motion is termed libration. In this example, the libration is about 0',
and we can writc d as
r = a ( . 1 - e c o sM ) , (6.12-'t)
d = () a) - 2(\ )/) = 0' + oscillation. (6.1l-4)
wilhM=\ t;t.
-
Difierentiating Equation 6.12 7, we have At a conjunction of the satellites, ) I' = 0, and so ,\ a rvill be close to zero also,
and hence all conjunctions occur near the pericenter of thc inner satellite orbit. As
Lr= acosMA,e - aesinML.at might be expected, this particular type of resonance has a significant perturbing
( 6 . 1 2h )
, efiect on the longitudes of the two satellites, and on the longitude of pcricenter
and eccentricity of the inner satellite. If the libration amplitude is small, then the
It is the normal practice to ignore these terrns in e and a, and to determine thc libration can be represented adequately by a single periodic term: e g,
value of the semi-major axis from observations that wiil absorb the constant tcrrn.
This must be allowed for when computittg the mass ol the planet from Kepler's iarv. 0 = 0' + B sin({Jt+ e). (6.r3 5)

Ifthe amplitude is large. e.g., for thc Titan-Hyperion and Nlimas Tethys resonances,
6.13 Perturbations
dueto Commensurabilities then a Fourier selies is needed to represent the libration. The anrPlitudc B anJ
phase e of the libration are the arbitrary constants associated with e and a The
There are many instances of pairs of satellites with mean motions. r and z'. having Iibration frequency B is a function of the masses of the satellites, the closerressr'f
a relationship of the form the commensurability, and the libration amplitudc if it is large.
( 6 . 1 3 - l) The principal effccts of this resonanceon the orbit are to causc perturhatiotts
of the longitudes of the satellites with the period of the libration, and to cause
where p and 4 are two small integers. Such a sitriation is called a conrmensura- a forced componcnt of the eccentricity, rvhosc magnittrde depends on fhe rnass
bilitv of mean motions or, more simply, an orbital resonance) because it causes of the pertrrrbing sateliite and the closenessof the commensurabilitv This forced
increased mutual perturbations. This resonance eflect is caused by a repeated gco component is usually the largest conponent of the eccentricity'
metric relationship of the positions of thc two satellites (their longitu<les) relatix: The cornmensurabilities that arc of most significancc in thc satellile systems
to a particular point in their orbits. This point is usually the pericentcr of one of are listed in Table 6.13.1. In addition rnanv ol the featrrres of the rings of Saturn are
the satellites, but for Mimas-Tethys it is the midpoint of their nodes orr the equator associateclrvith commensurabilities with satellites. Some of the newly discovered
of Saturn. We shall take as an example the case rvhere the commensurability is small satellites of Saturn are in 1:1 commensurabilities \a'ith the major satellites
These contrnensurabilities cause the small satellites to follow or precede the major
satellites by 60" in their orbits. 'Ihe 2:1 commensurabilities alnong the Galilean
2n' n b =O. (6.13-2)
satellites Io, Europa, and Ganym<:dearc closely associated with the Laplace com-
mensurability that affects all three satellites, as a result of which the long-term mcan
The prime refers to the outer satellite, and d is the secular rate of change of the
value of (n1 3n2+24) is exactly zero, and the angular argument (.\1 3)2 +2)r)
longitude of the pericenter of the inner satellite, caused by the oblateness of the
ooz EXPLANATOF]Y
SUPPLE[4ENT
4 6 / OBBITALEPHEMERIDES
AND RINGSOF SATELLITES 333
Table6.13.1 This effcct is particularly noticeable for Rhea, as the forced eccentricity es is
The PrincipalCommensurabilities
AmongSatellites particular case a1 is the proper
much larger than the proper eccentricity eo. In this
proper eccentricity and
Argument Satellites Libration Amplitude longitude of pericenter of Titan. and ei is related to the
mass of Titan
D,'-),-a Enceladus-Dione 195 In the satellite systems ol Jupiter and Saturn there are several sateilites of large
lo-Europa
Europa-Ganymede arasscompared to that of the other satellites, and so they are liable to cause sizable
D,'-).-a' Enceladus-Dione c forced eccentricities or inclinations on each other's orbits. These forced componcnts
lo-Europa can themselves cause significant forced components on other satellites. atrd thc re
Europa-Ganymede c
4^' 2^ A-91 Mimas-Tethys sulting terms can have frequencies quite diflerent from the natural frequencies of
97"
4\' 9. a' Titan-Hyperion 36" the system due to the oblateness, etc. To compute these perturbations the ei6leq
2^,'_3^,+^ lo-Europa-Ganymede0?07 r.alue theory used for the secular perturbation theory of the planets must be used
2ll/ - 3)'+ ) c
l\,4iranda-Ariel-Umbriel (seeBrouwer and Clemence,1961)
Nole: Circulationmotion is denotedby C. A prime relersto the outer salellileof a
pair orto lhe middlesalellileof a lriplel. Double-primesreterto the outer satetiiteoJ
l ple commensurabililies. RectangularCoordinates
6.15 Planetocentric
The theory of the mc,tion of a satellite gives expressionsfor the osculating elements
as functions of time. In nlany casesit is more convenient to represent the position
has a mean value of 180", and oscillates about this value with amplitude 0:066 and of a satellite in its orbit in rectangular coordinates rather than osculating elements
period 2070 days. Let the osculatingclementsof the satellite be a, e, i, \, o, and f/. Then,

(1) calculate the mean anomaly M tiom

6.14 Long-Period
Perturbations
by OtherSatellites
M=\-a. (6.15-l)
The pericenters and nodes of satellite orbiis move slou'ly compared with the ofbital
speed, and so the relative motion of the pericenters or nodes of a pair of satellites is (2) Determine the eccentric anomaly E from Kepler's equation
slow, and significant perturbations can arise from the arguments a o' ot 9, !2'.
In the planetary system these are called secular perturbations, but thev are in fact (6.1s-2)
E-esinE=M.
of very long period 25,000 years and longer. In the satellite systerns the periods
are much shorter, and the perturbations arc usually represented by periodic terms. (3) Calculate the coordinates t,l in the orbiial plane (where the x-aris is torvard
The magnitude of this efieci depends on the mass of the perturbing satellite and the pericenter) from
the size of its eccentricity or inclination. Hence, for example. in the Saturn systerrr x=a(cosE-e),
only the effects of Titan are significant. Its efiect on the osculating eccentricity e (6.15-3)
and longitude of pericenter o of Rhea are of the form Y = a(l - e211t2
sinE'

Methods for solving Kepler's equation can be found il standard celestial

e cos @ = ep cos ap + €icos ''T
(6.r4-t) mechanics texts.
e sin o = eosin oo + e6sin ra1 (4) Next calculate the coordinatcs xr , )r, zr relative to the reference plane and
x-axis direction used for a particular satellite lf, as in Figure 6 11 1, the
'where orbit is related clirectly to the reference plane by the orbital elements, then
@T is the longitude of the pericenter ol Titan, e1 is a forced component
of the eccentricity caused by its resonance with Titan, and er, uto are the proper x1, y1, 11 are given by
eccentricity and longitude of pericenter of Rhea, so that ao and the phase of op
are arbitrary constants, and the rate of ao is due principally to the oblaterressof
Saturn.
l:ll = Rr( O)Rr ( l)R:(-or)
lill (6.1s-4)
 334 EXPLANATORY
SUPPLEMENT ANDRINGSOF SATETIITES
EPH€MERIDES 335

Direction of
inqeasing declination
(o north)

Ifgul! 6.15.2
of thc sat€lliterelativeto the planer p = position angle,,r= angulardistance,
f = Ac cos 6, and4 = 46. Aa = differencein right ascension(satellitemirus plarct),
Figue 6,15.1
Ad = differcncein de{lination (satelliteminusplanet),andf = declinationof planet,
The satolliteorbit refered to atr intermediateDlane.where
P is the pericenterof the satelliteoftir andy'{= A, + AB

: If the coordinates\, tr, zr a.rereferredto the ecuptic and equinoxof 81950,

the coordinates .r2, y2, a of the satellite relative to the center of ma^ssof the
in the frame of the Earth's eouator and eouinox of 81950. a.re
where q., = o - Q, and, &(a) for i : 1, 2,3 is a rotation through an
l,l = Rr(-e)
angle a about the current r-, y-, or z-axis respectively (see Section 11.41 for
expressions for these rotation matrices). This particular traasformation is
given explicitly as l:) l;il
1_zt
)
(6.1s-7)

e = 23?4457878'1
is the obliquity of the ecliptic at 81950.
rl = (rcoso - ysinor)cos O - (r sin i..r+ycosr..r)sin O cos l, If the coordinates must be referred to the equator and equinox of some epoch
yl = ("rcos@ - y sin cu)cos f,?+ (x sin@+ y cosq./)cos f2 cos i, (6. l5-5) than 81950 (e.g., of date or of J2000), then precession should be applied to
21 = @sin c..r
+ycosr..,)sini. y2,a2.T}lLeprecessionmatrix from 81950 to J2000is given in Sectiou3.591.
We now,consider the appareut position of the satellite as seenby an observer
In a more complicated case, such as Figure 6.15.1, where the orbit is referred the Earth. Let f = Aocos6 and 4 = 46 be the differential coordinatesof the
to a.n intermediate plane (e.g., the equator of the planet) which is itself referred to in right ascensionaud declination respectively(see Figure 6.15.2).The
the reference plane a.ndx-axis direction. The coordinates,rl, are in the sensesatellite minus centerof planet' Then, to first order,
)1, zt are given by

{ = (y2coso --r2sino)/r,

f;11 = R3(-Oc)Rr(-,e)R3(-rV + Oe)Rr(-7)R3(-r.l)

f'l , (6.15-6) 4 = (22cos 6 -.r2 sin6 cosa - )z sin 6 sin a) / r,
(6.ril)

lzrJ LNJ o a.nd 6 are the right ascension and declination of the planetl r is the distance
where i. a.nd J7. a.re the inclinatiou and node of the intermediate pla.ne, and and the planet from the Ea,rth, in the same luaits a.sx2, y2,zz (usually eu); aud f aad
7
lV are the inclination and node of the orbital plane, as shown in Figure 6.18.1. in radians.
 336 ExpLANAToBy
suppLEMENr
The position angle p and angula.r distance s of the satellite relative to the planel
axe given by
North pole of
tanp=€/n, Earth'sequator
(6.15-e)
s=((2 + n2)rt2

iv - 90'
6. t 6 TheApparentOrbit

The apparent orbit of a satellite on the geocentric celestial sphere is an ellipse that
is the orthogonal projection, in the direction of the line of sight, of the actual orbit
ir space. In a circula.r orbit, the orbital diameter that is perpendicula.r to the line of
sight is projected into the major axis of the apparent ellipse. At its extremities, the
satellite is at its greatest elongations from the primary. In planning observations
of satellites with short-period orbits, times of greatest elongations are rnore useful
than difierential coordinates.
At times of greatest elongation the satellite and planet center are separated by
an appa,rent angular distairce a I A, where a is the apparent semi-major axis in arc
seconds at a distance of one astronomical u.nit and z1 is the geocentric distance of
the primary. The orbital diameter that lies in the plane formed by the line of sight
and the normal to the orbital plane, projects into the minor axis. At the further
extremity of this diameter, the satellite is in superior geocentric conjunction with Figure 6,16.f
the primary, and at the nearer extremity it is at inferior conjunction. The ratio of Planetocentriccelestialsphere.The Earth hasright ascension,180' .r a, and
the semi-minor axis to the semi-major axis is the absolute va.lue of sinB, where B declination,-6.'l .r = 180' + o andy'r = 180"+ d - (N 90') = o - lV - 90"
is the angle between the line of sight and the plane of the orbit.
On the pladletocentric celestial sphere of the primary, the path of the satellite is
a great circle, and the positions of the satellite on the planetocentric sphere at any
time are obtaiaed ftom the orbital elements. To represent these positions, the same
coordinate systems are adopted as those defined on the geocentric celestial sphere
by the equator and the ecliptic. Becauseof the mathematically infinite radius of the
celestial sphere, these reference circles are in identically the same positions on the a-N-90"
planetocentric sphere as on the geocentric sphere. The Earth on the planetocentric
sphere is diametrically opposite the geocentric position ofthe planet, and, therefore,
at right ascension o I 180' and decliuation -6, where o and 6 a,re the geocentric 90'+6
coordinates of the planet.
Referred to the celestial equator, the position of the great circle that the satellite
describes on the planetocentric sphere is represented by its inclination ("/) and the
a-U-180'
right ascension (N) of its ascending node, or by the right ascension (N - 90') and
declination (90' -"/) of the pole of the orbit. The major axis of the appa,rentelliptic
orbit that the satellte describes on the geocentric sphere is parallel to the plane of tr'lgure6.16.2 Figure 6.f63
this great circle. triangleusedto computp Sphericaltriangle usgdto compute
In Figure 6.16.1 the following angles are indicated: ,8, andP oandp-P
338 EXPLANATORY
SUPPLEMENT AND RINGSOF SATELLITES
EPHEMERIDES 339
P: The position angle of the minor axis of the apparent orbit, measured fro6
North. The axis selected is the one pointing toward the pole of the orbit
relative to which the satellite motion is direct.
B: the pla.netocentric latitude of the Earth relative to the orbital plane of the
satellite; positive towaxd the pole of the orbit for which motion is direct.
the geocentric longitude of the planet measured around the orbit of the satel-
lite, from the ascending node on the Ea,rth's equator. Hence the planetocentdc
longitude of the Earth, measured from the same point, is U - 1800.

Flom spherical rriangle QRE shown in Figure 6.16.2 we obtain

cosBsinU = +cos,Icos6 sin(a - 1f) + sin.lsin 6,

cosBcosU = +cos6cos(a- AI),
sinB = + sin"/cos6 sin(o - A) - cos"/sin6, (6.1G1)
cosBsinP = - sinJcos(a* I),
cosBcosP= + sin.Isin6 sin(a - Af + cos./cos6.

The position of the satellite is described by a the pla.netocentric longitude of

the satellite measured in the same way as U and o the planetocentric angular 6.16.4
orbit of a satelliteof the primary P asFojected on the geocentricspherc.This
displacement of the satellite ftom the anti-Earth direction; o takes the value 0o to shows the situation when the Eanh is nonh of the orbilal plane of the satellite.
180", and has the value 180' when the satellite lies between the planet a.nd the
Earth.
tr\om spherical triangle EIS shown in Figure 6.16.3 we obtain suftcient accuracy we may write:

ra- (6.164)
sino sin(p- P) = sin(u- I)),
r=z;i;Tsrno=zsmo,

sinocos(p* P) = sinBcos(r - {-/), (6.r6-2) a is the apparent radius of the orbit of the satellite at unit distarce' ex-
in seconds of arc. Rigorous formulas for p and s in terms of the difierential
cosa = cosBcos(z- {4.
of the satellite, and also in terms of the planetocentric coordinates, are
later.
Let r be the distance of the satellite from the planet, 4 be the distance of the planei At greatest elongation, u-U =! 90", o = 90', a.nd s = alA in position
from the Earth, 4" be the distance of the satellite from the Earth, and s be the the e p = P ! 90'. At the extremities of the minor a-ris of the apparent orbit, the
geocentric distance of the satellite lTom the planet. r, A, arrd 4. a,re measured in is at inferior or superior conjunctionl n - U is 0' or 180o, o = B' and the
units of length, and s is measured in units of arc. tion angle is P or P + 180'. Evidently, when the Earth is north of the orbita'l
Then, from the plane triangle formed by the Earth, the planet, and the sateUite, the satellite is at inferior coniunction at the southern extremity of the minor
;and is at superior conjunction at the northern extremity (see Figure 616'4);
positive if the motion is direct, negative if retrogra.de. When the Earth
4" sins = r sino, of the orbital plane, superior conjunction is at the southern extremity of
(6.16-3)
4"coss=rcoso+6. axis, and B is negative in direct motion, positive in retrograde motion'
340 EXPLANATORYSUPPLE[,{ENT SN D B I N G So F S A T E L L I T E S
E P H E M E R I D EA
5 / OFBITAL 341

Irrespective, therefore, of whether the motion is direct or retrograde, the positiori 6f the satellites is TDB. The timescale of thc ephemerides
it The Astronomi.cal
angle at superior conjunction, when B is positive, is p = P; q'hen B is negative, it is is UT. The value of TDB - UT adopted lor The Astronontical Almanac
Alrnanac
P: P+ 180". ftom 1990 2000 is apProximated bY
The apparent orbit becomes increasingly elliptical as the Earth approaches tlre
orbital plane, and reduces to a straight line when the Earth is in this plane. As TDB U T = 5 6 i 0 5 9 + 0 i 2 8 2 2 0 )- t 9 8 8 . 0 ) + 0 " . 0 2 2 2 3 0 1 9 8 8 . 0 1 r . (6.17-l)
the Earth passesthrough the plane and B changes sign, each superior and infelior
conjunction occurs at the opposite extremity of the minor axis from that at q'hich it
previously took place. Similarly, when the Earth is in the plane that is perpendicular 6.171 fimes of Gr€atest Elongation The method for calculating the inclination,
to the orbital plane and contains the celestial pole, the minor axis is exactly in thq orbital longitude, and longitudc of ascending node rvith respect to the Earth's
north-south direction, and the major axis is exactly east-west. As the Earth passes equator, "/, a, and N, can vary from satellitr: to satellite. The descriptions for these
through this plane, the extremity of the major axis that formerly rvas the morg calculations are contained in Sections 6.2 through 6.7. Given that "/, tr, and N have
northerlv becomes the more southerly. been computed for 0n on a given date. and that the right ascensionand declination of
Only when "/ is not too greatly difierent from 0o or 180' is the directiorr of the primary, a and 6. are known, U, B, and P arc calculated from Eqlrations 6.16 1.
The angle to the nearest elongation is given by
the minor axis necessarily nearly enough north and south for the elongations to
be strictly and unambiguously described as eastern and &'estern. When "/ is in 1,hc
neighborhood of 90', as in the case of the satellites of Uranus and Pluto, the dirt:c- d u = 9 0 "- ( u - U). ( 6 . 1 7 11 )
tion of the minor axis on the celestial sphere ranges from north-south to east $,est.
This introduces confusion in the terminology for the elongations, but in general The time for the satellite to transit this angle is
they are more appropriately regarded as northern and southern than as eastcrn
and western. dt=duln. (6.r'tt-2)
In the ephemerides for flnding the apparent distance s and position angle p. the
factor sin o giving the ratio of s to the apparelt distance at greatest elongation is If dr is greater than one day, -/, ir and N, and then U. B, and P are redetermined for
denoted by F, and therefore the apparent distance of the satellite from the prin:u v 0h on the next day. A new dr is then computed. The light time correction 01138614
is eiven bv is then applied to givc the time of eiongation. z1 is the Earth planet distance.
s=FalA. (6.l6-5)
6.172 Apparent Distance and Position Angle Given "/, r.t, and N on the date
With Pq denoting an arbitrary fixed integral number of degreesnear the r'alue of P of opposition, P is determined from Eqrrations 6.16 1. P6 is the integral part of
at oppositiol, the value ofp at any time is expressedin the form p1 +p2. p] is the P. Now from "/, a, and N at Ohon anv date. U, B and P are computed from Eqtra-
sum of the approximate position angle at elongation and the amount ol motion it tions 6.16 1. p - P and sino = F are found from Equations 6.16 2. p1 is the sum
position angle since elongation. p2, depending on the date, denotes the corrcction ofp - P and P6. p2 is the difTerenceP - Pe, where P is the value for 0n on the date
P - P6. In calculatingF and p1, the value of the eccentricityof the apparcnt orbit at in question, not the value at opposition. The determinatiorr of a f A is a straight-
opposition is used; conseqrrently,in the values of s and p that are derived from therD forward computation from the semi-major axis, tt, and thc Earth planet distance,
the effect of the variation of the eccentricity of the apparent orbit is neglected. A.

6.17 CalculatingTabulated Values 6'173 Differential Coordinates To first order. the differential riglrt ascensionand
declination in the satellite minus planet are
The apparent positions of the satellites are represented by their positions relativtr
to the primary, expressed either by the apparent angular distance and position
angle, or by the differential spherical (or rectangular) coordinates in right ascensiou ul<r=ssinpsec(€+d0),
(6.r73-1)
and declination. The tabular values are corrected for light time, and are directl]- AP = s cosP'
comparable with observations at the tabular times. It is usually sufficiently acculatu
to use the light-time from the planet to the Earth. The timescale of the theories vthere.tis calculatedfrom Equation 6.16 4 arldp = pt + p2.
342 EXPT.ANATORY
SUPPLEMENT ANDRINGSOF SATELLITES
EPHEMERIDES 343
6.18 Notation Phobos

Many different notations have been used in the literature for the orbital elements a = 6.26974x lO-5 tu
of the satellites. The notations for the a.ugula.r elements used in the bala.nce of this / day
n= 112898,14556
chapter are described below.
=
e 0.0150
For those satellites for which the orbit is referred directly to the ecliptic and
equinox of 81950.0 or J2000.0 the notation is 7 = 1:10
0 = 327?90- O?43533tl
.l = mean longitude,
=
L 232?4l + nd + A?CfIZ4f
a = longitude of the pericenter,
,f2= longitude of the ascending node, P = n8?96 + O?43526tt
i = iaclination. N'=47?39 -O?O0l4y
J, = 37?27+ O?0008y
Where an intermediate reference plane is used (usually the equatorial plane of
the planet or the Laplacian plane) the uotation used for the angular elements is

I = mea,n longitude,
P = longitude of the pericenter, a = 1.56828x 10-aeu
d = longitude of the ascending node, n=2853161888/day
? = ilclination.
e = 0.0004
When the orbital elements a,re referred to the Earth equator a,nd equinox of 7 = 1?79
81950.0 or J2000.0, the notation used for the angula.r elements is 0 = 24O?38- 0:01801d
L = 28?96+ nd - O.27sinh
I = mean longitude,
ur'= longitude of the pericenter, P = rrr?7 + 0?01798
d
N = longitude of the ascending node, h=196?55-0:01801d
./ = incliuation.
N. = 46?37-O.Nr4y
The reason for using the ecliptic as the reference pla.ne for the orbits is for "I"=36362+030008y
conveniencerather than any good dynamical or observational reason. The planes of
dynamical signiflcance are the equatorial a.ndorbital planes ofthe planet. The plane d=JD - 2441266.5'
of observ"ational significa.nce is the Earth's equatorial plane. The orbital elements of y = d 1365'25.
the planets a.re, however, referred to the ecliptic, and so it was probably a convenient
frame in which to calculate solar and planetar5r perturbations. ; These elements a.re referred to the fixed Laplacian planes of the satellites' The
elements a,re shown in Figure 6.2.1' The orbital elements a're defined as
the semi-major axis a; the mean daily motion n; the eccentricity e; the
6.2 THESATELLITES
OF MARS 'y; the longitude of the ascending node, d, measured from the ascending
of the Laplacian plane on the Earth's equator along the Laplacian plane to
Ephemerides of the greatest eastern elongations, and tables for determining the node of the orbit; L, the mean lougitude reckoned from the vernal equinox along
approximate apparent distance and position angle are given for the satellites of Earth's equator to the ascending node of the Laplacia.n plane, then along the
Mara it The Astronornical Ahnanac, The ephemerides of the satellites are computed lacian pla.ne to the ascending node of the orbit, and then along the orbit to
from elements given by Si.nclair (1989). mdan Dosition of the satellite; P, the longitude of the pericenter, measured in
 344 EXPLANATORY
SUPPLEMENT EPHEMERIDES
AND RINGSOF SATELLITES 345
from
u=FC+L-N"-0+(J-It), (6,2-3)
Satellite
orbital (f-M) is the equationof centerthat is computedvia Equation 6.11-1,where
plane
Iaplacian
plane M=L-P. (624)

Mars computation of the tabulated values follows as described in Section 6.17.

orbital
plane

THESATELLITES
OF JUPITER

The GalileanSatellites
Eanh equator
ephemerides and phenomena of Satellites I-[V given in The Astronomical AI-
are basedon the theory of Lieske(1978),with constantsdue to Arlot (1982).
theory is a revisionof the theory by Sarnpson(192f). A shortenedversion
Lieske's theory is presentedbelow, which will give the positions of the satellites
Flgure 6.2.1
Refqercc syslemfor PhobosandDeimos.D = position of satellitcpericenter,d = Ar, about 500km. The constants used below a.refrom Lieske (1987):
L=TA+AB +BC+ CD+ meananomaly, P =TA+AB+BC+CD,1V" = TE, and
Iv = TF. lO' $ = -41279 - 2L);
"os(ZLy
10?v1= -5596 tin1p, - Pr) - 2198sin(P| + P2 - 2IIr - ?G)
the sa,me manner as Z fiom the vernal equinox to the pericenter; N", the lougitude + 1321sin(dr)- l157sin(Zr-2L2+P4)
of the ascending node of the Laplacian plane reckoned from the vernal equinox - l9zl0 sin(trr - zla + P) - 791 sin(Lt - 2L2+ Pz)
along the Ea.rth's equator; Jo, the inclination of the Laplacian plane to the Earth's + 791sin(Zr - 214+ P) + 82363sir(2Lr - 2ld:
equator; ft, is reckoned along the Laplacian plane frorn its ascending node on the
107(=1 +7933r 1 n 1 1-1d 1 + v 1 ) + l 8 3 5 s i n ( Z-r0 2 + v ) i
orbital plane of Ma.rs to the ascending node of the oatellite orbit.
Though the theories for the motion of the satellites contain several periodic 107{2= -3137"ot1La- P) - 1738cos(Lz - Pa)
terms, the only periodic perturbation of importance for Earth-based observations + 93748cos(Lr- h\;
is I = -0. 27 sin ft for Deimos. 107v2= -1159.in1-2It + ztL)+ 1715sin(-Zllt + fu +i! - 2G)
J and N, the inclination a.nd ascending node of the satellite orbit on the Eaxth's - 1846sin(G)+ 2397sin(P: - Pa)- 3 172sin(6)
equator can be fouad by considering the spherical triangle EFC in Figure 6.2.1. Flom
- 1993sin(t2 - b) + l844s'n(la - P2)
spherical trigonometry we have
+ 6394sin(Ia - P) + 345lsin(la - Pa)
sin.f sin(1v- Na) = sin'Ysin d, +4159sin(lr - 2la + Pi + 757| sn(Lr - 2Lz+ Pi
(6.2-t)
sinJ cos(N- lva) = cos'y sin"Ir'+ sin ? cos./acos,. - 1491sin(Zr -214+P)- 185640sin(Zr-la)
- 803sin(Zr - L) + 915 sin(2Lr - 2ld;
Now find the arc length FC of triangle EFC from
107(2= .131575tin (14 - 02+ v) + 4512sn(la - 0g+ vz)
sin"/ sinFC = sinJ, sin I, - 3286sin(Il - i! + v);
(6.2-2)
sin"IcosFC = cos"/osin ? + sinJacos? cos0. 107{3= -14691"o(I4 - P, - 1758cos()h -)Li

t*tu
 346 EXPLANAToRY
SUPPLEMENT
AND FINGSOF SATELLITES
EPHEMERIDES 347
+ 6333cos(la - h)i Lt, = 84?455823000+ 2l:57107 087517961t,
10?v3= -1433 si(2111+ 2t!) + 411sin(-h + tP)
h = 184?415351U&+ 0?17356902t,
+ 346sin(-qa+V) - 2338sin(G) Pt = 82?380231O00
+ 0?16102275t,
+ 6558sin(Ps- PD + 523 sln(Pr + P3 - 2nr - 2G\ Pz = 128?960393000
+ 0:04645&4 t,
+ 314 sin(O) - 943sin(h - La) P: = 1879550171000+ 0:00712408t,
+ 29387sin(\ - P3)+ 15800sin(23- Pa) P4 = 335?309254000+ 0:00183939t,
+ 3218sin(ZLs- zLD + 226 sin(3h - zLa) IL = 13?470395000,
- 12038sin(I,2- \) - 662sin(L1- 214+ P4)
= 3O8?365749N0- 0:13280610/,
- 1246srl.(Lt - 214+ P) + 699sn(Lt - 2h + P2) - 0?03261535
0z= 1003438938000 t,
+ 2!7 sin(Lt - h): 0s= 1189908928000- 0?W7r7678t,
1d 6 = -2793 tio6t - e2+ v) + 32387sn(\ - fu + 4) - 0:00176018t,
0c = 322?746564000
+ 6871sin(l: - 0a+ 4) - 16876sin(Z:- V + 4); I = 3163500101000- 0:0000028 t,
107{a= a1656"o.gr - \) - 73328cos(La- Pa) Gt = 3l?9785280244+ 0?033459733896
t,
+ 182cos(I,+- II) - 541cos(h + Pt - 2IIt - 2G) G = 30?2380210168+ 0?08309256r789694s3
t,
- 269 cos(2Lt - 2PD + 974 cos(h - h):
6z = 52?1445966929
,
107va= -467 5in1- 2Pa+ 2\!) + 309 sin(-2Pa + 0a+ V)
t= JD - U43ON.5. (6.31-2)
- 4840sin(-2Zy + 2V) + 2074sin(- +V)
- 5605sin(G)- 204sin(2G)
P1, and, fi a.re the mean longitude, proper periapse, and proper node of
- 495sin(5G' - 2G + 6) + 234sin(P+- II) i. /1 is the libration phase a.ngle;Zy is the longitude of perihelion of Jupiter;
- 6112sin(Pz- Pq)- 33l8sin(la - Pz) the longitude of the origin of the coordinates (Jupiter's pole); G/ and G are the
anomalies of Saturn and Jupiter; /2 is the phase angle in solar (A / R)r with
+ 145573sin(Lt - Pa)+ 178sin(Z+- IIt - G)
se -2G.
- 363sin(Ia - ZJ) + 1085sin(Za+ Pt - 2h * 2G)
To convert {, v, and ( to recta"ngularcoordinates i, y, and Z in the moving Jovian
+ 672 sln(2Lt - 2PD + 218 sin(2Lq- 24 - 2G) ftame. Equations 6.31 3 and 6.31-4 can be used. Satellite subscripts have
+ 167sin(2La- 0c - iP) - l42srn(2Lq - ?V) omitted for clarity.

+ 148sin(Zr - 2Lq+ PD - 390 sin(Lz* Li

- 195sn(2Lt - 2la) + 185sin(31:-7Lq+4Pq); i=a(l+e)cos(L-V+r),

lOj(a = a773sin14- 2Il + V -2G+va) y=a(1 +€)sin(Z- V -v), (6.31-3)

- 5075sin(Zr - fu + v) + 4430Qsin(La- 0a+ va) 2= ae.
- 76493sin(lq - if + v); (6.31-r)
coefficients a a,re
where
ar = O.00281934'l
tv
Zr = 1063078590000
+ 203?4889553630643
t, a2 = 0.0044858?2AU
Ia = 175?'133787
0N + 101:3747245566U5t, as = 0.W155352 tu
L: = 1201561385500
+ 50?31760915340462t, q+= O.OL2585436l-u.

'{4qb.!
 w
348 EXPLANATORYSUPPLEMENT 6 / O F B I T A LE P H E M E R I O EASN D
R I N G SO F S A T E L L I T E S 349
The formula for finding the coordinatesin the Earth equatorial frame of 1950.0is For Satellite IV, none of the phenomenaoccur $'hen the plane of the orbit of
the satellite essentially the same as that of Jupiter's equator is inclirred at more
r = Rr(-e)Rr(- O)Rr(-"/)Rr(-O)R3(-4i, (6.31*4) than about 2' to the line from Jupiter to the Earth (for occultations and transits)
or to the Sun (for eclipscs and shadow trarsits).
Owing to the finite disks of the satellites, the phcnomena do not take place
where
instantaneously the times refer to the center of the disk. In ccrtain favorable
situations of the orbital planes of the satellites relative to the Earth, one satcllite
- .---.7 pay be eclipsed by the shadow of another, or may occult another. No prcdictiorrs
r = (x.y. z,
1 = (x,1,z)r of these phenomena are given in Thc Astronomical ALrnanac, but predictiorrs arc
given h Handbook B.A.A. at the relevant times. A description of the rlethod of
e = 23" 26' 44i84
prediction is given by Levin (1934).
a = 99?99754 The configurations of the four satellites relative to the disk of Jupiter are shorvn
I = r?30691 in The Astronomical ALmanacin graphical form, on the pages facing thc tabrrlar
@=A - A ephemeridesof the times of the eclipses and other phemomena.

1= 3110401.
6.32 The Fiith Satellite,Amalthea
e is the obliquity of the ecliptic; O is the longitude of the ascending node of Jupil cr:
The data given in The Astronomica,l ,Alrnanoc consist of the Universal Tiurcs ol
J is the inclination of Jupiter's orbit to the ecliptic; 1is the inclination of Jupiter's
every 20th eastern elongation. The elongations are comprrted from the following
orbit to the Earth's equator. The rotation matrices R; are defined in Section 6.15.
circular orbital elements, relerred to cquator of .lupiter. given by van \Voerkom
The data tabulated in The Astronomical Almanac consist of: the approxirnate
(1e50):
times of superior geocentric conjunction; the times of the geocentric phenomena;
and the approximate configurations, in graphical form, of the satellites relativc tcr
to = 1903September1.5UT : JD 2116359.0:
the disk of Jupiter.
The Universal Time of each superior geocentric conjunction is given for each
a = 249!55',
satellite to the nearest minute. The phenomena for which times are givetr are 1 = 24'.1,
eclipses, occultations, transits, and shadow transits. The UT of the beginning and 0 =82?5- 9r4?62t1
end of each phenomenon (disappearance and reappearance for eclipsesand occull a- n = 722?63175
I day
tions, ingress and egress for transits and shadow transits) are given to the nearest
lo = 194:98;
minute for all phenornena that are observable. When Jupiter is in opposition, tbe
shadow may be hidden by the disk and no eclipses can be observed. In gencral, dfu = 03113 0:0076/ + 0:00035rr;
eclipses may be observed on the western side of Jupiter before opposition ancl orr r = (JD, tdt365.25, ( 6 . 3 2 -)l
the eastern side after opposition. Before opposition the disappearance only of Satel-
lite I into the shadow may be obsen€d since it is occulted before it emerges frolr a is the mean elongation at unit distarrce; 1, the inclination to the ctluator of
the shadow; after opposition only the reappearances frorr the shadow are visiblt:. Jupiter; d, the longitude of thc ascending node; ru,the mean motion pcr solar day;
The same is true in general ol Satellite II, although occasionally both phenomcna {o, the mean longitude at epoc}r t6; and 6t9, the correction to mearr longitude.
can be seen. In the case of Satellites III and IV, both phases of the cclipses arrr I - 16is reckoned in Julian years. The longitudes are measured in the plane of the
usually visible cxcept near certain oppositions. Similarly the occultation disappcar'- equator of Jupiter, from the ascending node ol thc mean orbital plane of Jupiter on
ances and reappearanccs of a satellite cannot be observed if, at the time concerncd, the plane of the equator of Jupiter as shown in Figure 6.32.1. From these orbital
the satellite is eclipsed. For Satellites I and II thcre are, in general, cycles of six etements:
phenomena consisting of both phases of both transit and shadow transit, one phasr: siderealpcriod 0949991083 = 111997860,
of the eclipse, and the other phase of the occultation. mean svnodic period 0149996851= 111999244.
 350 EXPLANATORY
SUPPLEMENT EPHEMERIDES
AND RINGSOF SATELUTES 351
sinJcosly'= +sin 60sin? + cos60cos7 cos(0- 4), (6.32-3)

0-4-the arc alongJupiter's equatorfrom its nodeon the Ea.rth'sequator

node of the satellite's orbitis given by

0 - a--219?8- 2'.54057487(t
- to)
(6.324)
= 219?8- 2?50405737
d,

d is the numberof daysftom the epoch(JD 2416359.0).

the quantities N, ,f, and ry' are calculated from the preceding equations at
at any date, and axe then used to form U (B and P are not required) from
6.16-1. The orbital longitude, z, measured from the node on the Earth's
, at 0i on that date is derived from

u=h+nd-0+4,+614,
(6.32-5)
= { + 112?35
+ 2'A003567
d.

Figulc 6.32.1 the times of the elongations on that date are

Refercncesystemfor AmaltheaandThebe-A = 180" - AB,0 = 180" +AC,
e - A = AC, tlt = DC,lt/= TD, andoo ='fA - 90o
eaaternelongation 0lO333n(U + 90o-r.,)+0i138614
westernelongation eOT3n (U + 2700- lr) + 0 13861^4

In determining these elements, wn Woerkom adopted Souillart's elements of the which ^A is the geocentric distaoce of Jupiter. The terms in 4 a.re the correc-
equator of Jupiter, but for the calculation of the elongations in The Astronomical for light-time; for strict accuracy they should be interpolated to the times of
Almanacthe right ascensionand declination of Jupiter's pole of Davies, et al. (7989) elongation which the first terms represent.
are used: The calculation of €longation time need be made for only a few of the ephemeris
ao =268305-0:009I the others being obtained by mea.ns of multiples of the period. For strict
(6.32-2)
6o= 64?49 + 0'003 T, , the calculation should be repeated with the values of the quantities at
calculated times instead of at CF;but for the tabular accuracy of CDl, this is
where o6 and 6s are referred to the equator a.ndequinox of J2000.0

T = (JD _ 2451545.0)
r 36525.
The FourteenthSatellite,Thebe
The right ascension, N, of the node of the orbit on the Ea.rth's equator; the times of every 10th easternelongationfor Thebe a.representedin The As-
inclination, J, of the orbit to the Ea.rtb's equatorl and the a.rc, ry', along the orbit
Alrruanac.The orbital elements a.re taken from Synnott (1984) and are
ftom the node on the Earth's equator to the node on Jupiter's equator are given by
to Jupiter's equatorand autumnal equinoxof 1950.0.For the calculationof
ephemerides,the inclination of the satellite orbit to Jupiter's equator is ignored.
sinJ sin(N- oo) = +cosd0cos7 + sinr0 sin?cos(A- .4), The elements at eDoch axe
sin"Icos(1v- o0) = - sin r sin(d - ^4),
cos,l = + sitr60cos? - cos66sin 7 cos(o - 4), to = JD2443937.817065,
sin.Isind = +cos60sin(o- A), a = 305i95,

$r,:
 352 EXPLANATORY
SUPPLEMENT EPHEMERIDES
ANDRINGSOF SATETIITES 353
? = 0.015, of the satellite-perturbingbody distancel4" is the satellite-perturbing
n = 533?7l day, distance;4i is the Jupiter-perturbing body dista.nce.Planetary coordinates
from the DE200. The reciprocal masses,in solar mass units, and the "/2
7 = 0:8, usedin intecrationsare as follows;
h = 3s4?3,
Po = 344?0, (6.33_l) Su:r 1.0 Mercury 6023600.0
Venus 408523.5 Eaxth 332946.038
wherec is the semi-majoraxis at unit dista.nce;e is the eccentricitylz is the mean Mars 3098710.0 Jupiter 1047.35
daily motion; 7 is the inclination with respectto Jupiter's equator;Ia is the mea.1 Saturn 3498.0 Uranus 22960.0
longitude reckoned from the Jupiter's autumnal equiaox; and P6 is the longitude Neptune 19314.0 Pluto 130000000.0
of perijove reckoned aloug the Ea.rth's equator to Jupiter's autumnal equinox and J2 of Jupiter 0.01475
then along Jupiter's equator to the perijove (seeFigure 6.32.1).
The times of elongation for Thebe are calculated itr the sarnemamer as those
for Amalthea. In this case the orbital longitude (n), is derived from The starting planetocentric coordinates and velocities from Rohde (tggO) are:

v = la a nd - 0 + r! + (f - li4), (6.331) = 1989 December30

Epoch JD 21147890.5
i' (Au) v (Au/day)
where (/ - M) is the ecuation of center that is computed from Equation 6.11-1,
x -0.039,t627l36618 +0.00'13683275790
where -0.0722678670569 -0.0002593437765
v
M=h+nd-Po. (6.33-3) z -0.0170810682702 -0.0010464833518
Elara x -0.0619697i{57552 +0.0004i}92130964
v -0.0044124460A79 -0.0020946808809
6.34 The SixththroughThirteenthSatellites z +0.026946659756 -0.0003974001515
Differential right ascensions and declinations of Satellites VI-XIII are presented Pasiphae x -0.04747221
440A8 -0.00101825s7362
in The Astrvnomical Almanac. The coordinates v -0.10008854i)0216 +0.001| 925451174
are referred to the equator and
z -0.0009486933348 +0.0009171895371
equiaox of J2000.0. They a"re computed with a va"riable-order, variable-stepsize
numerical integrator developed by Shampine and Sinope x -0.0579232252532 +0.001'101u1867506
Gordoa (1975). The equations of
motion in .r, y a,nd z a.re: v +0.1464131750776+0.0002877970106
z +0.0885266909368 -0.0003615702389

#=-trl,-Hoo-',]
.*"? -u"' Lysithea x {.06€8709/0421 2 +0.0006111658964
v -O.O1'l4622.412925 +O.@1
2&69404782
z -0.0304090473403 {.0013,t66750538
Carme x -0.1221787634075 -0.0007075200606

#=?1,-fi,,,-,1 6
(6.34-l)
Ananke
v
z
x
-0.10481364978s5 +0.0009285971336
-0.08€6824702747 +0.0001468741809
+0.12441797,€900 -0.0006562509816
I
E=-il'-"ft,'r'-rtJ.D'u'f- nr'
&z M,z l. 3Jzt v -0.0256709692336 -0.001351,1562258
z -0.02038554t1615 10.0002939680931
L€da x +O.O3560tl434i(xl +0.0013645601411
z v -0.0810815597479 +0.0006286869081
l"=;,
z +.O.m,52592795262 +0.0005799993600

where My is the mass ofJupiter; the M1's are the massesofthe perturbing bodies; 4 is
the ra.dius ofJupiter; rjs is Jupiter-satellite distance; .1, n, and zr are the rectangular coordinates are computed ftom Equations 6.15-8.

{tt.
 354 EXPLANATORY
SUPPLEMENT ANDRINGSOF SATELLITES
EPHEMEBIOES 355
6.4 THE RINGSANDSATELLITES
OF SATURN the inclination, "I, to the mean equator of date a.nd the right ascension, N, of
node mea.sured ftom the mea.n equinox of date are obtained from
6.41 The Ringsof Saturn
J=90'-60
The rings of Saturn lie in its equatorial plane. The inclination and ascending nodq (6.41-2)
N=oo+903
of this plane on the Ea.rth's equator axe determined from the north pole of Satu11
as defined by Davies ef ol (1989). The dimensions of the outer ring atrd the factors
for calculating the relative dimensions of the rings a.reftom Esposito et ot. (1984).
axc d from the ascendhg node of the ring pla.neon the Ea^rth'sequator along
The outer, inner, and dusky rings correspond to the A, B, and C rings, respectively. ring plane to its ascendiugnode on the ecliptic is obtained from
The Cassini division lies between the inner and outer rings.
The ephemeris of the rings contains the following quantities that determine the sinNcot d = cosNcos€ - sin€cot"/, (6.4t-3)
Saturnicentric positions of the Earth a.nd Sun referred to the plane of the rinss.
upon which the appea.rance of the rings depends: e is the mean obliquity of date. From the elements referred to the equator of
Earth, and from the geocentric equatorial coordinates of Saturn, the ephemeris
U = the geocentric longitude of Saturn, measured in the plane of the rings , B, and P is calculatedby Equations6.16-1.Thesequantitiesa,redefinedwith
eastward from its ascending node on the mean equator of the Earth; to the ring pla.ne in the same way as with reference to the orbital pla.ne
the Saturnicentric longitude of the Earth, measured in the same way, is satellite; but the efiect of nutation must first be removed from the apparent
U+ 180". ascension and declination of Saturn, to refer the position of the planet to the
B = the Saturnicentric latitude ofthe Eaxth referred to the pla.neof the rings, equinox a,sthe elements of the reference plane.
positive towa.rd the north; when I is positive, the visible surface of the Ut , B' , arl.dP' may be obtained by formulas exactly analogous to Equations 6.16-1
rings is the northern surface. the ecliptic elements of the plane of the rings a.nd the heliocentric longitude,
=
P the geocentric position a.ngle of the northern semi-minor axis of the and latitude, D, of Satun, referred to the mean equiuox of date:
apparent ellipse of the rings, measured ftom the north toward the east.
7 = the heliocentric longitude of Saturn, measured in the plane of the rings cosB/ sinP' - - sini cos(,- O),
eastwaxd from its ascending node on the ecliptic; the Saturnicentric lon-
cosS'cosP = +cosicos, + sintsinAsin(,- O),
gitude of the Sua, measured in the same way, is U/ + 180'.
8'= the Saturnicentric latitude of the Sun refened to the plane of the rings, sind = - cosi sinD + sini cos, sin(,- O), (6.4t4)
positive towa.rd the north; when B' is positive, the northern surface of
cosE/sinU' = + sini sinZ,+ cosicosDsin(,- O),
the rings is the illuminated surface.
F = the heliocentric position angle of the northern semi-minor axis of the - O).
cosB'cosU/ = +cosDcos(l
rings on the heliocentric celestial sphere, measured eastwaxd from the
circle of latitude through Saturn. rings become invisible when

The right ascension and declination of the north pole of Saturn are according (a) The ring-plane passesthrough the Sun, since neither side of the rings is
to Davies et ol. (1989), then illuminated.
(b) The ring-plane passes between the Sun and the Earth, since the unillumi-
ar = 40358- 0:036r,
6.41-r) nated side of the rings is facing the Ea.rth.
dr= 83154- 09004Z, (c) The ring-plane passes through the Earth, since the rings axe too thin to be
visible edce on.
where oy and 6J are referred to the equator and equinox of J2000.0 (T = (JD -
245t54s.0)t 36525). during each revolution of Saturn around the Sun-near the times when
Adding the lariatious due to precession, calculated by the formulas in Sec- passesthrough the Sun-the Earth usually crossesthe ring-plane
tion 3.21, gives the elements referred to the ecliptic and mean equinox of date; from times, though occasionally it crosses only once. On 1995 May 22, the Earth

''n',
356 EXPLANAToRY
SUPPLEI,IENT
OBBITALEPHEMEBIDESAND RINGSOF SATELLITES
again crossesthe ring-plane. The rings will remain continuously invisible until l9g5
August 11 while the Earth is on the unilluminated side. The Earth crosses the
ring-plane for a third time on 1996 February 11. The last occasion on which only
one passageof the Earth through the ring-pla.ne occurred was in 1950.
Pericenrcr
6.42 The Satellitesof Saturn
The Astronornical Alrnanac presents times of greatest eastern elongation for Satel-
lites I-V, times of elongations and conjunctions for Satellites VI-VIII, apparent
distances and position angles for Satellites I VIII, orbital elements referred to Sat-
urn's equator for Satellites I VIII, and difierential right ascensionsand declinations
for Satellites VII IX. The orbital elements published are as follows: l, the mean
longitude reckoned from the ascending node of Saturn's equator on the Earth's A Earth equator
equator along Saturn's equator to the ascending node of the satellite's orbit then Equatorof Satum
Orbit of satellite
along the orbit to the satellite; M, the mean anomalyl 9, the longitude of the as-
cending node of the orbit on Saturn's equator measured from the ascending node
of Saturn's equator on the Earth's equator; ?, the inclination of the orbital plane Figure 6,421.1
to Saturn's equator; e, the eccentricity; and a, the semi-major axis. Reference TethysandDione.dr = TB + BC,
planefor Mimas,Enc€ladus,
P =TB + BC +CD,N ='lE,andI,r = TB + BC + CD + meananomaly.
Eclipses, occultations, tra.nsits, and shadow-transits of the satellites occur dur-
ing a limited period each time the Ea.rth passes through the plane of the rings.
Methods for calculating these phenomena are described by Comrie (1934). See also
Taylor (1951).
Ephemerides a.re not provided for Satellites X X\{II though orbital, physical,
a,ndphotometric data are Iisted on pages F2 and F3 of The Astronomical Almanac. The orbital elementsof Mimas are as follows:
Ja.nuswas discovered in 1967. Epimetheus, Helene, Telesto, and Calypso satellites
were discovered by ground-based observations during 1980. Atlas, Prometheus, and d=JD-2411093.0,
Pandora vrere discovered by Voyager. Many of the satellites are coorbital. Janus t=d/365.25,
and Epimetheus follow horseshoepaths in a reference frame rotating at their mean
T=5:0616((JD - 2433282.423)
/36s.25)+ 19s0.0- 1866.061,
orbital speed. Telesto and Calypso orbit a,t lhe La and Z5 Lagrangian points of
a = 0.OOl24l'11^u,
Tethys. Helene is nea.rthe Z4 Lagrangian point of Dione. Pandora and Prometheus
are the shepherd satellites of the F-ring. n = 381?994516/ day,
e = O.O1986,
6.421 Mimas, Enceladus,T€thys,and Dione The orbital theories for Mimas, Ence- 'v = 1?570,
ladus, Tethys, and Dione are based on Kozai (1957). The values of the elements for
h = 49?4- 365?o25t,
Mimas, Encela.dus,Tethys, and Dione are from Taylor and Shen (1988), with mean
motion and secular rates for Mimas a.nd Enceladus from Kozai and corresponding Lt = 128?839+ nd - 431415sinZ - 0?714sin3T
quantities for Tethys and Dione from Garcia (1972). The elements of Mimas, Ence- - 03020sin5 1.,
ladus, Tethys, and Dione are referred to equator of Saturn (seeFigure 6.42I.L), The P = 10?:0+ 365:560/, (6.42r-r)
only significant periodic perturbations on these satellites a.rethe mutual perturba-
tions of Mimas and Tethys, and those of Enceladus and Dione. In each case, they
a is the semi-major axis; n the daily mean motion; e the eccentricity; 7 the
are caused by a close 2:1 commeruurability of mean motions, These perturbatio[s
of the orbital plane to Saturn's equator; d1 the longitude of the ascending
a.fiect the longitude and are denoted by AZ.
11 the nean longitude; and P the longitude of the pericenter.
 358 EXPLANATORY
SUPPLEMENT AND RINGSOF SATELLITES
EPHEMERIDES 359
^t = 0'" 0289'
The orbital elements of Enceladus a,re given by:
= 228" - 30?6197t,
d=JD-2411093.0, Lr = 255?1,183 + nd - 0:88sin(59:4+ 32:73/)
t=d1365.25, - 0175sin(l1992+ 93918t),
a = 0.00158935
au, P = 173:6+ 30:8381t, (6.4214)
n=262?73190521day,
e = O.O0532, the va.riablesare defined in the same manner as for Mimas.
To determine the times of eastern elongation we must first find the orbital
7 = 0:036'
z; the longitude of the ascending node N of the orbit on the earth's
= 145" - 152?7t,
; a.rrdthe inclination of the orbit to the Earth's equator, J. From spherical
Lr = 2OO?155 + nd + 15:38sin(59:4+ 32:73t) we havefrom triangle BFC in Figure 6.421.1,
+ 93:18t),
+ 13:04sin(119:2
(6.42r-2) sini sinBF = sin? sin(Or- g.),
P = 312?7+ 123?42t, (6.42r-5)
sinicosBF = cos7 sinL + sin"ycosiecos(01- O"),
where the 'r,rariablesare defined ir the same manner a,sfor Mimas.
The orbital elements of Tethys a.reas follows: can be solved for 8F'and i. The a.rc length FC can be found from the equation

d=JD-2411O93.O, sini sinFC = sin(Or- Jze)sinL. (6.42r-4)

t = d./365.25,
r = 5:0616((JD -'2433282.423)/ 36s.2s) + 1950.0 - 1866.061, , from the triangle TEF we have
a = 0.001970691u,
sin"/ sinN = sin i sin(Oe+ BF),
n = LX)?6979202781 day, (6.421-:l)
sin"IcosN= cosi sin€ + sinicose cos(A. + Bn,
e = 0.N0212,
1 = 1: ll2l, ca,nbe solved for "I a.ndN. Then the atc EF can be found from
= 1ll?41 - 72?24754
t,
sin,IsinEF = sin(O"+ BF) sine (6.421-8)
Lt = 284?9982
+ nd + 230751sin7 + 0?0341sin3 f
+ 090010sin5 7,
follows from
P = 97"+72?29t, (6.421-3)
u = P - 0r + EF + FC + ft:ue anomaly. (6.421-9)
where the variables a,redefined in the same manner as for Mimas.
The orbital elements of Dione are given by: tabulated orbital elements'v. c. and e are listed above. The remaining elements
be computed. Flom the spherical triangle Tr4-Bin Figure 6.421.1 we have
d.= JD - 2411093.0,
t = tl /365.25, sin,l" sinAB = sin e sin O.,
(6.42r-r0)
a = 0.0025:413,xu, sin "/ccos /-8 = cos € sin L + sin e cos recos J2e,
n = 131?53492N261 day,
can be solved for 4.B.
e = O.0o1715,

*j1,
 360 EPLANATORY
SUPPLEMENT AND RINGSOF SATELLITES
EPHEMEBIDES 361
Then
0=0r-4,+AB,
L=Lr- Q+0, (6.42r_r1)
M=Lt-P.

6.422 Rhea, Tltan, Ilt?erion, and lapetus The orbital theories for Rhea and Ti-
tan a.reftom Sinclair (1977). The orbital elements a.refrom Taylor and Shen (1988)
with the mean motions and secular rates used frorn Garcia (1972). The theory and
orbital elemeuts for Hyperion a.re from Taylor (198a). The theory for the motion
of Iapetus is taken from Sinclair (1974) with some additional periodic terms from
Haxper e, al. (1988). The orbital elements for Iapetus are from Taylor and Shen
(1e88).
Because of the oblateness of Saturn and solar perturbations, the orbit plane of
Rhea precessesat constant inclination around the Laplacian plane of Rhea. Because
we a,re using the ecliptic as the reference plane, this motion around the Laplacian
plane appears as periodic va,riations in ), o, Q, and i . The osculating elements are
given by:
FigL]rc6.422.1
Referercesystemfor Rhea,Titan, Hyperion,Iapetus,and Phoebe.
a=0.00352400,xu, @=
^lF
+ FD and\ = afF + FD + meananomaty.
n = 79?6TJ04OO700
/ day,
7o = 013305'
esin@= 0.000210 sinzr+ 0.0010Osinar,
=
e cosa 0.000210cosfl' + 0. 00100cos@r,
). = 359?4727+ nd
,| t = d.1365.25. (6.422--2)
- lO?2O77
+ ,r sint0 tan;ie sin(356:87 t),
z
r=L_0:0455 c is the semi-major axis; n is the daily mean motion; e is the eccentricity; @
+ ,icos(356:87- l0?2O77r) +0?0201cosN1, ) are the longitude of pericenter and the mean longitude, respectively, reckoned
- 1O?2O'1'1 the vernal equinox along the ecliptic to the ascending node and thence along
A = A, - 0?OO78
+ [rcsinlo sin(356987 t)
i and 'f) a"rethe inclination and ascendins node of the orbit with respect
+ 030201sinNrl / sinr", (6.422-r) ecliptic of 1950.0. The angular elements are shown in Figure 6.422 1'
'The coefrcient
0.00100 in the expressions for esino ar.Ldecosa is a forced
where due to Titan. The algebraic expression for this is given in Sinclair
The inclination and node of Rhea's Laplacian plane relative to the ecliptic
ur= 305' + 10?2077t, ) a r e i = ; - 0 1 0 4 5 5a n d Q = A . - O ? O 0 7 8 .
or = 276?49 + O?5219(JD- 2/11368.0)| 365.25, a,rea number of significant solar perturbations on Titan's orbit ard, as
), 'f2, and i are afiected by periodic variations due to the reference plane
Nr = 44?5 - 0?5219(JD- 2411368.0)I 365.25,
the Laplacian plane. The sola,r perturbations are functions of the mean
n = 57?29578, of the appaxent orbit of the Sun about Saturn. The sola.relements needed
d=JD-2411O93.O, , and i" (which are defined in Figure 6.422.2), and 1", the mean anomary'

\r
 ANDRINGSOF SATELLITES
/ OFBITALEPHEI\,IERIDES 363
362 EXPLANATORY
SUPPLEI\,,IENT
quantities (ie - 0?62M) and (J?"- 0:1418) definethe inclination and node of
Laplacian plane of Titan. The auxiliary angles 0, f, 9, g, and l-are given

Pedcenter cos| = cosiscosia + sinissini" cos(Oa- O"),

sinI sinV = sini, sin(O, - O"),
sinf cos{ = cosissinio - sini, cos4 co(O. - f/'),
sinf sin(o- J?')= sinL sin(J?.- J2.),
MeanPositionof Sun sinlcos(d - J?,)= - sini. cosiu
+cosi" sinrucos(O"- O"),
z"=,\"_(0-a)-4",
g=@^- Qa-V' (6.4224)
To,,
The osculating elements of Titan are given by:

Orbit of Sun
a = 0.00816765 AU,
Flgare 6.422.2 n = 22?57697385 | day,
Anglesneededto evaluatesolarperturbationson Titan. e = 0. 028815- 0. 000184cos28 + 0. 000073cos2(2. - 8),
^l'A
a" = TD,0 =TA +AB,@" = TD + Dc, and), = +,48.
@ = @a+ K[O.OO630 sin2g+ 0.00250sn2(L"- d\
I
\ = 261?3121+ nd + xlsin16ran sin(4| :28
2i"
- 0:5219t) - 0.000176 sinl. - 0. 000215
sin2Z'
+ 0.000057sin(2zs+ g)1,
+ q),
i= iu + 0.00O232ncos(2zs
Approximate values of i, Q, ar.d.o are also needed, and these are denoted by L'
O = l-2a+ 0. 000503rc
sin(21,+ [), (6.422-s)
Qu, and o". These quantities are given by the following expressions:

whered=JD-2411368.0; t=d1365.25i andn = 5'1?29578.

T = (JD - 241sO?0.0) t 36s2s
The algebraic expressions for the coefficients of these periodic terms are giwen
=
ls l'15?4762 + 1221?5515T in Sinclair (1977).
t" = 2?4891+0." 002435I The osculating elements of Hyperion's orbit are given by:
A" = 113?350 0?2s97T
a = 0.0O99040- 0.00003422 cost AU,
\" = 267?2635 + 1222?1136T
t = (JD - 2411368.
O)t 365.2s n = 16?9199514lday,

n = 57?29578 e = 0.lm4l - 0.0ol101cos7 + 0. 00009cos(( - r)

iu = i. - O?62C4
+ n sin092990
cos(41928- 0:5219t + 0.02321cosC - 0. 00009cos((+ z) - 0. 001l0 cos2(

Q^ = Q" - O?1418+ rcsin032990sin(41128- 0:5219r) / sinL + 0.00013cos(31:9+ 6l?7524n,

i=i.-O?747 + O?6200 cos(105:31- 2?3927-)
@A= 275:837+O:5219t (6.422-3)
 3& EXPLANAToRY
SUPPLEMENT ANORINGSOFSATELLITES
EPHEMERIDES .to5
+ 03315cos(38373- 0:53537)
\=76?19854+nt,
- 03018cos(13'+ 24344I),
o = 352?905+ 1l? 65 tc,
O = Oe+ ((-0:061 + 036200 - 2=9Zr)
sin(105931
A = l43?1294- 3?797t, + o: 1164 + O?N8 4. (6.422-:7)
+ 03315sin(38373- 0:5353r)
- 01018sin(13' + 24?447)) / sin(i"- 0.747), auxiliary angles needed to compute periodic terns are calculated as follows:
\ = 176?7481+ d + 0:1507sin(105?31 - 2?392'f)
0 = 4?367- 0?1957,
+ 99089sinr + 03007sin2r - 01014sin3r
I = 1463819 - 33918
r,
+ 03192sin(( - r) - 0:091sin( + 0:211sin((+ r)
- 03013sin(176'+ 12?227-1+O?O17
\" = 26'l?263
+ 1222.11,47,
sin(8'+24?447),
o = 69?993- 18?67027+ 0?1507sin(105931- 2?3927) o" = 9l?796+0?5627,
\r = 261?319 - 2411368.0),
+ 22?576974(JD
- 0947sinr - 13336
sin( + 2916sin2(
ar =277?1O2 - 2411368.0),
+O?001389(JD
+ 0:07 sin(31:9+ 61"7524T"), (6.4224)
O= 60?470+ 1?s2rT'
where @=205305s -2:09rT. (6.422-8)

d=JD-A15020.0, these we compute the following:

T = (JD - 2433282.42345905)
/ 365.2422+ 50.0,
l=),-q,
r = 93?13+ 0?562039d,
e=M8?72_19?184r.
h = )r - 'zr,
Here again the elements are reckonedin the same way as for Rhea. 8=@-Q-0,
The orbital plane of Iapetus is inclined at about 8o to its Laplacia"nplane, 8s=@s-O,
which is itself inclined at about 15' to the equatorialpla.neof Saturn, and 12" to gr=@r-@,
the orbital plane of Saturn. As a result, the motion of the orbital plane of Iapetus
g=@-Q-0. (6.422-9)
is quite complicated, aud is described by cubic polynomials in time tor i and,L
The osculating elements a,recomputed as follows:
ofthe expressionsfor the perturbations involve e and i. For these the following
a"reused.
t=JD-2409786.O,
h = t 136525' e = 0.0288184+ 03000575
t",
(6.422-rO)
T = (JD _ 2415020.O)I 36s2s. - 0995sstc - O?O72ot+ 0:00544.
t = 18345959
The elementscomputedfrom constantand secula,rterms axe:
,following perturbations are to be added to the elements

q = O.O238O984
^u, = 10-5c[7.87cos(2I + 2e - 21"- 28")+ 98.79cos(l+ g - l'r - 8r))
n = 4?537957
ll / day, = 10-5[-140.97cos(8r- gr) + 3?.33cos(21" +zgs- 29)
e = 0.0288184+ 0.000575 r", + 11.8ocos(l+2g - zls - 28J + 24.O8cosl+ 28.49cos(21+g - lt' 8r)
t = 1834s959- 01955stc - O?o72o
t + o. 0os4e, i + 6l.gocos(lr + g.1- g) + 0.496cos(3ls+ 28s- zdl

\"
 AND RINGSOF SATELLITES
oRBITALEPHEMERIDES
367
366 EXPLANAToFY
SUPPLEMENT
bY the formula
Ai = 0304204
cos(2/s+ 2gs+ 0) + 0:00235cos(l+ gr + rr + 8r + d) 6es be determined
+ gr - ,r - 8r + d) + 030005cos(41.+ 2g,)
+ 0300360cos(/
sinlsind = sin"lsin(N- N"). (6.422-17)
+ 090058cos(3/,+ 29")- 030024cos(/,
+ 29,)
A) = -0104299 -
sin(l+ & lr -gr) - 0900789sin(21+2s 21"-2g") length FC from
is found by first finding the arc
- 0i06312sinls - 0300295sin21,- 0302231sin(21"+ 29.)
+ 0100650sin(2/,+ 2g"+ 0) sin FC sin 7 = sin(O - f2") sin i. (6.422-18)

A@ [0:08077sin(er- gr) + 0302139sin(21"

= + 2g" - 2s)
- 0300676sin(l + 21,- 21"- 2g")+ 0?01380sinl J, is then given bY
L= ),- A -FC+0. (6.422-19)
+ 0?01632sin(21+ g - fu - 91) + 0303547sin(/r + gr - gr)
+ 0:00028sin(3rs+ 29" - 2g)l / e The differential coordinates of Hyperion and Iapetus are computed
as follows:
in the orbital plane a.re found
AJz = [0:04204sin(2/"+ 29,+ 0) + 0:00235sin(/+ 91+fu+g+@) the rectangulax coordinates r and y of the satellite
tra'nsformed to the Ea'rth equa-
+ 0100358sin(l + st - h - gt + 6) - 010006sin21" from Equations 6.15 3' These coordinates are then
torandequinoxbyEquations6.ls_4arrd6.15_S.Thedifierentialcoordinates{and
+ 0:0003sin(4/"+ 28")+ 0:0028sin(31,+ 29,) Equations 6 15-8'
4 are then computed through
- 030012sin(/.+ 29,) - 0?0142sinl,l i sinl. (6.422-rr)
axe those
6.423 Phoebe The orbital elements used in ?i'e Astronomical Almdnac
right
To determine J, u, and N, we can use the spherical triangle TEF in Fig- determined by Zadunaisky (1954). The ephemerides consist of difierential
to
ure 6.422.1to obtain ascensionsaud declinations given at 2-day intervals' The elements are referred
the ecliptic a.nd equinox of 1950.0. The values of the elements are as follows:
sinJsinN = sini sinO,
(6.422-12)
N = cosisin e + sini cose cosf2, a = 0.0865752l.u,
sin"/cos
), = 277?872- 0:6541068
t,
which can be solved for N and J. The arc d can be found from the equation e = 0. 16326'
q = 280?165- 0:19586I,
sin"/sinS = sin J?sine. (6.422-t3) i = r'13?949- 0?oz0T,
- o:4r3s3T,
A = 24s?998 (6.423-r)
r then follows from
u=o-Q+6+ttueanomaly (6.422-14)

The tabulated orbital elements a and.e are shown previously. The remaining t -- JD - 2433282.5'
elements are computed as follows. To calculate d we first find the arc AC in Fig- T = t t365.25.
ure 6.421.1. From spherical trigonometry we have for the triangle /EC:
a is the semi-majoraxis; ), the mean longitude measuredfrom the equinoxalong
sin(N - N") cot d = cos(N - Ne)cosJe - sin "/ecot"/, (6.422-15) tlu ecliptic to th" *c.nding node then along the orbit; e, the eccentricitylo' the
longitude of the pericenter-reckonedfrom the equinox along the ecliptic to the
which can be solved for d. M is given by node then along the orbit to the pericenter; i, the inclination of the orbit
the ecliptic; and O, the longitude of the ascending node of the orbit on the
reckoned ftom the equinox.
M=),-4. (6.422-16)
 358 EXPLANAToRY
SUPPLEMENT EPHEMEBIDESAND RINGSOF SATELLITES oov
To compute the differential right ascension and declination, the .r and y cool- + 8.6159x 10 a expi(2.179316
x l}-sT +2.07),
dinates in the orbital plane are found using Equation 6.15 3, where
e = 3.5825x 10-aexpi(-4.782474x lD-sT+0.40)
M=A-co. (6-423_2) x 10-5?+0.59).
+2.9008x 1o-aexpi(-2.156628 (6.5-2)

The coordinates a,re transformed to the Earth equator and equinox using Equa-
tions 6.15-4 and 6.15-5. Finally, { and 4 a.recomputed from Equation 6.15 8.

mlMu=1.45x10-5,
6.5 THERINGSANDSATELLITES
OF URANUS n = 1.51595490
radians/day,
L = 2.28540169
+ 1.51614811
?
Orbital elementsare given for the rings of Uranusas determinedby Elliot el ot. + 6.6057x 10-asin(-2. L8167x lO-47+ 1.32)radia.ns,
(1981). Ephemeridesare given for the elongationsand for the apparent distance
and position a,ngle,of SatellitesI through V; Ariel, Umbriel, Titania, Oberon,and z= -2.2795 x 1o-aexpi(4.727824x 10-57+2.41)
Miranda. The orbital elements of these satellites are from Laskar and Jacobson + 3.90469x 10-3exp(2.179132 x lO-sT+2.07)
(1987).The theory is developedin elliptical elements.Secula.rterms and periodic + 3.091?x 10-aexpi( 1.580524 x 1O-57 + o.'14)
terms with a.rrplitudesgreaterthan 0105at oppositiona.requoted below. x 10-6?+ 0.43)
+ 2.2lg2x !O-a expi(2.9363068

Miranda + 5.4923x 10-aexp!(-0.01157T + 5.71),

( = 1 . 1 1 3 3 6 x 1 0 3e x pi ( - 2 . 1 5 6 6 2 8+7O . 5 9 )
m / Mu = O.075x 10-5, i ( - 1 . 4 0 1 3 7 x3 l o 5 T + 1 . 7 5 ) .
3 . 5 0 1 4x 1 0 - a e x p (6.5-3)
n = 4.44352267- 136.65cos(O.621
7 + 5. 89)radians/day,
L = -0. 23805158+ 4. 44519055
T
+ 0.02547217 sin(-2.18167 x tO-4T+ 1.32)
- 0 . 0 0 3 0 8 8 3s 1i n -( 4 . 3 6 3 3 6x l O 4 T+ 2 . 6 4 1 nlMu=3.97 xlO s,
- 3. 181x 10-asin(-6.54502x 10-4f + 3. 97)radians,
n = 0.72166316ra.dia.ns/day,
z = 1 . 3 1 2 3 8 xl O 3 e x pi ( 1 . 5 2 7 x3 1 0 - 4 T+ 0 . 6 1 ) Z = 0.85635879 + 0.72171851 Tradians,
- 1.2331x l0-a expi(0.08606 ? + 0. l5) x lo sT + 0.74)
z = 9.3281x 10-aexpi(1.580524
- 1.9410x l0-a expi(0.709? + 6.04), + 1. 12089x lO 3 expi(2.9363068x 10 6r+0.43)
( = 0.03787171exp i(-1.54449x 1Q 4T+5.70). (6.5-l) +7.9343x 10-aexpi(-6.9008 x 10-32+ 1.82),
C = 6.8572x lo-a expi(-1.401373x 10-57+ 1.'15)
Ariel
x rc-67 + 4.21).
+ 3.7832x 10-aexp!(- 1.9713918 (6.54)
mlMu=1.49x10-5,
n = 2.49254257rcdians/day,
L = 3. O98O4641
+ 2. 49295252T
- 1.86050x10-3sin(-2. 18167x10-4i.+1.32) mtMu = 3.45x l}-s ,
+2.1999x 10-asin(-4.36336x 10-4f + 2.64)radians, n = 0.46658054radians/day,
z = l. 18763x lO-3 expi(4.727824x 10-57+ 2.41) Z = -0. 91559180+ 0.46669212TraAians,

\e'
 370 EXPLANATOBYSUPPLE[,4ENT 6 / OBBITALEPHEIiERIDES AND
RINGSOF SATELLITES

z = - 7 . 5 8 6 8x l 0 ae x pl ( 1 . 5 8 0 5 2x4 1 0 5 f + 0 . 7 4 ) the interval from then until 2050 exemplifies the cycle during
the 84-1'earperiod of
+ 1.39734x l0-3 expt(2.9363068 x 10-67+ 0.43) Uranus.
- 9 . 8 7 2 6 x l 0 a e x p i ( - 6 . 9 0 0 8l x0 - 3 f + 1 . 8 2 ) ,
positive
1966-2008: Earth north of orbital plane; B
( = 5 . 9 6 3 3x 1 0 a e x p i ( 1 . 4 0 1 3 ' 7x 3l 0 5 ? " +1 . 7 5 )
Superior conjunction at position angle P.
+4.5169x.1a
0 e x p i ( 1 . 9 7 1 3 9>
1<8 t O6 T + 4 . 2 t ) , (6.s-s) 1966-1986: Northern elongation at P + 90", when a U 90'
=

1986: Earth passed throrrgh planc perpendicular to

with Z = JD 2444239.5 and i = r/ 1. orbital plane, at U = 210" .
- =
Here m / My is the ratio of the mass of the satellite to the mass of Uranus; 71is 1986-2008: Northern elongation at P 90", when ir U 270" .
the daily mean motion; Z is the mean longitudc measured from the ascendilg rrorlq 2008: Earth passes through orbital plane, north to
of Uranus' equator on the Earth's equator to the ascendinglode ol the satellitc south; B = 0'
orbit on Uranus' equator along the orbit to the mean position of the satellite. z alcl 2008-2050: Earth south of orbital plane; B negative
( are converted to classical elcments with the following formulas Superior conjunctiou at positiorr angle P + 180'
-90", when I- U =2'10"
2008-2030: Northern elongatiorrat P
z = e exp(iP), 2030: Earth passesthrough plane perpendicrriar to
(6.s-6) orbital plane, at U = 90'.
(=sin]exp(td), 2030-2050: Northern elongation at P+90", wherr r- U=90'

where e is the eccentricity; P is the longitude of the periceutcr reckoned frorrr the At the passage of the Earth through the plane perpendicular to the orbiial
ascending node of Uranus' equator on the Earth's equator ol 1950.0. along l-lrarius' plane, the position angle of the northerrr elongatiou changes by 180o, because the
equator to the ascendingnode of the orbit on Urantrs'equator, and then along the more northerly and more sorrtherly extremities of the rnajor axis are interchanged.
orbit to the pericenter; 1 is thc inclination of the orbital plane to Uranus equator; The angle P is the position angle of the pole ofthe orbital plane that lies nolth of the
d is the longitude of the ascerrdingnode of the orbit on Uranus' equator measur{rd celestial equator; but eithcr the easterrr or western cxtremity of the niDor axis rnaY
from the ascending node of Uranus' equator oll the Earth's equator. be directed toward this po1e,according to circumstances. Nortlteln elongation rnaY
'Ihe
Refcrred to the Earth's cquator the motions of the satellites are direct. Horvevcr, be to either the east or west of north. elongations cannot be unambiguously
referred to the ecliptic they are retrograde, having an angle of inclination of roughlv designated as eastern and western.
98'. The orbital pla[e is inclined at such a large angle to the Earth's cquatol To calculate the values of "/, a. anrJ N. consider spherical triangle ABD in Fig-
that the semi-major axis of the apparent orbit usually lies nearly north and sotlth. ure 6.5,1.Sphcricaltrigonometry gives us:
Consequently, the greatest elongations are designated as northern and southcrrr
elongations instead of eastern or western. Only when the Earth is near the plane
through the celestialpole perpetrdicularto the orbital plane, as during 1986 1987, sin"/ sin(N - N") = sin f sin d,
(6.s-'7)
is the minor axis of the apparent orbit directed approximately north and sorLLh
sin"/cos(1y' N.) = cosI sin/" + sinl cosJccosd,
Even then, the north pole of the orbital plane may lie to the south of lhe gcuterLtIi,
position of Uranus. The north pole of the orbital plane is that frotn lvirich tlie
satellites' notion appears to be counterclocku'ise.This pole correspondsto the which can be solved for ./", the lorrgitude of tht:
"/ and N. The vir,luesof N" and
physical ephemerissouth pole of Uranus, which is dellncd as lying south of th(l ascending node of Uranus' equator and its inclination to thc Earth's equator' are
invariable plane of the solar svstem. found from the coordinates of Uranus' south pole: c4 = 181?43and 6o = -15910
During the course of one revolution of Uranus, the Earth passestwice through (Davies,1989):
the plane perpendicular to the orbital planc when the apparent orbits of the satel-
lites are almost circular; ald likcwise, it passest\4'icetluough the orbital plane when
N"=oo+90o,
the apparent orbits become straight lines. The Earth passed through thc orbital (6.s-8)
plane from south to north irr 1966, and the sequenceof geometric relatioDs durilig "/"=90'-60.
 372 EXPLANATORY
SUPPLEMENT ORBITALEPHEMERIDESAND RINGSOF SATELLITES 373
THESATELLITES
OF NEPTUNE

are given for the elongations and for the appaxent distance and po-
sition angle of Tliton, calculated from the orbital elements determined by Harris
(1984). Difierential right ascensionsand declinations are given for Nereid from or-
bital elements computed by Mignard (1981).

6.61 Triton
The orbit of Triton is retrograde. The orbit, as far as it has been determined from
obsenations, is circular. The orbital plane is defined with respect to a flxed reference
plane. This reference plane is defined such that the projections of spin angular
momentum of Neptune and the orbital angular momentum of Triton on the plane
are equal and opposite. The pole of this fixed plane with respect to the Earth's
equator and equinox of J2000.0 are

ap = 298."72 +2.58 sinN - 0.04sin2N,

6 p= 4 2 . "6 3 - 1 . 9 0 c oN
s + 0 . 0 1c o s 2 N ,
(6.61-1)
N = 359."28 + 54."3087,
Figurr 6.5.1 T=(JD - 2451545.0)
t36525.
Reference N = TD, P = AB + BC,
systemfor theUmniansatellites.
andL = AB + BC + meananomaly.
The orbital elements of titon referred to the fixed plane (see Figure 6.61.1) are

to = JD2433282.5,
a = 488i49,
The arc length DB is found from
e=0,
sinDBsin1 = sin(N- N") sin"/.. (6.5-e) n = 61?2588532I
day,
h = 200?913,
Then
f = 1583996= 213004retrograde,
u= L- 0+DB+(J- ItO, (6.s-lo)
- to)t 365.25,
0 = l5ll40l + 0157806(JD (6.6r-2)
where (l - 14) is the equation of center which can be found from Equation 6.11-1,
and where a is the greatest elongation at unit distance; e is the eccentricity; n is the
daily mean motion; la is the mean longitude reckoned from the ascending node
M=L-P. (6.5-11)
through the inva.riable plane at epoch; is the inclination of the orbital plane to
1
Orbitat, physical, and photometric data are presented on pages F2-F3 of fhe the invariable pta.ne;and I is the angle from the intersection of invariable plane with
Astronornical Almanac for 10 small satellites discovered by Voyager. The orbits of the Earth's equatorial plane of 1950.0 to the ascending node of the orbit through
these satellites all lie inside the orbit of Miranda. Cordelia and Ophelia are shepherd the invariable plane.
satellites for the e-ring. It has also been proposed that Cordelia is the outer shepherd Because the retrograde direction of the motion is represented by an orbital
for the d-ring, and that Ophelia is the outer shepherd of the .y-rins. greater than 90o, the ascending node of the orbit is the point at which
 374 EXPLANATORY
SUPPLEMENT EPHEMERIDESAND RINGSOF SATELLITES o/5

The computation of J, r, and N is as follows. Flom the spherical triangle in

6.61.1 we have

sin"/ sin(N - Ne)= sin i sin P,

(6.61-3)
sin,I cos(N - Ne) = cosi sin,/e+ sin i cos.Iecosd,

be solved for "/ and N. N. and "I", the longitude of the node and the
of the invariable pladre to the Earth's equator, are found from

N'=oP+90"'
(6.61-4)
,r.=90'-6p.

arc length CB can be computedfrom

Earthequator sinCBsini = sin(N- N") sinJ". (6.61-s)

follows from
u=h+DB+n(JD-h), (6.61-{)

Figure 6,61.1 JD is the Julian date of interest.

'fC.
Referencesvstemfor Triton. ,V =
Nereid
orbit of Nereid is unique among the satellites in the solar system, becauseof its
the satellite crosses the equator from south to north. The pole of the orbit from
eccentricity and the great difierence in its inclination from that of Tliton.
which the motion appea,rscounterclockwise is the south pole, at position angle P.
orbital elements of Nereid with respect to the orbital plane of Neptune are (see
Elongation is at position angle P - 90', vrhen z - IJ = 27O' .
6.62.1):
Twice dudng the course of one revolution of Neptune, about 165 years, the
Ea.rth passesthrough the orbital plane of Tiiton, when the apparent orbit becomes
a = 0.036868AU,
a straight line. The Ea.rth crossedthe orbital plane ftom north to south near the end
e 0.74515 0.006cos2P + 0.N56cos(2w-2$),
= -
of 1952; for an interval during 1950 1954, Tliton transited the disk of Neptune and
was occulted by the disk during each revolution. Before 1953, B was negative, and cos./= cos10:041- 090094cos 2{,
iuferior conjunction was on the southern arc of the appaxent orbit at position angle 0 = 329?3- 2?47 + 19?7sin2i! - 3?3sin4r!
P. Since the passage through the orbita.l plane, B has been positive, and inferior + 037sin6V + 03357sin2S+ l?276 sin(2a- 2Q)'
conjunction is on the northern arc at P * 180'. The Earth will pass through the
P =Q - 19?25sin?iZ+3?23sin4V
orbital plane again in 2035.
- 03725sin6V - 03351sin2/ - 037sin(2u - 2Q)'
When the numerical value ofB reaches a maximum as the Earth passes through
the plane perpendicula.r to the orbital pla.ne, as in 1987, the minor axis of the + nt - o?38sin}iZ + 1?osin(2u- 26),
M = 358?91,
appareut orbit lies exactly north and south, the major axis lies east and west, and
n=0?9995521day, (6.62-r)
the position angle of the more northerh elongation cha.nges by 180'. Before 1987'
the western elongation was the more northerly; since tben, the ea,stern elongation 68 T and2$ = lO7?4+0901196r. I is reckonedin Julian centuries
itrr= 282?9+2?
has been the more northerlv. ! in days from JD 2433680.5' The inclination, 7, is referred to Neptune's orbital

'q*
376 EXPLANATORY
SUPPLEMENT ORBITAL ANDRINGSoF SATELLITES
EPHEMERIDES 377
The coordinates 12, y2, z2 are then used to compute { and 4 with Equations
#.

Pericenter

7 THESATELLITE
OF PLUTO

ofnorthern elongation, apparent dista"nces,and position angles are presented

The Astronomical Almanac for Charon. The orbital elements from which these
h.recalculated are taken from Tholen (1985). The elements, referred to the Earth's
equator and equinox of 1950.0, are

a = 0100O12788,
n=56?3625/day,
Earth equator
A J = 94?3,
N = 223?'t,
Orbit of Neptune
Orbirof Nereid lo ='78?6,
Figurc 6.62.1 to = JDZMSOOO.S, (6.7-l)
Referencesystemfor Nereid.N = TD andP = AB + BC.
where a is the semi-major axis; z, the mea.ndaily motion; ,/, the inclination of the
orbital plane to the Ea.rth equator of 1950.0;N, the longitude of the ascending node
reckoued from the vernal equinoxl .16,the mean longitude at epoch reckoned from
the ascending node along the orbit; and 16,the date of epoch.
plane. The longitude of the node, 0, is measured in Neptune's orbital plane ftorrr
The orbit of Charon is retrograde. The pole of the orbit from which the motion
its intersection with the 1950.0 celestial equator. The orbit of Neptune is taken to appears counterclockwise is the south pole, at position angle P. Elongation is at
have inclination, J. = 22?313, and longitude of ascending node, N. = 3i522 q'ith position angle P - 90o, when u - U : 27O".
respect to the 1950.0 celestial equator. M is the mean anomaly; r,l, the argument of Twice during the course of one revolution of Pluto, about 248 1'ears,the Earth
pericenter; e, the eccentricity; and z, the mean daily motion. passes through the orbital plane of Charon, when the apparent orbit becomes a
The differential right ascensionand declination x andy coordinates in the orbital straight line. The Earth crossed the orbital plane three times during 1987 1988,
plane are computed from the Equations 6.15-3. These coordinates are transformed the end result being a crossing from north to south. In the interval 1985-1990,
to the orbital plane using Charon transits the disk of Pluto and is occulted bv the disk durine each revolution.
1987,8 was negative, and inferior conjunction was on the southern arc of
f''I = & ( - d ) R , ( - 1 ) R , , - " (6.62-2)
appaxent orbit at position angle P. Since the passagethrough the orbital plane,
l/'l
y z r) [;]
has been positive, and inferior conjunction on the northern arc at P+ 180o.
When the numerical value ofB reachesa maximum as the Earth passesthrough
plane perpendicular to the orbital plane, as in 2050, the minor axis of the
These coordinates are then transformed to the Earth equator and equinox oi
orbit lies exactly north and south, the major axis lies east and west, and
1950.0usine
position angle of the more northerly elongation changes by 180". Before 2050,
ftl =n:t-ru"tn',-.,
f',.l (6.62-3)
western elongation is the more northerly; after 2050, the eastern elongation will
the more northerly.
lr:l L:;l
378 E X P L A N A T O RSYU P P L E [ 4 E N T
ANDRINGS
EPHEMERIDES OFSATELLITES 379
6 I OFBITAL
Since the orbit is referred to the Earth equator ald equinox of 1950.0,J ancl ,ry w.S. (1892) "The orbit of Hyperion" Astron J. II' 145 157'
Eichelberger,
are already known. r is found sirnply fronl W.S. and Newton, A (1926) "The Orbit of Neptune's Satellite and the Pole
Eichelberger,
of Neptune's Equator" ,4 P.,4.8. s,275 337.
u = l o+ n ( J D r0), (6.1-2; Elliot, J.L., French, R.G.,
Frogel, J A., Elias, J H, Nlink, D J ' and Liller'' W. (1981).
,'Orbits of Nine Uranian Rings" Astron. J.86, 444 455
E.A, Tyler, G L, and Porco, C.C.
whereJD is the Julian date of interest. Beposito, L.W., Cuzzi, J.N., Holberg, J H., N{arouf,
(1984). "saturn's Rilgs: Structure, Dynamics, and Particle Properties" in Solrrrn T.
Gerhels and M.S. Matthews, eds. (University of Arizona Press, Tucsol, AZ)
H.A. (1972). "The Mass and Figure of Saturn by Photographic Astrometry of Its
Garcia,
6.8 REFERENCES Satellites" Astron. J 77, 684*691
Grosch, H.R.J. (i9a8) "The Orbit of the Eighth Satellite of Jupiter" ,4slron. -/. 53, 180
Adams, J.C. (1877). "Continuation of Tables I and III of Damoiseau'sTables of Jupiter's 187.
Satellites" Nautical Almanac for 168l Appendix, 15 23.
Hall, A. (1878). "Observations and Orbits of the Satellites of Mars" lfnsi.ington Obser-
Aksnes, K. (1978). "The N'Iotionof Jupiter XlIl (Leda), 1974 2000" Astron. J.83. 1'21.9 oations Jor 1875 Appendix I, l-46.
't
256.
Hall, A. (1885). "The Orbit of Iapetus, the Outer Satellite of Saturn" Washington Obset-
Andol'er, H. (1915). "Sur ls Qalsrl ,JesEphcmerid"s d"s Qrralres Ancicns Satclliles de Dations lor 1882 APpendix I, I-82.
Jupiter" -Bull. Astr.32- 177 221.
Hall, A. (1886). "The Six Inner Satellites of Saturn" I'l/aslringlott Obseruations Jor 1883
Arlot, J.-E. (1982). "Ne*' Constants for Sampson-LieskeTheory of the Galilean Satellites
Appendix I, I-74.
of Jupiter" Astron. Asttophgs. 107,305 310.
Hall, A. (1898). "The Orbit ofthe Satellite of Neptune" Astron J. 19, 65 66.
Bessel,F.W. (1875a). "Untcrsuchungeniiber den Plancten Saturn, seinenRing und st:inen
Harper, D., Taylor. D.B.. Sindair, A.T., and Kaixian, S. (1988). "The Theorv of the
vierten Trabanten" Abhandlungen I, 11O 127.
Motion of Iapetus" Astron. AstropLgs. 191, 381 384.
Bessel,F.lV. (1875b). "Bestimmung der Lage unrl Gr6ssedes Saturns nges und der Fugru'
Harington, R.J. and Christy, J.W. (1981) "The Satellite of Pluto III" Astron.. J 86,
und Gr6sse des Saturns" AbhandLungen l, 15O 159.
442 143.
Bessel,F.W. (1875c). "Ueber dic Neigung der Ebele des Saturnsringes" Abhardlungen l.
Harris, A.W. (1984). "Physical ProPerties of Neptune and Triton Inferred frotr the Orbit
319 321.
of Triton" NASA CP 2330.357 373.
van Biesbroeck,G. (1957). "The N{assof Neptune from a New Orbit of Its SecondSatellilc
Nereid" ,4sfron. J. 62,272 271. Harris, D.L. Unpublished work on the satellites of Uranus, quoted by Kuiper (1956' p'
Bobone, J. (1937a). ''Tablas del VI (sexto) Sat6lite de Jupiter" Asl. Nacl'262,321 3]6 1632).
Bobone, J. (1937b). "Tablas del VII (s6ptirro) Sat6lite de Jupiter" ,4st. Nacfi..263. '101 Harshman. W.S. (1894). "Th€ orbit of Deimos" Astron J.14. 145 1'18.Harshman's el-
412. ements of Phobos "communicated by Prof Harkness" are attached (in manusr:ript)
Born, G.H. aDd Duxburyj T.C. (1975). "fhe Nlotions of Phobos and Dcimos frotrr \lrrilrcr to a copy of this paper. It is believed that they were never published See also Pzbl'
9 TV Data" CeL.Mech. 12. 77 88. U.S.N.O.6Appendix I, 821 (1911).
Brourver,D. and Clemence.G.N4.(1961). trlethodsof CelestialMecAonics(AcaciemicPrcss, Herget, P. (1938). "Jupiter XI" Hannrtl Announcement Co,rd. 163
London). chapter XVL Herget, P. (1968). "Outer Satellites of.Iupitcr" Ast|on. J. 73,737 742.
Christy, J.W. and Harringiurn, R.J. (1978). "fhe Satcllite of Pluto" '4slron. ./. E3. 1005 Herrick, S. (1952). "Jupiter IX and Jupiter XII' P.A.S.P. 64,237 211
1008. Kendall, E.O. (1877). An Extension for the Year 18E0of Damoiseau'sTables o[ the Sate]-
Cohn, F. (1E97). "Bestiurnuug der Bahnelementedes V. Jupitersmontles" Ast. Noch 112' lites of Jupiter (probably never published).
289 338. Kozai, Y. (1957). "On the Astronomical Constants of Saturnian Satellites System" Anz'
Comrie, L.J. (fo:+). "Phenonrenaof Saturn s Satellites" Mem. R.A.A.30, 97 106. Tokgo Obs. Ser. 2 5,73 106.
Crawford, R.T. (1938). "The Tenth ard Eleventh satellites of Jtpitcr" P.'4.S.P 50. 3'11 Kuiper, G.P. (1949). "The Fifth Satellite of Uranus" P.A.S.P.6l, 129.
317. Kuiper, G.P. (1956). "On the Origin ofthe Satellitesand the Trojans" Vistas xn Astronomlt
de Damoiseau. M.C.T. (1836). "Tabl"s ieJiptiques dcs Sat"llii". dc Jupiter" Paris 2, 1631 1666.
Davies, M.E.. Abalakin, V.K., Bursa, N{., Hunt, G.E., Lieskc, J.H., N{orando, B., RapP' de Lalande, J.J. (1792). Astronotnie3rd edition, Paris.
R.H., Scidelrnanrr.P.K., Sinclair, A.T., ard Tjullin, Y.S. (1989). "RePort of tln' Laskar, J. and Jacobson,R.A. (1987). "An Analytical Ephemeris ofthe Uranian Satellites"
IAU/IAC/Cospar \\'orking Croup o Cartographic Coordiuates and R.otation.r"lEle- Astron. Astrophys. t88, 212 224.
ments of thc Plancts aDd Satellites: 1988" Cel. trIech.. 46, 1a7 2O1. Levin, A.E. (193i, 1934). "Mutual Eclipscsand Occultations ofJupiter's Satellitcs" J'B A'A'
Delambre, J.B.J. (1El?). "Tables Eclipi.itluesdes Satellites de Jupiter" Paris. 4 2 , 6 1 4 ( 1 9 3 1 ) .M e m . B . A . A . 3 0 , 1 , 1 91 8 3 ( 1 9 3 1 ) .
Dunham. D.W. (1971). "Thc Nlotions of the Satellites of Urarrus" Dissertation. (Yale Lieske,J.H. (1978). "Theory of Motion of Jupiter's Galilean Satellites" Ast|on AstrophYs
llnivcrsit5', Nerv Haven, CT). 56, 333 352.
 380 EXPLANATORYSUPPLEI\/ENT 6 / ORBITAL
SN D R I N G SO F S A T E L L I T E S
E P H E M E R I D EA 381
der sat rrnrstfal>a]rlten" Publications da I'obserttu.toirc
Lieske, J.H. (1987). "Galilean Satellitr: Evolution: Observational Evidence for Secula" strlrve, H. (189S). "B*ol,achtunHeD
7 - " n 1 7 6l1t l ' , , ' l n s S ' r t ' I I I I . | : \ : \ 7
Changes in \lean Nlotions" Astron. Astrophgs.176, 146 158.
g cler l,ibrzrtion \limas-Tethvs' ,'1st Nd'ir' 162 325
Marth, A. Nlarth published a lolg series of ephemerides for physical obsen.ations of th. Struve, H. (1903). "Ncue Bestimmu
planets and for observations of the satellites. The first and la^stof oa(:h series are alon. 344.
quoted here. clnrve. H. (1911). a'ber die Lagc der \larsatrhsc und die Konstantclt irn \larssvstenl'
" llissenschalten Jiir 191t
Marth, A. (1870 1894). "Ephemcris olthe Satellitcs of Uranus" Mon. Not. Ro!. Astron. Sitrurqrb,,r"hte ,to Ktinighch Preus.sisc/renAskademieder
Soc. 30, 150 15'1,171. 1056 1083
,,Bahnen der uraDusl ral)aDtclr.Abteilung I: oberolr ulrd Titarria ;16
tr{arth, A. (1873 1895). "Ephemeris of the Satellitcs of Saturn' Mon. Not. Roy. Astn.trt. struve, H. (1913).
Soc. 33, 513 555, 164. hand,lungett der K6nigich Preussisdrcn Akademie der WisscnschaJtut'
of Jupiter" llarus 58, 178 181
Marth, A. (1878 1891). "Ephemeris of t he Satellites of Neptunc" MotL. Not. Rog. Astr an. svnnott, S.P. (13x4). "Orbiis of the Small Inner Satcllites
(19E1) A Comparison of the fhcory of the \4otitxr of t{v1x'tion with ob
Soc. 38, 475 451, 563. ii.vlot. D.B.
servations \'lzrde DuriDg 1967 1982' Astron Astrophys l4l' 151 lsfl
Marth, A. (1883 1896). "Data for Finding the Positiorrsofthe Satellites of Nlars" ,{Ion
D.B. arr,l Shcn. K.X. (19S8). "AIr:rlvsisof Astrometric Observatiorrsfronr 1967 to
Not. Roy. Astron. Soc. 44,29 56, 135. Tavlor,
1983 of thc \lai',r Satellite. of Saturrr" Astron Astrophljs 200 269 278
lUarth, A. (1891 1896). "Ephemeris for Corrrputing the Positions of the Satellircs of 'On the Shado* of Saturn on Its Rings Astro1t l 55 229 230
Jupiter" ,{Ion. Not. Roy. Astron. Soc. 51. 505 556. .13,1. Taylor, S.W. (1951)
'l'he 'I s0 2353 2359
Tholen, D.J. (19u5). Orbit of Plrtto s Satcllite
l,lignard, F. (1981). "The \Iean Elements of Nereid" Astron. .1. 86, fi2a 1729. "lstron
Ne*'corrrb,S. (1875). "The Uranian and Neptuuian Systems" l4laslringt on Obserxati,:tns Todd,D.P.(187tj)."ACortinuationofdcDamoiseau'sTablesoftlx'SatcllitesofJupiter
to the Year 1900" (\Vashington Bureau of \avigation, \\hslingtorr)
for 1873 Apperdlx 1.7-74.
Nicholson, S.B. (1944). "Orbit ofthe Ninth Satellite of Jupitcr" Asttrtph|s. J. 100,57 (j2 Wargentin,P'W.(17'16)...Tabrrlaeprocalclrl:rn<|iseclipsibussatellitttlll.]lrr,is..in-,lcfaSoc'
-reg.
Pottier, L. (1896). "Addition aux'Iablcs Ecliptiques des Satellit"s do .lupiter de Darloi sci. Llpsolienstsa(l annutn 1!7 (Lartrcntii Salvii' Stockholnt)
P.A S'P 51 241
seau" Bull. Astr. 13,67 79 and 107 112. Wilson. R.H., Jr. (1939). "Revised Orbit iuxl EpherDcrisfor Jupiter X"
Robertson, J. (1921). "Orbit ofthe Fifth Satellite of Jupiter,, Ast.,:.nt. 242.
.J. j5. 190 193. .!ao \Moerkom.A.J..l. (1950). "The\lotionof .lrrpiter'sFifth Satellite lug2 1919- l'P i I]
Rohde. J.R. (1990). UDpublished.
13,1 77.
Rose, L.E. (1974). "Orbit of \ereid anrl the llass of \eptuDe' ,4stron. .,I.29.{89 rg0. 'The \lotion of Hlperi<xr" Annalen uan de Steneu'ttlt tc Leiden 16'
Woltjer, J. (1928).
Ross, F.E. (1905). Investigations on the Orbit of Phoebe" Hana.rd. Annals b3. 101 lll.
part 3, I 139.
Ross, F.E. (1907a). 'Serrri-definitir,.eElements of Jupiter's Sixth Satellite,, Lick Obs. Bull. 'Woolhouse.
W.S.B. (1833). "Ne* Tables for Computing the Occultaliolrs of Jupiter's
4. 110 112.
Satellites bY Jupitcr. the Transits of fhc Satellites and Their Shadous Over the Disc
Ross, F.E. (1907b). "Ne$' Elements of.lupiter's Seventh Satellitc,' ,4sa. Neth. 174.3'Dl)
ol the Planet. and the Positions of thc Sat'cllites\\'ith ResPs t to lupitcr at Any
362.
T\me" Nauti.atl Almanac for 1835 Appcldix. 1 39'
Sampson, ll.A. (1910). ?o6les of the lrour Great So.tellitesof Jteit(\' (William \\ieslel &-
Zadunaisky, P.E. (1951). "A Detcrminntiorr of Ne1\ Elements of thr' Orbit of Phocbe'
Son, London). ' Astrotr' J 59' | 6
Ninth Satcllitc of Satur
Sanrpson,R.A. (1921). "Theory ofthe Forrr Great SatellitesofJupiter" lVem. Rog. Astratt.
Soc. 63. '\ 27O.
Sharnpine, L.F. and Gordon, \LK. (1975). Cornptrter SotutionsoJ Ordinarg DilJertntiul
Equations (W.H. Freeman. San Francisco.CA).
Sinclair, A.T. (1974). 'A Theorl of the \'lorion of Iapetus" ltlort. Not. Roy. ,lstr. Sor.
169. 591 605.
Sinclair, A.T. (1977). "The Orbits of Tcthys, Dione, Rhea, Titan anrl lapetrrs', ,il,Ion.Itt.
Rog. Astr. Soc. ti0. 147 4bg.
Sinclair, A.T. (1978). "The Orbits ofthe Satellitesof Mam,, yisras .in Astrcn.22.1:lS 110.
Sinclair, A.T. (1989). "The Orbits of the Satellitcs of Mars Detcrmincrl fronr Earth-basecl
and Spacecraft Observations" Astron. Asttophus. 220. 321 32a.
Struve, G. "Ncue Untersuchungen im Sarurnsvste nt:' Veriiffen ttt"h ttttgr n tler Linruersitiits-
sternuarte zu Berhn-Babelsbe.rg 6. Parts 1.4,5. (1924) Part I Die Bahn ron Rhco
1 16. (1930) Part 4 Dte Systeme ltlirnas-Tethys und. Encelarlus-Dtone I 6I. (193:Jj
Part 5 Die Beobachtungen der iiussen:n Tiabanten und. die Bahncn ton Titan und
Japetus 7 q1.
Struve, H. (1888). "Beobachtungen der Saturnstrabanten" Suppl'mcnt I aux C)bsertatior,s
de Poulkoua | 132.
 CHAPTER

PhysicalEphemerides
of the Sun,Moon,
Planets,andSatellites
by JamesL. Hilton

7.1 INTRODUCTION

The physical ephemerides of the Sun, Moon, planets, and satellites give information
on the apparent physical aspects of the disks of those objects. The physical ephem-
erides are used for making and reducing observations of the surface markings of an
object and for determining the exact center of an object's disk from observations
of the limbs of the object. AII the tabulated values are geocentric and have been
cotrected for the effects of aberration. Hence the information tabulated gives the
appeaxanceof the object as seen from the geocenter at the tabulated time.
In general, the ephemerides can be broken into two groups: (1) information
depending on the disk of the object alone, the rotational elements, and cartographic
coordinatesJ and (2) the phase and magnitude of the object, which depend on the
position of the Earth and Sun relative to the object and the object's physical
appeaxance,
Unless otherwise stated, all coordinates given in this chapter refer to the mea.n
equator and equinox of J2000.0. I is the time in Julian centuries of 36525 days from
J2000.0, and t is the time in days (86400 SI seconds) from J2000.0. The J2000.0
coordinate system is defined by the FKs star catalog and has the standard epoch
of 2000 January 1.5 (JD 2452545.0), TDB.

7.11 RotationalElementsand Cartographic

Coordinates
The cartographic coordinates of an object consist of the orientation of the north
pole and the position of the prime meridian of the object with respect to an inertial
tefetence frame. These elements are given by the three quantities: o0, 60, and lry as
b function of time. These parameters are shown in Figure 7.11.1.
 384 E X P L A N A T O BSYU P P L E [ , 1 E N T 7 / P H Y S I C A LE P H E M E R I D EO
S F T H E S U N ,l \ r O O N P
, L A N E T SA, N D S A T E L L I T E S o65

prime meridian is chosen to follow the cartographic prime mericlian

the ephemeris
as possible, thcre may be errors in the rotational r:lernents of the object
as closely
the cartographic position to drift away from the epherneris position
1[at will cause
positively counterclock\\'ise
by a snrall amount AW. The angle AW is measured
the ephemeris position of W as viet'ed from abovc the north pole. For a prime
1ep
Planet'sequalor meridian that has a cartographic definition, the ephemeris definition of lV niay bc
of date
changed as more accllrate information on the motion of the caltographic primc
meridian is obtaincd
Earth's equator The definition of the north pole requires the rotation about the polc t6 be
ofJ2000.0
classifiedas either direct or retrograde. Direct rotation is counterclockrviseas vier,"rd
from above thc planet's north pole, while retrograde rotatiou is clocklvise. For a
planet with direct rotatioI1 the angle W increasesu'ith tirne. For most of the satellitcs
it is assumed that the rotation period is equal to the mearr orbital pcriod.
Information on the point defining the prime meridian and pole for the Sun can
be found in Carrington (1863) and Section 7.2 of this supplement. The delirrition
for the prime meridian and pole of each planet can be found in Davics cl al. (1991)
and Sections 7.41 7.48 of this supplement. Information defining the prirnc meridian
and pole of the primary planetary satellites can be founcl in Davies el al. (1991) and
Figure 7.11.1 Sections7.3 and 7.51 through 7.56 of this supplement.In general,the expressions
The positionofthe northpoleandprimemeridian defining a0, 69, and W should be accurate to one tenth of a degree. Tlvo decimal
of a planetin Earthequatorialcoordinates
places are given, ho*-ever, to assure consistency rvhen changing coordiriatc slstcrns.
Zeros are added to the rate values (W) for conputational consistencv arrd are not
an indication of significant accuracy. Three signilicant digits bcyoud the decimal
The quantities o6 and dj are the right ascension and declination of date o1 ihe
point are given in the expressions for the Nl[oon and \'{ars, rcflcctirtg the €lreatcr
object's north pole. The IAU definition of the north pole of a planet, frorn Thc IAI.I
confidence in thcir accuracy. The recornmcrrdcd coordirrate systeni for the Nbon
\irorking Group on Cartographic Coordinates and Rotational Elements of Plant:ts
is the mean Earth-polar axis system in contrasi to thc principal axis systcru. ll'he
and Satellites(Davies el ol.. 1991), is the rotation axis that lies on the norlh sid('
Earth-polar axis systcm uses the polc of rolatiori of the Nlc,on and the nrcan axis of
of the inr'ariable plane of the solar systeu]. a0 arld 60 maJr vary slowl)' witir linre
the Moon, r'hich points ton-ard the (enter of the Earth. The principal axis sJstem
becauseof the precessionol lhc objcct about its polar axis. In the absenceof othcr
usesthe axes of the rotation and thc delincd 0o meridian of the obiect.
inforrnation, the axis of rotation is assurncd to be rrorrnal to the rr ean orbital plarxr'
Both planetocentricarrd planctographicsystemsof coordinatesalc useclin the
This assumption is used for N{crcury and most of the satellites.
study of the planets arrri sartcllites.Both systems are based on the sa[le fundamcn
The angle W is measured along the object's equator, in a counterclock$is{l
tal axis of rotation but diflcr, as explained belorv, in their definitions of latitude,
dircction when viewcd from the north pole, from the asccrrdirrgnode of thc ob icct s
longitude, ancl range. Planctoccrrtric coordinates are used lbr gen(r'al purposcs and
equator on the Earth's mean equator of J2000.0 to the point where tlle pr-llnt'
are based on a riglrt-hande<l system of axes with its origiu at the cerrter of mass of
meridian of the r.rbjcct crossesits equator. The right ascensionof 1,herode (0) is trt
the objcct. Planetographic coordinates are used for cartographic purpos(rs and de-
6h+oe,and the inclirrationof the obiect's equator to the standard cquator is 90" - 60. pend on the adoption of a reference sulface, usually a sl)heroi(I. that approxirnates
Generally, the prime rneridian is assumed to rotate unifolmly u.ilh the objcct. henct'
an equipotential surface of the object. The latitudc and ranges for both of these
lV varies linearly with timc. For objects without obselvablc, frxed surfacc fealurcs
coordinatesystemsiue shorvnin Figure 7.11.2.
the adopted expression for W defines thc prirrrc meridian and is not subje<:t 1rr
The reference surfaces for rtost of the planets are sphcroids for which the equa-
conection. However, if a cartographic position for the prime melidian is assignc.l torial radius, 4, is larger than the polar radius, b (sec Figure 7.11.2).For son.reof
to an object, that is, il the position of the plimc rneri(lian is dcflrrr:dby a srtitablt: rhe planets and rnost of thc satellites the rcference surface is a sphere. Fbr somc
obserr.able surface feature, the expression for W is chosen so that the cpherrL.rj' satellites, such as Phobos. Deimos, ancl Hyperion, the reference surface should be
position follows the rnotiorr ol the cartoglaphic position. Although thc defrnitior of
 386 EXPLANATORYSUPPLEi,4ENT
387
S F T H E S U N ,M O O N ,P L A N E T SA, N D S A T E L L I T E S
z / P H Y S I C A LE P H E M E R I D EO
PIanet's
axis of rotation y-axis is then orthogonal to thc other tlro axes attd tnakes a right-halrrir'<lcoordi-
these two s)'stemsale latilttde and
nate systcnl. The angular coordill.rtesused in
systcrn is the dist;rnce from
Iongitude. Tltc range of the plartetoccntric coordinatc
planetogral>hic system is the hoigltt lbove
the center of mass, and thc rangc of the
or below thc reference surfacc.

,2,
Latitrrde (sce Figure 7.11.2) is mcasured uorth and south of thc ctltrator rvith
the northern latitudes being clcsignatedas positive and the southern latitttdes as
Tangentto pltuttrl is the
ellipseat P' negative. The planetocentrit:Iatitule ol a point orr thc strrfaceof a
angle d bet$'ccn the equatorial l)larrc and the line segmeut connectirrg the Point to
Plrnet s equJtor Equatorialradius the center ol mass. The planctographic latitucle of a point is thc angle o/ between
the line segment normal to the r:qrripotential surfacc from the point in qttcstiou to
the equatorial plane. Since the r:rltipotential surfacc is usrrall.va spheroid. the shapc
of the spheroid is related to that of a spherebv its flattening. /'

f=(a-b)ts, ( 7 . 1l - l )
Figure 7.11.2
Planetocent c and planetographiccoordinatesystems.Thc quantitiesshown are: ,2,rne
planetocentriclatitude; r, the distancefrom the center of mass;q6lthe planetographiclatitudc: and
ft, the height of a point above or below the planelographicreferencesurface.Also included are a, where a is thc equatorial radius ol the object and D is tlx: Polar radius of 1hc object
the equatorialradius of the objecr; and b, the polar radius. For a sphcricalobject. the plattctocr:trtricand planctographiclatitu'les ar(' the san-re
The ralrres for the radii of tlle planets and ax{'s given in Sections 711 7 '18
and in ?Ae Astrctnomical Abnano.c ane derived bv diffcrent mcthods. alt(i do not
always refer to common definitiorts. Some radii Llsc stirr or spaceclaft oc(:llltation
triaxial. Triaxial ellipsoidsrvouldmake man) colnputationsmore corrrplicated,esp(l
measuremeut, some use linrb fittillg, and some uso ()rl1 rol-netlvork conUnltations'
cially lhose related to map projcctions. Therefort:. spherical or $plroroidal referencc
For examplc, the sphcroid fbr thc lJarth rcfers to rlnrarl sea lcvel, a defirtitiou that
surfaces are usually uscd irr nrapping prograllts. The polar axis for cach r.efer-etrct:
can be uscd onlv for lhis plarrrtrl.The radii and axcs of the large gasni'rl\I)lanr.ts'
surface is assumedto bc tlrc rnean axis of rotatioD as defined by tlx, aclopl<:cl ro- 'l'hc ladii
Jupiter. Satttrn. Ulanus. and \t'pttttre refcr to a ottebar pressuresrllfacc
tational elelnents since tltt: present accuracY ol rneasuremerttdoes llot r-cr.eala
in the tables are not necessiuilvthe appropriatc talrtr:s to be use'l itr tlrttionical
deviatiorrbetrveenthe axis of rotation and the axis of figrrrefor nost of the plaltets.
studies;thc ladii actualll.'usctl to tlerir-ea raluc of J1, fol cxample. slrottltl al*a1-s
The radii and flatiening of 1.hcplanets are forrnd it The Astrononticol Alrnonac ^trd
be used in conjunction rvith thal vzrlue.
in Scction 7.4 of this supplr:rnent.The rnearrratlii of the naturtll silt(tlitcs of the
Longilrrtle is mcasured trlortg tlrr: equator frorrl tltc prirne merirlialt l)ktrtetor:en
plalcts iue also found ol pagc F3 of The Aslron.ortirn.l,Alrnanac.
tric longitu,dt,s, ,\, are measrtrtrl positively in th<rrrrttnterclockwise'lir''' li"rr when
N'Iany small bodies in the solar svstenr (sa1.r:lJitcs. asteroids, artrl corlet rn1- viewed fronr thc north pole. PIon(.togftryhiclonttitttrlts,.\',are meas'tr'"1P'rsitivell
clei) have verf irregular slr:r1>cs. Sornetirrrossphcrical referencesurfaccs a1.eLlse(l in the dircctiou opposite to th('r'olation of the ol;jett. For an objcct rvith a direct
to prescr'\'eprojection propcrties. Orthogr.aphicprojections are oftcn ^dol)tcd lbr
rotation, thc planetographic lorrgitrtdeincreasesin tlte clockl'ise (lirr(ti"n $'hPn
cartogrzrphicpurposcsbeclusc thel.prescrr,.ethc irregular appear.z11x,e of the bo.h viewed froln abor.ethe north polc. Planetocentritrlongituclesare trrcasrllcdfroln
withoul artificial distortiorr.
the ephernerisposition of thc prirrrcrneridian as rlclirtcdll1' the rotational clcnlents'
Itr both systems.th<:fiur<leunental refercrx.ez-lxis is the mc.rn axis of rotatiol but the planetographic lorrgilrrlt:s are measulcd frotlr the cartographi< position of
I'ith thc positive directiorr bcirrg toward the lorth pole. The equir,t.or is the plale, the prime rnclidian as <lt:finecllrr, the aclopted lorrgilrt<lc of sorne cicarlv obscrvable
passing tltrough thc center of mass of the plaDot..pcrpendicular to tltc z zrxis.The
surface featrrlc. For the Ealth. thc Sun, and tltt: Nloon. Iongitudes arc llleilsured
r-axis is defined by thc intcrsrr.tion of the platrt:of the equator ancl t.hc plane cotl- from 0' to l80o east and west. artd east longitudc (countcrclockrvise$'lrnlrI'i'rvc'l
tainittg the prime rneritliarr. lhe choice of tlrt: prime meridian is arbitrarr'. The ftom above the north pole) is cottrllonly corrsidcrtxl to be positive
 OF THESUN' MOON PLANETS'ANDSATELLITES 389
388 ExPLANAToRY
sUPPLEMENT PHYSICALEPHEMERTDES

In the planetographic system, a point P not on the reference surface is specifled

by the planetographic longitude and latitude of the point P/ on the reference sur-
face at which the normal line segment connecting P to the equatoria.l plane passes
though the reference surface and the height of P above or below the reference
surface.

