See

discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/236623946

Progress in understanding El Nio

ARTICLE in COMPLEXITY JANUARY 1987
Impact Factor: 1.03 DOI: 10.1016/0160-9327(87)90284-5

CITATIONS

DOWNLOADS

VIEWS

10

81

50

1 AUTHOR:
David B Enfield
University of Miami
72 PUBLICATIONS 2,855 CITATIONS
SEE PROFILE

Available from: David B Enfield

Retrieved on: 17 August 2015

Progress
in understanding
El Nino
David B. Enfield
Prior to the work of JacobBjerknes the El Nino phenomenonwas regardedas an aperiodicclimatic event
confinedto the Pacificcoastof South America. Spurredby a growing consciousness
of the oceans'role in
global climate, there has been an explosionof El Nino researchin the last two decades,El Nino is now
recognized to be an integral part of a Pacific-wide ocean relaxation, with global climatic impacts and
economically important ecological consequences.However, we are still groping for the final prize: the
ultimate causeof this climate anomaly and the ability to reliably predict its onset and intensity.

In the latter half of this century scientists

have gained much greater insight into
the true nature of El Nino. It is a regional
but important manifestation of i very
large-scaleinteractive process involving
the tropical ocean and global atmosphere. With our increased knowledge
we have come to realize that El Nino is
not merely a curious but isolated phenomenon, but rather an important link in
the physical processes that affect our
global climate from one year to another.
By understanding the role of El Nino in
these processes we can hopefully gain
the ability to forecast short-term climatic
changes, in particular, and appreciate
the role of ocean-atmosphere interaction in climate changeson all time scales,
in general.

early research [2] has shown that the

surfaceatmosphericpressurein regions
dominated by tropical convection (ascending air) and rainfall, such as Indonesia, is inversely correlated with the pressure in regions typified by subsidence
(descending air) and dry conditions,
such as the eastern South Pacific (figure
1). Air is continually transferred at low
levels - through the zonal trade wind
circulations - from the subsidence regions to the convective regions. The air
returns at upper tropospheric levels,
completing a seriesof zonal cells around
the globe that compromise the Walker
Circulation (figure 2). The SO has been
defined as a fluctuation of the mass
exchange in the dominant Indo-Pacific
Walker cell; that is, between the eastern
(Indonesian and Indian subcontinent)

The Southern Oscillation

Research durins the first half of this
century focusedhainly on the meteorological aspectsof the ocean-atmosphere
system, embodied in the global-scale
atmospheric pressure fluctuation identified by Sir Gilbert Walker as the
Southern Oscillation (SO) t1l. The SO
was recognized as a coherent variation of
barometric pressures at interannual intervals that is related to weather anomalies in many different regions, particularly in the tropics and subtropics. The

lf,:.;rJrr*r

(southeast
Pacific)hemisThe slateofthe SO pressuresee-sawis
characterized by the Southern Oscillation Index, or SOI (figure 3). The SO
reachesits maximum development (high
index) when the Pacific cell of the
Walker Circulation (including the trade
winds) is strong. At these times the

pressures in the Indonesian region are

lowest and the convection and rainfall
there reach maximum intensity; then
also, the South Pacific subtropical high
pressure region is most intense and a dry
zone extending along the equator from
South America toward the dateline is
best developed. The SO enters a low
index phase when the Walker Circulation weakens:at such times the barometric oressures rise in the Indonesian
region and fall in the southeast Pacific.
There are periods when low pressure
regions (such as Indonesia and northeast
Brazil) undergo drought conditions. At
the same time, the equatorial dry zone in
the Pacific contracts eastward and the
Amazon convective regime appears to
shift toward the desert region in the
northwestern portion of the South American continent. The El Nino (east Paci{ic warming) typically occurs following
a prolonged period of high index, just as
the SO is entering a low index phase.