7.12 PhasesandMagnitudes
The tabulated magnitude of an object is the integrated total visual magnitude.
The magnitudes for the planets a"rebased on Harris (1961). The magnitude of a
planet is given in the section on that planet in this chapter. The magnitudes for
the brightest asteroids are taken ftom Lumme and Bowen (1981). The basis for the
photometric radii of the asteroids is based on Lunme, Karttunen, Piironen, and
Bowell (1986) and is described in Section 7.6. Rotation rates for the asteroids are
not included in The Astronomico,I Almanac,bul inlormation on how rotation curves
are determined can be found in Lumme, Karttunen, Bowell, and Poutanen (1986).
The apparent magnitude of an object is a function of its distance from the Sun.
r, its distance from the Earth, d, and the solar phase angle, i (see Figure 7.12.1).
The solar phase angle is the angular distance at the object between the Sun and
the Ea,rth. For Mercury, Venus, and the Moon, i varies from 0'to 180". For \'lars
and its satellites i can be as large as 47o. For the rest of the planets, satellites. and Ficu€ 7.12.1
iiJ^i" -a theangleof illuminationThesequantitiesaffecttheillumination
most of the asteroids i reaches only a few degrees. Earth-Sun-planetplane.
"""4^
oiu ofun"tu, ."rn f.om the The
Earth. figureis drawnin the
Neglecting variations from rotation or other intrinsic causes,the observed mag-
nitude for an object is given by

Y= Y(1,0) + 5 logro(O+ Am(i), ('1.12-I)

where V(1,0) is the magnitude of the planet seen at a distance of 1 AU with a phase markings
the planet, and (2) changes in the planet's atmosphere or volatile surface
angle of 0o, d is the Earth-planet distance in eus, and Az(i) is the correction for give rise to brightness
such as Mars' ice caps. The surface markings on an object
the variation in brilliancy of the planet with phase angle. 'rariations that depend on the rotation of the object The second source of intrinsic
The quantity An(i) is measured empiricall3.. It is the result of two efiects. Tbe varra-
variability gives rise to variations of an irregular nature The atmospheric
long periods
first source of variation arises from the fraction of the illuminated disk visible from tions may be short-lived, such as sand storm" on Mars, or persist for
the Ea.rth as it varies with i. The second eflect is due to the properties of diffuse the atmospheric variations
of time, such as the Great Red Spot on Jupiter' Because
reflection from the object's surface and/or atmosphere. The parameter Ant(i) is be predicted with great
change erratically, their effects oo t pl.,t"t'" albedo cannot
expressed as a power series in i, retaining as many terms as necessaryto represerlt a regular cycle'
accuracy although some, such as Mars' ice caps, generally follow
the observations. For the outer planets a single term is usually sufrcient, heDce magnitude' V6'
.' For some purposes it is convenient to use the mean opposition
Az(i) is usually referred to as t"lte phase cofficient, The series or coefficients for x'hich is related to V(1,0) by
each planet and the Galilean satellites of Jupiter are presented in Sections 7.J and

Vo = V(1,0) + 5log16[a(a- l)1,

('7.r2-2)
If the intrinsic properties of a planet cause its brightness to vary, the problem
of predicting the apparent magnitude at a given time is considerably more compli-
units'
cated. The causesof intrinsic variability are (1) the existence of surface markings o[ where a is the semlmajor axis of the planet's orbit in astronomical
 390 EXPLANAToRY
sUppLEMENT OFTHESUN,MOON,PLANETS,
ANDSATELLITES 391
EPHEMERIDES
The surface brightness (,98) of an object is the average brightness value for 1Lu
illuminated portion of the apparent disk. The units for the surface brightness 61u
visual magnitudes per square arcsecond. This is represented mathematically by

SB = V + 2.5logrc(krab'), (7.124)

where a is the equatorial radius of the object, Dr is the appa,rent pola.r radius of
\
the object that will be given in equation 7.12-6, and & is the fraction of the objecl
Sub-Earthvector
illuminated that will be given in Equation 7.12 32.
The apparent disk of an object is always an ellipse. The apparent flatteninE
is always less than or equal to the flattening of the object itself (depending ol
the appaxent tilt of the objects axis). To derive the apparent pola.r radius and the
apparent flattening of an object, start with the equation for a.n ellipse. The ellipse
can be represented parametrically by
%**
tA$s'

,=(i +f)1,2,
x = acos -1, (7.r24)
)=bsin7, Ftgare 7.12.2
of the apparentpolar mdius, ,'
Thl geometricappearance
where x is the length of the projection of the point along the major axis from the
center of the ellipse, y is the length of the projection of the point along the minor the binomial expansion:
axis from the center, a is the length of the major axis, b is the length of the minor
r-l^
axis, and 7 is the angle between the line segment connecting the point with the b' = 4ll - 1 - /'? sin'z1) + sin' 'v - /' sin' r)' - l. (7.12-8)
center of the ellipse and the semlmajor axis (see Figure 7.72.2). For a planet, the ){zI "in' ;(2/
apparent polar radius is the length of the radius, R, whose planetocentric latitude negligible'
is the same as the planetocentric colatitude of the sub-Earth point, and , for the planets / << 1, so terms of order f2 or larger will be
? is the
latitude of the sub-Earth point. Using the definitions for x and y in the equation
for R
R= a(1- f sn?i
R = (a2 cos21 + b2 sir? 1)rt2- Q.12-s) = a(l - f cos2e)
= q.11- fl(b' 'b) /( lb'llbl)l'), (7.rz-e)
Rearranging the definition for flattening, Equation 2.11-1, gives

b/ is the vector from the planetocenter to the apparent polar axis, b is the
b=a.(r*f). (7.124) from the planetocenter to the true polar axis, and d is the planetocentric
of D/. The apparent flattening of the planet will then be
So
f'=(a-b'rta.
(7.r2-ro)
R = (a2 cos27 + a2(1- fl2 sin21)rt2
= a(cosz? + sin27 - 2f sb?1 + J2 sin21)12 re remaining tabulated quantities are most easily understood from Figure 7'12 3'
shows the appa,rent disk of an object as it is seen ftom the Earth' At the cen-
= a(l -2f s:rir2I+f 't)'''. Q.12-7) apparent disk is the sub- Earth point, e. Other points of reference are
the
"
 392 EXPLANATORY
SUPPLE[4ENT EPHEMERIDES
PHYSICAL ANDSATELLITES
OFTHESUN,MOON,PLANETS, 393

Nonh heliocentric ecliptic rectangular coordinates for the planet, r = (r,y,z), and the
te = @".y.,z) at a given Terrestrial Julian Date (TJD). The vector from the
to the object is then determined by subtracting re from r:

d=r-re
= 1x - xa,! - !e,z - zc)

= \xd,yd,zd. (7.r2-t t)

Ne)d, the light-time correction is found by determining the length of d and dividing
suusolar "sub-E*h by the speed of light, c, to obtain the time:

Geometric
terminator ldl= (d d)'i2=1fi+yf,+fi1t12,
t=ldltc. ('7.r2-t2)

The difierence I"/D - r is calculated to approximate the time at which the image
South of the planet visible at the Earth at TJD left the planet. This time, L/D - r, is used
Figure7.12.3 to determine the approximate position of the planet, rl = (rl,)l,zl), as seen ftom
Thediskof a planetasseenby an observeron theEarth.Thequantities shownare:e, thesub_Earth the Earth at time TJD. If desired, the previous process can be iterated to produce
poinl J. the_subsolar
point;4 theapparent distancefrom thesub_Earth pointto thesubsolarpoint; a more accurate planetary position; however, planetary motions are slow enough
th: posit.onalgle of rhesubsolarpoint;'l, the northpoleofthe planit; d, theapparent
distance
fA,
lrom that to the accuracy of The Astronomical Almanac orrly one iteration is required.
thenorthpolefrom thesub-Earth poinr;pA,, thepositionof thi nonhpole;ande, the
geometricmaximumdefectof illumination. The position of the Earth, r", and the apparent position of the planet, r1, are
then transformed from the mean equator and ecliptic of J2000.0 to the equator
and ecliptic of date using the precession routines described in Section 3.21 and the
nutation routines described in Section 3.22. The precessed and nutated position
subsolar point, s, and the north pole, r. For most of the plauets, the planetographic
rtectorc are designated r"4 for the Earth and 16 for the planet. A new vector d is
loagitude and latitude of the points e and s are tabulated. The apparent positions
computed by subtracting ra from r"6 as in Equation 7.12-11 and then correcting for
of the subsolar point, J, and the north pole, n, with respect to the apparent cenrer
aberration in position given in Section 3.25. The unit vector along the planet-Sun
of the disk (point e) are tabulated by the apparent distances, d., and.d,, and posi-
line, j, is a vector of unit length denoting the direction of the Sun from the planet.
tion angles, PA^ and PA, for z and s, respectively. The position angle
is neasured The vector j is found by dividing -16 by its length, lrdl. The unit vector along the
eastward from north, N, on the celestial sphere. The direction north bei[g defined
planet-Earth line, j6, is found in a similar manner.
by the great circle on the celestial sphere passing through the center of the object's
The position of the north pole of the planet, (oe, 6s), is then found from Davies
appa.rent disk and the true celestial north pole of date. The apparent distancc of
el ol. (1991). These coordinates are then precessedand nutated to the ecliptic and
a tabulated point is listed as positive if the point lies on the visible hemisphere of
equinox of date, (oa, 6a). A unit vector, ff, in the ecliptic rectangular coordinates
the object and negative if the point is on the far side of the object.
Thus, as point is then calculated:
n or s passes from the visible hemisphere to the fa.r side, or vice versa, the sign of
the distance changes abruptly, but the position angle varies continuously. fr = (cos 6dcos od, cos 6asin oa, sin 6a). (7.t2-13\
However,
when the point passes close to the center of the disk, the
sien of the distance re-
mains unchanged, but both the distance and position angle ir.y ,ru..yrapidly and This vector gives the planetocentric direction of the north pole of the planet in the
may appeax to be discontinuous in the fixed interval tabulations. coordinates of the equator and ecliptic of date.
The planetographic coordinates of the sub-Earth and subsolar points are calcu- Finally, the ecliptic rectangular coordinates of the unit vector pointing from the
lated in the follorving manner. First, the planeta.ry ephemeridesare used to 's center to the intersection of the planet's prime meridian and the equator'
calculate
 394 EXPLANATORYSUPPLEMENT
7 / PHYSICALEPHEMERIDESOF THE SUN, IUOON,PLANETS,AND SATELLITES 395
Since g runs from +90o to -90" (see Figure 7.12.3), there is no loss of generality in
pot directly determining cosg. The right ascension and declination of U a"rethen

n ou = od +900+ tan-t(sins/ cos,s), (7.12-17)

\ nonh pole
6u = tan- t (sin8 / cosg). ('7.12-18)

The coordinates for the sub-Earth and subsolar points are then computed by using
the vectors for the north pole of the planet and U to determine the rotation from
Planet'sequator the equatorial ecliptic coordinate system to give the planetocentric latitude and
of date
solarpoint planet,;'sub_ the planetographic longitude. (The planetographic longitude is supplied from the
Earthpoint definition of IV, and the planetocentric latitude is determined becausethe angles are
measued with respect to the axes of the coordinate system at the center of mass
of the planet and not the local normal on the equipotential surface.) The flattening
of the planet, f, is used to convert the planetocentric latitude to the planetographic
latitude. Using n as the z-axis, U as the.r-axis, and defining y as the y-axis of the
coordinate system,

X=Uxn' (7.12-19)

Figre 7.12.4
Planetocentricunit vectorsfor points of intereston the disk of the planet,andthe anglesbetweenthe The unit vector for the sub-Earth point is
Earth'sequatorofJ2000.0andtheplanet'sequatorof date.Thequantities depictedare:z, thenorth
poleof theplanet;n, thenorthpoleofthe Earth'sequatorandequinoxofJ2000.0;i, the
planetocentricunit vectorfor the northpole:J, thesubsolarpoint;j, theplanetocentric unit vectorfor (7.12-20)
jd=6d.U,jd.y,jd.n).
thesubsolarpoint;jd, theplanetocentric unit vectorfor thesub-Earth point;U, theplanetocentricunit
vectorpointingto theintersection of theobject'sprimeme.idianandits equator;W, theanglealong
the object'sequatorfrom the ascendingnodeon the Earth's equatorof dateto the object'sprime
meddian;g, the anglealong the greatcircle from rhe Earth'sequatorof J2000.0to the Earth'snofih The vector, ia = Qa,,jay,jd is converted into the latitude and longitude of the
pole_ofJ2000.0,whichgoesftom theEarth'sequatorto theintersection of theplanet,sprimemeridian sub-Earth point:
andits equatorofdate;and.S. theanglealongthe Earth'sequatorfrom the ascending nodeof the
planets €quatorto thepoint wheretheangleI reaches theequator. Theplanetocenfticunit vectorfor
thesub-Earth point,jd pointsdirectlyat theobserver. ,
@e= sln
-t_,
Vd:J, (7.r2-2r)
6L= tar-tltand l(l - f)21, (7.12-22)

)!.=ran I QayI ia). (7.12-23)

The planetographic latitude and longitude of the subsola.rpoint are found by

r, is needed. To obtain this vector, first determine the intermediate angles g and S
point unit vector, i6,
(Figure ?.12.4): .'.using the unit vector for the Sun, i, instead of the sub-Earth
the unit vector for the Earth, in Equations 7.12 2O throtgh 7 .L2 23.
The position angle and apparent distance for the subsolar point and north pole
sin g = sin I4lcos 6d, (7.12-14) are computed by determining the vectors specifying a coordinate system in which
the yz-plane is the plane of the sky, the x-axis points towa.rd the Earth, and the
sin S = sin Wsin 6a/ cos g, (7.12-15)
y-axis points toward the west as seen on the pla,ne of the sky. In the equatorial
cosS = cosW/ cosg. (7.12-16) of date, the direction for the north pole of the Earth is j, = (0,0, 1)' So
 396 EXPLANAToRY 7 / PHYSICALEPHEMERIDESOFTHE SUN, MOON, PLANETS,AND SATELLITES 397
SuPPLEMENT
the directions for axes, (u; u1,u), of the coordinate system in the ecliptic coordi141p. The geometric terminator is used in computing the phase and the defect of
of date are illumination. The geometric terminator is the plane orthogonal to the direction of
the Sun passing through the center of mass of the object (see Figure 7.12.1). The
r-axls, ur = jd; (7.12-24) maximum geometric defect of illumination, p, is the length of the portion of the
line segment across the face of the planet passing through the points s and e that
y-axrs, u7=j6 x j,; (7.t2_2s) is not illuminated (see Figure 7.12.3). The defect calculated does nol include non-
z-a-tis, ur = u; Xja. (7.12_26) geometric efiects such as the refraction of light by the planet's atmosphere. The
apparent equatorial diameter of the planet in seconds of arc is
The coordinates for the north pole vector in the preceding coordrnate system
are
found lrom the equatorial coordinates by s.c=206264.8062at|d| (7.12-33)

n " = p p ( n .u , , f r. u j , i . u * ) , The ratio of the apparent polar diameter to the equatorial diameter thai depends
(7.12-2'1)
on the latitude of the sub-Earth point (see Figure 7.12.2)
where po is the polar diameter of the planet and fr is the unit vector pointing
in thc
direction of the north pole for.urd in Equation 2.12 13. The position angle b'/a=l*f\l-sin2t), ('7.12-34)
of the
north pole is then
where / is the flattening of the planet and 7 is the latitude of the sub-Earth point.
PAn= tan-t 1rr , rurr. (7.12-28) The defect of illumination is then
The length of the projection from the center of mass to the north pole
on the plan<rt - PA, + 90')l(l - k).
Q = 2s"cll - (l - b' t a) sinz(PA" (7.r2-3s)
is
4 = 1"?^+ n2*)t,2 t1 l)-)Q\ The position angle for the defect of illumination, measured from celestial north,

Finally, the apparentdistancein arcsecondsis

PAg= P4' a 1gg" PA"< t80",
d, = 206264. / (ldl _ n*)1.
Bo62td (7.12-30)
= PA' - 180" PA"> 180: (7.t2-36)
Similarly, the position angle and apparent distance of the
. subsolar point is
calculated using the unit vector for the Sun, j, and
the apparent radius of thc
planet at the latitude of the subsolar point rather 7.2 PHYSICALEPHEMERISOF THE SUN
than fr and the apparent pol:rr
radius. The appa.rent radius of the planet at the
subsolar point is derived using
Equation 7.12 9 and substituting the planetocentric latitude The elements used to calculate the physical ephemeris of the Sun are from Carring-
of the subsolar point. ton (1863). The values of the physical constants for the ephemeris are:
0,.
The phase angle, i, of a planet is the planetocentric elongation
- of the Earth (l) mean sidereal period of rotation, P, = 25. 38 days;
from the Sun. This can be found from the dot product
ofj anJj6: (2) inclination of solar equator to the ecliptic, / = 7925;
(3) longitude of the ascending node of the solar equator on the ecliptic,
cosi=j.ia. (7.r2-31) tl = 75?76+ O?O1397T;
(4) the equatorial diameter of the Sun, 4 = 696000km.
The phase, /(, is the ratio of the illuminated area of the
disk to the total area of the
disk, as seen from the Earth: The flattening of the Sun is thought to be extremely small (f = 0). Because the
Sun is a gaseous body, it does not rotate rigidly. The sidereal period given is for
/<= 0.5(l + cosr). the meat rotation rate of the Sun.
(7.12-32)
 398 E X P L A N A T O RSYU P P L E M E N T S F T H E S U N ,M O O N P L A N E T SA N D S A T E L L I T E S
7 / P H Y S I C A LE P H E M E R I D EO 399
From the inclinatiorr of the Srrn and the longitudc of the ascendingnocle.1lr. position
of the Earth s st'lcnograPhiclolgitude. I.. and latitutlc, rl., and of the
position of the lort h l)ole of thc Sun is anglej C, of the axis of lol,ation. Irr The Astronotnical Almarnc' the contlilltltiolts
from the physical libratiolts are tabulated separa.tr:lyA third corttribution to the
.to = 28601300,60 = 6318700. (7.2-t) librations, the topocclttri<:optical librations, are a rcsult of the diflerencebetwcerl
the viewpoilts of aIt obsclvrlr on thc surface oI the Earth zrrrtlthe hypothetical ol>
pcriocl of the Sun, the position of the prime meridian is
From the mean siclerrctrl serverat the ccnter of Inassof the Earth. The topoccutrit:optical librations na1' be
as large as 1o and h:rvc ilrlpoltaltt efiects on the appilrclrt <:otrtourof the limb Be-
lV = 84110+ 14:1844000I. (1.2-)) causethe nragnituclt:of t hc topoccutric optical librations <lcpettclon tltc obsorver's
position, thel'are not labultrted tn The AstronornicaLAbttattat. Instead. a methorl
The heliocentlic lorrgitudeand latitudc of the Earth. thc sub-Earth poinl. arxl for estimating theur is givelt at the end of this sectioDaIr<iitr The AstronomicaLAl-
the position auglc of tlx, nolth pole are determined frrrnr the heliocentricposjtjr)r) mdrroc.The \Iultivo.rr Interactive Cornputcr Almanac (\lIC.'A). horvevet.computcs
of the Earth. r", thc l)ositi(nrsof the north pole. an. 6e. ancl thc prime rncrirliarr the topocentric opl ical lilrratiorrsfrom rigorous fornnrlas.The dilTerencebefrveerr
of the Sun. W. Thesc <lrrautitiesare deterrnilctl usirrg tlrc rrrothodsdescribedirr MICA and the giverrtttt:tltoclfor estirnatingthe topocclltric opt'icallibrations is, at
Section 7.12. most, 1".
The horizontal parlllax is the differcncein directiou l)ctweeu the topoceutt.i. The selenographi<cooldirtatesof the Earth and Srtlt spccify therpoint on thc
and the geocentric( oordinates of an object when tlrc colrtcr of the objcct is orr lunar surface u'herc lltc Eartlt and the Sun, respectivcly, altr irr thc sclcnographic
the astronomical}torizotr.'l'ltehorizontal parallax (HP), irr arcseconds.is compnterl zenith. The tabulateclscloltographiclongitude aud latitu(lc of tl(: Earth include the
usrng total geocentric.opti<trl, lnd physical librations itt longitrrtlc artcllatitude' respcc
sinHP = rR"/ lr" l. (i .2 3 ) tively. \\ihen the librertion irt selenographiclongitrrcleof thc Earth is positi'e the
mean certer of the clisk is displacecleastivardon lltt'<clcstial spllere exposing to
'ilherc R" = 6378.l37krn. thc lAI- cquatolial raclirrsof tlc Earth (Davics r'1 al..
view a region of thc $cstcrn lirnb. \\-hcn the librirtion irt sclcttoglaphiclatitudc of
1991) and r" is the Earth Srrrrdistnnce. the Earth is posilive. 1]rc Inctrtt center of the disk is <lisplzrcccltorl'ard the south'
Similarll'. the sclni(liarlctcr (SD) of the Suu ir ar'<sccorrrlsis cornputed usirg and a region of tltc nortlt liutb is exposcclto vieu'.
The principal ll()llx'rll of irtertia ol the \Ioorr ( loscsl t() lhe Eari h Nloon axis
SD=2o6264i8062d,t2r", (1.24) is offset 21412rvcst of t,ltomoirrtdirection of the Earth flrrllt llte \lool'
'l'he AstrononticoLAI
The selenogrir.lrlri<rcolongitrtcleof the Sun tabltlatrrl ir
where d" is thc cliarrrclcl of thc Surr. The semiclianx,lcl is therr conveltcd fi-otl manac is tlte easl,sclcttographi<r longitude of the tnomittg grrrtnetlic terminator' It
arcseconds to arcnrilllt{)s an<l scconds. is determined by sttlrtractilrgihe selerrographiclolgitrrtlt'ol thc Srrn from 90' an<i
adding 360' if tlrc lt'srrlt is ncgative.Colongitudcso1270'. 0', 90"' and 180' corre-
(]uarter'
sponcl.approxirratclr'. to \cn'\[oon. First Quartcr'. Frrll \lrxut. and Last
7.3 PHYSICAL
EPHEMEBIS
OFTHEMOON respectivel]-.Thc rgc of tltc \[txru is cornputed ns thr] lilln't'lapsc'l itL rla-ts sirrte
the most reccnt lrlorll('rlt zlt rvhich thc selenographi<cololtgituck'of tlte Sun llas
ln The Astrononicnl Abttanat clata for the \loorr is talulaterl to an accuracr',rf 2'10".
iotr
09001in longitucle.latilrrrlc. arrd position anglc. The position anglc of t It' trtirlpointof the bright lirrtb is l lrt' satrre:ts t he posil
On the ar'eragr'.thc siurrr:hcmisphereof the Nloon is alnavs turned torvald lltc angle of thc sttbsolar poitrt. trtrri it is computctl iu tlrc Irtntutt:ldescribed for thc
r-'f
Earth, but there arr: pt'riodi<roscillationsor librations of l,lrc appalent positiol (t subsolarpojnt in Scc:t,iorr 7.12. l,ike thc other positiorrarrglcs,th<:pt,sition angle
the lunar surface thzrl, allows aLrout 59% of the surfacc to b<:scetr from the Earth. the midpoint ol thc lrrighl lirtb is measrrredcastwattl tr'rotttttlthe disk from north'
quantitv
The librations ale chrc partly to physical rotational librltions oscillations of tlrc The fraction of l ltc Nlootrillurninated,&, is the sattxr:ls l'lx: pltasc This
actual rate of rotatiou of t,lurN4oonwith respectto il,strc:lr r'o1,ill,ion
ratc. A second, is computed as dtscrilrrxl irr Sccl,ion7.1.2.
much larger part of llrc lilrrntions are the geocentrit r4rtir';rllibrations, rvhich arc Thc analytic thcotv for llre libratiorr of the Nloorr uso<litt ?/rc Aslrcnomicol
the result of the nonuniforrn rnotiol of revohrtion of th(: Nloon ar.oundthe Earth Almanac is that of Eckltalclt (19s1' 19s2) for the ph1-si< al librafion of the \'Ioon;
NIoon barvcenter. Both of thesc cffectsare taken into accouut in the cornputati()u however. the IAU r,.alucfor thc inclination of the mean h liu eqllator to the ecliptic,
 OFTHESUN,MOON, ANDSATELLITES
PLANETS, 401
400 EXPLANATORYSUPPLET!{ENT 7 7 pHYSICALEPHEMERIDES
of the observer and I is the local houl angle of
l'32'32:7 , is used instead of Eckhardt's value. In addition, there are printing err111. arbercO is the geocetrtric latitude
in Eckhardt's published values. so any attempt to use the theory must be done with 1[e Moon' given bY
great care. Although the valrres of Eckhardt's constants differ slightly from thos. h=LST-o^, (7.3-9)
of the IAU, the diflerence is of no consequencefor the precision of the tabulatiorr.
except for the inclination of the lunar equator. First, the geocentric optical libratiols $/here LST is the apparcnt local sidereal time'
are calculated from rigorous formulas; then, the combined geocentric optical nnrl Second difierences nust be taken into account in the interpolation oftho tabulnr
physical Iibrations are calculated using the results frorn the optical librations alone. qeocentric libratiorrs to the time of observation
Included in the calculations are perturbations for all terms greater than 010001. The semidiameter of the Nfoon in radians is simply,
Since the apparent positions of the Sun and Moon are used in the calculatious.
aberration is already included, except for the inappreciable difference betv-een the, sD=ran-r(R,'./lr.l) (7.3 l0)
light-time from the Sun to the Niloon and from the Sun to the Earth.
The differential corrections to be applied to the tabular geocentric librations where Rm = 1738km the radius of the moon and irml is the Earth \Ioon distance.
to form the topocentric librations are derived in Atkinson (1951). The formulas 1b1
the topocentric corrections arc 7.4 PHYSICAL OF THEPLANETS
EPHEMERIDES

a^m = - HP',(in(O - C) secb, (7.3 l) In The Astronomi,calAlmanac, the planetsN{ars,JuPiter. Satl1rll,Urarms' Neptuue'
Pluto, and all of the planets in MICA have a tabulated elcrnctrt in their physical
Af.^=+HP'cos(Q-C), ( 7. 3 2 \ ephemeridesin additiou to the quantities describedin Scctiou 7.1. This quaul,itv
AC = +sin(i + Ar3)l HP' sinQtan6, (7.3-3) is the planetocentric orbital longitude of the Sun (t5) L5 is rneasured eastu'ard in
the planet's orbital pl.rnc from its vernal equinox. The planct's vernal equinox is
where ,\. is the selenographiclongitude. d'n is the selcnographiclatitude. C is th0 the plane of its
the point where the ascert<lilgnode of the planet's equator crr'rsses
position angle of the axis of rotertion,0 is the geocentricparallar:fic angle subtcrrdcrl orbit about the Sun. Thc iustantaneousorbital and equatorial planes are used irt
by the Moon, and F1P'is the parallax between the point of obserration on the Eart[ -q computing as.
surface and the geocenter commonly called the topoccntric horizontal parallax. Ls = cos-r[(rp. r,") / (lro lr," )], ('7.4-t)
The horizontal parallax, HP, occasionally called thc geocetrtric horizontal par.-
allax, is found from the Earth Moon distance. r.l, and the c(luatorial radius of I hr' where ro are the ecliptic rectangular coordinates of thc plzrnet arrd ru" are thc eclil>ti<
Earth, R" = 6378.137km. rectangular coordinatcs of the planet's irlstantancous vt:rDal equinox The corrctrl'
quadrant is determined frorn the sign of
sinAP = lrml / Re. Q .34J
rP x rve. \7 .4-21
The topocentric parall:rx. HP', is computed frour thc horizontal parallax bi
using Values of L5 of 0", 90'. 180", and 27Oo corrcspond to the beginniug of sPtirtg'
s i nH P ' = s i n : s i nH P l ( 1 - s i n zs i n H P ) , (7.3 5) summer, autumn, ancl u'inter, respectively.for the Pl^net's northern hemisphcrc'
Unlessothern'isetroteclthc positionsof the pl:rttctarl-polesand the priute rnctid-
where z is the geocentric zenith angle of the \'Ioon. 'fhc values of Q and ; ar-c ians and the planetary radii and flattenings arc takctt liont Davies el ol' (1991)'
calculatedfrom the geocentricright a.scension, The values for the magnitudc of the planets at .t distartce of 1-tu and at a phasc
o., and clet:lirration,
6., ofthe N{oor.
ustng angle of 0", as well as the value for the phase cocflit:icrtt arc from Harris (1961)'

sinzsinO=cos@sinr, (1.3-6)
7.4.t Mercury

sinzcosQ = cos6. sin@- sin6mcos@cosr, (7.3-'7) The position for Mercury's polc, prime meridian, radius, V(1,0), phase coefficicttt'
and flattening alc givcn in Table 7.41.1.
cosz = sin6. sin@+ cos6. cos@cos,h, (7.3-8)
 402 EXPLANAToRY
SUPPLEMENT OF THESUN,IUOON,PLANETS,ANDSATELLITES
EPHEMERIDES 403
Tabl€7.41,1 phase coefficient (Danjon, 1949). The expression was deterrnined for the
PhysicalEphemeris
Parameters
tor Mercury range 0:9 < i < 170?7.For i < 099 and i > 17O?7,Venusis too closeto the
be observed photometrically. Venus'change in brilliancy with rotation is
d o = 2 8 1 9 0-10 9 0 3 3 I
6o= 61945- 09005I small.
W = 329971 + 691385025 t
Am(t - 3.80(i/100o)- 2.73(i/ 1oo")2
+2.00(l/ loo')3
v(1,0)= -0.36 Mars
F = 2 4 3 9 . 7 11 . 0 k m
f=0 basic parameters for Mars a.regiven in Table 7.43.1.

Table7.43.1
for
PhysicalEphemerisParameters
Mars
Since Mercury's orbit about the Sun is interior to the orbit of the Earth, Mer-
cury is seel to go through the entire range of phase angles. Becauseit passesthrough ao = 3'17968'l - 09108I
such a la,rgerange in phase angle, the phase coemcient must be replaced by an an- do= 529886- 09061r
W = 1769868+ 35098919830 t
alytic expression (Danjon, 1954). The expression for the phase coefficient is good y ( l , 0 ) =- 1 . 5 2
for the phase angle 30 < i < 123'. For i < 3' and i ) 123", Mercury is too close to Am(4= 0.016i
the Sun to be observed photometrically. B=339714km
The flattening is not appreciable so the equatorial and pola"r diameters are the J=0.0065t0.0017
same.
The cartographic prime meridian of Mercury is defined so that the 20' meridian
Although the phase angle of Mars ca.nbe as la.rgeas 47", the phase relation for
passesthrough the center of the crater Hun Kal.
magnitude is satisfied by a single term, Mars' surface contains a la,rge number
visible markings. The rnarkings cause the brilliancy of Mars to vary about 0.15
7.42 Venus over its oeriod of rotation. However. the amplitude of the variation ls
fixed. and is modified bv other factors such as the transpa,rencv of the atmo-
The basic paxameters for Venus are given in Table 7.42.1 from planetwide sand storms, There appears to be an annual variation as
Venus rotates in a retrograde fashion, so the sign for the motion of the prime All these additional factors can cause variations in the visual magnitude of up
meridian is negative. Like Mercury, the flattening of Venus is negligible; so the .1 magnitudes.
equatorial and pola"r radii of Venus are the same.
. The cartographic prime meridian of Mars is defined so that it passesthrough
Also, like Mercury, Venus'orbit is interior to the Ea,rth's orbit, so it has a phase center of the crater Airy-O.
angle that r,aries ftom 0o to 180o, and so an analytical function is used rather than

Jupiter
Table7.42.1 basic pa,rametersfor Jupiter are given in Table 7.44.1.
PhysicalEphemeris
Parameters
lor Venus
Multiple longitude systems a.re defined for Jupiter. Each system corresponds
ao = 272?72 a different apparent rate of motion. System I applies to the mean atmospheric
60= 67915 rotation of the planet. System II applies to the mean atmospheric rota-
l4l = 160926- 194t113596 t nofth of the south component of the north equatorial belt and south of the
v(1,0)= -4.29
a/r(r) = 0.09(i/100o)+ 2.39(i/ 1oo.)2- 0.65(l/ 100.)3 compotrent of the south equatorial belt of Jupiter. System III applies to the
B = 6 5 0 1 . 9 11 . 0k m of ra.dio emissions on Jupiter. The rotation rates for both System I and Sys-
J=o a,re uncerta.in and subject to change. The sub-Earth and subsolar longitudes
in The Astronornical Almanac are based on the Svstem III longitudes.
 404 E{PLANATORy
SUppLE[4ENT
EPHEMEFIDESOFTHE SUN, MOON, PLANETS,AND SATELLITES 405
Table7.44.1
PhysicalEphemeris There is, however, a,rr additional term for variation in the brightness of the
Parameters
tor
Juoiter with the aspect of the rings:

do = 268905- 0P009r (7.45-r)

6o= 64949+ 09003I % = -2.60sinlel+1.25sir]e,
System |'V, = 679t+877:900t
Systernll ry,, = 4393+ 8709270t
Systemlll ry,/ = 284995+ 87095360000
t V, is the visual magnitude of the rings and e is the Saturnocentric sub-Earth
y0,0) = -e.25 There is little evidence for variations in the magnitude in the brightness
Am(0 = 0.005i with its rotation or over extended periods of time.
R =71492! 4km
f = 0.064t|7I 0.00015

Uranus
'basic
Jupiter is the most luminous of the planets. Variations itr the brilliancy of the parameters for Uranus a.re given in Table 7.46.1.
planet with rotation are small, -0.01 magnitudes, and irregular. There a.re also
long-period r,ariations correlated with changes in the surface features of the planet. Table7.46.1
for
Parameters
PhysicalEphemeris
Uranus
7.45 Saturn
as = 257?43
The basic paraqreters for Saturn are given in Table 7.45.1. 6 0= - 1 5 9 1 0
For Saturn only System III fobserved rotation of radio emissionsl Desch and Systemlll t44l= 203981- 50191600928
t
v(1,0= ) -7.1e
Kaiser (1981)] rotations are defined. AIso, like Jupiter, the mean rotation rate of
Am(il = o. oo28i
the atmosphere is uncertain a,nd subject to change. The tabulated sub-Earth and F=25559i4km
subsola.rlongitudes in The Astronomi,cal Almanac use the System III longitudes. I = 0.02293t 0.00080
The previously defined System I rotation systern has been found to be of little use
since the Voyager encounters with Saturn, and is no louger supported by the IAU.
Because of the brilliance of Saturn's rings, the brilliance of the planet alone is Like Venus, Uranus rotates in the retrograde direction. There is evidence for
not well known; so the combined planet-ring magnitude and surface brightness are of the brightness of Uranus with rotation, but it has not been confirmed'
tabulated. The phase coefficient for Saturn includes the rings. The visual magnitude term with a period of 84 years is expected in Uranus' magnitude due to the
in Table 7.45.1 is the extrapolated value for the planet alone at lAU and 0o phase ming of the plaDet and the high inclination of its equator to its orbital path'
as for Saturn, only the rotation rate for ra.dio emissions (System III) is defined

Table7.45.1
PhysicalEphemerisParameters
lor
Saturn
Neptune
ao = 40958- 09036f
6o=83954-09004I basic pa,rameters for Nepture a,re given ln Table 7.47.1'
Systemlll ryi, = 38990+ 8'1097939024
t
Y(1,0)= -7. 19 The phase angle of Neptune at its greatest elongation is very small, so the
a/I?|lil=o.o44i in brightness of the planet from its phase angle is unknown' There is
F=60268t4km evidence of variation in the brightness of Neptune with its rotation' There is
J=0.0979610.00018
of long-period variations in its brightness Only the System III rotation
radio sources is defined for Neptune.
 ANDSATELLITES
OFTHESUN,MOON,PLANETS'
EPHEMEBIOES 407
406 EXPLANATORY
SUPPLEMENT / PHySICAL

fable 7.47.1 the Sun. The rotational difference between the minimum and maxirnum
PhysicalEphemerisParametersfor Neptune was 0.11 magnitudes in 1954 (Harris, 1961) and was 0 291 magnitudes
rotational variations is 6.3867days.
i" ieat (Mur.iuti", 1988).The period of the
oo = 299936+ 0907sinN
i,,;ns the eclipseseasonthe Pluto-Charon mutual events causechangesin the
60 = 43946 - 0951cos N
deepas 0'50 magnitudes(Tholen et ol ' 1987)'
Systemlll Wi, = 253918+ 53693128492 | - 0:48 sinN Irug"it"a" as
v ( 1 , 0 )= 6 . 8 7
R=24764t15km
J = 0 . 0 1 7 11 0 . 0 0 1 3
where 7.5 PHYSICAL OF THESATELLITES
EPHEMERIDES
N = 357985+ 529316f

ThedisksofalltheplanetarysatellitesasidefromtheMoonaresosmallthat
th"y be directly observed, except from interplanetary probes Therefore,
"".rrrrotof information tabulated for the physical ephemerides is mttch smaller
the amount
7.48 Pluto
than that for the Planets'
peliod of the satellites are
The basic paxameters for Pluto are given in Table 7.48.1. Un]ess otherwise Stated, the values for the sidereal
(1991), and the value for the phase coefficients are from Harris
The position for the north pole and the prime meridian of Pluto is derived from from Davies et aI.
Harrington and Christy (1981). The motion of the prime meridian and the position (1e61).
of the pole for Pluto are given under the assumptions that the satellite's orbital
plane is coincidental with Pluto's equatorial plane and that the orbital motion of
7.51 SatellitesoJMars
Pluto's satellite is synchronous with Pluto's rotation. Like Uranus, the rotation of
with
Pluto is retrograde, and the inclination of Pluto's axis to its orbit is nearly 90'. Both of Mars' tq/o small satellites, Phobos and Deimos, are triaxial ellipsoids,
Because of Pluto's great distance from the Sun and its small size, its flatteniDg is their longest a-xispointed toward the planet The source for the reference surfaces
unknown, but is presumed to be small because of its slow rotation rate. of these satellites is from Mariner I measurements by Pollack et al' (1973)'
The given brilliancy at a distance of 1 ,tu and a phase augle of0o is the combined The basic parameters for the Martian satellites are given in Table 7 51'1'
magnitude of Pluto and its satellite, Charon. The brilliancy given is from Harris
(1961); however, the brilliancy of Pluto has been changing over time as differerrt
Table7.51.1
albedo features of the planet become visible (see Marcialis, 1988). The phase coef- RotationParameterstor lvlars'Satellites
ficient given was determined by Binzel and Mulholland (1984). There is a chauge
in the combined magnitude with rotation of the Pluto-Charon system due to the lPhobos: do = 317968 09108f+1:79sinM1
6 0= 5 2 9 9 0 0 9 0 6 1f - 1 9 0 8 c o s M l
differing albedos of the two bodies and surface markings on Pluto. The rotational -eP - 1942sinM1- 0978sinM2
w = gsloo* ltze:a+45850 f+ 096644x 1o
change in magnitude, like the mean magnitude! has been changing with tine as ll Deimos: ao = 316965- 09108I+ 2998sinM3
different albedo features of the planet are exposed to the Earth in Pluto's orbit 6 6 = 5 3 9 5 2 - 0 9 0 6 1r - 1 9 7 8 c o sl ' t
^'t3
W = 79941 + 2a5?1618970t - 09390 x 1O-1oP - 2958 sin M3 + 0919cos
The values tor M^, are
M1 = 169951 - 094357640t
et2
Table 7.48.1 Mz= 192?93+ fi2A? 4096700f +0:6&4 x 10
PhysicalEphemerisParametersfor Pluto nfi = 53947- 090181510 f

ao = 313902
6o = 9909
', Both of these satellites have rotation periods that are synchronous with their
w = 236977 - 56?3623195 t orbital periods. Also. both satellites have secular accelerations in their rotation
Y ( 1 , 0 )= - 1 . 0 1 which apoear as the I terms in the motion of their prime meridians' These
An = 0.041i of the
F = 1 ' 1 5 11 6 k m t"a-a ur" the result of secular changes in the mean distances
motion
satellites' orbits. The difierent signs for the quadratic terms for the
408 EXPLANAToRY
suppLE[4ENT
of the satellites' prime meridians shows that Phobos' orbit is contracting with time
while Deimos' orbit is expanding. satellites
]n5#"t"t"r" fo,Jupiieas
The massesof Phobos and Deimos a.re derived from a redetermination by Siq.
rvrM€tis:
'- ao = 268905- 0:009 7
gren (1983) of the perturbations of the Viking Orbiters during their close flybys of do= 64:49 + 09003r
the satellites. Information on the magnitudes of Phobos and Deimos, their albedos, w = 302?24+ 1221?2489660
t

and their colors are derived from Viking photometry of the satellites that has been Yv Adrasl6a: oq = 268905- 09009r
"' 5^ = 64949+ 0:0m r
analyzed by Klassen, Duxbury, and Veverka (1979). Because the satellites are verv g95o4oo
ti = 5:7s + 12069 t
near Mars, information on the phase coefficients for both satellites and the 1U - g; f - 0:84sinJj +0:01sin24
v' Arnalthoa: ao = 268:05 - O:OO9
color for Phobos are uselessfor ground-based observations and are not given. .^=9949+O9oO3r-0936cos4
w = 231?67+722?69145'60l+ 0:76sin4 -0:01sin24

xfvThebe: ao =26s905-o9oogr -2?126in4+0?o4sin2h

of Juoiter
7.52 Satellites ,; = 64:49 + 0:003 T - o:91 cosJ2 + 0901cos24
tv = g:gt + sgg9zoossgot + 1:91 sinJ2 - 0:04 sin24
Physical ephemerides are tabulated for sixteen Jovian satellites. The most complete llo: oo = 268:05- 09OOgr+ O:Og4sin4 + 09024sinJ4
r+ 09040cos4 + 0:011cos4
6; = 64950+ O9OO3
and accurate information is that for the four Galilea.n satellites and four inner
w = zoogss* zos:a€ss$a | - 09085sin.-b- 0:022sinJ4
satellites, JV (Amalthea), JXIV (Thebe), JXV (Adrastea), and JXVI (Metis). Anr(i=0.O46i-0 0010i2
The basic parameters for Amalthea, Thebe, Adrastea, Metis and the Galilean do=268:08-090ogr+1:086sinJ4+09060sinJ5+0:015sinJ6+0:009sin'-h
Eurcoa:
satellites are given in Table 7.52.1. ai = e+3st + ogoosf * o9468cosJ4 + 09026cosJ5 + o:007cosJ6' 09002cos4
vi = 35?72 + 1o1?g747235t oggSosinJa - o:o54sinJ5-09014sinJB-0:008sjn4
These satellites all have periods of rotation that axe synchronous with their -O OO125t2
AnIn = O.O312i
orbital periods.
lll Ganym€d€:
' do = 268920- ogoogI - 09037sinJ4+ 09431sinJ5+ 0'oo9lsinJ6
The values for the visual magnitude at 1AU and 0o phase angle, the albedos, 6; = 64957+ o:oo3 7 - o9ol6cosJa+ 09186cosJ5+ o9039^cosJo
and the colors ofthe Jovian satellites a.retaken from Morrison (1984). The Galilean vi = 43914+ 5093176081 I + o:o33sinJ4 - o:389sinJ5 0:082sinJ6
satellites have been observed often enough that well determined phase relations have Am(rl = o.323i - o.ooo66F
been derived for them. lv callisto: oo = 268:72- o:oo9 r - 09068sinJs + 0:590sinJ6 + 09010sinJg
6; = 64983+ 090037 - 0:029cosJ5+ 09254cls J6 0:004cos4
The reference surfaces for Amalthea (Smith et al.,1979); the Galilean satellites f + o:061sinJ5 09533sinJ6- 0:009sinJg
ti = 25g?67 + 21?57107-15
(Davies et ol., 1991); and JXIV, JXV, and JXVI (Thebe, Adrastea, and Metis; = -
A.']/i) o.07ai 0.0027 4i'
Burns, 1986) were derived from measurements of the Voyager spacecraft pictures *iero th€ valu6sol J, are
of the satellites. Estimates of the size of the outer Jovia.n satellites are from Tholen 4 = 73932+ 91473997
J" = fiA? 54 + 44243?A T
and Zellner (1984) and are based on the albedos of JVI and JVII (Himalia and ,4 = 283390 + 4a5097r
Elara). Ja= 355:80+ I191:3r
J 5= 1 1 9 9 9 0 + 2 6 2 : 1 7
Galilean satellite massesa,recalculated from the perturbation in the paths ofthe J6=229?80+64?37
Pioneer and Voyager spacecraft flybys (Campbell and Synott, 1985). The massesot J7 = 352?25 + 23A2?6 T
the rest of the satellites are based upon their individual sizes and the mean density 4=113935+607090
of the compositions inferred from their albedos,
Three ofthe Galilea.n Satellites have cartographic coordinates defined by surface
fable 7.52.2
craters, The cartographic definitions are given it Table 7 .52.2. StandardCartographic
Longitudes for Jupiter'sSatellites

7.53 Satellitesof Saturn Satellile Crater Merldlan

Europa Cilix 182.

Physical ephemerides a.re tabulated for 17 Saturnian satellites. The rotation pe-
Ganymede Anal 128"
riods are tabulated for 13 of the Saturnian satellites. The rotational elements in Callisto Saga 326'
Table 7.53.1 for 1l of the satellites a.resupplied by Davies eJ ol. (1991).
 Table 7.53.1 7 / PHYSICALEPHEMERIDESOF THE SUN, MOON, PLANETS,AND SATELLITES 411
FlotationParametersfor Saturn's Satellites
Table7.53.2
StandardCartographic Longitudes
oo = 4096 - 09036r for Saturn'sSatellites
60 = 83:53 - 09004r
ry=4a:8 + 62690440000l
ao = 409€€ - 09036r S.tellite Craler Merldlan
60 = 03:53 09004r
,v= 137:88+ 598:3060000 r irimas Palomides 162"
oo = 40958 - 09036r Enceladus Salih
60 = 83:53 - 0:oO4r Tethys Arete 299.
W = 296:14+ 5879289oooor
Dione Palinurus 63'
ao = 40958 - 09036r
60= 83953 - O9oO4 r
Bhea Tore 340'
w= 162:92+ 5729789t0OO I lapetus Almeric 2764
o o = 4 0 : 9 r B- 0 9 0 3 6 1- 3 9 1 5 3 s i n q+ 0 9 0 8 6 s i n 2 q
60= 83:52 - o9oo4r - o9356cosq +o:oo5cos2q
W = 293:87+ 51894907239 r + 39133sinq - 09Os6sin2q
ao = 40958 - 09m6r - t9623sin S 2+ o 9 O 2 3 s i n 2 q
6 0= 8 3 : 5 3 - 0 ! 0 0 4 r o : 1 8 3 c o s 3 + 2 O 9 O Oc Io s 2 q
w = 5 8 : 8 3 + 5 1 8 9 2 3 5 9 8 7 61r9+6 1 3 s i5n2 o 9 o 2 3 s i n 2 q
Wisdom, Peale, and Mignard (1984) have shown that one of the major Satur-
ao = 40:66 - 09m6r+ 13956sinq
60 = 83:52 - 09CD4I- 195,3cosq nian satellites, Hyperion, is tumbling chaotically rather than being in synchronous
! y = 3 3 7 9 4 6 + 3 8 1 9 9 9 t 5 5 5 01I 3 9 € s i n q - 4 4 9 8 5 s i n q
rotation. This tumbling is a result of Hyperion's triaxial shape and its rather ellip
ao = 40:66 - 09€6 r
,'0 = 83952 - O9OO4 r tical orbit with a mean distance near the 4/3 Titan resonance.
w = 2:82 +26297318996r
The reference surfaces for the larger satellites are derived from Smith el ol.
ao = 40966 - O9o3or+ 9966sin54
60 = 83:52 090041 1:ogcos54 (1982). The size of the smaller satellites are provided by their geometric albedos
w = 10p45 + 190:6979085r - 9960sinS. +2:23sin.g
(Thomas et o,., 1983).
do = 50:50 - 0!036 r
do = 84906 - O9oO4 r Morrison, Owen, and Sonderblom (1986) determined the magnitudes, albedos,
w=56:8a + 19096979330t
and colors of Saturn's satellites. Iapetus is especially interesting, becausethe leading
oo = a0:58 - 09036r+ 139943sin q _ t:686sin2Ss
60 = 83:43 - 09004r - 19572coes5 + 09o95cos2ss hemisphere is over two magnitudes dimmer than its trailing hemisphere. Windorn
=
W 149936+ 19096742373 r 13:849sin55 + i96g5sin2S5 (1950) gives a parametric equation for the brightness (not the magnitude) of the
ao = 40966 - 09@6r
60 = 83:52 - 09004r satellite of
W=357900+131953a$161
B = O . 5 7 1- 0 . 4 2 9 s i n o , (7.s3-r
)
d o = 4 0 9 5 8 - 0 9 @ 6 r + r 9 6 6 2 s i n q+ O 9 0 A 4 s i n 2 q
6 0= 8 3 : 5 2 - 0 : 0 0 4 r , 0 9 1 8 7 c q + O 9 o 9 s @ s 2 q
,y = 245:39+ 131!61740s6 I 19651sinq + oPO24sin 2q where q is its orbital longitude, with the maximum occurring at greatest western
do = 40938 - o:mor+391osinq
6 0= 8 3 : 5 5 0 ! 0 0 4 r o : 3 5 c o s S r elongation.
,v = 235:r6 + 7996900478 r 3908sinq Tyler et al. (i981, 1982) determined the tabulated masses for Mimas, Ence-
do =36:41 - 0:036r+2:66sinSs
60= 83:94 - 0!004r o93ocosq ladus, Tethys, Dione, Rhea, Titan, and Iapetus from the deflections of the Voyager
ty = 189964+ 2295769768 I - 2964sinq trajectories. The masses of Enceladus and Mimas are of lower accuracy and are
d o = 3 1 8 : 1 6 3 9 9 4 9r
6 0 = 7 5 : 0 3- 1 9 1 4 3 I
contradicted by resonance theories (e.g., Greenberg, 198,1).
w = 350?20 + 495379572| Six of the major Saturnian satellites have cartographic coordinates defined by
do = 355916
6 0= 6 8 9 7 0 - 1 9 1 4 3r
surface craters. The cartographic definitions are given in Table 7.53-2.
W = 304970+930:$34720r
whor€ lhe valuesol S, aB
q = 353932+7570697r 7.54 Satellitesof Uranus
52 = 2A?72 +75706?7 T
q = 177:40- 3650595r
Sa= 3OO:00- 722599r The basic parameters for the 15 known satellites ofUranus are given in Table 7.54.1.
S s= 5 3 : 5 9 - 8 9 6 a 9 A r
q = 1 4 3 : 3 8- 1 0 5 $ : 5 r
The radii for the Uranian Satellites (Smith et al'. 1986) are derived from mea-
s 6 = 3 4 5 : 2 0- 1 0 1 6 9 r3 surements of Voyager photographs. The massesfor the five large satellites (Viellet,
q = 29980 52?1T
q = 3 1 6 : 4 5+ 5 0 6 : 2r
1983) are derived from mutual orbital perturbations. Smith el ol. (1986) determined
 EPHEI\'IERIDES
PHYSICAL OFTHESUN,I\4OON,
PLANETS,
ANDSATELLITES 413
Table 7.54.1
Botation Parametersfor Uranus' Satellites magnitudes, albedos, and colors for the large Uranian satellites from Voyager 2

Coddla: ao - 257931- 09153lnUl

60 = - l s9l E + 0914cosUl
W = 127?69 - 107495295730r - O9O4sln U1
Ophslia: ao = 25793r- 0909dnl4 .86 Satellitesof Neptune
6 0= - 1 5 9 1 8 +0 9 0 9 c ots4
w = 130935- 9509{b8150I - 0903sinU2
has eight known satellites. However, the physical ephemerides of Nereid
do = 257931- 09r6sln(,3
60= - 15918+ 0916costA and Thiton are the only ones presently included in The Astronomical Almanac-'lhe
W = 105946- 82893914760r - O9O4sin 14
basic paxarleters for seven of the saiellites are given in Table 7.55.1.
oo = 257931- O9O4si.U.
60 = - 15: 18+ 0904cosU.
W - 5 9 : 1 6- Z 6 : 5 8 1 6 3 2 0-t 0 9 0 1 s i U
n4
o o = 2 5 7 9 3 r- 0 9 1 7 s i U n5
6 0= - 1 5 ! 1 8+ 0 9 1 6 c o 6 4
w = 95:08 - 760:0531090 r - 0!04slnUs Table7.55.1
ao = 257931- 0906slnU6
for Neptune's
RotationParameters Satellites
6 0= - 1 s 9 1 8 +0 9 0 6 c o s 1 4
W = 302956- 73o:1253660 t - O9O2 sinUs Naiadi oo = 299:39+ 0:70sinN 6:49sinNl +0925sin2/v1
ao = 257931- 0909slnt, do = 43:35 - 0:51 cosN - 4:75 cosN1+0909cos2N1
6 0= - 1 s 9 1 8 +0 9 0 9 c o4s w =254?0a+e22:81A1209 t 0948sin N+ 4940sinNr -0927sin2N,
l,v- 25903- 701:,rE65870I - 0902sini-t
Thalassa: do = 299939+ 0970sinN - 0928sini/2
do = 257931- 0929slnl4
60= 43:,14 - 0:51cosN- 0:21cosA/2
60= - 15918+ 0P28coEUs
w = 314:90- 6,14:6311260 r - 0908lintb
w = 102906 + 115597555612r - 0944sinN + 0919sinN,
oo = 257931- 0903sini4 D€spinal oo = 299939+ 0:70sinN - 0:09sinw3
6 0= - 1 5 9 1 8+ 0 9 0 3 c o s 1 4 60 = 4{}944 0:51 cosN 0907coslrl3
w - 29794$ - 5l/9.62A170I- O9o1sinlt l4 = 306:51+ 1075:7341562t - 0949sinN+ 0906sinNr
ao = 257931- 0933sinU1o
6 0= - 1 5 9 1 8+ 0 9 3 1 c o460 Galatea: oo = 299939+ o9Tosinw - o:o7 san/v4
w =91:24 - 472?545c690 r - 0909sinqo 60= 439,13-0951cosN-o:o5cos/V4
ao = 257:43+4941si. Ufi - O904sin2ulj l - 094asinN + 0905sinN!
l,v = 258:09+ 839:65976a6
6 0= - 1 5 9 0 8 + 4 t 2 5 c o s U 1-r o 9 O 2 @ s 2 U 1 1
w = 30:70 - 25496906892 l - 1927sin142
vtl Lanssa: ao = 299:38+ 0970sinN - 0927sinN5
+ 0 : 1 5 s i n2 q 2 + l : 1 5s i nq 1 - 0 9 0 9 s l n 2 U r r 60= 43940 - 0:51cosN - 0920cos N5
t/v= 179941+ 649905344701- 0948sinN + 0919sinN<
ao = 257943+ 0929sinUt3
6 0= - 1 5 9 1 0+ 0 9 2 8c o s4 s v l oo = 299:30+ 0:70sinN 0:05sinN6
r,l/= 156922 - 14298356681 I + 0905sh U12+ 09o86inUr3 60= 42:90 -0:51cosN 0:04cosw6
a o = 2 5 7 9 4a 3 0 9 2 1 s l nq . t - 0948sinN + 0:036sinN6
W = 93:38 + 320:7654228
6 0= 1 5 9 1 0 +0 9 2 0 c oq84
yy = 108905- 8698688923 r - O9O9 sinU12+ 0906sinq4 ao = 299936-32935sinit-6928sin2w7 2:08sin3lV7 0974sin4At -
lll o:28sin5At- o911sin6& 0907sin7tq- 0902sin8Ah -
l-n€nia: ao = 257:43+0:29sin q5
6 0= 1 5 : 1 0+ 0 9 2 8 c o 8 U r s 0:01 singtg
W = n9 74 _ 419614316 t + 09OOsin Ui5 do- 41917 +22955cos/g+291ocos2tg+0955cos3At +0916cos4^t+
lv Ob€ren: do = 257:43+ 0916slnUlb 0905cosstq + 0: 02cos6lv7+ 09o1cos7tg
do = - 15910+ 09t 6 cosirr6 l,v = 296: 53 - 6112572637t + 22?25 sintt + 6:73 sin2tt + 2:05sin3/g+
W=6?n - 2fj7394932t + o?u sinUft 0:74sin4^t + 0928sin5^h+ 0911sinGih+ o:05sin7Ig +
rn6re lhs valu€s ol U" are 0902sinSlt + 09ol singlvz
ul = 11597s + s4991:87t tb = 101:81r 12872963 r |fhorelhe valuesof /V, are
t4 = 141:69+ 41887966 t U1o= 138964 + 8061:81r N = 35798s+ s29316t
14 = 135903+299279357 u11=102923-2024?227 rvr = 323992+ 6at0696 r
u4 = 6197r + 25733?s9 T Up = 316?41 + 2863996r l'12= 220?51 + 5fi64? 2 T
Us = 249?32 + 24471?46 T ul3 = 3(x:ol - 51:94 r It = 314327 + 465649s r
U6= 43?& + 22278?41 T Ur4= 308:71- 93317r w4- 75931 +2610994 r
th = T7?6 + 20289942 T u - 75?32 T
u1s = 3ts,o9 N5= 35:36 + 1432594 7
us - 157936+ 16652:76r U16= 259:14- s04981r N6= 142961+ 2824967
,V7= 12985+ 52:316r
414
The mass. radius, arrd colors of Tritort are determined from Vol'ager 2 4;,1n.
E X P L A N A T O RSYU P P L E \ 4 E N T v
t' , / pHystcAL

oroduct
EPHEMEFiDES
O F T H E S U N ,I ' I O O N .P L A N E T SA. N D S A T E L L I T E S

of its g(urrrr'tricalbedo and the squareof its mean diarueter.If one ol these
415

Little is knoNn of Nereid except its magnitudc at l Att and 6o phase arrglc (\lo1- n. or a relation bctween tlte t$'o factors r::rnbe determined. the other
?actors kno..
is
rison and Cruikshank, 1974). The cliarrieteris based on an assurrtcdalbedo of g..1. bc dcternincd. At thcrmzrl infrared rvavelengths, the dependence of the
factot ctLt
Recent studies by Williams et al. (.1991)of Ncreid iudicate that it ma].h4r.e. albedo is cornplementarv to the dependence at visible waveleugths.
brightness on
non-svnchronous rotatioll. take two objects at the saric distanrt u'ith the same visible rnagni-
For example,
tude. 1f one is dark arrd large, and the other is small but highly reflectilg. thel in
infrared emitted radiation. the first object will be nuch brightcr than the second
7.56 The Satellite of Pluto
oneJbecause of both its larger size and its liigher tcrrrpcrature. Thc complernen-
Charon, Pluto's orrly knou'n satellite. is assumedto orbit in tltr: plane of PluLos taxy nature of brightrrcss with t bedo in the trvo spectral regions is thc basis lqr
equator and in synchronousrolation. The position of its nolth polc arrcl1>rir1. the radiometric methocl of determining diameters. The equations relating the ol>
mericlianarc given in Table 7.56.1. servableparametcrs (visible and infrared rnagnitudes) to the unknorvns (diameter
and geometricalbedo) reqrrirecomplicatedassurnptionsto bc made concerrringthe
Table7.56.1 kmperature distribution ovcr thc surfacc of thc a"steroidsand cannot bc solve<l
RotationParameters
for Pluto'sSatellite a,nalvtically.
The V magnitude for an obiect at phase angle I and R = I AU from thr: Sun is
do = 312?02 related to the radius. r. and geometric albedo, p" bv
6o= 9909
W = 56?77 5693623195
I
V = V . - 5 l o g r - 2 . 5 l o g p , ,- 2 . 5 l o g ! " ( i ) + l o g R , (7.6 1)
The mass, diameter, and nagnitude of Charon alc bascd otr obselvzttiorls o1
Charon-Pluto eclipses (Dunbar and'fedesco. 1986). where V. is the magnitude of the Sun at 1r\u and d,,(i) is the phase coefficient in
intensity units, normalized to unitv at i = 0'. The brightness of the sarne object at
10pm (N magnitude) is
7,6 SF THE ASTEROIDS
P H Y S I C A LE P H E I V I E R I D E O

N = Krv 5 log r 2. 5 log(fy) 2. 5 log er,'(l), ("7.6,2)

At preserrt. thc only physical data published in Thc Aslronontical Alxnan. ..: Iot tlrc
astcroids are H, the absolute visual rnagnitude at a,dista[ce of 1AU an(l 0" l)hase
anglc; G, thc slope parameter that dependson the albcdo: the visual magrlitrrdc'at where rKlyis a constant cleterrrinctl by the :rbsolutc calibration of the N rlagnitude
6ystem; IN is the infrared surface brightrrcss in thr: N band, obtained by arcralling
opposition; arrd the estimated diarnctr:Lsfor 139 asteroids.Ihese data are strpPli<'tl
the emission from each surface clcrrrcrrt; ancl e1,,(i)is llrt: effective crnissivity in thcr
by the Institute for Theoretical Aslronorny. St. Petersburg,Russia. AtlditiorraLttt-
'l N band at phase anglc i. Note that er,(0) cau bc greater than unity to allorv for
formation on the asteloicls can be fouucl in the Institutc lor heorctical Aslrottoltl.v
peaking of the infrarecl
p u l r l i , - "iro r r .A 4 i n n rP l on e t s . emission at snrall phast: auglesJbut at the sane tinre €N(i)
rnust be substantially
The visrral magnitudes of the asteroids arc based on thc lrttrk of Bolvcll, [lart ts' lcss than unity at large I to satisfv thc totai heal, balan<re.
If it is assumed that
and Lumme (19tt8). the planet is seen at zero phase and thc surla(e tcrnperal,trre
rs everylvhere in
-Before1970, the cliarneter measurements lvcre based orr visual rnicrotttt:tet lttl'z' ecluilibrium r.ith the absorbed soiar energl, the distribution of
' remperature over
sure[rcntsor occultation observations.Reliablc rneasuremcntsof asteloid r-adii(t:'g the surfacc $,ili be qiven bv
Dollfus, 1971;\Iillis el al.. 1987) existed for on)y fir''eastcroids:Ccres.Pallas' JrtrxJ
Vesta. anclDros. Sincethcn, iurprovcdcstimatesofthc diametels ltavc resulttrrihotrt
'I'hc T ( o ) = T o n r t 2 1 1-l8 ) / . 1 r / t c o s " o n , ( 7 . 63 )
devcloprrerrtof lhc ra(liolnetric and polarimetric mcl.hods. diamcterspublished
'l'lrc
rr't use the radiomctric obser'l-ationsof \'Iorrison (1977)-
A str()n.)ttLic(LIAl,nL.LrL(Lc
1ro ao t" the eqrilihrium blackbocly tcmpcraturc at 1AU (= 295'K). A is the
Thc racliometric method is sirnple in its physical principle, but corrtPlex Ihu*
sond albedo,
tliscuss quantitativr:ly. The visible brightncss of an asteroid is proportional to the e is t-]realerage emissivity, and a is the local solar zenith angle.
416 EXPLANATORYSUPPLEI\IENT z i P H Y S I C A LE P H E M E R I D EO
S F l H E S U N ,M O O N ,P L A N E T SA, N D S A T E L L I T E S 417
A first-order solution may be obtained if the infrared brightness is assumed 1q trhs Galilean satellites (N{orrison and Cruikshank, 1973; Flansen, 1973) har.e shown
be proportional to the absorbed energy: that all these objects have surface thermal conductivities lower than that of the
Moon. As a result of these lorv conductivities, the surfaces all respond very rapidly
tN =Tfi-2\l - B), ti .64) to changes in sunlight; so they remain close to equilibriurn even for a rapidly rotat-
ing asteroid such as Eros, with a period of lessthan 6 hours. So the basic assumption
where ?1 is the mean surface temperature for the body. When this substitution is of thermal equilibrium appears to be valid.
made and Equation 7.6 2 is subtracted from Equation 7.6 1. the result is

y - N = (% - Kry) 2.5log(pvI eNTt)- 2.5loglp"/(l qp,)1, (7.6-5)

7.7 REFERENCES
where 4 is the phase integral adjusted to correct for the difference between p, and its Atkinson, R.d'E. (1951). "The Computation of Topocentric Librationsi' Mon. Not. Rou.
average or bolometric equivalent. To solve for pu from the observables V and N, it is Astr. Soc. 111, 448 454.
necessary to knovr the photometric parameters V,, K1y,and I1 (or its mote gerelal Binzel, R.P. and Mulhollartd, J.D. (1984). "Photometry of Pluto During the 1983 Oppo-
equivalent 7e). The ratio of the phase functions in both the visible and infrart:cl sition: A New Determination of the Phase Coefficient" Astron. J.89, 1759 1761.

must be known also. These four quantities can be obtained from observations olt:r' Burns, J. (1986). "Some Background About Satellites" rn Satelli,tesJ.A. Burns and X{.S.
Matthews, eds. (University of Arizona Press, Tucson, AZ), p. 1 38.
a range of phase angles. Finall]'. the phase integral, 4, must be specified to obtairr
Campbell, J.K. and Synott, S.P. (1985). "Gravity Field ofthe Jovian Svsl,emfrom Pioneer
a solution. The values of all these parameters must cither be assumed or treated as and Voyager Tracking Data" Astron. J.90,364 372.
variables in a calibration of the method based upon objects of knorvn diameter. Carington, R.C. (i863). Obseruat'ionsoJ Spots on trle &rn (Williams and Nolgate, Lon-
Since the phase integral is not generally knorl'n, it is important to consider hrtrv don).
sensitive the derived values of r, p", and B are to changes in 4. Of particular sigrif- Danjon, A. (1949). "Photom6trie et colo m6t e des planbtes \Iercurc ct V6nus" Bull.
Astron.14,2e Serie,315 345.
icance is the behavior of the solutions for the low-albedo objects that predominate
Danjon, A. (1954). "Magnitude et albedo visuels de la planb te N{ercure" Bull. Astron.
in the asteroid belt. In Equation 7.6 5 as p" approaches zero, the final term :r1r 17, 2e Serie, 363.
proaches-2.5logp", so that for dark surfaces,the solution is virtually irrdepenclort Davies, M.E., Abalakin, V.K., Brahic, A., Bursa, M., Chavitz, B.H., Lieske, J.H.. Seidcl
of the value of 4. However. a similar exercise for B shows that, even at 1'er] 1orl mann, P.K., Sinclair, A.T.. and Tjuflin. Y.S. ( 1991). "Report of thc IAU/IAG/COSPAR
albedos, the final r'alue of the Bond albcdo remains inversely proportional to lhe Working Group on Cartographic Coordinates arrd R.otational Elenteuts of the Planets
assumed phase iniegral. and Satellites: 1991."
Davies, M.E., Hauge, T.A., Katayama, F.Y., and Roth, J.A. (1979). Cantrol Netuolks lar
In this simplified solution, the expression for the radius is
the Cahleo,nSatellites: Noxernber'7979Thc Rand Corporation, R-2532-JPL/\ASA.
Desch, M.D. and Kaiser, X'II.L.(1981). "Voyager Measurement of the Rotation Period of
5 l o g ( r / R ) = K N - N - 2 . 5 l o g ( e 7 1 T r )2 . 5 1 o g ( 1- B ) (7.h-61 Saturn's JVlagneticFicld" Geophgs.Res. Lett.8,253 256.
Dollfus, A. (1971). "Diameter lleasurernents of Asteroicls" in Physit:al Studit:s oJ Mhnr
This expressiondoes not contain any of the parametersof the photonetlic slstellr Planets "l. (lerhels, ed. NASA SP-267 (NASA). pp. 25 32.
Dunbar, R.S. and Tedesco.E.F. (1986). "\todeling Pluto-Charon \'Iutual Eclipse Events:
including the observed V magnitude. Hou'ever, the derivecl radius is sensitive tt)
L First-Orrler N4odels",4stron. J.92. 12Ol 1,209.
the infrared parametersr r(1,,,e1,,,and Ir, in addition to the obscrved /V rlLagrrilrr'1" Duncombe, R..L., Klepczynski. W.,1.. aud Seirielrrrarrrr, P.K. (1971). The \lasses of the
Becausethe albedo enters only via the (l -B) tcrm, r is only wcakly dcl,en,lell "LL Planets, Sateilites, and Astcr.oids" Furul,. Cos. Ph\. 1,'119 16it.
the albedo, and for low albedos the final term is negligible. For this reason, a tnuclL Eckhardt, D.H. (1981). "Theory of the Liblation of the NIooD" '/)rc lVoon o.n,ltlrc Planets
better estifratc of the diameter can be made from a measuremerrt of N alone iharl 25. 3 49.
from one of V alone, even when the infrared magnitude is less well deternincd thatt Eckhardt, D.H. (1982). "Planetart' and Earth Figure Pcrturbations in the Libr:ition of the
the visual rnagnitude. Furthcrmore, cven a r'ery rough value of V. rvhert cornbirLecl Moon" in /{igA-Prr: cision Earth Ratation anrl Earth Moon D gnanLics O. Calamc ed-
(Reidel, Dordrecht. Holland). pp. 193 198.
with a rneasurement of N. is suficient to define a leasonably precise diamcter'. Gatley, 1., Kieffer, H.. tr,'liner, E., and Neugelrauer, G. (197,1). "Infrared Obsen'ations of
One of the assumptions of the radiometric rnodels is that the local surfacc t,'rrL Phobos from X{arincr g" Astropllgs. J. 190, .197 503.
^
perature is approximatcly in instantaneous equilibrium with the incident surligltl' ureenberg, R. (1984). "Orbital Resonancesamong Saturn's Satcilites" irr Sol rn T. GeI-
Infrared observations of Phobos (Gatley ef at., 1974), Eros (Morrison, 1976). alnl hels and tLS. N,Iatthcws,eds. (Urriversityof Arizona Press, Trrcsor, AZ), pp. 593 608.
418 EXPLANATORY
SUPPLE[.4ENT EPHEMERIDES
./ PHYSICAL OFTHESUN,MOON,PLANETS,
ANDSATELLITES 419
Hansen, O.L. (1973). "Ten Eclipse Observations ofIo, Europa, and Ganymede" lcorys 16. 8.A., Sonderblom,L.A., Beebe,R., Bliss, D., Boyce,J.M., Brahic, A., Briggs,
237 246. G.A., Brown, R.H., Collins,S.A., Cook, A.F., II, Croft, S.K:, Cuzzi,J.N., Daniel-
Harris, D.L. (1961). "Photometry and Colorimetry of Planets and Satellites" in Planets son,G.E., Davies,M.E., Dowling,T.E., Godfry,D., Hansen,C.J., Harris,C., Hunt.
and, Satellites G.P. Kuiper and B.A. Middlehurst, eds. (University of Chicago P1qsr. G.E., Ingersoll,A.P., Johnson,T.V., Krauss,R.J.,MasurskyH., Morrison,D., Owen;
Chicago). pp, 272-342. T., Plescia,J., Pollack,J.8., Porco,C.P., Rages,K., Sagan,C., Shoemaker, E.M..
Sromovsky, L.A., Stoker,C., Strom, R.G., Suomi,V.E., Synott, S.P.,Terrile,R.J.,
Kuiper, G.P. (1954). "Report of the Commission for Physical Observrations of the Planets
and Satellites" Trans. IAU 9, 250. Thomas, P., Thompson, W.R., and Veverka, J. (1986). "Voyager 2 in the Uranian
Systemi Imaging ScienceResults" Science233, 43-64.
Lumme, K. and Bowell, E. (1981). "Ra.diative Transfer of Atmosphereless Bodies, Il:
Interpretation of Phase Curves" Astron. J.86, 1705-1721. Scitb, B.A., Sonderblom. L.A., Beebe, R., Boyce, J., Briggs, G., Carr, M., Collins, S.A.,
Cook, A.F., II, Danielson, G.E., Davies, M.E., Hunt, G.E., Ingersoll, E., Johnson,
Lumme, K., Karttunen, H., Bowell, E., and Poutanen, M. (1986). "lnversion of Aster- T.V., Masursky, H., McCauley, J., Morrison, D., Owen, T., Sagan, C., Shoemaker.
oid Light Curves Using Spherical Harmonics" in Asteroid,s, Comets, Meteors II C.-1.
E.M., Strom, R., Suomi, V.E., and Veverka, J. (1979). "The Calilean Satellites and
Langerkvist, B.A. Lindbla.d, H. Lundstedt, and H. Rickman, eds. (Uppsala University, Jupiter: Voyager ? Imaging Science Results" Science 206, 927 950.
Uppsala, Sweden) pp. 55 59.
Thomas, P., Veverka, J., Morrison, D., Davies, M,, and Johnson, T.V. (1983). "Satuln's
Lumme, K., Karttunen, H., Piironen, J., and Bowell, E. (1986). "Simultaneous Solutions Small Satellites: Voyager Imaging Results" J. Ceophgs. Res. E8, 8743 8754.
for Pole, Shape, and Albedo Variegation of an Asteroid" BuLl. Am. Astron. Soc. lE, Tholen, D.J. and Zellner, B. (1983). "Multi-color Photometry of the Outer Jovian Satel-
801. lites" Iccrus 53, 341 347.
Marcialis, R.L. (1988). "A Two Spot Albedo Model for the Surface of Pluto" ,4stron. J Tholen, D.J., Buie, M.W., and Swift, C.E. (1987). "Circumstances lor Pluto-Charon Mu-
95,94L 947. tual Events in 1988" -4slron. J. 94. 1681 1685.
Millis, R.L., Wasserman,L.H., Franz, O.G., Nye, R.A., Oliver, R.C., Kreidl, T.J., Jones, Tyler, G.L., Eshleman, V.R., Anderson, J.D., Levy, G.S., Lindal, G.F., Wood, G.E., and
S.E., Hubbard, W., Lebofsky, L., Goff, R., Marcialis, R., Sykes, M., Frecker, J., Croft, T.A. (1981). .Radio Science Investigations of the Saturnian System: Voyager
Hunten, D., Zellner, B., Reitsema, H., Schneider, G., Dunharn, E., Klavetter, J., 1 Preliminary Results" Science 212,20I 206.
Meech, K., Oswalt, T., Rafeit, J., Strother,8., Smith, J., Povenmire, H., Jones. B., T$er, G.L., Eshleman, V.R., Anderson, J.D., Levy G.S., Lindal, G.F., Wood, G.E., and
Kornbluh, 8., Reed, L., Izor, K., A'Hearn, M.F., Schnurr, R., Osborn, W., Parker, Cloft, T.A. (1982). "Radio Science with Voyager 2 at Saturn: Atmosphere and lono-
D., Douglas, W.T., Beish, J.D., Klemola, A.R., Rios, M., Sanchez,A., Piironen, J., sphere and Masses of Mimas, Tethys, and lapelvs" Science 215, 553 558.
Mooney, M., Ireland, R.S., and Leibow, D. (1987). "The Size, Shape, Density. and Viellet, C. (1983). "De I'Observation et du Mouvement des Satellites d'Uranus" Ph.D.
Albedo of Ceres from Its Occultation of BD+8"417" Icarus 72.5O7 518. Thesis. (Universit6 de Paris, Paris).
Morrison, D. (1976). "The Diameter and Thermal Inertia of 433 Eros" Icaru,s 28, 125 132 Williams, LP., Jones, D.H.P, and Taylor, D.P. (1991). "The Rotation of Nereid" Mon.
Not. R. Astr. Soc. 1p-2p.
Morrison, D. (1977). "Asteroid Sizes and Albedos" Icarus 31, 185 220.
Wisdom, J.. Pealp. S.J.. and Mignard. F. (1984). "The Chaotic Rotation of Hyperion"
Morrison, D. (1982). "Introduction to the Satellites of Jupiter" in Satellites oJ JupiterD Ic&rus 58, 137 152.
Morrison, ed. (University ol Arizona Press, Tucson, AZ), pp. 3-43.
Morrison, D. and Cruikshank, D.P. (1973). "Thermal Properties ofthe Galilean Satellites"
Icarus lE,224-236.
Morrison, D. and Cruikshank, D.P. ( 1974). "Physical Properties of the Natural Satellites"
Spac. Sci. Rea. 15, 647 739.
Morrison, D., Owen, T., and Sonderblom, L.A. (1986). "The Satellitesof Saturn" in Satef
liles J.A. Burns and M.S. Matthews, eds. (University of Arizona Press, Tucson, AZ),
pp. 764 801.
Pollack, J.B., Veverka, J., Noland, M., Sagan, C., Duxbury, T.C., Acton, C.H., Jr', Born'
G.H., Hartmann, W.K., and Smith, B.A. (19?3). "Ma ner 9 Television Obserrations
of Phobos and Deimos, 2" J. Geophys. Res. 7a, 4313-4326.
Sjogren, W.L. (1983). "Planetary Geodesy" Betr. Geophgs. Space Phgs.21,528-537
Smith,8.A., Sonderblom, L., Batson, R., Bridges, P., Inge, J., Masursky, H., Shoemaker'
E., Beebe,R., Boyce, J., Briggs, G., Bunker, A., Collins, S.A., Hansen, C.V., Johnson'
T.V., Mitchell, J.L., Terrile, R.J., Cook, A.F., II, Cuzzi, J., Pollack, J.B., Danietson'
G.E., Morrison, D., Owen, T., Sagan, C., Ververka, J., Strom, R., and Suomi, v tr
(1982). "A New Look at the Saturn Systern: The Voyager 2 Images" Science 215,
504 537.
 CHAPTER

Eclipses
of theSunandMoon
by AlanD. FialaandJohnA. Bangert

8.1 INTRODUCTION

All bodies orbiting in the solar system cast a shadow away from the Sun. In the most
general terms, an eclipse occurs when the shadow of one body falls upon another
a,ndtemporarily blocks out a portion of the solar illumination. An eclipse is distinct
from an occultation, which is a block in the line of sight. Before this distinction was
understood, however, a long history of observation from Earth created difierent
names for the phenomena, depending on what objects were involved. The shadow
of the Moon falling upon the Earth was called a solar eclipse; the shadow of the
Earth falling upon the Moon vras called,a lunar ecltpse;and the shadow of a planet
falling upon the Earth was called a transit (across the Sun). By strict application
of these definitions, a solar eclipse is really an occultation.
The basic methods by which eclipsesof the Sun and Moon and transits of Venus
and Mercury a.repredicted and described were introduced by Bessel and extensively
developed by Chauvenet (1891). The discussion here will be limited to describing
the calculation of data that are necessarv to observe anv of these phenomena from
the Earth.
Bessel's method of calculating eclipse predictions is based on the following con-
cept: A bodv casts a shadow, the a-ris of which is a line extended from ihe center of
the Sun through the center of the body. Observations, by necessity, are constrained
to the topocentric coordinate svstem, However, the geometry and kinematics are
-,easier
to describe in a svstem oriented to the cone of the shadow that is, one in
which the z-axis is either parallel to, or coincident with, the shadow axis, and the
llplane passesthrough the center of the body on which the shadow falls. This plane
is also known as the Besselian plane. A description of the progress of the eclipse is
calculated in this coordinate system, and then is transformed to a topocentric sys-
tenr as necessary.Elements describethe geometry of the configuration as a function
t'ime circumstances describe distinct, observable events by time and place.
422 EXPLANAToRY
SUPPLEiTENT / ECUPSESOF THESUNAND MOON
423
Diagrams in treatises on eclipses by necessity are exaggerated to illustralo of the total or annular phase, and the altitude of the Sun on the central
points. Hence, thc diagrams tend to disguise the fact that, in reality, the boclics are tabulated at intervals of ten minutes or lcss throughout the eclipse. For
involved are very small, far apart, and moving relatively slowly. Thus, the shadows eclipses, the circumstances and their times are the same for all parts of the
cast are long and slender, there axe many small angles, and the relative short-tern particular phase is visible from any point at which the Moon is above
barth; any
motions are very nearly linear. These qualities were used to advantage by peolrkr the horizon'
equipped only with tables and a calculator, as it was necessaryand adequate to irs. For years in which a transit of Mercury or Venus occurs, general circumstances
approximate local circumstances may easily be
approximations and intermediate trigonometric constructs. In the era of high-spcr.cl snd a map are given, by which
programmable computers, the use of approximations and simplifying constructs is €stimated. Formulas are also provided for the computation of more accurate local
reduced but not eliminated. Also, it is helpful to newer generations of astrononers circumstances.
to discuss spherical trigonometry by incorporating more vcctor and matrix notatiol The information that appears in The Astronomical Almanac is prepared a min-
provisional value of Af to the nearest q'hole
instead of direction cosines. lglum of three years before the event. A
provisional Universal Time (UT)
second is assigned, so that the time argument is
or, more precisely, provisional UTI (see Chapter 10). More details are given in
8.11 EcliDse fromthe Nautical
DataAvailable AlmanacOffice
Sections8.12 and 8.363.
8.111 International Agre€m€nts Ever since astronomical ephemerides and alrrLn- The predictions given in ?be Astronomical AImanQc take no accorrnt of the
nacs have been published, eclipse information has been included. This was a nal,ural efiects of refraction, although refraction may be significant for the reduction of
action because the extensive calculations required are based on predicted ephcnr observations intended to give precise positions. The Besselizrnclements, however,
erides. In modern times, the situation is much eased, but the Nautical Almarac are rigorously independent of refraction.
Ofrce continues to provide information on every eclipse of cach year. The infor
mation conforms to IAU standards and is not changed nithout adequate notice &113 Data Provided in Other Publications Material from the section of ?he '4s-
to the International Astronomical Union. This may cause some misunderstanclings hpnomical Alman,Jc on solar and lunal eclipses and transits is also published, in
(e.g., in the ralue of /<,describedin Section 8.12), or occasionaldifficulties l hen a advaoce of publication of The Astronornical Almanac itself. in the small booklet
convention changes (for instance, longitude is now considered positive to the eirst Astronomical Phenomena' issued annuall5' by the Nautical Almanac OfEce' U S'
of Greenwich). Some changes that improve production without changing the inlor Naral Observatory.
mation are adopted by decisions within the Nautical Almanac Office, such as tlte In 1949,the U.S. Naval Observatory began a publication seriescalled the U S'
change in the solar eclipse nap curves to an older convention, which made mcch;rtt- Naral Observatorv Circulars. The series tralsmits astronomical data resulting from
ical computer-controlled plotting feasible. Other special publications for individunl continuing programs but not available in any othcr publications. Two of the con-
eclipses are designed to assist the scientific observer as much as possible. Sinctr tinuing programs that have contributed many issues to this series are solar eclipse
' predictions
1949, at the request of the appropriate Commissions of the International Astro- and sunspot counts. There are two kinds of solar eclipse circulars First'
nomical Union, the principal circumstances of all solar eclipses have been routitreL.l there are circulars issued well in advance that give just the priucipal circumstances
calculated and madc available several years in advance of the publication of the an- of all solar eclipsesin a multiple-year period. These are listcd in Table 8113'1'
nual almanac. This working agreement continues and is carried out b1: the speclal Second, the Observatory issues circulars for specific solar eclipses; these circulars
publications describediu Section 8.113. contaiD extensive information for observers to choose a sitc artd calculate accu-
rate local circumstances. Belore the circulars were begun in l949, this information
8.112 Data Pmvi ded in The Astronomical Almanac Elernents and general circunr- was issued in Supplements to the annual American Ephemeris Until 1991, it was
' the policy
stances are given in ?[e Astronomical Almanac for a]l solar and lunar eclipse'' of th; U.S. Naval observatory to issue a circrrlar for each total solar
including the penumbral lunar eclipses that occur during thc year. For the solar eclipse,with a very few exceptions, and for selected annular solar eclipsesthat have
eclipses, maps are given from which approximate local circumstances may be ob- some observationally interesting features. Typically these circulars contained gen-
tained for any particular place, and the Besselian elements are tabulated at interr'ats eral information and circumstances, Besselian elements, a discussion of anticipated
of ten minutes for the calculation of accurate local circumstances for any point on or' lneteorological conditions along the eclipse path, local circumstances for selected
above the surface of the Earth. For total or annular solar eclipses, the latitudes anq locations. the lunar limb profile and limb corrections, path of the central phase tab-
longitudes of points on the central line and on the northern and southern limits, the by time and longitude at sea level and at flying altitudes, and detail maps of
 424 E X P L A N A T O RSYU P P L E [ 4 E N T O F T H ES U N A N Ol t l O O N
I / ECLIPSES 425
Table8.113.1 flith its center of mass; to allo$'for this difference, empirical corrections ale made 1o
U.S,N.O.SolarEclipseCirculars
the mean equatorial coordinates of the Moort. referred fo the center of mass of thc
Cir. No. Date Ti e Bu11l. Unpublished investigationshave shown that thc corrections depend upol
the lunar librations at thc tirrrc of an eclipse. The range of the corrcctions is less
1 1 9 4 9J u l y1 8 Tracksof total solar eciipsesin 1952, 1953, 1954 than 0'i5, and for the calculations in The Astronomi,cal Almanac, otrly thc minirnum
2 1949September19 Tracksof totalsolareclipsesin 1955,1956,1957 = +01'50and AB = -01'25 for the LE200 luuar
co ections are applicd. These are A)
1950N4ay4 Tracksof total solar eclipsesin 1958, 1959, 1960
40 1952December8 Annularand padialsolar eclipsesin 1955,1956,and 19s7 epbemeris. The correctiorrs are made drrring the apparent place computation by
Non-centraltotal eclipsesol 1957 rotating the mean equatorial geocerrttic coordinates of the N4oou into the ecl\rtic
1954August20 Annulareclipsesin 1958,1959 system, applying the corrections in the ecliptic syslern, and rotating back into the
59 1955June 28 Solareclipses,1960-1963
85 1958April
'16
Solareclipses,1963-1967
equatorial systcm. The corrections are applied as constants for each eclipse.
B9 1 9 6 0J u n e1 5 Solareclipses,1968-1970 The semidiameters of lhe Sun and the X,{oonused in the calculation of eclipscs
101 1964July 29 Solareclipses,1971-1975 do not include irradiation. Thc adopted semidiameterof the Sun (s,) is cornputerl
113 1966December15 Solareclipses,1976 1980 directly from the ephemeris ol the Sun, thereby rrsing the ephcmeris rcnstants.
142 1973April20 Solareclipses,1981-1990
170 1986November7 Solareclipses,1991-2000 However, the apparent semidiameter of thc Nloon is calcrriated bv putting its sine
equal to k sinrr., q'bele rn is the horizontal parallax and l, thr: ratio of the \'Ioon's
Note: No. 170 was the first lo conlain eclipsemaps. The others listedonly numericaldala. radius to the equatorial radius of the Earth, is an adoptetl constant. The lAIl
adopted a new value of k (k = 0.2725076) in August 1982. Br:forc that time. t$'o
diferent values ol i wcre rtsed in computing a solar cclipsc. Thc valuc takcn fbr
the lunar ephemeris r.as rtsed for the most part, but in pra.ctict: it \ras too large for
the portion of the path or.erland. Thc U.S. \aral Observatory planned to rt:lt:ase total phase in total solar cclipsesbecauseof the eflectsof thc rorrgh lunar limb. As
each circular 12 to 18 lnonths irr adrance of a particr ar eclipseand to ircorl)ola1(l a,n approximate corrcction f<rl this efiect, a smaller valuc of k rvas adopted solelv
the best r''alueof AZ aiailabh at the time of preparation. However, the lra,l tirlc for calculating duration orr the central line of total solar- r:clipses. This smaller
and contents rveresubjcct to regular revielv rvith respect to meetilg t]re report.(l value caused numerical incoltsistencies as t'ell as nrisundelstandings on lhe Part
needsof the scientificcomDnnitl-. In 1992.the eclipsecirculars rvereclisco]rtinutrl. of users of thc infbrmation. Therefore, n.hen the rrcrv vahte was adoptcd in 1982
to conform to thc valur: used in occultation prcdictions. it rvas agrcc<l implicit ly
that limb effe<{s arc no longcr accounted for. but are averagecl,and if an obsclver
8.12 Corrections
to the Ephemerides
considersthern to be irupoltant. then corrcctions flust be calculated and appli<:cl
The basic quantities fbr calculationsof solar and lrrnar eclipsesare: the al)palelt separately.It is possible for these effects to advance or retard predit:led sc,'rrrl
right ascension(a.). declination (6.). and clistance(R.) of the Sun, and the apparcrll or third-contact tines ou the central line by as much as t$-o seconds apicr:e. Tlte
right ascension(n.), declination (6.), and distance(R.) of thr: \'loon, for everr.hour same limb corrections can be applied in thc case of annulzrt solal cclipscs, or e\-en
of TDT during 1,heeclipse;and the ephernr:rissidorealtimc at 0h TDT for the c1a1ol lunar eclipses,but the efiects lr,ill not be obscrvable. It is onif in the case of a tot:rl
the oclipse and for the following day. Ihc apparent places of the Surr and tlte Nloorr solar eclipse that the presence or abscncc of auy lighi frorr thc solar tlisk is ct*ily
are computed rigorously using t hc same ephemerides for the Sun arrd the NIooIr as detectedand is hence critical to delicato obscrvations.
are used for The Astronom.ical Alrn.anac,and using the reduction methods cles<rribecl The calculation of the occurrertccof etlipses does llot d(rpend upon the tinlc
elsewherein this book (seeChaptcr 3). At the time of this writing. thr: ephcner-i{lcs scalefor the portions restricted to thc spa.tialrelationshipsof the Eartlt, NIoon, arrcl
are based upon the DE200/LE200 produced by the Jet Propulsion La,boral,or'1'. The Sun. However, lor calculation of phcrronrenaas they arc relatql to the sttrfar:eof tltr:
ephemeris sidereal time at 0h TDT, rvhich is the local apparent sidercal tirne ott tlie Earth for time and place of visibility, it is necessarlto rtseUriversal Time (UT1) in
ephemeris meridian, is the same turncrically as the (Greenwich) apparent sidercaL the best approxirrratiorr possible. This approxima,tion is achievecl by adjusting tht:
time at 0h UT. as tabulated irt The Astr.ort.om,icLrl ephemeridesat tlre start of the calculatiols. The predictions it the A stronomicol
Alm.ona..
Gravitational epherncridcs refer to the positions of the centers of nrass of the Almonac were given in ephencris time (ET) until 1981,when AI had becornelatge
bodies concerlcd. Eclipses, hou'ever. are governed by the positions ol the centcr-sol enough that it u'as cleemed more useful to publish thc predictious in provisiorral
figures of the Sun and the N,Ioon.The centcr of figure of the N,Ioondoes not coirlci(le UniversalTime (seeSection !t.363for dctails).
 426 ExpLANAToRy
suppLEMENr OF THESUNAND MOON 427
8.2 THEOCCURRENCE
OF LUNARANDSOLARECLIPSES situation is illustrated in Figure 8.22.1.This is a generalrepresentation;Bm
be positive (above the ecliptic) or negative (below the ecliptic)' Correspond-
8.21 Overview ,1V may be either the ascending or descending node of the lunar orbit. For a
p eclipse, S represents the antisolar point at opposition. For a solar eclipse, S
The purpose of this section is to develop conditions that can be used to determine the center of the Sun at conjunction, For either case' M represents the
whether an eclipse will occur at a pa,rticula,r conjunction or opposition of the Sun of th" Moon. Take the absolute value of B., which is the latitude of the
"l
and Moon, The approach is, first, to develop a general expression for the geocentric at opposition for a luna,r eclipse, or at conjunction for a solar eclipse. The
least angular separation of the Moon and either the Sun or the antisola,rpoint, when values, M' and 9, represent positions at a time when angle SMS/ = 7' and
one of the latter is in the vicinity of one of the nodes of the Moon's orbit. Second. separation of Mt and S/ is o. .I is the inclination of the Moon's orbit to
expressionsfor the apparent sizes of the shadows are found. Finally, discrimination
tests are established. Speciflcally, to develop the conditions for a lunar eclipse, the wish now to derive an expression for o in terms of klown quantities. Define
appaxent sizes of the umbral and pem.rmbral shadows of the Ea,rth are compared .1,a ratio oI t he longitudinal motions:
to the geocentric least angular separation of the Moon and the a.ntisolar point.
To develop the conditions for a solar eclipse, expressions for the geocentric least SP Moon's motion in longitude
l=::- = (8.22-1)
angular separation of the Sun and the Moon are derived from the geometry of ss/ Sun's motion in longitude
each particula"r type of solar eclipse and are compared with the geocentric apparent
angular semidiameters. the geometry:
o2 = (s'p)2 + (M'p)2, (8.22--2)

8.22 GeocentricLeastAngularSeparation
Consider the Sun and the Moon at a time of conjunction or opposition in longitude. S,p = Sp - SS = (.\ - l)Bn tan7. (8.22-3)
In order for an eclipse to occur, the Sun (or the antisolar point) and the Moon
must each be near the same one of the nodes of the Moon's orbit. Thus, the ecliptic
MtP = NP tirrI = (SN- SP)tanl = B-(1 - .\ tan? tan0. (8.224)
latitude of the Moon (B,o) will be small, and plane geometry may be used.

o2 = l^11s - 1)2tan2r+(1- )tanltanr)21. (8.22-s)

'order
to find the value of 'y at which o2 is a minimum, take the derivative of
8.22 5 with respect to 7 and set it equal to zero. After some algebra, the

ltanl (8.224)
*' '- -
() 112+ ^2 t5r2I'

Ecliptic Equation 8.22 S into Equation 8.22 5 to obtain:

Moon'sorbit
o = g^0 - Dl() - I)2 + )2 tan24-rt2. (8'22-:7)
Figu€ E.22.r
Geometdcconstructfor determiningwhetheranykind of eclipsewill occur whenthe Moon is neat
oneof the nodesof its orbit on the ecliptic. In this figure,ly' is the descendingnodeasseenfrom equation may be simplified by the following change of va"riables'Let
outsidethe orbit, looking towardEarth.Let S representthe Sunor its aniisolarpoint, andM theMoon'
at conjunctionor opposition,respectively,whenthe apparentseparationis the lunar latitude 8.. At (8.22-8)
somelater time whenthe two points havemovedto S', M', then the instantaneous separationis o.
tanl = tanr.
;f
 429
EXPLANATOBY SUPPLEI4EN.T OFTHESUNANDMOON
ECLIPSES
// is the angle at which the moving Moon would 429
appear to intersect the c'
from xhe point of view of an observer on the moving point ""' eel;^.,
rrprrc T
s. Th.n

o = B^cosl'. (8.22_s)
Equations 8.22_7 and g.22-g give expressionsfor
the geocentric least angular sepa_
ration between the Moon and the Sun at conjunction,
o-rthe geocentric lea"t angulal
separation of the Moon and the antisolar point at
oppositioi of the Sun and N,Ioon

8.23 Occurrence
of LunarEclipses FlSur€ 6.23f .2
Ceometricparameters of theEanh'sumbralshadow,relatingits angularsizeto observed
In thisdiagram,ZGME = 1t-,ZETG = r,, ZSEi = s", unaZWU = t,.
quantities.
Now that we have an expressioa for least angular
separation of the Moon and
the antisolar point near opposition, let us examine
to* to.,." lt to test for the
occurrence of a lunar eclipse. From the geometry
of the shadow cone, we shall find
a relationship between the size of the cone and f,'. Angle EVM is an intermediate angle, so that angle EVM = angle ETG + ans,le
other well_known quantities. JEL We want "fr = angle LEM = angle GME + angle EVM. 'Ih*,

8.231 Apparrnt Semidiameters ofthe Shadow

Cones The penurnbral shadow cole
of the Earth is defined by the interior tangents Jt =zrm+7rs+Js. ( 8 . 2 3 1 -)l
between t't" su" ona the Earth.
Figure 8.231.1 illustrates the geometry. In the
diagram, S is the center of the Sun, E The cross-section of the shadorv is not a true circle, because the Earih is not
is-the center of the Earth, and M is a point
on th" penumbral cone at the distanr:e a true sphere. To compensate, a mean radius for the Earth is substituted, which
of the Moon. The Iine LM is the semidiameter
of the cone at the distance of the is equivalent to substituting for n,o a parallax ?rr reduced to latitude 45o, so that
Moon, The known quantities are angle
GME, whichis the lunar horizontal paral_ zq = 0,99834Or^. Moreover, observation has shown that the atmosphere of the
lax (zrrn)based on the mean radius
of the Earth; angie Sff, *f,i"f, is the apparent Earth increases the apparent semidiameter of the shadow by approximately one_
semidiameter of the Sun (s, and angle
); ETG , rvt i*r L tle parrita* of the Sun (?r. . ffiieth. Hence.
we wish to find' I)4 as the angurar quantity )
rEM seen from the Earth. desienated
ft=1.02(tn+n,+s.). (8.231-2)

The umbral shadow cone of the Earth is defined by the exterior tangents be-
tween the Sun and the Earth. The vertex falls beyond the
Earth and beyond the
Moon's distance as well. Figure 8.231.2 illustrates the geometry.
M --' In this diagram, S
and E represent the centers of the Sun and the Earth,
respectively, and M is a point
ou the umbral cone at the distance of the
Moon. All anEles are as described in the
previous section, except that angle
VEM, the appa,rent semidiameter of the umbral
6ne (/z), replaces angle
/1. From triangle VEM, f2 = angle GME * angleEVM, and.
rrom triangle TEV, angle EW =
angle EVM = angle SET - angle EIG. Thus,

fz=LO2(q+lrs-rs) (8.23
r-3)
&fter compensation for the
efiect of the Earth,s atmosphere on the semidiameter of
the shadow.
Figure 8,231.1
Geometricparameters of rheEarth.spenumbral
quantiies.ln thisdiagram.ZGME = rn, shadow.relatingits angularsizelo observed t232 Condltions for a Lunar Eclipse There a.reseveral types of lunar eclipse, de-
ZETC = r".2567 ='r". u;;-2LVt;:
Jr, upon what portion of the Earth's shadow the Moon pa,ssesthrough. We
 43O EXPLANAToRY
SUPPLEI\4ENT OFTHESUNANDMoON 431
I / ECLIPSES
will consider the colditions for each. All conditious shorrld be tested to flncl tLo this tilnc. / = /2. A partial eclipsc
casefor a partial (utubr:rl) lttnar cclipse,except
maximum possibler:lipse. partial lunar eclipseis
will occur if o is lessth.ut J2 + sm.Thus. the condition for a

8.2321 Conditions for a penumbral lunar eclipse A pcuumbral lunar eclipseocr,1115 o < I . O 2 ( n 1 +r " . t r )+ s m . ( 8 . 2 3 31 )
when the Nloon orrtcrspartially or completely into thc perrunrbralcone,but uot th(,
umbral cone. The riglrt side of Figure 8.2321.1illustratcs the lirniting case for ;r occurs rvhen t[e
E.2331 Conditions for a total lunar eclipse A total lunar cclipse
Moon passes through thc penumbral cone and conpletcly into and through thc
umbral cone. The left si(lc of Figure 8.2321.1illrrstratcsthc limiting casefor a total
(umbral) lunar cclipse. uherc f = f2 h this diagranr, S rcprcsentsthe zrntisolar
Panial
point and M reprcsentsthe center of the \Ioon. Let SM be tllc gcoccntric least
Total
(f=f) i angular separatioD(o) given 11' Equation 8.22 9; MT is i.. A total eclipse *'ill
(f=f) Penumb.al I tm. Thus, the condition for a total lunar eclipseis
occur if o is lessthan /2
ff=fn
o < 1.02(.trt+ n. - ss) .sm. (8.2331-r)

of SolarEclipses
8.24 Occurrence

The geometr)' to test for the oc<rrrrence of a solar eclipsc is sitrrilar to that for a
lunar eclipse.In this sittlation. \rc are examining tlte shatlou of the \Iool instcad of
that of the Earth. The shadon is. of course.smaller and short(, . the lertex being at
about the same distatttr: from thc \Ioon as the Earth. \Ioreortr'. thc Ealth is much
Figure 8.2321.1
Composite diagram showing limiting conditions for ditTerenttypes of lunar eclipses,nor ro scale. larger, so that it ciul lle\'or be colnpletell' withiu the \Ioott s shado* In searching
Limiting conditions that separatetypes of eclipsesoccur when the separationof the centerof thc for the occurrenccoI a solar eclipse.$€ use the cortel of t lx' !]arth as thc referencc
Moon and the shadowaxis is equal to the sum or difference of the apparentradius olthe Moon and point of observatiort. 'fltr: exprcssion for least angular scpitt:ltion rrf thr: centers ot
the radius of the cross-sectionof one of the shadow conesat the distunceof the Moon. Let the small
the Sun and Moon st,ill zrpplies.
circle representthe Moon, with radiusr., the shadedcircle representeither the umbra, with raciiusl.],
or the penumbra,with radius Jr, and .' representthe distancebetweencenters.t and M or M'. On the
left side,ifo < J -.r., the Moon is rornllr,rmmersed in lhe shadow.On the right side,if o >.f + f,..
the moon is just out of the shadou If the Moon lies betweenthesetwo Iimits, iiis
8.241 Conditions for a Solar Eclipse There are also sc\rcral t.ypesof solar eclipsc'
2arridllr eclipsedin depending upon $'h.rt portiot: of the shadou' reachos tllo Earth's surface \\ie rvill
the shadow.The type ofeclipse is named for its maximum phase-rotat (umbral), partial (umbrai).
total penumbral,or partial penumbral. examine the conditiorrs for c{ch. It is not necessarvto (orrect f"r' atmusplrerir
augmentationof thc \loolt s shados': the \Ioon has l)o atllrosl)here lt is also tLot
penumbral lunar er:lipsc, u'hcrc / = /1. In this diagrzrlr. S r(T)rescnts the atrtisolar necessar!'to conect thc ptuallax for shape: this is tlolte lrv lintb 'otrectiotts if
point and M' represcnts the (.crrter of the Nloon. Let SMt l>c. thc gcocentric k':r:,I needed(Heralcl.1983).
angular separation (o) giverr by Equation 8.22 9; MtTtis sn,. A penurnbral eclipst'
will occur if o is less than /1 +s.. Thus, the condition for 4 lxrnurnbral lunar ecliPsc
8.242 Conditions for a partial solar eclips€ A pa|tial solar cclil)stro(curs \\'lren
ls
the Earth passcsthrouglt the penurnlrra ol the N{oon's sha<krlr',Lltt rr"t the rtnbra'
o < 1 . 0 2 ( n 1 +7 r .+ r . ) + s , n . (8.232l-l) (Also, partial pDase is sccn by an obscrver in the 1>ctltuttl>rtrduring any fype of
solar eclipse.) Figurc 8.242.1 illustrates thc lirniting cas<rlirr a partial solar eclipse
8.233 Conditions for a partial lunar eclipse A partial luuar eclipse occurs wirelr In this figure. S rrprcsetrts the center of thc Sun, M thc tlenter of the Moon'
the Moon passesthrough the pcnumbral cone and partly, but rrot completell', irrto and E the ceDter of the Earth. Point O, on the Earth, lies on the edge of the
and through the umbra. The right sideofFigure 8.2321.1also illustratesthe limitirrg penumbral cone, dcpit tcd by line segment OI' which also sctves a's the horizon at
 432 EXPLANAToRY
SUPPLEMENT OFTHE SUNANDMOON 433
Equations8.242-7 and 8.242-3 ir:to Equation 8.242-2,neglectingthe
of the semidiameters, and setting cossr equal to unity yields

z.=90'-sr-urn,,
z"=90"+s"-n'. (8.2424)

angle MES equal to the geocentric least angular sepa'ration (o) yields

o = Zs' Zm = Js * Jm *?rm - ?rs' (8.242-s)

eclipse will occur when the apparent least augula,r separation of the centers
Sun and the Moon is less than the sum of their apparent sernidiameters'
Flgur€ 6.22.1
Known geometricparametersin the limiting configurationfor a partial solareclipse,whenthe the condition for a partial solar eclipse as seen from the geocenter is
Moon is extemallytangent,As seenby an observerat O, the apparentdisks of the Sunandthe
Moon areextemaflyrangentalongthe line 07. ln rbts diag/Iam,ZZOM = /^, ZZ1U = z^, (8.2424)
d<.ts+sm+7rm-lts.
ZZOS = z!, ZZES = z6-ZMON = s'-, andZSOT = s:.

Conditions for a central solar eclipse If the observer on Earth Passes

the penumbra of the Moon's shadow and into the umbra, he or she expe-
a central solar eclipse. If the reference observing point is in the umbra on
O; point Z is at the zenith for an observer at O. The topocentric zenith distances inside of the vertex, the eclipse is total. If the reference point is in the umbra
of the Moon and the Sun aX O ate: the outside of the vertex, the eclipse is annular. If the reference point is in the
and.passesthrough the vertex, it is a,nannular-total eclipse' In the rare event
pass
z:^=w' - lMoN = 90' - r;, an observer nea,r a pole seesa total eclipse, but the shadow axis does not
reh the sufare of the Earth, the eclipse is a noncentral total eclipse' A central
z: = 904- zTos =goa+st. (8.242-t) and
r eclipse will occur when the apparent separation of the centers of the Sun
Moon is less than the absolute value of the difierence of the apparent semidi-
where sl and sl a.re the topocentric semidiameters of the Moon and the Sun, re-
The condition for the occurrence of an annular solar eclipse is specifically
spectively. The geocentric angular separation between the Moon and the Sun is
ved in this section, but the same condition may be used to differentiate between
angle MES, which is equal to the difiereuce between the eeocentric zenith distances
occurrence of any t5.pe of central sola,r eclipse (total, a.nnular, or annular-total)
of the Sun and the Moon, z, - zn . The geocentric zenitb distances (z) are related geome-
the occurrence of a partial solar eclipse. Figure 8.2421 1 illustrates the
to their corresponding topocentric values (a,) by of Figure 8 242 1'
The definitions in this fieure are identical to the definitions
topocentric zenith
that point O now lies ou the edge ofthe umbral cone The
z= zt - Tt, (8.242-2) alOarc
zl^= 90" + IMON = 90" + r;, (8.2421-1)
where url is the geocentric parallax. Assuming a spherical Earth, rt./is given in terms
+ zros =90'+sl.
z!"=90"
of the hodzontal parallax (zr) by
the
aa in the preceding section, the geocentric angular sepaxation between
zr' = sin-l(sin rr sin z'). (8.242-3) r and the Sun is angle MES, which is equal to lz" -2.1, th" difference between
geocetrtric zenith distances of the Sun and the Moon Employing the same
 435
ANDMOON
EXPLANATORY
SUPPLEMENT EOLIPSESOF THESUN

NCP
Point of a
axlsoI
shadow Fundamental
plane

Greenwich

Figurc 8.221'l
Klown geometdcparametersin the limiting configurationfor a partial solareclipse,whenthe Moon
is intemallytangent.As seenby an observerat O, the apparcntdisksof theSunandtheMoonare
intemally tangentalong the line OI. In this diagtarn,IZOM = zi^, IZEM = 26,ZZOS = z:s,
/zES = z\, ZMON = s'^,nd ZSOT = s',.

procedure and approximations as in the preceding section, the condition for an

Takethe{-axis
to thefundamental.plane'
annular solar eclipse is fisff"Hil;t" "r
r".centricequarorial *nro'[a'[e
coordinates new
a'round'lhe
o<ljs-sm+?rm-?rsl, (8.24212)
Jfffrffi:iiff#"131
ffi##i;T,f;;';;1.::ilJijy;trf
which, as stated before, may be used as a general condition for any central solar
eclipse. If s. > s-, the eclipse is total; if s" < s., it is annular. If o is near zero. the
eclipse may be annular-total, but that determination requires further calculation.

8.32 BesselianElements
To useBessel'smethodfor
8.3 SOLARECLIPSES 8321 TheFundamentatCoordinateSystem:Ov€rview
of the Moon with respeci to
solar eclipses,the geometrrcposition of the shadow
8'321 1) wherein the origin
Section 8.1 described considerations for ephemerides to use in calculating er:lipses. the Earth is describedi,. u "ooJinut" system (Figure and
of coordinatesis the geocenter,il;; is parault to' the axis of the shadow
Section 8.2 described how to find approximately when an eclipse of any kind ma1' ihe. equa'tor and is positive
occur. This section describes how to calculate the details of a solar eclipse. positive toward the Moon, tf.'" tlt*i" pu"'"" through
the north This is tlte Jundamen-
toward the east,and the y-axis o oo"n*" toward
plan,e'The intersection
tal coordinatesystem, andtft" ty-ptu"" is rhefundam-ental
concentricabout the axis'
8.31 Fundamental Introduction
Eouations: of the shadowwith the fundamJntulplat'" is frvo circles Earth is
The surfaceof the Earth -tt;;;;;;'o"ct to this plane' becausethe
Bessel developed the method used to calculate and describe precisely any eclipse' rotating on its axis and moving in orbit'
based on using a coordinate system oriented to the shadow axis. The basic steps are:
elements are calculated in the Besselian system to describe geometrical quantitiles; to the Fundamental
the observer's position is transforrned to Bessel's coordinate system; equations ol 8.322 Calcutation of the BesselianElements: Transformation
coordinate system' Iet R" be the
condition are formed; circumstances are derived that describe the time and pla<t CoorrlinateSystem In trt" g"oi"ni'i" equatorial
G = R' - Rm be a vector
of observable events or conditions in the Besselian system; the circumstances are position vector of the Sun, utta n' tfttt of tne Moon' Let
transformed back to topocentric or geocentric coordinates. collinearwith the shadowaxis' In astronomicalunits'
 436 EXPLANATORY
SUPPLEMENT ECLIPSESOF THESUNAND IT4OON
the lunar position vector into
lR'l=n=t' By $,,orkingout Equation 8.322 3 or by resolving
R- sin zrn using Equation 8.322-5,
r-= ---:::= "
"' & R,sinr-'
(t
, I cos6. sin(o- - rz) l
5--'
r , n F m=d = _ sin6'cosd-cos6'sindcos(o' o) . (8.322-6)
J r r r" m | I
,(s
c o \ d c o s ( o m- a )I
ao = horizontal parallax of the Sun at mean distance, L s i n d ms i n d + c o s d .

fcoscr,cos4 I
r" = sine.cos6" , (r,y) are the coordinates of the intersection of the axis of the shadow with the
| |
I sin6" I fundamentalPlane'
. f c o s o - c o s d"-|l In presentingBesselianelementsof eclipsesfor practical use,the right ascension
S l n7 r o | by the Greenwich hour angle // of that
rm = :__ I sln o- cos d- | . (8.322-1) a of the point Z is conventionally replaced
1(.stn 7r_
- | bY
L srndm I point, given
By definition, p = Greenwich apparent sidereal iime - a'
fcosdcosc I
B = r s _ r m = g l c o s d s i n la, (8.322-2)
according io the
In practical calculation, the Greenwich sidereal time is evaluated
I sind I
where (a,Q is the geocentric equatorial right ascension and declination of the in_ precepts given in ChaPter 2.
tersection of the shadow's axis with the celestial sphere (point Z).
Recall from Section 8.321 the fundamental coordinate system (x, y,7) whose 2-
a-xisis parallel to the shadow axis, or to g. The plane z = 0 is the fundamental plane, 8323 Param€ters ofthe Shadow Cone The radii of the umbra and penumbra on
the .r-axis is the intersection of the fundamental plane with the equator (positive the findamental plane can be easily calculated given the generating angle of each
to the east), and the y-axis completes the triad (positive to the north). cone, and the perpendicular distance of each vertex from the fundamental plane
Hence. R.eferto Figures 8.323.1 and 8.323.2 for the plane geometry of a cross section of the
cone. This geometry relates the desired quantities to the known size and separation
(8.322-3) of the bodies.
Flom Figure 8.323.1,
in units of earth radii. (R1 and R3 are standard rotation matrices, positive direction,
about the -r-axis and z-axis, respectively.) d"+d^
Sln, | = --,
In the (x,),2) system, let the unit vectors be i, j, and k. Then " gR'

gxk=0. d"-d^ (8.323-r)

SlnJ2 = -----
g.i=0,
c i=0' To transform to units of Earth equatorial radii (see Figure 8.323 2),
fcosdcosaI
,^-
._ c cosdsina, (8.3224)
t I
I sind j
| 4=l=Rssin?rs,
(8323-2\
4 = R- ' s i n s ' = ' I=- l+ +
Sln iro
and the orthogonal unit vectors are Sln ns

where ns = z16/ Rs, ss = so/ R". For the Moon, set

f-sinal f-cosasindl
i=l cosal. 1 = l . s i n a s i nld. (8.322-s)
L0.l L cosd I d-=k earth radii, where k = d^ lde. (8.323-3)
 439

FlgureE,3232
amongatlgularsemidiameter
Relationships r, apparent
semidiameter
d, and
parallax
r. R, is exprcssed
in Eanhradii.

l,=1=R'nsinz'n,
sin Jm (8.3234)
d'n=&=R.stusn'=
sinzr-

R = & sin ?ro be the distance to the Sun in astronomical units. Substituting
8.323 2 through 8.323-4into Equation 8.323-1 and reducing yields

sin/1 = (sins6+ t sinrd / gR,

sinJ2 = (sins6- ft sin?r0)/ gR. (8.323-5)

numerators are elaluated using adopted values for k, s0, and 7r0.Equations
L a d 8.322-2 give g, a.nd R is from the solar ephemeris.
The distances cl, c2, of the vertices of the penumbral and umbral cones above
fundamental plane are shown in Figure 8.323.1. Figures 8.323.3 and 8.323.4
that

Fundamentalplane
cl =z+&cosec/1,

ngue &38.f c2= z- kcosecf2, (8.3234)

of shadgwconesin theBessetian fundamental reference a!
::T!:Y!
r. w h apparent system.TheSunis centelect hence the radii /1, 12of the penumbra a.nd umbra on the fundamental pla.ne are
semidiameier 4, arheightx ab"reth; tu;;;;i"f ,ii," frl*" ar
/,' wrthsemidiamerer o-1fii. is centered
d.. at heiehrr = t66e""-tn" f*0"-"ntuf-pf'-J. ii-" f--6"-"ntA pf-", ftom
throughthcE"nh\ cenier,andtfieobserver,s pr-" *" p".p"naluiar io ,ri" Ji, [,-ugr, srrz,,r,"
l"-!,.^|do_o
Eing (. Ttrepenurnbra hasis vertexyr athiight cr tj.,"iuoa..*tuf phne,radius
,i1.19:tu'ld"T*"t plane, Zrin rheobserv".',
(Jubscnpt I refersb peourDbra_
pil, ;;;;;; "Uo"" l3i#iilry ri,,,r,"un-'u,u.
f:, h = qtznfr,
2 to umbra_|
lz = cztanfz. (8.323-:7)
 440 EXPLANATORY
SUPPLENlENT OFTHESUNANDMOON
./ ECLIPSES 441
of the Earth and the fundamental plane may be calculated for any instant
the eclipse.

In practice, using a large-memory

1JZS A Note on Practical Calculation
begins the computation by constructing short ephemerides of the
computer, one
gun ard the Moon from the desired source' at a convenient interval, say, one hour,
possible with
and then subtabulates to smaller intervals as desired. However, it is
some ephemerides stored in polynomial form to take the beginning ephemeris di-
rectly at the desired interval. This directness is especially helpful in calculating
figure 8.323,3 points of cunes near the beginning and end of a central solar eclipse.
Vertexangleof thepenumbra,
detailfrom Figure8.323.1.sinfr = d^ I (ct d. Initially, for each eclipse, calculations are carried out using a ten-minute time
interval, bui for some instances additional calculations at an interval of 30 secondsor
lessare necessary.In all instances, the Besselian elements, derivatives, and auxiliary
quantities are calculated from the ephemeris entries and carried as an array; none is
assumedto be constant,. Derivatives are taken numerically. Most quantities change
so slowly that only a low-order scheme is required.

of the Observer
8,33 Coordinates

8.331 Geocentric Position Consider an observer located on the surface ofthe Earth
in longitude l, at geocentric latitude //, and at a distance p from the center of the
Figure 8.323.4 terrestrial spheroid. The longitude is not now corrected for z1? and is considered
Vertexangleofthe umbra,detailftom Figure8.323.1.sin/, =d./(z c). positive east (right-handed system). Careful attention must be paid to algebraic
eigns (see Figure 8.331.1). In the.x-, y-, z-axes of the fundamental plane system, in
units of the Earth's equatorial radius, let the observer's position vector be

ltl
The quantities zt k, cl, c2, 11,and,12are in units of Earth radii. The convention of
signs introduced in Equation 8.323 6 makes /2 negative for total eclipses -i.e., the ( 8 . 3 3 1I -)
vertex is below the fundamental plane. For annular eclipses, /2 is positive, and of
course /1 is always positive.
In the geocentric equatorial system, let the observer's position be

8.324 Sumrnary of Besselian Elements The quantities x,

), sind, cosd.,11.,fi, 121
tanfy, tanf2, p, and d. are conventionally designated as Besselian elements. (N.8. fcosd' cosAl
p6 = p I cos@/sin.\ (8'33r-2)
It is conventional in almanacs to use the notation z/. d' lor lhe derivatives with |.
respect to time.) In the publications, the first seven quantities are tabulated as a L sind' I
function of time at a short interval, or may also be given as a low-order polynomial
The transformation to the fundamental system is very similar to that of Equa-
as a function of time. The remaining four quantities, to the precision required, are
tion 8.322 3:
constaut for the entire eclipse and are given at the conjunction value. Using all
these quantities, the size and orientation of the shadow cone vith resDect to the ^=R,(;-r)*,(-Q-?)^ (8.331-3)
 442 EXPI-ANATORY
SUppLEMENT OF THESUNANDMOON 443
Ccocentric Velocity It will also be necessaryto know how the observer is
with respectto the fundamentalplane, The componentsof the motion are
about the same two axes used in the coordinate transformation: rotation
the polar axis of the Earth because of both the rotation of the Ea.rth and
of the Moon, and about the axis pointing to the equinox becauseof the
of the Moon. In the geocentric rectangula.r equatorial system, the rotation

to the fundameutal plane,

l;l (8.332-l)

Creenwich
'l
I d
pnme
meridian tetp = Rr [\ z) - , l l, . a = l| p c o s-dl l . (8.332-2)
Lpslndl

velocity of the observer with respect to the fundamental plane is

pr =,,trx pr = R, (I - a)." (8.332-3)

Observer's " *.
meridian
operation @Fx may be expressed as a matrix premultiplier:

0 -;i sind pcosd'l

n=fr,,.,",",.f= Psind -d (8.3324)
Ftgure 8.331.1
-it cosd
I l.
Angular quantitiesin the geocentricequatorialplane | d O I
of Equation 8.332 4 into Equation 8.332 3 yields
It is convenient to use the local hour
frr(-nsind+ -(cosd)I
angle in further calculations. Because by
curent cotrvention the Greenwich hour
in opposite directions, the local hour
angle (p) and longitude ()) are measured n=Am= I ir€sind dC I (8.332-5)
angle (d) is -it(cosd+
| dq I

g = p+ ). Coordinates of Observer with Respectto the Fundamental Plane, Corrected

(8.3314)
Flrttening Consider a spheroid of ellipticity e. Geodetic positions are specified
Applying Eqr.ration8.341 3 to Equation8.331-2, the surface, Calculations of eclipse tracks in the fundamental plane are no prob-
combiningterms by multiple_angle
lormulas, substituting Equation g.331-4, nor a,re local circumstauces for a given point. However, in translating any
and resolvingth1 componentsalong the
unit vectorssives from the fundamental plane to the surface of the spheroid, eraluation of (
knowing p, the distance of any point from the center of the Earth. This
{ = pcos// sind, is a function only of latitude, so that ( could be determined by successive
q = psing'cosd- pcosO,sli.rtdcosl, (8.331-5) However, Bessel devised a procedure to provide a direct computation of ( from
C= psinf' sind+ pcost', cosdcosd. fl' This procedure requires a set of auxiliary Besselian elements, based on the
element d a.nd the ellipticity € of the Bpheroid.
 444 EXPLANAToFY
SUPPLEMENT OF THESUNAND MOON 445
Let a and b be the equatorial and pola.r radius, respectively,of the
spheroid.
Then
sil af= p1sind1,
(l - e2)rt2cosd = p7 cosd1,
f l a t t e n i n gf , = o - b , ^ o d
cl
ellipticity, e = (l _ b2/ a2)u2. (l - e2)rt2sind = p2 sind2,
(8.333-1)
cosd = p2 cosd2- (8.333-8)
Define

g = 11- e2sin2 S = (1 - ez)C:

S)-1t2,
or € = cosdl sin0,
C = ( c o sO
2 + ( t - f ) 2 s i n z6 7 - u z , S=(t-f)2C. (8.333-2) q = n l A = s i nd r c o s d l- c o sd l s i nd l c o s0 , (8.333-9)
(= p2$it$1sind2+ cos/1cosd2cosd).
The relations between geocentric latitude // and geocentric distance of observer
o
and geodelic lat itude / are

e?=t-t"-n?' (8.333-10)
p s i n d ' = S s i nd
pcosA' = C cosd. (8.333-3) extensive algebraic manipulation and reduction, this becomes

To eliminate p and S', transform to a paxametdclatitude (r = sindr sindr + cosdt cosdr cos0' ( 8 . 3 3 3 -ll)
/1, definedby:

d.- desirable to eliminate 6r and'0. Then Equation 8'333-9 combined with Equa-
cosgl= pcosgt= --_1t =ccoso. (8.333+)
\r - e. stn q1". 8.333 11 mav be written

Then, from Equations8.934-2,8.3g3-3,and 8.333-4,

i€l fcosdrsindl (8.333-12)
l n ' ; = P ' , - o ' ,I s i n { 1 |
L('l fcos@ c o1s d l
s i n @=11 1 - c o s 2 d r )=r n, { f + n (8.333-5)
\t - s'7"
Inversely, f c o s / s1 i n o l lll (8.333-13)
=
Pcos// = sss4t = C cosd, I sin@r I Rrtdr)| "ir | .
(8.333-6) I cos@1 cosOl L(rI
psingt = (l - e2)l/2
sin41 = 55ipp.
components of Equation 8.333-13 to the expression for ( in Equation 8 333-9
Substituting into Equation 8.831-b, reduce to obtain

q = p2[j cos(dt- d) - ry sin(dt - d)1. (8.333-14)

€ = cosdl sind,
ry= sindr(1- e21tt2 cosd- cosglsindcos0, (8.333-:l) 4 Auxiliary BesselianElements and Summary The following defined quanti-
( = sindr(l - e2'1tnsind.+ cosSl cosdsinA. axeDart of a set knov/n as auniliargelementslmore aredefinedin Section8'3422'
elements are used to introduce the flattening of the Earth into the ecllpse
Now to eliminate the terms in d and e, deflne All. quantities a.re in units of Ea.rth's equatorial radius'
 446 EXPLANATORY
SUPPLE[,IENT OFTHESUNANOMOON 447

for umbra or penumbra, and L may be positive or negative'

h=(l_AcoszO1t2,
p., = 1l - e2 sin2d)tn,
sindt=sindlPt, Equations of condition We may now impose geometric and dynamic con-
consider two
c o sd f = ( 1 - e z ) 1 t 2 c o s d l p 1 , ,ns in order to calculate eclipse phenomena. We know r' We may
we wish to find the
sin(fi - d) = s2 sindcosd / p1p2, from Equation 8.342-2: otte, in which p is not known but
a condition as a function of time (curve); the other, in
of all points satisfying
c o s ( d r- d z ) = ( l -"')t'2 /ptpz. (8.334-l) !h p is fixed and we want the times of significant configurations (circumstances)'
case, the difference being the
Much of the following discussion applies to either
(The subscripts do zot refer to the umbra and penumbra.) In computation, all these and unknowns. First, consider the more general cases,in which p is
of knowns
qualtities are carried with the regular elements at the same interval, even though to both umbra and penumbra, so the subscripts
known. The discussion applies
^, p2, and, (d1 - d2) are nearly constant for the duration of the eclipse. Once the omitted
elements a.reavailable and a point ((, 4,0) is given in the fundamental plane, then
If the observeris on the edge of the shadow (penumbraor umbra) or, more
/ and ( are obtained as follows: point (€, nr, Cr) is calculated from the definitions , the surface of the shadow cone' the eclipse
is either beginning or ending'
of 41 and (1 in Equations 8.333-9 and 8.333-10. Next, Equation 8.333-13 gives @1 surface location may be defined by imposing the geometric condition
and 0; Equation E.333-6 gives @;and Equation 8.333-14 or 8.333 7 gives (.
(8.3421-1)
l A l= l r l , or A A-L2=0.
8.34 Conditional
and Variational
Eouations
evaluated at
8341 Introduction This sectionestablishes
conditionsby whicheclipsephenom- sider Equation 8.3421-1 now as a general expression that can be
imaginary, then the observer is
ena may be distinguished and calculated. insta^nt, The equation has two roots. If they are
the shadow. If they are equal, then the eclipse is simultaneously beginning
or southern
8342 Conditions Defining General Circumstances and Curves Consider now the ending-i.e., "grazing"-and the observer is on either the northern
relation between observer and shadow. We have, in the fundamental plane reference be defined by imposing the dynamic condition
; of the shadow. This limit may
system,the geocentricposition ofthe observerp=(t,n,O. The geocentricposition
of the intersection of the shadow axis with a plane through the observer parallel to !,t.d,-t't=0,
the fundamental plane is dt'

"=i'l A'it-LL=o. (8.342r-2)

(8.342-r)
Lil
Now define the vector in the observer's plane from observer to shadow axis as (8.3421-3)
lAl = ltl, to A'A - i = 0.
thisreduces

A=r-p. (8.342-2) to use

is always the case that d! t dt = 0 for fl and t2' so it is sufficiently accurate
In the pla.ne ihough the observer, let Q be the position angle of the axis, measured
from north through east, so that the unit vector L=i- Ctall.f,
(8.34214)
e=ez'p.
fsinOl
A= l*.b| ' (8.342-3)
Equation 8.3421 3, substitute
L0l
The radius (Z) of the shadow at height ( above the fundamental plane is L=i - P,
p=@FxPF=(.,Fx(r-A),
L=l-Ctanf, (8.3424)
 448 EXPL,ANATORY
SUPPLEI,,|ENT OFTHESUNANDMOON
/ ECLIPSES
to get Additionalauxiliary €l€ments With a largecomputer,Equation 8.3421-10
of (, p that satisfy it, and hence
A.i- A . ( q , Fx ( r - L ) ) + t a l f q t z . ( r . A ) - l = 0 . (8.3421_s) inay be used in this form to find combinations
however, it is conventional to define auxiliary
the limiting curves. For efficiency,
being a separate set for umbra and
In the second term, ouantities a, b, c wilh derivatives h, b, t, thete
o"or.-btu as distinguished by subscripting and use of the corresponding angle /,
x 1r- A)) = A. (c.,rx r).
A. 1c.rp (8.3421_6)
suchthat
rf
From Equation 8.332-4: f sino'l .l f sinol/f -acosal
l'*ol |-D|*c""/|'o'o| | a l+a=o' (8.3422-r)
0 -!sind acosdl o o I L o I
| fx l L J Lol L
o,xr= -d
| rrsina 9 llrl
l-r{cosd d 0 I L(l Thus, to make Equation 8.3421-10 identical to Equation 8-3422 I,

f - r s i n d-+ ( c o s d l a= -l - prcosdtarf+ydtanf,
= i) xsind ed.tp
| I b=-i+ilxsind +ldtanf, (8.3422-2)
I xcosd+yd/it)
-rsinal a=i + itysind +/ptanfcosd.
|
=slxsind
f-rcosal
l_(l d L (u.342r-1)
In older references, the approximations dtanf; = 0, sec'/i = I are used, which are
l-r.o'a-l L yd,C) reasonable and adequate for limited computer resources
From Equations 8.332 2, 8.332 4, and 8.3421 1: Now collect terms in sin P, cos Q:

I d 1 | r.o'a1 h + sinQ? - p( cosdsec2/) + cosQeb + (i sec2

h=o l8.3422-3)
wr= -.=l 0 A=A(/-(tan/). ( 8 . 3 4 2s1)
l i r c o s dl . l,
l/,sindl L 0 l Multiply by secQ to obtain

Substitute theseand Equations 8.342I_1,8.3427 6,8.3421 7 into Equation 8.3421 5 b-dCsec2f-asecQ

tar'Q = l8.34224)
and, through algebraic manipulation, collect terms as factors of A. and ( to obtain i - ir( sec2fcosd

The application of this fundamental equation to calculate limits of the shadows will
f [-r'ina1 ) be describedin Section 8.355. In calculationsfor ?i.e Astronomical Almanac, the
o ' j,'::lr] - ttan
r@, +4lsec2rA
c'r.
) auxiliary quantities are calculated along with the Besselian elements and carried in
t' l ]
an a,rray as a function of time.
- /+tan/o..r = 0. (8.3421-9)
8.343 A Note on Practical Calculation The Besselian elements, both regular and
Equation 8.3421-9may be written in componentform as:
auxiliary, are nearly linear in behavior during an eclipse, such that integration
.or numerical differentiation selflom requires more than a third-order Lagrangian
| -.r,'i'{
i43l'{lil.,l
':""] echeme,and linear interpolation is often adequate. In preparing computer programs
srnd. I -,,un
r (8.3421-10) and carrying elements in an array, howevet, it is always necessary to be sure that
|
L rcosd | ) angular quantities change monotonically, especially when passing through 360" It
l r
f s i n P l f -pcosdl -p cos
d'l is also necessaryto carry enough entries at the beginning and end of the eclipse to
.c*.,r | fxl

llol I 1l
ul
l+ta\f
LtlL;l allow interpolation, but not so many that the behavior of the tabulated elements
acceleratesor reverses and potentially creates specious extrema.
 oFTHE sUN ANDMOoN 451
450 SUPPLEtt4ENT
ExPLANATORY / EOLIPSES

Even though in calculations for The Astronomical AlmQnac all quantities q1.
carried in double precision (15-16 digits), the published data should be considered
accurate only to the number of significant figures presented.

of GeneralSolarEclipsePhenomena
8.35 Calculation

8.351 Intmduction The Besselian and auxiliary elements are the basis of calcu-
Iating all further phenomena associated with each eclipse' utilizing, of course, the j-(e - a
" i
necessaryephemeris data given in the eclipse section or elsewhere in The Astronom-
ical Almanac. The Eclipse section also gives positions on the surface of the Earth
as a function of time, where well-defined pha.sesor phenomena occur according to
certain geometric or dynamical conditions-e.g., geocentric beginning and end of
each phase, These data are given in tabular and graphical form, and are known
as general circumstances, elements of trhe eclipse, and curves of several kincls. The
independent variable, unless stated otherwise, is Universal Time. Conversell'. for a
given imposed condition, a time may be either deduced or assumed, and the coordi
nates (4, ri,0) determined.Using the auxiliary elementsA, p2, dt, d2, the quautities
( and ({, 41,(1; may be calculated and then transformed to \' 0 ' $ ' 6. General cir-
cumstances a.reusually calculated before the curves, but for purposes of exposition,
calculation of the curves is described first, Fisure 8.352.1
[iation.ttip l"t*.en rectangular andpolarreference systemswith theoriginsat the
Eclipse curves are the loci of points on the spheroid of the Earth at sorne referenceplane Anglesare
oUt.rver'sp'roiected point,in thefundamental
specified radius, usually the surface. These curves delineate set boundary conclitions' ""*"ntioi,ft"
f; il;;ntr. oiy-ixis ln the geocentric system'the observer's projecredpoint
;-;;; (r'y) or
and hence define areas within which ceriain eclipse phenomena are visiblei they also maybe sDecified as(€, a) orip. I )i theshadowaxisinlerceptmaybe specifiedas
il;; .;f*""i; syste- cente.edat the projected
observer's point' the positionof the
show time of occurrence. They are produced by going through a table of Besselian i;:d
elements as a function of time, calculating a table of derived discriminants and shadowaxisinterceptis (.r - €,) - 4) ot (A'Q).
evaluating whether imposed criteria are satisfied.
General circumstances (as opposed to local circumstances) are individual points
and times at which extrema of the curves occur on the spheroid of reference'

path of
As derived in Section 8.3421, if the point of the observer is "tangent" to the
8.352 Basic Geometry In general, eclipse phenomena are treated and calculated
1' the shadow, such that the eclipse is beginning and ending simultaneously' then the
using the fundamental triangle in the fundamental plane shown in Figure 8 352
secondfundamental relation. wnich is a condiiion on Q, is (from Equation 8.3422-3)
By imposing various conditions and finding corresponding points in the funda-
mental plane, it is then but a relatively easy step to project the point to the surface
ai - bicosQ+q sin0+ ((1 + tan2f i\(dcosQ - l'cosdsin0) = 0' i=lor2.
and convert to latitude and longitude as described in Section 8.33 ; (8.353-2)
P is usually not unity; it contains flattening and height above the geoid'
pro-
E.353 Relationships The fundamental relationship of the geocenter (0, 0), the These relationships axe also used:
jected axis of the shadow (.t,y), and the projected position of the observer (€' 4) rs:
fll = fx-4sinPl - [asinr'l (8.3s3-3)
p=m-4. (8.353-l) L n l L ,- A c o s Q ) [ r " o . r l
 452 EXPLANATORY
SUPPLEII,{ENT 8 / ECLIPSESOF THE SUN AND MOON 453
Furthermore,the cosinelaw for plane triangles (dot-product) gives (b) tan11t= pt sin 1 / cos 1;

(") 'vl = tan l(tan?');

-
A.A=(m-p).(m-p) (+cos(r ^ - m 2+ p 2 a 2 ,
zmp (d) p=sin?//sin1. (8.354-4)

p. p = tm- a). (m- A)<+cosle

-, = ^' *r2l; n . (8.3s3_4)
Repeat the steps of Equation 8.354 4 three or more times, until convergence.
A ' ?+ = 4 - n 2 Then { = sinl', ?l = cost/ (and Cl = 0).
m . m = ( p + a ) . ( p + A ) < +s i n 1-1p \ = .
zlJp
E355 Curves The order in which curves are computed is not importanti compu-
The sine law for planar trigonometry (cross-product) gives
tations a.reall done independently from the tabulated elements. Holv'ever,it helps
to leave computation of the outline curves until last.
m x A = [.4(rcosO -ysinO)
= kAn srn(e - M),
E3551 Rising and setting curves An observer who is at a point of intersection of
pxA=(m-A)xA=mxA the penumbral cone with the fundamental plane is also on the sunrise/sunset ter-
= kpA sin(t - e). (8.3s3-5
) minator and may, weather permitting, see partial eclipse beginning or ending at
sunrise or sunset. The locus of all such points on the Earth's surface, known as the
Another usefulrelation is rising and setting curve, is shown graphically on eclipse naps in ZAe AstronomicaL
Almanac with lhe label Ecli.psebegins at sunrise, Eclipse ends at sunset, and so on.
(l-m+p)(l+m-p) The rising and setting curve is the locus of end points of the outline curves for
2sin' ;(? - M) = | - cos(.)- rt4)= (8.3s35)
z 2mp the penumbra as it intersects the terminator. At a selected instant, there are tlvo
poiuts on the curve or no points. If there are two points, each belongs to one of two
8.354 Flaftening Flattening is taken into account by altering p and 7 slightll to
series. For a successionof instants, the two series of points trace different branches
rp and 1'. Assume that €'*rfi+C? = l. No* if (1 = 9, 15"n €, *,fi = P-L"t
of the same curve and are joined at frrst contact (penumbra emerges from the
terminator and intersects the Earth's surface) and last contact (penumbra merges
{ = rp sin ? = rsin ?', completely into the terminator and leaves the Earth's surface). The track of the
I)
(8.35,1- penumbra always generates at least one limit curve (Section 8.3553), either northern
n=rpcosl=rAcos1',
or southern. but if both the northern and southern limits exist for an interval of timc
ry=n/ A = r c o s1 / .
(i.e., the central path is in equatorial regions), then during that interval the rising
al'-rdsetting curve does not exist the penumbral shadow is completely emerged
Then
from the terminator. The result is that the rising and setting curve forms two
tanl' = € I rh = A€ / n = pttwt.y, (8.3s4-2) sepaxateloops like elongated teardrops, each loop being generated in tn'o segments'
Otherwise, if there is only one limiting curve, then the loops are joined, and the
and
curve of rising and setting forms a distorted figure-eight. In either case, one loop
sin
- 7' = p = p t cos
_ - , 1' (8.35,1-3) or lobe is for rising, the other for setting. In a series of similar eclipses,both forms
srn7 cos 1
. occur. When the first separation of the Ioops occurs in the series, the break into
' separate
The following iterative procedure accounts for flattening. Ordinarily, three itera- loops does not occur at the apparent node, but near it.
tions are sufficient, for each of ! \1 - m), In the fundamental plane, the circle of the penumbral cone' at any instant'
To start, assume lpl = l, where p = E,nl from Equation 8.353 3 and Fig- intersects the surface of the Earth in no points or in two points (which may occa-
ure 8.352.1 sionally be coincident). In the fundamental triangle,

(a) sinl={l p, cosl=nl pi C=0, A=h are imposed conditions. ( 8 . 3 5 5 1 )- l

454 EXPLANAToRY
SUPPLEMENT / ECLIPSESOF THE SUN AND MOON +cJ
If R: is the matrix for positive rotation about the z-axis, then Central line, duration ofcentral eclipse,and width ofpath The central line
the locus of the points of intersection of the axis of the shadow with the surface
the Earth. The following conditions must be satisfied:
p = Lf,.3(! (t - r0)m. (8.3s51-2)
m
p.m= pm,
Pxm=0,
I cos(+i1 - ,Yr;; 1 sin(i (r - n) 01 €=x, n=y, ry=ytpb
Rr= sin{to - a0r cos(+
(" - .44)) (8.3s51_3) (2 +r72+(2 = /P2'
l+ ?]
t'z+rtl+q!=l (8.3s53-1)
where one uses all top signs or all bottom signs.
Equat ions 8.3553-1 and 8.333 l4:
If, by the cosine lavr (Equation 8.353 4), cos2l! (/ - m)l < 0, then there are no
points on the curve for the selected insta.nt.
A=+(/ -€2 -n?)1t2,
( = p2[Clcos(dr - d) - th sirL(dt- d)]. (8.3s53-2)
8.3552 Curves ofrnadmum eclipsein the horizon Any locus ofpoints that satisfies
the fundamental relation given in Equation 8.353-2 (a dynamic constraint) is a on the opposite side of the spheroid may be found by taking the negative
curve of maximum eclipse. If the geometric conditions are also imposed that ( = 0, of (t'
4 = lt, then we have the curve of maximum eclipse at sunrise or sunset, or maximum The duration of central eclipse may be estimated by disregarding vertical mo'
eclipse in the horizon. On the map this curve approximately bisects the lobes of aad considering only the instantaneous size and velocity of the intersection of
the rising and setting curves. It is continuous if the rising and setting curves a,re shadow with the plane through the observing point and parallel to the funda-
joined, otherwise not. It is also generated in two bra,nches,as are the rising and pla.ne.
setting curves. From the section on conditional and variational equations (Equation 8.3421 5):
Flom Equation 8.353 2.
PF=@FxPF
(=0 components
q4 - brcosQ+ cp sinQ = 0, t)
(8.3s52- t = p(-ysind+ (cosd),
i = iu sind - de. (8.3553-3)
and from Equation 8.353 5,
sing the definitions
p x A = m x A = psin(7- Q) = xcosQ- ysnQ. (8.35s2-2)
#=@-€)2+O-ilz,
From Equation 8.3552 1 find two values of Q, then for each find a value of 7, then h = b - eanfz,
1/ as before (Equation 3.354 4).
Points exist on these curves onlv if both these conditions are satisfied: duration is
2Iz I n (with appropriate sign). (8.3ss3-4)
sin(7-@)<1, (8.3s52-3)
units must be converted as neceasa.ry. Normally the conversion is ftom hours
(x-O2 +(1 -q1p)? ! l2r. (8.35524) of time. Before 1982, Lz had to be adjusted for different values of &.
In local circumstances. however, the accurate duration is the difference between
pairs of conta,ct times.
 456 EXPLANATORY
SUPPLEMENT , ECLIPSESOF THESUNAND MOON 457
The width of the path, perpendicular to the direction of motion, may be esti_ coordinates ({,41, (1) are then easily converted to (d, l).
mated by a formula derived by Mikhailov (1931):
Outline curves An outline curve of the shadow (umbra or penumbra) on
= 2Lz{c* - n. - points at which the corresponding central or partial
width Jhe spheroid is the locus of all
lf,, }o ot)'}-'''. (8.3s53-s)
iclipse is beginning or ending at a given time. Outline curves are usually
published
penumbra, showing
only in the form of special maps. Outline curves for the the
83554 Northern and southern limits of umbra or penumbra Each curve limiting places where the partial phase is beginning or ending at stated times, are plotted
a shadow path is the locus of points on the surface of the Earth's spheroid tangent on the small-scale maps presented in The Astronomical Almanac and other publica-
to the path of the shadow, such that first and last contacts are simultaneous, or else tions. (See also Section 8.361.) Outline curves of the umbra are not published. For
maximum occurs in the horizon. The umbra nearly always has both limits on the all practical purposes, it is acceptable to neglect the flattening of the Earth in these
surface because it is so small, but the penumbra often has only one. The northcrl calculations, but for the maps in The Astronomical Almanac, it is not neglected.
and southern limits of the umbral track define the central path; the northern ancl/or The angle Q becomes the independent variable. At the instant for which the
southern limits of the penumbral track, plus the curve of rise and set, define the curve is calculated, if both the northern and southern limits of the penumbra exist,
Iimits of the eclipse. then @ has the complete range of 0' to 360'. If not, the extreme values must be
Impose the condition that Equation 8.353-2 must be satisfied. Choose a vaiue found; they are two points on the curve for which ( = 0. Because these points are
of (. If a complete set of limits is being generated, it is usual to begin with ( = 0. on the limiting curves, in the computer program an initial test is made for the
Then the coefficients of sin p and cos O are evaluated and the left-hand side of availability of points already computed for the limiting curves.
Equation 8.353 2 is used as a discriminant. This discriminant is evaluated for;r Those points are then verified from the equations
run of p, starting frorn a first guess obtained by setting a; = 0, and proceeding a
degree at a time until two zero points are found (by inverse interpolation when tlrr: e = x - LisinQ'
discrimina.nt changessign). Then for each Q, iterate on ( as follows, using Besselian 7,tpt=y LicosQ,
elements:
C'+4i+(i=r, (8.3555-
I)
Li = li - Ctanf i,
in which (1 = 0. Once the end points are found and the correct portion of the
t = x - LisinQ, arc which intersects the illuminated surface is established, points are calculated at
ry=6)-LisinQ)lpt, one-degreesteps of Q, iterating as described for ihe northern/southern limit curves
C?= / - t' - n? [f <?< 0, thereexistsno point.) and requiring (1 > 0, If desired for special purposesr points below the horizon may
Normally choose (1 > 0 to continue. be found by using the solutions for (1 < 0.

(1 < 0 may be used to continue the curve below the horizon.

83556 Other curves For 1960 1980, rnaps in the American Ephemeris and Nau-
C = p2l|t cos(dt - d) - rtr sin(dt - d)1, |)
(8.355+ tical Ahnanac showed curves of equal middle and equal semiduration of eclipse'
i = umbra or penumbra. These curves were constructed graphically from the intersections of a network of
outline curves, as there is no way to calculate them directly.
Put this value of ( into 8.353 2, and repeat until it converges to the required
Curves of maximum eclipse are not prepared for regular publications, but may
tolerance. In The Astronomical Almanac, that is l0 5 earth radii. ln qllrsrnss of .be readily calculated, Such calculations are an extension of the procedure for the
these curves, the iteration may begin oscillating about some mean value, in which curve of maximum eclipse in the horizon, in which the condition ( = 0 is removed,
case the mean is taken to continue the iteration. and ( becomes the independent variable i.e.,
When the iteration converges, then the point is assigned to a limit by the
discriminant: Choose (;
LicosQ < 0 for northern limit Find Q (two values) from Equation 8.353 2;
LlcosQ > 0 for southern limit. (8.3s54-2) Find 7 ftom sin('y - Q) = (xcos Q - y sinQ) I rp (fundamental cross-product)'
 458 EXPLANATORY
SUPPLEI\4ENT / ECLIPSESOF THE SUN AND MOON 459

Curves of equal semiduration or equal magnitude must be found by inverse in-

(See Section 8.36 on local circumstances
lslpolation on curves of maximum eclipse
for furlher of
dis,-ussion magnit ude.)

points known as the general

8356 General Circumstanc€s A collection of event
ci,rcumstancesare the important extrema of the curves of the eclipse. They define
join
the bounds of the eclipse on the spheroidal surface and in time. They also
together the various curves and are essential in preparing the eclipse map. General
circumstances are computed from the Besselian elements through a set of discrimi-
Figurc 8.3556.1 narrts. For each discriminant, a table of values is calculated for a sequenceof times,
Definitionofauxiliaryquantity{. Z is radiusofthe shadowin rheplaneof theobservet and used as described in Sections 8.3561 8.3565. In practice, general circumstances
.4 is distanceof observerf.om theaxis,r is speedof separation,
J is semi-duration.
a,recalculated before the complete curves are generated. If circumstances are re-
quired for high altitudes, the calculations are repeated for a larger spheroid.
The first and Iast contacts of the penumbra occur for every solar eclipse. They
axe the points where the shadow cone of the Ntloon first and last encorrnters the
If neither value of Q gave an imaginary result, only one may be an eclipsed point spheroidal surface. These points are shown on the eclipse r.:'apsas First Contact
in most cases.It is a-lsoreouired that arLdLast Contact, and always lie on the rising and setting curve. The tirnes at which
they occur define the date(s) of the eclipse.
lzll < 'Zl. or {x - {r2+ g \} > Ll. /R 1556-l I If the eclipse is central, then there are first and last contacts of the umbra. These
axe the extreme end points of the central line, and fall on the curve of maximum
If this test is met, iteration may proceed as usual. eclipse in the horizon.
From maximum eclipse, semiduration of the partial phase may be estimated as For a central eclipse, there are also extreme points of the northern and southern
follows. Define ty' in Figure 8,3556.1, where s is the semiduration and r?is the speed limits of the central track. It is a ra.reoccurrence when only one limit exists These
of shadow. The respective units are hours and equatorial radii per hour. Then points also fall on the curve of maximum eclipse in the horizon.
The extreme points of the northern and southern limits of the penumbra fall on
both the rising and setting curve and the curve of maximum eclipse in the horizon
5 = (Li cos t'b)I n'
ard, thus, mark the joining points. The extremes need not be known accurately;
where
they are the points at vrhich an observer theoretically could "see" the limb of the
sir*=AlL'
Moon brush the limb of the Sun for an instant lf the eclipse curves have been
n 2= ( i _ t ) 2 + O _ 2,
calculated taking flattening of the spheroid into account, the same must be done
{=pcosqlcosd, for these points or the curves will not link.
The point ofcentral eclipse at local apparent noon or midnight is self-explanatory'
i = i4sind.. (8.3s56-2)
i.e., maximurn occurs on the meridian. Meridian passageat midnight may occur in
the polar regions.
The exact duration is found from the diflerence in times of first and last contacts,
The point of greatest eclipse is calculated only for partial eclipses and shown
as describedin Section 8.3621.
on the eclipse map. It is the point on the spheroidal surface that comes closest to
The magnitude of the eclipse is found from
the axis of the shadow, and it lies on the curve of maximum eclipse in the horizon'

Lr- A
(8 3556-l) 8'3561 Firstand last contacts ofthe penumbra First and last contacts ofthe penum-
t1+t2
bra occur when the shadow cone is tangential to the spheroid. Thus at two instants,
If Z2 is not available, the approximation Io = Lt - 0.5459 is used. (Subscript I for
penumbra, 2 for umbra.) I + f = Q 1 +p 1 2 . (8.3561-l)
 460 EXPLANATORY
SUPPLEMENT ECLIPSESOF THESUNAND MOON 461
Taking the flattening of the Earth into account, let
€=.r-lisin0, i=1,2,
n 2= ? + f , ry=O+licosQ)lA, i=1,2,
n?=i+Y?, q-+?i=1. (8.3563-t)
Yt=y/ p,
that the sign of cos Q determines whether the point is on the northern or
p=mlml. (8.3561-2) limit. Note also that there axe two cones and, therefore, four discriminants
the form
Then * + 1\ = 14 + m I m)2 is the coudition; the discriminant is
- t:oso)'z
D i : @ - t s i n e ) 2+ ( Y - t, i =3,...,6. (8.3s63-2)
oy=*+fi-(\+rutm)2. (8.3561_3)
In the table of discriminants, the times of the contacts the times a.re found by inverse interpolation, aqd the coordinates, with
are calculated by inverse
interpolation when Dr = 0. These times are used to obtain taken into account, by the usual procedure.
x, y, and p1. At the times
of contact,
Central eclipscon the meridian Central eclipseon the meridian noon or
€=xlmr the point on the central line at conjunction of the Sun and the Moon
y1=
1y1 /m1=y/AmL right ascension. The discriminant is x, and the condition is x = 0. Test the sicns
in polar regions, to distinguish noon from midnight.
(r =0 (8.35614)
gives the data to compute latitude and longitude Greatest eclipse In pa.rtial eclipses, the maximum magnitude, or greatest
by successiveapproximation.
occurs at the point on the surface of the spheroid that comes closest to the
83562 Bcginning and end ofcentral eclilxc Central of the shadow. The magnitude changes very slowly, so high precision is not
eclipse begins and ends when
the shadowaxis is tangentialto the Earth,s surface_i.e.. The eclipse will occur in the horizon at this point, and it is sufficient to
the time when the shadow axis is closest to the center of the Earth or when
rate of change of separation goes to zero. Hence, the discriminant rs
t+fi.=1,
D7=xi+yi. (8.3s6s-l)
qt = yr,

(r =0' the time is known. position is found as before. The distance of the observer
(8.3s62-r)
the axis is
The discriminant is A=m_p. (8.3565-2)
D z = i + y ?- t , (8.3562-2)
magnitude of the greatest eclipse for this case is
and the procedureis the sameas for the penumbra.
(8.3565-3)
E3563 Extreme points ofumbral and penunbral lnlt
lirnits The extreme points of the
limits of umbra and penumbra are found
by setting (1 = 0 and then calcu latingtane
from Equation 8.3422-4. LocalCircumstances

cosO=t(l proceeding, the reader should review Section 8.34 on conditional and vari-
+ t a n 2Q ) - t D ,
€quations. Local circumstances provide a description of eclipse phenomena
siaQ = cosQtalQ,
to a ffxed reference position, which is equivalent to fixing p as constant.
 462 EXPLANAToRY
SUPPLEMENT OFTHESUNANDMOON
I / ECLIPSES 463
8.361 Dclipse Maps The eclipse maps currently given in ?he Astronornical Al- 8.3622 Position angles The projected shadow is a reflection of what the observer
mq,ne,cs\ow the region over which various phases of each solar eclipse may be seel seesin the sky, so the position angle of contact, p, may be found from
and the times at which phases occur. Each map is a plot of the curves described
in Section 8.355, except that outline curves are limited to those of the penumbra | 1- Srn (-, I
tl L- c- o^ sl L , (8.3622-1)
every half hour. m=
t-l
I
The outline curves are divided into leading edge (in short dash) and trailing L0j
edge (in long dash). Except for certain extreme cases, the shadow outline moves
generally from west to east. For a given location, first contact (the beginning of The angle p is measured eastward from the north, on the solar Iimb as the observer
partial eclipse) occurs when the leading edge of the shadow arrivesl similarly, last sees it, or the projected shadow circle on the imaginary plane of reference. From
contact (the end of partial eclipse) occurs when the trailing edge arrives. First or the observer's point of view, however, the vertex (up) is not usually north. The
last contact may be estimated from the map to within a few minutes. parallactic angle C, defined by tanC = { / 4, is the measure of the difference. The
position angle V of the contact point from the vertex is

8.362 Precise Calculations There are many refinements to be considered in per- v=Q c. (8.3622,2)
forming precise calculations: choosing a consistent ephemeris, corrections for Earth's
rate of rotation, and, for the Moon, ofiset of center of figure from center of mass;
8.3623 Magnitude and obscuration Magnitude is commonly confused with obscu-
correcting for irregularities of the lunar limb; and effects of the Earth's atmosphere
ration. Magnitude is defined as the fraction of the solar diameter covered by the
(refraction in solar eclipses and the effect on the shadow in lunar eclipses).
Moon at the time of greatest phase and is expressedin units of the solar diameter.
Some elements (e.g., the radius of the shadow cone) used to describe an eclipse
Obscuration is the fraction of the surface of the solar disk obscured by the Moon.
remain nearly constant for the duration. Elements that change do so very nearly In the case of total eclipse, the magnitude can be greater than 1.0. To derive
linearly. In efiect, an observer seesan eclipse as two disks of fixed size, one crossing the expression, however, the annular case is shown in Figure 8.3623.1. An observer
the other in a straight line at constant speed. at point B in the penumbra seespart S'B' of the solar diameter blocked ofi, whereas
In conventional notation, define l, v such that: a,u observer at point E in the umbra sees the Moon's entire disk projected in DD'
on the sun. JJ .

'-o=f.-€l=f,l =- (8.362-1)
The known quantities are

L';'] l;] P'O = PO = Lt,

A'O=AO=Ia,
The distance from the observer to the nearest point of the shadow axis is lrnl = la.
BO=m. (8.3623-r)
Let the relative velocity be n = rh, and then the relative speed is lnl = n. The rate
of sepa.rationis m . n, and maximum eclipse occurs when the shadow axis is closest
t o t h e o b s e r v e ro, r m . n = 0 . At B, magnitude Mt = S'B' / SS'
= PB / PA' by simple proPortion

E.3621 Contact tim€s and duration At the beginning or end of penumbral, or par-
= (Po BO)t(PO+oA')
tial, phase, the condition is m = Lt. Similarly, if the chosen point is in the ceniral =(Lt-m)l(L1+L). (8.3623-2)
path, then at the beginning or end of umbral phase m= L2.
There are methods for finding the times of these phenomena by successiveiter- At E, magnitude M2 = DD' I S S
ation, using approximations and auxiliary angles. However, it is simpler to tabulate = (s/D- s/D/)/ s/s
a, v, Z (where I is either radius) as a function of time, and also the discriminant
= PE/PAI AE/PAI
u2+v2 -L2, When the discriminant goes to zero, inverse interpolation gives the time
of the local contact. Duration is the difierence between the two contact times for = PA/PA'
the umbra. =(Lr_L)/(L\+L). (8.3623:3)
 +oc

tial at B

D'B,D
Sun

-*tion of theareaof thesolardiskobscured bv theMoon'

[tff#f;t":'t:'*
CDCF to theareaof thesolardisk'
i"i" it it tn" .utioof tft" arcaof segment
Moon

side of the vertex' A and A'

.a total eclipse, in which the plane is on the other
is identical' Note also that this
,".,"rr"d, .o thtt OA = -h, and the expression
of the solar disk'
dia.rneter of the luna"r disk in units of the diameter ^- . r m . I n F i g -
usrng r'l' r-2' zuru
The degree of obscuration is now ca'lculated
of the solar disk'
ilurii, o is the center of the lunar disk, B is the center
j"--" tl:
disks overlap, with intersections at C, C The line AB "":':t: the"idll-
Let luna"r
th" and the chord CC at the indicated points
"lr"n-f"r"n""s 8C = | arld AC = M2 = s'
be s = AC. The sola"rradius is talen as unity; hence
= or DF = 2Mt. Filal'|y'
deffnition also, and in units of solar radius, M1 DF l 2,

(8.36234)
AB=l+s-DF'

of known quantities,then
Observer's
plane
BC=1,

Ftgure8.3623.1 AC=Mz="#=t,
L1+14
Magnitude is thefractionof thelineardiameter of thesuncoveredby theMoon.In thisschematic
annularegurye_in theobserver's plane,.A/A
is thezoneof aroJtJf io aootser"e,*ltf,in DF=2"-4
3f-ll LtlLa
Moon'scross-section
|iffl""*ry"j 'jf 4,rhe*er point MMl projectsontosS,^ dnl H"*"G ."g.iila" L' - I4 2(Lr - m)
=
2m (8.3623-s)
at Il in thezoneof partialeclipse''98' is obscured,
hencethe n=t+
i"fiJri"l; 3,i". ,ur- Lnt LnL,
 466 EXPLANATORY
SUPPLEI,4ENT / ECLIPSESOF THE SUN AND MOON 467
The area of the solar disk covered by the Moon is given by g353 Differential Corrections in Spaceand Time Differential corrections were in
preciselycalculatedlocal circumstancesto nearby loca-
fimes past usedto adjust
has reached such a stage of advancement
S = segment CFC + segment CDC , (8.3623_6) , tions. However, computing equipment
the whole calculation
that difierentials have fallen into disuse in favor of repeating
for the new site.
segment CFC = 2(sector ACF - triangle ACE), Calculations are normally provided in provisional Universal Time (UT) or, more
purposes
orccisely, UT1 (see Chapter 2) using a predicted value of AT. For record
segment CDC' = 2(sector BCD - triangle BCF). (8.3623-?) convenience of users
this is not necessarily the best practice, but for the it was
deemed significant enough to be necessary. If a later value of AT is adopted-such
Using angles in radians:
that the ofiset becomes AZ + 6Z then tabular quantities may be corrected quite
easily bY aPPlying
Area of sectorACF = sA,
Area of sectorBCD = B, 6I = - l. 002738df (longitude measured eastward) (8.363-l)
CE=ssinA=sinB,
AE = s cosA, to all longitudes in the tabulated phenomena, and subtracting 67 algebraically
ftom all tabulated times expressedin UT. Finally, interpolate the table back to the
EB = cosB. (8.3623-8)
original a.rguments.
For calculation of circumstances at elevations above the spheroid, the assumed
Substitutinginto Equations8.3623-7and then 8.3623-6:
radius of the Earth is increased accordingly and the calculations repeated'
/t ^ | ^
s e g m e n tC F C ' = 2 1 ; s ' A - - s r s i n A c o s A) .
\Z 8.4 LUNARECLIPSES
sesmenr
cDc'= 2 f+B ,ina.ora).
\z 2] / 8.41 lntroduction
Ji= (fA + B) (s2sinAcosA+ sinBcosB) The calculation of lunar eclipsesfollows the same principles as that of solar eclipses'
=1s2A+B)-ssinC. (8.3623-e) The fundamental plane is perpendicular to the axis of the shadow, and the origin
of coordinates is the shadow's axis. The z-coordinate is not used Since the observer
The area of the solar disk is z; hence the obscuration as a fraction of the disk is is on the body that is casting the shadow, the circumstances are the same for all
S'=S/2. parts of the Earth from which the Moon is visible.
The angles A, B, and C may be evaluated from fundamental rules of trigonom- The criteria for the occurrence of a lunar eclipse, and the definitions of the
etry. The working relations are: tJrpesthat may occur) are given in Section 8.23

c o s c =( a ?+ L Z - 2 m 2 ) t ( L 2Lt -4 ) , O< C < n , 8.42 Comoutations

s s s! = ( L 1 1 2+ m 2 1 l m ( L 1 + L ) , 018z-n, (8.3623-lo)
Equations 8.2321-1 and 8.233-1 define specific values of the separation of centers
A=r (B+ C). at which contacts occur (see Figure 8.2321.1). By conventional notation deriving
from solar eclipses, the angular distance between the centers of the Moon and the
If an observer wishes to construct a diagram showing the exact relationships of the shadow is designated L and at times of contact has values as follows:
disks during the course of an eclipse, it is useful to have calculated these quantit,ies
at a convenient tabular interval. Note that during central phase, all intersections at beginning and end of the penumbral eclipse, L1 = Jl + sm;
cease to exist and the quantities are undefined. During that phase, for an anntllar at beginning and end of the urnbral eclipse, L2= f2+ sfr;' (8.42-l)
eclipse I = s2; for a total eclipse, S7= 1. at beginning and end of the total eclipse, L3 = f2 - s6'
 468 EXPLANAToRY
suppLE[.,tENT OFTHESUNANDMOON
/ ECLIPSES 469
As is the case for solar eclipses, for dates when a lunar eclipse is to be calculated. The Besselianelementsfor a lunar eclipseare.r,y, i, j', rz, L1' L2' \, f 1.f 2, There
published in
apparent ephemeridesof the Sun and the Moon are generated at a suitable intervai. 11owidely accepted convention, however, and the elements are not
by subtabulation if necessary.The usual interr,al is 10 minutes. The ephemeris is
fhe A stro nom ical A lm an ac.
corrected for the offset of the center of figure from the center of mass, as appropriate
for the chosen ephemeris, and the argument is changed to UT by applying 4?. The contact tirnes can now be found by either of two meth-
8.42,2 Contact Times
ods.
8.421 BesselianElements As in Section 8.322 and Equation 8.322-1, the geocel-
One method is to choose a time 16 near opposition, with corresponding m6, rhe.
tric unit vector to the Sun, in an equatorial system, is 'For
a contact at time 7 = 7s + /, in which / is positive or negative and L is chosen
from Equation 8.42 1'
Icoso"cos4l
rsG -- I sincr.cos6. | . (8.421-I )
(mo+mot) (mo+rho,=L L' (8'422-l)
L 'i"0, J
It follows that the vector of the antisolar point, along the axis, is ru6 = r,6. This Equation 8.422-I may be solved as a quadratic equation in t; use the two roots
vector corr€sponds to g in Equation A322-2, with d = as + l2h, d = -6., for a to estimate the two contacts. The two times may then be used to start a second
geocentric system and eliminates the step of solving for (a,d), To transform the iteration, and convergence is very rapid.
Moon's position to a geocentric system in which the z-axis is parallel to the axis Alternatively, the following discriminants may be tabulated as a function ofl
of the shadow, apply the same rotations as given in Equation 8.322 3, but for. time:
units use secondsof arc. This calculation gives Equations 8.322-6, without the unit
conversion factor, and from which only (-r,y) in the fundamental plane are used. Lr-m,
To summarize: Lz-m,
- coso. cosd,'l (8.422-2)
fcosacosd l [
raG = s ' n " , c o s 6 ,. (8.421-2)
l s i ns rancd o sl =
d-",o= | -srnd5 I for a timespan starting before opposition such that initially m = I ' 511, and contin-
L J L l
uing until the first discriminant has passed through zero twice'
No approximations or manipulations are necessary At each time step, the dis-
c o s d ' s i n ( o -m4 ) criminants are examined for a change of sign from the preceding step; whcn a change
tT \ / ,r\ | l
rmr = Rt ( t dJn, (a+ r . c = s i n l . c 3 s { c o s d m s i n d c o s ( o- m
d) occurs, find the time of the zero by inverse interpolation. The casesthat may occur
1,) | |
L S r no ms r nd + c o s0 mc o sd c o so( m- d l l appear in 'lable 8,422.1, and must occur in the sequenceindicated
= l.x, 7 In special caseswhere examination indicates that one phase might have begun
y zl , (8.421-J)
and ended within a single tabular interval, subtabulation or the alternative method
where of polynomial iteration may be necessary.

d=-6* 8.423 Time of Great€st Obscuration The time of greatest obscuratiort occurs wben
the axis of the shadow is closest to the center of the Moon, or when m is a minimum:

In the fundamental plane or m rh=0. (8.423-r)

i.e., where z = O-the separation is *rr^'^r=o
n=(?+y2)i, \8.42t4) The quantity
m.rh=xi+$, (8.423-2)
Call the vector form m. The hourly variation, rh, is found by mrmerical difierenti-
ation or, if a polynomial representation is used, by difierentiating the polynomial. nav also be tabulated as a discriminant and the zero found by inverse interpolation.
 470 EXPLANATORYSUPPLE[,4ENT oF THESUNAND|\4OON
g 7 ECLIPSES 471

Table8.422.1 For a given position (do, Io) found from this, the horizon circle is the locus of points
Sequences andConditions
lor ContactTimes
on a great circle that satisfy the condition:
Condition
tan d = - cot do cos()0 )). (8.426-2)
No eclipse L 1- m < 0 No eclipse
Penumbral eclipsebegins L1 m=0 Penumbral eclipseends
Penumbral eclipse Lj m>OandL2-m<O Penumbral eclipse
Partialeclipsebegins L2 m=0andL3-m<0 Partialeclipseends Values of ) may be assumed and coordinate pairs calculated. The hemisphere de-
Partialeclipse \ m<O Parlialeclipse fi1ed by this central point and horizon circle is where the Moon is above the horizon
Totaleclipsebegins L3-m=O Totaleclipseends at the instant.
Totaleclipse L3-m>O Totaleclipse

Note:The evenis in the leftmostcolumnot the lable occur in sequencefrom the top down, and in the
dghlmoslcolumnof the lable, kom the bottomup.

8.5 TRANSITS

E,424 Magnitude The magnitude is the fraction ofthe Moon's diameter covered
by shadow, in units of the lunar diameter. As published, magnitude corresponds As mentioned in Section 8.1, a transit occurs when the shadow of one of the infe-
rior planets falls upon the Earth. Information concerning transits of Mercury and
only to the maximum obscuration, but for use in special calculations, it mav be
Venus is tabulated it The Astronomical Almenac. This information consists of the
calculated as a function of time for both umbra and penumbra.
Thp ewnrpceinn ic
elements of the transit (positions and motions of the Sun and inferior planet at
time of conjunction), geocentric phasesl and short formulas that are functions of an
Li-m
(8.424-1) observer's longitude and latitude to generate local circumsta,nces.Additionally, a
world map (Mercator projection) is provided from which a user can determine the
in which Z1 is used for penumbral magnitude and Z2 for umbral magnitude. NIag- region of visibility and approximate local circumstances.
nitude may be greater ihan unity, but not negative. Prior to and including the 1986 transit of Mercury, an extension of Newcomb's
(1882) heliocentric method was employed in the transit calculations for The As-
tronomical Almanac. The required heliocentric ephemeris of Mercury was based
8.425 Position Angl€s The position angle of contact P on the limb of the Moon is on Newcomb's Tables of Mercury, and the heliocentric ephemeris of the Earth was
the position angle M of m at the instant of contact, measured eastward (clockwise) derived from Newcomb's ?aDlesof the Sun. The corrections made to these ephemer-
from the north; i.e., ides and the actual method used to compute the transits is described in the original
Explanat ory Supplement.
taJrM = x ly In theory, the general methods used to compute solar eclipse phenomena may
be extended to predict the circumstances of a transit once the inferior planet has
P=M for interior contacts been substituted for the Moon. Chauvenet (1891) derived a special method, based
= M+ 180' (8.425-l) on improvements to Lagrange's method, which takes advantage of the small paral-
for exterior contacts.
laxes ofthe Sun and the transiting planet. This method requires apparent geocentric
ephemeridesof the Sun and planet as input. The apparent ephemerides can be read-
8.426 Sublunar Points The latitudes (S) and Iongitudes ()) of places that have ily computed from a modern solar system ephemeris, such as DE200, using methods
the Moon in the zenith at given times are given by described in Chapter 5. Chauvenet's method also allows for the generation of short
formulas, functions of longitude and latitude, from which local circumstances may
6=5^ be computed. The interested reader may use the referencesgiven in Section 8.6 for
) = Greenwich apparent sidereal time - o-. (8.426-l) firther details.
472 EXPLANATORYSUPPLEMENT / ECLIPSESOF THESUNANDMOON
473
8.6 REFERENCES Mikhailov,A'(1931)'..UberdieBerechnungderBreitederTotalitetszonebeiSorrnenfin.
stetnissen A 5I ron o m t srhes N aeh ric hl en 243 5 I'
of the Sun, fr'fth ed (Columbia Universitv Press' New York) '
Further information on calculations aIrd pledictions is often difficu]t to obtain. The fol_ Mitchell, S.A. (1951). Eclipses
This book describes observing expeditions in the first half of the twentieth century'
lowing is a guide to the most important mate al. (1983) Cdnon ol Solar Eclipses -2003 to +2526 (Astronomical
Mucke, H. and Meeus, J.
" eclipses, in the form of
Offi"", Vienna). This book contains elements of 10774 solar
Chauvenet, W. (1891). A Manual of Spheri.cal and, Practical Astronomg filth ed. (J.B. computer listiogs and small plotted charts The information is based on Ncwcomb's
Lippincott Co., Philadelphia) reprinted 1960 (Dover Publications, New York) I, pp. and the lLE, and is arranged for use by Bessel's method Correction for
solar theory
436 542 (solar eclipses);pp.542 549 (lunar eclipses);pp. 591 601 (transits of inuer is not aPPlied.
Af
planets). This reference contains an account of the formulas necessary for the pre- (Adam Hilger Ltd, Bristol' England)'
Muuay, C.A (1983). Vectorial Astrometry
diction of the stated phenomena for the Ealth generally, and for a particular place. S. (1882) "Discussion and results of observations on transits of \'lercury, ftorn
["*"o-U,
Various corrections derived from observations are also considered. The forrnulas ancl part VI (U S Naval Observatory, 1&'ashington)
1677 to i3a7" APAEI,
constants are suitable lor use with logarithms. An a.daptation for modern use, upo11 T.R.v. (18a7). Canon der Finstentisse (Imperial Academy of Sciencc, Vienna)
Oppolzer,
which much of the development in this chapter is based, is in the technical report by reDlirtted1962(DoverPublications,NewYork)withEnglishtranslation.Thisbook
Williams. -1207 November 10 (Julian proleptic
co;tains elements of 8000 solar eclipses from
Her Majesty's Nautical Almanac Office (196I). Ecplanatory Supplement to the Astronom- date) to 2161 November 17 (Cregorian date) and of 5200 lunar eclipses from -1206
ical Ephemeris and the American Ephemeris and. Nautical Alnanac (H.M. Statiorery. Aoril 2t to 2t63 O.tober 12. Maps show the central paths of total and annular eclipses
rnethod,
Office, London). Chapter 9 on eclipses and transits is the predecessor of the prese1t iLthe northern hemisphere Data are arranged for calculation by Hansen's
exposition. using logarithms, and have known errors'
Green, R. (1985). Spherical Astronomy (Cambridge University Press, Cambridge). This Ssrart, li.lt. (1977). TextBook on Spherical Astronomg sixth ed' (Cambridge University
book and the book by Murray are extremely helpful in presenting spherical trigonorn- Press, Cambridge).
National
etry in vector notation. Williams, W. Jr. (197i) Prediction and Analgsis oJ Solar Eclipse Circumslonccs
Herald, D. (1983). "Correcting Predictions of Solar Eclipse Contact Times fol the Ef- TechnicallnformationServiceTechnicalReportNTlsNo'AD726626(Springfield'
fects of Lunar Limb Irregularities" J. Bri.t. Astron. ,4ssoc. 93, 241 246. This paper Virginia).
gives precepts for constructing charts to correct solar eclipse contact times for the
efects of lunar limb irregularities. A useful discussion estimates the total effect and
its components.
Link, F. (1969). Eclipse Phenomena in Astronomg (Springer-Verlag, New York). This
book contains a wealth of information on lunar eclipses and transits; the ilfornra-
tion is available almost nowhere else. The volume presents primarily physir:al and
atmosphedc effects ol the eclipsing body, but also includes computational and histor-
ical information. It also treats eclipse and occultation efects involving other planets,
natural and artificial satellites, radio wavelengths, and relativistic efects.
Liu Bao-Lin (1983). "Canon of Lunar Eclipses ftom 1000 B.c. to A.D. 3OOO"Pub. Pur\tle
Mountain Obs. 2, 1, This volume contains data on 9800 lunar eclipses, including
penumbral. It contains the data in a different form from that of Meeus and luucke.
and uses conventional augmentation of the Earth's shadow. The primary use ofcanons
is for chronological and historical research, but they are also useful for planning and
for seeking long-term cyclic features.
Meeus, J., Grosjean, C., and Vanderleen, W. (1966). Canon of Solar Eclipses (Pergamon
Press, Oxford). In the style of Oppolzer, but with improved calculations, this canon
covers only solar eclipses from 1898 to 2510. An introduction contains the theory,
general description, and a discussion ol eclipse cycles. Maps show the central tracks
over the entire Earth.
Meeus, J. arrd Mucke, H. (1983\. Canon of Lunar Eclipses, -2002 to 2526 second ed. (As'
tronomical Ofrce, Vienna). This book contains data on 10936 lunar eclipses, including
penumbral. The data are in the form of computer listings and plotted reference charts
The data are in a diferent form from those of Liu, using Danjon's augmentation of
the Earth's ra.dius. Correction for zI7 is not applied.
 CHAPTER

Astronomical
Phenomena
by B.D.Yallopand C.Y.Hohenkerx

NIGHTSKY
9.1 GENERALASPECTSOFTHE

The configurations and phenomena of the Sun, Moon, and planets as seen from
the Earth arise from their apparent movement in the sky. Most of the phenomena
published in the "Dia.ry of Phenomena" for example, are geocentricl the remainder
are heliocentric.
' The Earth rotates about its polar axis once a day and produces an apparent
, motion of the night sky about the celestial poles. Iu the Northern Hemisphere the
north celestial pole is elelated above the horizon. Facing away from the elelated
celestial pole, an observer seesthe Sun, Moon, planets, and sta,rsrise in the east and
set in the west. They reach their highest altitude as they cross the local meridian.
When the observer turns to face the elevated celestial pole, stars nearest the pole
neither rise nor set. They become circumpolar and cross the meridian each day
once above the pole at their highest altitude and once below the pole at their
lowest altitude. In the Northern Hemisphere, circumpolar stars appeax to rotate
obout the north celestial pole anticlockwise. In the Southern Hemisphere the effect
is reversed and they appeax to rotate clockwise.
Superimposed on the diurnal rotation is au annual rotation caused by the
Earth's orbiting the Sun. Since the stars a.re seen by the naked eye after sunset, the
appear to move from east to west, and to return to the same position
after a yea,r, Relative to the Sun, the stars rise and set roughly four minutes ea,rlier
bach day. In the course of a month, the night sky appears to move two hours in
right ascension to the west. Also because of thjs orbital motion of the Earth, the
circumpolar staxs in the Northern Hemisphere appear to rotate once a yeax in an
dnticlockwise direction around the north celestial pole and in a clockwise direction
about the south celestial pole.
 476 EXPLANATORYSUPPLE[,,IENT 9 / A S T R O N O M I C APLH E N O M E N A 477
Ihe Nloon movesin an orbit inclincd to the ecliptic b1-6.; the Nloon nrakesorLl Thc time t of thc phenomenotr derives from a frtnctional cxprcssiotl involving
revolution about the sky from u.est to east in about a nronth. During this preriorl the paranctcr-s associated t'iih the phcnorrtctton. The iitne of the Phcnourerrol is
the phases of the N4oo[ corrrplete a cyclc from neu,' to full ancl back to ne$.. Tl]{, calcnlirted from thc cquation l(r) = 0 o. f'(/) = 0, rvhere f'(0 = df(J) I dt.
orbit of the Moon is moving around the ecliptic, so that other aspectsof the \loou s
position in thc sky, such as its maxinlum and rnirrimum declinatiorr, change fi.o1r
one moDth to the next. 9.21 InterestingPhenomena of the Sun, Earth, and Moon
It is important to know when thc planets are in thc rnost favorable position firr.
obserr.ation.Thc outer planets,for example.are best seenarounri opposilion. Tho.r. 9,211 Equinoxes and Solstices The times of thc equinoxes and solstict's aro clt:
are in their most unfavorable position around coniunction. The inner planets a1c, flned to bc u'hen the Sun's apparent ccliptic longitude ,\5 is a nnltiplc of 90'; i.t:.,
- 0", 90'. 180". or 270'. 'fhus irr
different thev are in their most favorablc position near grcatest elongatiou. creu it is calculatcd from /(t) = 0, where.f(l) = ls
though they are not at full phase.At superior conjunction the phaseis around frrll. the northern hcruisphere,for the spring cqllillox.f(/) = Is. for the sunmer solstice
but the pianets are dificult to sec because they arc furthest from the Eurrth anrl
= ls 90'. lbr 1he autumrr eqrrinox/(t) = 15 180' and lbr the $'illter solstice
"f(r)
usually too closeto the Sru. At inferior co[julction the inner p]atretsare ncalcst f(t) = ^s 27O'. At thr: equinoxcsthe Sun crosscsthe etluator'"vhcrrthe length o[1hc
to the Earth, but again they are difficult to scc becausetheir phase is small. autL day exccedsthe length of the niglrt due to rcfritctiorr,semidiarneter',and parallax
they are too closeto thc Sun. of the Surr. At that tinrc the lengths of the cla,van<l light are apploxirnzitell' erlua1
Often the times of phcnomena need not have an\- great precision: sometinLcstlre every\\''hefc.
nearest hour or el'en thc nearest clay are sullicient for observatiorral purposes. Tlrl The tinre of the colltmencemcrrtof the seasollssho\\'s a progrcssi\1:cllarlge
dates and times, lroil'cvcr,usually depcnd on the coordinate systen. For hisl or.icirl becauseof thc ieap-vearc1cle.Becauscthe period of revolution rl1t h(r Flarth about
reasons the conjunctions and oppositious of planets har,'ealn'ays been calculal <:clin the Sun is not connensurate l'ith tht: Gregorian calendal J-ciu. it is onl.v-aftcr rr
geocentricecliptic coordinates.On thc other hancl,the conjunctionsof plalets n irlr complete c]'cle of four ccntllries thal the seasonsagitin com[l(llce at approx] Iatell'
other planets, bright stars, or the \Ioon har.e ah'ays beerr r:alculated using eqlial o the same tirnes. In the presentcentur]'the latest clatesfor thc seasonso((urred iD
rial coordinatesl the phenomena arc then observed more easily with an equatoriallr. 1903, and the earliest rvill be in 2000; by the tcar 2096 thc seasonsr'vill begirr at
rrrounted telescope. In sorne casesthc times of phenomena have been defined as thc their earlicst possibleiirres (seeTaLrle9.211.1).
maxima or minirrra of the .listarrccs fr.om the Sun or thc Ear.th or the elongatiorr
from another bodl'. In such cases.the phenornena are indcpcndent of the coordirato Table9.211.1
of the Seasons
Timeof Commencement
system.
Spring Summer Aulumn
9.2 CONFIGURATIONS OF THE SUN, N4OON,AND PLANETS

Latest 1903 M a r c h2 1 1 9 Jvne2215 Sept.24 06 Dec.23 00

The Universai Times (UT) of the principal astr.onomical phenomena invoh-ing con 2114
2000 20 08 21 02 2217
hgurations of the Suu, NIooD, and plancts are given in 7[e Astrono.n.ical Ahtrut.ttt: Earliesl 2 0 s 6 1914 20 07 21 23 20 21
pagesA1 A11 under tho heading "Pherromena".In rnost casesthe tinles alc ljivdr
to the lrearest hour. but for certaiu heliocentric phenonena of tlrc plallets. onlt Noie: The tota range in limes is aboul 54 hours in each case

the datc appoars. The times of the phenomena of greatest general i[terest t]rc
beginning of the seasonsand the pha,sesof the N{oon are to the uealest rnirlutc.
although the accuracy is of no significarrce. 9.212 Perihelion and Aph€lion of the Earth In lhe "Diary of Phr:nornerra"' the
Times may be calculated using a daily tabular ephemeris. For quantitics that times rvhen the Earth is at perihcliol (the Sun is at perigrre)are defined to be
vary slowly with time, such as planetary apsiclcs,a tabular inten'al of five dir,ysgives those for which the Sun's geonretricdistalce R is a rninimurn Likervisc,thc tirtres
better results. On the other hall<l,quaDtitiessuch as the phascsof thc N,loonvarl when thc Earth is at aplrtlion (the Surr is at apogee) are defined to be t'hen R is
rapidly, and a sholter interval of a half or quarter day is rnore appropriate. If t hc a maxirnum. Thus if r is thc time of periheliorr,it is calculated from the t:quation
epherneris is in TDT, then, to conr.crt to UT, add the correction AZ = UT TD I ./,(r) = 0 *l,ere .f(r) = R. The times do not al$,ays agree rvith those calculatcd lrorn
to the time obtained at the cnd of the calculation. the times whcn the Sun is at perigeeor apogeein its mean elliptical orbit (i e , the
 479
PHENOMENA
EXPLANATORY
SUPPLEMENT 9 / ASTRONOMICAL
479
Table 9.22.1
time t when the mean longitude of perigee | = 0. or apogee f - 180" = 0) be<ausc =
GeocentricPhenomenafor which J(f) 0
of perturbations by the planets.
Phenomenon fo Remarks
9,213 Phasesofthe Moon The times of the phases of the Moon are tabulated to
the nearest minute of UT in The Astronomical Almanac pages 41 and D1 anri ar1 Seasons:
Marchequinox,FirstPointol Aries
VernalEquinox ls
given to the nearest hour of UT in the "Dia.ry" orl pages A9 A11. They arc thc )s - 90' Junesolstice
SummerSolstice
times when the excessof the Moon's apparent geocentric ecliptic longitude )r/ o\tr Is - 180' September equinox
AutumnalEquinox
^s 27o' December solslice
the Sun's apparent geocentricecliptic longitude is 0", 90', 180", or 270' i.c., tl1. WinterSolstice
t i m e s w h e n " f ( r ) = 0 w h e r ef ( t ) = ^ M - . \ 5 f o r n e w M o o n . / O = ) a z - , \ s - 9 0 " for Conjunctions ot the planetswiththe Sun:
first quarter, f(t) = \u - )s - 180' for full Moon and /(/) = )y - A5 - 270' fol last Inferiorplaners:
quarter. Because the times are determined from geocentric qoordinates, thc]' ar6 interiorconjunction lp - ^s ,\p - )s changesfromPlusto minus
superiorconjunction trp- trs )p - )s changesfromminusto Plus
independent of location on the Earth.
Owing to the rapid variations in the distance and velocity of the Moon, the SuperiorPlanets:
coniunction p - ^s
intervals between successivephases are not constant: and it is not possible to <heck opposition )p - )s - 180"
these times by differencing. Moreover, there is no simple prediction formula u'ith
Phasesof the Moon:
which to make a comparison. Examination of the higher difierences of the successive NewMoon AM ^s
times of the same phenomenon provides a check. FirstOuarter ,\6 - '\5 - 9o'
The phases of the Moon do not recur on exactly the same dates in an1' regular Fulllvloon IM - )s - 180"
LastQuarter )M - )s - 27oo
cycle, but the approximate dates of the phases in any year can be found froln
the dates on which the phase occurred 19 years previously. Thus in the Metonic ol otherbodiesl
Conjunctions
planetwithPlanet dh - dP2 separation= 6p1-6p,
cycle, in which 19 tropical years are nearly equal to 235 synodic months (nell' ntool pfanetwithbrighlstar dP db separation=6p-6b
to new moon), the phases recur on dates that are the same or difler by ole or Moon\^/ithPlanet dM - dP separation=6M-6p
occasionally two days, depending on the number of intervening leap years and oIt Moonwithbrightstar dM - db separation=6M-5b

the perturbations of the Moon. For example, during 2000 the dates are the same
as in 1981 on thirty occasions, and differ by one day for the remaining ninetccn.
Another relevant cycle is the Saros, which consists of 233 synodic months antl
equals nearly 19 passagesof the Sun through the node of the N'Ioon's orbitl not Table 9.22.2
only will the moon phases recur but eclipses as well. Moreover in this cyck: lhc GeocentricPhenomenafor which/'(t) = 0
Moon's apse makes 239 revolutions and returns to nearly the same position. so tht) t(0 Bemarks
Body Phenomenon
durations of solar eclipses are similar as u'ell.
-28'<E<+28'
Mercury Greatestelongation -47'<E<+47'
Venus
9.22 GeocentricPhenomena (t+A+R)\r+A-n O.39AU<j<0.47AU
Venus Greatestbrilliancy
The times of geocentric phenomena are calculated from the expressions for /(1) At perihelionr is a minlmum
Earth Apsides
At aphelion r is a maxlmum
listed below. Table 9.22.I contairs the phenomena for which the timc is obtained
At perigee P is a mlnlmum
from the equation.f(r) = 0. Table 9.22.2 contains phenomena for which the tirne is Moon Apsides
At apogee P is a maxmum
obtained from the equation -f'(l) = 0. In this table, except for the stationary points At perigee 7ris a maxlmum
in right ascension, the phenomena are independent of the coordinate system. Al apogee r is a mlnlmum
A
The notation ,\, o, 6 refers to the apparent geocentric longitude, right ascension, Mars ClosestapProacn
Planet in
Slationary
and declination; and the subscripts S, M, p, b refer to the Sun, Moon, planet, or rightascension
bright star, respectively. The distances r, R, A, and p refer to the geometric distances
of the Sun-planet, Earth-Sun, Earth-planet, and Ea,rth Moon, respectivelyl r is
480 ExPLANAToRY
suPPLETTENT PHENOMENA
9 / ASTRONOMICAL 481

the horizontal pa.ralla-xof the Moon. The five bright stars used in the "Dia.ry" are Table 9.222,2
Aldebaran, Polhn, Regulus, Spica, and Antares. Long-PeriodCyclesol Mercury,Venus,andMars
The following formulas are useful for calculating elongation and longitude:
ForMercury 54siderealperiods 13years + 2 days
lSTsiderealperiods 33years 1 day
cosE = sin dpsin 6" + cos 6ocos 6" cos(oo - o"), (9.22-1) ForVenus lSsiderealperiods 8 years 1 oay
periods
359sidereal 243years I dav
sln € sm 0 + cos €cos 0 srno 8siderealperiods 15years
um.{= (9.221) ForMars + 17days
cos 6 cos o 17siderealperiods 32years I days
25siderealperiods 47yearc + Sdays
42siderealperiods 79years I day

vrhere e is the true obliquity of the ecliptic.

9.221 Visibility of Planets Table 9.221.1 shows the criteria adopted for the "Di-
a.ry" for the minimum elongations lrom the Sun at which Mercury, the bodies used the four brightest minor planets. The periods may be used for a systematic sea,rch
in navigation, and the minor planets can be seen with the naked eye. for geocentric phenomena.
The synodic period Z is calculated from the mean motion of the planet, zo, and
Table9.221.1 the mean motion of the Earth, lr". Thus,
VisibilituCriteriafor Geocentric
Phenomena
360' 360'
T= for inner planets, and 1 = - Jor outel planels. \y zzz-l)
Body MlnlmumElongation nP-n" n,-nn

Moon,lvlars,Satum 15' For the slowly moving planets (Jupiter to Pluto) and the minor planets, the
Minorplanels 15"
Venus,Jupiter 10" error in using these values is small, but for Mercury, Venus, and \4ars, the mean
Mercury 10o+visualmagnitude
of Mercury synodic period is Iess useful. Much more accurate estimates of when these planets
may be seen again in the same part of the sky may be made by using long-period cy-
cles that contain, with varying degreesof accuracy, integral numbers of revolutions
of the Earth and the planet. (See'Iable L222.2.)
9.222 Synodic Periods Once the date and time of a geocentric phenornenon are Any particular phenomenon of a planet repeats itself after each cycle at the
knowl, the approximate time of the next similar phenomenon may be estimated by same time of year and in the same part of the sky; However, repetition does not
adding the synodic period 7 of the body concerned. The eccentricities of the orbits occur after a single synodic period. For Venus a useful and accurate form of the
of the Eaxth and the body, and the perturbations by other bodies, will introduce relation is
errors in this estimate. Table 9.222.7 gives synodic periods ofthe major planets and
to'>1)_)\
5 mean synodic periods = 8 calendar yearc - 214,

unless the interval includes a century year that is not a leap year.
Table 9.222.1
SynodicPeriodsof the Planetsand FirstAsteroids
9.23 HeliocentricPhenomena
Days Days Asteroid Days
Certain heliocentric phenomena of the planets are given in The Astronomical Al-
Mercury 1 1 6 Saturn 374 Ceres 467
nqnac, page A3. The dates of perihelion and aphelion are those on which the
Venus 584 uranus 370 Pallas 466
Mars 7AO Neptune 367 Juno 474 geometric distance of the planet from the Sun is a minimum and maximum re-
Jupiter 399 Pluto 367 Vesta 504 spectively; the dates are thus the times when the first derivative of the distance is
zero-i.e., ft(t) = drldt = 0. Owing to the presence of perturbations in the planeta'ry
 482 EXPLANATORY
SUPPLEMENT / ASTRONOMICALPHENOMENA
483
Table 9.23.1 Air Nauigation. Within their range of declination (0'-29")' these tables may
HeliocentricPhenomena be used to find the times of rising and setting of staxs.
Section 9.33 gives algorithms that may ea,sily be programmed into a small
Phenomenon fltl Type Remarks
or computer to calculate the times of the phenomena.
J(t) = 0 At thedescending node,b is decreasing.
At the ascending node,b is increasing.
Apsides
Sunset,andTwilight
.31 Sunrise,
f'(t) =0 Ataphelion, risa maximum.
At perihelion,
r is a minimum.
data given in The Astronomical Almanac enable the times of sunrise, sulset,
Greatestlatitudes .f'(a)= 0 At greatestlatitudeNorth,b is positive.
the beginning and end of civil, nautical, and astronomical twilight fo be found
At greatestlatitudeSouth,b is negative.
any position between latitudes 66" north and 55o south. The times, tabulated
every fourth day, a.re the local mean times of the phenomena on the meridian
Greenwich and in the specified latitude; interpolation is necessary to obtain the
mean times for intermediate latitudes, for intermediate days, and for longitude'
motion, these dates may differ from those obtained from the angular elements of the a,rraccuracv of about five minutes this interpolation can generally be done by
mean orbits. The actual disturbed motion of the planets is also used to determine large changes
; neax limiting conditions, when interpolation becomes difficult,
the dates when they pass through the nodes on the ecliptic, and rvhen thev reach time correspond to only small changes in depression and accurate times have
greatest latitudes north or south. At the nodes, the heliocentric latitude is zero; i.e., real meaning.
f(t) = b = 0. The value of /(r) changes from negative to positive at the ascending Interpolation for latitude is nonlinear. Interpolation for longitude, which is
node, and from positive to negative at the descending node. These dates are given iustified. can be combined with the interpolation for date by increasing for
each year for Mercury, Venus, and Mars, but they occur less frequently for the west longitudes, or decreasing for east longitudes, the Greenwich date by the ap-
other planets, and in these casesare given as additional notes when aecessary.The propriate fraction (longitude in degrees/360); for sunrise and sunset, the error due
dates on which a planet has its greatest latitude north or south is determined as to neglecting the variation with longitude anounts to a maximum of tvr'o minutes
the times at which the first derivative ofthe Iatitude is zero i.e.)ft(t)=db/dt=0. in latitudes up to 60". The times so obtained are local mean times, which can
Table 9.23.1 shows /(r) for the various phenomena. be converted to universal time by applying longitude in time adding if west and
subtractine if east. Standard times are obtained by adding (subtracting) to the
local time four minutes for every degree of longitude west (east) of the standard
9.3 RISINGS,
SETTINGS,
ANDTWILIGHT meridian.
At the tabulated times of sunrise and sunset, the geocentric zenith distance
-50', of which 34'
The astronomer is concerned with the phenomena of rising, setting, and twilight of the center of the Sun is 90'50' ; i e., a geocentric altitude of
primarily in regard to the planning of observations. Precision ofbetter than a minute is allowed for horizontal refraction and 16'for semidiameter; the Sun's apparent
of
of time is not required for this purpose, and extensive tables of these phenomena upper limb is thus on the horizon. Corrections axe necessaryif some other value
value of -34')
a.rein ?he Astronomi,cal Almanac, pdges A74 AZZ. the altitude is required, (such as for the conventional meteorological
The tabulated times of the phenomena refer to sea level vrith a clear horizon and or to allow for the height of the observer and the elevation of the actual horizon'
an adopted correction for refraction under normal meteorological conditions. The In such casesuse Equation 9.31-1.
of
actual times of rising and setting may difier considerably, especially near extreme At the times given for civil, nautical, and astronomical twilight, the altitude
will
conditions when the altitude is changing slowly. This difierence can reach half a the center of the Sun is -6', -12' and -18' respectively' These tabulations
minute of time at midlatitudes and more at hieh latitudes. The illumination at produce times for any desired altitude between -50' and -18''
the beginning or end of twilight also varies greatly with meteorological conditions. The Nuutical Almanac and The Air Almanac tabulate the times of sunrise'
Precise times have little real significance, except in special circumstances, such as sunset, and the beginning and end of civil twilight (altitude -6") every three days
end
navigation at sea. for latitudes between 72" north and 60" south; times of the beginning and
No data are given for the times of rising and setting of planets. Data may be of nautical twilight (altitude -12") are also in ?he Nautical Almanac' The Air
of sunrise or sunset'
obtained fairly simply from navigation tables such as lhe Si.ght Reduction Tables Ahnanac provides graphs that enable conections to the time
484 EXPLANAToRY
SUPPLEN4ENT / ASTRONOMICALPHENOMENA z+t'5

These corrections give the times at which the Sun has altitudes between 0o and
-12' (that is, depressionsto l2o).
Sunrise or sunset at a height Il meters above the level of the horizon ocq115
when the Sun's altitude is approximately

- 50' - 2'.12JH (ll17 for Il in feet), ( 9 . 3 1 -)l

so that the same table gives corrections for height to the times of sunrise and sunset.
In The Air Almanoc the corrections can be obtained from graphs. The constant 2. l2
is really a slowly decreasing function of Fl, but below a height of about 5000 m it js
constant; it decreasesto a value of 2.00 at 100000m.
Times of rising and setting and associated phenomena change rapidlv fronl .lo. I Declinotion

day to day in polar regions or may not occur for long periods, the Sun being
continuously above or below the horizon; accurate times are therefore difficult to
tabulate. Diagrams are given in The Air Almanac that enable approximate times
that are sufficiently accurate for all practical purposes to be obtained.

9.311 A low Precision Ephemeris for the Sun The following algorithm gir-esthe Figure9.311.1
equation of time (E), the Greenwich hour angle (GHA), declination (6), and seniidi- Theanalemmic curve
ameter (SD) of the Sun, in degrees, to a precision of better than ll0. which cau
be used with the algorithm for rising and setting (see Section 9.33). The equation
of time is also useful for erecting sundials and determining the transit tirne of the (3) The equation of time E, the GHA, 6, and .9D are given by
Sun.
sin2G+ 2?466sin2.\ - 03053sin4),
E= -13915sinG O?020
(1) Using the Julian date, JD, and the universal time, UT, in hours, calculate GHA = 15UT 180'+ 8,
I, the number of centuries from J2000:
6 = sin-I (sinr sin,\),

T = (JD + UT I 24 - 2451545.0\/ 36525. (9.31t-t) S D= 0 : 2 6 ' lt ( 1 0.0l7cosG). (9.31l-3)

Plotting the position ofthe apparent Sun relative to the mean Sun (which moves
(2) Calculate the Solar argumentsl the mean longitude corrected for aberration, uniformly along the equator) produces the analemmic curve, seeFigure 9.311.1. The
L; the mean anomaly, G; the ecliptic longitude, ); and the obliquity of the displacement of the apparent Sun in longitude is given by the equation of time (E)
ecliptic. e: and the displacement in latitude is given by the declination (6). The progress of
the apparent Sun throughout the year is indicated at the beginning of each month
a = 2803460+ 36000?'7'70T, removemultiples of 360o, on the closed curve' The analemmic curve Dlavs an important role in certain types
of sundial.
G = 357?528+35999:050r,
) = Z+ 11915
sinG + 03020sin2G, 9.32 Moonriseand Moonset
e = 2 3 ? 4 3 9-3 0 : 0 1 3 0 0 I . e3rt-2) tables in The Astronomical Almanac give, for every day and for a range of lat-
itudes from 66o north to 55" south, the local mean times of moontise and moonset
 496 EXPLANATORY
SUPPLEMENT PHENOMENA
9 / ASTBONOMICAL 447

for the meridian of Greenwich. Interpolation for latitude and longitude is necessary ephemerisof the Sun (Section 9.311)may be used. For the Moon, the low-precision
to obtain local times for other places, In practice, tines are rarely required more formulas on page D46 ol The Astronomical Almqnac may be used.
accurately than to within about five minutes, and interpolation can be done me1- At sunrise and sunset the apparent altitude of ihe upper Iimb on the horizon
-5O I 6O - 0. 0353y'H
tally. Formal interpolation, using tables such as are given in ?ie Nautical Almanac is zero and hence the adopted true altitude in degrees is h =
and The Air Almanac, yields times accurate to about two minutes. The times so where 11 is the height, in meters, of the observer above the horizon.
obtained may be converted to universal time or standard time by applying lotgi- Similarly at moonrise and moonset the apparent altitude of the upper limb on
-0.0353vE.
tude in time-adding if west and subtracting if east. Times for latitudes between the horizon is zero and the adopted true altitude i = -34160+0.'72'75n
72o north and 60o south are given in The Nautical ALmanac and. The Air Almano,c. At rise and set of a star or a planet, the apparent altitude of the body is zero; hence,
In The Air Almanac times appear in graphical form above 72o north, right up to h = -341 60 - O.O353vq.
the North Pole. Other important casesto consider are twilights: civil twilight /r = 6". nautical
In calculating the times of moonrise and moonset, the true altitude l, of the twilight h = 12', and astronomicaltwilight ft = -18".
center of the Moon is The time of rise or set, UT, in hours, is found by solving iteratively the equation

h = -34' - semidiameter + horizontal parallax (9.32-r) UT=UTo-(GHA+)tt)/15, (9.33-l)

where 34' is allowed for horizontal refraction; depending upon the distance of the where the plus sign is used for rise and the minus sign for set, and r is the hour
Moon, A lies in the range 5' to I l'. At these times, the upper limb of the Moon is angle of the body at UTe, which is given by
on the horizon; no allowance is made for phase. At a height of F1 meters above thc
horizon, the altitude of the Moon when its upper limb is on the horizon is decreased sinh-sin@sind
cost= (9.33-2)
by 2'.l2tE (seeSection 9.331). cos { cos 6
The Moon revolves round the Earth and makes one conplete revolution relativc
to the Sun in a synodic month of mean length 29.53 days; in that time the \Ioorr and / is the latitude and ) the longitude (positive to the east).
therefore appears to lose one transit across any meridian and, in general, one rising
If cos, > +l,set , = 0";
and one setting. During each month there is therefore no moonrise on one local da1' (9.33_3)
(near last quarter) and no moonset on one local day (near first quarter). In higl If cos, < -1.set t = 180".
latitudes the times of the phenomena change rapidly from day to day and may not
occut for long periods, the Moon being continuously above or below the horizon; As an initial guess, set UTo = t2h, although any value in the range 0h to 24h will
in these extreme conditions the times of moonrise and moonset sometimes decrease do. After each iteration, add multiples of 24h to set UT in the range 0h to 24 '
from day to day, instead of the usual increase in lower latitudes, and it is possible Replace UTs by UT until the difference between them is less than 0!008 If for
to have two moonrises or two moonsets durins the samc local dav several iterations cosr > I, it is likely that there is no phenomenon and the body
remains above the true altitude ft all day. On the other hand, if cosI < l, the body
9.33 FormulasAssociated remains below the true altitude all day. In each lunation, around first quarter, there
with Risingand Setting
is always a day when the Moon does not set, and another, around la'st quarter, when
Consider first an algorithm for calculating the times of rising or setting of a bodl' the Moon does not rise.
at true altitude ft. At latitudes above 60' the algorithm may fail, and it is necessaryto use a more
The GHA and declination (6) of the body are required to about one minute of systematic approach. If fto is the adopted true altitude, the times when lr = l?owhere
arc as a function of UT. For the Moon, the horizontal parallax n is also required.
To speed up the calculation, which is iterative, it is convenient to express GHA, i, + l)
s i n l = s i n d s i n d+ c o s d c o s 6c o s ( G H A (9.334)
and rr as daily polynomials. For the Moon second-order polynomials are required for
GHA and 6, and first order for rr. For other bodies like the Sun and the inner planets, are t he r imes o[ rising or sctting.
first-order polynomials are sufficient. For stars and outer planets, a constant daily These roots, if they exist, will lie between the true altitude at 0h and 24" and
va.luefor declination may be adopted. Alternatively low-precision formulas for the the maximum and minimum altitudes during the day. The maximum and minimum
4gg ExpLANAToRy
SUPPLEMENT PHENoMENA
O/ ASTRoNOMICAL 489

altitudes occur at or very near upper and lower transit. These times are found bv Reftaction, the bending of light as it passes through the atmosphere, is illus-
setting t = 0 and t = 180 in Equation 9.33 1 and iterating. trated in Figure 9.331.1 for the rising or setting of a celestial object. The follow-
Ouly for the Moon at high latitudes is it necessary to calculate the maximq11 ing approximate formulas (Bennett, 1982), are valid for all altitudes and standard
and minimum altitudes more precisely. This calculation is done by fitting a second- aleteorological conditions. More precise algorithms are given in Section 3.28. The
order polynomial to the true altitude at points around upper or lower transit a11d amount of refraction, R, is given bY
differentiating with respect to time to find the turning point.
0?0167
R(h) = (9.331-l)
| ( h "+ 4 . 4 ) ) '
tan(h"+7.31
9331 The Effects ofDip and Refraction Dip is required for reducing sights made
with a ma.rine sextant and for calculating the true altitude of a body at rise ancl where l, is the apparent altitude. AlternativelS if h is the true altitude. then
set. Dip is the apparent angle between the geometric horizon (at right angles to the
vertical) and the visible horizon, For an observer at height l/ above the surface of 090167
R(l) = (e.33r-2)
the Earth, dip D(I/) = l'.75\/8, where f1 is in meters (O'.97vE rt H is in feet) (see tan(h+8.6/(h+4.4))
Woolard and Clemence, 1966, page 215).
A navigator using a marine sextant sets on the apparent horizon and wants 9.332 Time for a Specified Shadow Length The lengths of shadows depend upon
to know how much the horizon appea,rsto be below the horizontal and thercfore the apparent altitude of the Sun, llo. The true altitude /l of the Srrn is needed to
only needs to know the dip. On the other hand, when computing the times of risc find the required time where
or set of a body, we need to know the true altitude when the body appears to be h = h. R\h.) (9.332-l)
on the horizon, which involves dip and the refraction between the horizon and the
observerR(lI), where R(/t = 0:37t/E and H i5 in meters (or O'.20r[Eif H is irr feetl. and R is the refraction at apparent altitude h. r'i'hich may be calculated approxi-
In Figure 9.331.1, O is the observer at height F1above the surface of the Earth. mately using Equation 9.331 1. Alternatively, if l is known and ho is required use
S is the body at rise or set, and A is the observer's horizon. The true altitude of S Equation 9.331-2 for calculating the refraction at true altitude i anrl set
from the vantage point A is -34' due to refraction. The true altitude of S from O is
the true altitude at A - D(H) - R(I{) = -34' - l'.'75\/E - O'.3'l'/H = -34' - 2'.l2yE . h,=h+R(h). ra ?1r-r)

9.333 Rates of Change The rate of change of hour angle with altitude may be
obtained by difierentiating ihe altitude as shown in Equation 9.33 4. Assuming
that ), d, and 6 do not change with time, then the change in time, (Ar)' in hours'
for a given change in altitude (Ai), in degrees, can be calcrrlated from the simple
exDression
ah, (9.333-l)
A LHA cosd sinZ

where ALHA is the rate of change of hour angle per hour and Z is the azimuth'
For the Sun, A LHA = l5o per hour and for the Nloon, 4 LHA = 149493per hour'
Near rise and set when l = 0". this exDression can be further simplified to

Lt=ALh,
(9.333-2)
where A = 60 / (a LHA Gtd -.1"26-
),

Figure 9,331.1 and At is now in minutes of time. This correction is accurate to within a few minutes
Thehorizonat risingor setting provided A does not exceed 2O and Ah does not exceed a few degrees'
 490 ExPLANAToRY
SUPPLEMENT 9 / ASTRONOMICALPHENOMENA 491
9.334 fime for a SpecifiedAzimuth Another interestingproblem is to calculate
the time when the azimuth Z of a body ta.kesa particular value. The time UT is
calculatediterativelv from
lo*'
UT=UTo-(GHA+)trl15, (9.33+-17

where the plus sign is used if Z < 180' and the minus sign is used otherwise. / is
calculated from
-abty/li'-bz lo'2
cost= (9.33+-21
G7-,
where a = tanZsin$, D = -cosdtan6tanZ, and. the plus sign is used if90o < Z <
I
270o; otherwise use the minus sign. In the special casesZ = 9O' and,Z = 27O' or if
9
l+a2-b2<0,then

cost = -b I a.= tan6 I tan6i =

ro-'
If c o s t >+ 1 , s e t c o s t = + l ; /a 111-r ,

If c o s t <- 1 , s e t c o s t =- 1 .

Iterate Equation 9.334-1 until UT difiers from UTs by less than 01fi)8.
Storl;9ht
9.335 Times of Tiansit Times oftransit are required to low precision for planning
obsenations, and for setting instruments. The times for a particular place may bc
calculated for any body by iterating Equation 9.33 I and setting r = 0o. For tLe
Sun, the transit time may be determined ftom l2r - E I 15, where E is the equation
of time given in Section 9.311. Altitude

Figure 9.34.1
9.34 lllumination Groundilluminationfrom vadoussources

The ground illumination (i.e., the illuminance on a horizontal surface) from nat-
ural sources varies considerably during the day. The variation of the illumination
on a horizontal surface, in clear conditions, on the surface of the Earth is shown
in Table 9.34.1. The illuminance is measured in hix or lumens per square meter;
diagrammatically in Figure g.34.1 as a function of altitude.
however, the logarithm of the illuminance has been used because large ranges are
This section contains a simple method, suitable for use with electronic cal-
culators, of calculating ground illuminance from sunlight, twilight, and moonlight involved, and a,lso the sensitivity of the human eye varies in proportion to the
based on data published by the R-adioCorporation of America (RCA Electro-Optics logarithm of intensity. Thus the illuminance 1 in lux is given by
Hand.book,1974). A more precise method is available in I|nited. States Naual Obser-
uatory CircularNo. 171. This method is based on the extensive Natural Illumination l o g r 0 /= r 0+ / r r + r z x l * l g - , (9.34-r)
Cha.rts produced by Brown.
'where
Data from the handbook has been represented by sets of cubic polynomials x = hl9O, h is the altitude, a,nd ln, /1, 12,I are the appropriate set of
for the Sun and the full Moon at mean distance, for various ranges of altitudes, coefficients eiven in Table 9.34,1 for various ranqes of altitude. To obtain the total
 492 EXPLANATORYSUPPLEMENT
9 / ASTRONOMICALPHENOMENA 493
Table9,34.1
good cooditions and in the absence of other illumination) is usually described for
Coetficientstor Calculating
Groundlllumination
navigational purposes as illumination such that the brightest stars are visible and
Fromthe Sun Maximum the sea horizon is clearly defined; for the beginning and end of nautical trviliglrt,
Altituderange l2 error the corresponding statement is that the sea horizon is in general not visible and it
is too dark for the observation of altitudes with reference to the horizon.
20. 90" 3.74 3.97 -4.07 0.o2
520 3.05 13.28 -45.98 64.33 o.02
-0.8 5 2.aa 22.26 207.64 1034.30 0.o2 9342 Moonlight The illumination received from the N{oon varies according
-5 -0.8 2.88 21.a1 258.11 _858.36 o.02 to phase, altitude, and atmospheric extinction. From full Moon in the zenith, the
2.70 t2.17 -431.69 _ 1899.83 0.01
18 l2 13.84 262.72 1447.42 2797.93 intensity of illumination on a horizontal surface is approximately 0.27 lu-x. equiv-
0.01
Fromthe Full Moonat MeanDlslance alent to that from the Sun at an altitude of about -8". The amount of ground
20" 90. -1.95 4.06 4.24 1.56 o.02 illumination at night from moonlight logtoM, where M is the illumilance in lux
520 - 2.58 12.58 -42.58 59.06 0.03
-0.8 may be calculated from L1, the ground illumination from a Full \Ioon at altitude
5 -2.79 24.27 -252.95 1321.29 0.03 ft, at a mean distance that has been corrected for phase (22) and parallax (t3) a^s
follows:

l o g r u M= L 1+ L 2 + \ - (9.342,1)
ground illumination from the Moon, corrections for phase and parallax must also
be added (seeSection 9.342). Extracting the appropriate set of l's from the table and setting x = h 190, where h
The values for the Sun in daylight give the illuminance from direct and indircct is the altitude of the Moon, then
sunlight in a cloudless sky. The values for indirect sunlight are about a factor oI Lt=ht+ltx+l2f+ItI',
10 smaller. If the sky is overcast, the r,raluesfor indirect sunlight should be reducecl
L z = - 8 . 6 8 x l o - 3f - 2 . 2 x l o - e f , rO 1L)-)\
by about a factor of 10, and if the sky is very dark by about a factor of 100. A
correction of 2lo916(5/ So), where S is the semidiameter of the Sun and Sq is thc L t = 2 l o g r c Ql ro . 9 5 l ) ,
semidiameter at mean distance, should be added to the daylight illuminance to = -
where "f 180 E
allow for the va.riabledistance of the Sun from the Earth.
When the Sun is below the horizon, the sky is illuminated lrom the sourcr:s and cosE = sinds sin6,v+ cos6s cos6rl cos(dM- c!s).
discussed in the following sections.

934f Twilight This is caused by the scattering of sunlight frorn the upper layers Figure 9.3,1.1also shows the ground illumination on a horizontal surface from thc
of the Ea.rth's atmosphere, It begins at sunset (ends at sunrise) and is conven- stadrdard phases of the Moon at mean distance and at various altitudes. The half
tionally taken to end (or begin) when the center of the Sun reaches an altitudc Moon gives only one-ninth as much light as the full N{oon. For astronomical obser-
of -18". At an altitude of -18., astronomical twilight, the indirect illuminatiol vations, the position of the Nloon in the sky, rather than the general illumination'
from the Sun on a horizontal surface is about 6 x l0 a lux, rather less than the is the most important factor.
contribution from starlight and airglow. The actual brightness of the sky deperds
on where the observer looks and where the illumination is coming from, as well as 9343 Starlight, Airglow, Aurora, Zodiacal Light, and Gegenschein The total illu-
on meteorological conditions. mination from the stars contributes abort 2.2 x l0-4 lux, rather less than the Sun
In navigational practice and for certain civil purposes, two intermediate steps at the beginning and end of astronomical twilight. The illumination from the stars
in the twilight period are recognized and tabulated: ciui,l twilight ends (or begins) together with airglow contributes about 2 x l0 3 lux. The illumination from the
when the Sun reaches an altitude of 6" and nautical twili.ght ends (or begins) aurora (which depends upon solar activity) may in rare cases be comparable with
when the Sun reaches an altitude of - l2o. Before morning and after evening civil moonlight. The other sources are very faint and never give an illumination greatly
twilight, outdoor activities that depend on natural lighting require artificial illu- exceeding that from starlight. As with moonlight, the position of the source is the
mination. The degree of illumination at the beginning and end of civil twilight (in most imDortant factor for astronomical observations.
 494 EXPLANATORYSUPPLE[.4ENT PHENOMENA
9 / ASTRONOMICAL 495
9.4 OCCULTATIONS The formulas for ;r and l, the coordinates of the center of the shadow in the
fundamental plane, are
Only dates and general areas of visibility for the occultations of planets and bright
stars by the Moon are published in Th,eAstronomi,cal Almanac. Ranges of dates of .r sin ?r = cos 6 sin(a - a"),
(9.41-1)
occultations of X-ray sources by the Moon are also given. The International Lunar - -
I sin ?r = sin 6 cos 6s cos 6 sin 6, cos(o a'),
Occultation Centre, Astronomical Division, Hydrographic Department, Tsukiji-5,
Chuo-ku, Tokyo, 104 Japan is responsible for the predictions and reductions of
timings of occultations of stars by the Moon. The International Occultation Timine where a and d are the right ascension and declination of the Moon at a particular
Association (IOTA) provides predictions of occultations. Universal Time and r is the parallax.
For prediction purposes these reduce to

.r= ]j cos6, y= a " ' + 0 . 0 0 8 7 x 1 -o o ' ) s i n6 , , (9.4r-2)

'1f
9.41 Occultations
of Stars 1f

Whenever a star is within an angular distance from the Moon that is less than where d, d, and rr are in degrees.
the sum of the Moon's horizontal parallax and semidiameter, it will be occulted At conjunction
from some point on the Earth. The large number of occultations make it impossible
to publish maps of each occultation in The Astronomical Almanac. It is possible, 6-6,
however, to calculate and display these maps on a personal computer. The calcu-
(9.41-3)
Iations require the Besselian elements, which are defined in a similar way to those , a/ cosd 5t /
x = ---, y,=-_Y-
for eclipses. In occultations, the fundamental plane passes through the center of nTfT

the Earth and is perpendicular to the line joining the star and the center of the
Moon-i.e., to the axis of shadow. The origin of the coordinates is the center of the Here the primed quantities are hourly variations that arise from differentiating the
Earth; the x-axis is the intersection of the Earth's equator with the fundamental Moon's daily polynomial coefficients for cr, 5, and n.
plane and is taken as positive toward the east; the y-axis is perpendicular to that
The rectangular coordinates of the center of the shadow on the surface of the
of x and is taken as positive toward the north in the fundamental plane. The great Earth at Universal Time / from conjunction are given by
distance of the star implies that the fundamental plane is perpendicular to the line
joining the center of the Earth to the star, and that the Moon's shadow is esseDtially
a cylinder rvhose intersection with the fundamental plane is a circle of invariable
size, its diameter being equal to that of the Moon. The coordinates of the center of y=y't+Y, (9.414)
this circle-i.e., the axis of shadow are denoted by.{ and }. The adopted unii of
linear measurement is the Earth's equatorial radius. 0-?-v21,
The Besselian elements are given for one instant only-namely) the time of
conjunction of the star and Moon in right ascension, when x is zero. They are: where z is measured along the shadow axis from the fundamental plane toward the
Moon. The longitude and latitude (), d) of this point on the surface of the Earth
7'o= the UT of conjunction in right ascension; axe calculated from
t(y cos 6 + z sin 6),
I/ = the Greenwich hour angle of the star at Z0; d = sin
f=yaITp| '
h=tarl -ysin6+zcos6' (9.41-s)
t',)' = the hourly rates of change in.t and y;
cr,,6, = the right ascension and declination of the star. \=h-(H+15x1.@2738t).
 496 EXPLANATOHY
SUPPLEMENT 9 / ASTHONOMICALPHENOMENA 497
The rectangular coordinates of the edge of the shadow on the surface of the
Earth, ({, 4, O, are given by

t=x-ksinQ,
q=y-kcose, (9.41_6)
(1 - €2 - n2),

where k = O.2'725is the radius of the Moon in Earth radii and Q cycles through the
range 0o to 360". The longitude and latitude (), /) of a point on the edge oI the
shadow is obtained from Equation 9.41 5 by replacing (;r, y, z), with ({, a, 6).
The remainder ofthis section describeshow to plot a map of the area ofvisibility
of an occultation. Project the longitudes and latitudes of the continents, islands,
Figure 9.41.1
and places of interest onto the plane that passes through the center of the Earth
Fundamental planeshowingpathof occultation
and is perpendicular to the direction defined by a vector from the center of the
Earth to the place ,\", /.. Here (.\., /") is the Iongitude and latitude of the center
of the shadow at the time of conjunction. This plane rotates with the Earth and
coincides with the fundamental plane at conjunction. The rectangular coordinates
of a place at longitude and latitude (), /) are given by
The great circle arc AB between the polar angle d1 and 92 is then plotted by
settingx=cosd, l = sind, and z= 0, and using Equations 9.41*5 and 9 41 7. The
polar angles are found as follows: Calculate Q1 ar.d Q2 ftorn
u = cosd sin() - lc),
t) )2 _1
v = sind cosd. - cosd sind" cos(l - A"), (9.4r,7) s t n ( g t+ J ) = and Q2= 180"- 0r 25, (9.41-8)
2 tD
w = sind sind" + cos/ cos/" cos() - ).), t^ -Lto/x0).
where D= 1f@o+tZ) and J=tan

Thenfori=land2
vrhere a and v are in the plane (with the origin at the center of the Earth), Il is to
&=xo-ksin0r,
the east, y is to the north, and w is the height of the place above the plane; w ) 0
for all places to be plotted. 71t= y1t- kcos Qi, (9.4r-9)
To plot the edge of the shadow at any time during the occultation, calcu- I(riil€i),
and 9i=tan 0" <0i<360?
late the rectangular coordinates of the center of the shadow r0, y0, from Equa-
tion 9.41-4. Then calculate the coordinates {, 4, ( of the edge of the shadow frolrt
Equation 9.41 6 and obtain the corresponding(), /) from Equation 9.41 5. Finaill
project the (1, t') of the shadow outline onto the same plane as the continents, etc.' The total area of visibility of the occultation may also be plotted by repeating
using the coordinates (2, v) in Equation 9.41 7. the calculation for a series of times between the start and the end of the occultation
When {2 +\2 > 1, the shadow is only partially on the Earth. To find th<: on the same diagram.
values Q1 and p2 where the edge of the shadow cuts the surface of the Earth, see A similar procedure may be applied for predicting the areas of visibiliiy of
Figure 9.41.1, where O is the center of the Earth and S is the center of the shadow occultations of planets by the Moon.
 498 EXPLANATORY
SUPPLEMENT PHENOMENA
9 / ASTRONOMICAL 499
9.42 Occultations
of Planets
z.et\rth
The prediction of occultation of stars (or minor planets) by planets (or by minor
planets) Iargely centers on the search for conjunction in right ascension within the
limits of difierence of declination that make an occultation possible. These Iimits
are very small, being approximately the sum of the horizontal parallax and the
semidiameter of the planet.
The observations themselves yield useful information on the diameters of plan-
ets and about the shape of minor planets.
The actual prediction follows the basic principles underlying those for eclipses,
occultations, and transits; but more direct and less formal methods are used because
the angles involved are much smaller and the prediction much less precise.
An occultation will take place as seen from some point on the Earth's surface.
provided the difference 6 - 6" in apparent declination at the time of conjunction in
riEht ascensionsatisfies the condition

< ( z + s ) /l s i n p l ,
lr- 6"1 (9.42r)

where rr and s are respectively the equatorial horizontal parallax and semidiameter Figure 9.5.1
PZStriangle for Polaris
of the planet and p is the position angle of its direction of motion. Note must be
taken of the position of the Sun, as for lunar occultation, because it is generaily
uselessto predict occultations of stars in daylight.
Methods of prediction are described by Taylor (1955); the paper also described
the determination of the actual limits of occultation. The IOTA (see Section 9.4)
provides predictions of planetary occultations that are published in Skg and. Tele
scope and. Lhe Astronomical Journal. The predictions of planetary appulses and
occultations are also published in The Handbook of the British Astronomical Soci- An alternative to the Polaris table is to use expressions involving coefficients
et!!, for the GHA and the polar distance p (see Section 9.51). These coefiicients are
provided on page 863 and include the 5.5 magnitude star o Octantis, which lies
close to the south celestial pole.
9.5 POLE.STAR
TABLES If the pola.r distance of Polaris is denoted by p (of the order of one degree) and
its local hour angle by r, then its altitude lr and azimuth Z, as seen ftom an observer
The proximity of ihe second-magnitude star a Urse Minoris, (Polaris or the Pole in latitude /, are given by solving the spherical triangle (see Figure 9.5.1) formed
Star), to the north pole of the sky has given it a special significance for the con- by the North Pole, the zenith, and Polaris, S.
venient determination of direction and latitude. This is particularly so in the fields Becausep is small, the solution may be expanded as
of navigation and surveying, for which the Pole Star's constant availability for ob-
servations (in northern latitudes) and the simple methods that can be used for the I
reduction of observations are invaluable. For the more precise requirements of as- h = d + p cost - sine sinzttand+ ,
,o r O5 - 1 )
tronomy, the Pole Star's distance from the true pole is sufficiently large that the
-Z cos$ = p sint+psinp sint cosrtand +....
special methods of reduction no longer confer an advantage over standard meth-
ods. Thus Pole-Star tables are restricted to the precision required in navigation and
surveying. The principal table with the precision required by surveyors is included In each case, the next term of the expansion is of order p sin2p tan2/ and cannot,
in The Astronomical Almanac, pages 864 867. for many yea.rs,exceed 011 for latitudes up to 70'.
 500 EXPLANATORY
SUPPLEI/ENT 9 / ASTFONOMICAL
PHENOI\iIENA 501
For convenience of tabulation, these expressions are rewritten in the form The complications of these tabulations are unnecessary for astronornical use,
but valuable for navigational use, in which simplicity of tabular entry and of inter-
I
o - h = - (p0
- cos r0 - : p0
- sin'p0
- sin216tan ps) polation are of foremost importance, The Polaris table in 7[e Nautical Almanac
2
is essentially the same as the table in The Astronomical Almanac, except for the
I htervals of tabulation (l' in LST or LHA of the first point of Aries) and from a
+ srnposrn'to(tan 0 - tan Oo)
tpo frrther simplification for the user by adding constants (whose sum is one degree) to
(pcos t - p6 cos t9) = ao + at + a2, a6, a7, a2 to make them always positive. Since lower precision is required in azimuth
(9.s-2)
the b coefficients have been replaced by a simpler table of azimuth as a function of
Zcos$=- ( p e s i n t 6 + p 6 s i n p s i nt 0 c o st 0 t a nd 0 )
LHA Aries and latitude.
-p0 sinpo sin/0 cosrO(and tand0)
- (psinr - ps sin 16)= bo + br + bz,

in which p6 and 16are the polar distance and hour angle of a convenient point closc of the PoleStarCoetficients
9.51 Derivation
to the mean position of Polaris throughout the year, and Ss is a mean latitude.
usually chosen to be 50". The mean position of Polaris (which must not be confuse<.l The Greenwich hour angle (GHA) and polar distance (p) are expressedas series of
with its mean place) is usually chosen to have convenient exact values for its right polynomial terms that allow for precession and proper motion, and also as series
.rscensiono0 and polar distance ps. of trigonometric terms that contain the leading terms in aberration and nutation.
The first terms (a9,b6) in the modified expressions are functions ofa sinsle Thus
va"riable,local sidereal time, LST, because
GHA = ao+ arL + a2sinL + a3cosL + a4sir'M + as cosM
,o=LST-ao (9.5-3)
+ a 6 s i n z L +a j c o s z L+ l s U T ,
rq 5l-l\
and may be tabulated at a suitable interval of LST. L + q7cosL + a4sinM + a5cosM
p = au+ atL + a2sit.J'
The second terms (a1,D1) are functions of t6 (i.e., of LST) and of latitude and
+ a6 sin2L + a7 cos 2L,
must thus be tabulated in a double-entry table with arguments LST and latitude.
By incorporating a mean value (corresponding to latitude do) in the first term.
the magnitude of these terms can be kept down to about 015; terms may thus be where 40, at, ..., aj are two sets of constantsin degrees,UT is the Universal Tiqre
tabulated at wide intervals of latitude and LST. in hours, L -- L'n, M = 9/z (where L' and, Q' are the rates of change of the mean
Similarly, the third terms (a2,b2) aie functions of re and of the apparent positiorr aberrated longitude of the Sun and the mean longitude of the ascending node of the
of Polaris (i.e., of date). By proper choice ofps and os, the magnitude of these terrns Moon), and z is the interval from the beginning of the period to the time required.
can be kept down, during the year, to about 015; and they can also be tabulated at The coefficients are determined by the method of least squares. Annual coef-
wide intervals of date and LST. ficients are derived by taking daily values. Using all eight coefficients, the annual
As will be seen from the Polaris table in ?[e Astronomical Almonoc, the single- fit gives maximum errors of t O':2, If only the fiIst four terms are used in the
entry table of as and b6 is arranged in twenty four columns, each containing values least-squares approximation then the maximum errors increase to I 1'10.Pole Star
for one hour of LST; this anargement enables separate tables of c1 and b1, and coefficients have been calculated for the ten-year period 2096 to 2105, based on its
of a2 and b2, to be given for each hour of LST. In the column corresponding to FK4 position and proper motion. The north polar distance is found to decreasefrom
the hour of LST all these terms are taken from single-entry tables-the first with its present value of about 48t to 27t 10/' on 2100 March 20 and will then increase.
argument minutes and seconds of LST, the second with argument latitude, and The distance will reach 1' by about 225O and 2" by about 2450. (See Figure 9.51.1.)
the third with argument date. The error in using the tables for the hour, without The Pole-Star coefficients may be used to calculate the UT of upper culmination
interpolation for LST, is greatest for the second term bq (owing to its dependence of Polaris (UC) at the observer's longitude, the UT of the elongation west (EW)'
on sin2to) and may reach 0115for extreme latitudes; otherwise the error is small. and the elongation east (EE) at the observer's longitude and latitude,
 502 EXPLANATORY
SUPPLEI\4ENT 9 / AsrRoNoMlcALPHENoMENA 503

Taylor, G.E. (1955). "The Prediction of Occultation of Staxs by Minor Planets', J.8.,4.,4.
65,84.
Woolard, E.W. and Clemence, G.M. (1966). Spherical Astronomy 215 (Academic press.
60 New York).

50

-40
l--

Aro
z
20

10

00
2098 2rOO 2102 2104 2ro6
Yer

Figure 9.51,1
Polarisnorth polar distance2096-2105

Calculate GHA6, and ps at 0h UT. Then calculate the local hour angle LHAe =
GHA6 +,\, where ) is the observer's longitude. Upper culmination occurs at time

." 360 - LHAo

' (9.s1-2)
15.041069
and also at time th - 23.934 on the previous day. Elongation east and elongation
west occur at times
cos-t (tand tanpo)
rht (9.51-3)
15.041069
where the plus sign is used for EW, and the minus sign for EE, and / is the observer's
latitude. If necessary add or subtract 231.934to the time of the phenomenon to
obtain a time nearer the time of observation.