ElNifio
Apart from a few scientific analyses
based on oceanographic expeditions or

DavidB.Enfield,
A. 8.,M. S..Ph.D.
Graduated
in physics
fromthe
andgeophysics
(Berkeley)
in 1965and
University
of California
gainedthe M.S.and Ph.D.degreein physical
oceanography
at OregonStateUniversityin
1970and 1973.He workedtwo yearseachas
UNESCO
Exoert
in Ecuador
andasLiaison
Office
for the IDOECoastalUpwellingEcosystems
Analysis
Program
in Peru,1973-77.
From1977
to 1987heworkedasResearch
Associate,
Assistant Professorand AssociateProfessorat the
Collegeof Oceanography,
SCPS
OregonStateUniversity.Hisresearch
interests
includeEl Nino,
150"
180"
90"
120"
30"w o"
90p
60"
lgo"
150" 120'
t50"
largescaleocean-atmosphere
interactions,
-30"E 60"
and
coastal
andequatorial
dynamics.
Since1987he
hasworkedas Research
Oceanographer
at the
FigureI The globaldistributionof the correlationcoefficientbetweenthe barometric
NOAAAtlantic
Oceanographic
andMeterolog
ic- press.ure
variationsat Djakarta,Indonesia,and thoseelsewhere.The negative
al Laboratory
in Miami,Florida.
with centresover lndonesia
correlationbetweenthe easternand westernhemispheres,
the pressuresee-sawassociated
and the southeastsubtropicalPacific,characterizes
Endeavour,New Series,Volume 11, No,4,1987,
wrth the SO,an interannualfluctuation
in the strengthof the WalkerCirculation.(AfterH.
0163-9327/87
$0.00+. 50.
P.Ber lag e[ 2 ] ]
O 1987PergamonJournalsLtd. Printedin GreatBritain.

191

Zonot

( " Wo L k r " )

S Am er c o

circutotion

Arr Lo

To h i ti - D o r w i n
S{n o o th e d su r l o cc p r e ssu r e o n o m o ti e 6
( 1 2 m o n th r u n n i n g m e o n )

otong equotor

Indone:ro

S.O." 1.59mb
.A

-z

30

\-------'
ATw
EN

EN

EN

EN

EN

EN

t960
-2

teor
90' w

o"

W,
oe ,
ATw,

9C

tO.E

9 0 .W

Heot b u d g e t o f o n o t m o s p h e r i c c o l u m n ,
Plonet o r y o t b e d o ( K . u e 9 e r ,
l97O)
W ot er t e m p e r o t u r e o n o m o t y
lOietrich ond Kotte.1957l

F igu re2 Tran sver s e

v iew ( iook ingnor t hac r os st he equ a t o r )o, f
t he WalkerCircula t iont hat gir dlest he g lobein t he t r opi c s .
Individ ua lWalker
cellsoc c urbet weenr egionsof uplif ta n d
trooicalconvection.on the westernsidesof oceanswherewater
t empe ratu resare hig h,and r egionsof s ubs idenc eov er t h ec o l d e r
easternoceans.The convectionregionsshow a net heatgainfrom
late nthe atre lea seand hav ehigh albedoduet o t he r ef le c t i v e
cloudcover;the same propertieshaveoppositetendenciesover
the cool regions.(AfterH, Flohn,Atmosphere,13,96,19751.

regional data from South America [3, 4]

knowledge of El Nino prior to the early
20th century consisted largely of information from milita r y c am paigns m
. is s ionaries,privateers(ship logs), explorers,
historical compilations, and geographic
reports, as well as from the economically
oriented activities of construction ensinee rs. h yd rolo gist s .f ar m er s .guano administration officials. and so on. The
limited climatic information obtained
from these sources. dating back nearly
five centuries, has been extensivelyresearched and summarized bv W. H.
Quin n [5]. Thc El N ino. whic h we r ec ognize today as an anomalous east Pacific
warming episode, originally received its
name in reference to the Christ child.
because of the annual appearance of
warm water near the Ecuador-Peru border at about Christmas-time. The accompanying changes in coastal marine
fauna and seasonal onset of light rains
over this normally dry desertregion have
always oriented the lishing and agricultural activities of the local inhabitants.
who take maximum advantagesof them.
H owe ve r. most o[ lhes e s am e ac t iv it ies
are adversely affected when the occasional atypical E,l Nino occurs every few
years, greatly enhancing the normal annual changes. The more spectacular
caseshave prompted often colourful accounts of desert floods. infestations of
insects, disease outbreaks, exuberant
blooms of desert plants (figure 4) and
southward invasions of unusual trooical
marine fa un a. No waday s .people as s ociatethe term'El Nino'with the extraor-