9.6 REFERENCES

Bennett, G.G. (1982). "The Calculation ofAstronomical Refraction in Marine Navigation"

J. Inst. Nao. 35, 255,
Janiczek, P.M. and DeYoung, J.A. (1987). "Computer Programs for Sun and Moon Il-
luminance with Contingeut Tables and Diagrams" United, Stat$ Naltal Obserl&tof!
Circular 171 (U.S. Naval Obsen atory, Washington).
Radio Corporation of America (1974). "sources of Radiation" Electro-Optics Hand,book
Section 6, 61.
 CHAPTER
1O

StarsandStellarSystems
by D. Pascu,J.A.Mattei,C.E.Worley,H.C.Harris,J- Andersen,
W.P Bidelman,
J.H.Taylor,
G. Lynga,R.E.White,G. de Vaucouleurs, T.M.Heckman,
c.H. Kaplan,l.l.K.Pauliny-Toth,
andJ.F.Dotan

Section Author Affiliation

(10.1-10.15) Anonymous
(10.2, 10.21) D. Pascu U.S. Naval Observatory
(10.22) J.A. Mattei Am. Assoc. Valiable Star Observers
(10.23) C.E. Worley U.S. Naval Observatory
(10.24) H.C. Harris U.S. Naval Observatory
(10.25) J. Andersen Copenhagen Univ. Astron. Observatory
(10.26) W.P. Bidelman Case Western Reserve University
(10.27) J.H. Taylor Princeton University
(10.31) G. Lyngi Lund Observatory
(10.32) R.E. White University of Arizona
(10.33) G. de Vaucouleurs Unlversiry of Texas
(10.34) T.M. Heckman Johns Hopkins University
(10.41) G.H. Kaplan U.S. Naval Observatory
(10.42) I.I.K. Pauliny-Toth Max Planck Institute
(10.43) J.F. Dolan NASA/Goddard SpaceFlight Center

10.1 SOURCES OF DATA ON STARS AND STELLAR SYSTEMS

The data on stars and stellax systems are basically provided in catalogs. Positional
sta.r catalogs can be divided into types. Obseroational catalogs provide the results
of observing programs with a given instrument . Compiled catalogs are compilations
of observational catalogs, which are usually assembled over an extended period of
time so that proper motions of stars can be determined as well as positions.
 506 E X P L A N A T O NSYU P P L E M E N T
507
10 / STARSAND STELLARSYSTEMS
Obserr-ing programs and catalogs cau be classified as fundamental or diflererr-
10.12 StandardReferenceCatalogs
tial. In firndanrcntal observing programs! the coordinates of the observations i11p
determined independently as part of the observing progrml. For exarnplc, the polr..
equator, and erluinox might be determincd from observations of circumpolar sttrrs Referencecatalogs are of diflering types. Somc compiled reference catalogs are irrde-
and solar systern objects. Differcntial observations are made with respect to stilt pendent sourcesof positions and proper rnotions. Examples are the FK3, F K4, arr<1
positions from a reference catalog. Ail pholographic and snall ficld obserrations FKS catalogs. Other catalogs are dependent on the independertt catalogs to deter-
are diflerential. mine their reference positiorrs. Examples are the AGK2 and AGK3, u'hich require
Special purpose catalogs givc a specific type of data, such as parallaxes, rtrdial the FK caialog Positions.
Observational catalogs are fundalnental, or absolute' if the refelence liame can
velocities,and so on.
Heterogeneous catalogs are a collection of material from various sources. Io be determined from the observations themsclves. The independent compiled cata-
many cases.thc catalogs are limitcd to a region of the sky, a range of magnitrrdcs. logs must include some funda.mental obsen'atiorral catalogs. They can also inclrrde
or a given type of object. The stars are identified by a catalog number or br. a difierential obsen'ational catalogs r.herc the obscrvations are made with respect to
constcllation name and dcsignation.A lirnited rrumber of the brightesl stars har,o some reference catalog.
specific narnes. Lists of star catalogs for the cighteenth and nincteenth centurics are given in the
volumes of Geschichte des Firstetnhimmels (Karlsruhe, 1922 1957; Berlin. 1952
1959).A further list for the period 1900 1925is given in Index der Sternorter'79OO
1925 (Bergedorf. 1928). Foremost among the observational catr,rlogsirr these lists
are those compiled under the auspiccs of the Astronornische Gesellschaft (A G ) as
10.1 1 Compiled Catalogs of Stellar Positions and Motions a cooperative effort by a nurnber ol obserratories. This series of volumes was bcgun
in 1863, and gir-es the positiotis of all stars shown in the Bonner I)urchmusterung
(8.D.) from declination +80" to -18" to magnitude 90 The observationsmade
Compiled catalogs arc designated b-v-a systcmatic referencesysteru orr the sk1..rrhicL
is basedon a selcctedlist of stars included in the catalog.Thc positions and proper by each observator-5rwere colfined to a narro\\' zone of declinations bost suitcd to
motions for all thc stars are on thc same system of coordilates. Obserratiolal the latitude of the observatorv. In mote recent times photography has been used to
data from a number of different instrunents are included to minimize sl'stematic reobservc the A.G. zones. The neu''positions are givcn in:
instrumental errors and to achicve the optimum accllracy. Compiled catalogs ar{l
given on a standard equiuox and epoch such as 1900.0.1925.0,1950.0,or 2000.0. AGK2 Zwe.iter Katttlog der Astronomischen. Gesellschaft fi)r das AEtinoktiunt
This is achicvedby applying proper motion and pre<:ession to the compiled obscrvc(l 195A; kn volLunes covcr declination +90' to +20" (Harnburg Bcrgedorf'
positions made at difercnt times. (Thus, the epoch of the catalog is not the a\1:r.age 1951 1951),ancl five volumes covcr-declinations+20' to 2' (Bonn, 1957
date of the observationaldata in the catalogs.) 1958).
LGKS iweiter Kataloq tler AstrortonLisclwrtGest'Llschaflf r rlas Aqu inoktittrn
The compiled catalogs can be grouped into tu'o series. The German serjes in
1g'/3; based on plates taken at Bergedorf betrvcen 1959 and 196l Proper
cludos ?Ae Fund,amcntal Clatalog. The Neucr Funda,mental-Catalog,and thr: FKjl,
notions \4'ere lbrured usirtg AGK2 as first-epoch plates AGK3 R is the
FK4. and FK5-Basi<rcatalogs.The Arncrican seriesof catalogsinclucleslleu,<olLbs
r:atalog of referencc stars lbr the AGK3.
catalog of fundamcntal stars. Eichelberger'sstandard stars, thc GC catalog, arrrl -60'
Yale Catalogsof the zoncs+85o to +90", +50' to +60"' 30' to +30', anrl
thc N30 catalog. In addition, there is the SAO catalog, n'hich is a, hcterogcncous 'ln Obsenntoru
to -90' are publishecl Trart srtctiont; the Astronomical
"f
combination of a uumber of catalogs.The methods for converlirrgthe positiorrsiionL of Yole fhtiuersiLy (Nerv Haven' 1925 onn'ard); the posititins are aiso lbr
one epoch and refercnce system to another epoch and referrlcc systcr:rris gir-en itr equinox 1950.0.
Chapter'3. Currenlly in preparation are the S1luthr:r1ll-cfcrcnceStar st,stert. tlrt:
International RefereuccStar systern,and Part II of the FK5 catalog.
Among uioclern reference ca.talogs,the following are representative
In the firture it is expectcd that radio sources or quasars will be the rcfererrce
sources to which astrononical positions rvill be referred. These sources have the
advantage of lackirrg apparent tnotion due to their distance. Thus they provid<: a GC General Catalogrte ttf 33312 Stars for tlte Epoch 1950' in 5 vohrrnes' Wash-
ington, 1937. These catalogs give positions and proper motions of all stars
type of inertial or space fixcd rcference system.
briglrter than magnitude 7, and also include thousands of faintcr stars'
509 EXPLANATORY
SUPPLEI,IENT 1O / STARS AND STELLARSYSTE[,1S 509

FK3 Dritter n d,arnentalkoto,l,og desBerliner AstronomischenJahrbtclts. I'l't.i.l: have been publishcd in the Grcr:tlvich Observzltiottsseries),Riiclio soulce lists, arrci
Veri)ffentlichun gen d,esA stronom.isr:lte.nRechen-Institrtt.s. no 54. 1937.This many other obserlations from other obscrvatories.
catalog gives positions for 1925.0 and 1950.0 of the 925 star-sof Aur,ers'
Neue Fund.arnentalkatalog (19\0). II Teil: AbLLandlungen dcr Preussischt:n
Akademie tler Wissenschaftt:n, Phys.-Math.. K[asse'' nct.3. 1938. l'his cat and Lists
10.14 OtherCatalogs
alog gives the positiorrs of 666 additiorral stars for equinox 1950 0.
FK4 Fourth Fundam,en.talCataloq (FKl), Ver6ffentlicltungen dcs Astronorttis Con.rpiled catalogs and obscrlational catalogs generallv dcarl ri'ith posiLions arrcl
chen Rechen-Instituts, Heidelberq,rro. 10. 1963 A revision of FKil: tlrr: proper motions of stars. h additiorr,there are catalogsof doublc stars attclcatalogs
re-examination of the arailarblc obserlations shori'ed that no changc itr tht' giving lhe radial velocities.parailaxcs,spectra, and other characteristicsol stars.
equinox rvas justified. In sorne cerscs.the catalog can be based on single observation of each slar': in otllcr
FKb Fi|th Funclanrental Catalog (FK5), Part 1: The Bosic FuntlarrtenloLSlnt s. cases,such as parallaxes, rlany observations oI the stars arc necessarv in otder 1o
VeriiJfentlichungcn AshonomischesRer:he,n-Institttt, Heidelberq,no 32.19a8. determine the parallaxes.
A revision of FK4: the rc-cxamination of the available obscrlations shorvecl
that changesto lhe constanl of precession and a cottection to the equitrox.
both at the epoch and as a function of tirtle, r'ere necessarli SeeSection11.5 10.15 Data Centers and Their Facilities
for the tlansformation betrveen the FK'1 and the FK5 s1'stcrtl.
N30 "Catalog of 5268standard stars, l950.0, basedon the normal sl stcD \31).' With the irrtroduction of clata-processing e(llripment,otganizationsuse the equip
Astronomical Papers of the Americart.EpLtenteris.13' part III. 1952 flre ment to der.clop a collectiort of data in rtachine-reaclable form The clata inclucle
positions arc derived from rnore than 70 catalogs with epochs of obserlal iou data crurentl1 being devcloped and also older data that rvas codecl into machjrrt:
bet$'€en1917 and 19.19. readable form fol analvses purposes. In gcrrt:r'al.these cla,tacollections wcr-clo.atecl
ZC "Catalog of 3539 zodiacal stars for tlte eqrrinox1950.0."Astronom.itaLPrt at the places actuallv makirrg the ollserl-alions or at titc ones most ticpt:rrdent uporr
pers of the AmerzcanEphemeris.l0 part 11,19'10.This catalog givesposi computer ana,l1'sesfor exantple, the U.S. N aval Ollsclvatorv. Rolal Creenu'ich
tions of all stars to magnitudc 7 in the zodiacal zone as rveil as the positiorr-' Obsenatorl', Astrononischcs R.echen-Irrslittrt, and Pulkova Observatorl'.
of many faintcr stars. It is intcnded for use in occrrltation r-ork. It is basecl The rrurnber of diflcr<rt locations for collectioris of data proliferateci. and thc
on 90 catalogs, and positions are rcduced to the FK3 system.
need for ccrrtral data centcls becamc cvident Thercfore. the Certtre dt: l)onntr:s
PFKSZ "A preliminary gerrcral catalog of futrdamental failt stars bclrveert rlc
Stellai,rest'as establishcd itr Strasbour-g, France Subsequentll'. a data { entcr ("as
clinations +90o and -20"," N{.S. Zverev attd D.D. Polozheutsev.Publitt
establishe<lat the NASA (ioddard Sptxr: Flight Curter. Thcsr: tu'o orgaDizltrons
tions of the Main Astronom.icttlObse,r,Latory of Pulhot:o.72. 7958.
have become central dcpositolies for data They obtain copies of the dala frotn
previouslyestablishedcollections.In nddition. they cross-indcxand <rrnsolidatctht:
10.13 Observational
Positional
Catalogs information from many difTerent systcrris into a sirrgle file of data on ea(rhsl al
Computer netr,orks allorv astronorners to clircctll acct:ss tlrt' dtt'la cenicr- io
Obscrvatiolal <lata lrorn a given instrurrrcnt over a pcr-iotl of time trrc usrrallr 1'rrl-
obtain the most currcllt inforrnation of the desirctl specilic data. Onc s]-stenl
lished as a catalog.In this case,the data have been reducedin a systelnaticlnanlr(rl- I)ata
the Set of Indentifications,\'Ieasurcrrtcnts, and Bibliograph.yfor-Asl,rotrotni<ral
and are prrblishcclo sonle references]-stcrn.either ou the s)sterll of tr c"trlf il''L
(SIMBAD) systerrr allows direct acccssto the computer at lhc Strasbourg dirl'a
cattrlog or on a systcrrras determined by the instruttcnt. Becauscthe lrrellrt"i "1
center and the capabilitl'to obtain all the current jnformatiort corLcel-1lin!! a sper:ilic
reduction and thc referencesystem can vary for obscrvationalcatalogs,the intlo
star.
duction or description of the catalog must be revietr:d carefulll-. and th(r rlse oI
observationalcatalogs and the combirringof observationalcatalogs mrrst h"l"rlt
n'ith care. Observationalcatalogs usually combine a number ol observati'rnsr'1 11
given star to determine the position of that star, and give a mean cl'och fol llL' 10.2 STELLAR DATA IN IHEASTRONOMICAL ALMANAC
tirne of the observatiorrs. This epoch may differ from the epoch of thc refelclic('
system on which tlrc stars are given. Exaurplesof obscrvationalcat.logs itlp tlL' nl'. The progenitorsof The Astronont.icalAtrnanac(The Astronon cal Ephcrneris[AE]'
selvations of the six-inch trarlsit circle in Washington (t'hich have becn ptrblishetl formerlv ?%e Nautico,LAlmanac, and The Arneri.can Ephenteris aru1 N.rutt' ol AlnLa-
in the Ptrltlicationsof the U S. Naval Observatorv),Greenwichobscrlatiotrs(ulLit h nac [AENA]) have ah'ays inclucleclstellar data in the lorm of lists of apparent
 510 E X P L A N A T O RSYU P P L E M E N T 10 / STARSAND STELLAB SYSTEI!4S 51 1
places and/or mealr places of bright stars. Such lists were rrseclbv asrlororners lo1 eflers fourtccrr lists in 13 tables. A rnixture of frequently observed objects as well 1s
the posiiional calibration of their instruments and formed the basis for navigatiorrql commollv used caliblational systerns is inciudcd. !}en in those lists not specifically
and surveying almanacs. The precision to rvhich these lists were published rarir:d defining a calibrational system, rnuch of the data can bc used for calibratiolal
fron 0ll in right ascension and l" in declination. to 0i001 in right ascension ancl purposes.Titles and short dcscriptionsof these ta.blesare:
0101 in declination. \Vith the publication ol Appo,r'entPlaces of Fundamental Stars
Bright Stars:1482 bright stars. listing positional (FK5, SAO), photometric
in 1941, the number of apparent places printed in the almalacs was greatly tli-
( I/B I frorn Nicolet. 1978) and spectroscopic data (from W.P. Bidclman,
minishcd, aDd by 1957, apparent places for stars were omittcd altogether. In the
1991). Also notes on duplicity. r'ariability, and reliability of data (from Yale
1960 editions, the ,,1-Eand ,4ENA n'ere unified in cortentl but kept tireir scparate
Bright Star Catalog, |tb. cd., Hoffeit and Jaschck. 1982).
names. The purpose aDd scope of the almatacs had also evolved cousiderabll'. Tlrl Bright MK Atlas Standartls: Included irr lhe Bright Stars list. Recent N'1KAtlas
emphasis was [o longer on the "practical requirements of navigators and survey Standards, r'hich are inciuded in Lhe Yale Bright Star Cri.aloq.
ors," but on the "requirementsof the practical astronomer" (Exp. Supp.pp. 10 11). LIBVRI Standarrls: 107 bright standards selected bv II.L. Johnsol.
For stellar data, this resulted in rnean places for stars briglrter than.1.75 (1078) to Striirngren four-color and, H-beto,standards:319 briglrt standards selected liom
an accnracy of 0ll in right ascensionand l" in dcclination. Visual maqnitudes arrrl list of Perry el al. (1987)
spectral types were also given, though they wcre not up-to-date. Rad,ial-Velocity Stan.clo.rds: Bright and faint IAU lists (IAU Trans. 1957. IALI
With the 1981 edition, the unificatiorr of the almanacs was compietetl bl the f r a i r . . 1 3 7 3 )U
. ndpr reli.inn.
adoption of a single printing u'ith one title The Astronomico,l Alm,anac. Th\s edi- Selected Variable,Stars. 181 bright variables of diflcrent typcs, sr:lccted by J.A.
tion was a complcte revision in contcnt as tvell as format. To rlake the alruanar: \'Iattei from the fourth edition of the GeneraL CataLog of \raliable Stars\
Skg Cutalog2a00.0. Volume,2;and data flles of the American Association
morc rcsponsive to modern astronomical research, the scope of this new rerisidr
of Variable Stars Observers.
u'as enlargcd to include an improved and expanded stellar data section.
Selected,Open Cluters: 288 operr r:lusters selected by G. Lvngri (1988) frorn his
The general precepts for establishing the contents of the section were as follorr.s.
computer-bascd Lr,nd Co,to,logof Open Cluster Data.
Sirrce Z[e Astronomico,l Alrn.anac ryas placed handily near most teiescopes.firxling
GlobuLarClusters:Data on 157 galactic globular clusters conipiled by R.E.
lists for the most frequentll. obserl.ed obiects q,as considered most useful. These Whitc.
stellar or nonstellar objccts could be the subjects of investigations or they miglrt bc Bright Galaxies:194 of the brighl<:stand largesi galaxies,selec:tecl by G. de
used for calibrational purposes. Categories for these obiects .$'ereidentiiied (son(' Vaucouleurs frotr the SecondReference Catalogue o.fBright GaktL:ies(7976).
fell in several categories) and the data most pertinent to the observations tr,cre RarlioSotrce Positions:233 radio sourcessuitable for positional calibration:
listed. list of Argrre et dl. (7984')
Since the narnes of the astronomical objects researched *'ere quoted in thc lil Radio FLux Calibrators: 14 radio sotrrcessuitablc for flux calibration, from Baars
erature, studies u'ith the Bibliography of Astronomical Objects of SIN{BAD (e.g.. et aI. (1977). as updated by thc authors.
Selectedldentificd X ray Soutces:194identified X-ray sourccs,selct:tecl by J F '
Ochsenbein and Dubois 1986) u,'erc indispensable in determining the most fre-
quently quoted lames. Ochsenbein and Dubois have shorvn that thc rnost popular' Dolau from his unpublishecl survey filc.
Selectetl,Qu,a,sars:99 QSOs sclccterl b"vT.N1. Hcckman fr o:n:A Catalttg of Quasars
obJects now are nonstellar, and include galaxies, QSOs. globultr,r clusters. X-ra1'
and Actiue Nizclei(V6ron-Cetty and V6ron. 1985)
sourcesrand pulsars. Calibratiotal objects were tlot as readily identificd sincc thc]'
Selected,Pulsars:92 Pulsars selected by .1.H. Taylor frorn t]te cornpilation ol
were not usually cited. However, calibratiolal systems tcrc ah'ays mentioned, iurd
Nlanchesterand Taylor (1981).
studies such as that of N{ermilliod (198a) u'erc useful irr determining which systcrns
were most frequently used. Ideally, lists lecommended by the IAU. such irs tle ljst o[
More information on thcse astronorrical objects and tables' q'ith references
radial velocity starrdards (IAU Trans. 1957) $,ere the rnost desirable. Urrforturratei.r'. to additional data, arc givcn irr the sections that follow. Altliough these tables
that rvas the only such list, ancl it was hopelessly out-of-date. include about 90% of the lOU most frequently quotecl objects in thc literature,
Once a selection of tables (lists) was madc, internationally recognized expelts there are some rtoticeable gaps, including a table of Debulae and a list of infrared
were contactedto help make thc selectionofthe objects and data to be listed. Thcse
objects. It has also been suggestedthat fainter calibratiorral objects, such as fainter
expets. in general, had themselves published catalogs or compiled unpublished
photometric standarcls, be listetl. Fainter calibratiorral objects I'ill be included in
data files of astronomical objects. Selections from thcse catalogs (files) 1r,.erethe!
proposed for publication in The Astronomical Almo.nac At present, the alrrtalac future editions, but the problern of finding charts has not yet been resolved The
 ct4, EXPLANATOFlY 10 / STARSANDSTELLAR
SYSTEMS 513
SUPPLEI\,4ENT

form and content of the section are incomplete and experimentall it is expected to the FKs, GC, or SAO catalogs- in that order of preference-and all positions t'ere
change continually to reflect the changing needs of astronomical research. reduced to the FKS system.
Photometric quantities orr the UBy system were obtained Irom Nicolet (1978)
except for some double stars for which Nicolet lists the combined magnitudes and
10.21 BrightStars
color indexes. In these cases,the magnitudes and color indexes were taken from the
Tables of bright stars have been published in the national almanacs from their third ediiion ofthe YBSC. Spectral types on the MK system (if available) or the lVlt
inceptions. This affirms the continuing usefulnessof bright-star data. The position5 Wilson system have been kept current by W.P. Bidelman (1991) from his extensive
of the stars a,rethe most important data to users. These data are used in navigatiol card catalog. Last, but not least' the notes give the duplicity and variability status
and topographical surveying and by observational astronomers for the positiona] of each star as well as indications of the reliability of the data. It is in these notes
calibration of telescopesand instruments. that the YBSC (fourth edition) has been most useful.
At present lhe Yale Bright Star Catalog (Hoffieit and Jaschek, 1982) is the The annual publication of The Astronomical ALmanac ptesents a considerable
most comprehensive hardcopy collection of data for stars brighter than 6.5 visual. opportunity for the Bright Stars list to offer the observer the most current and
It contains astrometric, photometric, and spectroscopic data as well as duplir:it.y relevant data for the brightest stars.
and variability information for the 9110 stars of the Haraard, Reoised,Photomet.nr.
The Sky Catalogue2000.0, Volume 1(Hirshfeld and Sinnott, 1982), is less con
prehensive than the YBSC, but contains information or 45,269 stars brighter than Stars
10.22 Variable
8,05. Neither tome, however, can compare with the computer-based astronomical
data bank of the Centre de Donn6es Stellaires at Strasbourg Observatory cailrtl The information in the list of variable stars has been compiled using the references
SIMBAD (Set of Identification, Measurements, and Bibliography for Astrorromical in the third edition of the General Catalogue of Variable S,or5 and its three sup-
'Ihis plements; lhe Shg Catalogue 2000.0, Volume 2; and the data files of the American
Data). data bank gives, for each of 600,000 stars, an exhaustive arrav of data
as well as citations for each star in bibliographical references to papers in 90 ma- Association of Variable Star Observers (AAVSO). Only bright stars from each class
jor astronomical journals since 1950 (Egret and Wegner, 1988; Wegner. Egret. arrrl of variable stars that also fit the criterion of having an amplitude of variation 0 5
Ochsenbein,1989). magnitude or greater have been selected. The limiting magnitude for each type of
When work began on the Bright Stars list in the late 1970s, only the thircl star selected is indicated in The Astronomical Almanac.
edition of the YBSC was available, and most of its data had been superseded. For. Complete, up-to-date information on about 28,'150 variable stars discovered
tunately, some of the more important new data $'ere already available in computer- and named by 1982, and on the constellations Andromeda through Vulpecula is
readable form. A selection of less than 1500 stars (30 pages) was made from the given in the fourth edition of llte General Catalog of Variable Stars (Kholopov
third edition: all stars brighter that 4.5 visual were included.In addition, FK4 stals et at., 1957 1987) (GCVS), published in three volumes from 1985 to 1987 bv the
brighter that 5.5 visual were included to insure that the positions of fainter stars Astronomical Council of the USSR Academy of Sciencesin Moscow This catalog
were precise to 1 arcsecond. Finally, a short list of bright MK spectroscopic starF contains the 1950.0position, precessionconstants, galactic coordinates, referltc'
given'
dards were incorporated in the list to avoid unnecessary duplication. These q'ere and finder chart information, t]'pe, range of variation, epoch for the elements
YBSC stars that were in the atlas by Morgan, Abt, and Tapscott (1978), and the period, and spectrum for each star. Each volume also contains a section of remarks
this
atlas of Keenan and McNeil (1926). on stars that require them. Two more volumes will be published as part of
Following the lAU-recommended practice of giving two identifications for eaclr edition, and li,ill contain positional information for the equator and equinox of
object, the Bayer/Flamsteed designations were selected becauseof their rvide-sprearl 1900.0 and a list of variable stars in order of right ascension'
tape'
use, and the BS (HR) number to facilitate access to additional data given in the The information in the fourth edition of the GCVS is also on magnetic
and may be be obtained from the Astronomical Data Center (\iaii Code 633' NASA
YBSC. The positions of the stars are still the most important data to the obser\ er.
and for most applications, a precision of one arcsecond is sufficient. To avoid listing Go<idardSpaceFlight Center, Grcenbelt, MD 20771,U.S A)' Bound copiesof the
the proper motions, yet insure the stated precision for the rvhole year, the positions first three volumes are also available from the AAVSO,25 Birch Street, Cambridgc'
are brought to the mean equator and equinox of the middle of the current year, MA 02r38, U.S.A.
the
with proper motions applied. Reduction to apparcnt places may be performed using Stars that have been discovered and narned from 1982 to 1988 are given in
of
the methods given in Section B. Starting positions for the stars were ta,ken from 67th, 68th, and 69th Name Lists of Variable Slors, published by Commission 27
514 EXPLANATORY
SUPPLEMENT SYSTEMS 515
10 / STARSANDSTELLAR

the IAU as Information Bulletin of Variable Stars (Numbers 2681, 3058 and 3323.
respectively). Dr. E. Makarenko
The nomenclature used in the list of r,rariablestars in both The Astronomical Odessa Astronomical ObservatorY
Alrnanac and. the General Catalog of Variable Stars is the conventional method Shevchenko Park
developedby F.W.A. Argelander in the mid-1800s.According to this method, the Odessa270014
first star discovered in a constellation is given the Ietter R, followed by the genitir.c USSR
form of the Latin name of the constellation. Subsequent variables in the sane
constellation are named S, T, and so on to Z, followed by RR to RZ; then SS to SZ.
10.23 DoubleandMultipleStars
and so on to ZZ, after which follow AA to AZ, and so on to QZ, the 334th variable
on two palalnetels: (1)
in a constellation. The next variable found in a constellation is named V335 plus The structure and evolution of a star is basically dependent
Double-star systems pelmit the
the genitive. Stars which have been assigned Greek Ietters prior to the start of this the initial chemical content, and (2) the initial mass.
in astrophysics. Their
system, such as Omicron Ceti, or small Roman letters, such as g Herculis, contirrue determination of the latter parameter, hence their importance
into the process of star formation'
to keep those names. numbers and distribution also provide insights
the stellar population'
Extensive references on variable stars may be obtained through the intcrla- Double and multiple stars form an important fraction of
by chance juxtaposition of
tional database SiMBAD maintained by the Strasbourg Centre de Donn6es Stel- The term d,ouble star inchtd€s the optical pairs formed
are physically related. For manv
laires (CDS) (11, rue de I'Universit6, F-67000 Strasbourg, France) unrelated stars as well as the true binaries that
determine whether the pair is phys-
The AAVSO is the largest organization of its kind-compiling observatiorrs d.oublestars there is not enough information to
Historically' double stars have
on variable stars from observers worldwide. These observations are digitized a:rd ical; therefore catalogs also include optical doubles
to discover and studv
processed and kept up-to-date. Since its inception is 1911, the AAVSO hzrs the also been difierentiated by the observaiional means used
spectroscopic, astromet-
largest database of visual variable-star observations, including data on about 3000 them. Thus, there are not only visual doubles, but also
pairs With the development of
variable stars having an amplitude of variation I magnitude or larger. The AA\rSO ric, photometric, interferometric, and occultation
are likely increasingly to
also has a database of photoelectric observations on 50 small-amplitude (srnaller modern ob"".ving techniques, however, these distinctions
than one magnitude) red variable stars. Copies of both visual and photoelectric become blurred.
generally are classed
observations may be obtained from the AAVSO on request by writing to Dr. Janet Multiple systems, containing more than two components'
a typical triple
A. Mattei, Director, AAVSO, 25 Birch St., Cambridge, MA 02138. as either ihierarchical" or "nonhierarchical " In the first instance'
at a distance man]' trmes
Other locations that act as repositories of photoelectric observations, togcther will contain a close pair and a third component located
separations of all the
with the contact Derson. are: the separation of the close pair. In the second instance' the
are Epsilon Hydrae
compon"nts are roughly the same Examples of these classes
Dr. C. Jaschek
(hierarchicall and the Trapezium {nonhierarchical)
Centre de Donn6es Stellaires visual pairs are usually
Nomenclature for double stars is extremely mixed.
Observatoire de Strasbourg although some ob-
designated by the discoverer's abbreviation and serial number'
11, rue de l'Universit6 pairs' Eclipsing pairs
s"rvlrs ha"," .regl"cted to number all or a portion of their
F-67000 Strasbourg to be no standard
are designated by the variable star nomenclature There appears
France or occtrltation birraries'
nomenclature whatsoever for spectroscopic, astrometric'
which are given a multitude of designations taken from o'ttio""
Mrs. E. Lake, Librarian -tut ?lo,g" , , ,^ ,"^
a visual double ls rne
With few exceptions, the primary (A component) of
Royal Astronomical Society are. system::i,:T
brighter star; lhe secondary (B component) is fainter' There
Burlington House distinction is arbitrarv Addttlonat
sm;ll magnitude difierence, however, where this
London, WIV ONL so on'.in order of increas-
i.t t -ultiple system are designated C, D, and
Great Britain "o-pon"nta is discovered in
ing separation. It sometimes happens that a new, close' companion
it as Aa' Bb' Cc'
an*already defined system. The convention is then to designate
etc.
 Jt/
516 EXPLANATORYSUPPLE[,1ENT IO / STARSAND STELLARSYSTEMS
ralues refer to intensities outside the atmo-
The Washington Double Star CattLlog, 1981.0 (WDS) u'as issued in tape for4. for atmospheric absorption standard
using the magnitude system traditionally used
It is the replacement for the Ind.ex Catalog of Visual Double 9tars,7961.0, issued ,ph".". ftt"y are defined and listed
as (-2 5log
by the Lick Observatory, and commonly called the IDS. The format of the \\rDS L! *trono-".r. The magnitude scale is logarithmic and is defined
is nearly identical to that of the IDS; both are cardimage. The WDS tape includes IntensitY).
---
a're defined using filters and
an introduction, the list of known double stars (numbering 73610), notes, and ref- M";; diflerent photornetric systems exist They
The first system, presently included in ?he
erences. The data file is being kept up-to-date, and a new tape edition should be d.etectorsthat give Jifierent passbands.
available in 1992. Currently, the WDS is ordered by J2000.0 coordinates, and t:g1- AstronomicalAlmanqcisthesystemmostwidelyusedbyastronomers.TheUBV
the first to be ttsed ex-
tain other format changes have been made to modernize the data base. Users can int.oa,t""d by Johnson and Morgan (1953)' was
effective wavelengths of
obtain tape copies of the WDS from either the Astronomical Data Center or the
"y"t"-,
il""*"ft. It has broad passbands (700 i000A wide) with
V, respectively lt is d"scribed bv Johnson
Centre de Donn6es Stellaires (addressesfor these are given in Section 10.22). S-6; A,"+aoo A, and ssoo A for U, B, and
to the system by Johnson with effective
The companion catalog to the WDS isthe Catalog of Obseruati,ons.This catalog ftgOr). t-u,"r, R and / filters were added
on fiI systems have
contains the differential measures of each pair. At the present time, the database l,r""f""r,n" of 7000 A and 9000 A' However' several variations
systems most widely used are those of
contains about 426,000 measures (means). It is complete for published observations f""t a"i"top"a and used by others The 'Rl
L/BVEl
since 1927, and largely so for those of earlier date, which continue to be incor- i"fr"*", et al. (1957) and of Cousins (1976)' The list of brieht
"ita-" was selected bv H L' Johnson
porated, This database is in a continual state of change, and consequentlf it is standards published in The Astronomical Almanac
(1966) An extensive list of faint UBVstandards was
not (and never has been) intended to be distributed as an entity. However. an at- iro- tft" fi"t of Johnson el ol
Se-
(i973) Landolt's standards were located in the equatorial
tempt is made to provide requestors with data on individual objects, provided that o"Uftrn"a Ot t-tdolt (1983)
A decade later' Landolt
"reasonable" amounts of data are requested. Iected Areas, convenient to astronomers worldwide
in the
UB l/Rl standards' These stars were also located
Eclipsing binaries are listed by \&bod et ol. (1980). Their list contains informa o"Ui*n.J " io, of 223 faint list The
of the stars in the first
tion on more than 3500 systems, and includes a small number of objects that, while same equatorial Selected Areas and included many
the 81data were tied to the R1
interacting and producing light variations, are not strictly eclipsing systems. The UBV datawere calibrated by the earlier list while
magnitude limit is 15 at minimum light. The left-hand page for each entry contains standards ol Cousins (1976)
Astronomica.I Almanac is tlae
identifications and approximate coordinates, maximum magnitude, depth of pri- The second system presently included in The
established bv Strijntgren
mary and secondary minima, spectral types of each component, epoch of priurarl. widely used system of four-color and H-beta photometry
centered
eclipse, and period. The right-hand page contains brief notes and references. (1963) and others. The passbands are of intermediate width (180 300 A)
b' and respectivelv' with a pair of
Spectroscopic binaries with orbits are listed by Batten e, dl. (1989). For the ;;;a A, ,;taA, *ioi, 5'1?0A for u, v, v'
""u the
the hydrogen Balmer line' In the 1981 tlrrough
1469 systerns included, the authors list on the left-hand page the identification and O.rrb""a. at 4680 A to measure four-
list of strcimgren
position, magnitude, spectral type, and reference. The right-hand page lists the 1990 editions ol Th.e Astronornical Almanac, the complete
Crawford and Barnes (1970) list' was publish"d
spectroscopic elements and a quality rating fcrr each solution. Following the ent ries color standards, included in the
(1966). Beginning with the
are very extensive notes and referencesconcerning individual s1rsterns.A some*'hat along with beta indexes liom crawford and Mander
(1987) was rrsed All 319
parallel effort by Pedoussaut el al. (1988) has resulted in many supplemenl,ary 1991 edition, the new list of Perry, Olsen, and Crawford
The new list contains
catalogs, which are valuable for their attempt to list all bibliographical referelces four-color standards are listed alon-gwith their beta indexes'
accurate data
to each object. Unfortunately, there is no modern catalog for known or suspe(ted several lainter stars (V = 6 7 mag ) as well providing more
,I.he Astronoii.c(rl et*anac are provided for convenience' but
spectroscopic binaries that have no calculated orbit. The tables in
always provide a complete
observers should keep in mind that the tables may not
photometric system is that it
list of stand'lrds. One requirement for a well-defined
in The Astronomt'cal
10.24 PhotometricSiandards have a high degree of internal consistency The tables included
data of
Almanac are chosen from *lti an efiort to provide consistent
"orr.."" selected from a larger sei of
Tables of photometric standard stars are listed $'ith their coordinattrs and $'ith their course, no set of data is perfect, and a sample of stars
Furthermore' observets
photometric standard values taken frorn the cited literature. Photornetric standard standards cannot completely define the standard system
of a system' such as the several
stars are widely used by observers because it is difficult to establish absolute cali- should keep in mind the difierences between variants
seek to match their own passbands as
brations of the telescopesand detectors used for stellar photometry and to accourrt fil systems mentioned previously, and should
 518 EXPLANATORYSUPPLEMEN. 1O / STARSAND STELLARSYSTEMS 519
closely as possible to the system passbancls.Data taken in one systcrrr can somctitD1.s Drrring 1985 1988, radial-velocity variations of tens to hundreds of ms-l n1d
be transforriccl to anotirer systenl as described by Bessell (1986), Bessell arrd \\:cis with periods ralging flom clays to 1'ears, har.e been shorvrt to occur irr so]ir,rarrd
(1987). and Taylor (1986) for exarnple.I1o*'ever,no gcneral tra[sformation exists later-type stars. both drvarfs(Campbell el al., 1988)and giants (Irwiu cl ol., 1989).
that can be a.pplied to a rvidc range of steliar tvpes and that is satisfactory at flr! Horvevcr. few stars irave been thoroughly exanlircd at this level of precision. Nlore-
1% level of precisiol that is desirablefor many purposes. over, since thc Iow amplitude variations may have very long periods. the constancv
As large telescopeswith n-roresensitive detcctors come into colllol I
use, talDtet of a given star to, say, t 100ms cannot be established quicklv. Finally, <leti led
stars are beirrg established as secondalv standards in many photometric systents. comparisons have revealed subtle systcmatic differences between the nraitr ilstlr
The needs of astronomers and the capabiiities of their irrstruments u,.ill corrtinuc to ments no$,in use. These must be understood bcforc delinitive velocitiesfbr indi
require cha[ges irr the stars used a,sstandards. vidual stars can be acloptcd.
The longest possible history of precise observations was the main factor iu se-
lecting candidates for a new standard systcrn. Hence, potential neu.standards rvere
10.25 Radial-Velocity
Standards chosenfronr existing standards within 1 20' of the equator. makiug them obserrable
from both hemisphercs, which eliminates proven variables. Tlte candiclatcs are indi
The list of radial-r'elocitv standard stars given in The Aslronomical Ahn.onat: inrc1 cated in the list giverr ir Th,e Aslronom,ical Almanac. Further obserr.ationsof these
their standard velocities are a subset of those adoptr:d by IAU Cornrnissiorr ilt) stars are being made, and a,rc strongly encouragedJto identily those thal, ir,resuffr-
(Trans. IAU. lX. 412, 1957; Trans. IAU. XVA. ,109.1973). Stars that have sincc ciently constant for later adoption as primary standards. and to define irrdividrral
been proved to be definite velocity variables havc been deleted, and a fet' velocitir:s mean radial velocities for the[r.
that are now kno$'n to dillcr lrom their IAU r'-aluebv rnore than lkms l har.c To avoid a proliferation of diffcrcnt velocitv systems, the olcl IAll s1'stem rvill
been markcd. Candidates for future adoption as primarv standard stars (see bckrrv) remain in force until the ne$,systcm (:an be adopted, even though thr: vclociLiesof
are also identified. V rnagnitudes are frorn Nicolet (1978) rvhcn possible: othernise. some stars are known to difler from thc IAU standard lalues listccl. Observations
they are estimated frorn photographic rnagnitudes and spr:ctral types. The latlcr ol minor planets have shorvn that thc lnean zero-point of the olcl svstem is colrect
are from the Bright Stars list (pp. H2 H31), the Yale Bright Star Catalog. Ltr'tltt' to within approxinately I 300ms I. Thcrefore. obser'.-ationsof scveral prinary
original IAU list, irr order of prefererrce.Positions are obtained by the plocecirrres standard candiclates should vield a fairly good meall \'aiue: rcliauce on one or t$'o
used for the Briglrt Stars list. stars only is cliscorrraged.
Some discrimination is neccssan in the use of these standard stars. for. tllo Finally, sirrce radial-r'elocitv obserlatiorrs of early-type stars are inherentlv less
follorving reasons. precise than of soliu and latel-tvpe stars. and becauseaccuratt-',indepenrleut veloc-
Radial-velocitv staudards ser"-e t1r'o basic purposes: (1) as colstarrt velo(rtl ity checks arc not readily al'ailable, satislaclory earlv-tvpe raclial-velo, ii.t sfar 'lards
stars with which to nonitor instr.ument long term stability, and (2) as stars of kDo\\l do not cxist. Effirtts to establish such stattdards are under rvav, but rvill tal<eycals
radial velocitl. rvith which to test instrumcnt zero-points. Over thc last dccadc or. to complete. Thc ptesent list is not suitablc frrl checkingthe zero point of radial-
so, the precision and cfliciencv of radial velocity measuremellts have irnpror-ecllrr' velocity rla,ta for early-type stars.
factors of 10 to 100 or more. A drarnalic increase irr researchbaseclon radial-velocil \'
data has followed. rnuch of the stellar r,ork relying on high prccision radial velocjties
10.26 SDectralClassification
(e.g., cluster dynamics, stellar pulsations,or lolv mass companions).
It u'as realized some tirne ago that the old IAU standard s-ystem lvas ilrad- A few worcls legarding spcctral classification may be helpful to uscrs ol the Brigltt
equate to meet tirese increase<ldemands (Batten 1985). Therefore, in 1985. lAI Stars list. A star's spcctrzrltype is au alphanumcric notatiorl that. rrhen appr"
Comrnission 30 crcated a \lbrking Grorrp rvhose goal u,as proposing a rrew ar1(l cau give quitc rcliable information on thc absr'lutt nragrritude,
priately calibrir,tc<1,
better system, il u'hich the zero-point and irrdividual velocities would be krrorvl to surface tempcraturc. ancl chemical composition of the outer lavers of a, star' This
approximatcly 1 100rns I . The plans for the ncq. systcm, and the results obtained notation generalllr involves a "spectral class" r'anging fiom O through Nl in or
during 1985 1988.are describedin thc reports of IAU Comlrission 30 (IAU Trans. der of decreasing surface temperatutc, and a "luminosity class" r'anging flom Ia
1988 aud IAU Trans. 1990). Unfortunatcl1..no new svstcm has yet been adoptecl. through Vb in ordcr of dcr:reasirtgintrinsic briglrtness. A few exceptiolally bright
The reasons for this are relevant to the use of the present standard stars, and alc supergiants are assigned luminositv class zero. Further, if a star's sp(itrunr apPears
described in the paragraphs that follorv. to shovu'a significant deviation in chemical composition from that of the Sun, an
 520 E X P L A N A T O BSYU P P L E M E N T 1O / STARSAND STELLARSYSTEIVS 521
indication of this abnormality is added. For the peculiar A stars, this involvcs slxrc_ idea of thc natlue of a star, allorving us to group objects of similal properties
ifying the elementsscen in abnormal strength (Si, Ntln.Eu. etc.). Thc sttrrswith a11 together for statistical trezrtrrcDt: and providing irnportant and intcrestiug objects
abtrormal heavy-elements/hydrogen ratio are often indicated by use of an clemenl for astrophysicai study.
abbreviation (Fe, Ca, etc.) follou'ed by a positive or rregative number denotiug tlrt The load liorn Secchi to Nlorgan and Keenan and beyond has becn long an4
degree of enhancement or rveaknessof the line of that element as seen in tlie spec_ winding, and its end is still now only faintly glimpsed.
trum. Stars of markedly urrusual composition, such as thc carbon stars (C) and tbc Thosc who wish to lealn more about the dcvclopment and fulurc trenrls of
very-heavy-element stars (S). as well as the \\rolf,Ral.et and white dwarf stars. all spectral classificatiorr should find the follorving relerences uselul:
have their own schemesof ciassification. R.H. Curtiss (1932)
Spectral classificationnotation can bc a bit confusing ,,n" or ,,s,' indicatcs \\r.\\r. N{organ, P.C. Kcenan, ard E. Kellman (19.13)
broad (rotation) or sharp lines respectively. and "e" indicates hvdrogcl emission. P.C. Keenan and W.W. Nlorgan (1951)
A "+" or u " follou-ing the spectral class indicates that a star is vcry slightly later. W.P. Bidelman (1969)
(cooler) or carlier (hotter) than the class given. while follou'ing thr: \\t.\V. N{organand P.C. Keenan (1973)
luminosity class indicates that the star is vcry slightlv brighlcr or fainter tharr thr: P.C. Kccnau and R..C.N{cNeil (1976)
class assigned.An "m" indicates that thc object is a metallic-line star, arrd a .\.." \\i.\V. Nbrgau, H.A. Abt, and J.\V. Tapscott (1978)
before the spectral classindicatesthat the object is a supergiant.giart. NI.F. \4cCarihy, A.G.D. Philip, and G.V. Covne (1979)
or dq'arf but has no actual \'IK ciassification. T. Schmiclt-Kahr (19E2)
The l,Iorgan Xccnan system (NIK) is defined b1, standard stars, of unir.ersali,y R.F. Garrison (198.1)
accepted spectral tvpe. Classiliers attempt to locatc their unknorvns rvithin therrnul R.O. Gray (1989)
tidimensional array defined by these standard stars. Unfor.tunatelv the results at
least as exentplified in the prcceding data is not entireh- satisfactorl., for ser,c:ral
10.27 Pulsars
rcasons. In the first place. thc types assig{red to rnost of the starrdar.d stars La\.{l
under6loneconsiderable change during the past half-centur1.,so that a gi\.en i[\'esti Pulsars are selected for Tht Astronornico,l AIrn anac list if their me:rrr .100NIHz flux
gator's deterurinations rnay be a function of tine. Second. the irrstrumentation use,:l density is greater tharr ,10nJ1'. Another handful are included thur.ttrre particularly
by the larious spectral type practitioDcrshas varied considcrably,rvith naturallv interesting for example, millisecondpulsars ol nemllers of binary slstems. \{ost
a very large attendalrt variation in thc accuracl- of the r:lassification attenpt.{l. of the data are taken from zr.cornpilation bv N4anchester and Tallor (1981). wlich
il not actually realizcd. For examplc, nost of the tr1>eslisted are from rather- in- contains references to the originnl literatrrre. Sone recent up<lates antl aclditions
hornogeuous slit spectra. but solne are from moderate-dispersion objective-prisnr are fron a nerv compilation now bcirtg readied for publir:atior (Lyne, trI:rnchesl<:t,
platcs. llany of the types u,ere dctcrmined long ago; othcrs are t-elv nelv. Thc') are and Tavlor).
certainly in no sensehornogencous. Pulsars are conventionally designatcd by their positions i1r the qluatorial coor--
In view of these situations, lhc best one carr rrou. do is to iist r.hat appear to bc dinate system,equinox and equaior of 1950.(1. given as light asceltsiollirl hours and
tho best ctrrrent spectral t1.pcs. Even the prr:scnt list r.ill be someu'hat out-ol cltrtc minutes and declination in degrees(or'. iu <:aseof arnbiguitv, dcgreesald tt:riths)'
\4'trenlt appears. In tnost cases,tire positions are measurcrdby treans of prtlse-timirrg obscr'\'atiorls'
Tlre calibration of \'Iorgan-Keenarr spectral types in terr|rs of absohrte Itagni- Arrival tirnes neasured on a number of clatcs, spread olcr'nlitlry months or k)nger'
ttrde, intrinsic color, and surfacetemperature has also beerrsubjr:ct lo r:onsitlcr:rblc can provide astrometlic accuracieswell unrft:r a secondof arc for "nortlral" pulsars'
change over tiure. The rrrost uselul rccent reference is Ludoubtcdly thc extt:nsivc and at, or below. the rnilli-arcsec range lbr rnilljseconcl puisars (Rarvlev. Tavlor''
work of Schmidt-Kaler (1982). Calibration of the third (aburr<lancc) pararncter of and Davis 1988). The same observatiorN yicld high-prctisiou vahres li)r tlrc ptrlsar
the system is currcntly in a primitive stalc. partly becarrsethr:r-r:is rrot yet cornplete period.P , , a n d p e r i o dd p l i r a r i r p .o r ' p i n , l , , r v rrra t e . .P .
agreementamong the specialistsin the field as to details of te<.hniqueand notalion. The group vclocity lbr netcr rvavelength signals trareling through tlte intr,'rstel-
It should be clcar that everr an accurate spcctral type is uo substitutt: fi'- a lar urccliurnis k:ssthan c (lhe specdof light in vacrnrm).bec:artse of the Prescnceof
cornpletely rcliable spctral analysis and absolute-magnitucle determination. But il an ioniz{xl gas component. Thc velocity is also frequency<lependent,according tr
the stellar astronomers' real-r'"'or.ld,spectlal t1'peshave proved, and still are provirrg. the relation
extremely useful in euabling us to obtain quickly a good. albr:it necessarilvrough. v = c(l - 41 L,')tt2, 7 l)
(1 0 . 2
 E X P L A N A T O FS
YU P P L E M E N T 1O / STARSAND STELLARSYSTEI4S 523
where i, is the plasura frcciucncv and ru is the frequency of the rrave (scc. fbr cx- bound. irr which case theJ' survive to become open clusters (the ages of $'lich range
ample, l{anchester and Taylor. 1977. Chapter 7). For a giverr pulsnr this efccl i5 from less than 1 million ycars to more tharr 1 billion years). Open clusters a,requite
directly measurable fron tiurirrg observations at separatcd lrequencies. It is qrrarF distinct from globular clusters, u'hich are ah'ays dynamically bound ancl are several
tified in a coustant called lbe dispersion neosure, rvhich amounts to thc integral ol billion ycars old. Open clusters form a. flat disk subsystem' u'hereas tlte globular
free electron dcnsity along the linc of sight. The conventional units are cm I pc. clusters fonl a[ extended spheroid subsystcm
Irregularities in the interstellar medirrrn, togclher $,.ith the extrenel-y small In the table of selected opeu clusters, therc are a nunber of associaLionsthat
eflective angular diarneters of pulsars, causc arrother propagation effect not gencralh. might not clevelopinto physicallv bound clusters. The Trapezium cluster and sorle
seel in radio obserlations of othcr types of sources.Interstclla.rscintillatiorr. t]1, of the Collinder obiects belong to this category. Othcrs, such as the l3lanco ancl
resuit of multi-path propagation through an irregular medium, can carrsc larger. Upgren clusters. are considercd physical entities becausetheir rnembers have similal
(order-of-magnitude) r,ariations in apparent flux dcrrsity as a function of botli tirnc motions and physical propertics, although thcy are only a small part of thc total
and frequency. Consequently, flux derrsity measurements made ovcr a short tinrc, stellar poprrlation inside the volumc they occupy. An effort has been nrurcleto cxclude
and a [arror4'' frequency range may differ substantia]llr from Lhc tme mean flrrx from thc catalog such specious clustcrs that are in fact only the appalent result of
densitl'. For this reason. rnost tabulated pulsar flrrx derrsities ale subject to ratht:r. the distribution of obscuring clouds.
large uncertainties.t1'picallya factor of 2 or so.
10.312 Open Clusters as Related to Galactic Structure Thc sysiem of open clus-
ters provides thc best rnaterial for the study of the disk of our Galaxy There
10.3 CLUSTERS
ANDGALAXIES are scveral reasorrsfot this. Distances,ages,alld interstellal cxtirrctionsarc bcl,ler
known for open clusters than for other objects; open clusters havc a rattgr: of apJes
10.31 OpenClusterData that nake them particularly suitr:d for studl' of the recent el'oluti(m of the galac
tic system: and Jounger clusters conglegate in spiral features ln fat:t, photomclric
The data selectionconlt ned in the table "SelecttxlOpcn Clusters" is a subsetof (i. studies of open clusters are ofte[ carried out irr the course of investigations of galac-
Lyngti's Zznd Co.talogof Open Cluster Dala (lifth cditidr. 1988). This computer., tic strlrcture. and lltis also shorvsin the selectiortof obiects Thrrs, a lot of I'ork has
based catalog is available frorn the \\iorl<i Data, Ccrrtcr A or.from Cerltre de Donn6cs been carried out on younger clusters and on \-erY old operr clustersrcveu though
Stellaires.The fifth edition of the Lrrnrl cai.rlog rcprcsentsa nerv approach to tlrl some quite close-by open clustcr-sof intermcdiate age are still relativr:l1' linknown'
prcsentationof availableclata.\Vhen severalclata valueshave been publishedin 1he
litcrature. the calnlog contains rveiglrtcd:rvcr.ages of these.The weights are clcrilerl 10.313 Open Clusters as Related to Studies of Stellar Evolution Nlanl oftht: break-
objcctivelv using methods deviseclbl Jarrcs,Tille1..ancl L1'ng5. (1988). Thc <lalrL throughs in the study of stellar cvolution havc been aided lly obscnatiorrsof L'peD
collcction lbr this rvork en<id in Scpternber 1986. 'fhe
clusters. importatt point is that the clustcr members are ali of approxilrrzrt"lr'
The Lund ca,talogcontairrssorle 20,000data values for a total of 1100cltlics. the same age ancl chernical cornposition. Thc ortly fundarnental differencc is in the
A scparate file incluclesuror-cthan 500 referencesto authos and titles of prirnatr 'fhe
mass of gas from t'ltich each mctnber is formcd fact that rnor(: nassi\c slirrs
sorLrces.?he -;lslronorn.knLALn.an.aclable containsa selectiorlfronl thc Lrrnclc:ataLrg get higher centlal temPeratrue. higher luminositl'. and a faster rate of cvolutioll
of rclevarrtdata valrrcsftr.the best studicd clustt:r.s. than lessmassivestals. ]ras beerr utilized for trost of the lundamerttal studics of
the principles of stellar evolutiorr. Practical, observationaladvantagesalso resull
10.311 Definition of Open Clusters The clumpirrg of stals in space is a pllsicrl from the fact that tltt: t:lusterstars can be assuru:dto have the sanredistan' c fr"trL
reality. Probably rrrorethan 5(l% of all stars are consiclereddouble or mult iplc st;rts. us and. usualll-. show the same intcrstellar extinction
Gloulrs of ten or nlore stars are associationsol clusters.Becausethe stiirs ilr ca(lt
such gloup sharr:origin and cr-rrrditions. ther- are of great use in the study of str-'liat 10.314 Descriptional Data for Open Clusters An operr cluster is dclined bv its po
properties and iu thc stud1rof propertics of the larger systernsof thich thel' lirtttt sition in the sky ancl bv its apl)arent angulal tliametcr. Each cluster-hasalso tradi
pafts. tionally bccn given one or several [arnes refelring to various catalogs or discoverers
Associ;rtionsare stellzrrgroups that are less than 1 nrillion years old. Ilrr:s<' In fact. lor some objects thero is quite a profusion of different denotations To hcip
groups iue ofteu clynamically unbourrrl systcrns t hat clisper-scinto single stius, (lou alleviate conlirsion. the chrster collnnission of the International Astronornical Llniorr
ble stars, and snall groups. Occasionallr'. irolvevcr_.the associations arc dyrrarni<allJ' dcciclecljl 1979 on a llolncnclature consisting of a "C" for cluster, four digits tor
 524 EXPLANAToRY
sUPPLEMENT 1O / STARSAND STELLARSYSTEMS 525

hours and minutcs of riglrt ascensiol (1950.0)arxl a plus or nriurrssign plus tlrlc,e distallt stars or slar s)stems. Fol tlte'clistance of zr nearbl clttstt'r rvith a lorlg main
digits for dcgreesaucl tenths of ckrgreesof declirration (1950.0).In the Lurrrl opt'rr seqlrcrlcc. alr imprccision as lorv as 10% carr bo achieved'
clustcr catalog.the IAU desigrrationirnd the most commonlv rrscclrlarucare eur(! c(1 Iu connectiorl rvith each plx)tolltctric stud\'. il dcterminatiolt of ther illtorst{rllar
for each cluster. The current equatorial position is also entered. extirlctidl in frotrf of the cluster also has to b<: Itrade Thc ttsttal tncthocl is 1o de
Thc published vnlues of anllular diametels ol clusters sornctimes diffor bl a tenninc thc colol excess due to thc extinction iut<l assume tltat the reddelrirtg ratio
large factor. This is partly due to cliflerentlimiting magnitudcs f<t different strrtlies betlveen the extirrtiolr in visual liglrt and thc tolor excess is:J.l A coutPli<trtiort
but also to diffcrent opinions abollt the extetrt of cach cluster. It is important for. is that the extinction varies oYer the cluster in sone cases. Allart fr<lm tllosc clus-
compara.tivestudiesto have a colrsisteutvalue fol thc diameter in the fifth cdiliol ters, and if the asstttnption aborrl reddening lalio is corrcct, the extincti(nl call be
ofthe Lund catalogtltis was accomplishedusing lorrgexposrrrclrlints and films fiorrr deterrnined to within about 0.1 rllagnitudes.
'fhc
the existing seriesof Sky Surveys. angular (liiunetersarr: in rninutesof arc atrtI Frequentll-, tlte same investigzrtions that clcttr ltitre the clistiurcesto (rlrrstols will
refer to the cluster coles. also provide estiDtatcs of their agt's lt1' using th<' turn-off poittt of the ttraitt srtlrt, trro
During the exanrilation of thc Skt' Surr-e1'Chiuts. Trumplcr classesrvcrc also from the zero-age lrlain sequencc. This is the lllost relial)l(' nrethod. artrl rcsults
estirnated. The classificationscheux:proposed by R.J. Trrrn4rlcr (1930) coulairs from it are given iIr ?hc Astronomical Almonac I able. To hitvc a ltomogc:Ix'otts set
the followins classcs: of cluster ranges .Ianes. Tillcy, arrd Lyngd dcrivr:cl a calibraliorr of tlx: Lrgarithm
of agc as a lunction of the turtt-off color' Tllc listcd values arorlrld 6'0 i c ' olrc
Concentrationof clustel stius million lears can. iIt most cnscs. bc considerc(l as rough uplx'r'litnits. c'lttslcls of
I dctathed. strong ceDtralconceDtratioD that age group u)iglll still be irr tltc process of star firrmatiolr arr(l thus ha\1' a largc
II detaclxrd,little cerrtralconcentraliorl spreacl ir age betw(lou lnernber stius. Clusters with logaritlrrrric ages higlx:t t'ltan
III detaclrrd. rro noti<cableconcentration 7.0 havc an impr<r:ision of a few l,crrths of unit.
IV not u.cll detached,appearing likc a star-field (orr<eltration Data for the lllzrguitucle of tllc brightest clttster tnembers ltrd lbr the spcctral
class of the earliesl-tlpc cluster lltcDlbers ar.e failll' urrambiguolts. The onll Purlrlem
Range of stellar- brightness to
lies iD the possiblt' inclusion of lx)nnrclnbers \aturalll'. effor-ls llate been Irttldc
I most sl,ars of nearlv the same briglrt[ess
avoid this source of t:rror. The tot;rl trragnitudcs of tlte cluster-sIlave been t::rlcrtlated
2 mediurrrrange in lrrightness 'fltis
t'as
by aclcling, Iogarithtnically, the tttirgrritudes of trll krtorl'n clusll<:t ttternbers'
3 bright arrd faint st:r.r'sin the clustcl thc author lrv B Skiff of Lou'ell
done an<l the clata privatell' colnnrullicated l()
thn
Observatorl'. Thc opert (luster tal)lc of The 'lstrottttntictr'l llnanoc also lirts
Richnessof clustor tltatt 30
value ol metallicil.v, t'hich rvas <lctcrrnined bt rarious mctllo(ls in ;norc
p poor' Tiilel''
difiercrrt investisa.l,iotrs and tlartsforred to a c(tlsistent set of clzrttt bv Janes'
m modcratcly rich ratio of the
and Lyngi.. The rlcfirrition of tlrt' rrretallicity r;rtio is thc logatilhmit:
r rich ratio fol tlr" st-'lar
abundan.es of irott alxl hvdrogctl dil'ided bv lhc <:orrespoudittg
sDectrulrr. In pra<ticr.. nretal lirrt's ()tl)el than lhost'ofirorr ilrfltt<'l1t' th" PaIjtlrLcf"r'
\ebulositl-
n clustcr is irrvolvedil nebulositv

10.315 Derived Data for Open Clusters Data ( ()rr(erningclistiurtcs. extin(li()rLs

and phlsical prop('rtics ofopcn clustcrs are dclcllrrined during <qrritr: ('xtensi\f ill-
oPer
vesti€lltioDs.It is forturrate that so r uch rvork is rit'r'oteclto srrchlasks all orlr'tltt' 10.316 The Reliability of Open Clust€r Data Dcspite greal {'ff{)rts to assoss
world. The rvork of l lrc <latacompilcr is to cornparr.t he publishr:<lvalucs,to alrirl\'zr' clustcr data criticallv, it is quitc clcar that thc sct can never lrt: quite hontogcltcous'
<rn-
their rdinbility, ancl 1o find thc Lrcsl available figur.c for each pru.ulcter. The variations itr brightness of tlrc objects irr cr'owdcd fiel<lstr'trclin observiltg
catalo!! scvctal
thr: distal<r' of an o1x:r ( luster is to ol)scl\'('
Tht: utost reliabk' rvay to detelrrrint-' ditions contribute to an intrinsi<rvariation in quality ln the Lurrd
its colot-tl:rgnitudr: rliagram ancl fit the dia€lratr l() the rnain sc(luellceof lrcarl)\ of the phl sical par-?lmetershalr' bcen assigne<ltrrights that <lcsr:ribein a cotttl>ara-
stars. \\'lten a cluster irrcludesa rurubcL ol uue\'()l\1'(lstars orr tll(' tlaill se(luclr(('. tive fa-shionthc precrisionof the <lnta.The prirrcil>les for *'eight assignmentalttl tlre
this nrcthotl of det<:r-ruirrilg
distalrccsis in facl, srrPctiorto othcr rrrcthodsusc<llirr' analvsis have been described in Janes, Tilley, and LyngA (1988)'
 vfi"
:;
EXPLANATORYSUPPLEI\.I
ENT ANDSTELLAR
10/ STARS SYSTE]\TIS 527

10.32 Globular
StarClusterData value) arc seen to be in ncarll circular orbits that are confined to the galactic <iisk.
The metal-poor group (the prepordcranceof the 146 knou'n globulars)appcar to be
The data on globular ciusters given in ?De Astronom,ical ALnt.ano,c ate derived liorrL mof ing irl highlv eccenh'ic elliptical pa.ths arorrnd the centcr of thc Galaxl'. in orbits
a variety of sourccs.The list of clusters itself is from the IAU's "Star Clusters with planes lhat nay be highly inclined to the gcneral plane of the clisl<(although
and Associations," r'hich incorporated the omcial IAU chrster designations(the the orientation of the chrsters'orbilal planesto the disk seemsto bc quite random).
four-digit, sign, three-digit nurrrbcrsprefixcd by a "C"). Thc original IAU list has The ciuslcrs orbit the galactic center in zrconventional right-hzrrrdedscnse.the only
been supplementett by R.E. \Vhite as a result of his triennial literature searchcs to exception (so far) being NGC 3201 (C1015-461),u'hich appearsto be in a tctrograde
prepare the "\Vork Published and In Progress" cluster-by-cluster summary for e.r,cll galactic orbit.
IAU Reports on Astronomy sirrce 1975. Thcsc rcports ale rrot compilcd after tlLe
1988IAU General Asscrrtbly.Instt-'ad.J.-C N4ermilliod,at the Genera Obserratorr'.
continues to maintain a database of literature o\-er open and globular clusters. A
-fucson. 10.322 The Importance of Globular Clusters for Galactic Structure I{istor-ically,
copy resides at the Natiorial Optical Astronomical Obserratories in AZ.
the globular clusters \4,ereused by Harlow Shaplel'to determine thc Sun's dista,nce
The clusters'cquatorial coordinatesarc taken frorn Shau'l and Whilc's (19lit-l)
from the center of our Galax.v. Shaplcy's rvork led to thr: realizatiorr that our Surt is
listing (preferred) or lrorrr thc cornpilatiorr by \Vcbbirrk (1985). Thc axial rartio clata
not in the Galaxy's ccntelJas had becn cheerfullYassutred11pto tltcn. but is located
(b / a) valrresale taken exclusivell from \\-hitc and Shal l s (1987) extensivelist.
about tri'o-thirds of thc radial distancc outt'ard from the ceuter. Tlrr: circunstances
that led to Shapley''s discovert'. namely the Presence of Rll L1'rac r,ar..izrbk: stars
within many of the globulars, still makcs these clusters intportant fol large-scale
10.321 The Definition ofa Globular Star Cluster Thc cclcstial objccts called qlol
dynamical and kinematical strtdies of our Galaxl'. Becausc rnany external Elalaxies
ular stdr clust.erscannot bc identified by a"casual irrspection of a photographic plate.
' also show cvidenceof a globrrlar star clustcr population, the Calaxr-'s clustcrs are
instead certain criteria nust be met. Foremost among the "physical-appearaDce
also useful as secondarl- distiuce estimators t() these more distant svsterns
critcria is a cluster's contrast against the local field: The globular star cluster shoulcl
be secn ir,sa higher irnage-density obiect than the surrounding neighborhood of ficlrl
stars. The cluster should also appear to be somewhat conpactlv organized $,itirilr jts
ot'n image. In addition. a gradient of decreasingareal star densitl'u'ith increasilg 10,323 The Importance of Globular Clusters for Stellar Evolutionary Studies Be-
radius from the cluster center should be apparent. causethe marry stars in a glolxrlar cluster arc at the samc distance frort thr: Sun'
Globular star clustersofterr coutain short period variablestars of the RR Lvra(' a globular clustcr CN,ID involvirtg apparent (r'isttal) magnitrtcle versus visual color-
class;open star clustersand stellar associatiousjhorvever.never contain such stars. index reveals thc same pattcnr !\'ithin that paralnetric spa(reas t}xr equir':rlent <lis-
The brightest stars in globular clustcrsarc rcd giants, arrclarc scernspectroscoyricallr' tribution of stellar characteristics lvithin the lumilltlsity vcrsus srulace-tcnlperiitur(l
to be rcak in their abunclanccof hcavl elemcuts whcrr coulDale(l to thc solar' parametlic spacc. l'herefole,thc globular clustcrsprovide the thcorist u'ith lest-becl
abun<lanceof same. situations for corttparingrrretal$,eak model stars:rgainst lot'rlletal obsclmtio[a]
However. amongst the globulerr clusters resiclent in thc N{ilky \Vav Galaxr'. data. The 6llobularclusters havc been re\-ealcd.rs being some of tht: oldest objects
cxceptions exist to all ol lhe aforementioned criteria. The orrlv unambiguous de- in the Galaxy, t,ith a€iesraugitrg frorn 10 to 15 billion voars. Hellce, obscr\?tional
termination of "globular cluster-ness"is made usinpJthe ciuster's color nlagnilrr(lc studies of stellar associationsprovide cornpar.ativedata ftrr yotrng. mtrtal-richstars;
diagram (C\{D) globular clustels har,'eCNID's that range in thc valrtes ol (B-V) aud open clustcls are the sourcc of data fbr rnetal-riclt stars illteunecliatein age'
color frorn, approxirnatcly. -0.6 to +1.6.'Ihere is, roughlv. a 2 to l3 rnagnil,urLt' Globulal clustcrs complete thc obst:r'r,ationalsct by providing asl,rophvsical pararrl-
difierence betq'een a globular cluster's main sequence turn-ofl and the apparcltt etcrs for thc old and metal-poor stars forrrcd in the initial evolutionary stages of
magnitudc level of its horizontal branch (HB). A sirnilar magnituclc, rangc exists our Galaxy.
betll'een the tip of the red giaut branch and the HB. although this difierence is
subject,to variation as a function of a chrstcr'sinl,rirrsicrnel.alabundance:lhe rnor-c
metal-abundant, fhc smaller the diference (but not lessthan 2 mag.). 10.324 Basic Referencesfor Globular Clusters Studi€s There has rtot bcen a surn-
Kineutatically. the globulars f<rrn a two-t:ornponent group, again as a funcliotr mary monograph writierl on the general subject of the giobular stars clusters for
of thcir metal abundances.The metal-rich group (between t 0 and 50% of the solal over 25 J,ears(Arp 1965; Hogg 1959); however, becausenerv information on (:lusters
 528 EXPLANAToRY
suppLEvENT 10 / STABS
ANDSTELLAR
SYSTEMS 529
is developing so rapidly. and often in such urrexpccted directions (r:.g.. X-ray emis- a, b) ol better. the luminosity inder L = (T + L) I l0 (dc Vaucorrleurs1979) is
sion lrom point sources n'ithin clustcrs), writing an ali crrcornpassing rnonograph correlateclu'ith absolutc rr.ignitude; both 7 and I are listed in the tabk:. The L
'fo
would be futilc at present. come to grips I'ith current states of rcscarch arcas scale. as used irr RC2, runs from l for a bright giarrt galaxy, to 9 for a faint drvarf
concerning the clusters, one must go prcferentiall.v to the <olloquia and svmposi6 systerrr. The precision of Z and I values is on the order of one rtlit or better. For
held under the auspicesof the IAU. details, seedc Vaucoulcurs(1977) and de Varrcoulerrrsel al. (1978).
General suDrmariesorr the subject are givcn by Arp (1965) ancl Hogg (1959). Tire isophotal apparent nrajor diametel of galaxies is given in 1he table for Lhe
Data collectionsarc found in Krrkarkin (197a), Hogg (1973), Philip et al. (1976), isophotc Ievel 25.0 B magnitude per square arcsecond(B-m/ss). Thc level corre-
Shawl and White (1986), and White and Sharvl (1987). Recent colloquia ard svnF sponds roughly to thc visible size of the galaxy image seen on the Palomar Ob-
posia are IAU ColloquiurrrNo. 68 (Philip and Hayes. 1981). IAil SyrnposiumNo. servatory Skr-Sun'ey paper prints (bhrc light), errrdto lhc diametcrs liste{ i1 the
113 (Goodman and Hut. 1985).NATO AdvarrcedStudy Institute. 1978 (Hanesarrrl Uppsala General Catalog (Nilson. 1973) or northern gz axies. Notc lhal isophot,al
Madore, 1980), IAU Svmposiunr lio. 126 (Grindlay arrd Philip. 1988). and IAU diamcters iLre loot,nLetric diameters. The former depend on the surfa<r: brightness
Symposium No. 85 (Hesser.1979). (specificintensity) of the galaxy image. The latter do not depen<lon srrrfacebright
ness,and whilc lhcy arc bcllcr indications of the trlre sizeof galaxies.1he-varc more
10.33 BrightGalaxies dilicult to measurc and .irc Dot ]'et available fol ntany galaxics. For urore delails,
see the Introduction to RC2. Note that to ar-oid ucgalivc logarithms. D25 is giverr
The 194 galaxies selected for inclusion in The Astro,rLm.icol Almanac are gencralil in the table in units of 0.1 arcminutc.
brighter than B-band magnitude 11.5 and/or largcr Lhzrn,1.5 arcminutes. Thesc The ratio of the major and ninor axcs itt thc isolopc lcr.cl 25.0 B-rn/ss (R25=
galaxies $'ere extracted from an up-dated version of tlrc SecondReference Ca.to,loqtLt D25I d25)is also listed in logarithmic form. The precisionis usually better than 0.1.
of Bright Galaxies (RC2) (dc Leucouleurs et o.l. 1976). The isophotal axis ratios given in the table are corrcctccl Ibr thc svstematic errors
A galaxy's rnorphological type is based on thc revised Hubble system clcscribecl that affect visual rneasurements of photographic plates (Holmbcrg 1946, dc Vau
by dc Vaucouleurs(1959a,1963). The system includes four classes:ElJipticals(E). couleurs 1959h). (For more details on systematic and accidcntr crror-sin diameter
Lenticulars (L), Spirals (S), ancl Irregulars (I). In the L and S classes(disk galax and axis ratios, seePaturel el ol., 1987.)
ies). two families. ordinary (A) and balrctl (B), are distinguished,with tralsition Thc total apparent B-band magnitudes, Bi, arc gcrrcralll' dcrived from photo
t1.pesdenoted AB. Each famill' can cxist irr ttr,o r,arieties. dependilg orr thc abserlce electric mcasurenleuts or detailed photographic surfacc photorr(r1.ry in the Johrtson
(s) or prescnce(r) of an inner rirg structure. with transition tvpes dcrrolcd (rs). Lr B syst.'lr. Thcse magnitudes are rveiglrted ncans of lhc bcst ar,ailable maguitudes
the L class, the rclative importancc of the trl.o major componelrts ol:r. galax.r' thc extrapolatcclto infinitl' to allol, for the generally srrtall fractiott of the iurrrinositl
spheroid (or buige) and the disk is denoted b.v the superscript " " (car-lv.i.t:.. contribufcrl bt tlre olttermost regionsof a galaxl' zrscxplained 1t the Introduction to
most similar to E), '0" (irrterrnediate),and "+" (late, i.t:., trore similar to S). TIre -llv:
RC2. pr-ccisiortis generallv0.1 mag or bcttcr. (For tletails. sec de \''aucorrleurs
stage along the Hubble scqucnccof spirals is denoted by a. b, c. and d in orcler-o1 and Bollirrgcr. 1977.) li)r sorneapplications. thc obs<:rveci magnil,rrdesn(xrd to be
<lecreasingbulge-to disk ratio and increasing devcloprncrrt of the zrrrns. The trau corre(t(l for galar:li<r ancl irrtelllal extinction. \\hich are hoth conttovcrsialanrl un
sition type bet*'eel L and S is denoted SO/a; betrveen S ancl magellanic irrt:gulars certain b] 0.1 to 0-2 rrag or rnore in extrerne cases(t:.g.,an crlge-ortspiral galaxv
(Im). by Sm. Thus. lhc Large NlagellarricCloucl is the protot\-pe of SB(s)rn. anri near thc galaclic Plarrc).
the Andromcda gal.xv (\{31) of SA(s)b. An outel ring structure is designateclb1' The total <rlor intlicr:s,(B [1 and (U ,B)7. in thr' Jo]tnsottU. B. V sl sl erns
the srrnbol (R) preccding the Hubblc-stage designation. The puzzling norrnlagel- are as obscrvccl:i.c., urrr:orrt:ctc<lfor galactic and intcr-lal extinction.'[he raltge is
lanic irregrrleus,of which NtI82is thc prototvpe, are desiglatr:d I0. Edge-onsl.stcnrs rouglrlf irom; *1.0 1o *0.3 lor (B V)r, and+0.6to -0.3 lot (U-B)7 tis onc rloves
are sigrraiedby the suIfix "sp" (for "spindle"). Peculiaritiesarc noted by tlrr: suflix along thc Hrrbbk: stxlu<lcr:frorl E (yellorv-orange)to Ini (n'hitish) through L and
"pec." For urore detir,ilcddiscussionsof galaxv morphokrg5,. Sarrdage(1975) ; ld
sr:<: S typcs. l31 an oversiniplification, the extreme tJ'pcs are o[1en dcscribr:d as "red"
Buta (1989). and "bhre" galaxies.The colors givc crudc incli<rations of the encrg.vclistribution
Because plrysical propertics. such as color iucir:x, irvdrogen index. bulgt: to-tlisk in the contimrum and. hcncc, of the stellar cornpositiou of a galnxy llorvevcr, thc
latio, and so on, are closell' corrclated with stage along the Hrrbble scquence,it is colors are oftcn rrrorlilied lrv linc etnission fiorrr HII regions (n'hich d<x:reast:st]x:
convenient to attach to thc stage a nurnt:ricralsca]c. I, frorn -5 at E. througb 0 coior iutlex, or incleasesthc blueness)and iuternal dusl,extinction (t'hich incrcases
at S0/a, to 10 at Im. The luminositv class,a, introduced by varr dcn Bcrgh (196t) the color indr:x, or increasesthe ledness).
 530 EXPLANATORY
SUPPLET
ENT 10 / STARSAND STELLARSYSTEMS
531
The heliocentric radial velocity, derived from optical observations and 21_cnr casesonly a single emission-
lines are used to determine the redshift, but in some
neutral hydrogen radio emission, is given in kms-l and calculated with the opticai be the hydrogen Lyo line at )o = 1216 A' the
line is present (usually assumed to
convention that A) / ) = V / c. For the small redshifts involved, this convention is In such cases, the redshift is
strongest emission line in a typical QSO spectrum)'
= 0' 1 to > 4' Note that
entirely adequate. For a discussion of systematic and accidental errors in radial ve_ evidently uncertain. Measured QSO redshifis range from
Iocities, see the Introduction to RC2. The ndial velocityr y0-corrected for galactic objects that (apart from their
nuclei of many galaxies at still lower redshifts contain
rotation according to the old IAU convention, Vo = V + 300cosA, where A - the are essentially identical to QSOs. For his-
Iow redshifts aIId modest luminosities)
angle to the solar apex-is also listed. For cosmological applications, it is better galactic nuclei (or ' more specifically'
torical reasons, these are usually called actiue
to replace the conventional solar motion of 300 kms-l toward I = 90., D = 0. (see Type 1 Seyfert Nuclei) and are not considered to be "true" QSOs'
Ttrans.IAU, XVIB, 20i, 1977), with the total velocity vector of 366 kms-l toward In addition to the broad emission lines, many QSOs (particularly those with
I = 265" , b = +554 relative to the cosmic background radiation (see de Vaucouleurs The most common lines are
large redshifts) have a host of narrow absorption lines'
abundart heavy ions
et al, 1987), For more detailed discussions of solar motion relative to difierent ex_ du" to hyd.ogen Lyo, but the UV resonancelines of cosmically
alwavs have redshifts
tragalactic frames of reference, see de Vaucouleurs and peters (19g4). (".g., C"3, Mg*l) are also often detected. These lines almost
More up-dated data on bright galaxies can be found in lhe Third, Reference the emission-line redshift. They are believed to be
ihut u." -.,"h smaller than
Catalogue of Bright Galanies (de Vaucouleurs et al. !99I\. gas clouds'
produccd in intervening galaxiesor intergalaclic
have been studied
Although the spectral-energy distributions of some QSOs
available (typically UBV
in considerable detail, most have only very limited data
10.34 QuasFStellarObjects magnitudes and colors). Radio-selected QSOs have radio flux densities at several
(o) also available The radio flux densities
ft.fu"nci"" and a radio spectral index
A quasi-stellar object (hereafter, QSO) can be defined as an object with a large 1 Jy = 10-26 Watts m-2
axe given in Janskys (Jy) or mrlli-Janskys (mJy), where
redshift whose optical morphology is ,,quasi-stellar,,' (or ,'starlike', in the sense is the flux density at a
Hr-i. Th" spectral index is defined by S(v)av-' , where S
that it appears as an unresolved point of light in a typical optical image). Note
frequency tr.
that the terms QSO and quasar are often used synonymously, but the latter refers
ThemostextensiveandrecentcompilationsofQSosandtheirbasicproper.
technically only to the subset of QSOs that are strong radio sources (quasar being
tiesareHewittandBurbidge(1987)andV6ron-CettyandV6ron(1989).Agood
a contraction of quasi-stellar radio source). weedman (1986).
introduction to the basic phenomenology of QSOs is given by
QSO names are as diverse as the surveys by which they were discovered. Most
QSOs are found either by virtue of their radio emission or in optical surveys designed
to select objects with strong emission-lines and/or unusual colors. The
eSOs are REGION
BYWAVELENGTH
CATEGORIZED
designated by a number internal to a survey (e.g.,3c 279 is object 229 in the third 10.4 SOURCES
Cambridge radio survey and UM 148 is object 148 in the University of Michigan
optical survey) or by appending the approximate 1950.0 coordinates to the acronym 10.41 Radio-sourcePositionalCalibrators
denoting a survey (e.g., PKS2203-18 is taken from the parkes radio survey and is from
The list of 233 radio-source positions, given in ?De Astronomical4lmanoc'
PG0026+12 from the Palomar-Green optical survey). Note that there is wide range group under IAU commis-
Argue er ol. (1984). This list was compiled by a working
in the accuracy of the published coordinates for of extragalactic objects that have both
eSOs, but a significant fraction sion 24 as a first step in defining u
(especially the radio-selected QSOs) have positions known to better than one arcsec. "u"log from a number of
radio and optical counterparts. -Jo"itioo" l" ttt" li"t were compiled
The most critical parameter for a eSO is its redshift, '-fhe origin of right ascension is-defined bv the right
previously published catalogs.
ascension ol 1226!o23 (3C2738) at epoch J2000 0, 12h26^O616997 ' as computed by
for the source
z = (^ ,\o) / .\0, (10.34-l) Kaplan el al. (1982), and based on the 81950 0 position determined
is given by the
by Hazard et al. (7971). An indication of a position's uncertainty
may be subject to
where 'l0 and,\ are laboratory wavelength and observed wavelength respectively nurnber of digits in the tabulated coordinatesl the end figures
in Janskys at 5 GHz'
of a pa.rticular spectral feature. QSO redshifts are almost always measured using revision. The column headed 5569" gives the flux density
flux is meant to
the strong, broad emission-lines of hydrogen, helium, and other heavier, cosmically Fluxes of manv of the sources vary' however, and the tabulated
abundant elements that are the trademark of a QSO. In most ca.ses.several emission serve onlv as a rough guide.
 SYSTEMS
10 / STARSANDSTELLAR 533
532 EXPLANATORYSUPPLEIT,{ENT

10.42 Radio-Fluxcalibrators 10.43 X-RaySources

The calibration of radio-source flux densities, in absolute terms, is an essential palt The X-ray sources given in The Astronomical Almanac rvere selected from those
of observational radio astronomy. The basic absolutc calibration requires the use of having optical, raclio, or infrarcd counterparts listed in an unpublished Cotologue
small antennas, the gain of which can be theoretically calculated or measured (so1 of X-Rag Positions (Dolan, 1983). The master list of sources in this srtrvcy file
Findlay, 1966). Such small antennas can be used only to determine the flux dcnsities is available on magnetic tape froru the Astronomical Data Center. Mail Code 633,
of the strongest three radio sources (Ctrs A, Cyg A, and Tau A). These sources are NASA Goddard Space Flight Center, Greenbelt, lUD, 2077l Access to the cat-
unsuitable for the calibration of large antennas, particularly at high hequencies. alog is also available through electronic mail via Bitnet at W3WH\\IIaSCFNIVS
since they all have finite sizes and are all partially resolved. It is therefore necessnry or via SPAN at NSSDCA::WARREN The original references for the position and
to have a list of secondary calibrators, the flux densities of which are deternrirred identification of each source are contained in the notes to the catalog
with reference to ihe strongest sources as primary standards. The flux densitios Three general surveys of X-ray sources are the 4U catalog (Forman el al , 1978)
of the sources given in ?[e Astronomical Almanac have been determined in this the 24. catalog (Cooke el ol., 1978), and the lH catalog (\trroodel al . 198'1) The
manner. aources in these catalogs are dcsignated by their X-ray location in the equatorial
The data in the table were taken from Baars et aL. (1977) and derived in t,lie coordinate systen (equator and equinox 1950 0). The locations are given as right
following manner: ascensionin hours and minutes and declination in degreesor dcgreesand tenths, and
(1) The absolute spectrum of Cas A, rvith its secular rate of decreasetaken into are prefixed by the satellite initial letter and the edition nunrber of the catzrlog (4U
: Fourth Uhuru catalog; 2,A = Second Ariel catalog: lH : Filst HEAO-1 catalog)'
account, was derived for epoch 1980.0 between frequencies of 300 NIHz arrd
30 GHz. Hence, 4U1656*35, 241655+353, and 1H1656*354 all designatethe samc source
(2) The absolute spectra of Cyg A and Tau A were derived from direct rncir" in the three difierent catalogs.
surements and from measurements of the ratios of their flux densitv to that The nomenclature of X-ray source designations is confusing' primarily for his-
of Cas A. torical reasons.The brightest sourcesin the X-rav region are designated by constel-
(3) An accurate spectrum for the somewhat weaker source Vir A was estalr lation in order of discovery The letters XR (Friedman et al ' 1967) or X (Giacconi
lished from direct, accurate ratios of its flux density to those of Cas A arrcl et at., 1967) are appended and followed by a number. Thus, Cyg XR-1 and Cyg
cvg A. X-l are the same source Certain sources in the Large and Small Magellanic Clouds
(4) This Vir A spectrum was used as a basis for obtaining accurate rclativc pre-
receive the sane designation \\'ith LN'IC or SNIC instead of a constellation as a
spectra of the sources in the table, these sources being suitable for routirtc
fix. Another comrnon usage designates sources by'galactic longitude and latitude
calibration of flux densities.
prefixed by the letters GX (Bradt el al., 1968). Still other sources are dcsignated
Onlv three sources are suitable for the calibration of interferometers and sytr- Ly prefixing the initials of the discover'ssatellite, institution' or name to ;r rrtnning
in
thesis telescopes-3c48, 3C147 and 3C286 (the positions for these are taken fiotrr catalog number or a 1950.0 equatorial position. IAU Contrnission 48 decided
Elsmore and Ryle, 1976). Some other sources may need a correction for angulat' 1976 that no scheme of nomenclature t'as preferred over alr.v other rvhen referring
used irt
size when used with the largest single antennas at high frequencies, but sinct'tltc to sourcesdesignated in more than one system. The discoverl'designatiorl is
is listed' the sorrrce is
sizes are well-known, the correction is easy to apply. The Astronomical Almanac. Wheu no discovery designation
The IAU rec-
It must be emphasized, first, that the given flux densities cover a rangc ftortr always referred to by the common name of the ideniified courltcrpart
by
400 MHz to 22 GHz extrapolation beyond this range is dangerous. Second. :rl- ommende<l in 1983 that sources newly discovered after that date be designated
though these sources have "normal" spectra, characteristic of tra[sparent sJ'll- their X-ray position in right ascellsion and declination in thc equatorial coordinate
chrotron sources, some do contain compact, opaque components that may be vnri- system of 1950.0.
able. For example, 3C48 appears to show some variability at a level betwecn I Tabulated positions are based on published positions of identified courlterparts
I (10 rawattm 2), is taken from
percent around 1.6 GHz and 30 percent around 20 GHz (Perley and Crane, 1985). The (2-6) keV flux, in units of 10-lr erg
"m-2s
3C147 showsvariations in its structure on milliarcsecondscales(Preussel o1..198J) the 4U catalog. For sources with variable X-ray intensities, the maximum observed
optical
and its flux density is also variable (Simon el ol., 1983,Andrew e, al., 1981).Accu- flux from the 4U catalog is tabulated. The tabulated magnitude is the
y filter, unless marked by an asterisk' in
rate calibrations of flux density measurements should make use of several sources magnitude of the counterpart in the
are denoted by
from the table, so that such effects may be detected and allowed for. which case the B magnitude is given' Variable magnitude objects
 534 EXPLANAToRY
sUPPLEvENT 10 / STARS
ANDSTELLAR
SYSTEMS 535

V for these objects the tabulatcd rrragnitude pertairrs to maxirnuln brightlcss Curtis, R.H. (1932). 'Ciassificatio| :Lnd Description of Stellar Spectra" n Hartdbrch.tl.cT
Codes specifying the class of obiects to $'hich the identilled courrterpart belorrgs ,4iirophgrsilPart V (J. Springer. Berlin). pp. 1 108.

are explained at the eucl of the table. de Vaucoulcurs, G. (1959a). "Classification and N{orphology of External Clalaxies' in
HanclbtLchder PhysikS- Fluggc, ed. (Springer-Verlag.Berlin), 53, 275 310.
de Vaucouleurs.G. (1959b). "Photographic Dimensions of the Brighter Galaxies' .4slron.
J. 64, 397 .109.
10.5 REFERENCES de Vaucouleurs.G. (1963). "Rcvised Classilicationof 1500Briglrt Galaxics" Astrophus. .1.
SqtpL 8. 31. 97.
Andre*'. B.H., Nlacleod, J.\L, and Feldman, P.A. (1981). "Standard Sourcesat 10.6Clllz de Vaucouleurs,G. (1977). "Quali{at,ive and Quantitative Classilicationsof Galaxics" in
and Variabilitl in 3C147" Astron. Ast.Tophys.99,36 38. Thc EuctLutton of Galaeies and StelLar Populations B.lt{. Tinsiey and R.B. Larson,
Argue, A.N., de Vegt, C., Elsmore, B.. Fanselow,J., H:r.rrington.R., Hemenrval-.P., John- eds. (Yale Univ. Obserratory, Nerv Haven), pp- 43 96.
ston, K.J., Kuhr, H.. Kumkora, L, Niell, A.E.. \\raltcr, H., artl \\ritzel. A. (198.1).,,\ de Vaucouleurs.G. (1979). "The Extragalar:ti<:Distancc Scale.V. Tertiary Disiancc Ildi
Catalog of Selectcd Corllpat:t, Radio Sourccs for the Construction of an Extragalar:tic cators" ,Lslropi.ys. J. 227, 380 390.
Radio/Optical Referencc Frame" Aslron. Astraplrys. 130. 191 199. de Vaucouleu$, G., de Varx:orleurs, A.. and Corwin. H.G., Jr. (1976). Secontl Re.Jerencc
Arp, H.C. (1965). "Globular Clusters jn the Gala,ry" ir Galactic Sfrzcture A. Blaaul Catalogue oJ Btight Calaxies, (RC2), (Universily of Texas Prcss. Atsti ). pp. 1 3alrj.
ard NI. Schmidt. eds. Vol. V of Stars o.nd Ste,Llo,r SgstcrnsG.P. Kuiper and B.\l de Vaucouleurs. G. and Bollingel, C. (.7977). "Contributions to Galaxv Photornetr."'. VL
N'Iiddlehurst, eds. (University of Chicago Press: Chicago). pp. .101 '134. Reviscd Standard lbtal Nf.rgnittdcs and Colors of 228 N{ultiply Obsened Galaxies"
Baars, J.\\'.\L, Genzcl, R., Pa.uliny-Toth. LLK., arrd \\'itzel, A. (1977). 'The ,,\bsolule Astraphgs. .1. SuppL.34, 169 177.
Spectrurn of Cas A; An Accurate Flrrx Densitl. Scalc and a Set of Secondary Calibra
de Vaucouleurs. G., dc Vaucouleurs. A., and Corwin. I1.G..JI. (1978). "SJStenlatic and
tols" ,4slron. Astrophlls.61, 99 106. Accidental Errors in Galaxy Luminositv Cla.ssi6t:ai,ion" Astron. J.83. 1356 1359.
Batten, A.H. (1985). "Radial-Velocity Standarrls" in Stellar Rad.ial Vr:lor:ities. IAtl Coll. de Vaucouleurs,C., Peters. \\r.L., Bottinelli, L.. Gougtenhcirn. L.. and Paturel, G. (1981).
No.88 A.G.D. Philip and D.\\'. Latham, eds. (L. Dar.is Press: Schencctady).
"Hubble Ratio and Solar ltlol,ion from 300 Spirals Har.ing Distarccs Dcrivcd fiom H
tsatten, A.H., Fletcher, J.trL, and N,lacCarthl, D.c. (1989). "Eighth Catalogue ol rhe
I Line \\iidths" Astrophys J.218,408 422.
Orlrital Elemcnts of Spectroscopic Birrary Systems" Pub. Dorninion Astrol rllsit.al
de Vaucouleurs,G-, and Pcters. \V.L. (198,1)."The Dcpendenceon Distancc and Redshift
Obseruatory XYII, 1 317.
of thc Velocit)' Vectors of the Sun, the Galaxy, and the Local Group rvith RcsPcct to
Bessell, NI.S. (1986). " trl?1 Photorrretrv III: Photograplic nnd CCD fi ard 1 Bands an(l
Diferent Extragalactic Flarnes of Refererrce" Astraphys. J. 287, I 16.
the Kron-Cousirls rRl System" Pub. Astron.. Soc. Pac. 98, 1303 1311-
de Vaucouleurs.G., de Vaucouleurs.A., Corrvin, H.G.. Buta. R.. Paturel. G.. and Fotque.
Bessell.NI.S. and Wiis. E.W. (1987). "The Cousins and Kron l/filSysrems" Pzb. ,,lstrzn.
P. (1991) Thi.nl R.eferenceCatalogue of Brillht Go.Iaxres.(Springer-Vcrlag. Nerv York)
Soc. Pac. 99, 642 611.
Dolan, F. (1983). Privatc communicati(nr-
Bidelrnan, W.P. (1969). "stellar Spectra and Spectral T1-pes" irL StellctrAstronornu. H. \'.
Chiu. R.L. \\iarasila, and J.L. Rerlo, eds. (Gordon ald Breach, Nerv York). Volurrr,:' Egret, D. ancl Wcgner, \1. (1988). "SINIBAD Present Status and Future" itt Astronomg
1, pp. 1'17 206. From Lar.oeDotahase3ES(J CoDlerence and \\iorkshop Proceedings l'tro 28 F. \'lurtaglr
Bidelman, W.P. (1991). Priv:Lteconmunication. and A. Hcck, eds. (European Southern Obserratorl.' \Iunich), Pp. 323 :128.
Bradt. H., Naranarr,S., Rappaport. S., and Spada. c. (196E). ,,CelestialPositioDsof X R.r] Elsmore, B. and Rvle. \4. (1976). "Further Astrometric Obscr\?tiorts rl'ith thc 5-km ltadio
Sourcesin Sagittarius" Astrophys. .-I-152. 1005 1013. Telestr4re' Matr. Not. R. Astron. Sor:.171, 111, 423.
Buta, R.J. (1989). "caiaxy XTorphologyarrl Classiiication,,in Tlte Wotltl o[ GakniesIL(l. Explanatctry Supy.rlernenttct the Epherneris (1961). Prepared jointlv b)' thc Nautical Alma-
Corrvin and L. Bottirelli, eds. (Springcr Verlag, Nen York), pp. 29 47. nac Oflir:es of thc United Kingdom and l,he Ulrited States of ,'\lrcric.r (Her \l:rjcstv's
Campbell, B.. Walker, G.A.H., and Yang, S. (1988). "A Searchfor Substcllar Conrpaniorrs Stationary Ofiice, London).
to Solar-Type Stirs" r\stroplgs. ,].331,9O2 921. Findlay. J.W. (1966). "Absolutc Intcnsitv C'alibrations iD R;rtlio Astro ornt Ann llt:u'
Cooke, B.A., R.icketts.NL,I., Nlaccacaro.T., Pye. J.P., Elvis, NI., \Vrtson. NLG., Criffiths. Astron. Astrophys. 4,77 91
R.E., Pounds, It.A., \'Icllardy. L, Nlaccagni,D., Seward,Ir.D., Page. C.C.. arrd TrurLcr. Forman, \4'., Jones. C., Corninsky. L., Julicn, P., Nlurral-',S., f'eters. G., Tananbaurl, H,
\'I.J.L. (1978). "The Ariel V (SSI) Caralog of High Galar:tic Latirucle ( b >10 dt.g.). and Gincconi, R. (1978). "The Fourl,h Uhuru Catalog of X-Iial Sourr:es"Astroph|s'
X-Rav Sourccs" Mon.. Not. R. Astron. Soc. 182, ,189 515. .1. Su,pp. 38, 357 412.
Cousins, A.W..l. (1976). " VlU Standards in thc E Regions" Metn. R.. Astrctn. Soc. El. Friedrnan. H.. Bryan. E.T.. and Chubb, T.A. (1967). 'Distribui,ion and \hriabilitv of
Cosrnic X-Ray Sources" Science156, 37,1 378.
Crawford, D.L. and Btrrrres.J.V. (1920). "Stanclard Stars for zr.'6yPhotonretry" ,4rfrorr' Garrison, R.F., ed. (1,98a). The |vlK Processand Stellar Classffication (Da'ticl Dunlap Obs .
-/. 75. 978 998. Toronto), pp. 1 423.
Crawford. D.L. and Nlander, J. (196G). "Standarct Stars for Photlrclcctric H-Bcta Ph(ts Giacr:oni, R., GoreDstein,P., Gursky, I1., and \4'aters, J.R. (1967). "An X-Rav Sulvel' ot
torrretry" lslron. J. 71, 1,14 118. the Cygnus Region" ,Aslruphgs .I. 14a, LIL9-L127
536 EXPLANAToRY
SUPPLEMENT 10 / srABs ANDSTELLAR
SYSTEMS 537

Goodman, J. and Hut, P., eds. (1985). Dgnamics of Star Clusters,IAU Symposium No. Johnson, H.L. and Morgan, W.W. (1953). "Fundamental Stellar Photometry for Standards
113 (Reidel, Dordrecht, Holland), pp. 1 6?2. of Spectral Type on the Revised System of the Yerkes Spectral Atlas" AstroplLgs. J.
Gray, R.O. (1989). "The Extension of the MK Spectral Cla.ssification System to the In- 1 1 7 ,3 1 3 3 5 2 .
termediate Population lI F-Type Stars" ,4slron. -/. 9E, 1049 1062. Kaplan, G.H., Josties, F.J., Angerhofer, P.E., Johnston, K.J., and Spencer, J.H. (1982).
Grindlay, J.E. and Philip, A.G.D., eds. (19aa\. Globular Clusters Sgstems in Galaries,lAtJ "Precise Radio Sorrrce Positions from Interferometric Observations" Astron. J. 87,
Symposium No. 126, The Harlow Shapley Centenary Commemorative Symposiurn 570-576.
(Kluwer, Dordrecht, Holland), pp. 1 751. Keenan, P.C. and McNeil, R.C. (1976). An Atlas of Spectra of the Cooler Stors (Ohio
Hanes, D. and Madore, B., eds. (1980). GlobuLar Cluslers Proceedings of NATO Advanced State University Press, Columbus).
Study Institute, Institute of Astronomy, August, 1978 (Cambridge University Press. (eenan, P.C. and Morgan, W.W. (1951). "ClassificationofStellar Spectra" \n Astrophgsics
Cambridge, MA), pp. 1 390. J.A. Hynek, ed. (McGraw-Hill, New York), p. 12 28.
Hazard, C., Sutton, J., Argue, A.N., Kenworthy, C.M., Morrison, L.V., and Murray, C.A. Kholopov, P.N., ed. (1985). General Catalogte of Variable Slors, fourth ed., Volune I,
(1971). "Accurate Radio and Optical Positions of 3C 2738" Nature Phys. 5c2.233. 1985, pp. 1 376; Volume II, 1985, pp. 1 360; Volume III, 1987, pp. 1 368 (Nauka
89 91. Publ. House, Nloscow).
Hesser,J.E., ed. (1979). Star Clusters,IAU Symposium No.85, (Reidel, Dordrecht, Hol- Kron, G.E., Gascoigne,S.C.B., and White, H.S. (1957). "Red and Infrared Nlagnitudes
land )r pp, 1 D10. for 282 Stars with Known Trigonometric Parallaxes" Astron. ,J. 62,205 220.
Hewitt, A. and Burbidge, G. (1987). "A New Optical Catalog of Quasi-Stellar Objects" Kukarkin, B.V. (1974). General Catalogue of The Globular Star Clusters oJ a Galaxy
Astrophgs. J. Supp. 63, 1 246. containing Phystcal Characteristica of 129 Objects Reduced.lo One Sgsfem (Nauka
Hewitt, A. and Burbidge, G. (1989). "The First Addition to the New Optical Catalog of Press, Moscow), pp. 4-136.
QuasiStellar ObjecLs" Astrophys. J. Supp.69, 1 63. Landolt, A.U. (1973). " UBV Photoelectric Sequencesin the Celestial Equatorial Selected
Hirshfeld, A. and Sinnott, R.W., eds. (1982). S&y Catalogue 2000.0, Yolurr'e 1, Stars to Areas 92-115" Astron. ,J.78, 959 981.
Magnitude 8.0, (Sky Publishing, Cambridge, MA and Cambridge University Press. Landolt, A.U. (1983). " I/B yR1 Photometric Standard Stars Around the Celestial Equa-
Cambridge, England), pp. 1 604. lot" Astron. J. 88,439 '160.
Hofleit, D. and Jaschek, C. (1982). The Bright Star Catalogue, Fourth Revised Edition
Lyne, A., Manchester, R.N., and Taylor, J.H. (1993). In preparation.
(Yale University Observatory, New Haven), pp. 1 472.
Lyngi, G. (1988). "The Lund Catalog of Open Cluster Data" in Astronomy From Large
Hogg, H.S. (1973). A Third, Catalogue of Variable Stars in Globular Clusters ComprisirLg
Databases, ESO Conference and \Yorkshop Proceedings No. 28 F. Ilurtagh and A.
2119 Entries Volvn;e 3, Number 6 (David Dunlap Obs., Toronto), pp. 1 75.
Heck, eds. (European Southern Obserr-atory,N{unich) pp. 379 382.
Hogg, H.S. (1959). "Star Clusters" in Handbuch d,er Physik,53 S. Flugge, ed. (Springcr
Manchester. R.N. and Taylor. J.H. (1977). Pulsars. (W.H. Freernan, San Fransisco).pp.
Verlag, Berlin), pp. 729 2O7.
1 281.
Holmberg, E. (1946). "The Apparent Diameters and the Orientation in Space of Extra-
Manchester, R.N. and Taylor. J.H. (1981) "Obsened and Derived Paramcters for 330
galactic Nebulae" Medd Lund Astror. Obs. Series II, 117, pp. 1 82
Pulsars" ,4slron. "/. 86. 1953 1973.
International Astronomical Union (1957). Transactions, Volume IX (Cambridge Universitl'
McCarthy. N{.F.. Philip, A.G.D., and Coyne. G.V., eds. (1979). Spectral ClasstJicationaJ
Press, Cambridge, England), p. 442.
the Future IAU Colloquium No. 47 (Vatican Obs. Vatican Citl ). pp. 1 375.
International Astronomical Union (1973). Transactions, Volume XVA (Reidel, Dordrecht.
Holland). o. 409. Mermilliod, J.C. (1981). "Compilation of Photometric Data: Presenl State and Projects'
Bull. Inform. CDS 26, 3 7.
International Astronomical Union (1988). Tr-ansa ctions, Volum e X X A (Reidel, Dordrecht.
Holland), p. 362. l4organ, \V.\\'. and Keenan. P.C. (1973). Spectral Classilicati<tn"Anrtual Ret oJ Astron-
International Astronomicai Union (1990). Transactions, Volume XXB (Reidel, Dordrecht. omg and AstrophAsics11, 29 50.
Holland), p. 267. Morgan, W.W.. Abt, H.A.. and Tapscott, J.\\'. (1973). Rexised IIK Spe-ctralAtlas for
Ir$in, A.W., Campbell, B., Morbrey, C.L., Walker, G.A.H., and Yarg, S. (1989). "Long- Stars Edrlier than the S.un(Yerkes Obserlator1, Chicago and Kitt Peak \at'1 Obs .
Period Radial-Velocity Variations of Arcturus" Pub. Astron. Soc. Pac. l1l, 147 159. Tucson).
Janes, K.A., Tilley, C., and Lyng5, G. (1988). "Properties of the Open Cluster Systenr" Morgan. \\'.W.. Keenan, P.C.. and Kellman, E. (19.13).An Atlas of StelLarSpect'ra(Uni
Astron. J. 55.771 784. versity of Chicago Press. Chicago).
Johnson, H.L. (1963). "Photometric Systems" it Basic Astronomical Data K.A. Strand. Nicolet. B. (1978). 'Catalogue of HomogeneousData in the UB l'- Photoelectric Photo-
ed. Volume III of Stars and Stellar Systems, G.P. Kuiper and B.N{. Nliddlehurst, eds metric System Astron. Astrophys. Supp.34, | 19.
(University ol Chicago Press, Chicago), pp.204-224. Nilson. P. (L973). Uppsala General Catalogue ol Galades (Ro1'. Soc. Sci. Uppsala. Upp-
Johnson, H.L., Mitchell, R.I., Iriarte, B., and Wisniewski, W.Z. (1966). "UBVRIlKLPho- sala), pp. 1 456.
tometry of the Bright, Starc" Communications of the Lunar and, Planetary Laboratory Ochsenbein,P. and Dubois. P. (1986). "Some Aspects of the Bibliographi' of Astrono ical
4, 63,99 241. Objects' Bzl1. htforrn. CDS 31. 137 'I1O.
538 EXPLANAToRY
suPPLEMENT 10 / STARSANDSTELLAR
SYSTEMS 539

Paturel, G., Fouque, P., Lauberts, A., Valentijn, E.A., Corwin, H.G., and de Vaucouleurs. Wegner, M., Egret, D., and Ochsenbein, F (1989). SIMBAD User's Guide, (English ver-
G. (1987). "Standard Photometlic Diameters of Galaxies III. Reduction of the Diam- sion by Wayne H. Warren, Jr', of NSSDC, and Joyce M Watson, of SAO) (CDS,
eters in the ESO-B and SGC Catalogues to the Standard Diameter System at the 25 Strasbourg)
mag arcsec 2 Brightne"" Level" Astron. Astrophys.184, 86 92. White, R.E. and Shawl, S.J. (1987). "Axial Ratios and Orientations for 100 Galactic
Pedoussaut,A., Carquillat, J.M.J Ginestet, N., aod Vigneau, J. (1988). "Binaires Spectre- Globular Star Clusters" ,4slrop[gs. J.317,246 263.
scopiques. 15e Catalogue Complementaire" Astron. Astrophgs. Suppl. 75, 441-443 Wood, F.B., Oliver, J.P., Florkowski, D.R., and Koch, R.H. (1980). "A Finding List for
Perley, R. and Crane, P. (1985). "3C 48 Variability" NRAO Newsletter no. 25, p. 1. Observers of Interacting Binary Stars" Pub University of Pennsylvania, Astron. Ser.
Perry, C.L., Olsen, E.H., and Crawford, D.L. (1987). "A Catalog of Bright uubg(betal XII, pp. 1 349.
Standard Stars" Pub. Astron. Soc. Pac.99, 1184-1200. Wood, K.S., Meekins, J.F., Yentis, D.J., Smathers, H.W, McNutt, D P., Bleach, R.D.,
Philip, A.G.D., Cullen, M.F., and White, R.E. (1976). "UBV Color- Magnitude Diagrams Byram, E.T., Chubb, T.A., and Friedman, H (1984) "The HEAO A-1 X-Ray Source
of Galactic Globular Clusters" Dud,LeyObseraatoryl Reports (Dudley Observatory, AI- Catalog" Astrophys. J. Supp.56,5o7 649.
bany), pp. 1 186.
Philip, A.G.D. and Hayes, D.S., eds. (1981). Astrttphgsical Po,rameters for Gbbular Clus-
lers, I.A.U. Colloquium No. 68 (L. Davis Press, Schenectady), pp. 1 614.
Preuss, E., Alef, W., Whyborn, N., Wilkinson, P., and Kellermann, K. (1984). "The NIii-
liarcsecond Core of 3C 147 at 6 cln" in VLBI and, Compact Rad.io Sources, R. Fanti.
K. Kellerrnann, and G. Setti, eds. IAU Symp. No. 110 (Reidel, Dordrecht, Holland).
p. 29.
Rawley, L.N., Taylor, J.H., and Davis, M.M. (1988). "Fundarnental Astrometry and \'{i1-
Iisecond Pulsars" Astrophs. J.326,947 953.
Sandage, A. (1975). "Classification and Stellar Content of Galaxies Obtained from Direct
Photography" rn Galaxies and the Uniuerse A. Sandage, M. Sandage, and J. Kristian.
eds. Volume IX of Stors and StelLar Sgslems (University of Chicago Press, Chicago),
pp. 1 35.
Schmidt-Kaler, Th. (1982). "Calibration ofthe MK System" in Landolt-Bornstein, Net'.'
Ser., Gr. VI, 2b, pp. 14 24 arrJ 451 456. (Springer-Verlag, Berlin).
Shawl, S.J. and White, R.E. (1986). "Accurate Optical Positions for the Centers of Galactic
Globular Clusters" Astron. J.91, 312-316.
Simon, R.S., Readhead,A.C.S., Moflet, A.T., Wilkinson, P.N., Allen, B., and Burke B.F.
(1983). "Low Frequency Variability and Predicted Superluminal Motion in 3C 147"
Nature 3O2, 487-490.
Strijmgren, B. (1963). "Quantitative Classification Methods" in Basic AstronomicaL Data
K.A. Strand, ed. Vol III of Stars aniL Stellar Sgstems G.P. Kuiper and B.M. Nliddle-
hurct, eds. (University of Chicago Press, Chicago), pp. 123 191.
Taylor, B.J. (1986). "Transformation Equations and Other Aids for l/fi1 Photometry"
Astrophgs. J. Supp. 60, 577 599.
Trumpler, R.J. (1930). "Preliminary Results on the Distances, Dimensions and Spac€
Distribution of Open Star Clusters" -Lic,t Obsematory Bulletin 14, 154 188.
van den Bergh, S. (1960a). "A Preliminary Luminosity Classilication of Late-Type Calax-
ies" Astrophgs. J. 131,2\5 223
van den Bergh, S. (1960b). "A Preliminary Luminosity Classification for Galaxies of TyPc
Sb A'trophys. J. l3l, 558 573.
V6ron-Cetty, M.-P. and V6ron, P. (1989). "A Catalog of Qrasars and Active Nuclei" 4ttr
ed. ESO Scientiffc Report No.? (European Southern Observatory, Munich), pp. 1-143-
Webbink, R.F. (1985). "Structure Paramete6 of Galactic Globular Clusters" in Dgnamics
oJ Star Clusters, J. Goodman and P. Hut, eds. IAU Symposium No. 113, (Reidel'
Dordrecht, Holland), pp. 541 577.
Weedman, D.W. (1986). Quasar Astronorny (Cambridge University Press, Cambridge,
England), pp. 1 217-
 CHAPTER 11
Computational
Techniques
by R.L.Duncombe

11.1 INTRODUCTION
TO COMPUTING
TECHNIQUES

In this section we will examine several computing techniques of special relevance to

astronomical calculation. For a more general treatment, seetextbooks on computing
methods and numerical analysis. A short note on elementary computing principles
is also included 1n Interpolation and Allietl Tables, which covers the sources of mis-
takes, the nature of checks, and the nature and effect of the inevitable errors due to
rounding-ofi and other causes, These topics, although of fundamental importance,
a,re not discussed here.
An important principle of computation is that the data used determincs the
maximum precision of a calculation. This precision can be reduced by poor comput-
ing-for example, by a poor choice of formula or by the failure to retain an adequate
number of figures in the intermediate stages but it can never be increased. Gen-
erally no simple relation exists between the errors (absolute or relative) of the final
result and of the data, though a relation can readily be seen numerically by follow-
ing through the calculation step by step, and often geometrically. It is misleading
to quote a result to more figures than is justified by the data on which it is based,
and it is erroneous to do so if an inadequate number of figures have been retained
in the intermediate stages.
Retaining more figures than necessary is a waste of effort if it significantly
increases the work involved. In some operations, however, especially in those done
on a computer or a personal computer, extra figures involve very little additional
work; but in other operations, an extra figure may require a more elaborate formula
or extra word length, resulting in much more additional work. A suitable number of
figures is that which offers the greatest convenienceconsistent with the building-up
error, due to accumulation of rounding-ofis, not exceeding the error of the data;
this must be judged in relation to each calculation.
541
 EXPLANATOBYSUPPLEMENT TECHNIQUES
11/ COMPUTATIONAL 543
542
both x and y are small (y by the cancellation of two nearly equal components) if n is
small. It is not legitimate here to use extra figures in the trigonometric functions in r
and y, and B can be found with a precision only proportional to coseca1this precision
is, however, clearly adequate to fix the point C (from BA). Similar arguments hold
for the angle C. Neither B nor C can individually be determined as precisely as b.
c, or A; but the sum B + C can be so determined, as may be seen from gcometric
considerations and from the eauation

c ol2t( 8 +o = ' o ' i 1 l-* ' l o nlzo . (l l.l-3)

COS;\D c)

Important problems in astronomy may require the sum (or difference) of tu.o angles
to be known more precisely than is possible with either (for example. the elenlents
Figure11.1.1 of a planetary orbit); therefore, great care is required in handling them. The rec-
Sphericaltriangle ommended method is to determine one of the angles and then treat it as cxact in
finding the other, which must not be found independently.
Although there may be some uncertainty of prccision in determining an angle,
there should be no uncertainty of quadrant. This can always be achieved by ade-
quate choice of formulas. For examplc, the following formulas for the solution of a
Mathematical formulas, if correct, adequately give a desired precision in the
spherical triangle, in which two sides and the included angle are given, are alrvays
result; but they may be inefficient, inconvenient, and misleading. One exanplc is
adeouate if the sidesare lessthan 180'.
the inefficiency of determining a small angle ftom its cosine; for a stated precision of
angular measurement the number of decimals in the cosine is inversely propol i ional
sina sinB = sinb sinA =x,
to the angle; and for a fixed number of decimals in the cosine (the rrsual and
convenient case) the precision obtainable decreasesas the angle decreases.Cortsidcr, s i n a c o s B= c o s b s i n c- s i n r c o s c c o s A = y ,
for example, the formula for the third side of a spherical triangle, (Figure 11 1 1) ( rr . l - 4 )
c o s n = c o s b c o s c+ s i n D s i n c c o s A = ; ,
given two sides and the included angle,
sina cosC = sinb cosc - costsinccosA,
cosa = cosbcosc + sinr,sinc cosA= z, (||.]-r) sinasinC=sincsinA.

js
in which c is to be determined from measured values of D, c, and,4. Wht'l a The check ,2 +y2 +22 - I = 0 is useful. The sign of z determines tho quildrant
small, it can be found to the same precision as the data only by the apparentll' of a, and the signs of .r and y that of B; similarly with C.
incorrect procedure of retaining more decimals in each trigonometric fuuctiorl ou In astronomical problems, the sides and angles of triangles on thc celestial
the right-hand side (z) of the equation than the data appear to justify. This is sphere may be of any magnitude, and it is undesirable to solve them by methods
legitimate because both b - c and,A must be small so that their cosines ate knowu that restrict the sides to arcs less than 180". The preceding general formulas are
to extra figures. There are circumstances, however, in which the inconvettiurce "[ all valid for triangles with sides of any length, and may be applied immediately t'o
such a procedure is outweighted by other factors. the general triangles of spherical astronomy without any restriction to values less
The apparent failure of a mathematical formula should not be confused with than 180'. Ftrthermore, to make the solution determinate, it is only necessary to
real geometrical limitations of precision. In the spherical triangle ABC, find the algebraic sign of both the sine and the cosine of each arc or angle that may
exceed 180' to fix the quadrants in which they lie. Any of the casesof the general
sinasinB = sinb sinA = -t, (l | .l-2) triangle is determinate when, in addition to the three given paxts, the algebraic sign
s i n 4 c o s B= c o s r s i n c s i n b c o s c c o s A= ) ; of the cosine of one of the required parts is also given. In most practical problems
 544 EXPLANATORYSUPPLEMENT 11 / COMPUTATIONALTECHNIOUES
545
T a b l e1 1 , 1 , 1 T a b l e1 1 . 1 . 2
Precisionof AngleandNumberof DecimalsIor Trigonometric The Methodof InverseUse
Functions
0. 45. 90. 1350 180. 225. 270. 3150 360"
l', =0.00029 1" =0.00007 0001 =0134= 1i4
'l' =0.0000048 G1 =7 x 10 6 x 1 0 6= o t 2 = 0 : 0 1 4
++++
0ll =5 x 10 7 0:01 =7 x 1O-7 x 1o-7= ol02 =coo14
0 1 0 1 = 5x 1 0 - o 0 9 0 0 1 = 7x 1 0 - 8 x 1o-8=o:oo2=0:00014 v ++
lan cot cot
Ian or cot tan tan cot cot tan col
DotCarc DCDCDCDCD

the conditions of the problem supply this sign. In general triangles the utmost care according to which ratio is less than unity. The quadrant is determined by the signs
should be taken to specify unambiguously the direction of measurement of angles of .r and1, since,if a is lessthan 180', sina is positive.
and arcs. llost trigonometric tables are arranged semiquadrantally, with dircct (D) and
Collectively, the preceding formulas are sufficient for the solution of the general complementary (C) arguments. The Table 11.1.2shows the method of inverseuse
triangle without restriction on the rnagnitudes of the parts, but, in practice, the in finding B. There is no difficulty in systematic computation.
additional formulas The range of precision of the inverse determination of an angle frorn its sine (or
cosine) and tangent (or cotangent) is indicated in Table 11.1.3. The tabl<ris arranged
cosA = - cosB cosC + sinB sin Ccosa, to show the range of angle for which the alternative trigonometric function should
sinA cos, = cosB sin C + sinB cosCcosa, be used, and gives the error in the angle corresponding to an enor of I x 10-6 in
sinA cosc = sinB cosC + cosB sin Ccos 4, the function.
The technique of inverse interpolation, referred to briefly in Section 11.23. is a
sinAcotB = sinccot, - cosccos./4
powerful tool in the solution of transcendental equations and of equations in which
sinA cot C = sinbcotc - cosbcosl^, (11.1-5) the algebraic solution is complicated. The fundamental principle of such methods
Iies in the tabulation of a discriminant, deflned so that it attains a predetermined
are very useful. (Collected formulas are given in Table 11.3.1 p. 550.) value (usually zero) when the original equation is satisfied, and the calculation of
In practice, the trigonometric functions of r, c, and ,4 are taken out with a the unknown argument corresponding to the predetermined value of the discrimi-
number of decimals depending on the precision of the data, and the capacity of the nant by the process of inverse interpolation. Although these methods rr)a! involve
calculating machine to be used. Table 11.1.1 gives corresponding precisions of angle more calculation than direct methods, they have two considerable advantages. First,
and number of decimals. they are usually independent of theoretical developments, of anl. approximations
If no additional work is involved, it is clearly advantageous to use more decimals that may be necessary in such developments, and of extrapolations. Second, the
than the precision warrants, since the effect of rounding-off errors is then much
reduced. For instance, seven decimals give just sufficient coverage for data known
to about 0'.'02and for results to be rounded off to 011; but interpolation in eight- T a b l e1 1 . 1 . 3
flgure tables &t interval 1" ofiers little, if any, more dificulty than interpolation irr Range of Precisionof the InverseDelerminationof an Anole
seven-figure tables at interval 10", and eight decimals can be used with little extra
work. Errorsc-orresponding lo 0. 45" 90. 135" 1800
1 x 10-o in tunclion 180" 225. 270. 315" 360"
In the solution of Equations 11.1 4 and 11.1 5, a is found either from cosa =:
or from sina = (l+y2)l/2, whichever is the smaller. Assuming a is less than 180o,the Use funclion sin cos cos sin
ambiguity of the second formula is resolved according to the sign of z. I is found Error in angle o]2 0:29 0':2 0t29 0';2
from
Use function tan cot col lan
Error in angle 0:014 s014 0:014
t-r= I or t =1,
c oB (11.1-6)
v x
 EXPLANATORYSUPPLEMENT
,I1/ COI.,|PUTATIONAL
TECHNIOUES 547
546
correctness of the calculations and of the required answer, and the precision of that T a b l e1 1 . 2 1 . 1
Ditferencesin TabularArguments
answer, a.re directly under the control of the computer. Illustrations of the use of
these methods are to be found in the preceding sections, particularly in the calcu-
lation of the local circumstances of eclipses (Section 8.23) and the derivation of the Argument Function Diflerence
times of moonrise and moonset (Section 9.3). lst 2rd srd 4th
The availability of high-speed computers has changed the relative importance
I-z 6vz= fi fo
to be attached to the various factors entering into a computation. It is no Ionger 6-an o'-ltz a6=012-o-112
desirable to restrict the amount of calculation to a minimum: many repetitions of t1 =11 _2fo+f 1
a simple iteration a.re often preferable to a more sophisticated direct calculation. 6- t p 63+6?=!2-fi-lo+f 1
In particula.r, many of the transformations of coordinates a.rising in astronomy can to fo d5 6t a?- aB
a?'r=
6tp = f2 -3h +3fo _ f-1
be efficientlv handled by the direct use of the accurate formulas, instea<i of by A2
t+t l+l 6t=61p-63!2
approximate series expressions, These often involve multiplying a colurnn matrix 6gtz 04,2 = f2 - 411+6fo 4f_t + f z
representing the direction cosines by a transformation matrix, as is done, for exarrr- 6t+61 = fs 3fz+2f1 +2lo -Bf _1 + ! 2
ple, for the correction for precessionand nutation. Close attention to the precision of
computation is generally unnecessary,since many extra figures can be kept without
any extra work.
Although such machines are used for the computation of the data in The As-
tronomi.cal Almanac, thev may not be generally available to all users. The previous The symbol for the function is usually omitted in the notation for the differences.
remarks on the techniques of computation, therefore, are included to facilitate cir Tables are given for use with Bessel's interpolation formula for p in the range 0 to
cumstances in which desk calculating machines are used. +1. The differences may be expressedin terms of function values for conveniencein
Computing techniques adapted to electronic hand-held, desktop, or larger com- the use of programmable calcularorsor compulers.
puters axe given in most modern texts on computing (see Hildebrand, 1987 and
Press el al., 1986).
p = the interpolating factor = (r - 16)I (tt - td = (t - td / h \1r.21-2)

ANDSUBTABULATION
11.2 INTERPOLATION
11.22 Interpolation
Formulas
11.21 Introduction
andNotation Bessel's interpolation formula in this notation is

The interpolation methods described in this section, together with the accompany-
f p = f o + p6n + 82(63
+ 6I + h51D+ B46t+ 6!)+..., (r1.22-r)
ing tables, are usually sufrcient to interpolate to full precision the ephemerides in
The Astronomi,cal Almanac. Additional notes, formulas and tables are given in the
where
booklets Interpolation and. Allied' Tables arld Subtabulation and in many textbooks
on numerical analysis. It is recommended that interpolated values of the Moon's
82=p@-l)/4,
right ascension,declination, and horizontal parallax be derived from the daily poly-
nomial coefficients that are provided for this purpose in The Astronomical Almanac' Bz=p@-u(,-)) t,,
The term /, denotes the value of the funciion /(l) at the time t = t0 +pft, where
, is the interval of tabulation, t6 is a tabular argument' and P = (t - l0) / , is kno$/n Ba = (p + 1)p@- 1)(p- 2) | 48. (tr.22-2)
as the interpolating factor, The notation for the difierences of the tabular values ls
shown in Table 11.21.1; it is derived from the use of the central-difference operator The maximum contribution to the truncation error of /r, tor 0 < p < 1, from
6, which is defined by: neglecting each order of difference is less than 0.5 in the unit of the end figure of
6fo=fo*rn-fp-ttz. (rl.2r-r) the tabular function if 62 < 4, 63 < 60, 6a < 20, and 65 < 5oo.
548 EXPLANATORY
SUPPLEMENT 11 / COMPUTATIONAL
TECHNIOUES 549
Everett's interpolation formula in this notation is which may be evaluated in the nested form

f o= fs +p6y2
+fiafi+fi,a1.
+fia[ +fial.+..., tl1.22-3) f p = \\(aap+ a1lp + a2lp + a tlp + ao (r r.24_2)
Expressions for the coefficients ao, at, elc,) may be obtained from Stirlinq's inter_
-p(p - l)(p - 2) / 6,
where 4 = polation formula, neglecting fifth-order differences:
4=e,+1)p@-1)/6,
qo = fo,
4= -@+r)p(p,r)(p- 2)(p- 3)| 120,
a 1 = ( 6 1 1 26+- 1 p )l 2 - a 3 ,
4 = @+ 2)(p+ l)p(p- t)(p- 2)/ r20. (rr.224)
az=63/2-aq, ( 11 )/--\\
LaGra.nge's interpolation formula to second order in this notation rs
a3 = @i,.+ 6',r,) / 12,

t =t -, (io' - )r) . r*' - p2)

+r, ()t . )r) (rr.22-s) uq=6412+'

This is suitable for use in the range -l /2 < p ! +l /2, and it may be adequate
This formula has the advantage that the interpolation can be done entirely in terms
in the range -2 < p < 2, but it should not normally be used outside this range.
of the functional values. The disadvantages are that the coefficients are required
Techniques are available in the literature for obtaining polynomial representation
to the sarne number of significant figures as the function, and that the number of
that give smaller errors over similar or larger intervals. The coefficients may be
r,'aluesof the function to use in any pa.rticular case is not easily determined without
expressed in terms of function values rather than differences.
knowing the size of the difierences.
Economized polynominals, such as Chebyshev, provide efficient and accurate
expressionsthat may be easily evaluated with a small computer. The coefficients a;
'11.23 InverseInteroolation
of the Chebvshev expansion

Inverse interpolation to derive the interpolating factor p, and hence the time, for
, ao \-. -- (tt
which the function takes a specified value /, is carried out by successiveapproxi- It=-;+ ) aitt\x) )/,-.4\
mations. The first estimate D, is obtained from

are computed for prescribed time spans, where /, is the function represented, Z;(x) is
the Chebyshev polynomial of the i-th degree, and x is the normalized time variable.
To evaluate a Chebyshev expansion, it is necessary to normalize the time variable
rhisvarue orpisused,, ,, r.,rromthe.,,,,..,t:i;:ll
"ri:"::-,':':::'
otherwise, and hence an improved estimate ofp from
on the interval for which the series is valid. Attempts to use these series beyond
the specified time intervals for which they are valid will produce erroneous results.
An application of Chebyshev polynomials may be found in Section 5.5. Further
p = pt 82t63
+ 6l)5t/2. ( r 1.23-2)
information is given in Lanczos (1956) and Fox and Parker (1972).
This last step is repeated until there is no further change in 82 or p; the effects of
higher-order differences may be taken into account in this step.
11.3 PLANEANDSPHERICAL
TRIGONOMETRY
1'l .24 PolynomialRepresentations
In astronomical problems, the sides and angles of triangles on the celestial sphere
It is sometimes convenient to construct a simple polynomial representation in the rnay be of any magnitude, and it is undesirable to solve them by methods that
form restrict the sides to arcs less than 180'. The general formulas given in Table 11.3.1
fp = ao + a1p + a2p2+ a3p3+ a4Po+ "' , (rr.24-r) are valid for triangles v/ith sides of any length, and may be applied immediately to
 'I1 / COMPUTATIONAL
TECHNIOUES 551
Table11.3'l Table11,3,1,continued
Formulaslor PlaneandSphericalTriangles Formulasfor PlaneandSphericalTriangles
of thetriangle
fihe angles byA,B, C;theopposite
aredenoted sides,bya, b, c. Otherformulas
may
be obtainedby cyclic changes of A, B, C and a, b, c.\
Ouadrantal Triangle: a = 90o

PlaneTriangle SphericalTriangle sinAsinD=sinB sinbsin,4= sinB tanA = tanB secc

sinA cost = cosBsin C sinbcosA = - cosB cOSc tan,4= tanCsecb
asinB=bsinA sina sinI = sinbsinA
cos,4= cosBcos C cosb = cosSsinc cosA = - cotbcotc
acosB=c-bcos,4 sinacosB = cosbsinc - sinbcosc cosA
sin,4cosc = sinBcosC sinbcosC = cosc tanB = sinCtanb
*=b2+€-zbccosA cosa = cosrcos c+ sinbsinccosA
sin,4sinc=sinC sinbsinC = sinEsinc tanc = sinBtanc
acosC=b-ccosA sina cosC = sinbcosc - cosbsin ccosA
asinC=csinA sina sinC = sincsinA
sinl(a+C)=cos;,4 cos!asln |{a+ O = cos,Acos l(b c) SphericalTrianglein Whlchb ls Small
cosl(B+c)=sin;A cos!acos !{a* c) = sin;/4cos}(b + c) a - c = - b c o sA + ! G c o t c s i n z A + .
a s i nl ( a - q = @ c)coslA sin;asin l(B - c) = cos]/sin ](b c) B sinC = +bA+ LF colcsin2A+. .
acos l(a - c) = (b + c) sin ;,4 sin]acos ] (B - c) = sinlr sinI (b + c) 1 8 0 " C , 4 = + b c o t c s i n A +I F ( 1 + 2 c o t 2 c ) s i n 2 4. +
s=(a+b+c) s=l(a+b+c)
l=(s-ax-bxs-c)/s /rP= sin(s- a)sin(s- b)sin(s- c)sins
Area=sf=;bcsin/4 Area=/4+8+C-180"
bcsin2;,4 = (s - bxs - c) sinbsincsin2;/ = sin(s- b)sin(s c)
bccoszlA = s(s a) sinDsinccos21,4= sin6sin(s- a)
the general triangles of spherical astronomy without any restriction to values less
tan12A=tt(s-a) tanlA=m/sin(s-a)
than 180". Furthermore, to make the solution determinate, it is only lecessary to
find the algebraic sign of both the sine and the cosine of each arc or angle that may
exceed 180o, to fix the quadrants in which they lie. Any of the casesof the general
triangle is determinate when, in addition to the three given parts, the algebraic sign
AdditionalFormulas Right-angled Triangle: A = 90'
of the sine or the cosine of one of the required parts is also given. In most practical
sln4sinb = sinBsina sinasinB = sinb problems it happens that the conditions of the problem supply this sign. In general
sina cosI = cosb sinc triangles the utmost care should be taken to specify unambiguously the direction
cosa = cosbcosc of measurement of anqles and alcs.
sinacosC = sinbcosc
sin4 sinc = sinCsina sinasinC = sinc
cosacosB = sinacotc- sinBcotC sinBcosa = cosbcosC
cosacosC = sinacol b sinCcotB cosB = cosbsin C
cosbcosA= sinbcotc - sin4cot C sinBcosc = cosC
cosccosA = sinccot b - sin/4cotB sinBsinc = sinbsinC
s=t(A+B+Cl tana = tanbsecC
M2= cos(S A)cos(S- B)cos(S- C)/ coss tana = tancsecB
sinBsinCsin2l€ = - cosScos(S-,,4) cosa=cotBcotC
sinBsinccos2;a = cos(S- B)cos(S- C) tanb = sinctan I Figure 11.3.1
Arc of circle
tan la = cos(S- ,4)/ M tanc = sinb tanC
 552 EXpLANAToRy
suppLEMENT 11 / COMPUTATIONAL
TECHNIOUES
JbJ
In Figure 11.3.1 the arc P/P subtends the small angle b, measured in radians, where the matrix product is
r is the radius and PR is perpendicular to OP at R. The radius of the circle times
the angle subtended in radians equals the arc length, hence r x b = ptp- 'l
cosj
I 0 sinj
When D is small enough that P/R approximates p/p then sinb = p/R / r and PtitQQt= sinisinj .:'i -sinicosj .
| | fi|.414)
b= PtP I r. Thus, sinD = b and, tanb = b.If in the small angle of D radians there are
L - cosI sln stnI cosicosj I
4 minutes of arc (4'), then sin 4/ = 4 sin l'.
For three successivepositive rotations (i), (j), (k) about the l-,
)_, and z_axes,
respectively,(at,yt,7t)= (x,y, z)p(i)0u)R(k). and the matrix product is
11.4 MATRIXAND
VECTOR
TECHNIQUES
cosj cos/< _ cosj sin &
f sin i I
P()Qa\Rrk')= sin i sini cosk + cosi sin t - sin i sinj sint + cosi cos[
-sinicosj l.
11.41 Rotationof AxesUsingMatrices |
| - cosisinjcos,t + sin lsint cosi sinj sint + sin i cost i"or; J
In practical computations it is often necessaryto change the orientation of reference "o,
(11.4t1)
frames. Given a right-handed system of rectangular coordinates x, y, z, and rotating This case comprises a rotation (i) about the.r_axis, a rotation (,1)around
the
the system positively through the angle i about the x-axis, the new coordinates x/. new position of the y-axis, and a rotation (t) around the final position of the
z-axis.
y/, z/ are related to the original coordinates by The matrix denoting any number of rotations can be derived by multiplication
following the precepts below.
x'=y, Two matrices A and B form the product AB, where the superscript indicates
y/=/cosi+zsini, the row and the subscript indicates the column.

z, = _t sini + zcosi. (l1.41-l)

attl
lai 4 lbi bt, bll
In matrix notation this becomes(.t,,),,2,)= (x,:y,z)p(i), where A=lai qi a3l. B=lbl bi b3l. ( i l , 4 r8 )
Lal a3, al J lol u1, ut,)
fr 0 0 I
P(,= -sintl. (t r.4r-2)
l0 cosi qibl+ atbl + ata3,alo)+ a)ur,
+ atu3,alut,
+ a)a!+ a\u]1
l0 sini cosiI I
AB= qlbl+ olbl+ albl olotr+alo]+ alu] alot,+ alul+ aluj . (l 1.41-9)
Similarly, for a positive rotation j about the y-axis, (.r/,y/, z/) = (r ,1, z)e|),
| I
where Lolol+t3rtl+a3,b] + alb?,
alot, + ai,b3,
,tu1+a)tl+a]o])
cosj 0 sinj'l
I
o0= l 0 r 0l ( l l.4l-3)
0 cosjJ
11.42 Spherical
Coordinates
UsingVectors
l-sinj
Again, for a positive rotation,t about the z-axis,(xt,y,,2t)= (x,y,z)R(t), where Consider a triangle ABC (Figure 11.1.1) scribed on the surface of a sphere of unit
radius whose center is denoted by O. Each side cbc is a portion of a great circle
- sinr 0l
fcosk
defined by BOC = a, AOC = b. and AOB = c.
n(t)= cost 0J. (l1.4r4) A is the angle between the planes BOA and COA, B is the angle between the
l,int 0 lJ planes AOB ar'd COB, and C is the angle between planes AOC arld BOC.
Take l, m, n, as nonorthogonal unit vectors along OA, OB, and OC respectively.
The equation for two positive rotations may be written as The cross product I x m is a vector of magnitude sin c directed orthogonally to the
plane AOB. In the same manner, the cross product I x n is a vector of magnitude
(xt , yt ,2t) = (x,y,7)P(i)QQ), (rr.4't-s) sin& orthogonal to AOC. A is the angle between these two vectors.
 554 EXPLANATORYSUPPLEMENT 11 / COMPUTATIONAL
TECHNIOUES c55
The dot product of these two vectors is

(l x m).(l x n) = sincsinbcosA
=t.[mx(lxn)]
= l ll(m.n) - n(l.m)l
= (m. n) - (l . n)(t . m)
= cosa - cosbcos<, (r1.42-t\

which gives
cos4 = cos, cosc + sinb sinccosA. ( 1 14 ) - ) r

Similar equations for cos b and cos c are easily obtained.

From the definition of the vector product. we find

srnA = Figure 11.42,1

llxnl llxnl Triangleon unit sphere
l-l[m,l,n]+ m[,1,n]l
sinDsinc
ll,m,nl ( I I L)_\l
sinDsinc'
The problem can also be solved by the use of vectors. Assume axes l,m,n
Thus, (forming the conventional right-handed triad) with I pointing to the vernal equinox
sinA sinB sinC and n directed toward the north pole of the equator (@). Also assume the right-
(tr.424)
sinc sinb sinc handed triad u, v, w with u directed to the vernal equinox and w toward the pole
of the ecliptic (E).
Other formulas of spherical trigonometry (see Table 11.3.1) may be obtaincd
Then
from the two above.
Another application ofvectors to the solution of a problem in spherical trigonour-
etry concerns transforming the coordinates of a star from right ascension (a) and u=l
declination (6) to ecliptic longitude (,\) and latitude (B). v=mcos€-nsln€
Given the triangle EQS (Figure 11.42.1)on a sphereof unit radius with ccntcr w=msin€+ncos€. (r r.42-6)
at O, Q is the pole of the celestial equator, E is the pole of the ecliptic, and S is tlrc
stax. The components of OS projected onto the equatorial and ecliptic axes are
The side EQ is the obliquity of the ecliptic (€), the side eS is 90" - d, and thc
side ES is 90'-d. Two great circles passing through E and cos d cos a, cos 6 sin a, sin d,
Q, each being orthogonal
to side ES, intersect the equator and ecliptic planes at the vernal equinox. Then the cos B cos ^, cos B sin ) sin B,
angle EQS = 90' + o, and the angle SEQ = 90. - ,\.
Using these quantities the problem is easily solved by the equations previously Ieading to the original trigonometric equations.
derived:

s i n B= s i nd c o s€ - c o s 6 s i n a s i n e , 11.43 SpecificCoordlnateTransformations
cosBcos) = cosd cosa,
For geocentric spherical coordinates there are thus the four practical reference sys-
cosB sinI = sin6 sin6 + cosd sinacos €. (rr.42-5) tems of:
556 EXPLANATORY
SUPPLEMENT 11/ COMPUTATTONAL
TECHNTQUES 557
(1) Azimuth (A) is measured from the north through east in the plane of the where t, is the local sidereal time.
horizon, and altitude (a) measured perpendicular to the horizon. In as- tansformation of local hour angle/declination to azimuth/altitude involves a
tronomy, the zenith distance (z = 90" - a) is more generally used, but the rotation of the referenceframe through an angle 90" -@ in the plane of the meridian,
altitude is retained in the formulas for reasons of symmetry. where { is the geocentric latitude, followed by a rotation about the new vertical bv
(2) Hour angle (ft) is measured westward in the plane of the equator from the 180'. Thus,
meridian, and declination (6) is measured perpendicula.r to the equator,
positive to the north.
f c o s a c o s A lf - t -0l 0'l f sino 0 -cosdl fcos6cosll
(3) Right ascension (a) is measured from the equinox eastwaxd in the plane of
lcosasinAl=10 0ll 0 I 0 llcosdsinnl. (il.43-4)
the equator, and declination (6) is measured perpendicular to the equator,
positive to the north. I sina I L0 0 l . l l c o s d 0 s i n dI I s i n d I
(4) Longitude ()) is measured from the equinox eastward in the plane of the
ecliptic, and latitude (B) is measured perpendicular to the ecliptic, positive In this transformation the diagonal matrix is its own inverse and, if [R6] denotes
to the north. i h e r o t a t i o n i n t h e m e r i d i a np l a n e w i t h i n v e r s el.h e n

Formulas for transforming equatorial right ascensionand declination to ecliptic I sin/ 0 cos/l
longitude and latitude may be written using direction cosines (unit vectors) and t R dIt= l o r o I ( r 1.43-s)
frotation matrices) as follows: l-cos@ 0 sin/l

and
f cos6cos.\'lIl 0 0 I fcos6coso'l
0 0l fcosacosAl
=
l c o s l s i n l| l 0 - s i n r i n rI l c o s 6 s i nl o.
c o s / s (1,1.43-t) fcosocosfl [-l
=tR,l ' 0 -l 0l \r1.43-4)
I sind I l0 c o sI' I s i n d l I c o s d i i n i| | lcos4sinAl.
I sin6 I L0 0 llL sina I
The inverselransformalion, using [R.] lo denote lhe rolation matrix. is
The corresponding equatorial rectangula,r coordinates and distance are denoted
fcosli cos,\I fcosdcoso'l by X, Y, Z, and R for the Sun and by {, q, (, arrd A for the planets; they are derived
=Ln.l-' (1r.43-2) from the soherical coordinates bv the formulas:
l c o s t i s i n )| lcos6sino I.
I sinB I L sin6 l
XI -A = cos 0 cos a,
The transformations correspond to simple rotations about the x-axis through o,
n
e, which is the obliquity of the ecliptic corresponding to the particular equator and
! o, L=cosdsino.
ecliptic used. RA
Tlansformation from right ascension/declination to Iocal hour angle/declination ZC - (11.43-:7)
or =slnd.
may be carried out by substitution of R

ft = local sidereal time - o. Geocentric ecliptic rectangular coordinates are rarely (if ever) used.
For heliocentric coordinates there axe only the two practical referencesystems
The two systems are identical apart from the origin and direction angular mea- the equatorial and the ecliptic. In the equatorial system only rectangular coor-
surement. The transformation may also be considered as a rotation about the axis dinates are used. The relationships between the ecliptic rectangular coordinates
perpendicular to the equatorial plane and positive in the direction of the North (x",y",u), the ecliptic longitude, latitude, and distance (l,b,r), and the equatorial
Pole. Thus, rectangular coordinates (x,y, z) arc

= 'cos,cos/l
fcosdcosh] f cosr, sinr, 0l fcosbcosal fx" f'l
= r c o s b s i n=i t R . l (11.43-8)
=
l c o s d s i n lh l - s i n r r c o s t r 0 l l c o s d s i n o
(11.43-3) l)" I lyl.
l.
I s i n d .l L 0 0 ll L sins I Lr. = rsinb I Lz)
558 EXPLANATORY
SUPPLEMENT

trl
11/ ooMPUTATTONAL
TECHNTQUES 559

lll (r1.43-9)

The conversion from heliocentric to geocentric coordinates is performed in terms

from a set of axes designated by x, y, z to a set designated by x', y/, 2,. The two
systems are connected by the formulas

x = h x ' + l z y '+ l z z t , xt=hx+mry+ntz,

of equatorial rectangular coordinates using
y = m f i i + m z y+
t m3zt, yt=l2x+m2y+n22,
t=x+x'
z= ntx'+nzy'+ntz', z t= b x + m 3 y + b 4 ( l 1.43-14)
71=y+Y,

e=z+z' (11.43-10)
where lr, ,nn nl; 12,m2, n2i and.\, m3, ry are the direction cosinesof r', y,, 1, referred
to the system x, ), z. The direction cosines satisfy the relations typified by
where X, Y, Z are lhe geocentric coordinates of the Sun.
The calculation of the spherical coordinates from the rectangular coordinates,
or from the known direction cosines, typified by
rt+ml+nl=t,
l2l+m2m3+n2n3=0,
4cos6coso={,
4cos6sino=4,
ft+Pr+$=t,

4sin6=(, ( 1 1 . 4 3 -ll) mtnt+n2nZ+m3n3=0. (11.43-rs)

is performed by These nine quantities can be expressed in terms of the Eulerian angles d, /, ry'
as
tana=\/€,
A=(€z+n2+(2),
/r = +cosd cosd cosI - sin@sinl,
cotrt=€/q,
l2 = - cosd cosd sintl.t- sin$ costlt,
sin6=C/4. (rt.43-12)
/3=+cosdsind,
The quadrant of a is determined by the signs of { and 4, and that of 6 by the sign mr = + sindcosdcos?y'+ cosd siny',
of (. 4 and A cos d a.re always positive. The formulas for a and 6 may be written m2= - sind cosI sin/ + cos/cos ry',
zts m 3= + s i n d s i n d ,
n r = - s i nd c o s l ,
o=tan 4/C, or arctan 4 / €,
,12= + Sind sinly',
=cot-telI/, or arccot€ / ?, ( l 1.43-16)
h = + cos0.
6=sin r(i Z1, or arcsinC/4, (11.43-13)

In this casethe conversion corresponds to a rotation / about the z-axis, d about

provided that the appropriate values, and not necessarily the principal values, ot
the new position of the y-axis, and ry' about the new (and final) position of the z-
the multivalued functions are takel.
axis. The transformation is equivalent to a single rotation about some line not in
Many of the conversions above correspond to a simple rotation of the frame of
general coincident with one of the axes; but such single rotations are not frequently
reference about one of its axes. These are special cases of the general conversion
encountered in astronomical practice.
 560 EXPLANATORYSUPPLEMENT 11 / COMPUTATIONAL
TECHNIOUES 561
11.5 NUMERICALCALCULUS Table11.51.1
Derivatives
to an Orderof l0

11.51 NumericalDitferentiation Derivativssai a TabularPoinl Dedvativesat a HalfwayPoint

Successivederivatives of a function are closely related to the successiveorders of h!'o = p6o- - $t1," + ffid1,"
[eai . gerl hf\p = 6ttz
difierences, Thus to obtain the exact expression for the derivatives in terms of
tf t( = a,o - gri +gaf t?!\lp - p61p - fiutl," * ffiual,"
difierences it is necessary merely to difierentiate a.ny of the interpolation formulas
(see Section 11.2) as many times as required. r,3 '=pd i,o| . #a,4 - n3/!,L = 6",,, - t01," * &6r.,,"
Difierentiation of Bessel's interpolation formula leads to h'tt =6t - *rB * zk63 haff,z = p4tz - {1utl,"* ffiual,"
firE = utl Jaaj.'gaa! -. " t 1t2
-
"1t2
- fit'1.,,"* ffits.,,"
=, , , + ; ( ,- )
hfL6 jo - at*ai)/j!,,*'..,
1 6+062f 1 + (n.sr-l) n6ft|=t$ - iu3 * 1t6;o na1i," = psl,"- $vel,"+ ffiutl,o"
h7r{ =p63 $u$ . {autll -. n7til" = 4,2 - fit;1,,"+ffitll2
where fl denotes the value of df I dt at the point t = to + ph. tf t{" = tB - Jalo +Sol'? nstii2 = pt?,,- tlual,o"* ffiuall"
This formula is intended for use in the range 0 1!p I l; the maximum value of rf t,i = ud| lrall * ' '' t 1t2 - "t2
6"1r2
the third-differencecoeffcient is l/12 and occurs atp=0 and 1. At p= 11216"
formula reduces to
rrlo{ = alo - . d d - + Jl," = paf,o-
n1o fiuall"* ..
11
h J ' r , z= 6 t . z - - 6 1 ' + -a6i' - (|I.sl-2)

Difierentiation of Stirling's interpolation formula leads to

In astronomical usage, the tetms uariation arrd motion are synonymous rvith
-i "derivative with respect to time," when qualified by an adjective defining the unit
hf'o= pbx+ pb( - - 3pz)p6Jo
+ .... ( 1 1 . 5 13 ) of time, and are usually evaluated for the tabular points. The term seczlar aariation
e(t
usually implies a second derivative with respect to time.
where, for example, pd! = The second derivative is obtained by differentiating the formulas for f,, for
i(63 vz+ 6i/2). This formula is intended for use in the "f/
range -l /2 < p < I / 2; the maximum value of the coefficient of the mean third Stirline's formula this leads to
difierence is I / 6 and occurs at p = 0. At p = I 1h" formula reduces to
h ' f L = 6 3 + p p o"i +
, ( l l . s1 - 7 )
hf'o=116o - rll5l-y'\
!u8- |ua| and, when p = 0, tq

The condition for a maximum or minimum is that /i = 0. For Stirling's formula h'f;=fi-rard**4- (I l.5l-8)
this condition may be expressedas:
In terms of central differences, the derivatives to order 10 at the tabular point
-rro+*(l - sp2)p63 and at the half-way point are listed in Table 11.51.1. Primes are used to denote
(ll.5t-5) differentiation with respect to x. For example,

)2 )2
This equation must normally be solved by successiveapproximations. The maxi- h'1fi= h2;rflo +ph = (r l.5l-9)
mum contribution from the third difierence is | / 6 p6e I 63, and if this is negligible hrf e.
p may be evaluated directly from
For higher derivatives the primes are replaced by italicized roman numerals, In
general formulas,
"ff;r is used to denote the zth derivative of / with respect to r at
p = - t r , 6 o t 6 i-=) - a,t,rto3. (l r.5r-6) the point "ro+ph.
 562 EXPLANATORY
SUPPLET
ENT TECHNIOUES
1'I / COINPUTATIONAL 563

The formulas in Table 11.51.1 may also be written by substituting for the dif- of the first summation J' and to extend the table in either direction bv applying
ferencestheir values in terms of the function. Additional expressionsfor derivatives successivevalues of /. For a double integration, a starting value for /" must be
in terms of the functional values are as follows: evaluated and successivevalues of /' applied. Expressions to evaluate the first and
second integrals are derived by integrating an interpolation formula, for example,
tl
h f ' o= t-f tt
itf 6pbJIo+...
lt
=
O( fz+9fr
-8f t+f )+ -
lt
*u6'fo
.[''I,^,or="Vr,.
)0u,,-
lr'uoor,*#trrl,].
^l
=
*$z-9fz+45fr
4 5 f- t + 9 f z - J ) -
*nu6'fo+. *,=.'llt*f,r,- fin. ffi,,t ]
I l" ro, (11.52-1)
h ' f i = r f t - 2 l o +J . t t -
u6"fo+...
ll
= r 30fo+16f ' -.f r) + Additional formulas are given in Interpolation and Allied Tables and in Handbook
OGfz+16f *:6o,fo
l . ^ ^ " I ," of M athernatical Fun cti on s.
= - 2 7 f 2 + 2 7 O I t 4 9 0 1 0 + 2 1 0 [--2t 7 f . r + 2 f r ) -
tS0,2/j 560^d/0+... In the integration of orbits the problem is more complicated because the fuuc
1 tion that is to be integrated is undetermined until the integral becomesknown. Thus
hft,,,= rf . fnr - :6rf ,,, + .
it is necessary to proceed stepwise using extrapolation and approximation at each
=
li-
- 2 1 f 0 +f r ) + - ... step. Because of the availability of computing machines, as well as the rectangular
Uefz+27f 1 *16'f uz coordinates of the principal planets, two methods for the numerical integration of
l - 5 -,- the orbits have proved useful, Cowell's and Encke's.
= (9ft- r25f2+22s0f| + r2sf-1 -9f )-
22sofo
,r 1ftr5'fto+. When one is utilizing Cowell's approach, the equations of motion are integrated
t5 to give directly the positions of the perturbed body in rectangular coordinates. For
h zI ' t?' = tt fiuboltt+...
2(lr-ft-fo+I a planet or comet, the origirr usually is the Sun and at each step the attractions of
= I + 39f2 - 34f1 34fo+39f| 5l_2,+
)ss
uz-.... the perturbing bodies would be added to that of the Sun.
Vr(-Sfi ffiut'f With Cowell's method, however, the origin may be any perturbing body, or even
( rr . 5 l 1 0 ) a double star. sincc no use is made of a conic section as the first approximation Th<:
single condition is that the motions of all perturbing bodies relative to the selectetl
For further formulas sce tlne Interpolation and, Alli,ed Tables and the l"Iandbook origin should be known. A disadvantage of Cowell's method is that the integrals
of M at h emati cal Functi on s. comprise nunlerous sigpificant figures that vary extensively with time. Thus, tlrt:
integration tables are prone to converge rathcr slowly, which may rcquire thc use
'11.52 of a short step intcrval. The basic equations in the Cowell approach are
Numericallntegration
When studying motions within the solar system it is necessary to use the soiutiotr
of ordinary differential equations of the second order. This normally invoh'es the
+lk2n1(;
i= k2(t+m1] (.t1.52-2)
simultaneous integration of three second-order differential equations to produce an il
ephemeris in rectangular coordinates of a planet, comet, or satellite $'ith time its
the independent argument.
In numerical differentiation, the derivatives of a function are expresscd in ternls with similar equations in y ancl z. The massesare exptessed in terms of the lrass oi
of the differences.In numerical integration, the opposite method is used. Derivatives the Sun, ,t equals the Gaussianconstant, m is the mass of the perturbed bodr. alL'l
of some function are computed and checked by differencing, to form a table of /, mi the massesof the perturbing bodies. The frrst term is the attraction of tbc StttL
6,62,63,64, and so on. In the simplest case,where the valuesofthe function to be on rn. and the first term in parentheses is the attraction ol mi ot m. The last term
integrated can be calcrrlated in advance, it is possible to compute a starting value in parerrthesesis the action of the perturbing bodies on the Sun.
 564 EXPLANAToRY
SUPPLEMENT 11 / COMPUTATIONAL
TECHNIOUES
In Encke's method the unperturbed orbit of the mass rn about the Sun is
assumed known. Hence only the difference between the unperturbed position and
f3= f eo+ h,xo+ hf),
the perturbed position needs to be determined. The equations may be written ( I I 5r-7\

i - io=
t =*0.^,(i - i).4r*, (; (l r.52-3)
Then
I
il xlto+ hl = \ = xs + ;(fn + 2f 1+ f2 + I.tl.
o

with simila.r equations for ri and (. These equations could be integrated directly, Similar equations exist in y and z.
but since { is small, many figures would have to be carried in the term (-16lfo_x/f) Equations to the fourth order are not sufficiently accurately for most orbital
to retain significance in the difierence. On the assumption that {, q and ( are so calculations but this serves as an example of the method. For modifications of the
small compared to t, /, z that their squares can be neglected, the equation can be method refer to Danby [Section 11.7].
transformed by the use of auxiliary quantities 4 and / where In all the step-by-step methods for numerical solution of difierential equations
some precautions apply:
-_xo€+lon+znC
q----;-'
ro (1) The interval of tabulation must be monitored in cases of close approach,
l-11+2n1-zrz since the approximations used in the iterative process may prove to be of
q
(tt.s24) insufficient accuracy. Automatic methods to vary the step size are now in
use with some methods.
The final Encke equations become (2) The accumulation of round-ofi errors must be monitored. The probable
error after n steps can be estimated.
(3) If the method of solution adopted makes use of truncated formulas either
. 2 € = . 2 P (*t^ r \ r 1 q ,- {-t + l , } p n'\1 ( r + - + \ (l l.sz-s) corrections must be applied subsequently or the number of figures in the
'd"' ? oi 1) solution must be reduced accordinslv.

Where o represents the interval of integration. Similar equations hold for

zj and {. Examples of numerical integrations may be found in Brouwer and Clemence
Encke's procedure normally allows a larger step size than Cowell's but when (1961) in Astronomical Papers American Ephemerides,vols. XII and XX, pt. 1,
close approaches happen, the Encke perturbations tend to grow
rapidly in size, and in Danby (1988) (Section 11.7).
requiring a smaller step interval.
An example of a multistep scheme is the Adams-Bashforth method. The order
Other approaches to the numerical solution of difierential equations are pro- of this method is limited by the number of terms the user is willing to compute.
vided by Runge-Kutta methods. These are single-step methods with
each step being The formulas for this method are
independent of those preceding.
Given: /v
dx x4+t=xn+n (11.52-8)
= x(to)= xo. (t r.s2-6) 22akv-1fu
dt I(t,xli *=O

A widely used Runge-Kutta method (to the foruth order) involves the following where
operations for one step. Here ft is the step interval.
N = the number of terms,
.fo = | 5 ? t5t q5
"=t'i'o'i'r^';A' '
"f(ro,r),
/,=r(ro*)r,^.1^r"), v k f ^ = v r - t f ,- Y r tf , - r ,
voi, = f,,
n=r (',.)n,^.)nr,)
,
1" = f(x",il.
 t l / COMPUTATIONALTECHNIQUES co/
566 EXPLANATORY
SUPPLEMENT
T a b l e1 1 . 6 1 . 1
This method can be started using a single step method such as the Runge-Kutta. Accumulationot Error in ArithmeticalOperations
To differentially correct the initial conditions of the motion, one integrates seven
orbits simultaneously. The first orbit uses unaltered initial conditions. Each of the Operation AbsoluleError Varlanceot Error(unit 1/12)
rest ofthe orbits usesinitial conditions for which one coordinate or velocity has been
incremented by a small amount. Earth based observations are right ascensionsand \-n r'.
Di=,a' Di=t't
declinations of the satellite. The residuals for these observations can be expressed
nAl|.,niAi nal|.,niAi nf nioi
"li-.,
A^ A^. A^. A^. n^. )oi ^+ (n large) jrPa if a1= a Je161
;
Ao1= ll AXs+ i#Ayo + #n + l--.'AXo+ ijAyo + !i, ":4i Li=1iat
olo olo oz4 l./'o olo ah"^'
ai a^. 45, 45, 46. fr,, *eli_., &ei ka+ el'l-., kiai+ e t(a + t li_., tlal + t
Adi= =Axo+#Avo+ # L + =Axo+ =AYo+=A .
OAO OIO O14 dxn Ofo O4l
ktA kn+ e $t(a + t
(l 1.52-9)
s@)A g\p)a+ e Il{s@)}'ap."tr
where o and 6 are the right ascension and declination respectively; X6, Y6,74, Xs,
to, and Zt are the initial coordinates and velocities; the subscript i refers to the \i=.' oiei @roi = o1 Li=.' P'a' * e ( 1+ l + | + . + f r ) a + 1
number of the observation. The partial derivatives are evaluated numerically, and At Az A2Q + A1a2+ € Alctl+Al,a2+1
can be written in the form
nAn-|a+e f.A2n 2o+1
)oi - cti(Xo,Yo,4,Xo,
a;(Xs+ A.Xs,Ys,ft,Xo,Yo,h) Yo,h)
( 1 1 . 5 21 0 )
AAO a"a 1
l-$\a+e (1 tAala+1
A
The differential corrections can be written as A1t2 ( | A 1 t 2 1 a e+ (lA)a + 1

LR = (ArA) |Ar Lr. (rr.szrt) \ (!^)a'lfrl+* e (jllot+{}l"z+r

where AR is the matrix of corrections to the initial conditions, A is the matrix of rlm differenceof A(x) if 't is constant
Di=1e1\t(i'tai +,?iu
the partial derivatives, and Ar is the matrix of residuals.
f(At,Az,. .) f f i a , * f f a z * "* e ( f , [ ) "" ' * ( ] f r ) ' . 2 +* t
11.6 STATISTICS J(A) by direct interpolation aft(A, + et + ei + e Ilt(Al]lza + vt + vt + 1

J(A) by inverseinterpolation la - et e!)f'(A) + e {ft(Al]}2ld+vt+vi+1

11.61 The Accumulation
of Error
The formulas listed in Table 11.61.1 indicate the way in which errors accumulate
in simple arithmetical operations. The distribution of accumulated error is usually
nearly normal, whatever the distributions of the original errors, so the standard
deviation of the error, calculated from the formulas, is of value in the examination @i is the va ance, or mean square value, of the error of A;. The standard deviation
of tlte differences of tabulated functions. The notation used and assumptions made of the error is the square root of variance.
in the table are:
i, m, arrd.n are integers. t is a known constant. r is equally likely to take any value
c; is the absolute error of A;, in the sense of given value minus true value. These
between -l /2 and +1/2. p is equally likely to ta.ke any value between 0 and 1.
errors are assumed to be unbiased and independent.
 568 EXPLANATORY
SUPPLEMENT 11 / COMPUTATIONAL
TECHNTOUES 569
k., r, an.d,p are considered to be ftee from error, that is, they are exact or are known mathematical theory of systematic errors, we must proceed initially on the assump-
to more significant figures than the quantities A;. tion that the errors are accidental. In this case if a set of random variables x are
independent of each other, the theory of probability postulates that the frequency
e is the error arising from the rounding of the final result; it is assumed that a distribution of the sum of such variables, each with its own frequency distribution.
guarding figure is kept in intermediate results. e is equally likely to take any
will slowly approach the function
value between ! 1 l2 so that its variance is 1 / 12, equal to the integral of e2
between 1 l/2, in units of the last place retained. For convenienceall va,riances h
Y= "^
exp(-h'x:), (| 1.622-I)
are expressed in this unit of I / 12.
Ji
In addition, e. is the error due to the rounding-ofi errors in the tabula,r values,
including their effects on the difierences; e; is the truncation error, and v. v, are where I is defined as the index of precision. This is related to the standard deviation
the variances of e, and e1. If all values of the interpolating factor p are equa,lly o, or the root-mean-square error by h2 = 1 / 2o2.
likely, v, varies from 0.67 for linear interpolation to 0.81 for interpolation with fifth The best fit to a series of n observations, according to the principle of least
differences. If /(A) is determined from a critical table its error is that of a simple squares, is the one that reduces the sum of the squares of the errors to a rninimum.
rounding-off error unless the number of figures in the argument is small or unless Assuming that the index of precision is constant in this set of observations and
rhe interval in the respondenlis grealer than uniry. letting the errors associated with each observation, O;, be .q, the probability, P;, of
occurrence of error ,ri becomes

11.62 TheMethodof LeastSquares

r, = (1r.622-2
)
11.621 Introduction When one is measuring the physical universe, certain con-
ftexpr-n2ft)tut.
stants must be determined from observation, such as the orbital elements of a
Assuming that P is independent, the probability that each error will occur in turn
satellite about its prima.ry or the rotational elements about its axis. Methods for aD
IS
initio determination of orbital elements from three or more observations are well-
documented in the literature (Herget, 1948). The more typical case assumes that / h\' / ^ -l- ^\
p= | i I expf -h' > x| ldxt drz ... d,c^. (11.622-3)
approximate values of the constants are available on which a provisional theory \ v,r,/ \ ii /
can be based. This theory is then used to construct positions against which the
observations can be compared, Thus, each observation minus the computed place
Thus it can be stated that the set of errors that maximize P in Equation 11.622 3
forms a difference, referred to as an O - C. The errors in these differences can arise
will produce the best fit to the observations because at that point !i=t;rf is a
from three sources: errors in the observations; errors in the constants on which the
minimum.
computed positions are based; and errors in the theory itself. Only the first two er-
If the index of precision varies with each observation, the introduction of the
ror sources will be treated in the following discussion, which concerns the analysis
factor llj requires knowing the correspondence between the ft; and the arbitrary
of O - C's to improve the initial approximation to the constants.
weight of an observation, based on unit weight. If the precision in6ls1 6611esponding
to weight 1is h, and the precision index correspondingto weight w1 is,1, the
11.622 Errors ofObservation: Fr€quency Distribution Errors of observation may probability that an error of size.r will occur in an observation of weight 1 is
arise lrom many sources. The size of these errors vary and, in general, follow the
normal distribution curve or probabilitv distribution curve wherein small errors h --
occur more frequently than large errors, and positive errors occur with the same P= n'x'tdr. (rr.6224)
f,ex
frequency as negative errors. In most problems, the frequency distribution of the
errors is not known in advance. However, if each observation is assumed to be af- Also the probability of an error of size .t occurring in an observation of weight w1 is
fected by the sum of severa,lsimultaneous, independent errors, we may assume the
distribution of the errors to be accidental rather than systematic, the latter be-
ing the case where all the errors have a common origin. As there is no satisfactory
p=+exp( h?x\dx. \l1.622-5)
 SUPPLEMENT
EXPLANATORY 11 / COMPUTATIONALTECHNIOUES
571
570
errorof the arithmetic
and the mean-square
Sinceol = llln,then fi =llln-l
Assumingthat P = Pr, that is that an error of size x has an equal probability of
mean in terms of the residualsis
occurring in both Equations Il-622 4 and 11.622-5, leads to
T._ (t r.622-rr)
v=(h/h)2,
wi=(hith)2, (rr.6224)
quantity measured is
11.623 The Method ofleast Squares In many problems the
one unknown, 4, to be
not the one that we wish to corlect. A simple case involves
which shows that the weights are proportional to the precision indices ln terms of = T husC=f(4) Let?.be
( ; ,
d e t e r m i n e df r o m m e a s u r e m e n t s w h e r ei l , 2 , 3 ' ",n
the precision index i and the weights w;, the value to be determined from
the provisionally known value of 4, and 4s be the
- = from 4" for each
the measurements, then the difference n0 nc A4. Calculating
a",...a'^.
, =("-o#)"' r^"*rl-of ,,,7]a,, (t t.622-1)
measurement i, gives A1( = e - G.Using a Taylor's series expansion' A( may be
written as a function of A4,

showing that for P to be a maximum, the !',!, w; -f must be a minimum'

Assume Oi (where i = l, 2, 3, ..., n) to be observations of a quantity for which r N =9!a,tnn . !ffioa'. ...
+)dfi<l'n,' (l1.623-l)
-
is the true value. The error ri of each observation is x; = Qt T Using the standard
deviation o instead of the precision index h, the best estimate 7 of the true vahte powerscan be ignored,
Assuming A4 is snall enough that its squaxe and higher
r is that which reduces then
ttzo2l(ot-^i2 / aJ\ Ln = L(; (rt.623,2)
\a?/,
- l) = 0, frorn
to a minimum. To do this requires that A DlAl = 0; i.e., 2 !(Oi a known numerical
which 7 = | I nD Oi, or the best estimate is the axithmetic mean' and we have r equations where in each case the (df / d4) is
condition can be written
Given n observations Oi, having normally distributed errors,t;, the error of the coefficient. Denoting the coefficients by A, the equations of
arithmetic mean is AAa = A(. The preferred value of the unknown is that for which

, = l y-r., (t.622-8)
(rt.623-3)
^? =)-rA;Aa-Ai(t2=0.
oAq
which is a linear combination of n random variables. If the mean error ot one
--
observation is o, then it can be shown that the mean error of the arithmetic meatr Difierentiating gives !A2A4= IAA(, thus 1r4=| A4ri'f UA2 an<lA'r1+q' r1s'
is the o of one observation divided by the y4. mote unknowns
In a more general problem it nay be necessary to determine
Once the estimate 7 of the true value r of the measured quantity is obtained, a planetary orbit C;
than one from a series of measurements The constants of
a residual r can be formed by Oi-1= ri, Also Oi-r = rr, where the,ti are the tme and can be used to
(where i = l, 2,3, ...,n ) may already be approximately known
errors of the observations. Then qo - rl. = A4' Let the
compute the 4. corresponding to each observation 7b, then
be E1' so that the
corrections to the approxrmate values of the orbital constants
!r,=f o , - n ' r = n 1- n r , corrected constants are Ci 1Ei. The equations of condition can
then be written

fru,*![u,*..*ffiu.=
t,.
1=r+Lxi/n,
(rr.6234)
ri=xi_ Dxi/n- (t | .622-9\

= C' and so on' the conditional equation

Let o, be the mean error of the residuals and o' be the mean elror of the errors r;, Setting d4 / 0C1 = A, 0n I \Cz = B, En I \Ct
then can be written
A\+BE2+Cfu+" + N E n =L n . (l t.623-5)
- r) t fi2fi + @- | t n 01= tu - | / fifi
ol = 11n (r r.622-10)
cll EXPLANATOBYSUPPLEMENT 11/ COMPUTATTONAL
TECHNTQUES 573
Supposing there are no errors in the A4, i.e., that the theory used is adequate, and When the rz equations of condition are not all of equal weight, each equation mav
that the observations contain no errors, then the only source of error in the solution be multiplied through by the square root of its weight before forming the nq1m61i.
comes from the approximate values of the constants. The theory of linear equations Methods of solution of the normal equations are numerous and are detailed in
states that z independent equations of condition are necessary and suficient to the referencesgiven at the end of this chapter.
determine the Ei. With the above restrictions as to the source of errors, ihe solution An example of the method of least squares is the problem of fitting a straight
will yield values of E that will exactly satisfy the z equations. Normally, however, line through a series of measured points, where the function A + ,84 = /(4) is real
accideltal errors in the values of A4 prevent the derivation of the exact values of and continuous.
Ei, Increasing the number n of observations and thus the number of conditional
equations will decrease the effect of the accidental errors. In some problems in Example 1 The data are
dynamical astronomy it is not unusual to have rz exceed n by a factor of 1000.
Since values of d are derived that do not exactly satisfy the conditional equations, ?lo 0.80 1.60 2.40 3.20 4.00
the procedure is to evaluate the left-hand side of the equations and subtract from
the corresponding A4 to produce residuals r. The best values of the E; are those /(?) I 1.01 3.8'/ 6.52 9.36 12.0'7 14.87

that reduce the sum of the squares of the residuals (r) to a minimum as discussed The coefficients and right hand members of the six equations of condition can be written
earlier. in tabular form:
In the instance where there are m equations of condition, with m > r, the A B f(n)
procedure is to form n normal equations for solution. The normals are formed as
1 0 1.01
follows. Given the conditional equations, 1 0_80 3.87
1 r.60 6.52
1 2.40 9.36
1 3.20 l2.O'7
A1E1+ByE2+CP3+ ..+ NlEn= [711, 1 4.00 14.87
A2E1+ B2E2+ C2E3+ ... + N2E" = A,q2,
Assuming all of the equationsare of equal weight, the normal equationsare
A3E1+ B3E2+ C3E3+ ...+ N3Ea= A?:, etc., (1r.6234)
6A + l2B = 47.70
lU+35.28=134.096, ( | L623-8)
multiply through each conditional equation by the coefficient A, and sum to form and the solution given by the least squaresapproximation is
the first normal equation in the axray that follows, where the brackets represent
summation. Then multiply through each conditional equation by Bi and sum to A = +1.04
fbrm the second line of the array. Repeat with C; to form the third line of the array. B = +3.455
The procedure is repeated through the coefrcient of En and the symrnetry of the or
resulting a,rray of n equations in n unknowns reduces the complexity of the solution. (rr.6239)
f ('|i = z(n) = l' 04 + 3. 455n.

The valuesderived from the solution comparedto the measurementsare

lA?\4 + lAiBilE2+ IAiC ilr4 + .. + IAiNilE^ = lA7Lq l, 4 0 0.80 1.60 2.40 3.20 4.00
J(rt) 1.01 3.8'7 6.52 9.36 l2.Oi 14.87
lBtAiq + [B?I E2+ IBic i]h + ... + lBiN ilE, = IBiLn il, z(d I.M 3.11 6.54 9.30 12.07 14.83
r .03 .10 .02 .06 .00 .04
[CiAi]h + lCiBilEz+ lqw3 + '.. + lciNilE,t = lci\nil. 01.623-i)
The sum of the squaresof the residualsis 0.0165and the root mean squareerror of the
six data pointsis 0.052.
 574 EXPLANAToRY
sUPPLEMENT

11.7 REFERENCES

.Lbtamowilz, M. and Stegun, l. (1973). Handboole of Mathematical Functions, 9th ed.

(Dover Press, New York). CHAPTER
12
Bowditch, N, (1981). Section on mathematics. Arnerican Practical Na\igator (U.S. Gov-
ernment Pdnting Office, Washington, DC) pp.436-441.
Brouwer, D. and Clemence, G.M. (1961). Methods of Celestial Mechanics (Acadernic Press.
New York). Calendars
Chauvenet, W. (L850). Treatise on Plane and,Spherieal Tligonometry (H. Perkins, Philadel-
phia).
Chauvenet, W. (1892). A Manual of Spherical and, Practical Astronomg Philadelphia, 5th
by L.E.Doggett
ed. (Constable and Company Limited, London) (Reprinted Dover Press, New york,
1960).
Danby, J.M.A. (1988). Fundamentats of Celestial Mechanics 2d ed. (Willmann-Bel1, Rich-
mond, vA).
Duncombe, R.L. (1969). "Coordinates of Ceres, Pallas, Juno and Vesta 1928-200O,,Astro.
Papers Amer, Eph. XX,I (U.S. Government Printing Offce, Washington, DC).
Eckert, W.J., Brouwer, D., and Clemence, G.M. (1951). ,,Coordinates of the Five Outer
Planets 1653 2060" Astro, Papers Amer. Epn. XII (U.S. Government Printing Office,
Washington, DC). 12.1 INTRODUCTION
Fox, L. and Parker, LB. (1972). Chebysheo Polynomials in Numerical , aalgsris (Oxford
Univercity Press, Oxford, England). A calendar is a system of organizing units of time for the purpose of reckoning
Herget, P. (1948). Computatio4 ol OrAits Published privately by author
time over extended periods. By convention, the day is the smallest calendrical unit
Hildebrand, F.B. (1987). Introd.uction to Numerical Analysis (Dover Press, New york).
of time; the measurement of fractions of a day is classified as timekeeping. The
H.M. Nautical Almanac Ofrce (7956). Interpolation and, AIIied, ?a6les (H.M. Stationerv
Office, London). generality of this definition is due to the diversity of methods that have been used
Jennings, W. (1964), First Course in Numerical Method} (McMillan, New york). in creating calendars. Although some calendars replicate astronomical cycles ac-
Lanczos, C. (1956). Applied Analysis (Prentice-Hall, Englewood Clitr, NJ). cording to fixed rules, others are based on abstractr perpetually repeating cycles of
Press, W.H., Flannery, B.P., Teukolsky, S.A., and Vetterling, W.T. (1986). Numerical no astronomical significance. Some calendars are regulated by astronomical obser-
Recipes, The Art of Scientifi,c Computing (Cambridge University Press, Cambridge,
vations, some carefully and redundantly enumerate every unit, and some contain
England).
Scarborough, J.B. (1955). Numerical Mathematical Analgsis,3rd ed. (Johns Hopkins Uni- ambiguities and discontinuities. Some calendaxs are codified in written laws; others
vercity Press, Baltimore). a r e I r a n s m i t t e db y o r a l t r a d i t i o n .
Smart, W.M. (7958). Combination of Obseruatioas (Carnbridge University press, Cam- The common theme of calendar making is the desire to organize units of time
bridge, England). to satisfy the needs and preoccupations of society. In addition to serving practical
Whittaker, E.T. and Robinson, c. (1944). The Calculus ol Obseruations,4th ed. (Blackie,
purposes, the process of organization provides a sense, however illusory, of under-
London).
standing and controlling time itself. Thus calendars serve as a link between mankind
and the cosmos.It is little wonder that calendars have held a sacred status and have
served a.sa source of social order and cultural identity. Calendars have provided
the basis for planning agricultural, hunting, and migration cycles, for divination
and prognostication, and for maintaining cycles of religious and civil events. What-
ever their scientific sophistication, calendars must ultimately be judged as social
contractsr not as scientific treatises.
According to a recent estimate (Fraser, 1987), there are about forty calendars
used in the world today. This chapter is limited to the half-dozen principal calendars
in current use. Furthermore, the emphasis of the chapter is on function and calcu-
lation rather than on culture. The fundamental bases of the calendars axe given)
5/C
576 EXPLANATORY
SUPPLEMENT 12/ CALENDARS 577

along \Mith brief historical summa,ries.Although algorithms are given for correlat- In the preceding formulas, I is measured in Julian centuries of Terrestrial Dy-
ing these systems, close examination reveals that even the standard calendars are namical Time (TDT), which is independent of the variable rotation of the Earth.
subject to local variations. With the exception of the Julian calendar, this chapter Thus, the lengths of the tropical year and synodic month axe here defined in days
does not deal with extinct systems. Inclusion of the Julian calendar is justified by of 86400 seconds of International Atomic Time (TAI). (See Chapter 2 for further
its everyday use in historical studies. information on time scales.)
Despite a vast literature on calendars, truly authoritative references, partic- From these formulas we see that the cycles change slowly with time. Further-
ula.rly in English, axe difficult to find. Aveni (1989) surveys a broad variety of more, the formulas should not be considered to be absolute facts; they are the best
calendrical systems, stressing their cultural contexts rather than their operational approfmations possible today. Therefore, a calendar year of an integral number of
details. Parise (1982) provides useful, though not infallible, tables for date con- days cannot be perfectly synchronized to the tropical year. Approximate synchro-
version. Fotheringham (1935) and. the Encgclopedia of Religion and. Ethics (197o), nization of calendar months with the lunar phases requires a complex sequenceof
in its section on "Calendars," offer basic information on historical calendars. The months of 29 and 30 days. For convenience it is common to speak of a Iunar year
sections on "Calendars" and "Chronology" in all editions of the Encycloped.iaBri- of twelve synodic months, or 354.36707 days.
tannica provide useful historical surveys. Ginzel (1906) remains an authoritative, if Three distinct types of calendars have resulted from this situation. A solor
dated, standard of calendrica.l scholarship. References on individual calendars are caleru),ar,oI which the Gregorian calendar in its civil usage is an example, is de-
given in the relevant sections. signed to maintain synchrony with the iropical yeax. To do so, days are intercalated
(forming leap years) to increase the average Iength of the calendar year. A lunar

12.11 Astronomical
Basesof Calendars calendar, snch as the Islamic calendar, follows the lunar phase cycle without regard
for the tropical year. Thus the months of the Islamic calendar systematically shift
The principal astronomical cycles are the day (based on the rotation of the Earth with respect to the months of the Gregorian calendar. The third type ol calendar,
on its axis), the year (based on the revolution of the Earth around the Sun), and the lhe lunisolar calend,ar,has a sequenceof months based on the lunar phase cycle;
month (based on the revolution of the Moon around the Earth). The complexity but every few years a whole month is intercalated to bring the calendar back in
of calenda.rsarises because these cycles of revolution do not comprise an integral phase with the tropical year, The Hebrew and Chinese calendars are examples of
number of days, and because astronomical cycles are neither constant nor perfectly this type of calenda,r.
commensurable with each other.
The tropical Eeor is defined as the mean interval between vernal equinoxes; it
theWeek
Basesof Calendars:
12.12 Nonastronomical
corresponds to the cycle of the seasons. The following expression, based on the
orbital elements of Laskar (1986), is used for calculating the length of the tropical Calendars also incorporate nonastronomical elements, such as numerical cycles,
year: loca,l environmental observations, or decisions by societal authorities. In the Gre-
gorian calendar, the week and month are nonastronomical units, though the month
- 0.000006153597- 7.29x10-t0T2 +2.64 x 10-1013 (12.11-1)
365i2421896698 can be traced back to calendars that counted the phase cycle of the Moon'
The origin of the seven-day week is uncertain. As a continuously running, un-
where T = (JD - 2451545.0) 136525 and JD is the Julian day number (see Sec- interrupted cycle, it comes to us as part of Jewish tradition. However, Biblical
tion 12.7). However, the interval from a particular vernal equinox to the next may and Talmu[ic texts indicate a varietv of conflicting calendrical practices. System-
vary from this mean by several minutes. atic observance of the Sabbath every seventh day may have developed as late as
The sgnod,icmonth, lhe mean interval between conjunctions of the Moon and the Babylonian Exile in the sixth century n.c., followed by a period of gradual
Sun, corresponds to the cycle of lunar phases. The following expression for the syn- acceptance.
odic month is based on the luna.r theory of Chapront-Touz6 and Chapront (1988): The number seven had mvstical and cosmological significance ihroughout the
Semitic cultures. It was used in the Babylonian and Assyrian calendars, though not
2 9 i 5 3 0 5 8 8 8 5+301. 0 0 0 0 0 0 2 t 6 2 1 I 3 . 6 4x l O r 0 I 2 . (12.rr-2) as a continuous cycle. Although the seven-day week may have its origin in number
mysticism, astronomy may have contributed to the mysticism. The Sun, Moon,
Again T = (JD - 2451545.0)/36525 and JD is the Julian day number. Any particular and naked-eye planets comprise the seven "wandering stars." Seven is a useful, if
phase cycle may vary from the mean by up to seven hours, inexact, count of days between the four Moon phases, and four times seven is a
 578 EXPLANAToRY
SUPPLEMENT 12 / CALENDARS 579

useful. if inexact. estimate of da1'sfrom first to last visibilitv of tho \loon in its calendar maintaius s]'rt<llrolr_ rvith astronomical phcrrorncna. Horvevcr_.an obser-
phase cycle. vational calendar exhibits sltort-ternl uncertaintv, becattsc tho rratural phenornela
In Jewish practice, the days of the week were designatecl by nrrrrrbers rather are complex and thc obsclvations are subject to error,
than by names, except that day 7 was known as the Sabbath, Recognition of thc
week and the observance of the Sabbath as a rcst day gradually spread to the 12.14 Historical Eras and Chronology
Roman world. The use of names to dcsignate the days of the wr:ek developed irr
the Roman culture of the second and lirst centuries B.c. Thesc names come frorn The calendarstreated il this chapter, except for the Chincsc c:alendar(see Sec-
the astrological practicc of narning each day after the planct that governed the tion 12.61). havo courrts ol 1'earsfrom initial epochs. Irt tlto case of the Chinese
day, At that time, thc Sun and \Ioon u'ere included amorrg the planets. ,,Saturn s calendar and sone calcndirrs not included here, -vcarsarc corrnted in cvcles. r';ith
day" coincided rvith the .JervishSabbath. The planetarl'Damcs wcre gradualll-.if no particular c:'cle spccificclas the first c]-cle.Somecultures eschr:u'r'earcoults al-
reluctantly, acceptedby the Jervishand Christian culturcs (Garrdz. 19.19). together but narneeach tr':u:rftcr an event that chalacterizedthc \'ear- Ho11er-er. a
It is a reasonable,but ultirnately unpror,able.assumptionthat the cyclic conti- count ol1'carsfronr att ittitial cpoch is the most successfulrvzn'ofmaintaining a con-
nuity of the u,'eekwas naintained without interruption by rcligious authorities fron sistent chronology.Wlrcthel this epoch is associatedq'ith an historical or legendary
its origin in Biblical times to the present day. Although ten <layswere dropped from event, it must bc ticd to il sequenceof recorded historica,levcnts.
the Christian calendar in thc Gregorian Reform of 1582, the cyckr of weekdays u.as This is illustratecl bv thr: acloption of the birth of Clrrist as thc initial epoch
not disturbed. Thus, in 1582, Thrrrsday. October 4, of the .Irrlia,n calendar. ,"r,as of the Christian calendar'. This epoch rvas establishc<lby thc sixth <rcnturv scholar
followed by Friday, Octobcr 15, of the Gregorian caleudar. Dionysius Exigrrus,rvho u'as <rrrnpilinga table of datcs of Eastcr'.Arr existing table
Colson (1926) and Zcnrbavcl (1985) are standar.drefercrrccson the lieek. The coveredthe Dineteen-\'oarl)orio(l (lenoted228 2.17,whcrc ycars w(!\r r:ountcdfrom
latter reference also considers other rveek-like numer.ical cvclcs used bl past ancl the beginning of the rcigrr of thc Rornan empelor Dioclotiiur. Dioulsius continued
present cultures. the table for a rtirtetcerrrtar pcrirxl. s'hich he clesignatcd.\ruti l)ornini \ostri Jesu
'I'hus.
Christi 532 550. Diorrlsius' .{nno Domini 532 is cqrrivalcntto -\nno Dioclc-
tiani 2.18.In this sal a corrcsporrdcnce sas establisheclbct$r'('il the nel-Christiart
Era and an existing slslenl associaledrvith historical recor<ls.\\'hat Dionl'sius did
'12.13
Calendar
ReformandAccuracv not do is establishiur ?lccura.tc datc for the birtir of Chlist. .\lthorrgh scholarsgen-
crally believcthal, (lhrist rvasborn sonreyears before A.t). l. the historicalevidencrr
In most societiesa calcrrclarr.cform is an extraordinary event. A<loption of a calendar is too sketchy to zrllxr.a dofinitir-edatiug.
depends on the forcefulness with which it is introduced antl on the u,illingness of Given an inili;ll clxrrlr, ouc nust considerhourto record Prcccdirrgdates. Becle.
soclety to accept it. For cxample, thc acceptance of the Glr:gorian calendal as a thc eighth-centuryErrglislrhistolian, bcgan the practice of t ounting vt:arsbackrvarrl
worldn'ide standard spanrrcd nrore than three centuries. frotr A.D. 1 (scc Colgravc ancl Nlynors, 1969). In this svslcIn, tlic vear A D. I ts
The legal code of the United States does not specif\. an officiarlnational calenciar. precededby thc 1<rar- I Il.( .. $'ithout arr intef"ening year'0. Bocartst'oftbenurnelic:rl
Use of the Gregoriancalcndar il thc United Statessteutsfr.orrran Act of Parlianrent discontinuitl', this lrislorical" s1-stcnris cunlbersomofot cortpalittg .rllcieIrt alld
ofthe United Kingdonr in 1751. rvhich specifieduse ofthc Gregorian calendar il modcrn datcs. Ttxlar'. astrononrersu-se+1 to desiglate.\.1). l. TLell *1 is naturalll
England and its colonir.s.Horvcvcr.its adoption in thr' l.lrritctl Kinpidornand ot]rcl preccdedb1 1car 0. t lrit h is precctlcdb1 r'ear l . Siticotho rrscof ncliativenulnller-s
countries ivasfrauglrt tvith confrrsion,coltroversv. aud (]vcrlviolerrce(Bates. 1952: developedsklvlt' iu FJuropc.this ''astronotnical sl'stenlof rlatilrg rras dr:la1-e<l until
Gingerich, 1983; Hoskin, 1983). It also had a deeper cultur.al inrl)a(:tthrough the the eightccnth ct:nlrrrt,.u'lx'n it rv;rsintroduced b1'thc astrouorrrot.IacquesCassitli
disruption of traditional lcstivalsand calendricalprncticcs (trItL<rNcill. 1982). (Cassini, 174t)).
Because calcndars arc crcatccl to serve societal needs, thc qu<:stion of a calerr- Even as rrseof l)iottysirrs'C'hristianEra bet:antt:trrrtttttottitt ecclcsiasticalu'rit
dar's accufacy is usually rrrislcaclingor misguidcd. A ctrlcurlar l.hzrt is based on a ings of the \Iiddle Agcs. 1,ra(litionaldatirrg frorn regnal yoars coutinuedin civil use.
fixed set of rules is trccurate if the rules are consistentiv nppliol. For calendars that, In the sixteenth contury, .losrph .ltrstusScaligertricrl 1o t<:solv<: l,he patchwork ol
atteurpt to rcplicate astrorrornicalcvclcs.one can ask how ac<urately the cyclesarc historical eras by yrll<irrgcr'<:r.ythingon a singlt'systcnr (Scaliger. 1583). Instea<l
rcplicated. Hou'cver, astrorronrical cycles are not absolutely (.('lst;rDi.. and thel.arr: of introclucingnegative ycar counts. he sought an initial eJrocirin advanceof aly
not known exactlv (scc Ser.tiorr12.11).hr the long ter.m.onh a purely obser'\ational historical record. His llurncrologicalapproach utilized tlrlce ca.lcndricalcycles:thc
 EXPLANATORYSUPPLEMENT 12 / CALENDARS 581
580
28-yearsolar cv<:le.the rrittcteen-1-carc:1cleof GoklelrrNutnbers,atrd tlrt: filtcerr-year fable 12.21.1
'fhe Monthsof the GregorianCalendar
indiction c1.cle. solar clcle is thc period after rvhich $'eekdavsan<l calendar
peliod
dates repeat in the Julian calendar"The c1'cleof Golden \rtlnbers is the 1. January 31 7. July 31
after which mool] phases rcpeat (approximatelv) on the satne calendar dalcs fhc 2. Febtuary 28' 8. August 31
a year March 31 9. September 30
indiction cycle rvasa RoDru,ntax c1-cle.Scaligercortld thelelbte charactt'tize 3.
I through 2U. G frorl 4. April 30 10. October 31
by the combination of rtttlttbers(S,G,1). n here S ruus ftoln
5. May 31 11. November 30
1 through 19, and 1 from l through l5 Scaligcr noted that zr given cotltbinatiotr 6. June 30 12. December 31
would recur after 7980 (= 28 x t9 x 15) years l{c called this a Julian Pc'riod, b<r
' In leap year, Februaryhas 29 days.
cause it was based on the .lulian calenclar year. Rrr his initial epoch scaliger chos|
a

the year in rvhich S, G. and 1 n'erc all c<1ualto 1. Hc knerv that the l'ear 1 D C sas
characterized by the numbcr 9 of the solar c1-dc, by the Goldcn Nuntber 1, and lrv
the nurnber 3 ofthe indiclion cycle. i.e., (9, 1,3). He fourxl that thc tonbinatiotr of 365.2425 tiays per calcndar year, r'hich is a close ap1>t'oxirrratiortlo the lerrgllr
(l,l,l) occurred ill 4713 B c. or. as ilstlononlels tlow sa]', 4712 Tlris servesas of tho tropi<al year. Coutparisonrvith Equatiolt 12.11 1 rcvoalsth.lt llto Gregoliitll
year I of Scaliger'sJulian Pcriod. It Nils later acloptedas the initial eP()chfor thc calendar acctttnulatesan errol of otte daf in trbout 250{)lcars -\ltltortgh vatiotts
Julian day numbers (seeSr:ction12.7). adjustments1,othe leap-JearsysteDr)tavebeelr proposed,rtonc has bctrlt institut(xl.
Within cach year, diltes are spcr:ified accrrrclirrgto tllo count of tltr.vsfrom tlltr
'l'he
beginning of the month. order of nronths attd nuntbor of da1-spcr lttottih *r'te
CALENDAR
12,2 THEGBEGORIAN adopted fronr the Julian t alendar (scc Table 12.211 an<lScction 12 E).

The Gregorian calcrrdar toclay serves .ls an interrtiltional st.uldard for <:ivil use llr Rules
12.22 Ecclesiastical
additiou. it reelulatcsthe ccrernonialcr'<:leof the Roman Catholic and Protestall
churches.In fact. its original purposc strs ecclesiaslicaiAltliough a varictl ofotltt'r' The ecclesiastical calert<ltus of Clrlistian chur< ltes are bast'd on cy lcs of nlolirblc
calerldars are irt use today, thcl' arc reslricted t() l)itrticular rt:ligions or ctrlturcs' and immovable feasts. (lhristmas is the prirrciptr.l iuturovablc feast, rvilh its clir,lc
sct at Decenrlxrr 25. Eastcr is the prirrcipal tttovablc feast. trttrl dates ofmost otlrt'r'
novable feasts are detertttitred \\'itll rcspect to Easter. Hout:ter. the Iltorabic li'asls
12.21 RulestorCivilUse
of the Advent and Epiphrurl' seas(rls are SurrdaYs reckoncrl {iorn C-'luisl t as arttl l ltc
l-cars are counted lrom lho initial c1-rochdcfirtt:rl Jly Dion"vsius Exigults (see Stt Feast of the Epiphanv, rcspectivelv.
tiou 12.1.1),ancl are dividetl into trvo classcs:coltlltloll l'ear-sand le.rl) lears A In the Grcgorian calcttdar, the rlate ofEastot is definc<l to occul-oll the Sutr<llr'
'flris
colnnlon vear is 365 da1'sin length: a lt'trp]'ear is i366da1s. uitlt au intet calzrrl-tlar-' follou'ing thc ecclesiastical Full \Ioorr that falls on or rlcxt aftcr- \laIch 21
desigtratedFebruary 29. preceding}larch l. Lcap ycars aro rlcterminc<laccorclillg should not bt: confused rvith the popular noti(tl that Elstel is tltc first SuIl<llu
to the firllowing rule: aftcr the first Full Moou lbllowing tlte vernal (lrlinox ln l,lte filst placc, the v(rrxal
eqrrirrox docs not neccssarill' occllr on \Iarcll 21. In ad<litiorr, thc ('cclesiasti(al
Everl'1'ear that is cxactll'dir-isible br''l is a leap 1-cat.t'xr:eptfor lcar it is basctl on tabl('s lhat do lrol
Rrll Nloon is lrot the aslt'onomical Full \IooIr
that are cxactly divisible bv 100; these ccllturial r-c:rrsilre leap vears of lun:rr tltotion. As tl rr:srrlt' lltc date of alt
take into actuunt thc htll complcxit.l'
only if thoy are exacl,ly clivisible by 400.
ecclesiastic:al Full \'{oorr rnay diflcr frortr that of t}rc tru: Full Nloorr' llowever' tl()
As a result thc 1car 2000 is a leap 1'cat. rvhcrt'as 1900 ancl 2100 art'rtot lealr Gregorian st'stcIn of lezrp -vcarsand lttllar tables <loesprevcttt proglcssi!e clcpar-tllr('
years. These ruk-'sc:anlrc aPplied to tirDesprior to the Gregot'iturrefollll to creat(' of the tabulatod data froll the astrortomical pllcnomena
a proleptic Grcgorian cak:rrclar.In this case, year' 0 (1 u (r.) is consi(lercdto l)(' The ecclesiastical Rrll Nloon is tlefined as tltc fourtct:rtth clay of a t'abular lulta-

exactly divisiblc by 4, 100, :rnd 400; h<'nceit is a lca,p1'ear. tior, where cli\ 1 corresl>oncls to tltc trcclesiasticit l New N{oon. The tablcs are bastrl
The Gregorian calencl:u'is thus bascd on a crr le of 400 vt:ars. rvhit lt <rornpriscs on the Nlctorric c1'cle, in u'hich 235 rnean sylrodic months occur in ti939 688da1's
1,16097days. Sirrce 1,16097is eveuly divisible bv 7, the Grcgorian civil calendar Since nineteert Gregoriatr years is 6939.6075dirys, the dates of Nlootl phases in a
givcrr year will recur on uearly the same dates nineteen years latcr' To prevcnt
exactly repeats after,100 ycars. Diviclilrg 146097by 400 yieltls an avera,gelengtlt
F.99 EXPLANATORYSUPPLEMENT 12 / CALENDARS 583
the 0.08 day difierence between the cycles from accumulating, the tables incorpo- L=9 -2
rate adjustments to synchronize the system over longer periods of time. Additional M=3+(7+40)/44
complications arise because the tabular lunations are of 29 or 30 integral days.
D = ' 7+ 2 8 - 3 l x ( 4 1 4 )
The entire system comprises a period of 5700000 years of 2081882250days, which
is equated to 70499183 lunations. After this period, the dates of Easter repeat Thus in +2010 Easter will occur on April 4.
themselves.
The following algorithm for computing the date of Easter is based on the al-
gorithm of Oudin (1940). It is valid for any Gregorian year, L All variables are
12.23 Historyof the Gregorian
Calendar
integers and the remainders of all divisions are dropped. The final date is given by
M, the month, and D, the day of the month. The Gregorian calendar resulted from a perceived need to reform the method of
calculating dates of Easter. Under the Julian calendar the dating of Easter had
become standardized, using March 21 as the daie of the equinox and the Metonic
C=Yt1OO, cycle as the basis for calculating lunar phases. By the thirteenth century it was
N=Y-19x(Y/19), realized that the true equinox had regressed from March 21 (its supposed date
K = (.C 1"t)| 2s, at the time of the Council of Nicea, *325) to a date earlier in the month. As a
I = C - Ct 4 - G - n / 3 + 1 9 x N + 1 5 , result, Easter was drifting away from its springtime position and was losing its
relation with the Jewish Passover. Over the next four centuries, scholars debated
r=r,30x(n30),
the "correct" time for celebrating Easter and the means of regulating this time
I=I ( I / 2 8 ) x ( r - ( t t 2 8 ) x ( 2 9 t ( I + l ) ) x ( ( 2 1- I ) / 1 1 ) ) , calendrically. The Church made intermittent attempts to solve the Easter question,
l=Y+Yl4+I+2 C+C14, without reaching a consensus.
J=J 7 x ( , 1t 7 ) , By the sixteenth century the equinox had shifted by ten days, and astronomical
New Moons were occurring four days before ecclesiasticalNew Moons. At the behest
L= I - J,
of the Council of Trent, Pope Pius V introduced a new Breviary in 1568 and N{issal
M=3+(L+4O)t44, in 1570, both of which included adjustments to the lunar tables and the leap-year
D = L+28 31x(Mt4). system. Pope Gregory XIII, who succeeded Pope Pius in 1572, soon convened a
commission to consider reform of the calendar, since he considered his predecessor's
measures inadequate.
Example Compute the date of Easter in +2010. The recommendations of Pope Gregory's calendar commission were instituted
by the papal bull "Inter Gravissimus," signed on 1582 February 24. Ten days were
deleted from the calendar, so that 1582 October 4 was followed by 1582 October 15,
c=2010/100 thereby causing the vernal equinox of 1583 and subsequent years to occur about
March 21. And a new table of Nerv Moons and Full Moons was introduced for
/r' = 2010- 19x(2010t19)
determining the date of Easter.
K=(2O_1'7)t25 Subject to the logistical problems of communication and governance in the six-
I = 2 0 - 2 0I 4 - ( 2 0 0 ) / 3 + 1 9 x 1 5 r 1 5 - 309, teenth century, the new calendar was promulgated through the Roman-Catholic
world. Protestant states initially rejected the calendar, but gradually accepted it
1=309-30x(309/30)
over the coming centuries. The Eastern Orthodox churches rejected the new cal-
I = 9 - (9 t28) x (r - (9 t 28)x (29t (9 + l)) x ((21- 15)/ l1)) endar and continued to use the Julian calendar with traditional lunar tables for
=9-0x(1 0x2xl) calculating Easter. Because the purpose of the Gregorian calendar was to regulate
J = 2010+ 2Ol0I 4 + 9 + 2 - 20 +20 I 4 = 2508, the cycle of Christian holidays, its acceptance in the non-Christian world was ini-
tially not at issue. But as international communications developed, the civil rules
.l = 2508-7 x (2508/7)
of the Gregorian calendar were gradually adopted around the world.
 EXPLANATORYSUPPLEMENT 12 / CALENDABS
584 c65

Anyone seriously interested in the Gregorian calendar should study the col- Table 12.31.2
Monthsof the HebrewCalendar
lection of papers resulting from a conference sponsored by the Vatican to com-
memorate the four-hundredth anniversary of the Gregorian Reform (Coyrre el ol., 'L Tishri 30 7. Nisan 30
1983). 2. Heshvan 29* 8. lyar 29
3. Kislev 301 9. Sivan 30
4. Tevet 29 10. Tammuz 29
5. Shevat 30 11. Av 30
P3 THE HEBREWCALENDAR 6. Adar 29t 12. Elul 29

' In a cornpleleyear,
Heshvanhas 30 days.
As it exists today, the Hebrew calendar is a lunisolar calendar that is based on I In a deficientyear, Kislevhas 29 days.
calculation rather than observation. This calendar is the official calendar of Israel T In a leap year Adar I has 30 days; il is followedby Adar tt with 29 days

and is the liturgical calendar of the Jewish faith.

In principle the beginning of each month is determined by a tabular New Moorr Table 12.31.3
(molod) that is based on an adopted mean r-alueof the lunation cycle. To ensure that Terminologyof the HebrewCalendar
religious festivals occur in appropriate seasons,months are intercalated according
to the Metonic cycle, in which 235 lunations occur in nineteen years. Deficient(haser.)month: a month comprising29 days.
Full (rnale)month: a month comprising30 days.
By tradition, days of the week are designated by number, with only the seventh Ordinaryyear: a year comprising12 months,with a total of 353, 354, or 355 days.
day, Sabbath, having a specilic name. Days are reckoned from sunset to sunset, so Leap year: a year comprising13 months,with a total o1383, 384, or 385 days.
that day 1 begins at sunset on Saturday and ends at sunset on Sunday. The Sabbath Completeyear (she/emah):a year in whjch the monlhs ol Heshyanand Kisley both contain
30 days.
begins at sunset on Friday and elrds at sunset on Saturday.
Deficientyear (hasea):a year in which the months of Heshya, and Klsleyboth contain
29 days.
Regularyear (kesldrah)ia year in which Heshya, has 29 days and K/s/evhas 30 days.
12.31 Rules Halakim(singula\ helek ): "parts"of an hour; there are 1080 halakimper hour.
Molad (plu.al,moladol)t"birth"of lhe N4oon,taken to mean the lime of conjunctiontor modern
Years ate counted from the Era of Creation, or Era N{undi, which corresponds to calendricourooses.
-3760 October 7 on the Julian proleptic calendar. Each year consists of twelve or Dehiwah (plutal,dehiyof): "postponenent";a rule delaying 1 Tishtiuntil aflet the motad.
thirteen months, with months consisting of 29 or 30 days. An irrtercalary month
is introduced in years 3, 6, 8, 11, 1,1,17. and 19 in a nineteen-yearcyclc of 235
lunations. The initiat year of the calendar, A.i!,I. (Anno N{undi) 1, is ycar 1 of the
nineteen-year cycle. The months of Heshran and Kislcv vary in length to satisfl' requircrnents for
The calendar for a given year is established by determining the day of the rveek the length of the year (seeTable 12.31.1).In leap years,the 29-day morrth Adar is
of Tishri 1 (first day of Rosh Hashanah or Nen' Year's Day) and the nurnber of days designated Adar II, and is preceded by the 30-day intercalary month Adar I.
in the year. Years are classified according to the number of davs in the 1'ear (sec For calcndrical calculations,the day begins at 6 p.M., which is designated0
Table12.31.1). hours. Hours arc divided into 1080 halakim; thts one helefris 3 l/3 seconds. (Ter-
minology is explained in Table 12.31.3.)Calendricalcalculationsare referlcd to thc
meridian of Jerusalcm 2 hours 21 minutes east of Greenwich.
Table 12.31.1 Rules for corrstructing the Hebrew calendar are given in the sections that follo*.
Classificationof Years in the Hebrew Calendar Cohen (1981),Resnikoff(19,13),and Spier (1952)provide reliableguidesto the rulcs
of calculation.
Deticient Regular Complete

Ordinaryyear 353 354 12.311 Determining Tishri I The calendar year begins with the first day of Rosh
Leapyear 383 384 Hashanah (Tishri 1). This is cletermined by the day of the Tishri molad and. lhe
four rules of postponements (dehiygot). The d,ehiyyotcan postponc Tishri 1 until
560 EXPLANATORYSUPPLEI\,IENT 12 / CALENDARS co/

T a b l e1 2 . 3 1 1 . 1 (d) If the first molad following a leap year falls on day 2 at or after 15 hours,
LunationConstantsfor DeterminingTishri 1 589 halakim. then Tishri 1 is postponed one day to day 3.

Lunations Weeks-Days-Hours-rtarak m
E{ample 2 Find the day of A.M. 5760 Tishri 1.
1 4-1-124793 Since the initial epoch, 5759 years have passed, comprising 303 nineteen year cycles plus
sH-{H876
two years. Both of these years are ordinary years. The lunation constants given previously
54-5 2t {589
(with whole weeks eliminated) provide the weekday displacement, which are added to the
235 991-2 1H595
epochal moladl

Days_Hours_Halaklm
19-year cycles 303 x 2 16 0595 : 2 22 l,OOs
ordinaryyears 2x408 0876- 117 O672
l e a py e a r s 0x521 0589- 0 00 0000
one or two days following tb,e rnolad,,Tabular net' moons (maladol) are reckoned
epochal molad 2 '050204
from the Tishri molad of the year A.M. 1, rvhich occurred otr day 2 at 5 hours, 204 T i s h r i m o l o d ,A . \ r . 5 7 6 0 C 21 0&X
halaki.rn(i.e., 11:11:20P.M. on Sunday, -3760 October 6, Julian proleptic calen-
dar). The adopted value of the mean lunation is 29 days, 12 horrs,793 haLakim Since the molarl occurs on day 6, dehiygah (a) causes Tishri 1 to be postponed to day 7.
(29.530594 days). To avoid rounding and iruncation errors, calculations should be It may be noted tliat tlehiygah (,b) r.ould also cause a postponernent of one day, since the
molad, occtrs after 18 hours. DehiyEah (a) takes precedencc, however.
done in halakim rather than decimals of a da1',since the adopted lunation constant
is expressed exactly in halakim.
Lunation constants required in calculations are shor\''nin Tabie 12.311.1. By-
12.312 Reasons for the Dehiyyot Dehiyya.h (a) prevenis Hoshana Rabba (Tishri
subtracting off the weeks, these constants give the shift in weekdays that occurs
after each cycle. 21) from occurring on the Sabbath and prevents Yom Kippur (Tishri 10) fro r
occurring on the day before or after the Sabbath.
DehiyyaLL(b) is an artifact of the ancient practice of beginrring each month t'ith
Example 1 Find the weekday and time of the Tishri molad of e.lt. 2.
the sighting of the lunar crescent. It is assumed that if the m,oLad(i.e., the mean
Since it is an ordinary year, A.M. t has twelve lunations. Therefore, the shift in weekdays
conjunction) occurs after noon, the lunar crescent cannot be sighted until after
for twelve lunations can be added to the epochal molad:
6 r.u., which will then be on the following day. (For further information about
Days Houts-Halakim visibility of the lunar crescent.seeSection 12.411.)
epocial moLad 2 05 0204 Dehiyyah (c) prevents an ordinary year from exceeding 355 days. If the Tishri
+ 12 lunations 4 01J 0876 molad of an ordinary year occurs on Tuesday at or after 3:11:20 A.Nl., the next
f i . h r i m o l o d ,A . M . 2 6 I.1 0000 Tishri molarl will occur at oi after noon on Saturday. According to dehiyyah (b),
Tishri 1 of the next year must be postponcd to Sunday, which Ly dehiyaoh (a)
day 6 at 14 hours 000 ialalim corresponds to Thursday at 8 A.N{.
occasions a further postponement to N{onday. This results in an ordirrary year of
356 days. Postponing Tishri 1 from Tuesday to Thursday produces a year of 354
'lhe
d.eh'iyyotare as follows: davs.
Dehiyyah (d) prevents a leap year from falling short of 383 days. If the Tishri
(a) If the "Itshi molad. falls on day 1, 4, or 6, then Tishri 1 is postponed one rnolod following a lcap year is on Nfonday, at or after 9:32 43 ll3 A.M., the previ
d.y. ous Tislrri m,olo,d(t.hirteen months earlier) occurrcd on Tuesday at or after noon.
(b) If the Tishri molad occurs at or after 18 hours (i.e., noon), then Tishri 1
Therefore, by dehiygot (b) and (a), Tishri 1 beginning the leap ycar was postpon"d
is postponed one day. If this causes Tishri 1 to fall on day 1, 4, or 6, then
to Thursday. To prevent a leap year of 382 days, dehiyllah (d) postpones by one
Tishri 1 is postponed an additional day to satisfy dehiyyah (a).
(c) If the Tishri molad of an ordinary year (i.e., of twelve months) falls on day day the beginning of the ordinary year.
3 at or after t hours, 204 ho,Io,hirn,then Tishri 1 is postponed two days to A thorough discussion of both the functional and religious aspects ol the de-
day 5, thereby satisfying dehigyah (a). Aig3Tofis provided by Cohen (1981).
560 EXPLANATORYSUPPLEMENT 12/ CALENDABS 5g9
12.313 Deterrnining the Length ofthe Year An ordinary Jear consists of 50 \'ccks ofvegetation and aniural lifc. Although eight-1't'irr.
nineteen-vear.an(l lolgcr-period
plus 3, 4, or 5 da1s. The number of cxcessda1'sidentifrcs the 1-earas beirrg dcficient. intcrcal.tiou c1'clesrna'ha'c beeninstituted at 'arious times prior to Hillel II. thcre
regular, or complcte, respectivel;r. A leap 1'ear consists of 5.1*'eeks plus 5, 6. or 7 is little evidencethat thel'rvere ernplo5red
coDsistentlyorer long tirrrospnns.
davs, which again;:rre designated deficicnt, regular, or coruplctc, respectively. The For inforntation on the sevcn-dayrveek,seeSection 12.12.
length of a ycar can therefore bc dctcrmined by conqrnring the weekday of Tishri
1 with that of the ncxt Tishri 1.
First consider an ordinary vear. The t'eekday shift after tweh'e lunatious is 12,4 THE ISLAMIC CALENDAB
04 08-876. For exanple if a Tishri rrtolatlol an ordittarl'tear occurs on (la\' 2 at
0 hours 0 halakim (6 P.\I. on \Ionda) ). the next Tishri molad,v;ill occur on dar'
The Islamic calendaris a purely Iunar calendar in o.hich months cor.rcspondto the
6 at 8 hours 876 ltahkim. The first Tishri molad <loes lrot lequire applicntion of
lunar phase cycle. As a result, the cycle of twelve lunar months rcgresscsthrough
tlrc d.ehi.ygot,so Tishri 1 occuts on day 2. Because of dchi4yah (a), the follou'irrg
the seasorrsover a period of about 33 years. F<rrreligiouspurposcs.Mrrslims begirr
Tishri I is delayod by one day to day 7, five weekdays aflcr the previous Tishri 1.
the months with the first visibilitv of the lunar crescentalter c.ujrrrctiorr. For civil
Since this charactcrizes a complete vear, the months of l{eshlan and Kislcr, both
purposes a tabulated cale:rdar that approxiuratcs the lunar pltasc cvdc is olten
contain 30 da1's.
used.
The weekdal' shift after tllirteen lunations is 05 21 589. If the Tishri rnolcd
The scven-dayrveek is observedrvith cach dal.beginning at slllset. \\ cekclal,s
of a leap year occurred on day 4 at 20 horrrs 500 halakirn, the next Tishri rnolod
are specificdby number. u'ith day 1 beginning at sunset on Salru.rl:r1-andending
will occur on da1.3 lt 18 hours t halnftim.Becauseol delti;gyot(b), Tishri 1 of the
at sunset on Sunday. Day 5, rvhich is called Jrrnr'a.is the day ftrr <rrtrgregationa,l
leap year is postponcd two days to clay 6. Because of dehi.ygah(c), Tishli 1 of the
prayers.Unlike the Sabbath days of the Christiaus and Jews, hourrvcr,.lurn"ais not
follou'ing year is postponed t$'o davs to d.iy 5. This six-dal difference characterizes
a day of rest. Jrrrn'a begins at suuset on Thursdav and ends at suusct on Fridar-.
a regular ]'ear. so that Heshvan has 29 dtr5'sand Kislev ltn^s30 days.

'12.32 Historyof the HebrewCalendar 12.4'l Rules

The codified llebrew calendar as we know it toda]' is gcrrerally considercd to <late Years ol twelve lunar months ir,re reckoned frorn the Era of the Hijra, commem-
from -n.r't.4119 (+359), though the exact date is uncu'tain. At that time thc pa- orating the nigration of the Prophet and his lbllowers fron N,Ic< <ra lo Nledina.
triarch Hillel II, brcaking $'ith tradition, disseminatcd rules for calculalirrg tlte This epoch, 1 ,1.n. (Anno Higcrae) \'IuharraIr. l, is generallytaken b]'astr.onomers
calendar. Prior to that time the calendar rvas regardcd as a secret scieuccof the (Neugebauer.1975) to bc Thurs<la1', 'l'his
*622 Jul1 15 (Julian calendar). is callcd
religious authorities. The exact details of Hillel's calendar har.enot cornc clonn tcr the astronornir:alHijra epoch. Chronological tables (e.g., \Ia1r zur<lSpuler. 1961:
us, but it is gerrerally considered to irrclude rules for intercalation ol'er ninctccn-1-ear Freeman-Grern'ille, 1963) gencrally use Friday. .lrrly 16, u.hich is clcsiguatedI he civil
cycles.Up to the tenth century A.D., however,there was disagreemerrttrbout the epoch. Irr both casesthc Islarni<:day begins at slursetof the prcvioris rlay.
proper years lbr iutercalation and the initial epoch for rcckorring years. For rcligious purposesr ca.chrnonth begins irr principle with th(] Iirst sighting
Information on calendrical practices prior to Hillel is fragmentary turd oftcn of the lunar crescentaftcr thc Ncw Nloon. This is particularly irrrportant for es-
contradictory. The carliest evidence indicates a calendar based ol obsen'ations of tablishing thc beginning aucl <:tr<lof Rauradan. Bccause of un({rrtairltiesdue to
\'loon phases. Sincc the Bible meutions seasonal festivals. there must hzrye becn weather, horvever. a neu' nronth nrav be declared thirty days aftcr t lrt' beginning of
irrtercalation. Therc u.as Iikely an evolution of conflictilg calendrical practices. the preccding month. Although \arious predictivc l)rocedures havc br:eu used for
The Babylonian cxile, in the first half of the sixth celtury B-C., greatly inflLl- determining first visibility (scc Section 12.,111),they have always ha<lan equivocal
enced the Hebrew calendar. This is visible today in tlre narnes of the morrths. The status. In practice, there is disagreement amoug countries, religious leaders, and
Babylonian influcnce may also havc lcd to the praciic{r of intercalating leap rrrontlts. scientists about whether to rcly on observations, which are subject to euol, or to
During the period of the Sanhedrin, a committcc of the Sanhedril] tnet to use calculations, which may be based on poor ruodels.
cvaluate reports of sightings of the luuar crescent. If siglrtirrgs r.r'erenot possible, Chronologists employ a thirty-year cyclic calerrdar in studying Islamic history.
the new month n'as begun 30 days after the beginning of the previous rlronth. In this tabular calendar, thcre are eleven leap years in the thirty-year cycle. Odd-
Decisions on intercalation were influenced, if not dcterrnined entirely, by the state uumbered months have thirty days and even-numbered months have twenty-nine
 <on EXPLANATOBYSUPPLE[,4ENT 12 / CALENDARS 591
Table12.41.1 12.42 Historyof the lslamicCalendar
lvlonthsof TabularlslamicCalendar
The {orm of the Islamic calcndar, as a lunar calcndar without intercalation, l'as laid
1. MuharramS 30 7. Bajab$ 30 down bv thc Prophet in the Qur'an (Sura IX, verse 36 37) ald in his serrnon at thc
2. Safar 29 8. Sha'ban 29
3. Rabi'aI 30 Farewell Pilgrimagc. This was a departure from the lunisolar calerrdar commonly
9. RamadanT 30
4. Rabi'all 29 10. Shawwal 29 used in the Arab world, in which months were bascd or first sightings of the lular
5. JumadaI 30 11. Dhu al-Q'adahl 30 crescent, but an intercalary month was added as deemed eccssary.
6. Jumadall 29 12. Dhu al-Hijjah! 29* Caliph'Umar I is credited with establishingthe Hijra Era in .q.u. 17. It is uot
" In a leap year, Dhu al-Hijjahhas 30 days. known how the initial date was determined. However, calculations show that t|e
! Holy monlhs. astronornical New N'Ioon (i.e., conjunction) occuued on f622 July 14 at 04,14UT
r Monlh of fasting.
(assuming A? = L 0 hour), so that sighting of the crescent nost likely occrrrred on
the evening of July 16.

1 2 . 5 T H E I N D I A NC A L E N D A R

days, rvith a thirtieth day added to the twelfth rnonth. Dhu al-Hijjah (see Ta- As a result of a calendar reform in A.D. 1957, the National Calendal ol India
ble 12.41.1).Years 2, 5, 7. 10, 13, 16, 18, 21, 2,1,26. and 29 of the cycle arc des- is a formalized lunisolar calerrdar in rvhich lcap years coincide with those of the
ignated leap years. This type of calendar is also used as a civil calenclar in some Gregoriancalendar(CalendarReform Committee. 1957).Horvever.the initial epoch
\{uslim countries, though other years are sometimes used as leap years. The mean is thc Saka Era, a traditional epoch of Indian chronologl'. \'Ionths are named after
length of the month of the thifty-year tabular calendar is about 2.9 scconds less the traditional Indian morrths and are offset fron the beginning of Gregorian rru)rrths
r l r a nI l e ' ) n o d i cl , c r i o do f t h e \ l n o n . ( s e eT a b l e 1 2 . 5 1 . 1 ) .
In addition to establishing a civil calendar, the Calendar Rcforrn Committee
set guidelines for religious calendars, which require calculations of the motions of
12.411 Visibility ofthe Crescent Moon Under optimal conclitiolrs, the crescent the Sun and NIoon. Tabulations of the religious holida.vsare prepared by thc India
Moon has been sighted about 15..1hours alter the astronomical Neu' Moon (i.c.. Meteorological Dcpartment and publishcd annuallv in Tlrc Indian Astronomical
conjunctiorr) (Schaefer, 1988). Usually, horver.er,it is not seen rlntil it is more than Ephemeri,s.
twenty-four hours old. Babyloniau astronomers were the lirst to devclop methods Despite the attcmpt to establish a unified calendar for all of India, manl' local
for calculating first visibility. though no surviving tables are cxplicitly concerncd variations exist. The Gregorian calendar continues in usc for administrative pur
with this (Neugebauer,1975: 533 540). The earliest known visibility tables are by posesJand holidays are still determined according to regional, religious, and ethnic
al Khwarizmi, a nineth-centuryastronomerof Baghdad (King. 1987).These tablr:s, traditions (Chattcrjee, 1987).
and n.tany subsequent tablcs, rvcrc based on the Indian critcrion that the N4oonu.ill
be visible if the local hour arrglc of the N{oon at sunset is equal to or less than
12.51 RulesforCivilUse
78". With the developmcnt ol Islamic astronomyl more complex criteria were also
developed by N{uslim astrononers. Years are counted from the Saka Era; 1 Saka is considercd to begin with the vernal
Modern models for prcdicting first visibility incorporate ceiestial mechanics, equinox of l.l. 79. The reformed Indian calendar began with Saka Era 1879, Caitra
spherical astrorromv, selenology, atmospheric physics, and ophthaLnology. Bruin 1, which corresporlds to A.D. 1957 N{arch 22. Normal )'eals have 365 days: leap
(1977) was the Iirst to prepare such a model. Ilyas (1984), rccognizing that the years have 366. In a leap yearl arl intercalaly day is added to the end of Caitra To
Islamic calendar is used around the world. introduced the concept of an "Inter- determine leap years, first add 78 to the Saka year. If this sum is evenly divisible by
national Lunar Datcline," r'est of which thc N{oon should be visible under good 4, the year is a leap ycar, rrnlessthe sum is a multiple of 100. In thc latter case, the
observing conditions- Further theoretical work has been done by Schaefer (1988). year is not a leap year unless thc suln is also a rnultiple of 400. Table 12.51.1 gives
Extensive observing prograrns have been orgalized by Doggett, Seidelmann, and the sequence of months ald their correlation with the months of the Gregorian
Schaefer(1988) and Doggett and Schaefer(1989). calendar.
 ^ot EXPLANATORYSUPPLEMENT 12 / CALENDABS 593
Table 12.51.1 Table12.52.1
Monthsof the IndianCivilCalendar Solarlvlonthsof the IndianReligiousCalendar

Daysper Month Correlationof Indianand cregorian Months Sun'sLongitude Approximate Approximate

at Beginning Duration creqorian
1. Caitra 30' Cailra 1 March22' of Month of Month Date
2. Vaisakha Vaisakha1 April2l
3. Jyaistha 3'1 Jyaistha l May22 q
4. Asadha Asadha 1 June22 1. Vaisakha 23 15 30.9 A p r .1 3
5. Sravana Sravana 1 July23 2. Jyestha 53 15 31.3 l,4ay14
6. Bhadra Bhadra 1 August23 3. Asadha 83 15 31.5 J u n e1 4
7. Asvina 30 Asvina 1 Septembef 23 4. Sravana 113 15 31.4 J u l y1 6
8. Kartika 30 Kartika1 October23 5. Bhadrapada 143 15 31.0 A u g .1 6
9. Agrahayana 30 Agrahayana1 November 22 6. Asvina 173 15 30.5 Sept 16
10. Pausa 30 Pausa 1 December 22 7. Kartika 203 1s 30.0 Oct. 17
11. Magha 30 l\4agha1 January21 8. l\4argasirsa 233 15 29.6 Nov.16
12. Phalguna 30 Phalguna1 February20 9. Pausa 263 i5 29.4 D e c .1 5
10. Magha 293 15 29.5 Jan.14
"ln a leap year, Caitrahas 31 days and Cailra 1 coincideswiih [,4arch21. '15
11. Phalguna 323 29.9 Feb.12
12. Caitra 30.3 Mar.14

A seven-day week, exactl.t, correlated with the \i'estero wcck, is kept. Days run
from midnight to midnight.

Lunar molths are uleasured from one Nerv N{oon to the next (althorrglL sorne
12.52 Principlesof the Religious Calendar
groups reckon from thc Full Nloon). Each lunar month is given the name of the
Religious holidays are determined by a Iunisolar calcnda,r that is based on calcula- solar month in q'hich thc lunar month begins. Because rnost lunations are sholtel
tions of the actual positions of the Sun and Nloon. \4ost holidays occur on specified than a solar molth. lhere is occasionaliy a solar month in ri'hich trvo I'lelv Nloons
lunar dates (lzlllis). as is explained later; a feu'occur on specified solar dates. Thc occur. In this case, both lunar months bear the sarne name. but the first rrxnth
calendrical methods presented here are those recommended by the Calendar Re- is described u'ith the prefix adhiko.. ctt intercalary. Such a 1.ear has thirl,t:en lurar
form Committee (1957). They serl.e as the basis for the calendar published in Ti.r: months. ,4dAi,{:amonths occur every trvo or three years follorving patterus describcd
Indian Astronontical Ephemeris. However, many local calendar makers contimre to by the N4etoniccycle or more complex lunar phase cvclcs-
use traditional astronomical concepts and formulas, some of *'hich date back 1500 More rarell', a year n'ill occur in which a short solar rnouth rvill pass u'ithout
years. having a New N4oon. In that case, the name of the solar niorrth does llot occur in
The Calendar Reform Comrnittee attcmpted to reconcilc traditional calenchi the calendar for that yoar. Such a decayed (Asoya) month can occur only iu the
cal practices with modern astronornical concepts. According to theil proposals, months near the Eartlt's pcrihclion passage.In compensation,a mouth in the firsl
precession is accounted for and calculations of solar and lunar positiorr are baseci half of the year ri'ill havc hzrd two Nerv \4oons, so the vear rvill still havc trvelvc
on accurate modern methods. AII astronomical calculations are performed rvith lunar months. I(salya rnorrths arc scparatcd by as ferv as nineteen ]rcars aDd as many
respect to a Central Statiou at longitude 82"30' East, latitude 23'11' North. For as 141 years.
religiorrs purposes solar days are reckoned frorn surrrisc lo suririse. Lunations are divided into 30 liliris. or lunar days. Ei,rch lzlh.zis defined by tIrc
A solar rnonth is defirrcd as the interval required lbr thc Sun's apparent longi time required for the longitrrdeof the N{oorrto incrcaseby 12" or.cr the longitude oI
tude to increase by 30", corresponding to the passageof the Sun through a zodiacal the Sun. Thus the length of a lilhi may vary frorn about 20 hours to nearly 27 hours.
stgn (rasi). The initial month of the year. Vaisakha, begins when the truc longitude During tlx: rvaxing phases, tithis are counted from l to 15 lvith the clesignation
ofthe Sun is 23"15' (seeTable 12.5?.1).Becausethe Earth's orbit is elliptical, the Sukkt. f ithis for the waning phases are designated Krsno, and are again counted
lengths of the months \'ary lrom 29.2 to 3I.2 days. The short months all occur in from 1 to 15. Each day is assigned the number ol the tithi in cflcct a.t surrrise.
the second hall of the year around the tiure of the Earth's periheliorr passage. Occasirrrrallya shoft tithi will begin after sunrise and be cornpleted before the next
594 EXPLANATORYSUPPLEMENT 12 / CALENDARS cY5

sunrise.Similarll a loug lilh i lnav span tl'o sunrises.In thc forrucr case.a number Table 12.6'1.1
is omitted from the day coutrt. In thc latter. a da1' nunrbr:r is carried over to a Chinese SexagenaryCycle of Days and Years
second daY.
CelestialStems EarthlyBranches

jia
12.53 Historyof the IndianCalendar 1. 1. zi rat
2. yl 2. chou ox
3. bing 3. yin tiger
The history of calendars irr In<lia is a remarkably cornplcx subject o$'ing to the
4. dang 4. mao hare
continuity of Indian civilizatiorr and to the diversitl' of cultural influences.In the 5. wu 5. chen dragon
mid-1950s, n'hen the Caleltriar Reform Committee rnade its sttrvcl', thcre u'ere about 6. ji 6. si snake
30 calendarsin use for sctting religious festivalsfor HindLrs,Bucldhists.and Jain- 7. geng 7. wu horse
8. xin 8. wei sheep
ists. Some of these l.ere also usr:d for civil dating. Thcsc <'alendars rvere based 9. ren 9. shen monkey
on common principles, thorrgh they'had local characteristicsdotcrntined b1' long- 10. gui 10. you fowl
cstablished customs au<l thc astronomical practices of local <:abrrrlar makers. In 11 . xu dog
12. hai piS
addition, Muslims in Irtditr ttseclthe Islamic calendar, ancl the Irrrliaut govcrnmpnt
used the Gregorian calendalr for administrative purposcs.
Early allusions to a lunisolar ca.lcndarwith intercalat,trl rtrorttlts zrrefound in the
hymns from the Rig Veda, clntirrg liorrr thc second rrrillcnrrirrrrtIl.c. Litelature fion Yeat Names
1300 B.c. to A.D. 300. proviclcsirrforrrrationof a morc spccifi<:ttature. A five-1'ear
1. jia-zi 16. ji-mao 31. jia-wu 46. jiyou
lunisolar calendar coordirraieclsolar 1'earsrvith s.u-nodic ancl sirlereallutrar months.
2. yi-chou 17. geng-chen 32. yi-wei 47. geng-xu
Indian astronoml, rrndcr*cnt a general reform in the first fr:w celturies A.ll.. 3. bing-yin 18. xin-si 33. bing-shen 48. xin-hai
zrsadrances in Babvkrnian artcl Greek astronomv became knorvrr.\ew astrononl 4. ding-mao 19. ren-wu 34. ding-yoo 49. rcn-zj
i<al constants and nrodels for the rnotion ofthe \Ioon and Srur utre adapted to 5. wu-chen 20. gLri-wei 35. wu-xu 50. goichoo
6. ji-si 21. jia-shen 36. jihai 51. jia-yin
traditional calendric pra<:ticcs.This u'as convel'edin astronornicaltrcatises of this 7. geng-wu 22. yiyou 37. geng-zi ylmao
52.
period knorvn as Siddhanl.as. rnarry of u'hich have not survivcrl. The Suryo .9id- 8. xin-wei 23. bing-xu 38. xin-chou 53. bing-chen
d.hanta,which origirratcd iu thc fourth ceniury but ivas rrp<latc<lovcr t.hc folloiving 9. ren-shen 24. ding-hai 39. ren-yin 54. ding-si
10. gui-you 25. wu-zi 40. gui-mao 55- wu-wu
centuties, influenccd Inditu <:aloudricsup to and cvcrr aftr:r tlrt: r'zrlcndar rcform of jia-xu ji-chou jia-chen ji"wei
11. 26. 41. 56.
A . D .1 9 5 7 . 12. yi-hai 27. geng-yin 42. yi-si 57. geng-shen
Pingrce (1978) providcs a survey of the developrnentof rlathcrlatical astron- 13. bing-zi 24. xin-mao 43. bing-wu 58. xin-you
omy in India. Although he rloos not deal explicitly n'ith calcrrrlric:s. 14. ding-chou 29. ren-chen 44. ding-wei 59. ren-xu
this matcrial is
15. wu-yin 30. gui-si 45. wu-shen 60. gui-hai
necessart'for a full uncicrst;ur<tingof the historv of Inclia'scakrrtdars.

1 2 , 6 T H E C H I N E S EC A L E N D A R 12.6.1Rules
The Chinesccalendar is a lunisolar calendar based on carlcrrlal ions of thc positions There is no specificinitial cpoch for counting 1ears.Irr historictrllccords. dates $cre
of the Sun and Nloon. Nlonths of 29 or 30 days begin on davs of astronomical Ncrv sper:ified by counts of (lays iul(l yctus in sexagenary cy( los and by counts of t'eats
Moons. with an intercalalv urorrlh bein5qadded everv two or tlrrce.r'ears.Since the fron a succcssiorr of cras cstablishcclby reigning rnonalchs (scc Scction 12.62).
<alendaris based on the tnrc positions of the Sun and N,Ioorr,1l:<:zrccura,cy of the Thc sixty-ycal cyt:k: <orrsistsof a set of ycar namcs that arr: r:reated by pairing a.
r:alcndar clependsou tllrc a( ( luacy of the astronomical theorics zurd ca,l(:ulations. name from a list of tcrr (lelcstial Stems with a nanrc frorn a list of twelve Terrestrial
Although the Grt:goriarr<zrl<rr<lar is used in tlrc Peoplcs' Republic of China Branches. follon'ing thc ordcr spccificd in Table 12.61.1.TIrc Cclcstial Stems are
for adrninistra,tive purposcs, thc traditional Chinese calentlar is rrst:d for setting spccified by Chincsc charn ters that have no English trarrslatiorlj the Terrestrial
traditional festirals and for tirning aglicultural activitics in thc countryside. The Branches are named after twch,e animals. After six repctitiorrs of the set of stems
Chinesecalendar is also usetl bv Chinesecommunities arounclthe lvorld. and five repetitions of thc branches, a complete cycle of pairs is completed and
I
I
I
596 FXPIANATORY
SUPPLEMENT 12 / CALENDARS 597
a new cycle begirrs.Thc irritial ycar (jia zi) of thc currcnt cyclc bcgan on 1984 T a b l e1 2 . 6 1 . 2
FebruarY 2. ChineseSolarTerms
Days arc measured from midnight to midniglrt. The first dav of a calendar
Term* Sun's Approximate Approximate
month is the day on which the astronomical New \'Ioon (i.e., conjunction) is cal- Longitude GregorianDate Duration
culated to occur. Since the al'erage interlal betrveen successiveNew N4oonsis ap-
proximately 29.53 days, rnorths are 29 or 30 days long. I\,lonths are specified b1. S-1 Lichun Beginningof Spring 315 Feb.4
P-l Yushui Rain Water Feb.19 29.8
number from 1to 12. When an intercalary month is addcd, it bcars the rrunber S-2 Jingzhe Waking of Insects 345 Mar 6
of the previous month, but is designated as intercalary. An ordinary year of twelve P-2 Chunten Spring Equinox 0 Mat.21 30.2
months is 353. 354, or 355 days in length; a leap year of thirteen rnonths is 383, S-3 Qingning Pure Brightness Apr.5
P 3 Guyu Grain Rain 30 Apt.20 30.7
384, or 385 days lortg.
S4 Lixia Beginningof Summer May 6
The conditions for adding an intercalaly month trrc dctcrrnined by thc occur- P4 Xiaoman GrainFull 60 May 21 3 1. 2
rence of the Nerv N'Ioon with respect to divisions of the tropical ycar. The tropical S-5 Mangzhong Grainin Ear June 6
year is divided into 24 solar terms, irl 15" segments of solar longitude. These di- P-5 Xiazhi SummerSolstice 90
5-6 Xiaoshu SlightHeat 105 July 7
visions are paired into trvelve Sectional Terms (-/ieqz) and twelve Principal Terms GreatHeat 120 July23 31.4
(Zhongqi), as showrr in Table 12,61.2.These terrns are numbered and assignerl S-7 Liqiu Beginningof Autumn 135 Aug.8
names that are seasonal or rneteorological in nature. For convenience in the ex P-7 Chushu Limit of Heal 150 Aug.23 31.1
S-B Bailu While Dew 165 Sept.8
amples that follor', thc Scctional and Principal Tetms are denoted by' S and P.
P-8 Qiufen Autumnal EqLrinox 180 Sept. 23 30.7
respectively, followed by the number. Becausc ol the ellipticity of the Earth's orbit. S-9 Hanlu ColdDew 195 Oct.8
the interval between solar terms varics with thc seasons. P-S Shuangjiang Descentol Frost 210 OcI.24 30.1
Reference works give a variety of rules for establishing New Year's Day and S-10 Lidong Beginningol Winter 225 Nov.B
P-10 Xiaoxue SlighiSnow 240 Nov.22 29.7
for intercalation in the lunisolar calendar. Sincc the caleudar was originally based S-11 Daxue Great Snow 255 Dec.7
on the assumption that the Sun's motion n'as unilorm through the seasons, the P 11 Dongzhi Winter Solstice 270 Dec.22 29.5
published rules are frequently inadequate to handlc special cases. S-12 Xiaohan SlightCold 285 Jan.6
The follorving rules (Liu and Stephenson, in press) are currently uscd as the P-12 Dahan Great Cold 300 Jan. 20 29.5
basis for calendars prepared by the Purple Nlorrntain Obserratory (1984): 'Terms are cassiJiedas Secliona (Jeql) or Principal(Zhangqi),lollawed
by lhe numberol the lerm.

(1) The first day of the month is the day on which the New Nloon occrrrs.
(2) An ordinarv vear has twelve lunar rnonths; an intercalary year has thirteen
lunar months.
(3) The Winter Solstice (term P 11) ahvays falls iu rnonth 11.
(4) In an intercalary year, a month in which there is no Principal Term is Example 1 (typn:al r:ase):Firrd the Cregorian date of C.hineseNen'Ycal in A.D. 1991
the intercalarv morrth. It is assigned the nrrrnber of the preceding month. The Wirter Solsticc (Principal Tcrm 11) occurs on 1990 Decembcr 22, .rrrdthe Nc*'floon
with the further designation of intercalary. If tr-o rnonths of an intercalarl' whicir occurs on or prior to that date js on December 17. Hence we compilc the follorvirtg
year contain no Principal Term, only the first such month after the Winter tablc of phcnomcna:
Solstice is considered intercalarv.
(5) Calculationsare based on the meridian 120'Easi. Nerv l\'Ioon 1C90Dec 17 month ll
P 11 Dec 22 Wintcr Solstice
s12 Jan 6
The nurnber of the month usually corresponds to the number of the Principal
New N{oon 1991 Jarr 16 monlh 12
Term occurring during the month. In rare instances, however, there are months that P72 .Ian 20
have two Principal Terms, with the result that a nonirrtercalary month n'ill have S1 Feb 4
no Principal Term. As a result the numbers of the months rvill temporarilv fail to Ncw trIoon
correspond to the numbers of the Principal Terrrrs. These casescan be resolved b1' P1 Feb 19
strictly applying rules 2 and 3. The second example deals with such a case. S2 Mar 6
t N€w Moon NIar 16 month 2
(oR EXPLANATOBYSUPPLEI\,IENT 12 / CALENDARS 599
Thus Chinese New Year occurs on February 15. There are no intercalary months during Published calendrical tables are often in disagreement about the Chinese calen_
this inteNal' dar. Some of the tables are based on mean, or at least simplified, motions of the Sun
and Moon. Some are calculated for other neridians than l20o East. Some incor_
porate a rule that the eleventh, twelfth, and first molths are never followed by an

Dxample 2 (extraordinary case): Find the Gregorian date of Chinese New Year in A.D.
intercalary month. This is sometimes not stated as a rule, but as a consequenceof
1985. Both the Winter Solstice and a New Moon occur on 1984 December 22. the rapid change in the Sun's longitude whcn the Earth is near perihelion. Horvel,er.
this statement is incorrect when the motions of the Sun and Moon are accuratelv
New Moon 1984 Dec 22 month 11
calculated.
P 11 Dec 22 Winter Solstice
S 12 1985 Jan 5
P 12 Jan 20
New Moon Jan 21 month 12? 12.62 Historyof the ChineseCalendar
S 1 Feb4
P 1 Feb19
In China the calendar was a sacred document, sponsored and promulgated bv the
New \{oon Feb 20 month 12 Intercalary?
reigning monarch. For more than two millennia, a Bureau of Astronoury rnade as-
S 2 Mars
Nerv Moon Mar 21 month 1? tronomical observations, calculated astronomical events such as eclipscs, prepared
P 2 Mar21 astrological predictions, and maintained the calendar (Needham, 1959). After all,
a successlul calendar not only served practical needs, but also confirmed the con-
The first sign of difficulty is that month 11 has both Principal Terms 11 and 12. We will so[ance bet$reenHeaven and the imperial court.
also see that the month beginning on February 20 cannot be intercalary. This is discovered Analysis of surviving astronornical records inscribed on oracle bones reveals a
by finding the next \\rinter Solstice, which must occur in month 11 and by checking the
Chinese lunisolar calendar, with intercalation of lunar months, dating back to the
number of New Moons. In 1985, P 11 occurs on December 22, and the New Moon that
begins month 11 occurs on December 12. \\rorking backward, we lind New Moons in 19E5 Shang dyrrasty of the fourteenth century B.c. Various intercalation schemes were
on November 13, October 1,1,Septernber15, August 16, July 18, June 18, May 20, April developed for the early calendars, including the nineteen-year and 76-year lunar
20, March 21, February 20, and January 21. Since there are only twelve New N{oons from phase cycles that came to be known in thc West as the Metonic cycle and Callipic
month 11 in 1984 to month 11 in 1985, there is no room for an intercalary rnonth. So the cvcle.
correct sequence of months is as foliows:
From the earliest records, the beginning of the year occuued at a Ncw N{oon
New N{oon 198,1Dec 22 month 11 near the winter solstice. The choice of month for beginning the civil I'ear varied
P 11 Dec 22 Winter Solstice with time and place, however. In the late sccond century 8.c., a calendar reform
S 12 1985 Jan 5 established the practice, which continues today, of requiring the winter solstice to
P 12 Jan 20 occur in month 11. This reform also introduced the intcrcalation system in whidl
New X{oon Jan 21 month 12 dates of New Nloons are compared with the 24 solar terms (see Scction 12.61).
S 1 Feb4
However, calculations rvere based on the mean motions resulting from the cycli<:
P 1 Feb 19
New Moon Feb 20 month 1 relationships. Inequalities in the Nrloon'smotions were incorporatcd as early as the
S 2 MarS seventh century a.l. (Sivin, 1969), but the Sun's mean longitude was used lbr
Neq' Moon Mar 21 month 2 calculating the solar terms until 1644 (Liu and Stephenson, in press).
P 2 Mar21 Years were counted from a successionof eras established by reigning empcrors.
Although the accession of an emperor would mark a new era, an emperor might
Chinese New Year therefore occurs on 1985 February 20.
also declare a new era at various times within his reign. The introduction of a
new era was an attempt to reestablish a broken connection between Heaven and
In general, the first step in calculating the Chinese calendar is to check for Earth, as personified by the emperor. The break might be revealed by the death
the existence of an intercalary year. This can be done by determining the dates of of an emperor, the occurrence of a natural disaster, or the failure of astronomers
Winter Solstice and month 11 before and after the period of interest, and then by to predict a celestial event such as an eclipse. In the latter case, a new era might
counting the intervening New Moons. mark the introduction of new astronomical or calendrical models.
600 EXPLANATORY
SUPPLEMENT 12/ CALENDABS 601
Sexagenary cycles were used to count 1'ears,months. days. and fractions of a day to times bcfbre Caesar's reforrn. This provides a simple chronological systern f6r
using the set of CelestialStems and Terrestrial Branchesdescribedin Section 12.61. correlating other calendars and serves as the basis for the Julian rlal.rrumbers
Use of the sixty-day cycle is seen in the earliest astronomical records. By contrast (Section 12.7) .
the sixty-year cycle was introduced in the first century A.D. or possibly a century
earlier (Tung, 1960; Needham, 1959). Although the day count has fallen into disuse 12.81 Rules
in everyday life, it is stiil tabulated in calendars. The initial year (jia zi) of the
current year cycle began on 1984 February 2, which is the third day (bing-yin) of Years ale classified as rrornral years of 365 days and leap years of 366 diws. Lcap
the day cYcle. years occur in years that are evenly divisible by 4. For this purpose.venr.0 (or 1
Western (pre-Copernican) astronomical theories were introduced to China bv 8.c., seeSectiolr12.14)is consicleredcvcnly divisible by 4. The vcar is diviclcd into
Jesuit missionaries in the seventeenth century. Gradually, rnore llodern \VesterD tlr,elr.eforrnalized months that werc cventually adoptt-'tl for the Gregorian calcndar.
conceptsbecame known. Following the revolution ol 1911,the traditional practice
of counting years from the accessionof an emperor rvas abolished. 12.82 Historyot theJulianCalendar
Thc year -45 has bccn called the "1'ear of confusion,' br:cause in that year .Iulius
ANDJULIANDATE
12,7 JULIANDAYNUMBERS Caesar inserted 90 day's to bring thc months of the Rornan calerrclarltrck to their
traditional placc t-ith respectto thc sc:rsons.This rvasCaosar'sfirst step irr replat:ing
The system of Juliau day numbers is a continuous count of days elapsed since the
a calendatthat had gone badlv awry. Although the prc-Julian calcridarrvaslunisol:rr
beginning of the Julian period as defined by the sixteenth-century chronologist J.
in inspirat,ion. its months no longcr lblloti'ed the lrrrrar phases arrd its 1'ear tracl lost
J. Scaliger (see Section 12.14). Although Scaliger'soriginal idea was to irrtroducc
step u'ith thc c1.cleof seasons(scc llichels, 1967:Bickerman. 197-l).Follorvir.rg thc
a count of years, nineteenth-century astronomers adapted this system to create a
advice of Sosigenes,an Alexandrine astrononlcr, Caesar created a soiar calenclal
count of days. John Herschel (1849) thoroughly explained the system and proviclccl
n'ith tt'elvr: rnonths of fixerl )crrgths and a pror.isiotr Ibr an intelcalar.l day to be
a table of "Intervals in Days between the Commencement of the Julian Period, and
added e\.cr]-fourth vear. As a rcsult, the averagclcngth of the JrLlial cak:rrrlar.rr:ar
that of some other remarkable chronological and astronomical Elas."
u'as 365.25 da,]'s.This is consistcnt rvith the length of the tropical ]ear as it \r'as
Julian day 0 commenced at Greenwich noon on 4712 January 1, Julian pro
known al thc time.
leptic calendar (see Section 12.8). The Julian day nurrrber, expressed as an integer.
Folklving Caesar's death, lhc Roman calenrlrical authorities rnisappliecl1lr<'
denotes the number of complete davs elapsed sincc thr: initial epoch. The Juliau
leap 1'r:arrule, rvith the result that every tirird. ralhcr than er-ert'fourth. roar
date (JD) specifies a particular instant of a day by ending the Julian riay number
was iutcrc.rlary. Although delailccl evidence is lacking. it is gcncralll. br:lit:r,t'd that
with a decimal fraction. For exarnple,the Julian day numbcr of 1990 Jrrne25 is 24.1
Etupcror Augustus correctcdtlre situation br-ornitting iutcrcalalion liolr the Juliarr
8068, whereasthe Julian datc:rt noon is 2,148068.0. The midniglrt that begins the
years 8 through +4. After this the Julian calendar firrall,t'lrt'giui to function as
civil day is specified by subtraciirrg 0.5 from the .Julian date at rroorr.
plarrnr:tl.
A count of days (1 365) from thc bcginrrirrg of the 1'ear is sometirrrcs a useful
Tlirougir the \Iid<lle Ages thc use of the Julian calendar evoh'ecl ald accltirr:tl
tool for record-kecping. However, the dubious practice of calling this a Julian date
loca,lpcculiarities that corltiuue to srrtrrclhe unrvary historian. There w:r'e \-ariations
merely causes confusion with Julian day nurnbers.
in thc initial epoch for r:ornting ycars, the date for beginning tlrr: r'ear, trnd the
melhod of specifvingthe da1'of the rrxruth. Not onlv clid these varl' n'ith l,jrneattd
12.8 THE JULIANCALENDAR plar:c, but aiso ll'ith purpose. Different corlcutions ncrc sometirncs usetl Iirl clating
ecclcsiaslicr rccorcls.fiscal transactions,and pr:rsorill <rrrlcsl>orrtlt:nce.
The Jrrlian calendar, introduced by Julius Caesar in 45, was a solar calendar rvith Caesar dcsignated Januarl' 1 as the beginning of the Ho*'er'er. ollter corr
"vear.
months of fixed lengths. Every fourth year an intercalary day was added to mairrtain ventions florrrished at diflcrcnt tirrrcs errrd places. The most popular aiternativcs
synchrony between thc caicndar year and the tropical ycar. It served as a standard were NIarch l, Nlarch 25, and December 25. This coutinues to c:rus<: problems for
tbr European civilization rrutil thc Gregorian Reform oI +1582. historiaus, sirr<rc, for cxanrple.1.998 February 28 as recorded in a city that began
Today the principles of the Julian calendar continue to be use<lby chronologists. its year on \{arch 1, would be the same day as +999 February 28 of a city that
The Julian proleptic calendar is lbrmed by applying the rules of the Julian calendar lregan tire year on January 1.
 602 EXPLANATORY
SUPPLEMENT '12/ CALENDARS
603
Table12.82.1 performed by repeating the day VI Kalends March, i.e., inserting a day between
FlomanDatingin the JulianCalendar VI
Kalends March (February 24) and VII Kalends March (February 23).
January March Apill By the eleventh century, consecutive counting of days from the beginning of
August February May June the month came into use. Local variations continued, however, including counts of
Decembsr July September days from dates that commemorated local saints. The inauguration and spread of
Octobor Novembel
the Gregorian calendar resulted in the adoptiou of a uniform standard for recordins
1 Kalends Kalends Kalends Kalends dates.
2tv Cappelli (1930), Grotefend and Grotefend (1941), and Cheney (1945) offer guid_
3 lll t v l ance through the maze of medieval dating.
4tl ll lv
5 Nones Nones lll Nones
v l v t v l 12.9 CALENDAR
CONVERSION
ALGORITHMS
7 vtl Vll Nones vtl
VI vl v l
Conversion between calendar systems is perhaps most easily and generally accom-
't0 plished by use of Julian day numbers. For calendars based on fixed. consistent
1l Il t v l rules. calendar dates can be converted to Julian day numbers and Julian day num-
tv tl bers to calendar dates. Tables for this purpose were prepared by Harvey (1983). A
1 3 toes ldes lll ldes
generalized set of algorithms was prepared by Hatcher (1935). Parisot (1986) has
1 4 XIX XVI II XVIII
1 5 XVIII XV ldes xvtl contributed parameters for Hatcher's algorithms for a few additional calendars.
16 XVtl XIV XVII XVI The algorithms given in this section do not conform to Hatcher,s general system
'|7 XVI XIII XVI XV but were inspired by the algorithms created by Fliegel and Van Flandern (1968)
18 XV XII XV XIV for conversion between Gregorian calendar dates and Julian day numbers. Other,
19 xtv XI XIV x l
20 x l X XIII xtl possibly more efficient, formulations are possible. Integer arithmetic is exclusively
21 Xtl IX XII XI used in this algorithms; i.e., the remainders of all divisions are dropped.
22 Xl v t xl X Julian day numbers run from noon to noo!. Thus a calculated Julian day
23X VII X IX number pertains to the noon occurring in the corresponding calendar date. The
24 tX VI IX v l
25 V I vtl Dotation is:
JD = Julian day number
26 Vtl IV VII
27 Vl |||vl I : calendar year
28V V M = month
29 tV l D : day of month
30 l l
31 tl All variables are integer. Since Julian day numbers for current times consist of seven
digits, a word size of at least four bytes is required to use these algorithms.