198

F i g u r e3 A 4 5 - y e a r t i m es e r i e so f a n o m a l i e so f t h e S O Id e r i ve d
from differencesbetweenthe barometricpressuresatTahiti
( e a s t e r nS o u t hP a c i f i cs u b s i d e n c er e g i o n )a n d D a r w i n ,A u str a l i a
( l n d o n e s i a nc o n v e c t i v er e g i o n ) P
. o s i t i v e( n e g a t i v es) w i n g so f th e
S O l ,s h a d e dw h e n l a r g e ,c o i n c i d ew i t h c o o l( w a r m ) o c e an i c
conditionsin the easternTropicalPacific.Occurrences
of El N ino
a r ed e s i g n a t e da s 'E N '.

dinary but less frequent occurrence

rather than the annual event.
The early atmosphericworkers recognized El Nino as one of the important
climatic aberrations that occur durins
t he low- i n d e xp h a s eo f t h e S O . T h e y d i d
not, however, understand the causeeffect relationships between the SO and
El Nino; that is, that El Nino is the

regional manifestation of a large-scale

oceaniccounterDartto the SO. and that
the two are intimately related through
large-scale ocean-atmosphereinteractions.
TheBjerknesrevolution
The Iirst maior advance in our understanding of El Nino and its relationship

..@;;

Figur e4 D u r i n g t h e 1 9 8 2 - 3 3 E l N i n o , r e c o r d r a i n s f e l l o v e r c o a s t a l E c u a d o r a n d th e
northernmostdesertregionof Peru.Thisscenedepictsa portionof the Sechuradesertin
Nov em b e r1 9 8 3 a
, b o u tf i v e m o n t h sa f t e rt h e r a i n ss u b s i d e da n d a s h a l l o wl a k eh a dd r i e d
u p, par c hi n g t h e s o i l .M an y v e s t i g e sr e m ai n o f t h e c o p i o u sv e g e t a t i o nt h a t
s pont an e o u s l y e m e r g e d f r o m t hdee l u g e . ( P h o t ob y H . S o l i d i . l

Eq.