'12.91 ConvertingDayot the Week

Days within the month were originally counted from designated division points A problem common to several of the calendars is computing the day of the week.
within the month: Kalends, Nones, and Ides. The Kalends is the first day of the The following formula gives the day of the week for the Julian day number that
month. The Ides is the thirieenth of the month, except in March, May, July, and pertains at noon. Given: JD. Compute: Day of week, 1, where 1 runs from 1 though
October, when it is the fifteenth day. The Nones is always eight days before the Ides 7, with 1 being Sunday.
(see Table 12.82.1). Dates falling between these division points axe designated by
counting inclusively backward from the upcoming division point. Intercalation was 1=JD-7x((JD+l)/7)+2 (r2.9r-r)
604 EXPLANATORYSUPPLEMENT 12 / CALENDARS 605
Exrmple 1 Compute the day of the weekof JD 2451545. the Islamic calendar(seeSection12.41).If the astronomicalepochis preferred,in
which e.ri. 1 Muharram 1 correspondsto t.D. 622 July 15, then JDQ = 1948439
I = 2451545- 7 x ((2451545+ l) t7\ +2 should be usedin the equationsbelowl if the civil epochis preferred,in which a.u.
= 2451545-7 \350220 +2 1 Muha.rram1 correspondsto *622 July 16, use JDO= 1948440.Given Y, M, D.
=,7 Compute:JD.

Therefore JD 2451545is Saturday.

JD = (11 x y + 3) / 30 + 354x y + 30 x M - (M - t) | Z + D + JDO- 385 (12.93_t)

Given: JD. Qomptte: Y, M, D.

12.92 ConvertingbelweenGregorianCalendarDateandJulianDayNumber
These algorithms by Fliegel and Van Flandern (1968) are ralid for all Gregorian L=JD-JDO+1O632
calenda.rdates corresponding to JD ) 0, i.e., dates after *4713 November 23. Given: N = (L _ L) n0631
Y, M, D. Comptte:. JD. L= L* 1O631xN+354
-
JD = (1461 x(y+4800+(M* 14)tr2))t4+(367 x(M -2- 12x ((M- 14)tr2)))t12 "/ = (10985 z) / 5316)x ((50x L) t r7719)+ (L / s67o)x ((43x L) / 15238)
L = L - ((3o- J) / |s) x ((17719x J) t 50) - (J / 16)x (15238 x t) / 43)+ 29
- (3 x ((r+ 4900+ (M - 14)/ 12)t r00))t 4 + D - 32075 (12.92-t)
M = (24x L) t7o9
Given:JD. Qomputre:
Y, M, D. D=L-(709xluttl24
f=30xN+,I-30 (r2.93-2)
a=JD+68569
N=(4xL)/t46O97
12.94 Converting
betweenlndianCivilCalendar
andJulianDayNumber
L=L-(146O97xN+3)14
1= (4000x (t+ 1)/ 1461001 These algorithms a.re for the civil calendar recommended by the Calendar Reform
L=L*(146IxI)14+31 Committee of India. Years are counted from the Saka Era. The a.lgorithms are valid
for all ralues of Y 2 1, where Y is the year reckoned from the Saka Era, and JD )
J=(8OxL)/2447
1749995. Given: Y, M, D. Compute: JD.
D=L-(2447xJ)t80
L=Jl1l JD = 365 x v + (Y + 78 - 1 / M) / 4 + 3l x M - (M + 9) / 1l - (M I 1) x (M - 7)
M=J+2-12xL - (3 x ((r+78 - | / iO I l0o + l)) | 4 + D + 1749579 \12.94-r)
Y=l00x(N-49)+I+L (12.92-2)
Given: JD. Compute:Y, M, D.

12.93 ConvertingBetweenlslamicTabularCalendarDate Z=JD+68518

andJulianDayNumber
N=(4xL)t146097
While the religious calenda.r is based on first sighting of the luna,r crescent after L=L-(146097xN+3)14
New Moon, larious tabular calenda,rshave been used. In the system most often
1= (4000x (.+ 1))/ 1461001
ll€ed by historia.us, yea.rs 2, 5, 7, 10, 13, 16, 18, 21, 24, 26, and,29 of a 30-year
cycle a,re designated leap yea,rs. The following algorithms are valid for values of L=L-(1461x1)t4+1
I > 1 and JD > JDO, where I is the year of the Hijra era aqd JDO is the Julian J = ((L - 1)t 3t) x (r - L / r85) + (L t 185)x ((t - 1s6)| 30 + 5) - L t 366
day number of the Hijra epoch. Two difierent initial epochs have been used for D = L - 3 1x " I + ( J + 2 ) / 8 ) x ( J - 5 )
 606 EXPLANATORYSUPPLEMENT 12 / CALENDARS 607
L=J/rl Chatterjee, S.K. (1987). 'Indian Calendars" in HistorA of Oriental Astronomy c. Swarup
et al., eds. Cambridge, pp. 91 95
M=J+2-12><L Cheney, C.R. (1945). A Hand,book oJ Dates Jor Stud,ents of English llistory London.
1 = 1 0 0 x ( N 4 9 ) + L + I- 7 8 ( 1) gL_')\ Cohen. J.R. (19a1). Mishnah Tiactate Rosh llosrSanah.New York.
Colgrave, B. and Mynors, R.A.B. (1969). Bed,e'sEcclesiastical Historg of the Enqlish peo_
ple Oxford.
between
12.95 Converting JulianCalendar
DateandJulianDayNumber Colson, F.H. (1926). The Weei Cambridge.
Coyne, G.V., Hoskin, M.A., and Pedersen, O., eds. (1983). Gregori,an ReJorm of the Cat-
These algorithms are valid for all values of Y > 4712, i.e., for all dates with JD > endar Vatican City.
0. The formula lbr computing JD from I, M, D was constructed by Fliegel (1990) as Doggett, L.E. and B.E. Schaefer (1989). "Results ofthe July Moonwatch" Sky €4 Telescope
an entry in "The Great Julian Day Contest," held at the Jet Propulsion Laboratory 77, 373 375.
in 1970.Given: I, M, D. Compute: JD. Doggett, L.E., Seidelmann,P.K., and Schaefer,B.E. (1988). "Moonwatch Julv 1.1,1988"
Sky E TeLescope78, 31 35.
Encycloped,ia of Religion and Ethics (1910).
J D = 3 6 ' 7x v - ( 7 x ( v + 5 0 0 1+ ( M - 9 ) t ' 7 ) ) t 4 + ( 2 ' 7 5 x M ) 1 9 + D + l ' 7 2 9 ' 1 7 7
Fliegel, H.F. and Van Flandern, T.C. (1968). "A Machine Algorithm for Processing Cal-
endar Dates" Communications ol the Association of Computing Machines ll, 6sT.
Given: JD. Compute: Y. M. D. Fliegel, H.F. (1990). Personal communication.
Fotheringham, J.K. (193,1). "The Calendar" in The Nautical Almanac 1935 London.
J=JD+1402 Fraser, J.T. (1987). Time; the Familiar Slrazger Amherst.
Freeman-Grenville, G.S.P. (1963). The Muslim and, Christian CaleniLars London.
K=(J-t)/1461
Gandz, S. (1949). "The Origin of the Planetary Week or the Planetary Week in Hebrew
L=J 146lxK Literature" Proceed.ingsoJ the American Acad,em,gJor Jeuish Resea|ch 18, 213-2,54.
Gingerich, O. (1983). "The Civil Reception of the Gregorian Calendar" n Gregorian Re-
N=(L- t)1365 Lt1461
form oJ the Calenilar Coyne, G.V., Hoskin, M.A., and Pedersen, O., eds., Vatican
I=L 365xN+30 City, pp. 265 279.
J=(8OxDt2447 Ginzel, F.K. (1906, 1911). Hanrlbuch d,er Mathematischen urul, Technischen Chronologie
Leipzig.
D=I-(244'7xJ)/8O Grotefend, H. and Grotefend, O. (1941). Taschenbuch d,er Zeitrechnung iLes d,eutschen
I=Jt11 Mittelalters und, d,er Nelrzeit Hanlro,rer.
Harvey, O.L. (L9a3). Calend,ar Conuersions by Way of the Julian Day Nzmler Philadel-
M=J+2-12x1 . phia.
Y=4xK+N+l-4716 (r2.9s-t) Hatcher, D.A. (1985). "Generalized Equations for Julian Day Numbers and Calendar
Dates" Quarterlll Journal of the Rogal Astronomical Society 26, 151'155.
Herschel, J.F.W. (1a19). Outlines oJ Astronom! pp.633 637.
Hoskin, M. (1983). "The Reception of the Calendar by Other Churches" i,n Gregorzan
12.10 REFERENCES ReJonn of the Calendar Coyne, G.V., Hoskin, M.A., and Pedersen, O., eds., Vatican
REFERENCES City, pp. 255 264.
Ilyas, M. (1984). A Modern Cuid.e to Astronomical Calculations of Islamic Calend.ar,
Aveni, A.F. (1989). Empires of ?irne New york. Ti.mes aniL Qibla Kuala Lumpur.
Bates, R.S. (1952). "cive Us Back Our Fortnight', Sky and Telescope11, 267-268. King, D.A. (1987). "Some Early Islamic Tables for Determining Lunar Crescent Visibility"
Bickerman, E.J. (1974). Chronology of the Ancient Wortd, (Ithaca). From DeJerent to Equant: A Volume of Studies in the Historg oJ Science in the Ancient
Bruin, F. (1977). "The First Visibility of the Lunar Crescent,' Vistas in Astronomy 21, and Medi,eoal Near East in Honor of E.S. Kenned,g KirLg, D.A. and Saliba, G., eds.
331 358. Annals oJ the New York Acad,emg of Sci,ences,500, New York.
Calendar Reform Committee (1955). Repor.t of the Calendar Reform Committee (New Laskar, J. (1986). "Secular Terms of Classical Planetary Theories Using the Results of
Delhi). General Relativity" Astron. Astrophys. 157, 59-70.
Cappelli, A. (1930). Cronologia, Cronograf,a e Calendario perpetuo Milan. Liu Baolin and Stephenson, F.R. "The Chinese Calendar and Its Operational Rules" In
Cassini, J. (1740). Tables astronomiques d.z Soteil, de la Lune..., Paris. press.
Chapront-Touz6, M. and Chapront, J. (1988). "ELP 2000-85: a Semi-Analytical Lunar MacNeill, M. (1,982j. The Festhtal of Lughnasa Dttblin.
Ephemeris Adequate for Historical Times" Astron. Astrophgs. lg1,342 352. Mayr, J. and Spuler, B. (1961). WiistenJeld,-Maler'sche Verglei.chungs-Tabellen,Wiesbaden.
608 EXPLANATORY
SUPPLEMENT

Michels,A.K. (1967). The Calendar oJ the Roman RepublicWesLpott.

Neugebauer,O. (1975).A History of Ancient MathematicalAstronomyPantIII, New York.
Parise,F. (1982). The Book of Calend'orsNew York.
pa,risot,J.P. (1986). "Additif to the Paper of D.A. Hatcher: 'GeneralizedEquationsfor Ju-
lian Day Numbersand Ca.lenda.r Dates"' Quarterly Journal o! the Royal Astronomical
CHAPTER
13
Society27, 206-507.
Pingree,D. (1978), "History of Mathematical Astronomy in India" Dietionary ol Scientific
BiographyXY, New York, pp. 533 633. Historical
lnformation
Purple Mountain Observatory (198a). lfhe Newly Compiled,Perpetual ChineseCalend,ar
(1840-2050)(Popula.rSciencePublishing House,Beijing).
oudin, J.-M. (1940). "Etude sul la Date de PrA.ques" Boll- Astronomique(2) 12,39I-4I0. compiled
by J. Weeks
Resnikof, L.A. (1943). "Jewish Calenda.rCalculations" Scri,ptaMathernatica9, 191 195,
274-277.
Scaliger,J.J. (1583). ,e emenilationetemporumParis.
Schaefer,B.E. (1988). "Visibility of the Luna.r Crescent" Q. Jour. R.A.S. 29, 5ll 523.
Siviu, N. (1969). Cosmos and, Cornputdtiotuin Early ChineseMathematical Astronomy
Leiden.
Spier, A. (1952). The ComprehensioeHebrew CalendarNew Yotk. 13.1 HISTORYOFTHEALMANACS
T\rng Tso-Pin (1960). Chronological Tables oJ Chinese History Hotg Kong.
Welch, W.C. (1957). Chinese-AmericanCalendarfor the 40th through the 89th Year of From earliest times humans have been interested in the Sun, Moon, planets, and
the ChitueseRepoblicWashington. stars, and in determining their positions. In particular, they are interested in pre-
Zerubavel,E. (1985). The SeuenDag Circle New York. dicting the positions of solar-system bodies and interesting phenomena, such as
eclipses, planeta.ry groupings and directions, and times of rising and setting. The
results of these interests were numerous mathematical approaches to the ca,lcula-
tions, observations of limited accuracy, and attempts to predict the positions of the
"wandering" objects.
Eventually national publications of ephemerides appeared in various countries:
in France, the Connaissance d,es Temps (1679); in England, The Nautical Alma-
nac o,nd,Astronomical Ephemeri,s(1767); in Germany, the Berliner Astronom,isches
Jahrbuch (1776); in Spain, the Ephemerid,es AstrononLicas (1791); in the United
States, The American Ephemeris and Nauti,cal Almqnac (1855);in the Soviet Union,
the Astronomical Yearbookof the USSR (1923); in Japan, the JapaneseEphemeris
(1943); and in India, the Ind.ian Ephemeris arul. Na,uticalAlmanac (1958). Initially,
these publications were based on r,arious theories ofthe motions ofthe bodies, differ-
ent astronomica"lconstants, difierent geographical coordinate systems, a.nddifferent
timescales.
International cooperation led to the use of standard reference systems, systems
of constants, and common timescales. The brief histories that follow a.reconcerned
with the major changes of form and content.
^ "'
, ,-z1,'\to'
13.11 TheiAmericatEphemeris
"The Commissioners of Longitude, in pursuance of the Powers vested in
them by a late Act of Pa,rliament, present the Publick with the NAUTI-
CAL ALMANAC and ASTRONOMICAL EPHEMERIS for the Yea.r 1767.
 610 EXPLANATONY
SUPPLEI\4ENT 13/ HTSTORICAL
TNFORMATToN 61 1
to be continued annually; a \Vork which must greatly contribute to thc Im- Difference will be the Longitude in Time from thc Meridian by rvhich the
provement of Astronomy, Geography, and Navigation. This EPHE\{ERIS Watch rvas set; as near as the Going of the \Vatch can be depended upon.,,
contains every Thing essential to general Use that is to be found in any
Ephemeris hilherto published, with many other useful and interesting Par- Apart from many changesin the sourcesof the data in the tables from which the
ticulars never yet offered to the Prrblick in any Work of this Kind. The Moon's position was calculated, the rnain pages of the Almanac remained essentially
Tables of the Moon had been brought by the late Professor IIAYER of Got- unchanged until 1834. For that year, to quote from the preface:
tingen to a sufficient Exactness to determine the Longiiude at Sea, within a
Degree, as appeared by thc Trials of several Persons rvho made Use of them. "The NAUTICAL ALMANAC and ASTRONOMICAL EpHEN{ERIS for
The Difficulty and Length of the necessary Calculations seemed the onll. the Year 1834, has been construct-ed in strict conformity with the rec_
Obstacles to hinder them from becoming of general Use: To remove rvhich onllnendatiorls of the ASTRoNo\,IICAL SOCIETY of LONDoN, as con_
this EPHEMERIS rvas made; the Ntlariner being hereby relieved from the tained in their Report...; and will, it is believed,be found to contain almost
Necessil,yof calculating the Nloon's Place from the Tables, and afterwards every aid that the Navigator and Astronourer can require.,,
computing the Distance to Secondsby Logarithms, rvhich are the principal
and only very delicate Part of the Caiculus; sr.rthat the finding the l,ougi The changes werc both fundamental and substantial, and aimost doubled thc
tude by the Help of the EPHENITERISis now in a N'lanner reduced to the size of the book. The most fundamental change was to replace apparent time as the
Computation of the Time, an Operation..." argulnent of the ephemerides with mean time. In thc words of the reporl:
"All the Calculations of the EPHEN{ERIS relating to the Sun and
Nloon were made from Mr. NIAYER.'S last manuscript Tables, received bv "The attention of thc Committec was, il the first instance. <iir.ectedto a
the Board ofLongitude after his Decease,which have been printed under my subject of general importance, as affectirrg alnost oll the results in the
Inspection, and wiil be published shortly. The Caiculations of the Plarets Nautical Almanac; viz.. whethcr the quantities therein inserted should irr
were made from Dr. HALLEY'S Tables: and those of-.." future bc given for apparent tlrrte (as heretolbre), or lor meult solor time.
Considering that the latter is thc most convcnient, not only for ever-,vplu-
. pose of Astronomy, but also (from the best information t,hey have been able
These extracts from the preface to the fir'st cdition of The Nautical Alrnanac
antl American Ephemeris (7767) were written by Nevil l,Iaskelyne, then Astronomer to obtain) for all the purposes ol Navigation; at the same time tirat it is
Royal. Thc main incentivc for, and the main emphasis of, the pubiication u'as less laborious to the computer. and has already been introduced with good
the determination of longitude at sea rrsing the method of lurar distances. The efiect into the national Ephernerides of Coimbra and Berlin, the Committee
ephemerides were all given in terns of apparent solar time, lbr thc' reasorLsgiven irL recornmend tlre abolition of thc use of appa,r'enttine in a,ll the computations
the book's Explanation: of thc Nautical Almanac; excepting..."

"It may be proper lirst to premise, that all the Calculations ale nade The recognition of the irnl)ortance of practical astronorny a.lsoplayed a role! as
according to apparent Tine by the N{eridian of thc R.oyal Obser'\'atory itt noteclin the report:
Greenwich."
"And herc perhaps it rnay be proper to remark, that, although in tlicse
"What has been shown concerning ttrc Equation of Tirrte chiefly re-
discussionsthe Comrnittee have constantly kept iu vicrv the prirrcipal objcct
spectsthe Astronomer, the Mariner having little to do rvith it, irr computing
fbr rvhich the Nautical Almanac was originallv formed. viz, thc promotion
his Longitude from the Moon's Distances from the Sun and Stars observed
and advancement of n autical astronolny, they have not bccn rurmindful
at Sea with the Help of the Ephemeris, all the Calculations thcr-eof being
that, by a vcry siight exterrsiorrof the computations, and by a few additional
adapted to apparent Time, the same which he will obtain by thc Altitudes
articles (of no great expense or labour), the work might be rendered equally
of the Sun or Stars in the N{anner hereafter prescribed."
useful for all tlre purposes of pracfiicol astronomy."
"But if \\htches madc upon NIr. John Harrison's or other equivalcnt
Principles should be brought into Use at Sea, the appalerrt Time, decluced
The requirernents of the navigator wcre by no meatrs ovcrlooked; in particular,
from an Altitude of the Sun must be corrected by thc Equation of Time,
the number and presentation of "iunar distances," including distances from the
and the mcan Time found compared with that shewn by the Watch, the
612 EXPLANATORY
SUPPLEMENT 13 / HISTORICALINFORMATION
613
planets, was greatly improved. Hoq'ever. a completely neu'explanation was writtcn The Astronornical Ephemeris. The changcs are full1. dcscribecl
in thc prefacc to
q'as made to the use of the ephemerides for navigation.
in which little reference the edition for 1960. (Some adclitional notes on the historl,of
The Artronontit,al
Tables of refraction were excluded ald no example wns given of clearing an observed Ephe'nr'e'risarc given on pages ix xviii of the v.rume for 1962,
the two-rirrndredt,h
Iuuar distance for the efiects of semidiameter' parallax. and refraction. anniversary etlition.)
Apa,rt from the ornission of lunar distances in 1907, the first part of the almanac.
containing the ephernerides of the Sun and Moon, remaincd unchangcd in form
13.12 TheAmerican
Ephemeris
and Nautical
Almanac
until 1931. From tirne to time. tho dzrta was based on different data and tables.
and occasionally new material n'as added, sttch as ephemerides of the Nloon and During the first half of the ninetcenth century, ?he Nauticar
Alntanac rornained irr
planets at transit on the Grcenwich meridian and the apparent placcs of many general use on American ships nnd among astrorromers and
survcyors in the Unitcd
more staxs. Later, cphemerides for physical obserr-ationsrl'ere included; and in 1929' States. However, $-ith the continued der.elopment of the countr1,,
antl its grotvth
anticipating tlte redesign in 1931, cphemerides of the Sun referred to the standard as a nraritirne nation, an incrcasing need for a nationzrl alr.arrac
was felt. This
equinox of 1950.0were given for the lears 1928 and 1929. need eventually led to a bill in Congross, approved in 1g49. to establish
a Nautical
Much of the added matter t'as of no interest to the practical uavigator. In Almarrac Office in the Nar-v Departnent. The Office rvas set ul).
during the latter
1896, "Part I (contairring such data as are more partictrlar required for navigational part of 1849 in Cambridge, N{assachusetts.where library an<i pritrtirrg
Lacilities wer.e
purposes)" was "also published separately for the convenicrtce of sailors." This available. The first voluure of The American Ephemeris arrd Nau/tc:al
Ahnanac *,its
consisted of a straight reprint of thc monthly pages comprising the first part of for the year 1855, and u'as published in 1g52. The officc li'as moved
to \\iashington
the almanac, with selectiorrs frour thc other data and a fcw specially prepared in 1866, and therr u'as rnoved rLgainto the Naval Observatorv il 1893.
pages. In the Preface to Lhe 1914 edition. it rvas announced briefly that "Part I has For the vears 1855 1915 the r.oluureq,ascijvided irrtu ruo parts.
l.hcn. begirr,
been remodeled for the convenience of sailors": thus s'as introduced The Nadical ning with 1882, it was divided into tluee parts. The first part
cluring this entire
Alnr,anac, Abri.d,gedfor the Use of Seamen. wlich was specially dr:signed for its period u'as an ephemeris for thc use of navigators that was
also reprintr:d sepa_
purpose. This almanac rvas redesigrredin 1929 and in 1952. *,hen it u'as renalned ratell'. r'ith the inclusir.r' of a ferv pages from thc remairrcler of thc
rolur'e. as T6e
The Abrid,ged,Nautiatl Almanac; it was rearrangcd in a different form in 1958. Since Ameri,can No,uticalAlmanac.It comprised twclvc uronthlv sections, lbr the
meridiarr
1960, it takes on the appropriate poltion of l,he original title, namely The Nautical of Greenwich, each containing cphemerides of thc surl. \Ioon. iurd
runar distanccs
Atrnanac. for the month. Follou'irrg the monthll. sections rver.eephemeridc*
oI Verurs, \Iars.
Prior to the revision in 1931, a fundamental change in the measure of ntean Jupiter, and Satrrrn for the ycar, a,nd, beginrrilg q.ith 1gg2, of N,Ielcur3r,
Uranus,
solar time had taken place, rvhich required changes in the almarrac. Before 1925. the and Neptune.
astronomical day u'as considcred to start at noon. and the principal ephemerides The secon<lpart of the vohule containedephemeridcsof the Sur. \Ioon. pliur_
had been given for 0h (i.e. noon) on each day. As of 1925 .Ianuary 1 the tabular ets, and principal stars, for mcridian transit at Washington; and tlata on eclipses.
day n'as brought into coincidence q'ith the civil daf iurd r.as considered to st.rrt at occultations,and a few other pheDomena.In 1gg2.the sectionson phetorneDawcrc
midnight; the ephemerides*ere still given for 0n, now indicating midnight. formally grouped as a third part ivith the titlo ,.phcnonrena....l.he explanatory
The revision of 1931 was much more thau a rearrangement of thc same data sections and a lew misccllaneorrstables <:ompletr:dthc volume.
into a different fornr. Thc changes of page size. of presentation, of provision for' During the period lli55 1915. fen'changes rveremade in thc form or content.
interpolation, and of conterrt were less important than the conlplete break rvith the The nautical part rernairredvirtually unaltered; lunar distanccs were orlitted. be
century old layout designed primarily for navigation. The neu' form wtm designed for gtnning u'ith 19i2, but n page explaining how to calculatethen u,as
i'cl.decl. Th<r
the astronomer without considering the requirements of navigation. Its arrangettx:nt principal .e'isions in the other parts of the 'olume *erc in 1882
and 1912 r9r3.
remained basically rrnchanged.though there were freqrrent charrgesin content of the The rearrangerucnt of the 1882 volume was accornpaniecl by sonre additions arr<l
less fundamental matter. omissions. The principal ornission was thc epherueris of Moon_culrninating stars for
Major changes rvere introduced in the 1960 cdition q'hcn the almanac rvas determining longitude. The prirrcipal additions u,ere:thc physical epheureridesof
unified with The American Ephemeris,including the use of ephemcris tirne, in- Mercury and Vcnus for the reduction of meridian and photometric observations;
stead of Universal Time (mean solar time on the mcridian of Greenwich). a-sthe daily diagrams of the configurations of the four great satellites of Jupiter; and
argunent for the firndamental ephcrnerides. This change further enrphasized the ephemerides for the identification of the satellites of N,Iars, Saturn, Uranus, and
unsuitability of the volume for navigation, arrd led to the adoption of its new title: r Neptune. In the volume for 1912, the ephemerides of the satellites were extended
 EXPLANATORYSUPPLEIJENT 13 / HISTORICALINFORI\,IATION
614 otc
position arrgle and apparent dis- prepared cooperativell' and ri'ith the exception of a few pages prior to the tabular
to include tables for determining the approximate
\{oon' N'lars, and
tance; in 1913, physical ephemerides were added for the Sun, data, the volunres *'ere identical. The titlc The Astronomical Ephemeris replace4
Juptter. the title of The Nautical ALmanac and t.he Astronomiatl Eph,emeris.The title ZAe
Although the volume for 1916 (the first to be issued under the international Nauti,cal Almanac was used both in the United Kingdom and the United States
contained ferv content
agreements resulting from the Paris Conference of 1911) for the unified edition of the almanac for surface navigation, which had previously
that had
changes, extensive revisions were made in the form and artangement been titled The Abridged Nautical Almanac, arrd The American Nautical Almanac.
of the Green-
been reiained essentially uncharrged since 1882 The arrangement respectively. In accordance with the resolution of the IAU, the designation ephem-
and Moon by monthly sections was discontinued' and eris time was adopted for the argument of the fundamental ephemeris ol the Sun.
wich ephemerides of the Sun
The Arnerican' Nautical Alma-
replaced by annual ephemerides. At the same time, Moon, and planets.
was a
,ro" ."u""d to be a reprint of part of The American Ephemeris, but instead From 1960 onward, ?[e ATtparent Places of Fundo,mental Stars rl,'aspublishcd
sepa,ratelyprepared volume especiallv designed for the navigator' by the Astronomischcs Rechen Institute in Heidelberg, Germar5,.,under the auspiccs
In 1925, the astronomical reckoning of time frcrm 0h at noon rvas replaced bv of the International Astronomical Union.
r h e c i v i l r e c k o n i n gl r u m m i d r r i g h t The IAU System of Constants adopted by the IAU at its Tlv'elfth Gcneral
During the interval from 1916 until the fundamental revisions in 1960 when Tie Assembly in Hamburg, Germanf in 1964, were introduced with the edition lor
American Epherneri'swas unified with ?he Astronomical Ephemeris' the revisions 1968. The supplernent to the 1968 edition was bound in the back of the volrrme
of form and content were mostly only in details, althorrgh a few major changes and provided the details concerning the changes in the constants. An appendix
were incorporated. In the volumes for 1934 1937 a number of further subdivisions was bound in the front for the 1.ears 1968 through 1971 and gar,.eformulas and
and rearrangerlents of the contents were made. In 1937, the volurne was fortlally corrections for conversion to ihe IAU System of Astronomical Constants. Frorn
divided into seven parts; the part constituting the epherneris for Washington was 1972 through 1980, the editions wele published rvith only minor physical changes
reduced to only ephemerides o{ the Sun. Moon' and planets for meridian transit at in the arrangements of the publication.
Washington all the other material rvas transferred to other parts and referred to Beginning with thc edition lor 1981, the title ?Ae Astronorical Alrnano.cre-
the Greenwich meridian. placed the tille The American Ephemeris and Nautico,l ALmnnac and the title TAe
The usefulnessof the Washington-transit ephemerides was limited to observers AstrononticaL Ephemeris. Thls edition of The Astronornical Alrnanacwas publishcd
on the \\rashington meridian, therefore the publication of this part was discontinued jointly by the U.S. Government Printing Office and Her Majesty's Stationery Of-
beginning with the 1951 r,olume. The other principal changes in content during fice. Tlre physical content of The AstronomicaL Almo,nac changed in many respects.
1916 1959 were the following: in 1919, tables of the rising arrd setting of the Sun Howevcr, the bases for the ephemerides were unchanged, except that the funda
and the Moon were added; in 1941, the number of stars for wltich apparent places mental hclioccntric cphemerides of the Earth. \{ercury, and Mars s'ere computed
were given rvas decreasedto 212 (after reaching a high o1887) w\et Apparent Places directly from Newcomb's theories instead of from Newcomb's Tables, and for the
of Funtlamental Slars rvas first published; in 1957, apparent places were omitted Earth, thc IAU 1964 System of Astronomical Constants was used.
entirely, but precise mean places of 1551 stars, which had been given beginnirrg in ID 1984, the basesfor the ephcmerides r;r'erechanged. New fundarnental cphem-
1951, were continue<i. The departure of the N{oon from gravitational theory drre erides of thc planets and the \'Ioon were prepared at the Jci Propulsion Laboratory
to the variations in ihe rotation of the Earth made it necessary to snccessively and titled DE200/LE200. These ephemeridesrverc in gcncral accord rvith the recom-
extend the elements and predictions of occultations to more and fainter stars and mendations of the IAU 1976 System of Astronomical Coustants, except for minor
additional standard stations. An ephemeris of Phrto was added to the planetary modifications necessary to better fit the observations. The dynarnical times,-ales
ephemeridesin 1950;and ephemeridesol Ceres,Pallas, Juno, and Vesta in 1952' and the standard of reference of J2000.0 were introduced. The day numbers and
star positions were referred to the equinox and the equator of the middle o{ the
year instead of the beginning of the year. A supplcmcnt lo The A stronomi caLAl-
BritishandAmerican
13.13 TheCooperative (1960toPresent)
Almanacs
manac for 1984 was bound in the back of the volume, and gave the details of the
Beginning with the editions ol 1960, The Astronom'ical Ephemeris z:r'd The Amer- resolutions adopted by the International Astronomical Union and documentation
ican Ephemeri's antl Nautical Ahnanac wete joint publications The contents were concerning the ephemerides.
 SUPPLEMENT 1 3 / H I S T O B I C A ILN F O B M A T I O N
616
EXPLANATORY 6't7
COOPERATION Photographique du Ciel on the initiative of the Bureau des Longitu<les.Its pur_
OF INTERNATIONAL
13.2 HISTORY pose wa-s 'tl'etablir une entente permettant d'augDlenter, sans norrvt:aux frais.
la
masse dc donn6es numeriques fournies annuellement aux observateurs ct aux calcu_
Priorto the IAU
13.21 Cooperation lateurs." Although the CoDfererrce was primarilv coDcernedwith obtairringn greatly
International Merid- increasecllist of apparent placesof stars, it extcnded its attention t,r all the cphem_
Formal international cooperation began in October 1884 at the erides of bodies in the solar svstcn. Its comprehensive reco[rmenoaTrons co\€red
included:
ian conference held in Washirrgton The resolutions of that conference the distribution of calculationsbetweenthe fir.epri.cipal ephemerisoflices(France.
the center of thc transit Germany, Great Britain, Spain, and the United States). specifietlst;r.rrdards of cal_
"the adoption of the meridian passing through
of Greent'ich as the initial meridian for culation and presentation, arranged for publication of additional clata, ancl fixed
instrumerlt at the Observatory
the values of two further constalts to be rrscd irr the ephemerides:thc flattenine
longitude "
directions of the Earth (1,1297)and the semidiarneterof the Sun at unit distart.o for
"That from this uleridian longitude shall be counted in two "clipse
being plus and wcst longitude minus " calculatiorrs(15'59:'63).N{ost of these recommen.lationsar.estill in forrt.
up to 180 degrees, east longitltde
day for all purposes for which it nrat' be Ofiicial approval rvas iu sorne cases necessart'for the adoption of these rec_
"the adoption of a universal
ollmendations. as illustratcd by the follorring extract fiom the Act of Congressof
found convenient"
A u g u s t 2 2 . 1 9 1 2( 3 7 S t a t . L . , 3 2 8 , 3 4 2 ) :
"That this universal daf is to be a mean solar da1';is to begin for all the
world at the moment of mean midnight of the initial meridian' coittcicling
"The Sccretary of the Navy is hcrebv authorizccl to arrange for the cx<fimrge
with the beginning of the civil day and date of that meridian; and is to be
of data with such foreign alma.nac officcs as he ma1, from tinle to titno
corrnted from zero up to twentv-four hours."
dceur desirable,u'ith a vierv to reducing the arnount of duplicatiorrof l,ork
"That the Conference expressesthe hope that as soon as may be prac-
in preparing the differerrt natiorral nautical ancl astronomical alruana<.sancl
ticable the astronomical attd natttical days rvill be arranged everl-where to
increasingthe total data s hich mal' be of use to uavigatorsand astLouomeLs
begin at mean midnight."
availablefor publication irr thc AmericarrEpheurcrisand Nautical Alnrauac:
Provided..."
Although the other resolutions u'cre adopted. the convention of wcst lougitude
as positive was used in astronomy until 198'1 when tlte general use of East posi-
Here follorvs a number of plovisions. The nost iurportaut provisiou astr.onomi-
tive was introdrrced. The Astrontt'micaLAlmanac switchecl convention at thtrt tine,
cally was the repeal of ttre proviso in the appropriation Act of Septerrrber28, 1850
exactly one hundred years after the Intcrnational Nlelidian cont-etence'
(9 Stat. L., 513. 515) that "hercafterthe rneridianof t he observator1-
at \\ ashirrgton
At the iDvitation of the Bureatr des Longitudcs. the directors of the uzrtional
shall be adopted and userlas tlrr. American mericliiurfor all astrononricalpur.poses.
ephemeridesand other astronomcrs.met irr Paris in \14]'1896 for the Confercnce
and that the utcridian of Greenl ich shall be aclopteclfor all narrtical prrrposcs.'
Internationale des Etoiles Fondarncntales. In addition to adopting tesolutions con-
Such excltangeagreetucntsh:lvc been carried ollt in spite of intellal ional dili-
cerning the fundamental catalog, zrttd the calculatiotr and publicatiotr ol apparent
culties.
places of stars, the Conferencc atlopted the follorving fundarnental constallts:

Nutation 9':21
Abcrration 20':41 13.22 International
Astronomical
Union
Soltrr parallax 8'180
In 1911.trgrec'ments had been directed almost entjrclv to the reductiorrof tlle total
amount of work by the avoidanceof duplicate r.ak.rrlation.In 193i1,Conrurission
It also agreed to adopt Ncwcolnb's definitive values (which u"ere ttot theD in
4 reconrruoudcdthat this principlc should be cxtqr(lod to the avoidarrceof dupli-
final form) of lunisolar and plartetarv precession
cate prrblication by the collection in a single volurnc of the apparent places of stars
Active <rxrperation betwrrcn thc offices of the nertiortalephemericlesdatcs fiom
then printrrl in each of the prirr<.ipalcphemerides.This recommendation,coupled
the Congrcss International cles Ephemerides Astrorl(nniques hcld at thr: Pnris Ob-
with the adoption of lhe Drittcr Fund,amentalkataloq des Berliner Astronontischen.
servatory in October 1911. This confcrencervas called by B' Bailaud, Director of
Jahrbuchs (FK3), u'as implerrrentedfor 1941by the publication of the international
the obser.vatory and President of the comite InterllatioDal Permanent de la C'arte
 EXPLANATORYSUPPLEMENT 13 / HISTORICALINFORMATION
618 619
The IAU Symposium No. 21 (paris. May f963) conclrrded that a change
',tolutoe Apparent Places of Funda'ntental Stars (under ihe auspices of the Interna- irr
gave astronomers access to the apparent the conventio.al IAf.l systern of constants courd no longer be avoided. The incon-
tional Astronomical Union). This work
oflices {'ere savcd the sistencies and inadequacies of the system of that time, the better values of
places of stars in one volume, and the individual epherneris some
proofreading' as well as the cost of typesetting' constants from recent determinations, and deficiencies revealed by discussions
work of the compilation and of
of high accuracy observations indicated the need for new constants
Continuing the precedents of the 1896 and 1911 conferences' the Dircctor
a further conference that At, the Tu,'elfth General Assembly (Hamburg. 1964) a workirrg grorrp revrewed
the Paris Observatory (Professor A. Danjon) convened
in Nlarch 1950 to discuss the fundamental constants of astron- the report of the IAU Symposiurn. A list ol constants proposed by the Wbrkiqe
was held in Paris
no change be made in the conven- Group on thc System of Astronomical Constalts was adopted and re"61nnqn4.j
omy. The leading recornmendation &'as "that
with the most for use at the earliest practicable date in the.ational and international astronornical
tionally adopted value of any constantii But the recommendations
and brought ephemerides. The$e constants vrere introduced in the ephernerides fbr 196g. It was
far-reaching consequenceswere those u'hich defined ephemeris time
with the solar ephemeris in terms of ephemcris also troted that the constants of precession and planetary trasses had not been
the lunar ephemeris into accordance
to the International Astrottomital changed and that consideration should be given to their future iulJrove[rent.
time. These recommendaiions were adtiressed
of thc In August 197U, IAU Colloquium No. 9 cin the IAU System of Astlonom_
Union and formallv adopted b-v-Commission 4 and the General Assembly
ical Constarrts was held in Heidelberg, and r.ecommelded the establishmeut of
Union in Rome in SePtember 1952.
three working grolrps on planetary ephemeridcs, on plecession, and on units and
Commission 4 ha<i, at \arious times, made arrangernents for the redistribu-
for timescales. The recommendations were adopted and the u,orking groups were es_
tion of calculations between ihe ephemeris offices For example, the 1n"1i1q1e
places of stars to the tablished at the 1970 IAIJ General Assemblf in Brighton, England. A working
Theoretical Astronomv in Leningrad contributed apparcnt
meeting on constants and ephemerides rvas held in October 197,1.in \\rashington.
internatiorral volume for the years 1951 1959. with the availability of fast, au-
DC to draft a proposed report of the working groups. The chairmen ofthe *,orking
tomatic calculating rnachines, it $'as rlow both practicable and cfficient for large
groups met in September 1975, arrd June 1976, in Herstmonceaux and \!-ashing-
blocks of work. such as the calculation of apparent places of stars, to be done in
ton, rcspectively. The Report and Rccommendations, known as the Joint Relnrt of
onc omce. At the 1955 General Assembll' of the Union in Dublin' a generai redis-
the Working Groups of IAu Comrniss,ionI on Precession,Planeto,r.yEphemer.itles,
tribution of calculations along these iines was agreed upon by the directors of tl.re
Un.its and Tirne Sco,Iesu'r:re adopted by the IAU in August 197tj in Grenoble.
national ephemerides and confirmed by Cornmission 4 The reports of Comn.rission
In 1976. a r.orking group otr cartographic coordinates and rotational elements
4 in Transactions of the International Astronomical Llnion dcrctrmented the full
of planets and satellites n.as established by IAl.l Cornmissions 4 ald 16. In 1977, a
details of these agreements. the charrgesirr the bases of the ephemerides, ancl thc
working group on nutation was established bv Commission 4. These groups \4,ereex
discussions leading to the introduction of Apparent Places of Funr)ome'tftal Stars'
pected to pr-or.iderecorlmenclatiorrs for considcr.ation at thc 1929 General Assembiv
In the Draft Report for the Trvelfth Gtrreral Assembly' held in Hamburg in
in \4ontreal.
1964. D. H. Sadler indicated that the fnnctions of Commission '1 on Ephemerides
were twofold: "firstly, to cllsure that the published ephemerides fullv meet the re- The second prinary function of Commission 4 ol ephemeridcs the coolclina-
quirernents of astronomers arrd other usersl and secondll'' to coordinate the work of tion of the eflorts of thc various ofices was bcins fullilled wilh the move from the
the offices of the national ephemerides to ensure consisterrcy.economy of e{Iort' and distribution oI calculations arnorrg the separate offices. thc unifit:ation of printing,
and thc oxchange of reproduction proofs for printing by the various countrics. This
efficiency." Cooperation within the IAU on ephemerides has bcen clirected ton'artl
changc was made possible by the advent of high-speed courputcrs.
these functions. The primary emphasis of the first function has bcen thc establish-
ment of a consistent and urriversal set of constants and bases for the ephemerides' .A.tthe Fourth Gerreral Assembly of the IAU held in Carnbridge, N{A in 1932,
Dr. L. J. Comrie, Dircctor of the British Nautical Almalac OI}icc, sug€lested that
A brief review of the cfforts in these two areas may be helpful.
duplicate printing in thc national volumes of ephcmeridesbe discussed.particu-
At a meetine on fundamental constants for astronornl'held in Paris during
larly lvith lespcct to the apparent places of stars. At that time. Professor llcrrelo
March 1950, the dellnition of cphemeris time was recommended,and the lunar
srrggestedthat i'lre ideal would be an internntional alrnana,c.At the Fifth Gerrcral
ephemeris was brought into accotclance with the solar ephcmeris I'ith respcct to
Asscmbiy (Paris. 1935) an agreernent r'"-asreached for a single publication of the
ephemeris time. These recommendations t'ere adopted in 1952 by Comrnission'1 of
the IAU at the Eighth General Assembly in Rorne.
 SUPPLEMENT
EXPLANATORY 13/ HISTORTCAL
TNFOBMATTON
620 621
Apparent Places of Stars, initially to bc printed in Great Britain artd nou' priutc<l I'October 1974,plaru wrtr.ebegun to l.evisethe orgarrization,collte't, aDd
b.sis
fot The Arnet.ican.Ephemeris and, Nau,ticaLAlmanac/AstronorniatL Ephemeris.
in GermanY. It
At the Ninth General Asscnbly (Dublin, f955). an international fundamerttal *trs decidedthat a single.nificd printing in English u'ould be madc iu the
Unitccl
astronomical ephemeris (IFAE) r'as discussed. The IFAE was to bc a single pub- States, and that it would bc availablefronr both Her \Iajest1..sSlatiouerl,
Officc
Iication, under the auspices of the IAU, containing the fundamenttrl astronoutical in England a'd the superintende't of Do<uments in the united statcs. sincc
tlrt:
ephemerides to the fullesi accuracy The national ephemerides could then be mrt<rh United Statcs legal code requircd the title ?he Amerir,an Ephentens anrl. Nau.tical
smaller and cater more directly to the practical astronorner. Although nlany $'ele in Almanac, a bill was introriuqrd into Congress to modify the legal cocle to pcrrDit
a
favor of this proposal, several factors prevented its adoption: the practical dilicultl' change in the namc of thc ptrblication. Also. it rvas decided th:rt thc organizatirrr
that the sales were not Iikely to cover the cost of printing, because of the arrtic- and content would be changedfor the 1981 edition.
ipation of required, free distribution; the difliculty for astrononrets of purchasing
books published in other countries; the rcquired cooperation of alnrost all nationtr,l 13.23 Otherlnternational
Organizations
ephemerides; artd the loss of flexibility Horvever. various countties usitrg reproduc-
I'rprovemerrts i. accuracies and in knou,ledge concernirrg the motious aud rotatio.s
tion proofs of material prepared by a single source rvas an attra<tive alternative
of the bodics i' the solar systern,partic'la'ly the rottrtion of thc E.rlh ancl orrr.
At the same tine, it was announced that an agrcetnont had becn reached for the
unification of The American Ephemeris arul Nautical Almanac and The Nautit:al use of accurate time, has created a gron'iug need for doser cooperation betlr,et:rr
Almanac and Astronomical Ephemeris beginning with the 1'ear 1960. It I'as hoped th. different i.temational organizations.Thus. mau1, ilrternational or.ga'izatiorrs
that other ephemeris offices tl'ould save cousiderable composition arrd proofreading be<omeinvolretl in rariorrs .rsr)ectsof internatio'al stanclardizatio'. so're of thtsc
costs by using photolithography for reprodrrction. different organiz,tions and thcir functions are describetlirr the foll.rvirg pa.agraph.
At the Tenth General Assembly (Moscow, 1958) it was reported that th(] The International Associationof Geodcsy (IAG) is irrteresteclin tlxr rotational
Astronomisch Geod,atischesJahrbuch, introduced for the year 19'19, would ceasc elementsand physical chara<teristics of tht: planets. Thc Cornmittec orr SpaceRe
publication after the 1959eclition and that the Berlincr A stronontis<:hes Jahrbuclt. search (COSPAR) is interestcd in the ciutographic coorclinates\.stemsand the
introduced for the year 1776. would ceasc publication alter the editiorr for 1959 gravity 6elds of thc planets alld satellites.The Consultative Comnrittee on IDter_
As a result of these savingsof composition and printing costs,the Astronomisches national Radio (CCIR) is responsiblefor standardizatiorrof tirne arrd frequcrrcy
Rechen-Institute was able to take over composition atd publication ol the Appat as used in ratlio transmissions; its Stud). Clr.rup Sevcrr rcaches agreerucnts in this
ent Places of Fundamental Slors. Although it t'as sacl to see a publication of loug area. The CorrrntitteeInternationale cles Poides et \lesurcs (CIp\l) <rrordinates
standing cease, the change indicated an increased international cooperation iurd the standards of the units of measure.Thc Consultative Committrt ol the Dcf,
permitted a beneficial transfer of functions. inition ol the Srxud (CCDS) is responsiblefor the dcfinition of thc second arr<l
G. M. Clemence and D. tI. Sadler rcported at thc Eleventh Gctreral Assemblv the principles for ureasuring the second. The Committeo on Data lor S<riencean<l
(Berkeley, 1961) that an important step torvards unification of the national epheltt- Technologl. (COIIATA) r'orks on an interrlisciplinary basis to improvc tlx: qualitv,
erideshad taken place in the unification of the American and British volumes.Thcv reliabilitl.. prortssing. managcrnent,and accessibilitl.of rlata of irnportanceto sci-
stated further thtrt: ence and technologv. The Bur.cau Internationale des Poiclcs et Nlcasurcs (BIPN{) il
Paris, France,is rcsponsibkrfor.rnaintaining InternationzrlAtomic'firrru (TAI) arrd
"Many dearlyJreld, but cssentially urlitnportant. standards and prejudices providiug a sirrgleatomic tirucscrale. worldwide. The Intcrlational Flarth Rotatiorr
have had to be sacrificeson both sides; it is surprising horv quickll' these lose Service(IERS) rleternines thc rotation of tho Earth. and thus the UTC tirnescale.
their former importance in the satisftrction of a compreherrsive.rgreement. It also determinesrvheu leap secondsshorrltlbc insertcd into the UTC tiluescale.
That same cooperation, good will arrd confidence exists bctwccn all the
national ephemeris offices, and, although differenccs of languagc will intro-
13.3 HISTORICAL
LISTOFAUTHOBITIES
duce soure further difficulties, there is no obstacle to complcte urtification
that will not be overcorne in course of time." 13.31 Introduction
Between 1962 and 1964 the Japanese and Sovict ephemerides bcgan using the It is irnportant to know the basis of the ephemeridesto properly interpret the results
advairced proofs from The Astronomical Ephemeris. of pzrst discussious of observations. For this reason, and for purposes of historical
622 EXPLANATOFY
SUPPLEMENT
13/ HTSTORTCAL
tNFORt\.,tATrON
623
interest and record, the following list of authorities is given in the original form. In 1797 1804: Mayer,s tables with the mean motion correctedto the revrsed
addition, a ferv uncertainties and ambiguities, rvhich could onll'have been resolved preces-
sion of 501'2.
by excessiveresearch or recalculation, have been allowed to remain. 1805 1812: Delambre,stables as given in Lalande (1792), but rvith certain
The material is arranged in the following subgroups: up to 1900 The NauticaL (un_
specified) coefficicnts determined by Maskclyne.
Almanac and The American Ephemeris are treated separately, and after 1900 they 1813 1821: Improved tables by Delambrc (1806).
are combined. Each subsectionis divided according to body or srrbject (e.g., Sun, 7822 1832: The tables in Vince (180g, Volume III), ,,with the omrssior
Moon, precession. nutation, satellite constants). \\rithin each division the author- onlv of
some equations which do not materially effect the results.,. Ttre
ities, arranged chronologically, are preceded by a short narrative of the quantities tables
are stated by Vince (lblume III. page 2) to have been ,.constructed
tabulated. In these discussions "and" is used to indicate that the quantity is tabu-
bv M. de Lambre, frorn observations of Dr. Maskelyne, ancl the
Iated to r decimai places, and the term "precision" is used to indicate merely the theorv
of M. Laplace.SeeLes Memorics de Berlin, for 1784.1785..,(In
unit of the end figure. Usually, only names and dates are given for the unit of the 1832.
the position of the Sun for the calculation of the transit ,rf
end figure. Also, only narnes and dates are given for the authorities; full references flercurrl
was taken fronr Carlini,s tables.)
are given in Section 13.4. In some cases,for example for the adopted semidiameters
1833: The longitude was taken from Delambre's tables, improved by
of the planets, detailed referencesto the original publications have been omittecl. Airy,s
corrcctions based on Greenwich observations.
Some of the tabulated ephemerides are based on theories. dcrivations, and
1834 1835: Carlini's tables (1810) with Bessel,scorrections (1828) and nutation
constants given in appendices and supplements to ?r\e Nautical Alrnanac and to
rLsin the Astronomical Society's tables (Bail1,, 1g25). Thc elements
The American Ephe'neris. Lists of thcse appendices and supplements are given irr
used by Carlini are the same as those of Delarnbre (1806). but the
Section 14.6.
arrangement is better for the construction of an ephemeris.
1836 1863: Carlini (1832).
1864-1900: Leverrier(1858).
Almanac(1767-1900)
13.32 TheNautical
13.322 Moon The N,Ioon'slongitude, Iatitude. semidiamcter, and horizontal par_
13.321 Sun All ephemeridesfrom 1767 to 1833 were gir.en with the argumelt ap-
allax (each to 1") and its right ascensionand clcclination (each to l,) were tabulated
parent time. Quaatities tzrbulated for the Sun at irtervals of one day were longitudc
at inter\.?ls of 12h (apparent timc) for thc years 1Z6Z 1833. Lunar distances (fronr
and declination (each to l"), and right ascension and equation of time (each to l").
at least one star, and after 1770 from one or t$,o stars as well as the Sun rvhen the
Semidiametersin arc (to 0ll). in time (to 0:l), and log distancc (to 6D) llere given
Sun was conveniently placed) u,.cregiven to a precision of 1,,for every 3h. Frorn 1g23.
at intervals of6 days. From 1768,the equation oftirne (to 0:1) was given. ald fronr
the right ascensionand declinationnere given to 1". In 1g34,the argument becarne
1772, the right ascension (to 0:l) was also given. For the years 1815 1822, tho log
mean time and, with occasional rninor alterations, the tabulations werc given to aD
distance (to 5D) was given. In 1833, the semidiameters(to 0'i01, to 0?01)and log
extra figure until the y'ear 1900. The right a^scension(to 0:01) and the clcclination
distance (to 7D) r'ere given at intervals of one da1'. ( l o O : ' l ) w c r c g i v c r a t i n r e r r a l .o f o n e h o u r .
The 1834 almanac wils largely remodeled in accordance rvith the Ro1.al Astro-
nomical Society's report prirlted in that almanac, and thereafter the argument of
t (OI | / /tj:
tvlaycl last marrrrscr.ipt
iables,
most ephemerides was nrean time. Most quantities tabulated in time were given
1777 1788: Maver's tables, irrproved by Mzrsonunder Maskelyne's dircction, based
to Cli01,those in arc to 0l'1 (except the Sun's latitude, given to 0:'01),and the log
on Bradlel.'s obscrvations (the latter are printed ln The Na,u.ticalAl-
radius vector to 7D. In many cases,differencesor variations l€rc givcn. From 1848,
manac for 7774).
equational rectangular coordinates (to 7D) were inclrrded at intervals of one day,
1789 1796: Nlayer's tables were further improved by Mason (1280). Eight new
Iatitude terms being included for the frrst time in 1866; the valucs for 1845 1847
cquations were taken from Mayer's tables, the coefficients being de_
were given in the 1848 volume. No other substantial changes wcre made bcforc
termined from Bradley's observations. The eighteenth equation in lon_
1900.
gil ude u-asomir red.
1797 1804: The same set of tables, but adjusted (as for the Sun) for the corrected
1,767 1796: Mayer's "last manuscript tables" assumed an annual precession of
value of precession.
5013.
 EXPLANATORY
SUPPLEMENT 13/ HTSTORTCAL
INFORMATTON
624 625
Lalande (1792); the tables are the same as \Iason (1780) except for at intervals of 5 days, usuallv as an appendix to later Almanacs.
1805-1807:
the substitution of Laplace's acceleration and secular motion Transit ephernerides were introduced in lg39 for \,Iercury to Urarurs,
and in
Lalande's tables, with the addition of two further inequalities found 1861 for Neptrrne.
1808:
by Laplace.
r809-1812: The epochs, Laplace's accelerations, and "a particular equation of his"
Mercury, Venus, Mars, Jupiter, Saturn
were taken by Maskelyne from Burg's tables, and the rnean Iongitudes
were computed. The parallax was taken from Mayer. 1767 177Y Halley (1749).
1813 1817: Burg (1806) on Laplace's theory, the coefficients being determined 1780 1804: \4rargentin'stables, .annexed to NI. De Lalande,sAstrorrotrv.,,
from Maskelyne's observations, and the epochs from those of Nlaske- 1805 1833: Lalande (1792). These are the tables calculated by Delar.brr: on
the
lyne and Bradley. theory of Laplace. From 1822, the ta.blcs of Mars werc takcn from
1818-1820: According to Pond's preface, the tables of Burckhardt were rrsed. [Brrt "those of Lalande in Lhe Conn.aissancedes Tems fsic] f<rr the 12th
a note (initialed T.Y.) at the end of the 1820 preface statcs that those year 11803 04].,, The places of Mercury for the transit of 1832 q.ere
of Burg were used.] taken from Lindenau's tables.
1821 1833: Burckhardt (1812).
1834-1855: Burkhardt's tables, r'ith nutation from Baily (1825).
Mercury
1856: As in the previous years, but the parallax taken from Adams (1853b)
and the semidiameter taken as O.2725times the horizontal parallax. 1834 1863: Lindenau (1813).
1857 1861: The ratio of semidiameter to horizontal parallax was changed to 1864 1900; Leverrier (1859).
0.273174.
1862 1882: Hansen (1857).
Venus
1883 1895: Hansen,but q'ith Newcomb'scorrections(1878b)included iD the right
ascension and declination. 1834 1864: Lindenau (1810). For the 1-ears1837 1848 a correction of 2'1g,,l,as
1896: As in previous years, and with the substitution of Newcornb's Table
applied to the tabular longitude of the rrode.
XXXIV for Hansen's. 1865 1900; Leverrier (1861a).
1897-1900: Newcomb's corrections included in horizontal parallax and semidianl
eter.

13.323 Major Planets Ephemeridcs of the five 'classical" planets t'ere given at
1 8 3 4 1 8 6 5 : L i n d e n a u( 1 8 1 1 ) .
intervals of 6 days until 1832, except for Nlercury, which were given at intervals of 3
1866-1900: Leverrier (1861b).
days from 1778. Those of Uranus (at intervals of 10 days) were introduced in 1789
and again from 1791onwards. The adopted precision was 1" for declinatkrtt arrd both
heliocentric and geocentric longitudes and latitudes. When the right as<r:rtsionu'as Jupiter
added, in 1819,a precisionof l' was used.
1834 1877: Bourard (1821).
Heliocentric coordinateswere omitted in 1833, rvhile declination (to l"). geo-
1878 1900: Leverrier (1876a).
centric longitude and latitude (to t"), right ascension(to 0:1). and log clistance(to
5D), rvere all given at intervals of one day.
The intervals were changed in 1834 to one-day intervals for all planets. In 1861, Saturn
when Neptrrne was introduced, the intervals for Uranus and Neptunc werc changed
1834 1879: Bouvard (1821). lFbr the years 1852 1829, Bouvard,s Table 42 was
to 4 days. Right ascension(to 0:01), declination (to 0'll), heliocentric longitude
used in the corrccted form given by Adans (1849) and in The Nautical
and latitude (to 011), and log distance and log radius vector (each to 7D) rvere
Almanac 1851,xiv.]
included. Geocentric longitude and latitude rvere omitted for the year 183'1.A geo-
1880 1900: Leverrier (1876b).
centric (equatorial) ephemeris of Neptune *'as published between 1850 and 1860,
626 EXPLANATORYSUPPLEMENT 13/ HTSTORTCAL
TNFORMATTON 627
uranus ("The Georgian"in The Nautical Almanac 1789 1850) Juno

1789, 1791 1833: Computed from the same tables used for the "classical" planets. 1867-1893: Hind (1855).
1834-1876: Bouvard ( 1821). 1894 1900: Hind (1855), with correctionsby Downing (1890).
1877-1881: Newcomb (1873).
1882-1900: Leverrier (1877a). Vesta

1867 1900: Farley (1856b).

Neptune

1850-1857: Computed from elements given in various issuesof the B erliner Jahrbuch Astrqea
and The Nautical Almanqc.
1858-1870: Korvalski (1855). (This is a little uncertain for the years 1859-1860, 7867-1875: Farley (1856).
as the supplements to the almanacs containing these ephemerides do
not quote the authority.) 13.325 Satellites No predictions for satellites, except for the Galilean satellites of
1871-1881: Newcomb (1873). Jupiter, were published until 1899. Diagrams of the apparent orbits at the time of
1882-1900: Leverrier ( 1877b). opposition and elongations (to 0!l) for a limited period around opposition, were
given for the satellitesof N{ars,Saturn, Uranus, and Nepturre. The authorities for
13.324 Minor Planets Ephemerides of minor planets were given for the first time the latter are the same as that for 1901, and are listed in Scction 13.3{5.
in the almanac for 1834. That issue corrtained ephemerides based on elements by Diagrams of the configurations and predictions of eclipsesfor Jupiter,s satellites
Encke, at intervals of 4 days throughout the year for Ceres, Palles' Juno, and Vesta have been published in every issue ol The Nautical Almanac.
The right ascensionwas tabulated to 0T I . declination and heliocentric longitude and
latitude to 1', and log distance and log radius vector to 4D. For one rnonth on each 1767 18O4: Eclipsesto 1'; bascd on Wargentin (1746).
side of opposition, at intervals of one day, the right ascension was given to 0101, 1805-1823: Based on Lalande (1792) (quoting Delambre).
declination to 011, and log distance and log radius vector to 5D 1824-1833: Based on Delambre (1817).
Similar ephemerides for the years to 1849 were based on the same elements, 1834-1839: Eclipsesto 0il. Other phenomenato 1'; based on Delarnbre (1817).
with va.riations calculated by the method given by Airy (1835). 1840 1900: As in 1834 1839.From 1877,eclipsesto 1.. From 1896,tiures of con-
Between 1850 and 1866, the number of planets for which ephemerides at wider junction to 011, bascdon Damoiseau(1836),and extensionsbv Adams
intervals were published, was increased to as many as 36 in some years The elements and others.
used were from a number of different authors. (Thc almanac lor 1856 contains a
translation (by Airy) of papers by Encke (1852a and 1852b) on the computation of 13.326 Auxiliary Quantities Auxilia.ry quantities includc siclerealtime, nrean obliq-
special perturbations.) uity of the ecliptic, apparent obliquity of the ecliptic, precession, nutation, con-
From 1867, ephemerides for only frve planets were published, and from 1876, stants. and sernidiameters at unit distance.
only the first four planets wcre included, on the grounds that mote accurate ephern-
erides were to be published in the Berlin.er Jahrbuch.
13.3261 Sidereal time The sidcreal timc was not tabulated explicitly for the first
sixty or seventy years, but from 1833, values were given at intervals of onc day. The
Ceres
sidereal time at Dlean noon is stated to have been calculated from the followins
1867 1881: Schubert ( 1854). expresslons:
1882 1900: Godward (1878).
1833: Sun's mean lonsitude + 610 l6'i5 sin O - 0!917 sin2
Pallas 1834 1900: Sun's rnean longitude + nutation, where the Sun's mean longitude
at Paris mean noon of January 0 of the year 1800 + t is given by
1867 1900: Farley (1856a). Bessel (1830a, p. xxiv) as 279"54'01'136+27:605844 i2-
I+O"OOOL22I8O'
628 SUPPLEMENT
EXPLANATOBY 13/ HTSTORTCAL
TNFORMATTON 629
l4'471083/. t'here f is (for the 19th century) the nultrber of 1'carsfrorrr Bctrvccn 1834 and 1853 there is no spe(.ificstatement of the r,ahresusc<I,but from
the preceding leap year. 1854 to 1895 the aDrural (and daily) increments wcre given (to 01000t). aud the
preccssionfrom thc begiuni[g of the year rri,astabulatcrl (to 0l'01) a1 interrals
of
13.3262 Meanobliquity ofthe ecliptic The following valucs wcrc rtsedfor thc 1'ears 10 daYs.No authority is quoted for thcse frgures.FrolD 1896 to 1900.pclers. ralue
1767-1900 (r being measured in years): (18f2. p. 71) is stated to have been rrsedas the alrthoritl..
The following cornparison shorvs thc larious vahros of the arulral precession
1767 18O7 23a28'16"-x(t-1756). Mayer (1770,pp. 105 and v). Nlaskelyncstated that wcre used (7 being tncasuredin <crrturiesfroIn 1850.0):
in several almanacs that.{ was about half a secortd,but Mayer's table
indicates 0]',16.Tlte values seenr to have been adjusted occasiottally 185,11856 (deduced): 50:2357+0';0257.
by \laskelync. 1857 1895 (deduced): 50';2524+ 0:227 T.
1808-1833: Corrccted year by year, from Gleent'ich observations to a current 1896 1900 (Peters): 50';2524+ 0';227T.
date. 1901 1959 (Nen'cornb): 50:2453+O:O222T.
1834-1863: 23"21'54':8- 0:457(t - 1800.0).Bcssel (1830a'p xxvii).
1864-1900: 23'27'31':83- 0:476(t - 1850.0).Levcrrier (1858, p.203). 13.3265 Nutation Values for nutation have appearerl since 1767. Hontvcr-. the terrns
includcd have changerl over the l'ears.
The authorities for the r'alues of the obliquitv adopted for the conlersion of the
Moon's longitudc and latitudc to right ascensionand declination rvere (see A.\{.W. 1767 1833: The "Equation of the EqrrinoctialPoints" (rrrrtationin lorgitude) rvas
D o w n i n g ,M . N . R . A . S . ,6 9 , 6 1 8 , 1 9 0 9 ) : tabrrlatcd (to 0'll) for t:vcrv three rnonths,brrt rvithout anv irrdication
of the authority or of tho terms includcd. In the years 1817arrd 1818.
1862 1874: Hansen (1857, p. ,15iseeHansen and Olufseu. 1853,p. 5). the samc e<|ration in sidereal timc (or nutatiou irr r.iglrt ascension)
1875 1900: Leverrier (1858, p. 203). was gitcn (to 0:01) at intcrr.alsof ten da].s.
183.11856: The ttrbulated valucsin longitude (to 0:'01)and irr right :r.sccnsion (to
Reducing all these to a cornmon date, for comparison with Peters (18'12) and 0101)at intervals of 10 days, r,i'erebasql on Baily's (1825) valucs and
Newcomb (1895a),we have: inclutlcd the four terrns mrmbered 1,2,3, 14 in Table 13.341t.1 (page
654).
NIayer 23"27'29' 0146(t- 1850.0). 1857 18U0: The sarrrt'terrnsq'eretabulated. and tlre precisionand 1h. iDter\al of
Bessel 23"27'31';9501457(/ - 1850.0).
tabulatiorr rvereunchangcd.but the coefficientsrrerc l)asc(lol Peters
Leverrier 2 3 " 2 ' 7 ' 3 1 : 8 30 1 4 7 6 -( / 1 8 5 0 . 0 ) . ( 1 8 1 2) .
Hansen and Olufscn 23'27'31'i42 0146784(r - 1850.0). 1881 1892: Nutation in obliquity wrs also included, to a prccision of 0l'01.
Peters 23' 2't' 30:99 - 014645(t - I 850.0). 1893 1895: Two additional terms (5 and 7) were includr:<lin the talniations.
Newcomb 23'21', -
31:68 01468(r 1850. - 0). 1896: Long-pcriocland short-period terms in both longitude ancl obliquitl'
q'ere iu<lrrded in the tabulations. the intcrral of s'hich *':rs onc da1-.
13.3263 Apparent obliquity ofthe ecliptic Vahresofthe apparentobliquity (to0:'l) Term no. 15 r'as inclucleri.
at intervals of three months were published from the inception of the Almanac; ill 1897 1900: Nine ad(litional terms (6, 8 13, 17, 18) were includecl.
1817 and 1818, and again from 1834, the interval was changed to 10 days, and itt
1834, the precision was changed to 0'101.Frorn 1876 to 1895, thc short-period terms 13.3266 Constants fhe lollowing valrresofthe principal constantsIra\1'beenused:
of nutatiol were included in the aDDarent i?lucs. which q'ere tabulated at irtcrva]s
of one day. Solat' parallax

13.3264 Precession Mayer's value (1770, p. 52) of 5013 was used in the almanac 1834 1869: 8l'5776 Encke (1824, p. 108).
from 1767 to 1796. but was corrccted to 501'2fron 1797 to 1833, with a corre- 1870 1881: 8195 Leverrier (1858, p. 114).
sponding adjustment of N{ayer's values of the mean motions of the Sun and N{oon. 1882 1900: 81848 Newcomb (1467, p. 29).
 13/ HISTOFICAL
INFORMATIoN 631
630 EXPLANATORY
SUPPLEMENT
Jupiter ( equatorial)
Constantof Aberration
1837 1881: 99]704 Struve.
1834-1849: 20136 Baily (1825,p. x).
1882-1895: 98119 Leverrier.
1850-1856: 20142 Baily (1845,p. 21).
Struve (1844,p. 275). 1896-1900: 97136 Schur.
1857-1900: 20!4451

Constantof nutqtion Jupiter (polar)

1834-1856: 9l'25 Baily (1823,p. xiv). 1834: 93l'37.

1857-1900: 9:2231+ 0:0009T, whereZ is in centuriesftom 1800.0(Peters,1842, 1835-1856: 93'j4 Delambre.
D. 'D I.
L857-L887: 92!426 Equatorial SD x 0.927.
1882-1895: 921200 Equatorial SD x 0 939.
1896-1900: 91110 Schur.
13.3267 Semidiametersat unit distance Semidiametersare given for the Sun, Mer-
cury, Venus,Mars, Jupiter (equatorialand polar), Saturn (equatorialand polar),
Ura.nus.and NeDtune. Saturn ( equatorial)

1834: 88l'72.
Sun 1837-1881:811'106 Bessel.
1882-1895: 83131 Leverrier.
7767 laoT: 962i8 Mayer (1770,p. 56).
1808 1833: 1896-1900: 84'175 Meyer.
961"37 Mayer (1770,p. 56).
1834 1852: 960119 Bessel(1830a,p. L).
1853-1895: 9611'82 Airy (1855,p. xxviii). Saturn (polar)
1896-1900: 9611'18 Auwers (from observationsat Greenwich,1851-1883).
1835 1856: 751'25 Bessel.
1896-1900: 9591'63 Auwers (1891, p. 367) (This value was used for eclipse
1857 1881: 75119 Equatorial5Dx.0.927.
calculationsonly.)
1882 1895: 74'i56 Equatorial SD x 0.895.
1896 1900: 6188 Meyer.
Mercury

1834-1863: 3'i23 Lindenau. Uranus

1864-1900:3:'34 Leverrier.
1834: 37':20.
1835 1881: 37'125 Delambre.
Venus
1882 1895: 34]28 Leverrier.
1834-1864: 8l'25 1896 1900: 34i128 Hind.
Delambre.
1865-1895: 8'i305 Leverrier.
1896-1900: 8'i40 Auwers. Neptune

1899-1900: 341'56 Barnardl however,no valuesweretabulated.

Mars

18341 4:37. 13.33 ThoAmerican (1855-1900)

Ephemeris
1835-1865: 41435 Littrow.
1866-1895: 5155 [sic] Leverrier. The twelve morthly sections that formed the principal content of the first part of
189G 1900: 41'68 Hartwig. The American Ephemeris contained the Greenwich ephemerides of the Sun and
oa4 13/ HISTORTCAL
tNFORtiATtON 633
Moon, which remained virtually unchanged during the period 1855 to 1900. The 13.332 Moon In the Greenrvichephemerides of the \,Ioou, the right ascensiou(to
Greenwich ephemeridesof Venus, N{ars, Jupiter, and Saturnl similar ephcmerides of 0:01) and declination (to 0i'l) were tabulated lor everv hour; the s,nicliamerpr and
Mercury, Uranus, and Neptune; and heliocentric ephcmerides for all seven planets horizontal parallax (to 011) rvere given for noon ard midnight; and. frour 1860 to
were added in 1882. An ephemeris of the rectangular coordinates of thc Sun was 1900, thc longitude arrd latitude (to 011) $€re given for noon and miclnight.
also included, although for the period 1875 1881, it was relegated to the second In the Washington ephemcrides, the right ascension (io 0:01) and cleclination
part ol The Amcricon Ephemeris, (to 0'jl) were tabulated for upper and lower culmination frour 1g55 to 1g6,1.ancl for
The second palt of the book contained the ephemeris for the meridian of \\tash- upper culmination fton 1882 to 1900. The sidcreal time of semidiameter passilg
ington, and included further ephemerides of the Srrn, Moon, and planets, partly for the meridian (to 0i01) r.as given for both culminations during the perioci 1855
Washington noon and midnight and partly for meridian transit at \Vashington. 186,1,but onlv for upper culrnination aftcr 1864. The semidiameter tr,nclhorizo.tiil
These ephemerides, which were rcvised someu'hat from time to time, included a paralla-x (to 011) rvere tabulated for Washington firean noon and rnidrright fror.rr
tabulation of the obliquity, precession, and nutation until 1882, rvhcn it was trans- 1855 to 1881,and for upper culmination frorn 1882to 1900.
ferred to the first part of the volume. The times of the phases, apogee, and perigee were given both iu Grcenrvich
meatr time and in \\rashington rnean time. The mean longitude and the loneitude
13.331 Sun In thc Greenwich ephemerides of the Sun. the apparent right asccn- of thc ascending node $,ere given at inter\'als of 10 days.
sion (to 0101),the declination (to 0:l), and the equation ol time (to 0i01) were
tabulated at intervals of one day for apparent noon ald (except in 1855) for mean 1855 1856: Picrce (1853).Thesetables are basedon Airy (1848).rvith conections
noon. The semidiameter (to 0101) and the sidereal tirno of semidiameter passing by Air1. (18a9) and Longstrength (1853). The tables uscd bv Airl.
the meridian (to 0101) u'ere given for apparent noon. Except in 1855, the longi- (18,18)were <ierir,.edfrorn Damoiseau (i824), and wcr-esulrst:r,ntial]va
tude (to 0ll) referred both to the true equinox of date and to the mear equinox devclopment of Plana's theory (1832). modified to irrclucletrvo Venus
of the beginning of the year. The latitude (to 0l'01), and the log radius vector (to
incqualities discoveredby Hansen (1847).
7D) were given for rnean noon. In the Washington ephemerides,the right ascension i857 1882: Picrcc (1853).dcrir-esparalla,xfrom tablesbaseclon formulasof Adarns
and declination rvere given for mean and apparent noon. The equation of time, (1853a) and \\ralker (1848).
the semidiameter, and the sidereal time of semidiameter passing the meridian were 1883 1900: Hansen (1857) contains corrcctionsby Ncwcomb (1878b). lln the 1n-
given for apparent noon. From 1855 to 1881, the longitrrde, latitude, and log radiLrs
troduction, lo The ArrLe,rir:anEphemeris for 1912 and follor.ing ycars.
vector were given for \!-ashington mean noon and midniglrt. The tabular precisions
atteDtion n'as called to the thct that thesc corrections 'i,ere not pre-
were the same as that in the Greenn'ich ephemerides.
cisel.vin accorclance$'ith the statement givcn in the volurnes for 1883
During the period 1855 1881, the equatorial rectarrgular coorclirratesof thc
1911.artd the formula actue ly used rvasgivcn.]
Sun (to 7D) tabulated for Grecnu'ich rrcan noonl referled to the true equinox and
equator of date. For Washingtorr, nrcan noon and midnight, r'eferred both to thc
13,333 Major Planets AIter 1881, in thc Greenrvicir eplxrrrrcridcs of the pianets,
true equilox of date and the mean equinox of the beginning of the vear. In 1882,
the apparent right ascerrsion(to 0:01)and declination (to 011)q,'eretabulatcd for
these ephemerides were replaced by a tabulation (to 7D) for Greenwich mean noon
mean noon at interr,alsof one dav for Nlercury,Ventrs.Nlars, .rrlpiter, and Saturn.
and midnight, which referred both to the tme equinox of date and mean equinox
and ilt intervals of 4 clays for Uranus and Neptune. 'I'he tiure of meritlian passag<:
of the beginning ol the year.
(to 011) rvas also given. The semidiameterzrndhorizontal parallax (in general,1,o
The horizontal paralla-x (to 0l'01) and aberration (to 0101) of the Srrn $'ere 0l'1. and to 0l'01 for Uranus trnd Neptune il 1900,to 0l'01 for N1ercury.Venus.and
tabulated at intervals of 10 days, for 0h Washington sidereal tine frorn 1855 to
Nlars) rvere tabulated at \arious different inten,als for clifTerentpianets.
186.1,for Washirrgtonmean noon from 1865to 1881,and tor Greernich drean noon
In tht: \Virshington epherncridcs, r'alues'"vere tabulatecl for the apparent riglrt
thereafter.
ascension(to 0:01) and declination (to 0']1). for the inferior planets a,t \Vashirrgton
mean noon and nleridian transit, and for the superior planets at Washington sidcreal
1855 1857: Carlini (1810), q'ith Bessel'srevisious (1828). noon and meridian transit betrvccn 1855 and 1869. After 1870. mean noon and
1858 1874: Hansen and Olufsen (18b3). meridian transit tinle'r\,'creused for all the plauets, but after 1882 thc cphemerides
1875 1899: Hansenand Olufsen (1853),with aberration accordingto Struve (1844). for noon u.cre orrrittcd. The selnidiameterand horizoltal parallax (to 0l'01) were
1900: Newcomb (1895a). given for 0h \Vashington sidereal tirne from 1855 to 1864, fbr Washington mean
634 SUPPLEMENT
EXPLANATORY 13 / HISTORICALINFORMATION ooc

noon from 1865 to 1881, and (to 0:'1) for Washington meridian transit beginning Mars
with 1882. The sidereal time of semidiameter (to 0:01) passing the meridian was
1855-1899: Manuscript tables based on Lindenau (1811), with correctionsfrom
tabulated throughout.
Breen (1851) and Leverrier (1841), and various other corrections from
Included in the second part of the volume from 1855 to 1881 were also helio
time to time.
centric ephemerides of the planets. The quantities tabulated were the rectangular
1900: Manuscript tables, based on the elements derived by Newcomb (1895d).
coordinates (to 4D) for Mercury, Venus, Mars, and Neptune; to 5D for Jupiter, Sat-
urn, and Uranus, and the orbital longitude (to 0:1 for the inner planets; to 1" for
Jupiter
the outer planets). The attraction on the Sun was added in 1961. During the period
1855 1860, the rectangular coordinates were referred to the equinox and equator 1 8 5 5 1 8 9 7 : Manuscript tables, prepared from Bouvard (1821), with corrections
of date; beginning with 1861, they were referred to the ecliptic and mean equinox to make them agree with observation.
of a selected epoch, and the coordinates of the Earth were included. A table of the 1898 1900: Hill (1895a). (In the 1898 volume, it is incorrectly stated that Bou-
adopted masses and the orbital inclinations and nodes was also given. vard's tables were used for that year.)
In 1882, the heliocentric ephemerides were replaced by different ones and fol-
lowed immediately after the geocentric ephemerides in Part I. These heliocentric
Saturn
ephemerides contained the longitude and latitude (to O'il) referred to thc ecliptic
and mean equinox of date. AIso included were the reduction to orbit, log radius 1855 1882: Manuscript tables prepared from Bouvard (1821), r'ith various cor-
vector (to 7D); in 1900, to 8D for Venus and Mars, and log geocentric distance rections from time to time.
(to 7D), for Greenwich mean noon at intervals of 8 days for Uranus and Neptune, 1883 1899: N{anuscript tables prepared from a provisional theory by Hill (1890).
2 days for Mercury (one day in 1900), and 4 days for the other planets (2 days 1900: Hill (18e5b).
for Venus and Mars in 1900). Log geocentric distance was also given for the dates
between the tabular dates. Uranus
During the years preceding the completion of the planetary tables of Newcomb
and Hill, the ephemerides were calculated with tables that, for the most part, were t6DD 16/D: Bouvard (1821),with revisionsby Leverrier (1846)and Peirce(1848a),
constructed by applying corrections to the early tables of Lindenau and Bouvard and beginning with 1859,further correctionsby Runkle (1855).
that were based upon Laplace's theories. 1876: Manuscript tables constructed bv Newcomb.
1877 1900: Newcomb (1873). (An ephemerisfor 1873 1876calculatedwith these
tables is given in the 1877 volume.)
Mercury
Neptune
1855 1899: Winlock (1864), based on the theory of Leverrier (1845).
1900: Newcomb ( 1895b). 1855 1869: Tablesbasedon Peirce'stheory (1848b)and Walker's elements(1848)
(Ephemerides for 1853 and 1854 were given in the volumes for 1855
and 1856, respectively.)
1870 1900: Newcomb (1865). (An ephemeris for 1866-1869 calculated from these
Venus tables is given in the Appendix to the 1869 volume.)

1 8 5 51 8 7 5 : Manuscript tables prepared from Lindenau (1810) by applying cor- 13.334 Minor Planets Among the early tables constructed and printed for Zl'e
rections ba-sedon investigationsby Airy (1832), Breen (1848), and American Ephemeris were tables ofthe minor plalets ( 15) Eunomia, (40) Harmonia'
Leverrier (1841). (18) Melpomene, and (11) Parthenope. Ephemerides of these minor planets were
1876 1899: Hill (1872). (An ephemeris for I874-1875 calculated from these tables not included 1t The American Ephemeris, but an "Asteroid Supplement" in the
is given in the Appendix to the 1876 volume.) volume for 1861 contained opposition ephemerides for 33 minor pla.nets for 1859,
1900: Newcomb (1895c). and the orbital elements of (1)-(55).
636 EXPLANAToRY
SUPPLEMENT 13 / HTSTORTCAL
TNFORMATTON 637
13.335 Satellites Irr the volumes for 1855 1881, ephernerideswere givcn only for 1882 1899: 23"27'31'.'42-0':467841, rvhere r is reckorred irr years frorn 1850 (IJalscrr
the four great satellites of Jupiter, and the apparcnt elements of thc rings of Saturn. zrrrdOlufsen. 1853).
The ephemeridesof tlrc satcllites of Jupiter gave the superior geocentricconjunc- 1900: 23'27'08':26- 01468t, q.here 1 is reckonedin 1.earsfLom 1900 (Ncr,'-
tions, the phenomerrtr,the coordinates irr thc nrean apparent eclipses.arrcldiargrams conrb. 1895d).
of the phases of the eclipses.
In 1882, diagralr.rsof the configurationsof the four satellites of.Iupiter l'ere The apparent obliquitl. (to 0'l0l) sas tabulated at iDtcr\.alsof l0 dar.s(5 da1.s
added, and the forurcr cphemerides of the coorclinates in the appareut orbits rvere in 1900), for 0h \lhshington siderealtirne frorn 1855to 186.1.for \\hshington nrcarr
omitted. A diagram of the apparent orbits was also added. Ephemerides of the noon froln 1865 to 1881, and for Grcenrvichluean noon then'zrfte,.
elongations of the satellitcs of t\lars, Saturn, Uramrs, and Neptune, and diagrams The oblirlrity actually used in ctr,lculating the ephemerirlcs ol the Sun. [4oorr,
of the apparent orbits, were introduced. and planets irr the volunes for 1865 1899 was taken frorrr Ilansen arrd Olufscrr
The elongations ol the fifth satellite of .lupiter were added in 1898, but no (f853). hr 1900,it rvas taken fronr Neu'comb (1895a).
statement of authority was given from 1898 to i900. No authority rvas given for the
satellitesof \Iars frorn 1882 to 1900.The authorities for the othcr satellitesq.ere: 13.3363 Precession For the geleral precessionin longitrrrlc.tlrc cxpressiorrgiltrr
bv Peters (1812) 50:'2411 + 0'10002268 /. rvherer is reckorrcdin.\.carsfrom 1800
Jupiter was used until 1900. rvhen the value given by Neq,comb (1895d) rvas adopted
50:2482 + 0'.'000221,where / is re<rkorredin years from 1900.
1855 1881: Damoiseau(1836). Extended to 1880 by Kendall (1877). The aruountof precession(to 0:'01) sincethe beginningof thc yr:arwastabulattrl
1882 1900: For elongationsand eclipses,To<1d's(1876) continuation of Damoi- at intervals of 10 days (5 days in 1900) for 0h \l'ashington siderealtinre tiom 1855
seau;for occultations,transits, etc., Woolhouse(1833), r'ith Table II 1864,for \Vashingtonnlean noon cltrring1865 1881,anrl f<rrCreeul'ich ulcall no(Jn
for each satellite adapted to Damoiseau. thereaftcr.

Saturn 13.3364 Nutation From 1855 to 1899. thc terrrr 'equation of tlrt' crluinoxes il krn,
gitude" r'ias uscd for the mrtation in longitude. The Dutatiorrsill obliquitl'used for'
1855 1900: For rings, cxcept the dusky ring, Bessel(1875aand 1875b). the computations relating to the stars wcre calculated frorrr the forrnulas givcrr b1
1882 1900: For satellitcs, manuscript tables prepared by Newcomb. Peters (1842). The lormulas of Peters wcrc given in the appordix to the volurrrc 1or'
1855; during latcr 1'ears, provision was rrra,<lefor including trtlditional surall ternrs
Uranus and Neptune when requirt:<I.In the volume for' 1900.thc nutations of obli<yuitvu'erc taken frorrr
Newcomb (1895a).
1882 1900: Newcomb (1875). Fron 186j-) to 1891.the apparentobli<pit1'and mrtation rrst'dil the ephcntericlt's
ofthe Sun. Iloon. and planetstere takcn frorn Hansenanrl OlLrfsctr(1853).In f90{).
13.336 AuxiliaryQuantiti€s the obliqrrity and the long-period rlrtati()n \\'eretakeu lrorrr Ncu,r:ornb(1895n).
The eqrraiionof the equinoxcsin krrrgiturie(to 0:'01)alxl irr right ascensiou(t,o
13,3361 Sidereal time The sidereal tirne (to 0:01) was tabulated for every Wash- 0l0l) for 1855 1877, and to 0i001 for 1878 1900 were tabrrlllcd at intervals ol l{)
ington mean noon, and (cxcept in 1855) for every Greenwich mean noon. In the days (5 davs irr 1900)lor 0h.Washirrgtonsiderealtime in 1855 1864,for rfthshingtorr
Greenwich ephemeris,except in 1855, the mean of0h sidereal tirne (to 0l0l) r.as mean noorr irr 1865 1881, and lor Greenwich mean noon thcreaftcr. In 1900. thc
also given for el'ery da1'. nutation in oblirpitv (to 0:'01)rvas also explicitll' tabulatcd at intervals of 5 clal's.

13.3362 Obliquity ofthe ecliptic 13.3365 Constants The values ancl ar.rthoritiesused for solar parallax, constant o[
The adopted cxpressions and authorities for the
mean obliquity were: aberration, and constant of nutation ilro as follorvs:

Solar parnlLat
1855 1881: 23'27' 54i22- 0':4645t- 0:0000014t2, where tis reckoned irr years from
1800 (Peters, 1842) 1855 1869: 8:'5776 Encke (182.1).

t
 EXPLANATORYSUPPLEI,,IENT
1 3 / H I S T O B I C A ILN F O R M A T I O N 639
638
Newcomb (1867). The value for N,Iercury was taken from Leverrier; the othcrs were determined
1870-1899: 8'1848
8:80 Paris (1896). (This value was used for eclipse calculations by Peirce from observations in 1845 and 1846 with the mural circle at Washington.
1900:
from 1896 onward.) The values for Jupiter and Saturn *'ere the polar semidiarneters. It was statecl in
the 1869 r.olume thai 19'119was erroneously used in 1858 1869 for Jupiter in the
Washington ephemeris. In the volumes for 1869 1900, the equatorial semidiameter
Constant of aberration
of Jupiter was given as 201'00at Iog distance 0,70, and that of Saturn as 9l'38 at log
1855-1899: 2014451 Struve (1844). (In the ephemeris of the Sun for 1869 distance 0.95, without authority.
1874, the value 201255from Hansen and Olufsen (1853)
was used.)
1900: 20:47 Paris (1896). 13.34 The Nautical Almanac, and The American Ephemeris (1901-1983')

Constant of nutation The details given in this section and The Astronomical Almanac for 1901 to 1959
refer specifically to The Nautical Almanoc. Differences in the corresponding editions
1855 1899: 9:'2231+ 0:'0009f, where Zis in centuriesfrom 1800.0.Peters (1842) of The American Ephemeris are speciallv noted in brackets in q'hich "A.E " refers
1900: 9:'21at 1900.0 Paris (1896). o n l y t o f i r e A m e r t c o nE p h e m e r , s .
From 1960 through 1980, The Astronomical Ephemeris and Nautit:al Almanac
13,3366 Semidiamet€rs The values and authorities for the Sun, Moon, and plan- and The American Ephemeris and Nautical Almanoc wele completely in agreement,
ets are: scientifically. Beginning in 1981, TDe Astronomical Almonttc replaccd those two
publications and was printed in a single printing. For the edition for 1968' the IAU
System of Astronomical Constants t'as introduced The supplement to the 1968
Sun
publications gives the details of the changes to the ephemerides as a resuit of this
1855-1899: 16'02" at mean distance (Greenwichobservations). new value of constants.
1883 1899: In the calculation of eclipses,15'59178(Bessel1830a)-
1900: 15'59163at mean distance (Auwers 1891). This value was used in
calculating eclipses, but in the ephemeris of the Sun, li15 rvas added 13.341 Sun The quantities tabulated, generally \r-ith argument mean trme,
to it for irradiation. a,re substantially the same as before: iongitude, latitude, right ascension, decli-
nation, and log radius vector; the precision,normally: 0'J1,0101,0i01, and 7D; and
(except for the mean longitude and anomaly and the equatorial rectangular coor
Moon
dinates) the interval one day. The mean longitude (to 0?00001 for 1906 1911, a:Ld
1855 1868: Burkhardt's (1812) value increasedby 1/500 part. to 090001 since 1912) was given at intervals of 10 days since 1906 [In A E , the
1869 1900: 0.272274 - 215 for irradiation; the irradiatiou was omitted in the mean longitude (io 0:0001), at intervals of 5 days, rvas given silce 1934.] Equato-
calculation of eclipses and occultations. rial rectangular coordinates (to 7D) was referred to the tme equator and t'quinox
of date. with reductions to those of the beginning of the year, were givon at rnter-
Planets vals of 12nand until 1930, while similar coorciinates at intervals of one day for the
log distance equinox of the beginning of the year were tabulated since 1931, and those for the
1855 1900: Mercury 3':34 0.00 standard equinox of 1950.0 (Comrie, 1926) since 1928. (The coordinateslor 1928
Venus 8:',546 0.00 were given in the volume for 1929.) [In A.E., the interval of 12hwas retained uutil
Mars 2:842 0.25 1950: from 1931 to 1950 the coordinates were referred to the beginning of the year'
Jupiter 18178 0.70 and the reductions to the irue equinox of date were also tabulated. Coordinates for
Saturn 8:77 0.95 1950.0 were tabulated since 1938, at intervais of l2h until 1950, and intervals of one
Uranus 1168 1.30 day from 1951.]Since 1954 lin A.E., since 1953]the coordinateson the "standard"
1882-1900: Neptune t':28 1.48 10-day dates (IAU, 1950) were emphasized by the use of bold type.
 EXPLANATORYSUPPLEMENT 13 / HISTORICALINFORMATION
640 641
Longitude, referred to the mean equinox of the beginning of the year. u'as the interral for Uranus and Neptune rvas '1 davs until 1935.] Log distalce at the
included from 1931 to 1937 lin A.E., from 1901 to 1915 and from 1931 to 1951]. satne interral and to 7D q,as given until 1934 for 1!lercur1'and ultil 19.1Ufor thc
Longitude and latitude, which referredto the mean equinox (to 0300001)of 1950.0, othcr planets. [ln A.E., the intervals until 1915 wcre l2h for Mercury, ono day lor
were tabulated at intervals of one day from 1928 to 1937 lnot in A.E ]; from 1938 Venus and NIars, 2 days for Jupiter and Saturn and, from 1916, one day for cach of
to 1959 [also in A.E.] longitude (to 0':l), latitude (to 0'l0l) were given. lln A E., thesc fir'e; for Uranus and Neptune, it \a'as,1dal s fron 1901 to 1935.] \atur.al values
Iatitude referred to ecliptic of date given during 1901 1959.] Natural values of the of the distance (to 6D for \{ercurS-to Jupiter. and to 5D Saturn to \cpturre) rvere
radius vector (to 7D) were introduced in 1928 [in A.E., in 1938],the logarithmic given since 1935 for N,Iercluy and siur:c 1941 for the other six planets at illcrvals
value being omitted froln both almanacs in 1983. of one day.
From 1960 to 1980, the longitude of the Sun ll'as given rvith respect to the mean Heliocentlic longitude and latitude, (to 011). rcferred to the rnean eqrrinoxof
equinox of the beginning of the year. Starting in 1981, the ccliptic longitude was datc, and log raclius vcctor (to 7D) r'er.e tabulatecl at inter\,als of one to four dar.s
given for the mean eqrrinox of date. The geocentric rectangular coordinates of the until about 1915, and thcreafter thet' u'ere givcn, in the appurdices to various
Sun were omitted in 1981 but reinstated from 1982 onward, and given with respect almanacsbetween 1915 and 1920,at wider intervals (up to 40 dn1.s)for thc pcriod
to the mean equator and equinox of 1950.0through 1983. up to 19,10[in A.E.. at irrtervalsof oue to ten tlar.s throughout thc periorl 1901
19591.Tlre first two volurnclt of Planetary Co ordinates(H.II.\.A.O., 1933 an<l
1901-1980: Newcombtables (1895a). 1939) contaiuedsimilar coordinates(to 0?001) or 010001, but roferredto the neari
1981-1983: Newcombtheory (1895a). equirroxof 1950.0,and rratural values of the radius vector (to 4D or 5D) for the
period 1920 to 1960.except that no tabulatiorrs are given for \Ierr.url.and pluto.
13.342 Moon Values of the longitude and latitude (to 0:1), of the parallax (to The rolumcs also containeclcoordinatesof the four outer plalcts .lupitt:r. Saturn,
0l0l) [in A.E., (to 0:l) until 1912],and ofthe sernidiameter(to 0:'01) lin A.E., until Urarrus,and Neptune for 1800 to 1920.
19391were tabulated at intervals of l2h. Values of the right ascensionand tleclination Astrometric right ascetrsions,declinations, an.l distancesof Pluto havc been
(to trO1 and 0'11)were given at intervals of one hour. rvhilc lunar distances continued included since 1950,but no heliocentric.coordinatesof this planet have be<,rrgir.en.
to appear until 1906 [in A.E., until 1911],and examplesof their calculation u'crc lln A.E., heliocentricloDgitude and latitude (to 0il) and log rzr.diusfor thc periocl
given in the issuesfor' 1907 1919 lirr A.E., for 1912 1935].Elements of the mean 1960 1983, the riglrt ascensions(to 0:001) and declinations (to 0l'01) of tho outer
equator and orbit have been included througtrout. [The angrrlar distancc from thc planets were tabulated.]
Sun (to 0:1) was given irr A.E. from 1937to 1941,and to 1:0, thereafter'.]
In 1981, the hourly ephemeris of the Moon was replaced by a daily ephemeris Mer<:urg and, Ve.nus
in polynomial fornr for irrterpolation. The apparent longitudc arrd latitrrde (each to
0301) of the Moon were given only daily from 1981 onward. 1901 1980: Ncq'comb Tables (1895b altd 1895c).
1981 1983: \encomb Thcor-ics(tXSsb arrd 1895c).
1901-1922: Hansen(1857)$'ith \ervcomb's corrections(1878b)to right ascensiox.
declinatiou, parallax, and semidiameter from 1901 to 1914, and tt' Mars
longitude and latitude (before corrversion to cquatorial coordinates)
from 1915 to 1922. lA.E. for 1912 states that Nervcomb'sfornrula for 1901: Lcverrier (1861b). [In A.E.. tables iD Dranus(ripl (]onstnrctc(lfroDr
the correction to Hansen's mean longitude was not, in fact, used for' clctnt'rtts b1' Nervcomb ( 1895d).]
the years 1883 to 1911.The formula actually used was (lu')ted] l9O2: Lr:vcrrier (til6lb). [In A.E., Newcomb (1898).]
1923 1959: Brown (1919). It is to be noted that the valrrcsof the lorrgitude(antl 1903 1921: Ncr','comb(1898a).
correspondingly of the right ascension and dc<:lination) fol the )ear 1922 1980: \cu'cornb Tablcs (1898a) rvith Rosa's cor.rectious(1g17).
1923 requirc a small correction of +0108cos(r l'). 1981 1983: New<on.rbthcory (1898a) rvith Rosa's corr.ectiorrs(1917).
1959 1983: Brown's theory as givcn in the ImprouedLurtar Ephent,t is (1954)
Jupitcr arLdSaturrt
13.343 Major Planets Values of the appareut right zLscensionand <lcclination to
0?01and 0'11were tabulated at intervals of one <lay throughout the period. [In A E., l90l 1959: Hill (18954 and 1895b).
642 EXPLANATORYSUPPLEMENT 13/ HTSTORTCAL
TNFOR|\4ATION 643
Uranus Vesta

1901-1903: Leverrier (1877a). [In A.E., Newcomb (1865).] 1901 1906: Farley (1856b).
1904-1959: Newcomb (1898b). 1907-1915: Leveaau(1896).
1901-1902: Leverrier (1877b). [In A.E., Newcomb (i865).]
1903: Leverrier (1877b). [In A.E., Newcomb (1898c).] Ceres, Palkts, Juno, arul, Vesta
1904-1959: Newcomb(1898c).
1952 1959: Herget, Clemence,and Hertz (1950).
Pluto 1960 1971: Herget (1962).
7972 7983: Duncombe (1969).
1950 1959: Bower (1931).
13.345 Satellites Satellite information is given for Mars, Jupiter, Saturn, Uranus,
Jupi.ter, Saturn, Uranus, Neptune, and, Pluto Neptune, and Pluto.

1960 1983: Eckert, Brouwer and Clemence (1951). Geocentric ephemerides only
include perturbations by inner planets (Clemence,195.1). 13.3451 Mars Diagrams of the apparent orbits and times of elongations to a pre-
cision of 0?l, were given for a period of a month on each side of opposition from
1931 to 1980.From 1931 [in A.E. from 1920],tables were added for calculating the
13.344 Minor Planets Elements and ephemerides of the first four minor planets
position angle and apparent distance of each satellite from the planet at anv time
were published as in previous years lnot in A.E.] from 1901 to 1913; ephemeridcs
during the same period. From 1981, times of each eastern elongation for Deimos
alone were given in 1914 and 1915. Thereafter, no tabulations were given until
and every third eastern elongation for Phobos were given for the entire year. Tables
1952. Since that year ephemerides have been inclrrded to cover, at intervals of one
for calculating the position angle and apparent distance were also extended to cover
day, the periods during rvhich transit occnrs bet$'een sunset and sunrise at (most)
fixed observatories. Although in 1958, the period of tabulation was altered to thir.t the entire year.
period of time during which the plarct is "not within about 40' of the Sun." The
quantities given were apparent right ascension and declination (to 0:01 and 0:'1) 1901 1902: Elements by Hall (1878). [In A.E., those of Harshman.]
with corrections "astrometric minus apparent." and distance (to 6D). 1903 1915: Elements by Harshman (1894).
Starting in 1981astrometric positions (to 0i1) and declination (to 110),respec- 1916 1983: Elements by Struve (1911, p. 1073).
tively, were tabulated at two-day interr,als for Ceres, Pallas, Juno. and Vesta. In
addition, opposition dates, magnitudes, and osculating elements were given for the 13.3452 Jupiter Phenomena and configurations of Satellites I IV are given from
brighter minor planets coming to opposition during the year. 1901, with considerable variation from time to time in the precision, which is usually
The basis of the tabulation has been: higher for eclipses than for other phenomena. Elongations every twentieth on each
side of Satellite V were given in 1906 1980 fin A.E. 1901-i980]. From 1981,times of
every twentieth eastern elongation have been given for the whole year. Differential
Ceres
coordinatesof SatellitesVI and VII have been given since 1931 [in A.E. since 1912].
1901 1915: Godward (1878). Sidereal periods of Satellites VIII to XI, and XII have been included since 1953 and
1957. The sidereal period of Satellite XIII wm introduced in 1984. Differential
Pallas coordinates of Satellites VIII to XIII were introduced in 1981.

1901 1915: Farley (1856a). SatellitesI IV

Juno 1901 1913: Damoiseau (1836) and later extensions. [In A.E., also Woolhouse
(1833).1
1901 1915: Hind (1855), with correctionsby Downing (1890). 1914-1915: Sampson(1910). [In A.E., Damoiseauand extensionsand Woolhouse.]
644 EXPLANATORY
SUPPLEMENT 13 / HISTOFICALINFOBMATION o+J

1916: Sampson(1910), with Andoyer's (1915) modifications. [In A.E. until (1898). [In A.E.; 1901 1903,Hall (1886); 1904,Hall, except the el6n-
1930, the configuration are attributed to Pottier's (1896) continuation gations of SatellitesI and III (from Struve, 1898); 1905-1913,Struve
of Damoiseau.l (1898); 19i 1915,Struve (1888, 1898,and 1903).]
1 9 1 6 - 1 9 3 0 :Struve (1888, 1898, and unpublished corrections).
Satellite V 1 9 3 1 - 1 9 3 5 :H. Struve (1888, 1898,and 1903) and G. Struve (192,1).
1936 1983: G. Struve f1924 and 1930).
1901-1905: [In A.E. only, Robertson'selements(unpublished).]
1906-1915: Chon (1897). [In A.E., Robertson.] SatelLite VI (Titan)
1916-1959: Robertson(1924). [In A.E.; 1930-1933,"from ConnaissancedesTemps
for 1915"; 1934 1959, "from Connaissance des Temps every year be- 1901 1915: Newcomb'smanuscripttables. [In A.E.: 1901 1904,Hall (1886); 1905
'.]
g i n n i n gw i t h 1 9 1 9 1913, Struves (1898); 191a 1915, Struve (1888, 1898,and 1903).]
1960-1983: van Woerkom (1950). 1916-1930: Struve (as for SatellitesI V).
1936 1937: G. Struve (1933). [In A.E., H. Struve, as in 1931.]
Satellites VI V 1938 1983: G. Struve (1933).

1912-1930: [In A.E. only, Ross (1907aand 1907b).] Satellite VI I (Hsperion)

1931-1947: Ross (1907aand 1907b).
1948-1983: Bobone (1937a and 1937b). 1901 1915: Newcomb'smanuscripttables. [In A.E.: 1901-1902,Eichelberger(1892);
1903 1913, Struve (1898); 1914-1915,Struve (as for Satellite VI).1
1916 1935: Struve (as for Satellite VI).
Satellites VIII X
1936 1983: \4/oltjer (1928).
1981 1983: Herget (1968).
SateLlite VIII (Iapetus)
Satellite XIII
1901-1902: As for Satellite VII. [In A.E., Hall (1885).]
1981 1983: Aksnes (1978). 1903 1935: As for Satellite VII
1936-1937: H. Struve (1888, 1898.and 1903).
1938 1983: G. Struve (1933).
13.3453 Saturn Until 1980, diagrams were given for part of each year of the orbits
of Satellites I VII. The volume also included times of elongations and conjunctions
of Satellites I-VIII to a precision of 0ll. Beginning in 1981, the tables cover the Satellite IX (Phoebe)
entire year. Difierential coordinates of Satellite IX (Phoebe) were introduced in
1909 1930: [in A.E. only, Ross (1905).]
1931 [in A.E. in 1909],as well as those of SatellitesVII and VIII lin A.E. in 1960].
1931 1983: Ross (1905).
Elements for determining the distance and position angle were added in 1931 [in
A.E. in 1912]. During 1935 1959, The Nautical Almanac [not A.E.] contained, for
13,3454 Uranus Diagrams of the orbits of Satellites I-IV, and times ofelongations
about nine months in each opposition, quantities to assist in the calculation of the
phenomena (eclipses, occulations, transits, and shadow transits) of Satellites I VI. to a precision of 0ll were given for about nine months in each opposition in 1901
1980. Elements for the calculation of distance and position angle were included in
1927 [in A.E., in 1912]. F]om 1981, the times of elongations and elements for cal-
Satellites I V: Mimas, Encelad,us, Tethys, Dione, and Rhea culating the distance and position angle have been given for the entire year. The
sidereal period of Satellite V (Miranda) was added in 1953. In 1981, elements for
1901 1915: Tables in manuscript "prepared by" Newcomb, except tlte elongations the rings of Uranus were introduced, as well as times of eastern elongation and
of Satellites I (Mimas) and III (Tethys), which are from H. Struve tables for the calculation of the distance and position angle of Satellite V.
 13 / HTSTORTCAL
TNFORMATTON 647
SUPPLEMENT
EXPLANATORY
646
13.346 Auxiliary QuantitiesAuxiliary quantities given are sidereal time, obliq-
SatellitesI-lV uity of the ecliptic, precession,nutation (for both The Nautical Almanac and The
1901-1915: Newcomb(1875). American Ephemeris),and constants.

Satellites I and II 13.3461 Sidereal time The entries for sidereal time include:

1916 1983: Newcomb.

1901 1932: The tabulated values (to 001) at intervals of one day were based on
Newcomb's value (1895a) for the right ascension of the mean Sun af-
Satellites III and IV fected by aberration- 18h38'45i836 +8640184?5427+Oi092922. where
T was measured in Julian centuries from 1900 January 0 at l2h UT,
1913-1983: Struve(1913).
and included the effect of long-period terms only of nutation.
1933 1959: The precision was changed to 0?001, and the effect of short-period
R'ings terms of nutation was included also.
1960 1983: Tabulated separately for both apparent and mean equinox- Universal
1981 1983: Elliot el ol. (1978).
Time of transit of frrst point of Aries also tabulated.

Satellite V
13.3462 Obliquity ofthe ecliptic Newcomb's value (1895a) for the mean obliquity
l98l 1983: Dunham('|971). was used througho:ut 23'27'O8126- 461845I - 0:005912 + 0100l8l 73. [In addition,
A.E. includes values attributed to Hansen and to Peters for the years 1901 to 1915,
13.3455 Neptune Diagrams of the orbit of Triton, and times of elongations to a also those due to Leverrier for the years 1902 to 1915.]
precision of 0ll, were given for about ten months in each opposition in 1901 1980' Values of the true obliquity (to 0'101)excluding the effect of short-period terms
Elements for the calculation of distance and position angle were included in 1927 of nutation were given at intervals of one day until 1930 [in A.E., until 1933; and at
calculation of intervals of 5 days from 1934 to 1959]. From 1931 to 1959, only the mean obliquity
lin A.E., in 1912].From 1981, times of elongation and tables for the
distance and position angle covered the whole year. The sidereal period of Nereid for the beginning of the year and the daily nutation in obliquity were given. From
was added in 1953. Times of eastern elongation, and tables for the calculation of 1960 to 1980, the obliquity of the ecliptic was given daily with the Solar ephemeris
distance and position angle of Nereid were added in 1981. and after that it was given with the Besselian day numbers.

T\'iton 13.3463 Precession The values given were based, throughout the period, on New-
comb's determination (1897, p. 73) 5012564+ 0'.'0222 T [A.E. for the years 1901
1901: Newcomb (1875). [In A.E., Hall (1898).] to 1911 also gave values based on Peters (1842).]
1902 1929: Hall (1898). Values ofthe precession (to 0'l0l) ftom the beginning ofthe year were tabulated
1930 1983: Eichelbergerand Newton(1926). at intervals of one day [in A.E., 5 days until 1915], while the daily and annual
increments (to 0'j0001)were given until 1930 [in A.E., until 1915].
Nereid From 1931 fnot in A.E.] additional prccessional constants, and tables for reduc-
tion of star positions, were included also.]
1981 1982: Rose(1974).
1983: Mignard(1981).
13.3464 Nutation: The Nautical Almanac "Ihe terms used in The Nautical Alma-
13.3456 Pluto In 1981,times of northern elongationwereaddedfor Charon' nac from 1901 to 1959 were:

1901-1902: The tabulations of long-period and short-period terms (separately)

1981: Christy and Harrington (1978).
in both longitude and obliquity were based on the values given by
1982-1983: Harrington and Christy (1981).
 SUPPLEMENT
EXPLANATORY 13 / HISTORICALINFORMATION
648 649
Newcomb (1895a), as modified by the Paris Conference(1896). Six Solar parallax
long-period terms, and seven short-period, were included, as shown
in Table 13.348.1.The precisionwas 0101. 1901 1968a: 8]80 Paris(1896,p. 5.1).
1968 1983: 8'179405.
1 9 0 3 - 1 9 1 r : The coefficients of a few of the terms were modified in accordance
with the revised values given by Newcomb (1898d).
19121930: An additional term, (8), was included in the day numbem throughout Constant of aberration
this period, but only in the tabulation of the nutation from 1918.
1931-1936: Nutation in obliquity and the short-pe od nutation in longitude rvere 1901 1968: 20'j,17 Paris (1896, p. 5.1).
no longer tabulated explicitly, though the former was available as -B 1968-1983: 20]4958.
(the day number). Nutation in right ascension (to 0:001), including
short-period terms, was tabulateo. Constant of nutation
1937 1959: Tabulations were the same as in former yearsJbut 21 terms (1 4, 6
10. 14 25) rvere included. The coefficients used $'ere those given by' 1901-1936: 9121 Paris (1896, p. 54).
Newcomb (1898d) and are quoted in Tablc 13.348.1. 1937-1983: 9,i210+0:'00091' Newcomb (1898d,p. 2a1).

The terms uscd in ?he American lA.E. for the years 1901 1911 included tables for the Struve arrrl peters con-
13,3465 Nutation i The American Ephemeris
stants of aberration and nutation as used formerly. as well as those for the above
Ephemeris from 1901 to 1983 were:
Paris ralues.]
An IAU System of Astronornical Constants was adopted in 196,1arrd introduced
1901 1911: The tabulations of long-period terms (five-day intervals) and of short-
in the 1968 Astrononttcal Eph.emeris and, American Ephemet-is.The compk:te list
period terms (one-day intervals) were ba^sedon both the values of
of constants is given in Sectiorr13.35.
Peters (1842) and those of the Paris Conference(1896). The terms
included were 1 3, 5, 6, and 8 15. From 1901to 1911,the argunlert
Sun's true Iongitude was used instead of Sun's mcan longitude (L) in 13.347 Ephemeris for Physical Observations This section covers ephemerides for
the Peterscalculationfor terms 3, 5, 8, 6, 10, and 11, and from 1901to the Sun. Moon. Mercury, Venus,Mars. Jupiter, and Saturn. Irr addition, semicliam-
1907 in the Paris calculations for terms 8, 10, and 11. The coefficients eters are given for the years 1901 1983.
only difier in a very few casesfrom those shou'n in Table 13 3'18.1,attd
the individual discrepancies (usually of a single unit) are rrot listed.
1912-1936: The tabulations were basecl solelt' on the values of the Paris Corr 13,3471 Sun An cplrcrnerisfor physicalobserrationsof the Sun was first published
ference, and after 1915 were all at an intr:rval of one day; the terms in The NauticaL ALmartac for the Year' 1907. Previous to lhat tirne, observations
included were 1 4, 6 8, 14, 15, and 17 21. were reducedu'ith thc aid of tables priratell' printed b1'\Varren de la Ilue. (Scethe
1937 1959: Tabulation of long-period nutation in longitude, arrd short-period nu volumes of tlre Greenrvir:h photo heliographic observation.) Thc Ameri.mn Ephem-
tation in both longitude and obliquity, were givcll at intcrvals of onc eris first included obscrvationsof the Sun in 1913.
day, and included the terms I '1, 6 10, and 1'1 25. The coefficicrts
are those listed in Table 13.348.1.
1960 1983: The formulas adopted for computing nutation in lolgitudc and obliq- 13.3472 Moon Ephemeris for physical observatiorts ol thc Moon, calcula,tcd by
uity were obtained by retaining all terms for the coemcierlts as grcat NIartlr. appeared in Monthly Notices of the Royal Society d,uring the last quarter of
as 010002from the expression developed in trVoolard (1953). l,lre nirreteenth century. 'Ihe ephemeris was introdrrced ilotoTh.t:Nautical Alrn,anac
in 1907, and rvas first irrcluded 1t TIrc American Ephemeris in 1913. But the for-
13.3466 Constants The followine consta[ts wcrc used for the period 1901 1983' mulas and tables for the calculation of the optical librations, and the times of the
greatest librations, had been included with the ephemeris of the elements of the
mean equator of the Moon, since 1855.
 EXPLANATORY
SUPPLEMENT 13/ HISTORICAL
INFORMATION 651
650
Jupiter (polar)
13.3473 Mercury and Venus The physical ephemeridesof Mercury and Venus were
added to The American Ephemeris in 1882, and included in The Nautical Almanac
1901-1920:911'10 Schur(1896).
in 1907. Previously, only a small table of the versed sine of the illuminated disk
1921 1959: 911'91 Sampson(1910).
divided by the apparent diameter, for Venus and for Mercury, had been given.