Along equotor

Evolution of modern concepts

In the mid-1970s, as more data became
Iavailable and was analyzed, it became
--.-1=l-7-T*-on
o''
obvious that the coastal winds along
lmrnrs gPoc(r convl ,
i
most of the Peru coast do not decrease
Geoslr conv
___:1t--lduring El Niflo episodes (figure 6), als,/l\.N
t..
though the large-scale trades farther
offshore do indeed weaken [8]. At some
Eq.
c o a s t a l l o c a t i o n s t h e u p w el l i n g C otm
favourable winds actually intensify.
Clearly, the physical process of upwelling is not diminished and cannot explain
the well documented warming and decreased oroductivitv that occur in the
Peru coistal wateis durine El Nino.
Could it be that the larger scaleanomalies in the equatorial Pacific, described
n
F igure 5 Le ft,lo okingnor t hac r os s t heequat or int heP a c i f i c : S c h e m a t i c r e p r e s e n t a t i oand
explained by Bjerknes, were in
of the eq ua toria lflowp at t er nsunderc ondit ionsof nor m a l( p o s i t i v eS O I )a n d a b n o r m a l
some way transmitted to the coastal
(negativeSOI)trad ewinds .Right ,look ingwes t war dalon gt h e e q u a t o r :S c h e m a t i cv i e w
region?
of the normaland abnormaltransversecirculationpatternsnearthe equator.At top,
Beginning in 1975, a flurry of observapolewardinto either
westwardwind stressforcesa surface'Ekman'transport
tional and theoretical studies provided
hemisphere;sealeveland temperaturesare low at the equatordue to upwellingand a
compelling evidence that E,l Nino is,
convergenceat depthcom pensatesfor the Ekmandivergenceat the surface.At bottom,
indeed, remotely forced in the equatoEkmandivergenceand upwellingceaseasthe winds relax,the convergencecontinues
t h e t h e r m o c l i n e . rial Pacific [8, 9, 10]. Prior to El Nino the
t hro ug ho utthe u pp eroc ean,howev er ,r ais ings ealev ela n d d e p r e s s i n g
(From J. BferknesL6l,reprinted courtesYof the lnter-American Tropical Tuna
SO is in its high-index phase, with a well
Commission.\
developed southeast Pacific high pressure system and strong southeast trade
winds. Many oceanographers believe
this is an important precondition to El
Nino, because the strong trade circulaassumption of steady conditions, have tion amassesa large pool of warm, upper
to the Southern Oscillation came with
the wo rk o fJa co b Bj er k nes( l8S7- 1975) essentially captured the character of the layer ocean water in the western tropical
transient ocean responsewe have obser- Pacific. Then, as the SO enters its lowin the 1960s.The earlier SO researchhad
ved extensively since his time (figure 5): index phase, the normally westward blos eek ing
been p he no men ol ogic al,
a deceleration of the normally westward wing winds in the western and westcompound periodicities that coincided
with those of external agents such as equatorial surfacecurrents accompanied central eouatoridl Pacific weaken or reby cessation of equatorial upwelling; verse, allowing the accumulated warm
sunspot cycles. Bjerknes, on the other
accumulation of warm upper layer water water to move eastward. However, behand, sought to explain the SO through
in the equatorial zone; and consequent causeof the impulsive nature of the wind
the internal dyna'mical mechanisms of
large-scalewarming.
collapse. the ocean response is wavethe atmosphere and ocean, and he acThere are two principal shortcomings like, although physically very similar to
cepted the importance of their mutual
in Bjerknes' contributions. As already the mechanism proposed by Bjerknes.
interactions.
noted, his analysis assumes a steady
Initially, Bjerknes tried to explain El
ocean and coltrasts two quite different
Nirio as an occasional exaggeration of
but essentially equilibrium states: El
the El Nino-like state that occurs during
Nino and non-El Nino. This approach
every southern hemisphere summer,
T umbes
precludes recognition of the wave-like
centered about March. As had others
Tc lo ro
nature of the transient El Nino response
before him, he argued that El Nino is
Poi
{'\ 'Prto
ur o
that we are now familiar with. The
caused by local trade wind weakening
and north-south ocean density gradients oversight was probably due in large part
C hi c i oy o
\xt\\
to the second difficulty that he faced: a
within the El Nino region itself. As the
\\"
\
\ L0005\
lack of data. Because he had few obserTru j i U.o
normal equatorward coastal winds devations of Pacific-wide wind distribucrease,cease,or reverse, the ubiquitous
Ch i m ooie
tions, he assumed that the trade winds
upwelling of cold, nutrient-rich waters
weaken everywhere and that the El Nino
alons the Peru coast would be interrupc ondit ion c a n b e u n d e r s t o o d e v e ted; i=hewater would warm from the lack
rywhere, including the region off South
W i nd v ec tor s
warm,
because
also
and
of a cool source
R/V Wecomo
r\rit\V'
America, as an adjustment to locally
low-salinity water of low density north of
883
4 Feb:22Feb.,
c oi
r \ \'
led
This
assumption
reduced
winds.
the
in
o-_10
southward
flow
the equator would
m/sec
investigators
at
the
I
I
absenceof opposingwinds. Both mecha- Bjerknes and other
time [7] to erroneously conclude that the
nisms would explain the collapse of the
F i g u r e6 D i s t r i b u t i o no f w i n d s w i th i n
Peruvian El Nino is a locally produced
coastal ecosystem due to deficits in the
warming due to the cessation of coastal severalhundredkilometresof the Peru
nutrient supply.
upwelling under reduced (or reversed) coastobservedduringthe February1983
He quickly realized, however, that the
cruiseof the R/V Wecoma.Thecoastal
alongshorewinds, simultaneous with the
regional El Nino is closely related to a
w i n d s r e m a i nn o r m a la n d u p w e l l i n g
o
c
c
u
r
r
i
n
g
phy
s
ic
ally
r
e
s
p
o
n
s
e
similar
Pacific-wide relaxation of the equatorial
favourablesouthof6'5, in spiteofsevere
much farther west, along the equator.
ocean in response to weakened trade
E l N i n oc o n d i t i o n sp r e v a l e nat t t h e ti m e .
As we shall see, the Bjerknes scenario North of therewinds areweak and
winds on a much larger scale [6]. This
would need modification to adequately variableand coincidewith convective
was a major, new contribution that has
activityand rainfallovertheadjacent
been confirmed by subsequentresearch. explain what happens off South America
SechuraDesert.
during El Nino.
His arguments, though based on the
'

W i ( h eos t w i nd

\attt-\

)o',*j'i:

199

F igu re7 Lo okingd own on t he equat or f r om abov e,a n u m e r i c a l s i m u l a t i oonf a n e q u a t o r i a l l y t r a p p eidn t e r n a K

l e l v i nw a v e
that impingeson the SouthAmericancoast.Contoursapproximatelyrepresentthe anomalousrisein sealevel (centimetres)
and
depressionof the thermocline(metres).The
three panels depictconditionsat successivemonthly intervals.(FromD. B. Enfield I11I
after O'Brienetal.)

Because the wave response involves

an eastward perturbation of existing
currents. the earth rotation deflects the
flow just off the equator - in either
hemisphere - towards the equator. The
unleashed wave, or series of waves, is
thus trapped on the equator and advances eastward much like an elonsated
bubble of liberated upper layer water
(figure 7). Called Kelvin waves (after
Lord Kelvin. who first described such
waves mathematically), these pulses
propagate acrossthe Pacific to the South
American coast in about two or three
months. They are internal waves, supported by the density contrast at the
equatorial thermocline, a zone found
50-150 metres below the sea surface

where the temperature decreasesdownward most rapidly. The behaviour of

these waves is not altogether unlike the
sloshingof an oil layer overlyingvinegar.
As they pass(say, for example, an island
tide gauge or an array of submerged
instrumentation). the sea level at the
equator typically rises by 10-30 centimetres, the thermocline is depressed
downward by a comparable number of
metres, and the normally westward
flowing equatorial currents are reversed.
As Kelvin waves reach the South
American coast, several things happen.
Consistent with their behaviour along
the equator, sea level rises and the
thermal structure alons the coast is depressed, but due to the confining effect

F e b- Mor

Nov- D e c

of the continental boundary, the amplitude of these disturbances is increased

considerably(figure 7, third panel). Immediately, part of the energy begins to
reflect westward as pairs of slowly propagating, counter-rotatingeddies, called
planetary or Rossby waves. More importantly for the coastal ecosystem,much of
the energy continues poleward into both
hemispheres,as the Kelvin waves split
i n t o d i v e r g i n g c o a s r a lw a v e s l a l so w i th
Kelvin-like characteristics).
The impact of the wave arrivals on the
ocean thermal structure is well illustrated by a two-year series of temperature
sections off Ecuador, spanning the
strongl9T2-7 3 El Nino (figure 8) [11]. In
November-December 1971 conditions

A ug-S ep

Moy- Jun

>\-/^'
\--l-'

'-'--^--/"- to---z-

t972
o)
0)
c
E
o
o)

50

roo
200
300
400

1973
2"S

o"

r.N
2.S
Loti tude

o'

l eN

F ig ure S Eig htsucc es s iv eoc eant em per at ur e(s' C)

ec t i o n s a c r o s s t h e e q u a t o r o f f E c u a d o r a t t h r e e - m o n t h i n t e r v a l s b e f o r e ,d u r i n g ,
and afterthe 1972-73El Nino.The first (toplandsecond(bottom)yearsof the sequencebeginin November-December
(left)and end in
(right).(FromD. B. Enfield[11])
August-September