Satum ( equatorial)
13.3474 Mars and Jupiter Ephemerides for physical observations of Mars, calcu-
lated by Marth, appeared in Monthly Notices of the Royal Astronomical Society 1901 1920: 84'i75 Meyer(1883).
beginning in 1869, and for Jupiter beginning in 1875. They were continued by 1921-1959:83'133 Struve(1898).
Crommelin after Marth's death, and transferred Lo The Nautical Almanac in 19o7.
They were first included in The American Ephemeris in 1913.
Saturn (polar)
For 1968, the ephemeris for "Physical Observations of Mars" was introduced
based on the IAU System of Astronomical Constants.
1901 1920: 76188 Meyer(1883).
7927 7959: 74!57 Struve (1898).
13.375 Saturn The physical ephemeris ofSaturn first appeared in the ephemeris
in 1960. The stellar magnitude that is included had previously been given in the
Uranus
ephemeris for the rings.
1901 1930: 34]28 Hind.
13.3476 Semidiameters at unit distancet The NsuticalAlnttnac The values given in 1931 1959: 34128 Barnard(1896);See(1902);and Wirtz (1912).
The Nautical Almanac tor the period 1901 1959 are:
Neptune
Sun
1901 1959: 36'156 Barnard (1902).
1901 1959: 961118 Auwers(1891).
1901 1959: 959163 Auwers (1891,p. 367); for eclipsesonly.
13.3477 Semidiameters at unitdistanceiThe American Ephemeris The values given
in The American Ephemerisfor the period 1901 1959are:
Mercury

1901 1959: 3'134 Leverrier (1843). Sun

Venus 1901 1902: 960'178 Auwers (1891); includes 1'j15for irradiation.

1903 1959: 9 6 1 t 5 0 Harkness(1899); includes 1l'15for irradiation.
1901 1920: 8]40 Auwers ( 1891). 1901-1959: 959'163 Auwers (1891); for eclipsesonly.
1921 1959: 8141 Auwers (1894).

Mercury
IWArS

1901-1959: 3134 Leverrier (1843).

1901-1959: 4'i68 Hartwig (1879).

Venus
Jupiter (equatorial)

1901-1920: 97136 1 9 0 1 1 9 1 9 : 8'i55 Peirce.

Schur (1896).
1921 1959: 98i47 1920 1959: 8 1 4 1 Auwers (1894).
Sampson (1910).
 652 EXPLANATORY
SUPPLEMENT 'I3
/ HISTOBICALINFORMATION
653
Mars Venus
1901-1919:5:05 Peirce.
8141 Auwers (1894).
1920-1959: 4:'68 Hartwig (1879).

Mq,rs
Jupiter ( equatorial)

1901-1919: 100]24 Sampson(1910). 4168 Hariwig (1879).

7520-7959:98:47 Sampson(1910).
Jupiter (equatorial)
Jupiter (polar)
S8:47 Sampson (1910).
1901-1919:94112 Peirce.
1920-1959:911'91 Sampson(1910). Jupiter (polar)

Saturn ( equatorial) 91191 Sampson(1910).

1901-1911: 831'60 Ba,rnard (1902).

Saturn (equatorial)
1912 1919; 841'88 Barnard (1902).
1920-1959: 83133 Struve (1898).
83:'33 Struve (1898).

Saturu (polar)
Saturn (polar)
1901-1919: 781'16 Peirce.
l9L2 7919: 77'!47 Barnard (1902). 74:57 Struve (1898).
792O-1959: 74i57 Struve (1898).
Uranus
Ure,rurs
34':28 Bernard (1896); See (1902); and Wirtz (1912).
1901 1919: 33]b2 Peirce.
1920 1959: 34'i28 Barnard(1896);See(1902);Wirtz (1912). Neptune

Neptune 36156 Bernard (1902).

1901-1919: 38'166 Barnard (1902).

1920 1959: 36156 13.38 Coefficients of Nutation (1S34-1959) Table 13.348.1 gives the values
Barna.rd(1902).
adopted, beginning with each year shown, of the coefficient of sine (Argument)
in the nutation in longitude (Ary'), and of cosine (Argument) in the nutation in
133478 Semidiametersat unit distance (1961)-1983) From 1960 to 1983 the follow- obliquity (Ae). A blank indicates that the term rvas not included in the tabulations
ing constadrts were used:
for that year. A leader (...) indicates that the coefficient used was the same as in
the preceding entry. The subheading ,,Ae,' is omitted for those terms that do not
Mercuru occur in the obliquity. Terms nurnbered 1 to 13 all have periods greater than 100
days, and are known as long-period terms; terms 14 to 2E have periods shorter than
3:34 Leverrier (1843).
35 days, and a,re known as short.period terms.
 xd
ol'
Terml{o. I 14 )ao
Argument a 2n 2L 2( EJ

A1l) Nl) Ae A6 L1l) A€ < i')

ca!
3:
-17.2985 +9.2500 +0.2082 -0.0903 -1.2550 +0.5447 -0.2074 +0.0900 o
1834
1857 -17.2524 +9.2236 +0.2063 -0.0895 -1.2691 + 0 . 5 5 0 7 ... +0.0885
1896 0.2041 o
't901 -17.236 +9.210 +0.209 0.090 -1.257 +0.546 -o.2o4 +0.088
-17.235 -1.27O +0.551 o'
J
't903
1937 -17.2U +9.210 -1.272
1937 -0.017 T +0.0009f
z
TermNo. 4 5 6 7 P.
Argumenl I - f L+0 L+ f 2L- A

aIP N! (d) ^11) A€* ^4r A6

r893 +O.1476 82.02 +0.0093

1897 -0.0058 +0.0027 ... +0.0125 -0.0067
'1901 +o.110 -0.049 +0.021 0.009
1903 +0.107 74.4 0.0s0
1912 74.3 ... +0.022 ... +0.012
1918 +0.021 -0.007

: -:id-..-

TermNo. 10 11 12
F C
oG'
Argument 2lt a 2L-2r' 2L-29 f' + 90' YO
2r' AJ
^{ A.6 ^V L1b
---'\._
Atp A€ Nb AE < ?,o
3-f
1897 +0.00,f4 -0.0024 +0.0053 -0.0024 +0.0026 -O 0023 +0 0020 O.OOO8 FO
1901 =o
^l
1937 +0.005 -0.003 +0.O04
=o
T e r mN o . 15 16 17 18 19 20 o
Argument t-tt 2t-Zft 2t-O 3t_ft t_2L+ft t+fl
A{ Lt Lt A€ Arl] A€ LIb A,j) A€ z
c
D)
1896 +0.0677
1897 0.0339 +0.0181 -0.0261 + 0 . 0 1 1 3
1901 +0.067 0.034 +0.018 -O 026 + 0 0 1 1 +0.015 +0.012
1903 +0.068 +0.011
1937 ... +0.003

TermNo. 21 22 2g 24 25
A r g u m e n t2 t - 2 L t-l'+t2 t-tt-!) 3t_2L+f, 3t- tt+O
/_^-...- '
^4r L4, A€ LIJ Ae At Ae 1r1t Ae

1901 +0.006
1937 ... +0.006 -0.003 +0.006 +0.003 -0.005 +0.002 0.004 +0.002
'The
sign of lerm 7 jn Ae is given incofiscttyin the atmanactrom 1893 to 1900.
656 EXPLANATORYSUPPLEMENT 13 / HISTORICALINFORMATION
oc/
13.35 system of constants(1968-1983) Figurc of the Earth, continued.
One degree of latitude (m) I I 1 3 33. 5- 5 5 9 8. 4 c o s2 /
13351 SystemofAstronomicalConstants This systemof constantswas replaced + l. I7 cos4d
for the 1984 edition of the Astronomical Almanac by the IAU (1976) System of (d = mid-latitudeof arc)
Astronomical Constantsgiven on pagesK6 to K7 in the Astronomical Almanac One degree of longitude (m) 1 1 4 1 3 . 2c8o s0 - 9 3 . 5 1 c o3sd
and in Table 16.2.
+ 0.l2cos5d
Def.ning constants
The complete system of astronomical constants is given in Supplement
Number of ephemeris seconds in one 31556925.9747 to the
American Ephemeris 1968 (pp.4s 7s).
tropical year (1990)
Gaussian gravitational constant 0.017202098950000
= 354811876069651 13.352 Otd Constants The printed ephemeris of the Sun and inner
planets are
based on the following values of the constants that were in use prror
Pri,mary consto,nts to the intro_
duction of the IAU (1964) System.
Astronomical unit 149600x l0o m
Velocity of light 299792.5x 103m/sec
Velocity of light 299860x 103m/sec
Equatorial radius of the Earth 6378160 m
Equatorial radius of the Earth 6378388m
Dynamical form-factor for Earth 0.0010827
Constant of aberration 20';47
Geocentric gravitational constant 398.603x l0em3s-2
Flattening factor of the Earth r/2e7
Mass ratio: Earth/Moon 81.30
Polar radius of the Earth a(l-f)=6356911.946r.L
General precession in longitude per 5025]64
Solar parallax 8:'80
tropical century (1900)
Light-time for unit distance ,198:',38
Constant of nutation 9:210
Constant of aberration 20:47
Derfued,constants Mass ratio:
Solar parallax 81794 Sun/(Earth+X,Ioon) 329390
Light-time for unit distance 4991012 Earth/Moon (planetary theory) 81.4b
Constant of aberration 20]496
Flattening factor for Earth r1298.25
Heliocentric gravitational constant 1 3 2 ? 1x8 l 0 r 5m r s - 2 13.4 REFERENCES
Mass ratio: Sun/Earth 332958
Mass ratio: Sun/(Earth * Moon) 328912 Adams,J.C.(18.19)."OnanImportaitErrorinBouvard'sTablesofSaturn,,MemR.A.S.
Mean distance of the Moon 3844fi)x 103m 17, 1 2.
Constant of sine parallax for Moon 3422':457 Adams, J.C. (1853). "On New Tables of the Moon,s parallax,, Nautical Almanac
for 1856
Figure of the Earth Appendix, 35 43.
Adams, J.C- (1853a). "On the Corrections to be Applied to Burckhardt's and plana,s
Equatorial radius (primary) 6378160 m Parallax of the Nloon, Expressed in Terms of the Mean Arguments" M.N.R.A.S. ll,
Polar radius 4(1 - f) = 63567'74.7ri 262 266.
Squa.reof eccentricity 0.00669.154 Adams, J.C. (1877). "Continuation of Tables I and III ol Damoiseau's Tables of Jupiter
Reduction from geodetic Iatitude / d' - d = -ll'32')743Osin2S Satellites" Nautical Almanac for 7887 Appendix, 15 23.
to geocentric latitude /' + I I1633sin4d - 010026
sin6/ Airy, G.B. (1832). "On an In€quality of Long period in the Motions of the Earth and
Venus" P/ril. Tians. Roy. Soc.67 124.
Radius vector p = 4(0.9983270'13
Airy, G.B. (1835). "On the Calculation of the perturbations of the Small plaoets and the
+ O.O01616438 cos2$
Comets of Short Period" Nautical Almanac for 76j7 Appendix, 149-171.
- 0. O00003519 cos4/ Airy, G.B. (1a45). Red,uction of the Obsen)ations of the planets (1750 18J0) London.
+ 0. 000000008cos6d) Printed by Palmer and Clayton, and sold by J. Murray.
 SUPPLEMENT
EXPLANAToFIY 13 / HTSTOBTCAL
TNFOBMATTON
65g 659
(1750 1830) 2 r'olumes
Airy, G.B. (1848). nedtcrion of the Obsenations of the l|loon Carlini. F. (1832). "\uote Tarole de'\Ioti Apparenti del Sole', Efferne:ndi...di illtLanoper
London. Printed bv Pahner ard Clavton. and sold b1 J. \lurra1. l'Antto 1833 Appendix. I 10.1.
Divided from the
Airy, G.B. (1849). "Corrections of thc Elements of the \loon's Orbit, Christy, .1.\V.and HarriDgton. R.J. (1978). "Thc Satellite of Pluro" ,4stron. EJ. 1005
"/.
Lunar Observations \lade at the Roval Observatory of Greenrvich. ftom 1750 to 1830" 1008.
Mern. R A S. 17, 21 57. Clemencc.G.M. (1951). "Perturbations ofthc Five Outer Planets by thc Four Ilner Ones"
Airy, G.B. (1855). Obsentations Made at the Royal Obsetuuto4l, Creenuticlt' in tlrc Year Astrott Papers for the American Ephemeris XllI, Y
1853 (Her Nlajesty's Stationary Office, London). Cohn, F. (1897). "Bestimrnung der Bahnelernentedes V. Jupiters \{ondcs,' Asl. Nac/r.
Aksues, K. (1978). "The motion of.lupiter XIII (Leda). 197'12000" Astron l a3 l2l9 142. 289 338.
1256. Comrie, L.J. (1926). "The Usc of a Standard Eqrrinox in AstronornJi, nlN./?.,{.S. 86.
Andoyer, H. (1915). "Sur le Calcul des Ephemerides des Quatres Ancipns Satellites de 618 631.
J r - r p i t e r "B u l i A s l r . 3 2 . 1 7 7 2 2 1 Comrie, L.J. (193'1)."Phenomena of Saturn's Satellites" Mem. B.A.A. J0, 97 106.
Auwers, A. (1891). "Der Sonnendurchmesser nach den Bcobach-
und der Venusdurr:hmesser Crawlord, R..T. (1938). "The Telth and Eleventh Satellites of Jupiter" p.A.S.p. i0,344
tungen an den Heliometern dcr Deutschcn Venus_Expeditionen",4st. No.i. 128,361 347.
376. dc Darnoiseau. \l.C.T. (1824). I'abLesd.eIa Lune for-m6.espar la seule theo t: de I'attractiorr
Auwers, A. (189'1). "Benerkung zu den \littheilungen von Resultaten der Dcutschen (Bachelier, Paris).
Venus-Expeditionenir \r. 3066 und 3068 der A.\'" ,{st AIar:A 134. 359 3tj2 de Darrroisearr. \I.C.T. (1836). Tableseclzptiquesdes satelhtesde Jupitu' (liachelier. paris).
Baily, F. (1825). "Otr the Construdion :rnd Use of Some \el Tables for DeterruiDing the Delanrbre. .l.B.J. (1806). Tablesdu Soleil (Bachclier. Paris).
Apparent Placcs of Nearly 3000 Priut:ipal Fixed Stars ,\lem R.,4 S. 2, Appendix. Delambrc, J.B.J. (1817). Tables ecliptiquesd.essatcllit* rle Juptter (BacJrclier,paris).
i ccxxviii. Dor,vning,A.NI.W. (1890). "Corrcclions to the Elenrcntsof the Orbit of ,luxo" I,I.N.R..A.S.
Bessel,F.W. (1828). "Uber den Gegen*'arl,igenZustand Utrsorer Kennttriss dcl Sonlcn- 50, t87 497.
bewegung und die Nlittel zu Ihrer V'rbcsserurg" Ast Nadt 6.261 276 and 293 302. Duncombc. R.L. (1969). "Hulioccntric Coordinates of Ceres. Pallas, ,hrno. \'esta 1928
Bessel,F.W. (1830a). ?a6rla€ Reqiontor anae Konigsberg 2OO0" Astronomrcal Papers Jor the Arne.rican Ephemeris XX. II.
Bessel,F.W. (1830b). "Vorlaufige \aclrricht von Einenr auf cr Kc,nigsbergerSterr*arte Dunham. D.\\'. (1971). "Thc \lotions of the Satellitcs of Uranus Disscrtation. (Yalc
Befindlichen Crossen Heliometer" ,4s1.Nac,'r.E.397 '126. Universit5. \en' Har.en. CT).
Bessel.F.W. (1875a). "U tersuchungeniibcr den Planeten SaturD. ScilrenRing ulld Seinen Eckert. \\:.J.. Brourr'er. D.. arrd Clemence. G.\1. (1951). "Coordinates of the Five Outcr
Vierten Trabantcrr" Abhand.lungen l, 7lt0 159. PlaDcts" /slrcnomical Papers for the Ameri(an Ephemeris XIl.
Bessel,F.W. (1875c). "Ueber die Neigrrlrgder Ebene des Sattrnsringes" Abhandluryenl, Eichellrerger,W.S. (1892). "Tho Orbit of Hyperion" Astron. J.II, 145 t57.
319 321. Eichelberger. W.S. and Nervtorr. A. (1926). "'l'he Orbit of Nepturre's Satcllite and thc
van Biesbroeck,G. (1957). "The mass of Neptune from a Nerv Orbit of its SeconrlSatellite Polc of Ncptune's Equator" /slroaomical Pape.rs for the Ameritut Ephtneris IY.
Nereid" -4slrorr.J. 62.272 271. 27s 337.
Bobone, J. (1937a). "Tables del VI (scxro) Satellite de Jrrpitcr" ,{sf ,\hch. 262. 321 3lti. E l l i o t . J . l - . . F l c n c h . R . G . . F r o g e l .J . A . . E l i a s . . l . H . . \ l i n k . D . J . . a n d L i l l e r . \ \ ' . ( 1 9 8 1 ) .
'Orbits of \ine Uraniarr R.irgs'
Bobone. J. (1937b). "Tables del \rII (scpliuxr) Satellitc dc.luPiter" ,4sf. Nac/r. 263. {01 Astron. J. 66, {ll .155.
412. Encke. .1.1".(1824). Der Verntsdnrthgangoon 1769 (Bccker. Gotha).
Bouvard, A. fta21). TablesAstronotniqucs. ...Contenant lcs Titblcsde Jttpiter' de Soturtte Encke, .1.F. (1852a). "Uber eirx'Neue Methode dcr Borechnung der-l)llrx'tenstorurrgcn"
et d'Uranus (Bn<helieret Huzard, Paris). Astr. Nar:lt.33, 377 398.
Bower, E.C. (1931). "On the Orbit:rnd tr{assofPluto wilh alr EPhcmeris fttr 1931 1932' Enr:ke,J-F. (l8l,2b). "Zusatz zu dcnr Aufsatze: Ncur'\'lethode zur Berechnlug clcr Speciellclr
Lick Obs. Bull. 15. 171 178. Storullgcn in \r. 791 un<l 792 der Astron. \achri<:lrten" Ast. Noc/r. 34. 3.19 360.
Breen, H. (18.18)."Corrections of Lirrdenaus Elcments of thc Orbit of Venus Deduccd frorr Farlcl. R. (1E56a). "Correction of the Elements of Pallas" Nautlcal Alnatnt for 1860
the Greeuuich Plarretarl Observations. from 1750 1830 llrrl. -R.,4.S 1E.95 l5l. 570 :-t72.
Breen, H. (1851). On the Corrections of Lirrdenau's Elenrentsof llars" lvlem R.A.5.20. Farler. ll. (1856b). "Corrccliorr of the Elements of Vcsta" Nazfir:al Altnartoc for 1860
137 168.
Brown, E.W. (191,9). Tables oJ the Motion of Lht:MootL (Y.rle Ultivcrsit-"- Press, Now H?Lvcn). Farcly, R. (lfl56c). "Elenrents of Asl,raea" NautitnL Alrnanttt for 1860576.
Burckhardt, .J.C.(1812). Tablesde Ia Luru'ir TablesAstronontiqucspubliles par k: Iluttau Godcarcl, \\r. (1878). "Orr the Correction of the Elcrrrcntsof Ceres" M.N.8.,4.S. 3E, 119
des Longitudes tle France (NIme. Ve Cotrcier. Paris). 122.
Burg, J.T. (1806). ?o6les de Ia Lune in Tablcs Astrononmlut:s1nfilt€espar Le Burearl des Grosch. [[.fi..I. (19a8). "The Orl)it of the Eight Satcliifes of Jupiter" Ashon. J.53. 180
Longttudes de -Frunce(\Lne. Vc Court:ier. Paris). 187.
Carlini, F. (1810). E:xpostzione tli un nuouo lnctod.o dr cortstl uire Ic Tauole 'AslrortottLiche Hall, A. (1878). "Observations arrd Orbits ofthe Satcllites of \'lars" I4losiington Obser-
in Applicuto alle Tattole de Sole (Rcal<' Stamperia' \'lilan) tatzons for 1675 Appendix, I 4tj.
660 EXPLANATOBYSUPPLEMENT 13 / HISTORICAL
INFoRMAION
661

Hall, A. (1885). "The Orbit of lapetus. the Outer Satellite of Saturn" WashingtortObser- Kuiper, G.P. (1956). "On the Origin ofthe Satellitesand the Trojans' Vistas irt ,4stronomu
oations for 1882 Appendix I, I 82. 2, 1631 1666.
Hall, A. (1886). "The Six Inner Satellites of Saturn" WaslrrrrgtonObseruationsfor 1883 de Lalande, J.J. (1771). A stronomie 2rrd ed. (Desairrt,Paris).
Appendix I, 1-74. de Lalande, J.J. (1792). Astronomie 3rd ed. (Desaint, Paris).
Hall, A. (1898). "The Orbit of the Satellite of Neptune" Astron. J.19, 65 66. de Laplace, P.S. (1787). ,,SuI les Inegalitie Seculares des plan6ts,, Metn.oircs tle I'Academie
Halley, E. (17,19). Tabulae Astronomicae, acced.Lntd,eusu tabularum praecepta (Gulielmum d,esSciences de Paris 1781, pp. 1 50.
lnnys, London). de Laplacc, P.S. (1788). ,,Theoricde Jupiter et de Saturne,, Memoires de l,.4cademie
d.es
Hansen, P.A. (1817). "Aufzug aus Einem Briefe des Hcrrn Professors Hansen an den Sciences de Paris 1785, pp. 33 160.
Herausgeber" Ast. Nach.25. 325 332. Leveau, G. (1896). "Tablcs du l\louvement de Vesta. Fondees sur la Conrparison
de la
Haosen, P.A. (1857). ?ables de la Lune construites d'apress le principe neutonien de Ia Tlreorie avec les (Jbservations" Ann. de I'Obs. d,e paris 22, A.l A.317.
grauitation unil)erselle (G.E- Eyre et G. Spottiswoode, Lorrdon). Leverrier, U.J.J. (1845). "Theorie du Mo\,-emeDtde Mercure" Conna.issanctdes Tenrps
for
Hansen, P.A. and Olufsen, C.F.R. (1853). Tables d,u Soleil (B. Luno, Copenhagerr). 1618 Additions, I 165.
Harkness, W. (1899). fThe following is takel fiom the preface to The American Ephrcm- Leverrier, U.J.J. (1846). "Researchcssur les Nlouvements de la plarretr: Her.schel(dite
eris lor 19031 "oD a recent investigation by my self, which rests upon 35. 8'12 merid- Uranus)" donnoissane (les Temps lor /819 Additions, 1 254.
ian observations made at GreeDu'ich, Paris, \l'ashington, Konigsberg, NIadra^s.lr'lilan, Leverrier, U.J.J. (1858). "Theorie et Tables du \,Iouvement Apparent (lu Solcil'. Inn. de
Dorpat, N{odena, and Seeberg."] l'Obs. tle Paris 4, 1 263 and I 121.
Harrington, R.J. and Christy, J.W. (1981). "The Satellite of Pluto III" , stron. "/. 86, Ler,'errier,U.J.J. (1859). "Theorie et Tables du trIov€nrentde Mercure ,lrLn. d,e I'Obs. d.e.
442 443. Pdri.s 5, 1 195.
Harris, D.L. fRecent (unpublished) work on the satellites of Uranus quoted by Kuiper Leverrier, U.J.J. (1861a). "Theorie et Tables du N{ovementde Venus" Ann..rle l,Obs. d.e
( 1 9 5 6 ,p . 1 6 3 2 ) . 1 Paris 6, I 184.
Harshman, W.S. (1894). "The Orbit of Deimos" Astron. .1. 14, 145 1,18.lHarshrnan's Levcrrier, U.J.J. (1861b). "Theorie et Tables du Mouverrrenrde l",lars,,;l2rr. tle I'Obs. de
elements of Phobos "(comnunicated by Prof. Harkness)" arc attached (in nranuscript) Paris 6, 185 ,135.
to a copy of the above paper. It is believed that they $'ere never published. See also Leverrier. U.J.J. (1876a). "Tables de Jupiter" ,4nn. de I'Obs. de po.ris 12, | 76 ard 1 178.
P u b l . U . S . N . O . , 6 , A p p e r r d i xI , 8 2 l ( 1 9 1 1 ) . 1 Leverrier. U.J.J. (1876b). "Tables de Saturne" :lnn. de l'Obs. rle pons 12. A.l A.80 and
Herget, P (1938). "Jupiter XI" H.A.C. 163. A.1 4.286.
Herget, P., Clemence,G.M., and Hertz, H.G. (1950). "Rectangular Coordinates of Ceres. Leverrier. U..l.J. (1877a). "Tables du llouvement d'Uranus, Fonclees srrr la Cornparison
Pallas, Jrrno, Vesta, 1920 \960" Astronornical Pape,rsJor the American Ephemerisll, de la llreorie avec les Observatiolrs" Ann. aLeI'Obs. d,e Paris 14, A.1 A.92 and A.1
521 587. A.163.
Herget, P. (1966). "Rectangular Coordinates of Ceres, Pallas, Juno Vesta 1960 1980" Leverrier, U.J.J. (1877b). "Tables du tr4ouvementde Neptune, Fondeessur la Corrrparison
Astronomical Papers for the Am,erican Ephemeris XYl. dc la Thcorie avec les Observations" Ann. de I'Obs. d,eParis 14. 1 70 arrd 1 96.
Herget, P. (1968). "Outer Satellites of Jupiter" Astron. J.73,737 742. Levin. A. E. ( 1931,193'1)."\l ut rral Edipses and Occultatiom of Jupiter's Satcllites' J.8.,4 . t.
H.tr'I. Nautical Almanac Office (1933). Planetary Co-onlinates for the Yeats 1800 191A 4 2 , 6 1 . 1( 1 9 3 1 ) .A , I e m .I ] . A . A . 3 0 . 1 . 1 91 8 3 ( 1 9 3 4 ) .
Refet"red to the Equinor oJ 1950.0 London. Linderrarr.B. (1810). Tabulae Veneris nouae et cortectae(Bccker. Gotha).
H.M. Nautical Almanac Office (1939). Planetarg Co-onlinates Jor the Years 1910 1960 Lindenau, B. (1811). Tabulae Mol tis noLae et correctac(Schoeniana,Eiserrbcrg).
Referred to the Equinoz oJ 1950.t) Lo:rl<1,on. Lindenau, B. (1813). Inuesttylation. noua orbitae a Merrnrio cir:a Solem d,escriptae(Becker,
Herrick, S. (1952). "Jupiter IXX and Jupiter XLl" P.A.S.P.64,237 241. Gotha).
Hill, G.W. (1890). "A new theory of Jupiter and Saturn" Astronomical Papers Jor the Longstreth, \'1.F. (1853). "On the Ar:curacy of the Tabular Longitude of rhe \Ioon, to
American Ephemeris lY, 1 577 be Obtained by the Construction of New Lunar Tables" ?rons. Arner.. I'h.tl. Soc. 10,
Hilt, G.W. (1895a). "Tables of Jupiter Constructed in Accordancc rvith the \Icthods of 225 232.
Hansen" .Aslronomical Papers Jor the American Epherneris VII, 1 1.14. Nlarth. A. Ilarth published a long series of ephenrerides for physical obscrvations of the
Hill, G.W. (1895b). "Tables of Saturn Constructed in Accordance with the N{ethods of plancts and for obserr,ations of the satellites. Thc first and last of each serics are alone
H a n s e n " , 4 . P . A . E .7 , 1 4 5 - 2 8 5 . quotecl here.
Hind, J.R. (1855). "Revised Elements of Juno" Nautical Alrnanac for 1859 562 564. Marth, A. (1870 1894). "Ephcnrcris of the Satellitcs of Uranus" M.N.R.A.S.30, 150 54,
International Astrononrical Union (1950). Trans. IALI7,65 and 83. 171.
Kendall, E.O. ( 18?7). [An extension for the year 1880of Damoiseau'stables of the satellites Marth, A. (1873 1895). "Ephenreris of the Satellites of Saturn" M.N.R.A.5.33, 513 55,
of Jupiter (probably never published).] 16.1.
Kowalski, M. (1a55). Recherches sur les mouuements d.e Neptttue Kasan Marth, A. (1878 189r). "Ephemeris of the Satellite of Neptune" M.N.R.A.S.3E, 475 51,
Kuiper, G.P. (1949). "The Fifth Satellite of Uranus" P.A.S.P.61, 129.
ooa EXPLANATORYSUPPLEMENT 13 / HTSTOBTCAL
|NFORMAIoN
663
Marth, A. ( 1883 1896). "Data for Finding the Positionsof the Satellitesof \[ars" M. N. R. A. S. Peters. C.A.F. (1842). Numeru,s Constants Ntdalionts ex ascensionibus rectis
st.ellae po_
44,29 56, 435. Iaris in specuLadorpatensi 1822 1g38 obsentatis ded.tLctus;adjecta est trzsquisttio
te_
Marth, A. (1891 1896). "Ephemeris for Computing the Positions of the Satellites of oretica de lormuLa nlttationis St. petersburg.
Jupiter" M.N.-R.,4.S.51, 505 56, 534. Plana, J.B. (1832). Theorie d,u mouuement de Ia Lune 3 lolumes, (lmprimeric
Royale,
Mason, C. (77a0). Lunar Tobles London. (no publisher given in book) Turin).
Mason, C, (1787). Mager's Lunar Tables In'.prouedLondon. Printed by William Richardson Pottier, L. (1896). ,,Addition aux Tables Ecliptiq.cs dcs Satellites de Jupiter
dr: Damoi_
and sold by P. Emsly. seau" Bzl1. Astr. 13,67 79 and 107 112.
Mayer, T. (1770). TabuLae Motuum SoLrs et Lunae nouae et cor\ectac; puibus accedit Robertson, J. (1924). "Orbit ofthe Fifth Satellite of.lrrpiter" ,4p. J. :tb, 190
1g3.
Method,us Lonqitud,inum Prornota (Guliclmiet Johannis Richardson, London). Rose, L.E. (197a). "Orbit of Nereid and the \,lass of Neptune,, Astton. -r. Zg. .j89 .190.
Mignard, F. (1981). "The Nlean Elements of Nereid" Astron. J.86, 1728 1729. Ross, F.E (1905). "rnvestigationsof the orbit of phoebc" Haruard Annars 5r.
ror rr2.
Newcomb, S. (1865). '4n Intestigation of the Orbit of Neptune, urth General Tables oJ its Ross, F.E. (1907a). "seDi-definiti'e Elements of Jupiter's Sixth satelrite' Lick
Motion (Smithsonian, \\iashington, DC). obs. Btfl.
4. 110 112.
Newcomb, S. ( 1867). "Investigation of the Distance of the Sun alrd of the Elentents Which Ross. F.E. (1907b). ,,Ne\v Elemcnts ofJupiter,s Seventh satellite.. Asl. 174.359 362.
Depend rrpon it" Washington Obseruattons for 1865 Appendix II, I 29.
Ross, F.E. (1917). "Ne\\' Elcments of Mars and Tablcs f.r Correctirg the Helioc.rrtrir. po_
Newcomb, S. (i873). An Inuestigation of the Orbit of Uranus, It)ith GencraL Tables of its
sitions Derived from Astronomical Papers. V.rlunre VI, part IV" ,4slronorl ical po.pers
Motion (Smithsonian. \\'ashington, DC).
for the Ameriu.rt Ephemeris IX, 2b1 271.
Newcomb, S. (1875). "The Uranian and Neptunian Systems" Was,fringtonObserrations
Runkle, J.D. (1855). ,'New Tables for Determirring the Values of thc Cocfllcic'i,s. in the
/or 1873 Apperdix I, 1 7.1.
Perturbative Frrnction of Planetary Motion, Which Depeld upon the Ratio of the
Newcomb, S. (1878a). "Researches on tlto llotion ol the Moon" trlashirrytorL Obseruations
Mean Distancts' Srnithsonian Contributiot to Knowledgc IX. Appl (Smithsorrian,
for 7875 Appendix II, 1 280. Washington, DC).
Newcomb, S. (1878b). Correctiots to Hansen! Tables of the Moon, Prepared. and Prrnted.
Sampson. R.A. (1910). ?a6les of the Four Gr?at SatellitesoJ Jupiter (\\iilliarn \\bsLrr,&
lor the Use of the The American EplLetnetisand Nartical Almanac (U.S. Government
Son.Londorrl.
Printing Office, \Vashington, DC).
Schubert. E. (185.1). "On the Correction for the Eiements of (1) Ceres Astr.on..J.3.
Newcomb, S. (1895a). "Tables of thc \lotion of the Earth on its Axis arrd Around thc Sun'
Astronomical Papers Jor the Ephemuts VI. 1 169. fUsualll referrcd to as Tables of 153 t59 a|ld 162 lri5.
the Sun.] Seidelmann,P.K. (1979). 'The Ephemerides:Pa^st.PreseDtand Future Dynarntcsof the
Newcomb, S. (1895b). "Tablesofthe HelioceDtric\4otion oftrIercurl" AstrortotnicalPapers Solar SgstcntR.L. Duncombe, ed. pp.99 11,1.
Struve, F.G.W. (11311)."Sur le Coefficient coustant dans l Aberration des Etoiles Fixes.
for the American Ephemeris Yl, l7'\ 270.
Newcomb, S. (1895c). "Tables of thc Heliocentric Nlotion of Venus" Astronomical Papers Deduits dcs Observations Qui ont et6 Executecs a I'Obsertatoire de pulkovl par
I'Instrument dcs Passagesde Repsold, Etabli <lnrrsle Premier Verical.' Mert. d.t:I'Atatl
lor the American Ephemeris YI, 271 382.
Newcomb, S. (1895d). "The Elements of the Four Inner Plancts aud ihe Fundamental Imp. Sci. St. Petersburg6th series,5,229 285.
Constants of Astrorromy" Supplemnt to the American Ephem,erisJor 1897 1, 2O2. Struve, G. (192'1 1933). Neue flntersuchungen nn Satutnsystem. VerolJentliclwnqtn dcr
Newcomb. S. (1897). "A New Determination of the PrecessionalConstant" Astronomicdl Unixersitatssternuarte zu: Berlin-Babelsbcrv 6, parts 1, 4, 5.
Papers for the American ET.themerisllll. 1 76. Struve. c. (1924). Part 1. "Die Bahn von Rhea" t 16.
Newcomb, S. (1898a). "Tables of the Helioccntric Nlotion of flars" Astrcnomical Papers Struve. G. (1930). Part 4. "Die Sl.steme llirnas-Tethls und Enceladus-Dirxx) I 61.
Jor the American Ephemeris VI, 383 586. Struve. G. (1933). Part 5. "Die Beobaclrtungencler.ausserenTrabanten und dio Bah-
Newcomb. S. (1898b). "Tables ofthe Heliocentric Nlotion of\eptune l1i,roaoDical PapeTs nen Yon Titan und Japetus" 1 44.
for the Anrcrican Ephemeris Yll, 287 416. Struve, H. (1888). "Bcobachtungender Saturnstrabarrten,'Supplement 1 aux Obscrvations
Newcomb, S. (1898c). "Tables ofthe Helioccntric \{otion ofNeptune" /.ctrofl.ornicalPapers de Poulkova, I 132.
for the American Ephemeris \ll, 419 471. Struve, H. (1903). "Neue Bestirnmung dcr Liblatiorr Mirnas-Tcthys,' Ast. Nach. 162,325
Newcornb, S. (1898d). "Sur les Formules dc Nutation Bases sur les Dccisiols dc la Con 344.
ference de 1896" Bull. Astr'. 15,211 246. Struve, H. (1911). U6er die Lage der Marsachse urLd die Konstanten in Murc S1)stcnr
Nicholson, S.B. (1944). "Orbit of the Nilth Satellite of Jupiter" Ap. ,1. 100,57 62. SitzurLgsberichte t1.erK6niglich Preussischen Akadernie d,er Wissenschaften fii.r 1911
Paris (1896). Conference Internationale d.esEtoiles Fond,amentales: P|ocess- Vcbaux Pans. 1056 1083.
Peirce, B. (1848a). "Perturbations of Uranus: Investigations into the Ar:1ir-rn of \eptune Struve, H. (1913). Bahnen der Uranustrabanten. Abteitung l: Oberon untl Tikrnia. Ab-
upon Uranus" Pttc. Amer. -4cad. Sci (Boston) 1, 144 149 and 332 3,12. handlung en der K oniglich Preussischen Akadernie iler Wissenschaften.
Peirce, B. (18.18b)."Uber die StorungeD cles\eptune's Ast. Nach." 27. 215 2la. Taylor, S.W. (1951). "On the Shadow of Saturn on Its Rings" Astron. J.55,229 230.
Peirce, B. (1853). Tables oJ the Moon 2nd ed. 1865. (Bureau of Navigation. \\'a-shington, Todd, D.P. (1a76). A Continuatiorl of d.e Damoiseau,s Tables of the Sate ites of Jupiter
DC). to the Year 1900 (Bureau of Navigation, Washington, DC).

i*
 SUPPLEMENT
EXPLANATORY 13 / HTSTORTCAL
INFORMATTON
664 665
3 volumes 2d ed ' 1814-1825 The reports and recommendationsof commission 4 of the InternationarAstronomical
Vince, S. (1797-1808).A CompleteSYsten ol Astrotuotrly Union have been published as follows:
(University of Ca,urbridge,Cambridge)'
Walker, S.C. (l8l8). Repott of the llnited States Cot*t ond Geodetic Sunev for 18J8'
(1?46). 'Tabulae procalculaudis eclipsibus satellitum f,otis" Acta Soc'
W"rg""ti", P.W.
Transacllons IAU A$€mbly
reo, scd. IJpsali'ensisail annum lflI. Stockholm
51'
W so; R.E. (jr.) (19as) "Revi6ed Orbit and Ephemeris for Jupiter X' P '4'S'P' I,I59,207; 1923. Rome 1922
241-242. II, 18-19, L78,229i L926
and Not- Cambridge, England
Winlock, J. (l8f!). Tables of Mercury, for the IJse of The Ameticnn Ephe'meris III, 18, 224, 300; 1929 Leiden 1924
tical Altnonac (Bureau of Navigation, Washington, DC)' [V,20,222, 282i !933. Cambridge,MA
van Woerkom, A.J.J. (1950)' "The Motion of Jupiter's Fifth Satellite, la92 1949" Astro' V, 29-33,281 288,369-371;1936
1932
Paris 1935
nomical Pcperslor the Americtn Ephemeris XllI'' L-77 Vl, 20-25, 336, 355-363; 1939 Stockholm
Woolard, E.W. (1953). "Theory of the Rotatiou of the Eaxth Around its Center of Mass" VII, 61, 75-83; 1950
1938
Zurich 1948
Ashvnomical Papersfor the American EphemerisXY' pt l, VIII, 66-68, 80 102; 1954 Rome 7952
Woolhouse,W.S.B. (1s33). "New Tables lor Computing the Occultaiions of Jupiter's IX.80-91:1957 Dublin
Satellites by Jupiter, the Tlamits of the Satellites and Their Shadows over the Disc 1955
X, 72, 85-99; 1960 Moscow 1958
ofthe Planet, aod the PositionB of the Satellites with Rcspect to Jupiter at any Time"
XI, A, 1..8; 1962,B, 164-167,44L-462i 1962 Berkeley 1961
NauticalAlmanac lor 7835APpendix' 1-39.
XII, A, 1-10; 1965,B, 101-105,593-$25;1966 Hamburg 1964
Zaduaaisky,P.E. (1954). "A Determination of New Elements of the Orbit of Phoebe,
XIII, A, 1-9; lS67, B, 47-53, 178-182;1968 Prague 7967
Nioth Satellite of Sat'ttrt" Aatton, J. 59, 1-6.
XIV, A, 1-9; 1970,B, 79 85, 198-199;1977 Brighton 1970
XV, A, 1-10; 1973,B, 69-72; 1974 Sydney t973
XVI, A1, 1-7; 1976,B, 31, 49-67;1977 Grenoble 1976
XVII, 41, 1-6; 1979,B, 63-83; 1980 Montreal 1979
XVIII, A, 1-13;1982,B,67-72; 1983 Patras 1982
XD(, A, l-6; 1985,D, 93-96; 1986 New Delhi 1985
XX, A, 1-6; 1988,B, 105 108;1989 Ba.ltimore 1988

Protocols of the Procedings of the Interuational Confetence held at Washington for

the purpose of fixing a Meridian and a Universal Day. October 1884.Washiogton D.C.,
1884.
Process-Verbaux of the ConferenceInternationa.ledesEtoiles Fondamentalesde 1896.
Paris. Bureau des Longitudes, 1896.
CongressInternational des Ephemerides Astronomiques tenu a'l'Observatoire de Paris
du 23 au 26 Octobre 1911,Pa"risBureau desLongitudes,1912.A full a.count, with Erylish
translations of the resolutions,is given in M.N.R.A,S., 72,342-145, 1912,
Colloque International sur les Consta.ntesFondamentales de I'Asttonomie. Observa-
toire de Pa.ris, 27 Mars-ier Avril 1950. Colloques Iuternationaux du Centre National de la
RechercheScientifique,25, l-131, Pa.ris,1950.The proceedingsand recommendationsare
also availableit Bull, Astr, 16, parts 3-4, 163-292, 1950.
Proceedingsof IAU Colloquim No. 9 (Heidelburg, 1970) on .The IAU System of
Astronomical Constants," CelestialMechanics4, n,o.2, 128-280,1971.
Duncombe,R. L., Flicke, W., Seidelma"nn, P. K., and Wilkins, G. A.,1977\.
Proceedingsof the Sixteenth General Assembly,Grenoble 1976,p. 56.
 CHAPTER
14

RelatedPublications
compiledby M.R.Lukac

14.1 CURRENTPUBLICATIONS

14.11 JointPublications
of the RoyalGreenwichObservatory
andthe Uniled
StatesNavalObservatory
These publications are available from Her Majesty's Stationery Office and from the
Superintendent of Documents, U.S. Government Printing Office, Washington, DC,
20402.
In the United States, the Sight Reduction Tables may be obtained from the De-
fense Mapping Agency, which includes the former Hydrographic Department. The
Defense Mapping Agency's distribution center address is: DMA Combat Support
Services, Washington, DC, 20315-0010.
The Astronomi,cal Almanac contains the ephemeridesof the Sun, Moon, planets
and their natural satellites, as well as data on eclipses and other astronomical
phenomena. It includes information concerning timescales and coordinate systems;
data on staxs and stellar systems, including bright stars, standards for photometry,
radial velocity, and radio sourcesl and lists of bright galaxies, X-ray sources,variable
stars, quasars, and pulsars. It also contains a list of the observatories of the world,
information concerning the calendar, and the Astrouomical System of Constants'
The Nauticol Almanac contaiDs ephemerides at an interlal of one hour and
auxiliary astronomical data for marine navigation.
The Air Almanac contains ephemerides at arr interval of ten minutes and aux-
iliary astronomical data for air navigation,
Astronomical Phenomena conta.ins extracts ftom The Astronomical Almano,c
and is published aonually in advance of the main volume. It contains the dates and
667
668 EXPLANATORY
SUPPLEMENT 1 a / B E L A T E DP U B L T C A T T O N S
669

times of planetary and lunar phenomena and other astronomical data of general 14.13 OtherPublications
olthe RoyalGreenwjch
Observatory
interest.
The Star Almanac for La,nd Surueyors contains tabulations of R. declination,
Planetary and Lunar Coonl,inates, 1981 2000 provides low-precision astronom- and
E for the Sun for every 6 hours, and right asce'sion to 0: I and declinatio.
ical data for use in advance of the annual ephemerides and for othcr purposes. It of I ,, of
all stars brighter than magnitude 4.0 for each month. In addition.
coniains heliocentric, geocentric, spherical, and rectangular coordinates ol the Sun, thc ephemericles
of R, declination, and E lor the Sun are represented by polynomial
Moon, and planets; eclipse data; and auxiliary data, such as orbital elements and series for each
month. This volume is availa.ble from Her Majesty,s Stationery
precessionalconstants. Office and frorn
Bernan Associates, 9730 E. George palmer Highway, Lanham. NID,
Sight Red,uctionTablesfor Air Nauigoiion, (Pub. No. 249, reproduced as (British) 20706.
compact Data for Natigation and Astronomy
Air Publication, A.P. 3270). Volurnc 1 cortains the altitude to 1' and the azimuth for 1gg6 to 799, coltains data,
which are mainly in the form of polynomial coefficients, for use
to l' for the seven most suitable stars for navigation, for each degree of latitude by navigators and
astronomers to calculate the positions of the Sun, \,Ioon. navigatioual plarrets,
and for each degree of local sidereal time. Volumes 2 and 3 give similar data for and
bright stars using a small programmable calculator or personal colnpurer.
each degree of declination to 29' and for each degree ol hour angle. Tabulations
Interpolation and Allied ?116les contains tables, formulas, and explanatory notes
extend to depressionsof at least 5o below the horizon.
on the techniques for numerical interpolation, difierentiation, and integration.
Sight Reduction. Tables Jor Marine Nauigation, (Pub. No. 229, reproduced as In
particular, it contains extensive tables of Bessel and Er€rett
(British) Hydrographic Department, N.P.,101,1971onward). Thesetables give alti- interpolation coeri_
cients. This booklet is available from Her tr{ajesty,s Stationery Of6ce
tude to 011,r'ith variations for declination and azimuth to 091, rvith arguments for and Bernan
Associates.9730E. GeorgePalmer Highway, Lanham, MD, 20706.,l.he
latitude, hour angle, and declination, all at f intervals. They provide all solutions companion
booklet Subtabulalzon contains tables of Lagrange interpolation cotfficicnts
of the spherical triangle, given two sides and the included angle, to flnd a third side at in
tervals of 7/2O and 1/24, as well as details for two other techniques oi
and adjacent angle. systetratic
interpolation.
Royal Obse,r,tatory Bulletins (21 181) and Royal Greenwich ObseruatoryBul
lettnt; (I 2Oand frorn 182) are issued regularly and contain details of cuuent
astro-
14.12 OtherPublications
of lhe UnitedStatesNavalObservatory nomical research. These publications may be obtained, subject to availabilitl,,
from
the Ro1,al Greenwich Obsenatory, Madingly Road, Cambridge CB3 OEZ.
Almanac for Computers contains shori mathematical series that are used to repre- SelectedNAO Technical Notes of the Royal Greenuich. Obseraatorg are issued
sent the positions of the Sun, Nloon, and planets for eficient evaluation with small irregularly to disseminate technical information concerning astrorromical phenom_
computers or programmable calculators. Data for both astronomical and naviga- ena. ephemeridcs, and navigation.
tional applications are included. Discontinued. Greenuich Obseruations is a complete list of the appendices ancl special in_
Astronomical Papers of The Arnerim,n Ephemeris are issued irregrrlarly and vestigations included in thc annual volumes of observations macle at thc Ro1.al
contain reports ofresearch in celestial mechanicswith particular relevanceto ephem- Observatorv, Greenli'ich. A list of the separate publicatious of the Observatory arc
erides. given in the volume for 1946.
Uniterl States Naual Obseruatory Circzlors are issued irregularly to dissemilrate Royal Obseraatory ,4nnols This series of publications includes: Number 1. ,,Nu_
astronomical data concerning ephemerides or astronomical phenomena. tation 1900 1959," 1961;valuesbased on E.W. \Voolard,sseries.
The Multi-gear Interactiue Computer Almanac (\,IICA) is an integrated package Annals of Cape Obseruatory "fhis serics includes Dapers and observational data
of software and astronomical data on a floppy diskette. MICA v'ill produce to full t h a t c r e o [ r c j e v a r r creo r h e e p h e m e r i d e s .
precision most of the daia in The Astronomical Almanac, including both positional
and physical data interpolated to any date and time within the appropriate year. 14.14 Publications
of OtherCountries
Versions are available for microcomputers running under MS-DOS and Nrlacintosh
Systems. Apparent PLacesof Fund,amental ,9fars is prepared annually by the Astronomis_
Publications of the United States Nat:al Obse.rtatory, Second,Series, are issued ches Rechen-Institut in Heidelberg and contains mean and apparent coordinates of
irregularly and contain observational results or data related to obtaining observa- 1535 stars of theafih Fund.amental Catalogue (FKs). This volume is a'ailable from
irons. Verlag G. Braun, Karl-Friedrich-Strasse, 14 18, Karlsruhe, Germa.lv.
 SUPPLEIVENT
EXPLANATORY 14/ RELATED
PUBLICATIONS
670
671
Ephemerides of Minor Planets is prepared annually by the Institute of The- III Newcomb' simon "Measures of the velocity of Light N{ade u'der
of the U S S'R' Direction
oretical Astronomy, and published by the Acaderny of Sciences of the Secretary of the Navy During the years 1880 1882" 1885.
Included in this volume are elements, opposition dates, and
opposition ephemer- IV. Nfichelson, Albert A. ,,supplementary Measures of the Velocitics
of \\rhite
numbered minor planets. This volume is available from the Institute for and Colored Light in Air, Water, and Carbon Disulphid", l,I..tu
ides of all ,ith tl]
Theoretical Astronomy, 10 Kutuzov Quay, 191187 St. Petersburg. Aid of the Bache Fund of the National Academy of Sciences', 18g5.
v. Newcomb, simon "Discussion of observations of the Transits
of Venus in
1761 and 1769, 1890.
VI. Newcomb, simon "Discussion of the North porar Distances observed
with
the Greenwich and Wa_shingtonTransit Circles with Determtnatrons
PAPERSPREPARED
14.2 ASTRONOMICAL FORTHEUSEOFTHE Constant of Nutation', 1891.
of the
AND
EPHEMERIS
AMERICAN ALMANAC
NAUTICAL
Volume III
Astronomical Papers of The Arnerican Ephemeris ate available from the U'S Naval
I. Newcomb, simon "Deveropment of the perturbative Function and
Obseri.?tory, Washington, DC, 20392. Volumes I through X a.reout of print and no its Deriva-
tives, in Sines and Cosines of Multiples of the Eccentric Anomalies,
Ionger available. and in
Powers of the Eccentricities and Inclinations,' 1gg4.
Introduced in 1882, the series is published irregularly and contains reports of II. Hill, G.W. ,,Determinationof the Inequalitiesof the N{oon,s\lotion
Which
research in celestial mechanics with particular relevance to ephernerides' A full list are Produced by the Figure of the Earth: A Supplcment to Delaunay,s
of the papers published to date follows: Lunar Theory" 1884.
III. Newcomb, Simon ,,On the Motion of Hyperion. A New Case in Celestial
Mechanics" 1884.
Volume I IV. Hill, G.W. "On Certain Lunar Inequalities Due to the Action of .)upiter
and Discovered bv Mr. E. Neison" 1995.
V. Newcomb, Simon "Periodic Perturbations of the Longitudes and Radii Vec
I. Newcomb, Simon "On the Recurrence of Solar EcliPseswith Tables ofEclipscs tors of the Four Inner Planets of the First Order as to the Masses,' 1g91.
from g.c. 700 to A.D. 2300" 1879.
II. Newcomb, Simon, assisted by John N{eier "A Transformation of Hansen s Volume IV
Lunar Theory Compared with the Theory of Delaunay" 1880'
III. Michelson, Albert A. "Experimental Determination of the Velocity of Light Hill, G.W. "A New Theory of .)upiter and Saturn', 1890.
Made at the U.S. Naval Academy, Annapolis" 1880.
IV. Newcomb, Simon "Catalogue of 1098 Standard Clock and Zodiacal Stars"
Volume V
1882.
V. Hill, GeorgeW. "On Gauss'sN{ethod of Computing Secular Perturbations. I. Newcomb, Simon "A Development of the Perturbative Function in Cosinesof
with an Application to the Actiorr of Venus on N{ercury" l 88l ' Multiples of the Mean Anomalies and of Aneles Between the perihclia and
VI. Newcomb, Simon "Discussion and Results of Observations on Transits of Cornmon Node and in Powers of the Ecccntricities and Mutual Inclination',
Mercury, from 1677 to 1881" 1882. 1895.
II. Ncwcomb, Simon "Inequalities of Long period, and of the Second Order as
Volume II to the Masses, in the Mean Longitudes of the Four Inner planets" 1g95.
III. Newcomb, Simon "Theory of the Inequalities in the Motion of the Moon
Produced by the Action of the Planets', 189b.
L Newcomb, Simon, assisted by John Meier "Formulae and Tables for Express- IV. Newcomb, Simon "Secular Variations of the Orbits of the Four Inner plan_
ing Corrections to the Geocentric Place of a Planet in Terms of Symbolic ets" 1895.
Corrections to the Elements of the Orbits of thc Earth and Planet" 1883' V. Newcomb, Simon "On the Mass of Jupiter and the Orbit of polvhvmnia"
II. Safford, tuman Henry "Investigation of Corrections to the Greertwich Plan- 1895.
etary Observations from 1762 to 1830" 1883.
 EXPLANATORY
SUPPLEI\.,IENT '4 / RELATED
PUBLICATIONS
672 673
II. Robertson, James .Catalog of 3839 Zodiacal
VolurneVI Stars for the Equinox 1950.0,,
1940.
Tables of the four inner Planets.
Yolume XI
I. Newcomb, Simon "Tables ofthe Motion ofthe Earth on its Axis and Around I. Clemence, G.M. *The Motion of Mercury t76S
the Sun" 1895. 1987,, I9a3.
II. Clemence, G.M. ,,First-order Theory of-Mars,,
1949.
II. Newcomb, Simon "Tables of the Heliocentric Motion of Mercury" 1895. III. Morgan, H.R. ,,Definitive positions and proper
Motions of primary Refer-
III. Newcomb, Simon "Tables of the Heliocentric Motion of Venus" 1895. ence Stars for pluto,' 1950.
IV. Newcomb, Simon "Tables of the Heliocentric Motion of Mars" 1898. IV' Herget, Paul, cremence, G.M., and Hertz,
Hans G. ,,Rectangurar coordi_
nates of Ceres, Pallas, Juno, Vesta 1920 1960,, 1950.
Volume YII
Volume XU
I. Hill, George William "Tables of Jupiter, Constructed in Accordance with the
Methods of Hansen, and Prepared for Use in the Office of the American Eckert' w.J., Brouwer, Dirk, and clemence, G.M. ,,coordinates
of the Five
Ephemeris and Nautical Almanac" 1895. Outer Planets 165A,2060, 1951.
II. Hill, George William "Tables of Saturn, Constructed in Accordance with the
Methods of Hansen, and Prepared for Use in the Ofrce of the American Volume XIII
Ephemeris and Nautical Almanac" 1898. I. van woerkom, A.J.J. "The Motion of Jupiter's
Fifth satellite 1892 1949,,
III. Newcomb, Simon "Tables of the Heliocentric Motion of Uranus" 1898. 1950.
IV. Newcomb, Simon "Tables of the Heliocentric Motion of Neptune" 1898. II. Brouwer, Dirk and van Woerkom, A.J.J. ,,The
Secular Variations of the
Orbital Elements of the principal planets,, 1950.
Yolume YIII III. Morg-an, H.R. ,,Catalog of 5268 Standard Stars,
1950.0, Based on the Nor-
mal System N30" 1952.
I. Newcomb, Simon "A New Determination of the Precessional Constant with IV. Clemence, G.M. ,,Coordinates of the Center
the Resulting Precessional Motions" 1897. of Mass of the Sun and the
Five Outer Planets, 1g00_2060,,1953.
II. Newcomb, Simon "Catalogue of Fundamental Stars for the Epochs 1875 and V. Clemence, G.M. ,,perturbations of the Five
1900 Reduced to an Absolute System" 1899. Outer planets by the Four Inner
Ones" 1984.
III. Hedrick, Henry B. "Catalogue of Zodiacal Stars for the Epochs 1900 and
1920 Reduced to an Absolute System" 1905.
Volume XIV
Herget, Paul ',Solar Coordiuates 1g00-2000,'1953.
Volume IX

I. Newcomb, Simon "Resea.rcheson the Motion of the Moon. Part II. The Mean Volume XV
Motion of the Moon and Other Astronomical Elements Derived from Ob- L Woolard, Edgar W. ,,Theory of the R.otation of the
Earth Arouncl Its Center
servations of Eclipses and Occultations Extending from the Period of the of N{ass" 1953.
Babylonians until A.D. 1908' 1912. IL Hertz, Hans G. ,,The Mass of Saturn and the N{otion
of Jupitcr 1884 194g"
II. Ross, Frank E. "New Elements of Mars and Tables for Correcting the He- 1953.
liocentric Positions Derived from Astronomical Papers, Vol. VI, Part IV" III. Herget, Paul ,,Coordinates of Venus 1800 2000" 1955.
1917.
III. Eichelberger, W.S. and Newton, Arthur "The Orbit of Neptune's Satellite Volume XVI
and the Pole of Neptune's Equator" 1926.
I. Duncombe, Raynor L. ,,I\{otionof Vcnus t7b0 -1949',
1958.
II. Clemence,G.N,I.,,Theory of Mars-Cornpletion,,1961.
Yolume X III. Herget, Paul ,,Rectangular Coordinates of Ceres, pallas,
Juno, Vesta 1960
I. Eichelberger, W.S. "Positions and Proper Motions of 1504 Standard Stars 1980" 1962.
for the Equinox 1925.0" 1925.
674 EXPLANAToRY
SUPPLEMENT 14 / RELATEO
PUBLICATIONS
675
VolumeXVII 14.3 UNITEDSTATES
NAVALOBSERVATORY
CIRCULARS
Watts, C.B. "The NtlarginalZone of the Moon" 1963.
United States Naual Obseruatory Circulars are issued irregularly to dissemilate
astronomical data concerning ephemerides or astronomical phenomena.
Volume XVIII
Circulars are available from the U.S. Naval Observatory, Washington, DC 20392.
Jarnagin, Jr., Milton P. "Expansionsin Elliptic Motion" 1965.
Copies of the first 100 circulars are no longer available. Circulars from 101 onward
are availablewith the exceptionsof 103, 104, 109, 115, 11g, and 138. The content
Volume XIX of the first one hundred circulars can be summarized as follows:
I. Franz. Otto G. and \'Iintz, Betty F. "Tables ofX and Y- Elliptic Rectangular
Coordinates" 196,1. Subiocl Circular#
II. Eckert, W.J. and Smith, Jr., Harry F. "The Solution of the N{ain Problem
of the Lunar Theory by the Method of Airy" 1966. Sunspols,positions,areas, counts 3, 4, 6, 8-13, 15, 17, 19,21_26,2&_39, 41_4a,50_52,54_58,60-77,
7F84, 86, 87
Eclipses,1952-'1964 '1
, 2, 16, 27, 40, 53, 59, 78, 85, 88, 89
Minor PlanetEphemerides 5,7,14,20
Volume XX -Ilfi
Naval Observalory e Service 14,49
I. Jackson, Edrvard S. "Determination of the Equinox and Equator frorn Ilerid- Ephemeis of Mars 90, 95, 98
Coordinatesol the Moon 91
ian Observation of the Minor Planets" 1968. Radio Longiludoollhe Meridianot Jupiler 92,94
II. Duncombe, Raynor L. "Heliocentric Coordinates of Ceres, Pallas, Juno, GC and DM Numb€rs 93
Vesta 1928 2000" 1969. Distancsand V€locityof Venus 96
Polarisand Equalionot Time 97
III. O'Handley, Douglas A. "Deternination of the Mass of Jupiter from the MachineReadabloData 99
Motiorr of 65 Cybele" 1969. 100

Volume XXI
I. Janiczek,Paul N,L "The Orbit of Polyhymnia and the Mass of Jupiter" 1971. Unlted StataaNeval ObseruatoryClrcdats, *1O1onward
II. Fiala, Alan D. "Determination of the Mass of Jupiter lrom a Study of the
Motion of 57 Nlnemosyne" 1972.
III. Branham, Jr., Richard L. "The Orbits of Fir,e \linor Planets and Corrcc 101 S o l a rE c l i p s e s1, 9 7 1 - 1 9 7 5
tions to the FK4 Equator and Equinox" 1979. 102 Tolal Solar Eciipseol30 May j965
103 Observationsof the Sun, Moon, and planets/Six-lnchTransitCircle Rosults
104 Durchmusterungand Henry Drap€rNumbersol AtbanyGeneratCatatogStars
Volume XXII 105 Observationsof the Sun, Moon, and ptanets/SixtnchTransilCircteResults
106 RectangularCoordinatesof Mercury1800-2OOO
I. Cohen, C.J., Hubbard, E.C., and Oc.sterwinter,Claus "Elementsof the Outer 1O7 Sunlight,Moonlight,and Twitightlor Antarcljcaj966-1960
108 Observalionsof the Sun, Moon, and Ptanetvsix-lnchTransitCirct€ResLrlts
Planets for One Nlillion Years" 1973. 109 AnnularSolar Eclipseot 20 May 1966
II. Jackson, Edrvard S. "A Discussionof the Observationsof \epturc 1846 '110
TotalSolarEctipseof t2 Novomber1966
111 AslronomicalDala in Machin€BeadableForm
1970" 1971.
112 Phas€sot the Moon 180G..1959
III. Pierce,David A. "Star Catalog CorrectionsDeterrrrirredfrom Observations 113 Solar Eclioses.1976-1980
of SelectedMinor Planets" 1978. 114 AslronomicalDala in l\rachineBeadabteForm
IV. Laubscher,Roy Edward "The Nlotion of NIars 1751 1969" 1981. 115 Observationsof the Sun, I\,loon,and Ptanets/Six-tnchTransitCircteRosults
116 TotalSolar Eclipseol22 Sept€mber1968
1 17 Eph6m6risof lhe Radio Longiludeof lhe C€ntratMeridianot JupiterSyst€m I t (1957.0)
118 Observalionsof lhe Sun, Moon, and Planets/Six- Inch TransitCircteResutts
Volume XXIII
11I Phases of lhe Mcton 1960-2003
120 Sunlight,Moonlight,and Twilighllor Antarcticat969-1971
I. Gutzwiller, Nlartin C. and Schmidt, Dieter S. "Thc Motion of the Nloon as
121 A LinolronSystem Manuallor lhe Photo-Composition of Astronomicatand MathematicatTabtes
Computed by thc Method of Hill, Brown, and Eckert" 1986. i22 Annular Solar Eclipseot 18 lrarch 1969
123 AnnularSolar Eclipseol 1l Seplemb€r1969
676 EXPLANATORYSUPPLEMENT 14 / RELATED
PUBLICATIONS 677
Unllnd StateaNaval ObseNalory Ctrculars,conlinued
14.4 PUBLICATIONS
OFTHEUNITEDSTATES
NAVALOBSERVATORY
No. lllle SECONDSERIES
124 Obs€rvalionsoflhe Sun, Moon, and Planels/SixlnchTransilCircle Fesults
125 Tolal Solar Eclipseof 7 l\,,larch
1970 The Publicati,ons of the United States Naral Obsertatory, Second Series are issued
126 AnnularSolar Eclips€ol31 August-l S€pt€mber1970 irregularly to disseminate observational results. Thcsc publications, from Volume
127 Obsetuationsol the Sun, Moon, and Planels/Six-lnchTransil CircleResulls
128 AslronomicalData in MachineRoadableForm
XVII onward, are available frorn the U.S. \aval Observatory. Wa^shingtol. DC
129 AnnularSolar Eclipseol 16 January 1972 20392.
130 Perjgee and Apogee of the Moon 1959-1 999
'131 Total Solar Eclipseol 10 July 1972
132 Sunlight,Moonlighl,and Twilighttor AntarclicalgT2-1g74
Volume I
't3:l SunspotAreas 1907-1970
134 Nomaliz€d Obseruationsof Venus 1901-'1949 Harkness, W. and Skinner, A.N. "Transit Circle Observationsof the Sun,
135 TotalSolar Eclipseof 30 June 1973 Moon, Planets, and MiscellaneousStars, 1894 1899" 1900.
136 Obsetuationsof the Sun, Moon, and Plansts/Six-lnchTransitCircle Besulls
137 Ephemerisol the RadioLongitudeol tho CentralMeridianot Jupiler System lll {1957.0)
138 Geocenlricsolar Data Volume II
139 AstromelricEphemerisol Plulo 1970-1990
lzlo R€ctangularCoordinatesof lhe Moon 1971-1980 Skinner, A.N. "Zone Observationslrith the Nine-Inch Tr.ansit Circle, 1894
141 Lunar Ulnb Profilesfor Solar Ecliosos
1 9 0 1 "1 9 0 2 .
142 Solar Eclipses,1981-1990
I 43 Obs6rvationsof the Sun, Moon, and Planets/Six-lnchTransitCircl€ B€sults
144 TotalSolar Eclipseof 20 Jun€ 1974
145 NormalizedObservationsol Mercury1901 1937
Volume III
146 AstronomicalDala in MachineReadabl€Form
147 Sunfighl,Moonlight,and Twilightlor Anlarclica1975-1977
I. See,T.J.J. "Observationsof Eros with the Twenty-Six Inch Equatorial, 1900
148 PhysicalEphemerisol Mars 1901" 1903.
149 Fodran Aulomalic Typesetling systsm II. King, T.I- "Observations of Eros and Reference Stars rvith the \ine-Inch
150 Indexto Uniled Slales NavalObservatoryChcularsNos- 101-150
151 Appar€ntEphemerisof Mars 1960-19m tansit Circle, 1900 1901" 1903.
152 TotalSolar Eclipseot 23 October'1976 III. Eichelberger, W.S. "Observations of 495 Zodiacal Stars with thc Nine-Inch
153 Coordinatesof lJ.S-Naval ObssrvatoryInstallalions Transit Circle, 1900" 1903.
154 Th€ Dotominationol lJniversalTimeat th€ lJ.S. Naval Obssruatory
155 Almanaclor Compul€rs,1977
IV. Updegrafi, M. "Observationswith the Six-Inch Transit Circle, 1900 1901'
156 Solar Ecliosesol 1977 1903.
157 Tolal Solar Eclapseof26 February'1979 V. Ingersoll,R.R., Bowman, C.G., and Taylor, H. "Observationsrvith the Prime
158 Tolal Solar Eclipseol 16 F€bruary1980
I 59 Observationsol the Sun. Moon. and Planets/SixlnchTransatCifcle Results Vertical Transit Instrument, 1882-188,1"1903.
160 TotalSolar Eclios€ol31 Julv 1981
161 Ephem€dsol the Systom lll (1965) Longitudeof th€ Cenval Meridianof Jupiier
162 G€ocenlricEphemerisol Plulo 1975-1985 Volume IV
163 The IAU Resolutionson AstronomicalConslants,Time Scales,and the FundamentalFelerenceFrame
164 AstronomicalData in MachineReadableForm I. "Transit Circle Observationsof the Sun, Moon, Planets, and Miscellancous
'165
TotalSolar Eclipseot 11 Jun€ 1983 Stars, 1900 1903" 1906.
166 AnnularSolar Eclipseol30 May 1984
II. Eastman, J.R. "Ttansit Circle Observationsof the Sun, Moon, Planets, arrd
167 ProjeclM6rit Slandard
168 TotalSolar Eclipseof 22-23 November1984 Comets, 1866-1891" 190o.
169 Phasosof the Moon 2000-2049 III. Updegraff, M. "Observationswith the Six-Inch Transit Circle, 1901 1902"
170 Solar Ecljpses,1991-2000
1906.
171 CompulorPrognms for Sun and Moon llluminancewith ContingentTablesand Diag.ams
172 Tolal Solar Eclipseof l7-18 March 1988 IV. Appendices
173 Tolal Solar Eclips€of 22 Juty t990 L "Total Solar Eclipsesof May 28, 1900,and May 17, 1901" 1906.
174 TotalSolar Eclipseot 11 July 1991
II. Eichelberger, W.S. "Reduction Tables for TYansit Circle Observations"
175 AnnularSolar Eclips€of 1S-1 6 Januarv 199j
176 C€ntralSolarEclipsesol 1992 1906.
177 On High FrequoncySpectral Analysis oJ Very Noisy Slochaslic Procgssesand lts Applicationlo EFIP III. Frederick, C.W. "Reduction Tables for Equatorial Observations" 1906.
S€ies Delerminedbv VLBI and SLR IV, Hayden, Edward Everett "The Present Status of the Use of Standard
Time" 1906.
 14 / RELATED
PUBLICATIONS 679
678 EXPLANATORY
SUPPLEMENT
Volume XI
Volume V
,,MeteorologicalObservationsand Results, 1893 1902" 1903. Hammond, J.C. and Watts, C.B. ,,Results of Observations rvith the Six-Inch
Transit Circle, 1909 1918' 1927.

Volume VI
Volume XII
"Equatorial Observations,1893 1907" 1911.
Appendices L "Equatorial Observations,1908 1926" 1929.
I. "Miscellaneous Astronomical Papers by Nlembers of the Naval Observa- II. "Photographic Equatorial Observations, l9I2 1924 and Photoheliographic
tory staff" 1911. Observations, 1977 1927" 1929.
II. "Miscellaneous Reports on the Transit of Mercury of November 10, Appendix:
1894"1911. Littell, F.B., Hammond, J.C., Watts, C.8., and Sollenberger,p. ,,World
III. Horigan, Wm.D. "List of Publications Issued by the United States Longitude Operation of 1926 Results of Observations at San Diego
Naval Observatory, 1845 1908" 1911. and Washington" 1929.

Volume VII Yolume XIII

Eichelberger,W.S. and Littell, F.B. "Catalogueof 23521Stars between 13o35' Nlorgan, H.R. "Results of Observations with the Nine Inch Transit Circle,
and.45'25' South Declination for the Equinox 1850 from Zone Observations 1913 1926.Observationsof the Sun, Moon, and Planets. Catalogueof 9,989
Made at the United States Naval Observatory,1846 1852" 1911. Standard and Intermediarv Stars. Miscellaneous Stars" 1933.
Appendices
Volume VIII I. Vening Meinesz, F.A. and Wright, F.E. "The Gravity Measuring Cruise
of the U.S. Submarine S-21 with an Appendix on Computational Pro-
Littell, F.B., Hill, G.A., and Evans, H.B. "Vertical Circle ObservationsMade cedure" 1930.
with the Five-Inch Alt-Azimuth Instrument, 1898 1907" 1914. II. Lamson, Eleanor A. "Declinations of Stars Dcrived from Observations
with the Prime Vertical Transit Instrument, 1927 7925" 1932.
Volume IX III. "Total Solar Eclipses of January 24, 1925, January 14, 1926 and NIay
9 , 1 9 2 9 "1 9 3 2 .
I. Eichelberger, W.S. and Morgan, H.R. "Results of Observations with the Nine-
IV. Nlorgan, H.R. "Washington Observations of the N{oon, 1894 1922"
Inch Transit Circle, 1903 1911" 1920.
t932.
IL Eichelberger, W.S. "Observations Made with the Nine-Inch Transit Circle,
1903 1908" 1915.
IiI. Littell, F.B. and Eichelberger,W.S. "ObservationsN{ade with the Nine- Volume XIV
Inch Transit Circle, 1908 1911" 1918.
I. Hammond, J.C. "CataIog ol 3,52OZodiacal Stars Based on (Jbservations rvith
IV. Littell, F.B. and Morgan, H.R. "Observations Made with the Nine-Inch
the Six-Inch Transit Circle, 1928 1930 Reduced Without Proper Nlotion to
Transit Circle, 1912 1913" 1918.
the Equinox 1925.0" 1938.
Appendix:
II. N{organ, H.R. and Lyons, U.S. "Results of Observations on the Nine-Inch
Littell, F.B. and Hill, G.A. "Determination of the Difference of Longitude
Transit Circle, 1932 193.1.Positionsand Proper Motions of 1.117Reference
BetweenWashington and Paris. 1913 1914" 1918.
Stars in Declination -l0o to 20". MiscellaneousStars" 1938.
iII. Littell, F.8., Morgan, H.R., and Raynsford,G.M. "Vertical Circle Observa-
Volume X tions Made with the Five-Inch Alt-Azimuth Instrument, 1916 1933. Cat-
I. Hill, G.A. "Observations Made with the Prime Vertical Transit Instrument, alog of Declinations of Standard Stars. Declinations of the Sun, Mercury,
and Venus" 1938.
1893 1912" 1926.
II. "Total Solar Eclipsesof August 30, 1905, and June 8, 1918,with Aviators' IV. Morgan, H.R. "Proper Motions of 2,916 Intermediary Stars Mostly in Dec-
lination 5' to -30'" 1938.
Notes on the Total Solar Eclipse of September 10, 1923" 1926.
680 EXPLANATORY
SUPPLEMENT 14 / FELATED
PUALICATIONS
681
V. Watts, C.B., Sollenberger, P., and Willis, J.E. "\&brld Longitude Operation Volume XVIII
of 1933 at San Diego and Washington" 1938.
I. Franz, O.G., Gossner,J.L., Josties, F.J., Lindenblad, I.W., Mikesell,
A.H.,
Mintz, B.F., and Riddle, R.K. ,,photographic Measures of Double
St.r"ii
Volume XV 1963.
IIA. Mikesell, A.H. ,,Mechanical Improvement of the 26-Inch Refractor,,
I. Wylie, L.R. "A Comparison of Newcomb's Tables of Neptune with Observa- 1968.
IIB. Riddle, R.K. ,,A Study of the Optical properties of the 26_Inch
tion, 1795 1938" 1942. Refractor,,
1968.
II. Clemence, G.M. and Whittaker, G.C. "Observations of the Transit of Mer-
III. Worley, C.E. ,,A Catalog of Visual Binary Orbits" 1963.
cury November ll-I2, 1940" 1942.
IV. Walker, Jr., R.L. ,,Micrometer Nleasuresof 256 Double Stars,,
III. Wylie, L.R. "An Investigation of Newcomb's Theory of Uranus" 1947. 1966.
V. Strand, K. Aa. "Photovisual N{agnitude Differences of Double
IV. Burton, H.E., Lyons, U.S., Raynsford, G.M., Wylie, L.R., Browne, W.M., Stars,, 1969.
VL Worley, C.E. ,,MicrometerMeasuresof 1164 Double Stars,'
and Smith, J.L. "Observations of Double Stars Made with the 26-Inch 1962.
VII. Kallarakal, V.V., Lindenblad, I.W., Jostics,F.J., Riddlc, R.K.,
Equatorial of the U.S. Naval Observatory, 7928 44" 1917. N,{iranian,
M., Mintz, 8.F., and Klugh, A.p. ,,photographicN{easuresof Double
V. Morgan H.R. and Scott, F.P. "Results of Observations Made with the Nine- Stars,i
1969.
Inch Transit Circle, 1935 1945. Observations of the Sun and Planets. Cat-
alog of 5,446 Stars. Correctionsto GC and FK3" 19.18.
Volume XIX

Volume XVI I. Adams, A.N., Bestul, S.M., and Scott, D.K. ,,Resultsol Observatiors
X,iade
withthe Six-InchTransit Circle, 1949 1956.Observationsofthc Sun. N4oon,
I. Watts, C.B. and Adams, A.N. "Results of Observations Made with the Six- and Plancts. Catalog of b,965 Stars for 1950.0.Corrections to FK4. GC,
Inch Transit Circle, 1925 1941.Observationsof the Sun, Moon, and Plan- and N30" 1964.
ets. Catalog of 2,383 Stars for 1925.0.Catalog of 1,536 Stars for 1950.0. II. Adams, A.N. and Scott, D.K. ,,Resultsof Obsenations N,Iacle
Corrections to GC and FK3" 1949. u,ith the Six_
Inch Transit Circle, 1956 1962. Observations of the Sun. tr{oon, ancl plan
II. Watts, C.B. "Description of the Six-Inch tansit Circle. Instrumental De- ets. Catalog of 2,554 Stars for 1950.0.Correctionsto FK4, GC, arrd N30"
velopments,1932 1948" 1950. 1968.
IIL Watts, C.8., Scott, F.P., and Adams, A.N. "Results of Observations\tlade III. Adams, A.N., Klock, B.L., and Scott, D.K. ,,\,Vashington
with the Six Inch Transit Circle, 1941 1949. Observations ofthe Sun, Moon, \rleridian Obser_
r'ations of the }Ioon. Six-Inch Trausit Circle Results, 1925 1968' 1969.
and Planets.Catalog of 5,216Stars for 1950.0.Conections to GC and FK3"
t952.
Volume XX

\blume XVII I. Strand, X. Aa. "Thc 61-Inch Astrometric Reflector System" 1971.
II. Hoag, A.A., Priser, J.B., Riddle, R.K.. and Christ1,. J.W. ,,Installation.
i. Hall, John S. and \{ikesell, A.H. "Polarization of Light in the Galaxy as Tests, and Initial Performancc of the 61-Inch Astrometric Reflector,, 1967.
Determined from Observations of 551 Early-Type Stars" 1950. IIIA. Riddle, R.K. "First Catalog of Tligonometric paraliaxes of Faint Stars
II. Lyons, U.S. and Wylie, L.R. "Observations of Double Stars N4adeq'ith the Astrometric Rcsults,' 1970.
26-Inch Equatorial of the U.S. Naval Observatory, 1945 1948" 1953. IIIB. Priser, J.B. "First Catalog of Trigonometric parallaxes of Faint Stars
III. Hall, Jr., A., Burton, H.E., Lyons, U.S., Wylie, L.R., Raynsford, G.M., Photometric Results', 1970.
Browne, W.M., and Smith, J.L. "Observationsof SatellitesMade with the IIIC. Strand, K. Aa. and Riddle, R.K. ,,First Catalog of Trigonometric paral-
26-Inch Equatorial of the U.S. Naval Observatorx 1928 1947" 1953. laxes of Faint Stars Discussion,,1970.
IV. Mikesell, A.H. "The Scintillation of Starlight" 1955. IV. Ables, H.D. "Optical Study of Nearby Galaxies" 1971.
V. Markowitz, W. "Observations of Double Stars, 1949 52, and a Study of the V. Kron, G.E., Guetter, H.H., and Riepe, B.y. ,,A Catalog of Colorimetric
Optics of the 26-Inch Refractor" 1956. Measures of Stars on the Six-Color Systern of Stebbins and Whitford"
VI. Hall, J.S. "Polarization of Starlight in the Galaxy" 1958. r972.
VII. Hoag, A.A., Johnson, H.L., Iriarte, B.. Mitchell, R.I., Hallam, K.L., and VI. Routly, P.M. "Second Catalog of Trigouometric parallaxes of Faint Stars',
Sharpless,S. "Photometry of Stars in Galactic Cluster Fields" 1961. 1972.
682 EXPLANATORY
SUPPLEMENT
14/ RELATED
PUBLICATIONS
'UBV Sequencesin Selected Areas" 1974. 683
VII. Priser, J.B.
VI. Worley, C.E. "Micrometer Measures of 1,980 Double Stars,' 1928.
VII. Worley, C.E. and Heintz, W.D. ,,Fourth Catalog of Orbits of Visual Binary
Volume XXI Staxs" 1983.

Blanco, V.M., Demers,S., Douglass,G.G., and Fitzgerald, M.P. "Photoelectric

Catalogue. Magnitudes and Colors of Stars in the U, B, V and U", B, V Volume XXV
Systems" 1968.
I. Rydgren, A.E., Schmelz,J.T., Zak, D.S., and Vrba, F.J. ,,BroadBand Spec_
tral Energy Distributions of T Tauri Stars in the Taurus-Auriga Region',
Volume XXII 1984.
II. Walker, Jr., R.L. "Micrometer Measures of 711 Double Stars', 19g5.
I. Walker, Jr., R.L. "Micrometer Measures of 463 Double Stars" 1969.
III. Worley, C.E. "Micrometer Measures of 2589 Double Stars', 1989.
II. Worley, C.E. "Micrometer Measures of 1,343 Double Stars " 1971.
III. Harrington, R.S. and Mintz, B.F. "Positionsof Bright Minor Planets" 1972.
IV. Worley, C.E. "Micrometer Measures of 1,056 Double Stars" 1972. VolumeXXVI
V. Walker, Jr., R.L. "Micrometer Measures of 618 Double Stars" 1972.
VI. Josties,F.J., Dahn, C.C., Kallarakal, V. V., N{iranian,M., Douglass,G.G., L Scott, D.K. "Results of Observations Made with the Six-Inch Transit Circle,
Christy, J.W., Behall, A.L., and Harrington, R.S. "Photographic Measures 1956 1962. Catalog of 10,010AGK3R Stars for 1950.0.Comparison with
of Double Stars" 1974. GC and N30" 1987.
II. Hughes, J.A., Smith, C.A., and Branham, R.L. ,,Resultsof Observations
Made with the Seven-Inch Transit Circle, 1967 1973" Igg2.
Volume XXIII

I. Rhynsburger,R.W. and Gauss,F.S. "Catalog of Proper Motions for the 5,965

Stars of the Six-Inch Transit Circle Program, 1949 1956' 1975.
lI. Rhynsburger, R.W., Gauss, F.S., and Crull, Jr., H.E. "Motions and Phys-
ical Properties of 1,086 Cepheids and Early-Type Stars Observed in the
Washington Six-Inch Transit Circle Program, 1949 1956" 1980.
IIL Hughes, J.A. and Scott, D.K. "Results of Observations Made with the
Six-Inch Transit Circle. 1963 1971. Observations of the Sun, Moon, and
Planets. Catalog of 14,916 Stars lor 1950.0.Comparison with FK,l, GC,
and N30" 1982.

Volume XXIV

I. Harrington, R.S., Dahn, C.C., Behall, A.L., Priser, J.B., Christy, J.W., Riepe,
8.Y., Ables, H.D., Guetter, H.H., Hewitt, A.V., and Walker, R.L. "Third
Catalog of Trigonometric Parallaxes of Faint Stars" 1975.
IL Behall, A.L. "Micrometer Measuresof 267 Double Stars" 1976.
III. Dahn, C.C., Harrington, R.S., Riepe, B.Y.. Christy, J.W.. Guetter, H.H.,
Behall, A.L., Walker, R.L., Hewitt, A.V., and Ables, H.D. "Fourth Catalog
of Trigonometric Parallaxes of Faint Stars" 1976.
IV. Harrington, R.S., Dahn, C.C., Miranian, M., Riepe,8.Y., Christy, J.W..
Guetter, H.H., Ables, H.D., Hewitt, A.V., Vrba, F.J., and Walker, R.L.
"Fifth Catalog of Trigonometric Parallaxes of Faint Stars" 1978.
V. Josties,F.J.. Kallarakal, V.V., Douglass,G.G., and Christy, J.W. "Photo-
graphic N{easuresof Double Stars" 1978.
684 EXPLANAToRY
SUPPLEIjENT 1 4 / B E L A T E DP U B L I C A T I O N S o65
NAOTECHNICAL
14.5 SELECTED NOTESOF THEROYALGBEENWICH 14.6 LISTSOF APPENDICESAND SUPPLEMENTS
OBSERVATORY
14.61 TheBntishNautical
Almanac
'lhe NAO Techni.calNotes are published irregularly by the Royal Greenwich Ob- Many issues, especially the earlier years, of the almanac contain appendices on
various astronomical and navigational subjects. A list of these and of separatc sup_
servatory to disseminate technical information of speciflc interest.
plements, together with sections of a similar nature in the prefaces and explanations
No. Author to certain issues, is given belorv. Unaltered, or nearly unaltered. reprints in later
6 Hobden,
D.E. Computations for lhe Decca HyperbolicLattice 1966 issues are ignored.
System The appendices to the almanacs lor 1772 to 1778, 1787, 1788, arrd 179,1u,-ere
11 Sadler,
O.H. Interpolation
in H.O.299,SightReduction Tableslor 1966 collected and published in 1813 with the title Setectionsfrom the Ad.d.itions That
MarineNavigation
Haae Been Occasionally Annexed,to The Nautical Almanac from its Comm,encernent
12 Scott,W.A. Corrections, for lrradiation,to theObservedAltitude 1967
Sadler,D.H. of theSun to the Year -1879,while those to the almanacs for 1835 to 1854 *,ere similarly
18 Taylor,G.E. Predictions ol GrazingOccultations 1970 published irr 1851 with the title Append,icesto Various Nautical Alrnanar:s Bettueen
26 Sinclair,
A.T' The Representation of PlanetaryEphemerides for 1971 the Years 1837 and 1857.
use in the AutomaticReductionof Navigational
Sights
29 Taylor,G.E. TheVisualObservation of Occultations 1974 1766: Maskelyne, N., el al. "Tables Requisite to be Used wjtb The Asro-
3l Wilkins,G.A. The Future Publication of Astronomical 1974
Ephemerides nomical and Nautical Ephemeris." Separate publication. 166 pages.
36 A.T.
Sinclair, TheOrbitsol Tethys,Dione,Rhea,andTitan 1974 1769: Ntlaskelyne,N. "Instructions Relative to the Obserr,ation of thc En-
37 Sinclair,
A.T. TheTabulation of i/onthlySetsof Polynomial Coef- 1975 suing Transit of the Planet Venus over the Sun's Disk, on the 3rd of
ficientsfor Astronomical Datain the StarAlmanac
June 1769." 9 pages.
for LandSurveyors
39 Watson,F.G. The ZenithaiBlind Spot of a Large Altazimuth 1976 Ntlaskelyne.N. "Use of the Astronomical Quadrant in Taking Alti-
Telescope tudes." 38 pages.
40 Morrison,
L.V The Movements of theSun,Moon,andStarsin the '|976 7777: Douwes,C. and Campbell, J. "Tables for...Finding the Latitude of a
skv
46 Yallop,
B.D. Formulae lor Computing Astronomical Data with '1978 Ship at Sea." 77 pages.
Hand-HeldCalculators Maskelyne, N. "Determination lby John Bradley] of the Position of
47 Emerson, B. ApproximateSolar Coordlnates 1978 the Lizard." 6 pages.
48 Emerson, B. ApproximateLunar Coordinates 1979
53 Sinclair,
A.T. The Computationof PhysicalEphemeridesof Plan- 1980
Wargentin, P.W. "Tabulae Novae et Correctae Pro Supputandis Eclipsi-
ets and Satellites bus Tertii Satellitis Jovis..." 16 pages.
54 Yallop,B.D. Geocenlricand HeliocentricPhenomena 1981 1772: Maskelyne, N. "A Correct and Easy Nlethod of Clearing the fl-unar]
55 Yallop,
B.D. The Phase Correctlonfor Venus 1981
Distance..." 25 pages.
Emerson, B.
Emerson, B. ApproximateCoordinatesof Jupiterand Salurn 1981 "Eclipses of Jupiter's Third Satellite, and Tables of the Hour Angle
57 Yallop,B.D. Groundlllumination 1986 of the Sun and Jupiter." 6 pages.
Yallop,B.D. Formulae for Determining Carringlon's Elements 1982 Lyons, I. and Dunthorne, R. "Problems in Navigation." 8 pages.
and DifferentialSolar Rotation
59 Sinclair,
A.l The Effect of Atmospheric Refraction on Laser 1982 1773: "A Table of the Equations to Equal Altitudes." 24 pages.
RangingDala "A Catalogueofthe Placesof387 Fixed Stars,...(for 1760.0)."14 pages.
62 Yallop,
B.D. Coefficientsfor Calculatingthe cHA and DEC of 1985 1 7 74 : Mason, C. "Longitudes and Latitudes of the Moon, Deduced from
Hohenkerk, C.Y Stars
63 Hohenkerk, Dr. Bradley's Observations, Made Between September 13th 1750 and
C-Y The Computationof Angular Almospheric Befrac- 1985
Sinclair,
A.t tion at LargeZenith Angles November 2nd 1760,and Compared with a Set ofManuscript Tables."
64 Yallop,
B.D. Algorithmsfor Calculatingthe Dates of Easter 1986 36 pages.
Hohenkerk, C.Y Determinalionof Polynomial Coefiicientsfrom B- 1986 Maskelyne, N. "Elements of the Lunar Tables." 11 pages.
Spline Coefiicients
Maskelyne, N. "Remarks on the Hadley's Quadrant." 14 pages.
 14 / RELATEDPUBLICATIONS
686 EXPLANATORYSUPPLEMENT 687
1832: Airy, G.B. "Corrections of the Longitudes and Right Ascensions of
Lyons. I. "Astronomical Problems." 10 pages.
the Sun,.." 4 pages.
1778: Mason. C. "Right Ascensionsand Zenith Distancesof the \Ioon..."
Jenkins.H. "RecalculatedElementsof Delambre'sTables...ofJupiter,s
40 pages.
Satellites." 12 pages.
Lyons, I. "Astronomical Problems." 11 pages.
1833: Schumacher,H.C. "Ephemeris of...Lunar Distancesof Venus. Mars,
1779: Wargentin, P.W. "Tabulae Novae et Correctae Pro Supputandis Eclipsi-
Jupiter. and Saturn." 44 pages.
bus SecundiSatellitis Jovis..." 30 pages.
"Geocentric Places of the Planets." 75 pages.
t78L: Edwards, J. "Astronomical Problems." 27 pages.
1834: "Report of the Committee of the Astronomical Society of London.,'
Edn'ards. J. "Addition to the...Tables Annexed to the NauticaLAl-
ll pages.
manac of 1771." l0 pages.
1835: \\toolhouse.W.S.B. "New Tablesfor...Jupiter's Satellites...',39 pages.
1787: Edlv'ards,J. "Directions for \'laking...ReflectingTelescopesand...
\lbolhouse, W.S.B. "On the Computation of an Ephemerisof a Comet...,'
Polishing...Them.. ." 48 pages.
9 pages.
Edwards, J. and Maskelyne,N. "An Account of the...Tremors of Re-
"Comparison of. . . Burckhardt's and Damoiseau'sLunar Tables...',
flecting Telescopes..."12 pages.
4 pages.
1788: Blair. R. "Description of a...Nlethod of Adjusting Hadlel's Quad-
1836: \\:oolhouse. \\'.S.B. "On Eclipses." 96 pages.
rant." 20 pages.
1837: Airl', G.B. "On the Calculation of...Perturbations..." 23 pages.
1791: Nlaskelyne,N. "Advertisement of the Expected...Comet of...1788.
\\roolhouse,W.S.B. "On the Determination of the Longitude..."
and Relative to...Saturn's Ring in 1789 and 1790." 4 pages.
12 pages.
r794: Brinkley, J. "Tablesto Improve...the Method of Finding the Latitude.. ."
1839: Stratford, trV.S."On the Elements of the Orbit of Halley's Cornet..."
15 pages.
79 pages.
1798: Brinkley. J. "Tables to Improve...Latitude..." (Second edition. re-
1851: Adams. J.C. "On the Perturbations of Uranus." 29 pages.
vised and corrected), 16 pages.
1 8 5 3 1 9 1 4 : Each contains the elements and ephemerides of a number (from 4 to
1 8 0 9 1 8 2 1 : Each issue contains one or more catalogs of stars, some of Iongitude
36) of minor planets.
and latitude, some of right ascension and declination; the number of
1854: Challis. J. "On the Correction of a Longitude..." 23 pages.
stars varies between 9 and about 50.
1856: Encke, J.F. (trans. Airy, G.B.) "On a New Method of Computing the
1812: Pond, J. "On the Obliquity." 3 pages.
Perturbations of Planets." 33 pages.
1818: Brinkley. J. "Two Practical Rules for Reducing the Observed Distance
Adams. J.C. 'On \ew Tables of the \{oon's Parallax." 20 pages.
of the \loon from the Sun or a...Star..." 18 pages. ''Comparison of \Ioon's Places by Burckhardt's Tables rvith Similar
1862:
18221 Brinkley. J. "A Practical \Iethod of Computing the Latitude.'' 16 pages.
Ones by Hansen's Tables." 2 pages.
1822 1833: "Tables of. . .Refraction,. . ,of Second Differences, and, . .of Star Places."
1867: Breen, H. "Corrections...to...the [Tabulated]Values of the \'loon'e...
16-40 pages.
Parallax...1831 1839." 6 pages.
1 8 2 4 1 8 3 3 : "Elements of Occultations." 6 l7 pages.
t874: "Predictions for the Transit of Venus." 6 pages.
1826: "Rules for [Predicting] Occultations." 8 pages.
1881: Adams. J.C. "Continuation of... Damoiseau'sTablesofJupiter's Satel-
t827: Young. T. and Henderson, T. "Rules for [Reducing] Occultations.'
lites." 9 pages.
4 pages.
1883 1922: Between about 1850 and 1900, a series of "Nautical Almanac Circu-
1828 1833: Separately issued supplements for each of these years contained a
lars" was issued, mostly giving details and local predictions of total
number of quantities that were transferred, between 1832 and 1834,
eclipses.
to the pages of the almanac proper.
1883 1922: Newcomb. S. "Corrections. . . to Hansen's Tables of the Nloon." [These
1829: Lax, W. "An Easy Method of Finding the Latitude and Time at
contain longitude and latitude corrections only from 1883 to 1895,
Sea..." 23 pages.
those for right ascension and declination being included also from
1831: La-x, W. "An Easy N{ethod of Correcting the Lunar Distance..."
6 pages.
i 1896.]2 4 pages.
1897: "Approximate Placesfor 1900.0of 834 Zodiacal Stars..." 11 pages.
 EXPLANATOBYSUPPLEMENT 14 / RELATED
PUBLTCATTONS
688 6g9

1900: Downing, A.N{.W. "Continuation of...Damoiseau'sTablesof Jupiter's the Superintendcnt, sailing directions, wi'd and current charts, sprr:ificatio.s relat_
Satellites." 7 pages. ing to the new Nar,al Observatory, rrrathcrnatical and astrono[ri<,al tablcs, special
''Corrections to the Apparent Places...to Obtain Appareut Places reports, obselvations, circulars. and reports on the U.S. Naval Astronomical Ex
1901-1906:
Corresponding to the Struve-Peters Constants." Separate publica- pedition to thc Southern Hemispherc and the publicatious of thc Transit of venus
tions, 22 pages. Commission.
1907-1919: "Calculation of a Lunar Distance." 2 pages. Unlike nrost of the other national epheurerides. The American Ephemeris was
1907-7914: "Ephernerides for Physical Observations." 30 pages. never a medium for the publication of technical articles. The voluures from 1g55 to
1915: "Some Constants and Formulac." 4 pages. 1911, inchrsive, coutained an appendix, but ordinarily it compr.iscclonly the miscel
"Heliocentric lCo-ordinates]of... [Planets]." 68 pages ianeous tables rcgularly included every ycar, arrd the list of funclarner)tal constants
1915: "Corrections to...1532 Stars..." (Separate publication, by Pulkovo and tables used in preparing the ephemerides.In 1912,this list was transferredto
Observatory), 308 pages. the beginning of the volume. leaving thc se<:lionof miscellarreorrs tables at the erd.
1916: "Heliocentric [Co-ordinates]. . .of Venus." 82 pages. and the appendix was discontinued. Oct iuionally. appendices containing r.arious
1917: "Heliocentric [Co-ordinates]. - .of N{ars." 82 pages. ephemerides havc also been added to irrdividual volumes; but r.err. ferv technical
' contributions \rere e\rer included.
1918: "Derivation of Quantities Containedin the Nautical AImanac. 23 pages.
1920: "Ross'sCorrectionsto...Placesof N{ars..." 16 pages. Hou.ever, scparate supplements to Tie Am,erican Epherult.ts lrtrve beel issueci
L929: "Coordinatesof the Sun for 1950.0for 1928 and 1929." 32 pages. from time to time. Sornetimesthese tr,'ercseparatc printings of tnatt:riaJ lrom thc ap_
1931: Fotheringham,J.K. "The Calendar." 14 pages. pendicesbut rnoreoften they were in addition to the conterltsof tlxr volumes.Most
"Derivation of Quantiiies..." 26 pages. of them contain supplcmentary epherrreris data, especially lbr total sol:rr eclipses.
"Tables for Interpolation. . .by the End-Figure Process." 32 pages. but some havc also bcen several important lechnical publicatiols.
1935: Fotheringhan, J.K. "The Calerrdar." 17 pages. The appcndi<:csand supplemerrts thzrt are of interest for their tcchnical content.
"Interpolation Tables." 14 pages. other thau qrhonrcral data. are listo<lirr St:ction 14.621.A list of the supplenents
1936: "Interpolation and AIlied Tables." (Reprinted for scpalate sale) ,18pages. giving extcncled data and large sctrlc rnaps for total solar eclipsos is given in Sec
1938: "The Prediction and Reduction of Occultations." (Scparatepublica- tion 14.622.
tion) 50 pages. Prior to tlrc t'stablishment of the scrios of AstronrnticrLl Palx:rs Prepared for
1938: "The Total Solar Eclipse of 1940 October 1." (T'ypescript) 19 pages. the Use of Th.r:Anteri,can Ephemcris arul Nautical Almono,c (set: S(\.tioD 1,1.2for
1940: "Heliocentric Co-ordinates of Mcrr,ury." 4 pages. a complete list of contents), the principal tables constructed lirr' t.hc olice were
"Corrections FK3 - Eichclberger." 6 pages- printed, each as a single publication (a list of these is given in Scctiol 14.623);but
1941: "Occultation Reduction Elcrrrerrts...' (Separatepublication by Yale many tables rvcro plepared onlf in marurscript.
U n i r c l s i t . r( ) h s e r \ a t o r \) 3 7 p a g e s .
1950: "Ephernerisof Pluto." 2 pages.
'l'hc
1954: "Implovcd Lunar Ephemeris 1952 1959." (Separatcprrblicalion as a 14.621 Appendicesand Supplements lollorving are aplx.ldices a l sup-
"Joint Supplement to The Anrcricarr Ephemeris and Thc (tsritish) plem, rrt' t,' The Arn'ricon Ephentrtr.,:
Nauticrl Alrnanac" ) 435 pagcs.
1855: Clltaurvr:rrct's
tables for rorrccting lrrnar distanccs,with dircctions lor
ttsilrg thc tables, and explilnation of their <ronstnrctiol.Pages 13 7{)
14.62 TheAmerican
Eohemeris
of the appendix.
Appendix III of Vrlume VI of the Publications of the U.S. Naral Obst'rvatorl',Sec- 1 8 57 : Chaur,enet'stablcs for corrccting lunar distances,u'ith dircctions lor
ond Series,corrtainsa list by \\'illiam D. Horrigan of publicatious issurrl bl the U.S. ttsirtgthc tables. and explarration()f their constr.ll(.tion.Pages11 67.
Naval Observatory fiom 18.15to 1908. This lisls the publications of lhe Depot of Chauvenet. \\:. "Improved \lcthod of Finding t]rc Error and Rate of
Charts and Instrunlents, the annual volurrnrs of the \Vashingtorr Obst rvations, the a Chronometer by Equal Altitudes." Pages69 9.1.
Publications "Second Scrics,'! appcndiccs to tLc \'{hshingtol Observatiors, appen- \\ralker, S.C. "Logarithrls of the LeVerrier Coeflicicnts of the Pertur-
dices to the Publications "Second Series," ructcorological observatiorrs, reports of bativr: Function of Planetary Motion." Pages95 117.
690 EXPLANATORY
SUPPLEMENT 14 / RELATED
PUBLICATIONS
691
Pages 11 94 of this appendix were later reprinted as a separate publi- In addition, data for two other eclipseswere issued in Il,S, Naual ObseruatorECircu_
cation. (Chauvenet, W. "New Nlethod of Correcting Lunar Distances, lar no. 27 and U.S Naual Obser-uatoryCircular no.78. The former gives information
and Improved Method of Finding the Error and Rate of a Chronome- on the annular eclipse of 1951 September l, while the latter gives inforrnation
on
ter by Equal Altitudes." 1866.) the total eclipse of 1959 October 2.
1874: Coffin, J.H.C. "Tables for Finding the Latitude of a Place by Altitudes
of Polaris." (Supplement for 1874 1877).
14.623 Thbles Prepared for Th€ American Ephemeris and Nautical Almanac The
This article, which included the formulas from which the tables were
following tables were printed by the U.S. Naval Observatorv:
calculated, with instructions for using the tables, and an illustrative
example, was also put into the Appendix (pages 25 33) in The Amer-
Schubert, E. "Tables of Melpomene." 1860.
ican Ephemeris for 1877, the first volume in which these tables were
"Almanac catalogue of zodiacal stars." 1864.
glven.
Winlock, J. "Tables of Mercury." 1864.
A similar supplement for the years 1878-1881, inclusive, was sepa-
Peirce, B. "Tables of the Moon." 1st ed. 1853;2nd ed. 1865.
rately printed, and was also included in the Appendix in The Amer-
Schubert, E. "Tables of Eunomia." 1866.
ican Ephemeris for each of these years. In the volume for 1882, these
Schubert, E. "Tables of Harmonia." 1869.
tables were replaced by a simple one-page table which was retained
Schubcrt, E. "Tables of Parthenope." 1871.
until it, in turn, was replaced by the table that was given throughout
"Tables to facilitate the reduction of places of the fixed stars." 1st
the period 7972 7959.
ed. 1869;2nd ed. 1873.
1895: Newcomb. S. "The Elements of the Four Inner Planets and the Fun-
Hill, G.W. "Tables of Venus." 1872 (on title page; cover has 1873),
damental Constants of Astronomy." (Supplement for 1897.)
Todd, D.P. "Continuation of de Damoiseau'stables of the satellitesof Jupiter
1945: "Tables of Sunrise, Sunset, and Twilight." (Supplement for 1946.) to 1900." 1876.
1950: "Ephemerisof Pluto."
1954: "Improved Lunar Ephemeris 1952 1959." Published as a "Joint Sup-
plement to The American Ephemeris and ?he (British) NauticaL AI- .14.63 JointSupplements
andAppendices
manac.t'
1968: The 1968 editions of The Astronornical Ephemeris and. Nau.ticalAIma-
14.622 Supplementary Publications for Solar Eclipses The series of supplements nac and The American Ephemeris and Nautical Almanac corrlain The
to The American Ephemeris that were published for the occasions when total Supplementto the A.E. 1968, "'Ihe Introduction of the IAU Svstem
eclipses of the Sun u'ere visible in the United States began with a supplement of Astronomical Corrstants." The supplement gives the values for the
to the volume for 1869. Prior to that, pamphlets had been issued for the annular astrononrical constants. the theory of the corrections, and tlte changes
eclipseof 185,1N{ay 26 and the total eclipseof 1860 July 17. ft The Explanatory Supplement that result from these changes.
For the total eclipse of 1869 August 7, a supplement was issued containing 1968 1971: These editiorrs of Tlte Astronomical Eplrcmeris and,Nautical Alrrtanac
predicted data, and also one containing suggestions for observing the eclipsc, In and The American Ephemeri,s and Nautical,4lmanoc corrtairr appen-
1885, a publication containing reports of observations of this eclipse was issued by dices with the forrrrulas and corrcctiorrs for conversions to the IAU
the Nautical Almanac Office. system of astronomical constants. The ephemerides irr thc volume
Since the eclipse of 1869, supplements have been published for the total eclipscs lrrre not bascd on the new constants, so the appendices were neces-
that occurred on the followine dates: sary to give the corrections for the ephemerides to be on the system
of constants.
1878 July 29 1936 June 19 198,1: The 1984 edition ol The Astronomi,cal Almanac contains Supplement
1900 May 28 .t940Ocl ohr.r I to the Astronornical Almanac 1981. "The Introduction of the Im-
1918 June 8 1945 July 9 proved IAU System of Astronomical Constants, Time Scales and
1925 January 24 1947 NIay 20 Reference Frame into The Astronomical Almanac." This supplement
1932 August 31 1954 June 30 gives the various resolutions involved and the resulting equations that
692 EXPLANATORY
SUPPLEMENT

introduce the IAU (1976) System of Constants, the FK5 reference

frame on J2000.0, and the TDT and TDB time systems.

CHAPTER
15

ReferenceData

Table 15.1
FundamentalConstants(1986 RecommendedValues)

GeneralConslanls

Quantlty Symbol Value Units Relative

(ppm)
uncertainty

UnlveraalConsiants
speedof light c 299792458 ms-r (exact)
permeability
of vacuum 4r x 1o-7 NA 2
'12,566370614... (exact)
1o-7NA-2
permittivity
of vacuum,1 / poC 8.854187817... 10-12Fm-1 (exact)
Newtonian constantof gravitation 6.672s9(95) 10 11m3kg 1s 2 128
Planckconstant h 6.6260755(40) 10 sJs 0.60
in electronvolts,h/ {e} 4.1356692(12) 1O-15Js 0.60
h/2r 1.054s7266(63) 10-sJs 0.60
in electronvolts,fi/ {e} 6.s921220(2ol 10-16 Js 0.60
Planckmass,(hc/ G)1/2 mp 2.17671(14) 10-8 kg 64
Plancklength,h/ mpc lp 1.61605(10) 1o-35m 64
Plancktjme,/p/ c tp 5.39056(34) 10-ss 64

ElectromagnelicConstanls
elementary
charge e (49)
1.60217733 1 o - l ec 0.30
e/h (72)
2.41798836 1 O 1 4 A J1 0.30
fluxquantum,/l/ 2e
magnetic o (61)
2.06783461 1 0 - 1 5w b 0.30
2e/h (141
4.4359767 10l4HzV-l 0.30
quantized
Hallconductance,
zot / tloc e/h (17) 1 0 - 5 A V 1
3.a74046i9 0.045

Table 15.1 is continued on nexl page.