200

entire coast may be affected, much sooner than could occur under the action of
currents alone. Although the coastal
flow is acceleratedsouthward. water that
starts at the equator can travel only a
short distance in the time it takes the
waves to propagate southward. However, since the coastal upwelling of subsurface water from depths of 50*100
metres continues unabated. and because
the water at those depths is much warmer than normal, the surface temperatures of the upwelled water become
unusually high. Coastal surface temperature anomalies of 3-4"C above normal
are common for moderate El Ninos and
they reached as much as 8-10'C during
the very intense 1982-83 episode. More
signilicantly for the coastal ecosystem,
the upwelled water is also poor in nutrients and subsequently leads to a collapse of the primary productivity and of
the heavily exploited commercial fish
stocks that it normally sustains(figure 9)

[r2).

Such is the conceptual model'of El

Nino that has evolved since the work of
Bjerknes, a paradigm widely accepted
by oceanographers as a framework for
modern research. The rapid developments in El Nino research over the last
decade probably would not have occurred, however, ifthe effectsofthis phenomenon were geographically conlined to
the equatorial Pacific. Instead, we owe
much of our progress to the realization
by scientists,newsmedia, and legislators
that E,l Nino has far-reaching climatic
and economic repercussionsaround the
globe.
Teleconnections
It was B jerkneswho, usingdata from the
unusual 1957-58 event, advanced a major hypothesis to explain the geographic
extensionsof El Nino [13]. He argued
that the equatorial warming feeds back
on the lower atmosphere (troposphere)
through evaporative heat transport from
the warmed ocean surface. Thunderstorm activity then becomes frequent
and unusual amounts of rainfall assault
the normally dry equatorialzone. As the
F igure 9 Sche maticillus t r at ionof
upwellingc har ac t er ist i c s a l o n g t h e c o a sPt oefr u
atmospheric moisture condenses, unu. her m o c l i n ea n d n u t r i c l i n e
during n orma l(top )a nd El Nino ( bot t om ) c ondit ionsAt
sual amountsof latent heat are released.
separatethe wa rm, nu trient - def ic ient
upperlay er( lightc olo u r ) f r o mt h e c o o l ,e n r i c h e d
propelling additional heat and momenlower layer below (darkcolou rl. lArtwork by Allen Carroll basedon original drawing by
R. R. Barber@National GeographicSocietyl2l\
tum north or south toward higher latitudes at the upper levels of the troposphere. The jet streamsintensify, altering
were normal with a shallow tropical
time, most of the isothermsat depthsless significantly the extratropical atmosphethan 400 metres were 50-150 metres ric circulation and weather Datterns.
thermocline at 30'-50 metres and relatively isothermal water below. The therdeeoer than normal. and maximum tem- e s p e c i a l l y i n t h e w i n t e r h e m i s p he r e .
mocline deoression due to the first Keloeritu.e increasesof about 9-10'C were Bjerknes adapted the term 'teleconnect i o n s ' t o c h a r a c t e r i z e t h e l i n k a ge s
iound 50 metres below the surface. By
vin w ave arrival wa s det ec t ed in
March 1973 the system was returning to between the equatorial source region
February-March 1972. Taking the prenormal, and by the end of that year and these remote. but related weather
E l N ino con ditio n of Nov em ber appeared to have completely recovered. phenomena.
December l97l as a benchmark for the
Ironically, the first strong evidence for
As the southern hemisphere portion
'normal' temperature distribution, the
of the El Nino disturbance propagates teleconnections came from the ocean
maximum depression occurred one year
later when a second rise in coastal sea southward, it quickly spreadsthe deepe- and not the atmosphere. During E,l
ned thermocline and sea level rise along N i n o s . u n u s u a l i n c r e a s e si n m e a n se a
level signalled the probable arrival of
the Peru coast. In a matter of weeks the level occur simultaneouslv. not onlv off
another series of Kelvin pulses. At that

201

South America but also along most of

the Pacific coast of Central and North
America as far north as Canada and
Alaska (figure 10) [1a]. The data are
consistent with the continued poleward
propagation of the El Nino wave along

the coast. As sea level rises off Calilbrnia, the normal southward drift of the
California Current is decreasedor reversed within about 500 km of the coast-and
marine organisms tend to be found north
of their usual habitats [15, 16].

SEA L EVEL
5U

20

33

bU

b3

/u

Y AK U T A T

0
-z u
RU PE R T

MAZATLAN

A CA P U L C O

BA L BOA
B UE N AV EN T U R A

GALAPAG O S
T ALARA

C ALLAO
I.4 ATARANI

ANTOFAG ASTA

VAL PARAI SO

5i_*--_*_-5dFig ure l0 Time s er ies of s ea- lev elanom alies f or a 2 5 - y e a r p e r i o d a t l 9 t i d e g a u g e

stationsfrom Valparaiso,Chileto Anchorage,Alaska.Periodsof persistentpositive
a,and Callao( blac ks h a d i n g )c o r r e s p o n dt o E l N i n o
a no malie sa t Ga lapagosTalar
,
e pisod es.a nd arec oher ent wit hs im ilar f luc t uat ionast h i g h e rl a t i t u d e s(.F r o mE n f i e l d
a nd Allen[1 6])

202

Public and scientific awarenessof the

atmospheric teleconnections proposed
by Bjerknes was increased because of
severeweather abnormalities that occurred in North America during the 197677 winter. An exaggerated circulation
pattern, attributed to teleconnections
from a moderate El Nino that year [17],
brought a paralyzing drought to the US
west coast, while the east coast suffered
the numbing effects of repeated snow
storms and record cold. Extensive analysesof historical data setssoon revealed
that the interchange between ocean and
atmosphere during one of these events is
a continuous process occurring in both
directions, developing in phases seemingly locked to the march of the seasons
[18], and with major teleconnectionsto
the northern hemisphere occurring during the northern winter season(figure7).

The 1982-83 ElNifio

By 1982, a certain complacency had set
in amongst scientists, who felt that El
N i n o / S o u t h e r n O s c i l l a t i o n e o i so d e s
( E N S O s ) t e n d t o f o l l o w a r e p e a ta b l e
pattern and that further examination of
the teleconnection process could soon
lead to useful forecastsof maior climatic
a n o m a l i e s .H e n c e , w h e n t h e fi r st r e p o r ts
o[ anomaliesin the tropical Pacificu'ere
received in mid-1982, many scientists
refused to believe a true El Niflo was in
progress, based on the unusual time of
year and the lack of antecedent conditions thought necessary(for example., a
positive SO index). In spite of its nonconformity to the canonical pattern, the
1982-83 period subsequently proved to
b e a d i s a s l r o u sE l N i n o , m o r e i n l e n se
than any other in living memory. The
last episode of comparable magnitude
occurred in 1891, almost a century before. Although our basic concepts of
what happens during a ENSO episode
remained unchanged, our faith in the
feasibility of forecasting its occurrence
had been badly shaken.
There was extensive damase to the
m a r i n e e n v i r o n m e n t a l o n g th e So u th
American coast in 1983 [9]. Ironically,
not all of the fisheries impacts were
disastrous; certain commercial species,
most notably shrimps, but also dolphinfish, scallops, octopus, and many
others appeared in unusual numbers.
Others, of course, virtually disappeared
or their stocks were greatly depleted.
High temperatures and low concentrations of traditional plankton varieties
devastated many pelagic (mid-water)
fish stocks. such as anchovies and sardines. Many bottom dwellers like shrimp
and scallops fared well because of an
enriched oxygen supply, while other
demersal (near-bottom) species such as
hake were redistributed by the environmental changes. One can speculate, of
course, that the demise of some species
favoured the prosperity of others, direc-