693
 E X P L A N A T O RSYU P P L E M E N T
694 15 / REFERENCE
DATA oYc

Table 15.1,continued
Ili!?lt;li"i?l"1iXi,,?is(les6Recommendedvarues) Fundamental Constants(1986Recommended
Values)
Ouantity Syrnbol Value Units Relative
(ppm)
uncedalnty Quantity Symbol Value Units Relative
uncertainty(ppm
(311
9.2740154 1 O 2 4J T I 0.34
}ofi magnelon,ehl2me PB
1 o 1 sm
(52)
5.78838263 10 5 evr-1 0.089 prolon Comptonwavelength, 1.32141002/.12) 0.089
in electronvolts, la / {e} (19)
(42)
1.39962418 1 o 1 oH z T 1 0.30 ).6.p| 2tr 2.10308937 1 o - 1 6m 0.089
in herlz, 1ts / h '|.41060761
(57)
0.6717099 KT-1 8.4 proton magneticmoment 147) 1 O - 2 6J T - 1 0.34
in kelvins, pe I k
(14)
46.686437 m-rTr 0.30 in Bohrmagnetons, 1.521032202(15) 10 3 0.010
in wavenumbers'lalhc
in nuclearmagnetons, 2.792847386 (36) 0.013
nuclearmagneton,efi./ 2mp lN s.0507866 (17) 1 0 - 2 7J T 1 0.34
i n e l e c k o nv o l t s , ,N / { e } 3.15245166 (28) 1 o 5e v T - 1 0.089 diamagnetic shielding
in F,eftz,pN I h 7.622s914 (231 MHZT I 0.30 correctionfor protons
in wavenumbers,plv / nc 2.54262281(77) 10 2m 1T-l 0.30 In pure waler,
3.658246 (31) 10 4 KT 1 8.4 sphericalsample,25'C 2s.689(1s) 1o-6
in kelvins,/rN / t(
shieldedproton moment
AtomlcConstants (H2O,sph.,25'c) uo 1.41057s1
4 l47l 1 0 2 6J T 1 0.34
in Bohr magnetons, P b t P e 1.520993129(17) 10 3 0.011
fine-skuctureconstanl .} 7.29735308(33) 1o-3 0.045
0.045 in nuclearmagnetons, P;tPN 2.792847386\63) o.o22
137.03s9895(61)
(13) 0.0012 proton gyromagneticratio 267522128(81)' 0.30
Rydbergconstant R- 10973731.534
B-c 3.2898419499(39) 1 0 1 5H z 0.0012
Physico-chemicalConstanls
R*hc 2.1798741\13) 10 r8J 0.60
in eV R-hcl {e} 13.6056981(40) 0.30 Avogadroconstant 6.0221367(36) 1 o 2 3 m o l1 059

Bohrradius,a/4?r8- aO o.529177249\24) 1 o l om 0.045 atomic (unified)mass unit,

Hartreeenergy,e /4Teoao 4.3s97482 (26) 0.60 atomic mass constanl
-, in eV Eh / {e} 2 7 . 2 1 1 3 9 6(18 1 ) 0.30 1u=mu=1112m112C) 1.6605402(10) 1o 27kg 0.59
10 am2s 1 0.089 in electronvolls, rnuC / {e} 931.49432(28) 0.30
quantumof circulation hl2me 3.63694807(33)
muonmolarmass M(p) 1 . 1 3 4 2 8 9 1( 31 7 ) 10 4 kg/mol 0.15 Faradayconstant 96485.309(29) cmol r 0.30
(36) 1 O - 1 0J s m o l - 1 0.089
muon magneticmoment 4.4904514(15) 1o 26JT1 0.33 molar Planckconslanl 3.99031323
4.84197097171) 1o-3 0 .t 5 0 . 1 1 9 6 2 6 5(81 r ) Jm mol 1 0.089
in Bohr magnetons,
8.890s981(13) 0.15 gas constant 8.314510(70) J mol r K-r 4.4
in nuclearmagnetons,
Boltzmannconstant,,g / NA k 1.3806s8(12) 1o 23JK r 8.4
muon magneticmoment anomaly,
ur\entZmt,)-1 ap (84)
0.001165923 7.2 in electronvolts, kt{e} 4.617384\72) 10 5 eV/K 8.5
1 1 6 0 4 . 4 (51 0 ) KeV 8.5
muongJactor,2(1+ ap) sp 2.002331846\17) 0.085 {e}ik
muon-protonmagnetic molarvolume(idealgas)
T = 273?15K;p = 101325Pa 22.41410119l Umol 8.4
moment ratio 3.18334547(47) 0.15
proton mass 1 0 2 7k g 0.59 Loschmidtnumber,NA / Ym nO 2.686773\23) 1025m 3 8.4
1.6726231(10)
1.007276470112) 0.012 r = 273915K; p = 100kPa VN 22.711O8/.19) Umol 8.4
0.30 Loschmidtnumber,NA / Vm ,o 2.651625l22l io2sm 3 8.4
in electronvolts,npC / {e} 938.27231(28)
(37) 0.020 Stefan-Boltzmannconstanl 5.67051(19) 10 8wm 2K-4 34.
Proton-eleckonmass ratio 1836.152701
proton-muonmass ralro 0.15 first radaationconstant,2rhC C1 3.7417749\22) tO 16Wm2 0.60
La802444 (13)
second radiationconstant,hcl k 02 0.01438769(12) mK 8.4
Protonspecificcharge et mp 95788309(2S) C/k9 0.30
prolon molar mass M(pl 1.007276470(12) 10 3 kg/mol o.012 Wein displacementlaw constant,
b = \rr'a)<T= c2 | 4.96511423 ... b 2.a977s6 (24) 10 3mK 8.4

jn
Notei The digits in parenthesesare the one standardd€vialionuncertainty lh€ last digits ot the givenvalue.
 EXPLANATORYSUPPLEMENT 15 / REFEFENCEDATA
696 697
Table15.2 Table15.2,continued
tAU(1976)Systemof Astronomical
Constants IAU(1976)Systemof Asironomical
Constants

17. Ratioof massof Sunto thatof rheEarth (GS)/(GE)=S/E=932946.0

Unlts: . The unilsmeter(m),kilogram(kg),andsecond(s) are the unitsol length,mass,andtame
Systemot Units(Sl).
in the lnternational [332946.038...]
. Theastronomical unitot timeis a timeintervalof oneday(D)ol 86400seconds.An interval 18. Raliool massol Sunto thatof (s/ F) / (1+ p)=3289005
of 36525daysis oneJuliancenlury Earth+ Moon [328900.55]
. Theastronomical unitof massis the massof the Sun(S). 19. l\rassof the Sun ( G S ) /G = S = 1 9 8 9 1x 1 O 3 o k g
. Theastronomical unitof lengthisthatlength(,4)forwhich theGaussion gravilational constant 20. Systemol planetary masses
(k) takesthe value0.017202098 95 whenthe unitsof measurem€nt are the astronomical Batiosof massof Sunto massesof the olanets
unilsof length,mass,andtime.Thedimensions of k2arethoseof theconstantof gravitation It ercury 6023600 Jupiter i047.355 [1047.350]
(cl, i.e.,L3M-1T-2.rhe term"unitdistance" is alsousedforthelengthA. Venus 408523.5 Satum 3499.5 13498.01
. In the preparationof the ephemerides and the fittingof the ephemerides to all the obser- Earth+ Moon 328900.510.551Uranus 22A69
vationaldataavailable,it was necessary to modifysomeof the constantsand planetary iZZeAq'
Mars 3098710 Neptune .t9314
masses.Themodiliedvaluesottheconstants areindicated in bracketsfollowing the (1976)
Pluto 3000000 [130000000]
Systemvalues.
OtherOuantiliestor UseIn the prepara on of Ephemerides

]-tl"-I9991la!_."9tlar
or newepnemenoes.
thevatuesgivenin thefo owingtistshoutdnormafly be usedin theprepararion
Oetlnin9conslants: 21. Massesof minorDlanets
1. Gaussiangravitatjonalconstant k= 0.01720209895 Minorplanet Massin solarmass
2. Speedol light c= 299792458ms I ( 1 ) C e r e s 5 . 9x 1 0 - 1 0
(2)Pallas l.1x10-10[1.0814x1 r 00]
Prlmaryconstanls: (4)Vesta 1 . 2x 1 0 - 1 0 1 1 . 3 7x8170 1 0 1
3. Light-time
for unitdistance rA =499.004782s
22. Massesof satellites
[499.00478370...] Planet Saleltite Satellite/Planet
4. Equatorial
radiusfor Earth ae = 6378140m Jupiter lo 4 . 7 0x 1 0 - 5
IUGGvalue ae = 6378136m Europa 2 . 5 6x 1 O - 5
5. Dynamical
form"factorfor Earth Jz=O 001O42626 Ganymede 7 . 8 4x 1 0 s
6. Geocentricgravitationalconstant GE= 3.986005x 1or4mos-2 Ca isto 5.6x10 s
[3.98600448...x 1014] Saturn Titan 2 . 4 1x 1 0 a
7. of gravitation
Constant G = 6 . 6 7 2 x 1 0 1 1m 3k g 1 s 2 Neptune Triton 2 x 10-3
11]
[6.67259x 10 23. Equatorial radiiin km
8. Ratioof massof Moonto thatot Earth P=0 01230002 Mercury 2439 Jupiter 71398 Pluto 25OO
[0.012300034] Venus 6052 Saturn 60000
9. General precessaon in longitude,per P=502910966 Earth 6378.140 Uranus 25400 Moon 1738
Julian century at standardepoch lrars 3397.2 Neplune 24300 Sun 696000
J2000.0 24. Gravityfieldsol planets
10. Obliquityofthe
ecliptic, epoch
at standard e= 23'26'21! 44a 123" 26'21:4119) Planet Jz Js Ja
J2000.0 Earth +0.00108263 -0.254 x 10 5 - 0 . 1 6 1x 1 0 s
Mars +0.001964 +0.36 x 10-a
Dsrlvsdconstants:
'11. Jupiter +0.01475 0.58 x 10 3
Constantof nulation,at standardepoch N=912025 Saturn +0.01645 -0. 10 x 10 2
J2000.0 Uranus +0.012
12. Unitdistance cr1 = A= 1.49597870x 10rr m Neptune +0.004
x 10rr]
[1.4959787066 \Ma.s. C22= 0.000055, S22= +0.000031, .%1= +0.000026)
13. Solarparallax arcsin (as / A) = tro=81794148 25. Gravityfield ot lhe Moon
18.7941441 1=lB - A\ | C =0.0OO2278 C / MR2=O 392
14. Constant of aberration,for stanoaro n=20!49552 B=(C A)/a=0 0006313 r =5552':7 = 1432'32!7
epochJ2000.0 C2a= O.O0O2O27 Cso= -0.000006 Ca2=+0.0000048
15. Flattening
factorlor the Earth l=0.00335281 q2=+0.0000223 Q1 =+0.000029 q2 = +0.OOOOOIT
=1 t298.257 qr =+0.000004 Ca3=+0.0000018
16. Heliocentricgravitationalconstanl F k 2 t e = G S = 1 . 3 2 7 1 2 4 3 8 x1 0 2 0m 3 S - 2 si33= 0.00oo01
11.32712440.. x ldol
 15 / REFERENCEDATA
EXPLANATOBY
SUPPLEMENT 699
698

Table15.3,continued
Table15.3 TimeandStandardEDochs
TimeandStandardEPochs
Standard Epochs
1 day= 24 hours= 1440minutes= 86400seconds
'| Julianyear= 365.25days= 8766hours= 525960minutes= 31557600
seconds JulianYoarepochs
J.Y Juliandate
Length ot lhe Yearat 1990 J1900.0 2415020.0
d hm J 1950.0 2433282.5
Tropical(equinox to equinox) 365.2421897 365 05 4A 45.19 J2000.0 24s1545.0
'10
Sidereal(fixedstar to tixed star) 365.25636 365 06 09 J2050.0 2469807.5
Anomalistic (perihelion to perihelion) 365.25964 365 06 13 J2100.0 2498070.0
Eclipse(Moon'snodeto Moon'snode) 346.62005 345 14 52
1900 January0.5= JD 2415020.0
caussian(Kepler'slawfor a = 1) 365.25690 06 09 56
06 00 00
1925 January0.5 = JD 2424151.0
Julian 365 '1950 January0.5 = JD 2433282.0
2000 January0.5= JD 245i544.0
Lengnhol the Month
o d hm s 2050 January0.5 = JD 2469807.0
Synodic(newmoonto new moon) 29.s3059 29 12 44 03 2100 January0.5= JD 24a8069.0
Tropical(equinox lo equinox) 27.32158 27 07 43 05 Beginning year
of Besselian
Sidereal(tixedstar to fixedstar) 27.32't66 27 07 43 12
Anomalistic (perigee lo perigee) 27.55455 27 '13 18 B.Y JulianDate B.Y JulianDate
Draconic(nodeto node) 27.21222 27 05 05 81850.0 2396758.203 82000.0 2451544.533
81900.0 2415020.31382025.0 2460575.58a
Lenglh of the Day 81950.0 2433282.423 82050.0 2469806.643
41975.0 2442413.47A 8'2100.0 2488068.753
1dot meansolartime = 1900273790935
of rneansiderealtime
= 24h03m56:555368
of m€ansiderealtime
= 86636.555368
meansiderealseconds
ld of meansiderealtime = 0!99726956633
of meansolartime
= 23h56n04909054
of meansolartime
= 86164.09054
meansolarseconds

j
 15 / REFERENCEDATA 701
Table 15.4
Sun, Earth, and Moon
Table15.4,continued
IAUSystem Best Estlmate Sun,Earth,andMoon

Sun IAU System Best Estimate

Radius 6 . 9 6x 1 O 8 m Orbit ol Moonaboutthe Earth

'|
Semidiameter at meandislance 5'59163= 959163
1.9891x 1030kg Siderealmeanmotionol Moon 2.661699489 x 10 6 rads 1
Mass
1 . 4 19 c m - 3 l'reandistanceof MoonfromEarth 3.844x 10skm
Meandensity
Surfacegravity 2.74 x 102ms-z = 27.9g 60.27Earthradii
lnclinationot solar equator to 7015' 0.002570AU
ecliPtic Eauatorialhorizontaloarallax 57'021608
Longitude olascending node(f in 75'46' + 84' T at meandastance 3422':6Oa
centuriesf romJ2000.0) Mean dastanceof center of Earth
Periodof synodicrotation(d = 26.90+ 5.2 sin2d days from Earth-Moon barycenter 4.671x 'lO3km
latitude) l/|ean eccentricity 0.05490
Periodot siderealrotation
adopted 25.38 days l\.4eaninclinationto ecliptic 59 145396
tor heliographiclongitudes Meaninclinationto lunarequator 6'41'
irotionrelataveto nearstars Limitsof geocentricdeclination t 29"
apex:a=271o6=+30o Saros= 223lunations = 19passages
of Sunthroughnode= 65851/3days
speed:1.94 x 104m/s = O.0112AU/d Periodol revolutionof node 6798d
Periodot revolutionof perigee 3232d
Flgureand GravityFieldof the Eanh Meanorbitalspeed 1023ms-1 = 0.000591AU/d
Meancentripetalacceleration 0.00272ms 2 = O.OOO3g
Equatorial
radius a = 6378140m 6378136m
Dynamicalform factorfor Earth 4 = 0.00108263 O . 0 0 1 0 8 2 6 2- 62 . 8 x 1 O 1 1y r - 1 Rotation ol the Earth
Flattening f=1/298.257
Polarradius b = 6356755m 6356752m Periodwithrespectto fixedstars
Massof the Earth 5.9742 x 1o24kg 5. 9742 x 1024k9 in meansiderealtime 24h00.00:0084
Meandensity 5.5ZECm- in meansolartime 23h56'04:0989
j
Normalgravity(g) 9.80621- O.02593 cos2d+ O.OOOO3cos4d m/s2 Rateof rotation 15:'04106717866910s-
2 3s 7 . 2 9 2 1 1 5 1x 01 O 5 r a d s1
Geocentricgravitational
conslanl 3 . 9 8 6 0 0x51 0 1m
4 3s - 2 3 . 9 8 6 0 0 4 4 0 x 1 0 1 4 m
Annualaatesof orecession(f in
Fora point on the spheroidof the IAUSystemat geodeticlatitude(4): centuries
fromJ2000.0)
generalprecessionin longilude 50':290966
+ 0]0222226
T
1' of latitude 1 1 0 . 5 7+51 .1 1 o s i ndzk m precession
lunasolar in longitude 50:387784
+ 010049263
f
1o of longitude ( 1 1 1 . 3 2+00 . 3 7 3s i n 2d ) c o sd k m planetaryprecession 0.0188623- 010476128
f
Geodeticlatitude(C) - geocenkic 692:74sin2O - 1:'16sin 4d
latitude(d/) Moon
Orbitot the Earlh
Meanradius 1738km
Solar parallax 8':794148 a':794144 Semi-diameter at meandistance 15',3216
Constantol aberration(J2000.0) 20:49552 l,Iass 7.3483x 1022kg
Light-timefor 1 AU 499.004782s 499.00478353s irean density 3.34gcm 3
'1.49597870
1 astronomicalunit of length x 1011m 1.4959787066x 1011m Surfacegravity 1.62ms 2 = 0.179
Mass ratio-Sun/Earth 332946.0 332946.045
Sun/Eadh+ Moon 328900.5 328900.55
Moon/Earth 0.0123002 0.012300034
Mean eccentricity 0.016708617
Mean obliquityot the ecliptic 23026'21i448 19
23.26'21':41
Annual rate ol rotation on the o':4704
ecliptic
Mean distanceot Eanh from Sun 1.0000010178AU 1.00000105726665AU
l\,leanorbitalspeed 29.7859km/s 29.784766966km/s
Mean centripetalacceleration 0.00594ms 2 5.9301134387x 1O 3 ms-2
 Table 15.5
N)
Geodetic ReterenceSvstems

GeodellcReterenceSpheroids
Equatorial Reciprocalof Gravitalional DynamicalForm Ang.Velocity
Name and Oate Badius,a Flattening,
1 /J Constant,GM Factor,J2 ol Earth,(,
m 1 0 1 4 ms 32 1o-5rads-1

MERIT 1983 6378137 294.257

GRS 80 (rUGG,1980) 8137 294.257222 3.986005 0.00108263 7.292115
IAU 1976 8140 298.257 3.986005 0.00108263
South American1969 8160 298.25
GRS 67 (IUGG,I967) 8160 298.247167 3.98603 0.0010427 7.2s2;s1462
AustralianNational 1965 8'160 294.25
tAU 1964 8160 294.25 3.986; o.oo1o;7 7.2921
Krassovski1942 4245 298.3
International1924 8388 297
(Hayford)
Clarke 1880 mod. 8249.145 293.4663
Clarke 1866 8206.4 294.978698
Bessel 1841 7397.155 299.152813
Everest 1830 7276.Us 300.8017
Airy 1830 7563.396 299.324964

F
z
d
I

.o
m
m
z-l

6
Table15.5.continued
Reglonel Geodetlc Detums

GeodoticDalum Spheroid Odgin Latilude Longatude Cent€r Ollsel m

2
(Easl) xal Yo zn o

t
N€wArc1950 Clarke 1880 mod. Buflelsfontsin -33 59 32.000 25 30 44.622
AustralianGeodetic1966 Auslralian National Johnslon Memorial -25 56 54.55 133 12 30.08 -122 ,43 138
1965 Caim
European Intemational1924 HelmedTower
'l950 52 22 51.45 13 03 58.74 -84 -105 - 126
lndian
1938 Everest1830 Kalianpur 24 07 11.26 77 39 17.57
NorlhAmodcan 1927 Clark€ 1866 Meados Ranch 39 13 26.686 261 27 29.494 -22 154 176
Ordnance SurvoyGBSNSo Airy 183O Herslmonceux 50 5t 55.271 0O 20 ,15.842 372 -127 433
Picodo lasNi€v6s(Canaries) lniemalaonal1924 Pico d€ las Nieves 27 57 41.273 344 25 49.476
Potsdam B€ssel 1841 52 22 53.954 13 04 01.153
Pulkovo1942 Krassovski1942 PulkovoObs. 59 46 18.55 30 19 42.09
SouthAmerican 1969 S. Ameican 1969 Chua -19 45 4!.653 311 53 55.936 -75 5 -43
Tokyo B€ssel184'l Tokyo Obs (old) 35 39 17.51 139 44 ,t0.50 -14:] 514 675

Se6Mo lz, 19841

flUGG,1980.

Geodeiicand GeocentrlcCoordinatoa
O = centerof Earth
OA= equatodal radius,a
OB= polarradius,b
- a(1 J)
OP- geoc€ntric radius,ap
POois normallotherelerencespheroid
QQt = aS
ooQ = ac
d = geodeticlalitr.ide
d/ = geocenticlalitude
..1
704 EXPLANATOBYSUPPLEMENT Table 15.7
Planets:FlotationalData
Table15.6 (NorthPoleof Rotationandthe PrimeMeridian)
Planets:MeanElements
= 2000January1 5)
(ForepochJ2000.0= JD2451545.0
Sun ao = 286:13
60 = 63:87
Planet Incllnation(t Eccentrlclty(e)
l,Y= 84910 + 14:1844000d

Mercury 7000'17:95051 0.2056317524914 Mercury o o = 2 8 1 . 0 1 - 0 . 0 0 r3

Venus 3023'40:07a2A 9142
0.0067/1881 6 0= 6 1 . 4 5 - 0.005 r
Earth 0.0 0.0167086171540 W = 329.71 + 6. 1385025d'
Mars 1o50',59:01532 0.09340061
99474
Venus do = 272.72
Jupiler 1o18'11:77079 0_048494a51
2199
2'29'19:96115 7172
0_055508621 6o = 67. 15
Satun
0046'23150621 0.0462958985125 W = 16O.26 1 . 4 8 1 3 5 9d6
UranLls
Neptune 1"46'11182795 0.0089880948652 Earth .Yo=000 - 0.641 T
Pluto 17"08'31:8 0.249050 60 = 90.00 0.557 r
w = 1 9 0 .1 6 + 360. 9856235dl
lt6en Longitude Moan Longllud€ Mean Longllude
ol Perlhellon (@) al Epoch (L)
lvlars o o = 3 1 7 . 6 8 1 0 . 1 0 8f
ol ode (r2)
6o = 52.886 - 0.061 r
4€o19'51:21495 77"27'22:02855 252"15 03:25985 YY= 176.868+ 350.8919830 di
Mercury
V€nus 76o40'47:71268 '|31'33'49134607 1A1" 5a 47:28304 Jupiter ao = 268.05 -0.009f
Earth 0.0 102056'14145310 100" 27' 59:21464 6o= 64.49 + 0.003 f
Mars 49o33'29:13554336"03',36184233355025'59t78866 W = 2a4.g5 + 870.5360000db
Jupirer 100o27'5119863114419'52:71326 34421'05:?/211
sarurn 113039'55:88533 93' 03'24:43421 50004'38:89695 Saturn do = 40.58 - 0.036 I
Uranus 74"OO21:41OO2 173000'18157320 314003'18:01840 60 = 83. 54 _ 0.004 r
Nepiune 131o47'02:60528 4a"07'25:28581 304020'55:19574 W = 38.90 + 810.7939024 d$
Pluto 110017',49:7 224008'05:5 23ao 44'3A:2
Uranus do = 257.43
6 0= - 1 5 1 0
lrean Distance(AU) MeenDlatance(10' ' m)
w = 2 0 3 . A 1 - 5 0 1 . 1 6 0 0 9 2d85
0.3870983098 0.579090830 Neptune a6 = 299.36 + 0 . 7 0 s i nN
o.7233294200 1.08208601 bo = 43.46 - 0.51 cos N
Eanh 1.0000010178 1.49598023 W = 253.18 + 536.3128492 d 0.48sinN
1.523679U19 2.27939146 N = 359.28 + 54.308r
Jupiter 5.2026031913 7.78298361
Salum 9.5549095957 14.29394133 Pluto do = 313.02
Uranus 28.75038615 60=909
19.2184460618
Nepiune 30.1103868694 45.04449769 W = 236.77 - 56.3623195d
39.544674 59.157990

Sldereal Synodlc Mean Oaily Orbltal aO, ,O are slandardequalorialcoordinaleswith equinoxJ2000 al epoch J2000. Approxtmatecoordtnatesoi
P€rlod (Jullan years) Perlod (d) Molion (r) Veloclty (knv3) lhe nonh pole of the invariableplaneare ao = 273:85, 60 = 66:99-
locationoJtheprimem€ridianmoasuredalonglhe planets equatorin an eastertydirectionwilh respecl
lo the planel'snorth pole lrom the node (locatedal ighl ascension9Oo+ oo) of lhe planets equalor
on lhe slandardequator.ll l/ increaseswilh time, the plan€t has direct rotationand it W decreases
024044445 115.8775 4?09237706 47.8725 wilh lime, roialion is said to be retrograde.
0-61518257 583.9214 1960216874 35.0214 T= inlervalin Julian centuries(o136525days) i.om the standardepoch.
Eadh 0.99997862 0:98564736 29.7855 d= inlervalin days Jromthe slandardepoch.
Mars 1.88071105 779.9361 0:52407109 24.1305
Jupiter 11.85652502 398.8840 0908312944 13.0697 The slandardepoch is 2000 January 1.5,i e., JD 2451545.0TDB.
Salum 29.42351935 378.0919 0903349791 s.6724
83.74740682 369.6560 0:01176904 6-8352 Notes: * Th€ 20' me dian is
delined by the crater Hun Kal.
Neptun€ 163.7232045 367.4867 0:006020076 5.4774 I Th€ expressionlor W might be in error by as much as 0 2 becauseol uncertaintyin lhe t€nglhol the UT
Plulo 244.0204 366.7207 09003973966 4.7490 -
d a y a n dt h € T D T u T o n 2 0 0 0J a n u a r y ' t .
t Th€ O0mefidianls delined by the craterAiry-O.
Not€: The elsmontslor Plulo are based on a fit over one century The values tor lhe other ll The equationsJor !y lorJupiter, Saturn,and Uranusrefer to rhe rotationoJthejr magnetictiotds
plansls are based on an evaluaiionof the exorossionsfor lhe mean elemenlsevalual€dat {systsm
lll). On Jupitsr,System | { n1 = 67: 1 + 877: 900 d) refers to lh€ msan almosphericequatorialroralion;
the epoch.
Systemfl ( l 1l = 43?3+A7O?27Od) telerclo lhe mean atmosphericrctationnorlh ot the southcomponent
ol th6 nonh equatorialbelt,and southofths north componentol thg south equatoialbslt.
 {
Table15.8 o
Data
Planots:Physicaland Photometric
fa.| Rdlu. Anluht Dlrt nc. FbtLnlng an CoalficLr a
(cquJ Dlrm.ter lrom Earth (Stom) Dcn.liy ol Poll|rrbl
-,--
rda kg km (s€enote3) dcm' 1oo4 1oo4 looJ4

Mercury 0.3Xn22 24g,.7 11t0 0.613 0 5.{'

V€nus 4.8690 6051.9 y'" 0.277 0 5.24 0.U27
Eanh 5.9742 6378.140 - 0,00335364 5.515 t.08263 -2.54 -1.61

(Moon) 0.073483 1738 31:08 0.00257 0 3.34 o.m27

Mals 0.64t91 3i197 17:9 0.524 0.00647630 3.94 1.964 36
JuPitor 1898.8 71492 ,{6:8 4.203 0.OW744 1.33 14.75 _;
Satum 566,50 602€8 l9t4 8.539 0.0979624 0.70 16.45 -1000

Uranus 86.625 25559 3:9 18.182 0.W9273 L30 12

2:3 29.06 0.0171 't.76
N€ptuns [email protected] 24764
05 8.44 0 't.1
Pluto 0.015 1151

Sld.rfil Incllnation Gaomettlo VLual Colo.Indhaa

P.dod ot ot Equatoa Albcdo agnltud.
Rotallon to Orbft
d v(l.o) vn B-v u-B

MeGry 58.6462 o.o o.fo6 -o.42 - 0.*) 0-41

Venus -243.01 177.3 0.65 -4.Q - 0.82
0.99726968 23.45 0.367 -3.86 T' fiI
Earth .D

(Moon) 27.32166 6.68 0.12 +4.21 -12.74 0.92 0.,16 F

25.19 0.150 - t.52 -2.O1 1.36 0.58 z
Ma|s 1.0259567s
Jupit€r 0.41354 (Sysl6rnll0 3.12 os2 -9.,10 -2.70 0.&| 0.44 d
Salum 0.€75 (Syst€mlll) 8.73 o.47 -8.88 +0.67 1.04 0.58

Uranus -0.65 97.86 0.51 -7.19 +5.52 0.56 o.2a 'o

Neptuno 0.768 29.56 0.41 -6.47 +7.U 0.41 0.21
-6.3€67 -l.o +15.12 0.8o 0.31 m
Ptuto lla? 0-3
m
z

Table15.8,continued
Planets:Physicaland photometric
Data
m
'n
l{otaa: l. Th€vatu6storthe massosinctudethe atmosph6r6sbut exclud€satelites. m
2. Th€ meanequaloriatradiiare giv6r.. m
z
3. The angulardialhet€rscorrespondlo the distanc€stromlhe Eadh (in Au)glv6nin
$e adiac€ntcotumn: m
th6y r6tarto inferiorconjunctiontor Mercuryand Venusand to m€anopioslton lor
the other planets.
(110= 4.848micoradians.)
4. Theffafientngis the ratioof th€ difior€nceot the €quabrialand polarradiito the equalonat
radius.
5. The notalionfor th6 cosffcientsof th6 gravitEtionalpotentiali6 giv6nin lrrns. UU xl A, flg,ls,t[.
6. The pedodof rotationre{ersto the rotationar th€ eq!,arorwfth;spect to a fixeJ rramed
or reterence:a
negativesign indicatesthat the rotalionis r€tlograd€with resp€ctio the pol€that xes
to the northof tho
nvadabropraneof the sorarsyst6m.The periodis givenin daysot 8a$6 sr seconds.
Th6 btation daia
for the ptanetsare tabutatedin Tabl61S.7.
7' The dariaon equato.iarradii,fattoning,p€riodor rorationancrincrinalionof equarorto
orbitars basodon
Daviesel at i989.
8 Th€ g€omstdcarbedois the ratioor the ituminatonar the Earrhfromthe pranetfor phas€
angrezeroro
th6 iluninarionprqjuc€d by a prane,absoluterywhitetamb€i surtaceof the sameradius
plac€dat the sameDosilion. as the pranel
s. The quantityy(l , 0) is thgvisuarmagnitudeol lh6 planetreduced to a distance
of t Aurromboththesun
andEarlhandphas€anglezero:yo is th€ meanopposilion rnagnilude. Thepholomotric quantities
Satum1616r for
to thediskonly.

\
..1
Table15.9
Satellites:Orbital Data

Maximum Orbital Motion ol

Planet Satellite OrbilalPeriodl Elongation Semi-maior Orbital Inclination to Node on
R = Retrograde at Mean Axis Eccenlricity Planetary Fixed
Opposltlon EquatoJ Plane4
da x los km o
"/yl

Earth 27.321661 384.400 0.054900489 18.28- 28.58 19.346

0.31891023 25 9.378 0.015 1.0 158.8
Deimos 1.2624407 102 23.459 0.0005 o.9- 2.7 6.614
.rupner lo 1.769137786 2t8 422 0.004 0.04 48.6
Europa 3.551181041 340 671 0.009 o.47 12.O
lll Ganymede 7.15455296 551 1070 0.002 o.2t 2.63
lV Callislo 16.6890184 10 18 1883 0.007 0.51 0.643
0.49817905 59 181 0.003 0.40 914.6
Vl Himalia 250.5662 10246 11480 0.15798 27.63
Vll Elara 259.6528 1 04 10 11737 0.20719 24.77
Vlll Pasiphae 735 B 20426 23s00 0.378 145
lX Sinope 758 R 20931 23700 o.275 153
X Lysithea 259.22 10404 1t720 0.107 29.O2
Xl Carme 692 R 20331 22600 o.20674 164
Xll Ananke 631 R 15552 21200 0.16870 '147
Xlll Leda 234.72 10039 11094 o.14762 26.O7
XIV lhebe 0.6745 1 13 222 0.015 0.8
0.29826 42 129
o.294740 42 124 m
Salum | llimas o.942421413 30 185.52 0.0202 1.53 365.0
ll Enceladus 1.370217a55 38 238.02 0.004s2 0.00 156.25 tz
lll Tethys 1.887802160 48 294.66 0.00000 1.86 72.25
lV Dione 2.736914742 101 377.40 0.002230 0.02 30.855 o
4.517500436 '| 25 527.04 0.00100 0.35 10.16
15.94542068 317 1221.43 0.029192 0.33 0.52135 c
Vll Hyperion 2t2766044 359 1481.! 0.104 0.43
Vlll lapetus 79.3301825 935 3561.3 0.02828 14.72 m
lX Phoebe 550.48R 34 51 12952 0.16326 i72
X Janus 0.6945 24 15t.472 0.OO7 0.14 z-.1

Table15.9,continued
Satellites:
OrbitalData
fl

Saiurn Xl Epimetheus o.6942 24 151.422 0.009 0.34

Xll Helene 2.7369 1 01 377.40 0.005 0.0 z
Xlll Tetesto 1.8878 48 294.66 o
XIV Catypso 1.8878 48 294.66 o
XV Altas 0.6019 22 137.670 0.000 0.3
XVI promelheus 0.6130 23 139.353 0.003 0.0
XVll Pandora 0.6285 23 141.700 0.004 0.0
Xvlll Pan 0.5750 21 133.583
Uranus I A el 2.52037935 14 191.02 0.0034 0.3 6.8
ll Umbdel 4.1441772 20 266.30 0.0050 0.36 3.6
lll Titania 8.7058717 33 435.91 0.OO22 0.14 2.O
lV Oberon 13.4632389 44 543.52 0.0008 0.10 1.4
V Minnda 1.41347925 10 129.39 0.OO27 4.2 19.8
Vl Cordetia 0.335033 4 49.77 < 0.001 0.1 550
Vll Ophetia 0.376409 4 53.79 0.010 0.1 419
Vlll Bianca o.4u577 4 59.17 < 0.001 o.2 229
lX Cressida 0.463570 5 61.78 < 0.001 0.0 257
X Desdemona 0.473651 5 62.68 < 0.001 o.2 245
Xt Juliet 0.493066 5 64.35 < 0_001 0.1 223
Xll Portia 0.513196 5 66.09 < 0.001 0.1 203
Xlll Rosalind 0.558459 5 69.94 < 0.001 0.3 129
XIV Betinda 0.623525 6 75.26 < 0.001 0.0 167
XV puck 0.761032 7 86.01 < 0.001 0.31 81
Neptune I Triton 5.8768541R 17 354.76 0.000016 157.345 o.5232
lt Nercid 360.13619 4 21 5513.4 07512 27.63 0.039
lll Naiad 0.2s4396 2 117.6 < 0.001 4.74 626
lV Thalassa 0.311485 2 73.6 < 0.001 o.21 551
V Despina 0.334655 2 52.6 < 0.001 o.o7 466
Vt catatea o.424745 3 62.0 < 0.001 0.05 261
Vll Larissa 0.554654 3 50.0 < 0.0014 0.20 143
Vlll Proteus 1.122315 6 48.2 < 0.001 0.55 o.5232
Plulo I Charon 6.38725 <1 19.6 < 0.001 993
Notes: l Siderealperiodq exceptthal lrcpicalpe ods are given 4. Rate ol decrease(or increase)in the tongitucieof the
for satellitesof Saiurn, ascendinonode. \
2- Relativelo ecliplicplane. 5. Rate of rnirease In the tongitudeof th€ apse.
3. Refefiedto equatorof 1950.0.
(o
6. On the ectioticDtane.
 Table15.10
Satellites:Phvsicaland PhotometricData

Maf€ Radius SiderealPeriod Geomelric

Satelllte (1/Planel) (km) of Rot6tionl (d) Atbedo (vt3 v(1,0) B-V U-B

Moon 0.0'1230002 1734 s 0.12 +O.21 -12.74 0.92 0.46

Phobos 1 . 5x 1 o - 8 l3.5x1o-8xs.4 S 0.06 +11.8 11.3 0.6
Deimos 3x10-9 7.5x6.1x5.5 s 0.07 +12.89 12.40 0.65 0.18
to 4 . 6 8x 1 o - 5 1815 s 0.61 1.68 5.02 1.17 1.30
Europa 2.52X 10 5 1569 s 0.64 -1.41 5.29 0.A7 0.52
l Ganymede 7.80 x 1O-5 2631 o.42 2.09 4.61 0.83 0.50
Callisio 5.66 x 10 5 24oo s 0.20 -1.05 5_65 0.86 0.55
Amalthea 38 x 10-10 135 x 83 x 75 s 0.05 +7.4 14.1 1.50
Himalia 50 x 10-10 93 0.4 0.03 +8.14 ft.44 0.67 0.30
Elara 4x10-10 38 0.5 0.03 +10.07 16.77 0.69 O.2A
v l Pasiphae 1 x 10 10 25 +10.33 17.03 0.63 0.34
IX Sinope 0.4 x 10-10 18 +11.6 18.3 0.7
X Lysithea 0.4 x 10 10 18 +11.7 14.4 O.7
XI carme 0.s x l0 10 20 +11.3 18.0 0.7
xtl Ananke 0.2 x 10 10 15 +12.2 18.9 O.7
xlll Leda o.03 x 10-10 8 +13.5 2O.2 O.7
Thebe 4x10-10 55x45 0.05 +9.0 15.7 1.3
Adraslea 0. I x 10 10 12.5x10x7.5 0.05 +12.4 19.1
Metis 0.5 x 1o-10 20 0.05 +10.8 17.5
I\,4imas 8.0 x 10-s 196 S 0.5 +3.3 12.9
m
II Enceladus 1.3 x 10 7 250 s 1.0 +2.1 1 l .7 0.70 O.2A -o
l Tethys 1.3 x 10-6 530 S 0.9 +0.6 1o.2 0.73 0.30 tz
Dione 1.85x 10 6 560 s 0.7 +0.8 1O.4 O.71 0.31
Rhea 4.4 x 10-o 765 s O-7 +0.1 5.7 0.78 0.38 d
Titan 2.38 x 10 a 2575 S o.21 1.28 8.28 1.28 0.75
Hyperion 3x10 I 2o5x13ox11o 0.3 +4-63 14.19 0.78 0.33 c
vIl lapetus 3.3 x 10 6 730 s O.22 +1.5 11.1 O.72 O.3O
tx Phoebe 7x1o-10 110 0.4 0.06 +6.69 16.45 0.70 0.34 m
X Janus 1 1 0 x 1 0 0x 8 0 S 0.8 +4.4 14 3
z-.t

-atrF!!-

Table15.10,continued
Satellites:
PhysicalandPhotometric
Data
I
m
Mas,s Radlus Sidereal Period Geomelric m
Satellite (1/Planel) Il
(km) ol Rotationl (d) Albedo(t43 v(1,0) V4 B-V U-B m
z
o
Xl Epimetheus 70x60x50 0.8 +5.4 15 o
Xll Hel€ne 18x16x15 o.7 +8.4 18
Xlll Teleslo 17x14x13 0.5 +8.9 18.5
XIV Calypso 17x11x11 0.6 +9.1 1A.7
20x10 0.9 +4.4 18
XVI Prometheus 70x50x4{) 0.6 +6-4 16
XVll Pandora 55r45x35 0.9 +6.4 16
Xvlll Pan 10 0.5
I Ariel 1 . 5 6x 1 o 5 579 s 0.34 +1.45 14.16 0.65
Umb el 1 . 3 5x 1 O - 5 586 s 0.18 +2.10 14.81 0.68
4 . 0 6x 1 O - 5 790 s 0.27 +1.02 13.73 0.70 0.28
lV Oberon 3 . 4 7 t <1 0 5 762 S o.24 +r.23 13.94 0.68 0.20
V Miranda 0.08 x 10-5 240 s o.27 +3.6 16-3
Vl Cordelia o.o7 11.4 24.1
Vll Ophelia o.07 11.1 23.8
Vlll Bianca 21 0.07 10.3 23.0
lX Cressida 31 0.07 9.5 222
X Desdemona 27 0.07 9.8 225
Xl Juliet 42 o.o7 8.8 21.5
Xll Ponia 54 o.o7 8.3 21.O
Xlll Rosalind 27 o.o7 9.8 22.5
XIV Belinda 33 0.07 9.4 22.1
77 0.07 7.5 20.2
2 .0 9 x 1 0 - a 1353 o.7 1.24 13.47
't70 o.72 0.29
ll Nereid 2 x 10-7 o.4 +4.0 14.7
lll Naiad 29 0.06 10.0 24.7
lV Thalassa 40 0.06 9.1 23.A
V Despina 74 0.06 +7.9 22.6
Vl Galalea 79 0.06 +7.6 22.3
Vll Ladssa 1 0 4x 8 9 0.06 7.3 U.O
21ax2OAt2O1 0.06 5.6 20.3
I ..1
Charon o.22 593 0.5 +0.9 16.8

Notea:1. S= Synchronous,
rolalionperiodsameasorbitarperiod.2- Brighiside,0.5;raintside,0.05. 3. y (sun)= 26.8.
712 EXPLANATORYSUPPLEIT/ENT
15/ REFERENCE
DATA 713
Table15.11 Table15.12
PlanetaryRings ConstellationNames and Abbreviations

SaturnRiog Data Norninative Genilive Nominative Abbrev. Geniiive

Feature Distance(km) Distance(Rs)
Andrcm€da And Lacena Lac
Anllia Anl Anlliae Leo Leonis
radius
Equatorial 60330 1.000 Aps Leo Minor Ltli Leonrs[rinoris
D-ringinneredge 67000 1.11 Aquarius Aquarii Lepus Lepons
C-ringinneredge 74400 1.233 Aquila Libra Lib Librae
Lupus Lup Lupi
B-ringinneredge 91900 1.524 -Argo Aru Argus Lyn Lyncis
B-ringouteredge 117400 1.946 Lyra Lyl Lyrae
A-ringinneredge 121900 2.021 Aunga Auagae Mensa
A-ringgapcenter 133400 2.212 Eootes Boo Bootis Microscopium Ivlicroscopii
A-ringouteredge 136600 2.265 Caelum Cae Caeli
F-ringcenler 140300 2.326 Camelopardalis Cam Came{opardalis I\,{us
Cancer Cancri
G-ringcenter 170000 2.8 CanesVenatici CVn CanurnVenaircorum Octans Oct Octantis
E-ringinneredge - 180000 -3 CanislJajor CMa Canis lMajo s Ophiuchus oph Ophjuchi
E-ring outer edge - 480000 -8 Canislrinor CL4i Canis [rinoris Oon Ori Orionis
Capicornus Cap Cap corni
Rings of Uranus Carina car Carinae Pegasi
Cassiopeia Cas Cassiopeiae
Ring Semi-majorAxis (km) Eccentricily Azimuth of PrecessionRale (deg./day) Centaurus Cen Centau
Periapse(deg.) Cepheus cep Cephei Pictoris
Cetus Cet Ceti Pisces Psc Piscium
41870 0.0014 236 2.77 Chamaeleon Cha Chamaeleontis TPiscisAuslrinus PsA
5 42270 0.0018 142 2.66 Circinus Cir Circini Puppis Pup Puppis
4 42600 Columba Col Columbae Pyxidis
0.0012 120 2.60
Coma Berentces Com Comae Berenices Reliculum Rel Reliculi
44750 0.0007 331 2.18 t CoronaAusl na CrA CoronaeAuslrinae Sagitla sse Sagitlae
p 45700 0.0005 231 2.O3 CoronaBorealis CrB CoronaeBorealis Sagittarius sgr Sagitlarii
47210 Corvus Crv . Corvi Scorplus Sco Scorpii
Crater CN Crateris ScrJlptor Scl Sculptoris
1 47660
'too Crux Cru Crucis Sculum Sct Scuti
48330 *oo, Cygnus cyg Cygni rserpens Ser Sepentis
51180 0.0079 216 tau Delphinus Del Delphini Sextans Sex Sextantis
Dorado Dor Ooradus Taurus Tau Tauri
Epoch: 1977March 10, 2ohuT (JD 2443213.35/ Draco Dra Draconis Telescopium Tel Telescopil
Equuleus Equ Equul€i Triangulum Tri Trianguli
E danus Eri Eridani T angulum AustraleTrA TrianguliAuslralis
Tucana Tuc Tucanae
Gemini Gem Gemrnorum Ursal ajor UMa Ursae lrajoris
Gru Gruis Ursa[,4]nor Ul\,li UrsaeMinoris
Hercules Her Hercu/is Vela Vel
Horologium Hor Horologii Virgo Vir Virginis
Hydra Hya Hydrae Volans Vol
Hydrus Hyi Hydri Vulpecula Vul
Indus Ino Indi

Note:This lisl0l constellalionnames and abbreviationsis in accordancewith the resolulionsol the lnternational
AshonomicalUnion( r/ar s. tAl) 1,ls1t 4.22119,66 and 77). The boundariesol the conste ationsare tistedby E.
uelporle, on b€half of the tAU, in Delmtation scientiftque des constelations (tabtes et cattes)
l:ambtidge
unlversily Press. 1930):the areas ol the constellationsare gtven'n HandboohB.A.A.,1961.
*
In mocternusageArgo is dividedinto Carina,Puppis,and Vela.
' Auslralis
is somelimesused. in both nomtnativeand genilive.
+ Serp€nsmay bs dividedInto
SerpensCaput anclSerpensC.uda.
 f
5
- o -o - o - " " Fd
o ddts
E
o ":5
F;.
1r
6 do o
' ^x 4.1
ff ||nI=
Rs,
[[n
N(l)(,| NO NNq' l.)
[[tl
o.
bbb
u;B 8Bd Fiilsn;;; ? -a
- -4
661 E8 EeS
d Fiq €& H$$ "
6 .oc 6

*;H AEH
H$HHB
ll ll h:E €,q
*N d6 I
6ii
d
:::\
qo x x xro
EEEg
d6

do it
m

-NirE Iz
d
a
--l
I .D
.D

d -'q 6 m
x'qd =
ido m
e+E' =
Ei6

Tablel5.l4
Factors
EnergyConversion

lJ =r.0 o 11126501
x 10-15 O.50341125
x lds 1.5@18@7
x 1f 0]21?944x tF o. 62415064x lole 0. €7005300x 1O1o 0.22$?r0a x 1Or3
1l€ =&9a755179
x to16 1.0 4.52/t43470x tolr 13.56391401
x 1de 6.5@61579
x lo* 5-S0953616x lcf €.0@13671
x 1@ 2.0614S4r4i1fl

1r-1 =1.9€6{47t15
x lo-a 0.22102209
x 1O-al 1.0 2.99792454
x 1d 1./4870887x 101 1.2398,t21K
x 10-6 l.331O2s22
x lO-t5 x 1o-7
O.45563353
lHr = 0.66280755
x to-s O.07372503
x 10 { 0.33350410x10-3 1 . 0 0.47992164
t 10-7 o.41356692
x 1O'ra O,443€822x1g-a 0 . 1 5 1 9 8 2 x9 91 o - 1 5
lK =1.380€s780
x 10-6 x 1O €
O.t5361890 O.E95GAO7x 1O_1 2.08367351
x 107 1.0 O.8€17384,1
x 1O-7 x lO-g
0 . 9 2 5 1 l 4 O O1xO - 1 6 0.3165E292
lav =1.6@1733
x 1o-1e O.i78266e2x lo_s x iO6
O.EO€5t4lO 2.4!798836i'l01' t. 1E044?t93
x 107 1.0 t.O73tKt85x 1O-r 0.362{ts@ x 10 1
l|| =1.492418@xlo-ro x 10 6
0.160@tO2 x l0r5
0.75130C56 2.252U212x 1F 1.08094786
x 1or5 0.9314ea32
x ld l.o 0.342317rcx rd
lEh ={35s7421x lO-13 o 18508741x ro-s 2. 19zl7163l x 107 6.57s68€90x 1ot5 3. 15rrc266 x rop 2-721139€lx lO1 a 92'1262@
x lO-3 1.0
 -.t
Table 15.15
(t,
Units of Length,Speed, and Mass

MoLr lrllLr Astronomlcal Lbhr yee6 Pars€c3

(m) (ml) Unlt (AU) (1.y.) Oc)

lm 6.21371192. x 1O-a 6.€'845€7122x 10-12 1.057fina3 x 1O-10 3.24029269x lo-t7

1ma = 1609.3440 1 1 . 0 7 5 7 8 0 0 1 x8 1 O - B 1 . 7 O t O Z l 9 5x 1 O - 1 3 5.2155287(X
x lo-ra
= 1.4959787066
x lorr 9.2955ttr727 x 107 1 1 . 5 8 1 2 5 0 7 4x 1 O - 5 4.848136811
x 10-6
I Ly. = 9.46o73(X7
x 1015 5.87862537x lO12 e€241.0771 1 0.306601393
1pc = 3.085677582
x 1016 1 . 9 1 7 3 5 1 1 5 x8 1 0 1 3 206264.8062 3.26156378 1

lileterB lllles Aslaonomlcal Pa]a€ci Lighl year3

p€r aecono Pernour unil3 por alay Peacentury por yoar
(rn/s) (mUhr) (AU/day) (pc/cent) (volocltyot lighl)

1 2.236936292 5.Tt5483274 x 1O-7 1. 022712165 x I O- 7 3.33564095

x 1O-s
o.44704 I 2.581872043
x 1O-7 4.571932462x 1O'8 1 . 4 9 1 1 6 4x9 13 O - e
1731,156.837 3873t 5€.636 1 o.1770781971 5.2551833x 1O-3
97n9n.217 21472549.07 5.647202907 I 326.156374
299792458 670616623 173.1446327 30.660139s7 t
ol light m
1 (statul€)mile= 5280fe€t F
1 foot= 0.3048h6t6rs(exactly) U.S.standard z
I yr = 31557600 sec
c = 299792,158 rn/s= 1. 40261750x 1012lurlongs/lodnight d

'!
!
m

Yo' tr
NEr>FrW>
)io 4
4. atr u
gf:
3J
go do i
:a, - Bt
=D = - 5
q9 5d

3 <>X-@E

M'! Fo I r 1z

. * . f E . ' 9= . e

e gx g c I

3<Xe<-l

93 9B€ q
d---9-
39
.{
..1-
v18 EXPLANATORY
SUPPLEMENT 15 / REFERENCE
DATA 719
Ttble 15.17
InbmationalSystemof Units(Sl) Table15.17,contlnued
Systemof Units(Sl)
International
sl Bs unltr
Unlt8In Uservllhlhe InternatlonalSystem
Quantity Symbol
Symbol Valuein Sl unlt
bngrh mqer m
mass kilogram kg minute mln 1 min 60s
tme sacond s nour h th 60 min= 3600s
glactaiccun€nt ampere day d 1d 24h=88100s
thgrmodynamic temperature kelvin K oogree (?r/ 180)rad
amountof subatance more mol
mlnute (1/ 60)0= (?r/ 10800)rad
I
lumlnousintensity candela co
sscond (1 / 60)' = (?r/ 648000)rad
liler I,L 'tL 1 dm3= 1O-3m3
Sl Derlved Un[s wilh Speclal Namg6 metricton t 1t 103kg

Ouantity Symbol Expression Expression Sl Pl€tixes

Inlerms in terms
of other units of Sl baseunitrs Factor Prefix Symbol Factor Prefix Symbol
frequency hertz Hz s-t 1ot8 exa E 10-1 deci d
lorce neMon N m kgs-2 p€ta P 1o-2
1015 centi c
pressure,su€ss pas0al PA lVm2 m-l kgs 2 1012 tera T 1o-3 milli m
engrgy,woft quantity 1oe giga G 1o-E m|cro p
of heat ioule Nm m2kgs-2
power,radiantflux watt J/s m2kgs-3 1d mega M 10-e nano n
electriccharge, 1d kilo k 1o-12 pico p
quantityot electricity coulomb SA 102 hecto h 1o-15 femto f
el€c,tricpotgntial, 101 deka da 10-18 atlo a
potentialdifference,
electromotivoforce volt mz kgs-3A-r
capacitanc€ tarad CN m-2 kg-l s4./A2
electricresislance onm o m-2 kgs-3 A-2
elgctricconductance s|emens m-2 kg-1s3A2 15.1 REFERENCES
magneticflux weDer wb m2kg s-2 A-1
magneticflux density tesla T wby'm2 kg s-2l4 1 Taylor, B.N. and Cohen,E.R. (1986). "The 1986AdjustmeDt of the FundamentalPhysical
inductance henry H wb/A m2 kgs 2A-2 CoDstarts" CODATA Neusletter, November.
C€lsiustBmperalure degreeCelsius K Davies, M.E. et ol. (1989). "Report of the IAU/IAG/COSPAR Working Group on Car-
lUminous flux lumen tm cosr
lllumlnance tographic Coordinatesand Rotational Elements of the Planets and Satellites:1988"
lux tx lnlm2 m-2 cdsr
CelestialMechanics46, 187-204.
Moritz, H. (1984). "Geodetic R.eference System 1980" Bull Geod,esique
SS,388-398.
Sl Supplem€ntaryUnits -(1986). The Intentational Systernol Uzriis(SI). Goodman, D.T. and Bell, R.J.'
eds. U.S. Department of Commerce.
Ouantity Symbol

planeangle radian rao

solidanglg steradian sr
 Glossary

aberration the apparent angular displacement of the observed position of a celestial

object from its geometric position, caused by the finite velocity of light in
combination with the motions of the observer and of the observed obiect.
(See aberration, planetary.)

aberration, a.nnual the component of stellar aberration (see aberration, stellar) re-
sulting from the motion of the Earth about the Sun.

aberration, diurnal the component of stellar aberration (see aberration, stellar)

resulting from the observer's diurnal motion about the center of the Earth.

aberration, E-terms of terms ofannual aberration (seeaberration, annual) depend-

ing on the eccentricity and longitude of periheliou (see longitude; pericenter)
of the Earth.

aberration, elliptic see aberration, E-terms of.

aberration, planetary the apparent angular displacement of the observed position

of a celestial body produced by motion of the observer (seeaberration, stellar)
and the actual motion of the observed object (see correction for light-time).

aberration, secular the component of stellar aberration (see aberration, stellar)

resulting from the essentially uniform and rectilinear motion of the entire
solar system in space. Secular aberration is usually disregarded.

aberration, stellar the apparent angular displacement of the observed position of

a celestial body resulting from the mqtion of the observer. Stellar aberra-
tion is divided into diurnal, annual, and secular components. (See aberration,
diurnal; aberration, annual; abenation, secular.)
722 GLOSSABY GLossaRY 723
altitude the angular distancc of a celestial body above or below tltc hotizort, mea- atomic second see second, Systime International.
sured along lltc grcat cirt le passingthrough the boclv and the zenith. Altitucle
augmeutation the amount by which the apparent semidiameter of a celestial bo<11,.
is 90' minus zenith distance.
as observed from the surface of the Earth, is greater than the scnridianr,,rcr
anomaly angular meiuurement of a borly in its orbit frotrr its perihelion. that would be observed from the center of the Earth.

aphelion thc point iIr a plattotnrl' orbit that is at the gr'catest clist:rnce frotrt the azimuth the angular distance measured clockwise along the horizon from a specilicd
Sun. reference point (usually north) to the intersection with the great circle drawl
from the zenith through a body on the celestial sphere.
apogee the point at rvhich a bodv in orbit aroultcl the Eartir rcirchcs its farthest
distarrcc from the Earth. barycenter the center of mass of a system of bodies; c.g., thc center of mass of llrc
solar systcm or the Earth Nloon system.
ott a celestialsphere.tenteleclat the Earth. dctcrlninr:cl
apparent place thc 1>osition
by removing from the diroctly obscrvedposilion of a celestialbody thc cffects Barycentric Dynamical Time (TDB) the independent argument of ephemeridcs arr<l
that clcpeudoll the topocentric locatiorr ol the obscrverl i <:..refraction, di- cquations of motion that are referred to the barycenter of the solal systeDr.A
uural aberratjon (see aberration, diurnal). and geoccntric (dirrural) parallax family of timescales results from the transformation by various theories arr<l
Thus the positioll at which thc objcct wottltl actrialli be seeltfrom tlx'<clttel metrics of relativistic theories of Terrestrial Dynamical Time (TDT). TI)II
oftlrc Earth, displacedlry plarrclerryaberration (cx<r'pttlie tlittrnal piit't see differs from TDT only by periodic variations. In the terrninology of the gt,n_
aberration, planetaryt aberration, diurnal) and rcforred to tirc true equator eral theorl' of relativity. TDB may be consideredto be a coordinate time. (Sr,c
and equinox. dynamical time.)

apparent solar tirne the mcasurc of tirtrc basctl ott the dittrnal rttotiott of tlte' trut: brilliancy for Nlercury and Venus the quantity /<s2/ r:, r-here t = 0.5(1 +cosl). i
Sun. The rate of diulnal urotiott tttclergoosseasoualvariatiott ltccaust'of the is the phase angle, s is the apparent sernidiameter, and r is the helioceutr.ic
obliquity of tltc ecliptic aud bec:utseof tlrc eccentricity of tlrtr Earth s orbit distance.
Additiollal srtrzrllvariations rcsrill fLom it rt:gulalitit's in tll,' lotaliori of the
Earth on its axis. calendar a system of reckoning tirrre in rvhich dat.s are enurlerated accordirrgto
thcir position in cyclic patterns.
aspect thc apparelt, l)osilioll ofany ol lltc plern<'ls
or tht'Nloon relllive to lltc Sutr,
as sr.cn from Eiuth. catalog equinox the intersection of the hour circle of zero right ascension of a srirr.
catalog $ith the celestial equator. (Sec dynamical equinox; equaror.l
astrometric ephemeris al ephemerisofa solat system bo<ivin lti< lr the talrltlatcd
posil.ioDsare cssentiallY(orripalalrlcto calalog meaDplaces()l slars at ii stan- celestial ephemeris pole the referetrcepolc for nutation and polar motion: tlie 2rxis
dard epoch. An astror('1.ic positiou is olrt:rincd ltl aclding 1o tlle geometric of figure for the mean surfaceof a rnodel Earth in u,hich the frec urotion lras
position, comlrrrtedlirrtr gravil,aliorral tlu'ott', t,1n'(\)rrectiorr for light-time zero amplitude. This pole has no nearly diurnal Dutation q,ith rcrspectto t[
iotr (sor:aberration, annual: aber-
Priol t o 1981,1Ire E-t r:r'rus of arttlral abclt-41 space-fixed or Earth-fixed coordinate svstem.
ration, D-terms of) lcrc also addc<lto tlrc gcornetli( l)ositi(tl.
celestial equator the projection orrto the celestial sphere of the Earth s equator.
astronomicalcoordinates 1lx: krngitudt':trr) latil rrdeof a poilt otr l ltt: IJa.rtltlclativt' (See mean equator and equinox; equinox; true equator and equinox.)
to tlx'geoid. TIrcsecoor'<linates ll'local glavilr attomalics.(S*'
irrc irrflrrerrcr'<l
zenith; longitude,';errestrial latitude, tenestrial.) celestialpole either of the two points projected onto the celestial sphere bl tlrr:
extension of the Earth's axis of rotation to irrfiuity.
astronomical unit (AU) the latlius of a cir<ulal orbit itr rvlricha l>orlvol lrcgligible
urass,lncl froc of perturbations, rvould r-t'rrrhcatotttttl the Stttt in 2;/t da1's. celestial sphere an imaginary sphclp ofarbitrary radius uporr rvhich celestial bodics
rvherc I is thc Gaussiangravitational constant. This is sligJtlly less tltarr the may be corrsideredto be located.As cifcumstancesrequirt:,the celestialsphcr.r:
semi-rnajor axis of thc Earth's oll)it. may be centcred at thc observcr, at the Earth's center, or at any othcr locatiorr.
724 GLoSSARY
GLOSSARY
725
coqiutrction the phenourenon in which two bodies have the sane apparent celes-
this cffe(t rvrrichis indepcndent of *averc'gtrr. is incruclecl
tial longitudo (see longitude, celestial) or right ascension as viewed from a iu the rcdrrctio.
from mean place to apparent place.
third body. Conjunctions are usually tabulated as geocentric phenomena. For
Mercury and Venus, geocentric inferior conjunction occurs when the planet is deflection of the vertical theangle between the astrorronrical vertical
aud the gcoclc_
between the Earth and Sun, and superior conjunction occurs when the Sun is tic vcrtical. (See zenith; astronomical coordinates; geodetic
coordinates,)
between the planet and Earth.
Delta T (AT) the cliflercnce betwecn dynamical time arrd Universal
Time: spe<.if_
constellation a grouping of stars, usually !',,'ithpictorial or mythical associations, icallr- thc difference betrvcen Terrestrial Dynamical Time
(TDT) ancl UT1:
that servesto identify an area of the celestial sphere. Also, one of the preciselv AT=TDT UTI.
defined areas of the cclestial sphere. associated with a grouping of stars, that
the International Astronorrrical Union has designated as a <:onstellation. direct motion for orbital motion iD the solar sJ,stenl.motio[ thilr
ls counter(jlock_
q.ise in the orbit as seen frour thc
north polc of the ecliptic: for. an objer.t
Coordinated Universal Time (UTC) thc tirnescale available from broadcast tirne oltserved on the cclestial sphere. rnotion that is from \\,cst
ro cast. resltlting
signals. UTC difiers from TAI (seeInternational Atomic Time) by an integral from the relative rnotion of the ob.fectancl thc Earth.
number of seconds;it is maintained within +0.90 secondof UTI (sce Univer-
sal Time) by the introduction of one second steps (leap seconds).(Sec leap diurnal motion the appare.t daily motion of celestial b.dic,s across
th. sky fior.
second.) east to wcstJr:auscdb"vtlxt Earth,s rotatiotr.

cubninatiou pa^ssage of a celestial object across the observer's meridian; also callecl AUT1 the prcdicted l,alueof thc diflerencebetwcen UTI arxl UT.CI.
tr.tnsmitlcd in
"rneri<lian passago." Nlore precisel"v,culmination is the passagc through the code on broadcast tinrc signals:AUTt = UTI tiTC. (Sr.cUniversal Time;
point of greatestaltitude in the diurnal path. Upper culmination (also callod Coordinated Universal Time.)
"culmination abovc pole" for circumpolar stars and the \{oon) or transit is
dynarnical equinox the asccndi'g node of the E.rth s urca' orbit
the crossirrg r:loser to the obscr'"'er's zenith. Lower culmination (also callcd o. tho E.rth.s
equator: i.e.. 1.heilrterscctionof thc ecliptic ',ith tlrc r:ek.stialequ.tor
"culmination belorvpolc" for r:ircunrpolarstars and the \loon) is the crossing at *.hic.lr
the Srrn'sdeclination is changing from soulh to nolth. (Sec catalog
farthel from the zenith. equinox;
equinox.)
day an intcrral of 86100 SI secorrds (sce second, Systime International), unless
dynamical time tho falrily of tirnescales introduccd in 19g.1
othenvise iudicated. to r.r:pl:rceephemeris
time trs the irrclependentargunlelt of d1,n:r,mic.allheories :rrrd ephcrncricles.
day numbers quantities that facilitate hand calculations of the reduction of mean (See Barycentric Dynamical Time: Terrestrial Dynarnical Time.
)
place to apparent place. Bessclian dal nurnbers depend solelv on the Earth's
eccentric anomaly in urrtlistrrrbcdelliptic motiorr. titc aDgle
position and rnotion; secorrd-order day numbers, rrsed in higher prccision rc- nrcasure(lat the (..Dter.
of the ell\rse from pericelter to the
ductions, depend on the positions of both the Earth aucl tltc star. l)oinl orr the circrrrlst.ribirrg;urxilinrl. cilclc
fron-r which a per.pcndicrrlnr to thc nirjor axis rorrlcl inter.s(:ct th(, or.l)itina
declination angular distanc:c orr the celestial sphere north ol south ol the celestial bodl'. (See mean anomaly: true anomaly.)
equator. It is mcasured along thc hour circle passing through thc celestial
eccentricity a par.aluctortlrat specifics the shape of
object. Drxlirtation is usualll'givcn in cornbination $'ith right ascensionor a t.rlrricsct:l,iorr;orrc of tlto
starrdardclcnrontsuserlto desc,ribcan olliptic orbit. (Soeelements,
hour angle. orbital.)
eclipse the obscllrationof a <.r:lcstial
boclycatrscclltv its passag<:
defect of illumination the augrrlar amorrnt of the observed lunar or planctary disk througll th. sl)adoi,!,
cast by alother bo(h,.
tllat is nol, illumirrat<d to an observeron the Earth.
eclipse, annular a solar eclipse (see eclipse, solar) in nhich the
deflection of light the er,nglebv which the app.uent path of a photon is altered solar disk is nerer
cornplctell' covcrcd but is seen as arr atnulus or ring at rnaxirnur eclipso.
from a straight lirre bv the gravitational 6eld of the Sun. Thc path is deflccted An
annular eclipse occurs when the apparent disk of the Moorr is smaller
radially away fronr the Sun by up to ll'75 at the Sun's limb. Correction for tharr
that of the Sun.
 GLoSSARY GLOSSARY 727
726

eclipse, lunar an eclipse in rvhich the Nloon passes through the shadorv cast by ephemeris time (ET) the timescale used prior to 1984 as the independent r,ariable
the Earth. The eclipse may be total (the Moon passing completely through in gravitational theories of the solar svstem. In 1984. ET l,as replaced b3r
the Earth's urnbra), partial (the Moon passing partially through the Earth's dynanical tirne.
umbra at maximum eclipse), or penumbral (the Moon passing only through
ephemeris tra$it the passage of a celestial body or point across the ephemeris
the Earth's penumbra).
meridian.
eclipse, solar an eclipse in which the Earth passesthrough the shadow cast by the
Moon. It may be total (observer in the Moon's umbra), partial (observer in epoch an arbitrary fixed instant of time or date used as a chronological rcfercnce
the Moon's penumbra), or annula,r. (See eclipse, anmrlar.) datum for calendars (see calendar), celestial reference systems, star cataloes.
or orbital motions (seeorbit).
ecliptic the mcan plane of the Earth's orbit around the Sun.
equation of center in elliptic motion the true anomaly minus the mean anomaly. It
elements, Besselian quantities tabulated for the calculation of accurate predictions is the difierence between the actual angular position in the elliptic orbit and
of an eclipse or occultation for any point on or above the surface of the Earth. the positiol the body would have if its angular urotion rvcre uniform.

elements, orbital parameters that specify the position and motion of a body in equation of the equinoxes the right ascensionof the mean equinox (seemean equa-
orbit. (See osculating elements; mean elements.) tor and equinox) referred to the true equator and equinox: apparent sidereal
time minus mean sidereal time. (See apparent place; mean place.)
elongation, greatest thc instants when the geocentric angular distances of Nlercury
and Venus are at a maximum from the Sun. equation of time the hour angle of the true Sun minus the hour angle of the ficti-
tious mean sun; alternatively, apparent solar time minus mean solar time.
elongation (planetary) the geocentric angle between a planet and the Sun, mea-
sured in the plane of the planet, Earth and Sun, Planetary elongations are equator the great circle on the surface of a body formed bv the intersectior of the
measuredfrom 0o to 180o.east or west of the Sun. surface with the plane passing through the center of the body perpendiorlar
to the axis of rotation. (See celestial equator.)
elongation (satellite) the geocentric angle betr-een a satellite and its primary, mea-
sured in the plane of the satellite, planet and Earth. Satellite elongations are equinox either of the trvo points on the celestial sphere at which the ecliptic in-
measured from 0o east or west of the planet. tersects the celestial equator; also the time at $'hich the Sun passesthrouglr
either ofthese intersectionpoints; ie., when the apparent longitude (seeappar-
epact the age of the Moon; the number of days since New Nrloon,diminished by one
ent place; longitude, celestial) of the Sun is 0o or 180". (See catalog equinox;
day. on January 1 in the Gregorian ecclesiastical lunar cycle. (See Gregorian
dynamical equinox for precise usage.)
calendar and lunar phases.)
era a system of chronological notation reckoner.lfrom a given dat..
ephemeris a tabulation of the positions of a celestial object in an orderly sequence
for a number of dates.
fictitious mean sun an imaginary body introduced to define mean solar time; cs
ephemeris hour angle an hour angle referred to the ephemeris meridian. sentiallv the name ol a mathematical formula that defined mean solar time.
This concept is no longer used in high precision work.
ephemeris longitude lorrgitude (see longitude, terrestrial) measured eastward from
the ephemeris meridian. flattening a parameter that specifies the degree by which a planet's figure differs
from that of a sphere; the ratio / = (a b) I a, where a is the equatorial radius
ephemeris meridian a fictitious meridian that rotates independently of the Earth and b is the polar radius.
at the uniform rate implicitly defined by Terrestrial Dynamical Time (TDT).
The ephemeris meridian is 1. 002738ull east of the Greenwich meridian, where frequency the number of cycles or complete alternations per unii time of a carrier
A?'= TDT - UTl. wave, band, or oscillation.
728 GLOSSARY GLOSSARY 729
frequency standard a generator t'hose output is used as a precise frequencv refer- heliocentric with reference to. or periaini[g to, the center of the Sun.
ence; a prinrary liequency standard is one rvhose frequclcy corresporrdsto the
adopted definition of the second (see second, Systdme International), rvith its horizon a plane perpcndicular to the line from an observer to the zenith. The great
specified accuracl'' achievcd t'ithout calibration of the device circlc formcd by the intersection of the celestial sphere with a plarre perpen-
dicular to the line from an observer to the zenith is callcd the astronomical
Gaussian gravitational constant (,t = 0.01720209895) thc corrstarrt defining the as- horizon.
tronomical systern of units of length (astronomical r.rnit), mass (solar mass)
and tirre (day), bl rneans of Kcpler's third law. The dimensions of k2 arc horizontal parallax the difference between the topocentric and geocentric positions
those of Nct'ton's constant of gravit atiort L3M tT 2. of an object, r'hen the object is on the astronomicalhorizon.

hour angle angular distance on the celestial sphere measured wesdrvardalong the
gegenschein faint nebulous light about 20o across near the ecliptic and opposite
celestial equator from the meridian to the hour circle that passesthrough a
the Sun, best seen in Septemberand October. AIso called counterglorv.
celestialobiect.
geocentric rvith refercrrcc to, or pelterining to, the center of the Earth.
horu circle a great circle on the celestial sphere that passes through thc celestial
geocentric coordinates the latitude and longitucle of a point on the Earth's surfacc poles and is therefore perpendicuiar to the celestial equator
relative to the center of the Earth; also celestial coordinates given u'ith rc-
inclination the angle betrveen two planes or theil poles; usually the angle betu'eerr
spect to th. center of the Earth. (Seezenith: latitude, terrestrial; longitude,
an orbital plane and a reference plane; one of the standard orbital elements
terrestrial.)
(see elements, orbital) that specifies the orientation of an orbit.
geodetic coordinates the latitude and longitude of a point on the Earth's surface
International Atomic Time (TAI) the cortinuous scale lesulting from analyses by
determined frorn the geodeticvertical (normal to thc specifiedspheroid). (See
the Bureau International des Poids et \Iesures of atomic tirne standards in
zenith; latitude, terrestrial; longitude, terrestrial.)
many courltries. The firndamental unit of TAI is the SI second (see second,
geoid an equipotential surface that coincidcs with mean sea ler.el in the open ocean. Systdne International), and the epoch is 1958 January 1.
On land it is the level surface that ri'ould bc assumed by water in an imaginarl invariable plane thc planc through thc cerrter of mass of the solar system pcrper-
network of frictionless charrrrclsconnected to the ocean. dicular to the angular monrentum vector of the solar system
geometric position the geocentric position of an object on the celestial sphere re- irradiation an optical eflect of contrast that makes bright objects viewed against a
ferred to the true equator and equinox, but rvidhout the displacement due to dark background appear to be larger than they rcally are.
planetary aberration. (Seeapparent place; mean place; aberration, planetary.)
Julian calendar the calendar introduced by Julius Caesar in 46 B.c. to replacc the
Greenwich sidereal date (GSD) the number of sidereal days eiapsed at Greenwich Roman calendar. In the Julian calendar a common 1'ear is defined to cornprise
since the beginning of the Greenwich sidereal dav that u,as in progress at 365 days. and every fourth year is a leap year comprisirrg 366 days Thc .lulian
Julian date 0.U. calendar was superseded by the Gregorian calendar.

Greenwich sidereal day number the intcgral pari of the Greenwich sidereal date. Julian date (JD) the interral of time in days and fraction of a day since 4713 B.c.
January 1, Greenr.ich noon. Julian proleptic cal€ndar. ln precise work tlc
Gregorian calendar thc calendar introduced by Pope Grcgory XIII in 1582 to re-
timescalc, c.g., dynamical time or Universal Time, should be spccilied.
place the Julian calendar; the calendar now used as the civil calendar iu most
countries. Every year that is exactly divisible by four is a leap yearj except Julian date, modiffed (MJD) the .Juliandate minus 2400000.5.
for centurial ycars, which must be cxactly divisiblc by 400 to be lcap years.
Julian day number (JD) the integral part of the Julian date.
Thus 2000 is a leap vear, but 1900 and 2100 are not leap years.
Julian proleptic calendar the calcndric system employing the rules of the Julian
height elevation above ground or distance upu'ards from a given level (especially calendar, but extended and applied to dates preceding the introduction of the
sea level) to a 6xed point. (Seealtitude.) .Julian calendar.
 GLOSSARY 731
730 GLoSSARY
longitude, celestial angular distance on the celestial sphere measured east$'ard
Jglian year a period of 365 25 days. This period served as the basis for the Julian along the ecliptic from the dynamical equinox to the great circle passilq
calendar. through the poles of the ecliptic and the celestialobject.

Laplacian plane for planets see in'ariable plane; for a system of satellites. the fixcd longitude, terrestrial angular distance measured along the Earth's equator from
plane relative to which the vector sum of the disturbirg forces has no orthog- the Greenwich meridian to the meridian of a geographic location.
onal comPoncnt
luminosity class distinctions among stars of the same spectral class. (See spectral
latitude, celestial angular distance on the celestial sphere measured north or south types or classes.)
of the ecliptic along the great circle passing through the poles of the ecliptic
and the celestialobject. lunar phases cyclically recurring apparent forms of the Moon. New N{oon, First
Quarter, Full \Ioon, and Last Quarter are defined as thc times at which the
latitude, terestrial angular distzrnceon the Earth measured north or south of thc excessof the apparent celestial longitude (seelongitude, celestial) of thr: NIoon
equator along the meridian of a geographic locatiou. over that of the Sun is 0". 90", 180", and 270o,respectivell..

leap second a second (seesecond, SystEme International) :rdded betwccrt 60' and 0" lunatiou the period of time between two consecutive New Nloons.
at announced times to keep llTC t'ithin 0:90 of UT1. Generalll,, leap seconds
are added at the end ol June or Decentber. magnitude, stellar a measure on a logarithmic scale of the brighlness of a celestial
object considered as a point source.
librations variatious in the orientation of the Nloon's surface with respect to an
observer on the Earth. Physical librations are due to variatious in the rate magnitude ofa lunar eclipse the fraction of the lurrar diameter obscurecl by the
at which the \,loon rotates on its axis. The much largcr optical librations are shadorv of the Earth at the greatest phase ol a lunar eclipse (see eclipse,
due to rariations in the rate of the l{oon's orbital notion. the obliquity of luuar). measured along the common diameter.
the \'Ioon's equator to its orbital plane, and the diurnal changes of geometric
perspective of an observer on the Eartlt's surface. magnitude ofa solar eclipse the fraction of the solar diameter obscured by the
\{oon at the greatest phase of a solar eclipse (see eclipse, solar), tneasurcd
light, deflection of the bending of the beam of light due to gravity. It is observable a l o n g t h n r o m r n o nd i a n r e l e r
when the liglrt from a star or planet passesa massilc object such as the Sun.
mean anomaly in undistrrrbed elliptic motion, the product of the mean motion of
light-time the interval of time rcquired for Iight to travcl from a cclestial bodv an orbiting body and the interval of time sincc the body passedpericenter.
to the Earth. During this interval the motion of the body in space causcs Thus thc nrean anortrall is the anglc fi'om pericenter of a hypothetical botll'
an angular displacement of its apparent place from its geometric place (scc moving q'ith a constant angular speed that is equal to the mean motion. (Sce
geometric position). (See aberration, planetary.) true anomaly; eccentric anomaly. )

light-year the distarrce that light tlaverses iD a vacuutl clrrring one ycar. mean distance the semi-major axis of an elliptic orbit.

limb the apparent cdge ol the Srrn, \foou. or a planet or arrl other cclostial bodv mean elements elements of an adopted reference orbit (see elements' orbital) that
with a detectabledisc. approxirnatesthe actrral, perturbcd orbit. Nlcan elemeuts rna), serve rrs the
basis for calculating perturbations.
limb correction corrcction that must be made to the distance betwecl the center
ol mass of the \Ioon and its limb. These correctionsare due to the irregular' mean equator and equinox the celestial reference s1-steurdetcrnrined bl ignoling
surface of the Nloon and arc a function of the librations in longitude (scr: small variations of sholt period in the motions of the celestialequator. Thus
longitude, celestial) and latitude (seelatitude, celestial) and thc position angk: the mean equator and equinox arc affected only by precession. Positions in
fron the central meridian. star catalogs are normally referred to the mean catalog equator and equinox
(sce catalog equinox) of a standard epoch.
local sidereal time the local hour angle of a catalog equinox.
732 GLOSSARY
GLOSSARY 733
mea-nmotion in undisturbed elliptic motion. the constant angular speed rerlttired obliquity in gencrarlthe angle br:lween the equa,tolia.larrcl orbilnl plaucs of a bocly
for a body to complete one revolution in an orbit of a spccifled semi-major or. equir,aleni,lv,between the rot.ationaland orbital poles. Rrr the Earth tlro
axis. obliquity of llrt: ecliptic is thc angle bet$'eortthe planes of thc equator arrcl
the ecliptic.
mean place the coordinates, referred to the mean equator and equinox of a stan-
dard epoch. of an object on the celestial sphere centered at the Sun. A rnean occultation thc obs<rrration of one cclcstial bodv bv anothcr of groater ilpparcnt
place is detcrrnined by rcrnoving from thc directly observctl position thc ef- cliarneter; espr:cially the passage of the Nloorr in front of a star or planet. or
fects of refraction, geocentric and stellar parallax, and stcllar aberration (scc the disappearanceof a satellit<:behind the cliskof its priurarv. If 1,heprinrarv
aberration, siellar), and lry referring the coordinatcs to the mean equator source of illuuriuation of a refic<ting body is <:ut off by tho <xcrrltation, thr:
and equinox of a standard cpoch. In cotrtpiling star catalogs it has been the phenomenonis also called an eclipse.The occrrltationof the Sun b1.the \Ioon
practice not to remove the sccular part of stellar aberration (see aberration, is a solar eclipse(seeeclipse,solar.)
secular). Prior to 198,1.it u'ns additionally tlle practicc llot to remove the cl-
liptic part of annual abcrration (see aberration, annual; aberration, D-terms opposition a configruation of thc Srrn, Earth and a plarrct in ivhich tll-' appir,rent
of). geocentric krrrgitude (sec longitude, celestial) ol the plarrct rliflcrs by 180.
frorn the apptrrcut geocentrickrrrgitudeof the Srrrr.
mean solar time a neasure of titttc basedconccptualll'on the clittrnalmotion of the
orbit the path in space follorvedbv a cclestialbodr'.
ffctitious mean sun. urrdcr the assumptiorttltat the Earth s rate of rotatiorl is
constant. osculatingelements a sct ol paranlclors (see elements,orbital) thal specifiesthc
iDslantaneousposition ancl vclo<itv of a celestinlbody iu its lrt'r'lrrrbed orbil,.
meridian a great <rirclepassing through the celestial poles ancl through the zenith Osculatingelerrx.rrts describetlrr. unperturbecl(tu.o-bodl') (n bil tlrrt thc boclv
of ant location on Earth. For planetary obserrations a nreridian is half thc u ould follos' if perturbations $<.Ic to cease inst zurtaneousl\'.
great circlc passingthrough the planet s poles and througlt aur' location on
the planet. parallax thc diffelen<c irr appalent <lirottion ofan objcct as seenflorn 1t'o differcnt
locrtions: convcrscly,the anglc al the object lhnt is subtenrk'rl by thc lin<:
month the period of one complctc slrnodicor sidclcal revolutionof tho \'{oon aroltrd joirrirrgtrvo desigrra.lodpoints. Gtnccntric (diulral) parallax is 1hc differencrr
the Earth; zrlsoa calendricalunit that appfoxilnatesthe perioclof revolutiotr. geocentric
irr rlircction bctrv*l a topocentric obselvation arrrl a h1-potlrr'1icrl
olrscrvation.Hcliortntric or annutl palallax is thc diffcrcrr<r:lrt'tneen hr'pr>
moonrise, moonset thc times at rvhich the apparcnt uppcr limb of the \IooD is ou
thcti<al geocculric and heliocerrlricobserratious:it is thc arrglcsul)1.cD(le(l at
the astronornicalhorizon; i.e.. rlhen thc tnrc zenith distance, referred to tlt<r
1,lrcobserr-edobjr:ct b1'the semi-major a-ris of lhc Earth's orbit. (See alsrr
center of the Earth, of the <trrt,ralpoint of the disk is 90'34' + s - z. u'heres is
horizontal pa.rallax.)
the Moon's semidiameter. r is the horizontal parallax. ancl 34' is the ad<4rtcrl
valrrcof horizontal refractiorr. parsec llx.distancenl i\lrich one astronomicalunit sulrlt'rrclsarr artglt'ofoDc secolt(l
of ar<: cquiralt'nlh. the distarrrt to au objcct havirtg an artttrtalparallax ol'
nadir the poill on thc celestialsphere diaructricallv oppositc to the zenith. orlr:sttond of nt<.

node either ol the points on tho celestial sphere at which thc pJaue of an orbit penumbra ttrc porl,iottof a shaclorvirr rvhich light frorn an extetrdo<lsollr(:eis p2r1-
iltersects a u:fcrenceplanc. The position of a node is olc of the stancliud 1i;rllv but rrol <rrrupletell'crrtoff by an inter\'('lringbody; thc alczrof paltial
orbital clcux:nls (seeelements,orbital) uscd to specif-r'tlrc olicutation of atr shados'surrorrrr<ling the umbra.
orbit.
pericenter the poirrt il au orbit th;rt is nearestto tlrc ccuter of forct:. (Sct:perigee:
nutation the short-yrcriodoscillations in the rrxrtion of the pok. of rotation of n perihelion.)
freely rotarting bodv that is rrndergoingtolquc fiorn extclnal gravitafional perigee the point al w'hich a bocly iu orbit arotrtrl thc Earth nnsl closely ap-
forces. NutatioD of the Ezrrth's pole is cliscrtssr:d in tcrns of corDponentsin proaches the Ealth. Perigee is sotnctiures used u.ith reference to tlxr itpparent
obliquity zurtl longitude (scc longitude, celestial.) ()rbit of the Srur around the Ealth.
 GLoSSABY GLOSSARY 735
734
precessionthe uniformly progrcssing motion of the polc of rotation o[ :r freelr.
perihelion the point at rvhich a bodv in orbit around the Sun most closely ap-
proachesthe Sun. lotating bodl undergoingtorque from external gravitational forces.In the case
of the Earth. thc comporrcntof precessioncausedbv the Sun and \Ioorr acting
period the interval of time required to complete one rcvolution in an orbit or one on the Earth's equaiorial bulge is called lunisolar precessiolt the conrponent
cycle of a periodic phenomenon,such as a cycle of phases.(Seephase.) causedby the action of the planets is called planetary precession.The sum of
perturbations deviations between the acttral orbit of a celestial body and an as- lunisolar arrd planetary precessionis calle<lgeneral precession. (See nutation.)
sumed referenceorbit; also, the forces that cause deviations between the actual proper motion the projection orrto the celestial sphere of thc space nx)tiur of a
and referenceorbits. Perturbations, according to the first mearing, are usually star relative to the solar systern;thus the transversecomponent of thc space
calculated as quantities to be added to the coordinates of the reference orbit motion of .i star with rcspect to the solar system. Proper motiotr is usuallv
to obtain the precise coordinates. tabulated in star catalogs as changes in right ascension and declination per
phase the ratio of the illuminated area of the apparent disk of a celestial body to J,earor ceDtur)'.
the area of the entire apparent disk taken as a circle. For the Nloon, phase
quadrature a configuration in utich two cclcstial bodies havc apparent lorrgitudes
designations (see lunar phases) are defined by speciflc configurations of the
(see longitude, celestial) that difler by 90" as viervc<l from a tirir.d body.
Sun, Earth and Moon. For eclipses,phasedesignations(total, partial, penurn-
Quadraturcs are usualll' tabulated u'ith respect to thc Sun as r-iervcd fronr
bral, etc.) provide generaldescriptionsof the phenomena.(Seeeclipse,solar:
the center of the Earth.
eclipse, armrlar: eclipse, lunar. )
radial velocity thc rate of chtrnge of the distauce to an objcct.
phase angle the angle measured at the center of an illuminated body between thc
light source and the observer. refraction, astronomical the change in dircctiorr of travel (bcrrrling) of a liglrt lny as
it passesobliquely through the atmospht're.As a result of refractiou. tlrt: ob-
photometry a measurement of the intensity of light usually specified for a specific
serr,edaltitude of a celestialobject is grcatef than its goometric altitu<le.The
frequency range.
amount of rcfraction depentlson the altitutle of the oi.rjt:<:t
and on atnrospheric
planetocentric coordinates coordinatcs for general usc, where the z-axis is the mean conditions,
axis of rotation; the r-axis is the intersection of the planetary equator (normal
retrograde motion for orbital nlotion in the solar system, rrlotion that is clockwise
to the z-axis through the center of mass) and an arbitrary prime meridiar: and
in the orbit as seen from thc lrorth polc of the ecliptic: for trn object obselved
the y-axis completes a right-hand coorditrate system. Longitude (seelongitude,
on the celestial sphere. ulotioD that is from east to &'est. rcsulting froI the
celestial) of a point is rneasured positive to the prime meridian as defined by
relative rnotion of the objcct aud the Eiuth. (See direct rnotion.)
rotational elements. Latitude (see latitude, celestial) of a point is the angle
between the planetary equator and a line to the ccnter of mass. The radius is right ascension ertrgulardistarrce on the celestial sphere rneasrtred eastlvar<ltrlr>ttg
measured from the center of mass to the surface point. the celestial equator frolrt the equinox to the hour circle passing tluottglt the
planetographic coordilates coordinates for cartographic purposes depcndent on au celestial obje<:t. Right ascelrsionis usually given in combination rvith declina-
equipotential surface as a reference surface. Longitude (see longitude, celes- tion.
tial) of a point is nreasured in the direction opposite to the rotation (positive Saros a Babylonian lunar cycle of 6585.32 days, or 18 ycitrs 11.33 dat's, or 223
to the west for direct rotation) from the cartographic position of the prime lunations. at the end of which the centets of the Strn and \'Ioon retrtrn so
meridian defined by a clearly observable surface feature. Latitude (see lati- nearly to thc relative positions of the beginning that all the cdipst:s (see
tude, celestial) of a point is the angle between the planetary equator (norrnal eclipse) of the period recrrr approximately as before, but in longitudcs (see
to the z-axis and through the center of mass) and normal to the reference longitude, terrestrial) approxirnately 1200 to the west. (See lunar phases.)
surface at the point. The height of a point is specifled as thc distance above
a point with the same longitude and latitude on the reference surface. satellite natural body revolving around a planet.

polar motion the irregularly varying motion of the Earth's pole of rotation with satellite, artificial device launched into a closed orbit around the Earth, another
respect to the Earth's crust. (See celestial ephemeris pole.) olanet. the Sun. etc.
 736 GLOSSARY
GLOSSARY 737
second,SystbmeInternational (SI) the cluration of 9192631770ocles of radiation surface brightness (ofa planet) the visual rnagnitude ofan averagc square arc-second
corresporrtlilg to thc transition betnccn tu-o hvpetlinc k:r'elsof tire ground
area of thc illuminated portion of the apparent disk.
state of cesiurrl133
synodic period for plancts, the mcan inter\,.al of time between successiveconjrurc-
selenocentric with rcltrcnce to. or perla"iriirrg to, the centcr ol the Nloon. tions of a pair of planets. as obse.ved from the Sul; for satellites, the rnean
intelval between slrccessiveconjunctions of a satcllite with the Sun. as ob-
semidiameter the anglc art the obsen'er sublcrrded bv the equatorial radius of thc
s e r r " t l f r o r n r h e s a r e l l i r es p r l m a r l .
Sun, Nloon, or a planet.
synodic time pcrtaining to successiveconiunctions; successiveretrrrns of a planet
semi-major axis half thc lcngth of the major axis ol an eliipsr:; zr standard elemett
to the same aspect as determined by Earth.
used to describc an elliptical orbit (scc elements,orbital).
Terrestrial Dynamical Time (TDT) thc indcpendent argunlent for apparerrt geo-
sidereal day the iltenal of tirnc bet$'een tu,o consccutive transits of the catalog
centric ephemerides.At 1977January 1d00h00'00' TAI. thc vahreof TDT rvas
equinox. (Scc sidereal time.)
exactlv 1977 .lanuarv 190003725.The unit of TDT is 86,100SI secondsat
sidereal hour angle angular distance on the celestial sphere measured rl"'est',r,arc1 niealr sea level. For practical pulposes TDT = TAI + 32:184.(SeeBarycentric
along the celestial equator from the catalog equinox to the hour circle pass- Dynamical Time: dynamical time; International Atomic Time.)
ing through the celestialobjcct. It is equal to 360' rninus right ascensionin
terminator the boundary beirveen the illuminated and dark ir,rcasof the apparent
degrees.
disk of the N{oon. a planet or a planetary satellite.
sidereal time the rneasure of time delincd bl' the apparelt diurnal motion of the
topocentric rvith referenceto. or pertaining to. a point on tlrr: surla.ccof the Darth,
catalog equinox; hence a measure of the rotatiorl of the Eilrth $'itlt respect to
usuallv ri'ith reference to a coordinate system.
the stars rather than the Srrrr.
transit the passageof a celestial object across a meridian; also the passageof one
solstice either of the t$'o poirrts on the ecliptic at rvhich the apparent longitude
celestialbody in front of another of greater appare[t diameter (e.g.. the pas
(see longitude, celestial) of the Sun is 9Uo or 270"; also the time at u,hich the
sageof Nlercury or Venusacrossthe Sun or Jupiter's satellitesacrossits disk);
Sun is at cither point.
hor.ever. the passageof the \'Ioon in front of the larger apparent Srrrt is called
spectral types or classes catagorization ofstars according to their spcctra, primar- an annulal eclipse (see eclipse, annular). T'he passage of a body's shadon'
ily due to differing temperatures of thc stellar atmosphere.From hottest to across another bodv is called a shadorv transit; holvever. the passageof the
coolest.the spectral types are O. B, A, F. G. K. ald M. Nloon's shadorv across the Earth is called a solar eclipse. (See eclipse, solar.)

standard epoch a date and timc that specifiesthc relerencesystem to which cdcs true anomaly the angle, rneasuredat the focus nealest the pericenter ofan ellipti<:al
orbit, between the pericenter and the radius vector florr the lircus t,, thc
tia"l coordinatcs are referrcd. Prior to 1984 coordinates of star catalogs rvcrc
orbitirlg bodl r one of the standard orbital elements (sr:r:elements, orbital).
commonly rcferred to thc mean equator and equinox of thc beginning of a
(See also eccentric anomaly; mean anomaly.)
Bcsselian year (see year, Besselian). Beginning rvith 198.1the Julian year has
becn used. as denoted by the prefix J, e.g., J2000.0. true equator and equinox thc cclcstial <rror<linatc system detelnincd bt- t,he in-
stantancous positiols of the celestial equatirr and ecliptic. The rtrotion of this
stationary point (of a planet) the position a,t which thc ratc ol charrgc of the appar-
svstem is due to the progressive effect of precession an<l the short-terrn, peri-
ent right ascension (see apparent place) of a plaiei, is momcnt.irily zero.
odic rariations ol nutation. (See mean equator and equinox.)
sunrise, sunset thc timcs at which the apparent upper. Iirnb of the Sun is on the
twilighi the intenal of tirnc prcceding sunrise and follon'ing sunset (see sunrise'
astronomicalhorizon; i.e., whcD the tlue zenith distance,referrcdto thc center
sunset) during $'hich the sky is partially illuminated. Civil trvilight comprises
of the Earth, of the central point of thc disk is 90'50'. bascdon adopted values
the intr:rval rvhen the zenith distance, referrcd to the center of the Earth, of
of 34' for horizontal refraction and 16' fbr the Sun's semidiameter.
the central point of the Sun's disk is between 90'50' and 96', nautical twilight
738 GLoSSARY GLOSSARY
73g
comprises the interval from 96o to 102o, astronomical twilight comprises the zenith in general, the point directly overhead on the celestial sphere. The astro_
interval from 102' to 108'. nomical zenith is the extension to infinity of a plumb line. The geocentric
zenith is defined by the line from the center of the Earth through the observer.
urnbra the portion of a shadow cone in which none of the light from an extended The geodetic zenith is the normal to the geodetic ellipsoid at the observer,s
Iight source (ignoring refraction) can be observed. location. (See deflection ofthe vertical.)
Universal Time (UT) a measure of time that conforms, within a close approxima- zenith distance angular distance on the celestial sphere measured along the great
tion, to the mean diurnal motion of the Sun and servesas the basis of all civil circle from the zenith to the celestial object. Zenith distance is g0. minus
timekeeping. UT is formally defined by a mathematical formula as a func- altitude.
tion of sidereal time. Thus UT is determined from observations of the diurnal
motions of the stars. The timescale determined directly from such observa- zodiaca.llight a nebulous light seen in the east before twilight and in the west
tions is designated UT0; it is slightly dependent on the place of observation. after twilight. It is triangular in shape along the ecliptic with the base on the
When UTO is corrected for the shift in longitude (seelongitude, terrestrial) of horizon and its apex at varying altitudes. It is best seen in middle latitudes
the observing station caused by polar motiou, the timescale UT1 is obtained. (see latitude, terrestrial) on spring evenings and autumn mornrngs.
Whenever the designation UT is used in this volume, UT1 is implied.

vernal equinox the ascending node of the ecliptic on the celestial equator; also the
time at which the apparent longitude (seeapparent place; lougitude, celestial)
of the Sun is 0'. (See equinox.)

vertical apparent direction of gravity at the point of observation (normal to the

plale of a free level surface.)

week an arbitrary period of days, usually seven days; approximately equal to the
number of days counted between the four phases of the Moon. (See lunar
phases.)

year a period of time based on the revolution of the Earth around the Sun. The
calendar year (see Gregorian calendar) is an approximation to the tropical
year (see year, tropical). The anomalistic year is the mean interval between
successivepassagesof the Earth through perihelion. The sidereal year is the
mean period of revolution with respect to the background stars. (See Julian
year; y€irr, Besselian. )

year, Besselian the period of one complete revolution in right ascension of the
ffctitious mean sun, as defined by Newcomb. The beginning of a Besselian
year, traditionally used as as standard epoch, is denoted by the suffix ".0".
Since 1984 standard epochs have been defined by the Julian year rather that
the Besselian year. For distinction, the beginning of the Besselian year is now
identified by the prefix B (e.g.,81950.0).

year' tropical the period of one complete revolution of the mean longitude of the
sun with respect to the dynamical equinox. The tropical year is longer than
the Besselian year (see year, Besselian) by Gl48I, where 7 is centuries from
81900.0.
 lndex
This index doesnot include entriesfrom Chapter13 (Historical Information), Chapter14
(RelatedPublications),or the Glossary.Material in Chapters13 ard 14 can be located
more effectively tbrough the Table of Contents.The Glossary provides definitions of
nomenclaturelisted in alphabeticalorder

A Apparentephemerides,307-308
Abenation, 23, 127-135, l5O Apparentplacesof fundamentalstars,5 I 0
$n\al21 129-130 Apparcntposition,23, 145-160,165
constantof, 131,170 Apparentsidercaltime, .144
coFectionfor light-time, 23, 393 Apparentsolar time, 74
daynumbers,130,169
Aryumentof pericenter,27
differential, 134
diumal,129,132,160 Ariel (UranusI), 709,7 I I
elliptic (.reeE-terms) appare[tdistanceandposition angleorbital
E-terms,131,169 elements, 368
planetatry,23,127,133 physicalandphotometricdata,412
secular,129 Aries, first point of (ree Equinox)
stella\ 23, l2'1- 128, l3O
Ascendingnode,majorplanets,315-316,704
Adrastea(JupiterXV), 4O9,708,'l lO
Airglo% 493 Asteroids(seeMinor planets),3 I 0
Albedo Astrogmphicplace, 167
geometricof asteroids,415 Astrometricposition, 23, 166,308,3ll-313
Allan variance,59 Astronomicalconstants, 280, 289,302-303,
Altitude, 14
31'7*319
Altitude andazimuthformulas,208
Amalthea(JupiterV) , 4O9,'7O8,'7 coordinates,16,2O2--2O7
l0
oftital elements,349 Astronomicalunit, 27
AmericanAssociationof VadableStar Adas(SatumXV), 410,7O9,711
Observers(AAVSO), 513-5 14 Atmosphericloading,244, 248-249
Analemmiccurve,485 Atomic time, 39, 84
Analytical theories,27
Aurora, 493
Ananke(JupiterXII), 708, 710
stalevector,353 Axis of figure, 14, 109, I l0
Angular velocity, Barth,51 Azimuth, 14
741
 INDEX
742 INDEX 743

B planetographic coordinates,385-388 Constants, 693 Deimos(MarsII), 385,708,710

height,387 astronomical, 696 orbitalelements, 343
Barycentriccoordinates,I I laritude,387,395 fundamental,693 physicalandphotometicdata,4O'1
BarycentricCoordinateTime{TCBl.3. 45 longitude,38?,395 gravitational,696 reference surface,385
BarycentricDynamical Tim€ (TDB),2, 42 referencesurface,385 mathematical, 714 DeltaT, 54, 81, 83,265-269
Belinda(Uranus XIv), 412,'709,'l l l primemeridian,384 Constellations, Tl3 Desdemona (UranusX), 412,'709,7 | I
elements
Besselian (.te?
Eclipses) Mars,403 Conventional Intemational Origin(CIO), 15, Despina(NeptuneV), 709,711
solarYear,8
Besselian Mercury,402 53,tto,139,223
Diameter
Bianca(UmnusVIII), 412,'l09,'1| | Pluto,406 Conventional te.restrialreferencesystem,200,
apparentpolar, 397
BilliancY,479 223224
Sun,398 asteroids,radiometricmethod,4 I 4-4 I 7
Pluto,406 Coordinated UniversalTime(UTC),6, 53,76
Satum'ssatellites, 41I equatorial, 397
Sahrm'srings,404 Coordinates, astronomi cal,202-20'7
selenographic, 399-400 polar,396
barycentric, 11, 164
colongitude,399 Differential abenation,I 34
conversion from B1950.0to J2000.0,167,
Catalogddft, 293 184 Differential nutation, 120
Catalogs Differential precession,I 08
Eeocentic,202-207
c compiled,505 geodetic,202-207,2!7 218 Dione(SatumIV), 708,710
differential,506 84d,209 2t0 orbitalelements, 356
Calendar, 575 fundamental, 506 loca'|,2O'7-216 physicalandphotometric data,4l0
accuracy, 578 observational, 505,508 planetocentric, 31, 37, 385-388 Dip, 488
Chinese, 577,594,599 absolute,507 planetographic, 37, 385-388 Distance,apparent, 392,395.396
Gregorian, 577,580,583,604 differential,507 rcductionsbetweenapparent andmean Dopplermethod,62
Hebrew,577,584,588 fundamental, 501 places, 145,152 Dopplersatellitetacking,251 258
Indian,591,594,605 reference,507 referenceframe,8 Doublestars, 515-516
Islamic,577,589,591,604 Celestialcoordinates, 20 selenographic, 399 Dynamical time,39,41,54,63
Jewish(seeHebrew) Celestialephemeris pole,109,I 10,139 stateplane,214216
Julian,600,605 Celestialreference system,199 topocentric,10,160-165
lunar,577 Celestialsphere,10 transformations, 555-559
lunisolar,577 Ceres,697 Cordelia(UmnusVI), 412,'7O9,'7 ||
'111
Moslem,(seeIslamic) 60 Cressida(UranusIX) , 412,'7O9,
Cesium-beam standards,
relorTn,) /6
Chandlerwobble,257
solar,577
Charon(PlutoI), 709,7l I
E
Callisto(JupiterIV), 708,710
o6ital elements, 377
conjunctions, 348 D
physicalandphotometdc data,4l4 Eafih, apheliorl,4'7'7, 479
orbitalelements, 345
Chebyshev polynomials, 305-307 figure,200,700
phenomena, 348 Datacenters, 509
physicalandphotometric Chronological cyclesanderas,579 $avlty field, 224-233
data,409 Centrede DonndesStellaires, 509
Christian,579 mass,697
Calypso(SatumXIV), 410 GoddardSpaceFlight Center,509,513
Hebreq 584 orbitalelements, 316,700,704
statevector353,708,710 Datums'16,216-223
Came (JupiterXI), 708,710 lndian,59l orientation,199
Daycorrections, 292
Cartographic lslamic,589 perihelion,477,479
coordinates,
383-388 Day numbers, Besselian,157,169
Galileansatellites, 408 409 Circulation,331 49, 75 physicalandphotometric data,706
Greenwichsidereal,
longitudesystems,403 Clockperformance, 58-61 independent, 157-158 rotation,251
SysremIII, 403 Comets, 3l0 3l5 Julian,8, 55, 580,600,604 rotationelements, 701,705
planetocentric coordinates,
385-388 Commensurability, 330 second-order, 158-160 rotationmeasurements, 6143, 76
latitude,387,395 Configurations, 47fu82 Defectof illumination,397 rldes,233-238, 283-284
longitude,387 Conjunctions, 33,479 Deflectionof light, 24 Easter,581,583
range,387 Connected-Element 258
RadioInterferometry, Deflectionof thevertical,204 F.{:cenfi
i,city,27, 3 | 5-3 | 6, 704
 INDEX
744 INDEX
745
34,421 partial,431 Flattening,major planersand Moon, 398_407
Eclipses, Greenwichtransit,54
Bessel'smethod,434 positionangle,463 apparent,390 Gregorian calendar,7
centerof figureoffset,424 total,433 Mercury,
402 Gyroscopes,26 I -262
time scale(DeltaT), 425,467 Forcemodelevaluation,
2g7
Circula.s,423
umbm,429,438 Frequencies,standard,5g
circumstances, 421
E c l i p t i lcl , 1 1 4 Fundamentalephemerides,279 2gl
elements, 421
elements, orbital,315-316 Elara(Jupiter VII), 708,710 H
ellipsoid,200,202,206.218-220,223 statevector,353
Helene (Saturn XII), 410..709,71|
ephemeides,424425 Elements, Besselian (.r?eEclipsesand G Heliocentric
coordinates, II
lunar,35 Occultations)
Galatea(NepruneVt). 709, 7t I Heliocentric
planetaryphenomena, 4gl 4g2
elements,468 469 Elements, orbital,315 316,704
Besselian Hirnalia(JupirerVI), 708,710
Effipsoid,200,202,206,218-220,223 Galaxies,bright,5l I
contacts, 467,469 statevector,353
EIongations,33 Second Reference Catalogue of Bright
magnitude, 470 Horizonsystem,208 209
Enceladus (Saturn II), 708,7l0 Galcxies,St 1,528
obscuration,469 Third Reference Catalogue of Bright Horizontalparallax
orbitalelements, 356
occurrence, conditionsfor. 426'431 Galaries, S3O Moon,400
physicaiandphotometric data,410
paftial, 430,46'7, 470 X-ray sources,533 Sun,398
Energyconversion factors,715
penumbral, 430,467,470 calilean sarellires,
345
Hubblesequence, 528
Ephemerides, fundamental, 279
phases, 470 Ganymede(Jr.rpiter III), ?08,710 Hydrogenmasers. 60
initialconditions, 304
position angle,470 conjunctions,348 Hype.ion (SatumVII), 385,4l l, 708,7t0
Ephemeris meridian,54-55
total,431,467,4'10 orbital elements,345 differentialcoordinates,
367
Ephemeris time,2-3,42,'16,'19, 81 82
visibility,470-471 phenomena,348 orbitalelements, 360
Ephemeris transit,majorplanets,309
penumbra, 428,438 physical and photometric data,409 physicalandphotomerric dara,385,4t I
Epimetheus (Saturn XD, 410,709,7l l
publisheddata,422-424 Epochs,27,698 Gaussiangravitationalconstant,27
satellitesof Jupiter,348 Equationof center,327 Gegenschein,493
solar,35 Geocentdccoo.dinates, 10, | 26, 202-20.7
Equationsof motion,26
GeocentricCoordinateTime (TCG), 3, 45
I
annular,433 E q u a t i oonf t h ee q u i n o x e s . 4 . 4 8 .l5l 6J .
annular-total, 433 Geocentricplanetaryphenomena,47g Iapetus(SarumVItI), 4|,'t08,7 tO
Equationof time,5, 55, 74,485
Besselian elements(seeelements) Geodetic coordinates,16,126, ZO2 201, differential coordinates,367
EquatorI l, I 14,386,393
centralline,455-456 Equinox,catalog,48 49. 54,97, 293 2t'|-2t8 orbital elements.360
circumstances,
Geodetic
datums,702 physical and photometric data,4l I
general,421. 446,450, dynamical, I 1,55,97,288-289
Geodetic referencespheroids,701 Illuminateddisk, 388-397
459461 Equinoxconection, |67-l68
Geographiccoordinates,l6 Illumination, 490-492
circumstances, local,46l -467 Equinoxof date,75
Geoid,17 planets,401407
contacts, 459-460,462 datesof, 477
undulation,202 Illumination. defective, 397
curves,450,453-458 Enors,566_568
Geopotenrialheld, 22'l 233 I mprovedLunar Ephemeris, 81
duration,455,458 Eulerangles,284-285
Geopotentialsurface,201 Inclination,2T
elements,450 Europa(Jupiter II), 708,?10
Global Positioning Sysrem (GpS), g, 68, lnitial conditions,304
Besselian, 435-441 conjunctions,348
259-261 Intemational Aromic Time (TAI), 2, 40
Besselian, auxiliary,445,449 orbitalelements, 345 Globularstarclusters,5l l, 526 Intemational Earrh Roration Service (IERS),
elevation,correctionfor, 467 phenornena, 348 Gravitational model, 280 86, I tt,2N, 223_224,262,265
equations of condirion,446-450 physicalandphotometric data,409 Creek alphabet,7l7 Interpolation,546-548
limb conections,lunar,425 Greenwich civil time, 76 lnvariable plane, ll
rnagnitude, 458,463 464 Greenwich mean astronomicaltime, 76 Inverseinterpolation,545, 54g
maps,423, 462 Greenwich meantime, 7, 76 Io (Jupiter I), 708, 710
F Greenwich meridian, 384, 386, 393
obscuration, 463,465 466 conjunctions,348
observer's coordinates, 441446 Figure effects,282 Greenwich siderealangle, 2l obital elements,345, 708
occulrence, conditionsfor, 426-128, First point of Aries (seeEquinox), 12 Greenwich siderealdate,49, 75 phenomena,348
432434 FKs, 167 Greenwich siderealtime, 4,49-53, l16 physical and photometric dara,409, 710
 746
INDEX 747
J Librations,satellite,33I
Mals Moon
Light, deflectionof, 135-139,166
Janus(SatumX)' 410,708'710 velocityof, 693 defectiveillumination,403 defectiveillumination,399
Juliancalendar,7 Light-time,127,148 magnitude, 403 eclipses(seaEclipses,lunar)
Juliandate,8' 55' 600 corectionfor, 23,393 mass,697 eph€meris, basisof, 317
modifred,8 fractionilluminated,399
Local apparentsolartime, 4 orbitalelements, 316,704
Juliandaynumba$,8,55, 580,600,603 horizontalparallax,400
Localcoordinates, 207-216 physicalandphotometricdata,706
Julianperiod,580 librations,398-400
Localhourangle,14,55 physicalephemeris, 403
Julianyear,8 optical,398
Localmeantime,5, 76 rotationelements, 403,705
Juliet(UranusXI), 412,709,711 physical,398
Localmeridian,I l, 55 satellites,
342,403,708,710
Juno,697 topocentric,
399-400
Jupiter Localplace,165 Masses, 697 limb corrections,425
magnitude,404 Localsiderealtime,4,49,52-53
Mathematical model,28I mass,696
mass,697 Longitude(ree a/soCoordinates)
nlrnat^.enr;. le?
Matdces,552-559 meanlongitude,115
orbitalelements, 316,704 Maaix methods,180 midpointof brightlimb, 399
phase,404 planetographic,387,395
Maximumdefectof illumination,397 node,115
physicalandphotometricdata,706 Longitudeof node,27
Meansealevel,201-202 orbitalelements, 701,708
physicalephemeris,403-404 LORAN-C,66
Meansiderealtime,4, 48, 75 perigee,I 15
rotationelements,404, 705 Love numbers,31,233,250 phases,478-479
345,409,708,710
satellites, Luminositt Iapetus,4l I Meansolartime,3, 74-75,79
physicalandphotometric dara,398 401,
Jupiter'ssatellites,
708,710 Luminosityclass,brightgalaxies Mercury '101,706,110
I-IV 345,409 planetocentdc, 387 brilliancy,402 physicalephemeris, 398301
Y 348,409 planetographic, 387,395 elongations, 479 principalmomentof inenia,399
vr-xllr, 352,409 Luriareclipses(.reeEclipses,lunar) magnitude, 402 riseandset,485-486
XIY 35I Lunarephemeris,3l7 mass,697 secularacceleration,
77, 84
LunarLaser-Ranging, 62, 299-300 orbitalelements, 316,704 selenographiccoordinates,399-400
Lunarlibrations,284-286,398-400 physicalandphotomeric data,402,106 semidiameter,401
K geocentricoptical, 398
physicalephemeris, 401 403 topocentricparallax,400
physicalrotational,398 Moonlight,493
Kepler'slaw,26 rotationelements, 402,705
topocentricoptical, 399-400 Moonriseandmoonset,32, 485486
tuansit,471
Lunarsecularacceleration, 288
Lysithea(JupiterX), 708,710 Mercury ion frequencystandard,6l
L statevectot 353 Meridianpassage (seeTransit,meridian,33
Metis(JupiterXVI), 708,710 N
Laplacianplane,327
Minas (SatumI), 708,710 Naiad(NeptuneIII), 709,711
Larissa(NeptuneVII), 709,711
orbitalelements, 356 Nauticaltwilight,33,483,492
Laserranging,254-255
l-atitude (seedLroCoordinates) M physicalandphotometric data,410 Neptune
planetocentric,
387,395 Minor PlanetCenter,310 magnitude, 406
plaretographic, 38?,395 Magnitude Minor planets,310-315 mass,697
Lawsof motion,24 asteroids,414-416 geocentric phenomena, 4lMl7 orbitalelements,316,704
l-eapsecond, 7, 53 infrared,415 magnitudes, 3l I physicalandphotometric data,405-406,
Leastsquares, 300-301,568-573 visual,414 Miranda(UranusV), 709,711 706
lada (JupirerXUI), 708,710 majorplanets(seeindividualplanets) physicalephemeris, 405-406
orbitalelements, 368
statevector,353 at I AUand0o phase,388 rotationelements, 705
physicalandphotometicdata,4l2
Lengthof day, 19,265-269 meanopposition,389 373,4l3 414,709,7|I
satellites,
MK spectralclassification system,520 Nereid(NeptuneII), 709,7l I
Librations,Moon,398,400 minorplanets,414-416
ModifiedJuliandate,56 differentialcoordinates,376
geocentric optical,396 infrared,415
physicalrorational,398 Month,577 orbitalelements, 375
visual,414
topocenric optical, 399-400 i lengths,698 Newcomb's Tables,317
Satum'srings,404
synodic,486,576 Nonrigid-body effects,286
 748 INDEX INDEX 749
Northpole,384 385,392-393 Pasiphae (JupiterVIID, 708,710 PlanetographiccoordiDares,37, 385-388 geocentc, 10, 126,202-207
apparent,391 statevector,353 Planets heliocentric,1I
Sun,398 Perihelion,4TT aberration,150 kecession, 105,150
Numericaldifferentiation, 560 562 Perturbations, 28 apparentpositions, 145-l 52 differential,108
Numericalintegration, 2'l, 286.562-566 Phase,396 astrometricpositions, 166, 308 drifL,293
N u t a t i o n1 ,2 , 9 91, 0 9 l, l l , l 1 6 - 1 2 1 , 1 5 1 angle,388,396 brightness general,99
conections, 116 119 coefficient,388 Pluto,406 geodesic, 100
differential,120 Calileansateilites, 408-409 Satum's rings, 404 tAU t9'16,t',73
IAU theory,l ll-113 Mercury,402 brilliancy,479 luni solar,12,99
in longitude andobliquity, I l2-115 Venus,403 configurations,476-482 Newcomb's, 106,173
planetaryterms,I 17-119 corrections, 291 292 deflection of light, 149 planetary13,99
fractionof Moonilluminated,399 elements, 315-316,704 reduction,daynumberformulas.108
Moon,399,478 elongations,479 rotationmatrix,103
Primemeridian,384,386,393
o Phases andmagnitudes,
intrinsicpropefties,
388-397
388
geocentricphenomena,4'7848 1
heliocentricphenomena,48 1-482 Mars,403
Oberon(UranusIV), 709,711 Phobos (MarsI), 385,708,710 masses,697 Mercury,402
oblateness,l4 orbitalelements, 343 orbitalelements,315-316,704
Pluto,406
orbital elements,369 physicalandphotometric data,385,407 physical and photometric data.401 407, Sun,398
physical and photometric data,412 Phoebe (SarumIX), 708,710 Primevenical,204
706
Obiiquity of the eclipttc, 12, 1'71,289 differentialcoordinates, 368 Prometheus (SaturnXVD, 410,709,7l I
precession,150
Observationalcatalog, 179 orbitalelements, 367 Proper motion,22, 121,111.1'71,184
rings,325,712
Occultations,34, 494--498 physical andphotometric data,410 Proteus (Neptune VIID.709,711
rotation elements,705
Occultations,lunar, 494 Photometric standards, stars,5 I l, 5 l6 Puck(Uranus XV), 412.709,1 11
satellites, 325,407J1 4,'708-11|
elements(Besselian),494 UvBRI,511,51'7 P u l s a r5s1, 1 5
, 21
semidiameters,401 407
planets,498 ,r1rl'andHB,5l l, 517 stationary,479
Oceanloading, 244-248 Physical ephemerides, 36,383-419 synodic periods,480-481
Oceantide model, 240-244
Openstarclusters,51 |, 522
a s t e r o i d s , 441147
Moon,398-401
visibility,480
Plate motions, 249-250
a
Lund Catalog,51l, 522 planets,401 407 Quartz-crystaloscillators,60
Pluto
Trumpler classificationsystem,524 Jupiter,/t03404 Quasars, 511, 530
mass,697
Ophelia(UranusVII), 4 12,7 09,'j I I Mars,403 Quasi stellarobjects,530
orbitalelements,316, 704
Opposition,33 Mercury,401 402 activegalaclicnuclei,531
physical and photometric data,406, 706
Orbitalelements,27 Neptune,405 /t06 catalogs,53G-531
physical ephemeris,406
Pluto,406 Radar ranging, 294 297
rotation elements,406, 705 Radialvelocity,22, 121.51 l, 518
Satum,404-405
satellite,3'7'7 , 414,'709,'7 | I Radio interferomctry,62
P Uranus,405
Pointmassinteractions, 281
Venus,402-'403 Rarlionretric methodol diametcrestimation,
Pallas,697 Polaris,498-502 414- 41'7
references,4l7--'ll9
Pandora(SarurnXVII ), 4lO,'109,j 1l P o l a rm o t i o n ,1 4 ,1 8 , 1 0 9 ,1 1 1 , 1 3 9 1 4 0 , geometric albedo,415
satellites,407 414
P a r a l l a x2,l , l 2 l , 1 2 3 238 240,269 272 intiared magnitudes,,l l5
Galilean,408 409
Parallax,annual,22, I25 Polynomials,548 phaseintegral,416
Jupiter,2t08409
diurnal(geocenrric), Portableclocks, 65 Radio sources
22, 123 124,l60 Mars,40? .tr08
equatorialhorizontal, Moon, 400 Portia(UranusXII), 412,709,'711 astrometriccalibrators, 5l I, 531
N e p t u n 4e 1, 34 1 4
horizontal,Moon,400 Positions flux calibrators,5l l, 532
Pluto,414
horizonlal,Sun,398 angle,392,395 Radius of the planets,387
S a t u m , 4 0 8 , 441101
in altitude,125 bright limb, Moon, 399 apparentpotar, 390, 396
U r a n u s4,1 14 1 3
solar, 22 greatesldefect of illumination, 397 apparentsubsolar,396
Sun,397-398
topocentric, Moon,400 Planetocent c coordinares, apparenr, 23, 145-160,165 Ref-erence frame, inertial, 96
3l, 37,385 3{i8
ParameterizedposFNewtonianequation,281 Planetocenlric orbitallongitude.Sun,401 astrometric, 23, 166,308, 3l l-313 Referenceframes, 223, 280
 750 INDEX 751
Referencesystems,95-99' 199 Seasons, 477,479 St^rs(continued) Tidal friction,l9
Refaactior, 24, 140-145 Second,SI, 40, 84 iadialvelocitystandards, 5l l, 518 Time(seespecificnames)
risingandsettingPhenomena, 488 Selenocentric coordinates, 1I secularvariations, l2l equationof, 5, 55, 74,485
Saastamoinen, 143 Selenographic coordinates,399 spectralclassifi
cation,5 19 signals,65
Relativity, 4O,10, 96, | 49, 295 Semidiameter MK spectralstandards, 511,520 transfer,64
Rhea(SatumV),410,708,710 Moon,401 Morgan-Keenan system(MK), 520 zones,56-57
orbital elements,360 Sun,398 stellarpositions,184 Titan(SatumVI)
physicalandphotometric data,410 shatosphere, l4l orbitalelements, 360,708
Siderealday,48
Ringsof Satum,brilliance, 404 variable,5lI physicalandphotomerric data,410,710
Siderealhourangle,1l
Rising atrdsettingphenomena,482-49( Stationary points,479 Titania(UranusIII)
Siderealtime,3-4, 39,48,55
Moon,485-486 Statistics,566-568 orbitalelements, 369,709
Sun,,183-485 aPparent,4, 48
Stellardata,509 physicaland photometic data,412,'l I l
Rosalind(uftnus XIII), 412,7O9,7 | I local,4,49
Stellarpositions,184 topopause, 14l
Rotationalelements, 383-388 mean,4, 48 point,391-395
Sub-Earth troposphere, 295
deformation,238-240 relationto universaltime,50-53 Subsolar point,392-395 Topocentric coordinates, 10,160-165
direct,385 Sinope(JupiterIX), 708,710 Subtabulation, 546 Topocentric parallax,Moon,400
retrogade, 385 Shadowlengths,489 Sun,700 Transformations to FK5, 167-187
Sun,397 Statevector 353 eclipses(JeeEclipses,solar) matrixmethod,180
Rotationof axes,552-553 Solarcorona,295 ephemeris, basisof, 317 TransitacrossSun'sdisk,471
Rotationof the Earth,251 Solareclipse(seeEclipses,solar) low precisionephemeris, 484 planets,47l
Rubidiumclocks,6l Solarparallax,398 mass,697 Transitcircledata,290 294
SolidEanhtides,233-238 physicalephemeris, 397-398 Transit,ephemeris, 309
Solstices,timesof, 477 riseandset,483-485 Transits,34
S Spacecraftdata,297-299 rotationelements, 705 Trigonometric functions,544
Spacemotion,121 Sundials, 73 Triton(NeptuneI)
Saros,35 SunriseandSunset,32,483485
Space-Timecoordinates,45 orbitalelements, 373
SatelliteAlnm@c,325 Surfacebrightness, 390
Spectralclassification,519 physicalandphotometric data,413
Satellites Surfacebrightness, planets,390
MK spectralstandards, 5l l-520 troposphere, l4l
apparent distanceandpositionangle,341 Synodicperiods,48G481
Morgan-Keenan classificationsystem Tropicalyear,length,80
apparcnrorbits, 336 SystemIII longitudes, Jupiter,403
(MK), s20 Twilight,33,482,492
authorities,3T8
astronomical, 33,483,492
differcntial coordinates,34I Sphe cal astronomy, 543,549-552
civil,33,483,492
elongations,34l Sphericalcoordinates, 9, 553
invariableplane,327 Sphericaltriangles,542 T na]utical,33, 483,492
Laplacianplane,327 Sph€ricaltrigonometry, 554 TAI (IntemationalAtomicTime),2, 39,84
orbital data,708 Starclust€rs Telesto(SatumXIII), 410
planetocentriccoordinates,333 globular,5l l, 526
rectaDgularcoordinates,341
Televisiontransmissions,67 U
oDF-n, 5ll,522 Terminator,geometric,397
Satum UByRI photometric 51l, 5 17
standards,
Starlight,493 Terestrialcoordinates,
13,20, 126
mass,697 DynamicalTime(TDT).2.42. 54
Terrestrial Umbriel(UranusII)
Stars
orbiralelemenrs, 316,704 Terrestrial system,199,200,
reference orbitalelements,
369,709
physicalandphorometric apparent places,23, 152-160,510
data,404-405, 2t"t-2t8 physicalandphotometicdata,4I2, 7 | I
'706 bright,5ll-512
TenestrialTime(TT), 3, 45 27
Unit, astronomical,
physicalephemeris,404-405 double,515-516
Tethys(SatumIII) Units
rin$s,354,404,1 12 foreshortening, 122 IntemationalSystem,718-719
orbitalelements,
356,708
rotationelements,404, 705 meanplaces, 511 physicalandphotometdcdata,410,710 length,716
salellires,
356,408,410-41t,708-71 parallax, 121,123,398 Thalassa(NeptuneIV), 7O9,7 | I mass, 716
semidiameter,4(X photometric standards,5l I, 516 Thebe(JupiterxIV), 708, 7 I 0 speed,716
Schwiderskimodel,242-243 propermotion, l2l 351
orbitalelemenis, UniversalPolarStereographic,2 lO-2 | 4
 752
UniversalTime(UT), 3, 39, 50-54,76-77 X
rclationto siderealtime,51-52
X-raysources,511,533
ndal etrects,250-254
(UTC),6, 53,76, catalogs,
533
UniversalTimeCoordinated
85-a6
UniversalTransverse Mercator,2lU2I4
Uranus Y
mass,697 Year,576,698
orbital elements,3 I 6, 368, 704 lengths,576,698
physicalandphotometricdata,706 lunar,57'7
physicalephemeris,405 sidercal,698
dngs,368,712 ffopical,80,576
rotationelements, 405,705
satellites,411-413,709,711
semidiametet405 Z
zvry andHB photometric standards,
511,517
Zodiacallight, 493
ZaMl tlde tems, 252--253

V
Variablestars,511,513
AmericanAssociationof VariableStar
(AAVSO),5 l3-5 14
Observers
GeneralCatalog ofUariable Stars(GCVS),
513
NameListsofUariableStars,514
Venus
billiancy,4O2, 479
magnitude,402
mass,697
orbital elements,3 16, 704
physicalandphotometic data,402403,
,7M
physicalephemeris,402-403
rotation elements,402, 705
semidiameter,402
transitof,471
Very Long BaselineInterferometry(VLBI),
i
69,255J56
Vesta,697
Vinualplace,165

w
We€k,577,603
wwv, 86