tly or indirectly, through greatly altered

patterns of interdependent factors such
as available food types, larval survival,
predation, and so forth.
Mortality among the normally abundant guano birds and marine mammals
o f t h i s c o a s t a l r e g i o n w a s e x t e n si ve .
especially of their young, because their
normal food supply of schooling fishes
was greatly reduced and/or made unavailable to them. Since the hungry adults
abandoned their nests and broods in
large numbers in search of scarce food,
their populations were subsequently
slow to recover after environmental
conditions returned to normal in 1984.A
well documented microcosm of the bird
mortalities occurred near the eouator on
C h r i s t m a s I s l a n d . h a l l - - w a ya i r o ss th e
Pacific [20].
By far, the most extensive damage to
the South American coast came in the
form of rainfall. Unheard-of amounts of
rain fell on coastal Ecuador and Peru.
north of Lambayeque (6'5). Flas-hfloods
roared through the desert gullies, ripping out roads and bridges and damaging
the oleoducts that transport crude oil to
ports of embarkation. The important
coastal town of Talara, built in a canyon
cut through the Sechura plateau, was
inundated with mud that took more than
a year to remove. Talara and other
isolated communities had to be supplied
by helicopter until road links could be
re-established.Vast regions of the Sechura desert were revegetated by the
rainfall (figure 4), and large shallow
lakes appeared, often burying major
highways, some remaining for almost
two years. Agriculture, highly dependent on a well controlled system of
irrigation, was ravagedby untimely water damage, a lack of transportation
facilities and - not the least - bv infestations ofvoracious insectsspawndd by the
rogue vegetation. Because mean sea
level had risen by as much as two feet,
m a n y l o w - l y i n gb e a c hc o m m u n i t i e sw e r e
pounded by surf, streets flooded, and
beachfront buildings wiped out.
Many poignant laments of human suffering emerged from the 1982-83 El
Nino disaster[21]:

ff,>r!-;A-+\\
t-i-__
l)\
,,,,ftftu_o_

! '-'Y

,1\

---l

j.

--..a--l

tL

.'

-2,

J,,

rG,

30N
20N
to N

o
ros
20s
30s
tooE

r 20E

t40w

't20w

100w

80\Y

Figure11 An illustrationof the distributionsof key propertiesinvolvedin the

telec on ne ctio pro
n ce s sdur inga t y pic alENSOev ent ,( a)Up p e rt r o p o s p h e r ea n o m a l i e s
of geo po ten tialhe igh t f or t henor t her nhem is pher ewint e r( .b ) C o n t o u r s osf u r f a c e
barometricpressure,with surfacewind anomalies(arrows)in the tropics.Bottom:
D istrib utio nof an oma louss eas ur f ac et em per at ur e( "C) (. F r o mE . R a s m u s s o n[ 1 4 ] .
Reprinted courtesy of OceanusMagazine.l

In the center of the city of Sullana a

whole block became an island of land
on Alcedo street. Entire families were
imprisoned by the water which reached four metres. A merchant
committed suicide. A wall fell on a
young woman and caused her death.
The Navarro family prayed in front of
their religious statues. La Arena, a
country town, was totally inundated.
In the television interview with a
simple country woman: 'The water
took everything, our children are hungry. Oh dear Lord. I mean to say that
the rain is good but our lack of preparation to receive it is bad'.

203

..:

Figu rel2 Da mageduet oc oas t al

er os ionby highs u r fa t P a c i f i c a , n e a r S a n F r a n c i s c o ,
Californiain 1983.(Photoby James Sugar/BlackStar.l

The teleconnections of weather phePresent and future research

nomena to more remote regions began The 1982-83 ENSO was a humblins
in June-July 1982, when central Chile
ex per i e n c ef o r s c i e n t i s t s :a c o m p a r a b l a
suffered the effects of record rainfall and
El Nino may not occur for another 100
flooding resulting from an intensified
years,but this one left its mark. Fortunaso uth ern he misp her ewint er jet s t r eam . tely, the interest and funding generated
By December 1982the sea surfacetemfrom it has causedan explosion in ENSO
peratures along the equator had risen to
research around the world. as well as
4-6"C above normal and were pumping
widespread coverage in the popular
energy into the subtropical jet stream of
press. In 1985 a decade of ENSO rethe no rthe rn h em is pher e. Fr equent
search was inaugurated under the aegis
storms moved south of their normal
of an international effort called the Trotracks and battered the California coast pical Ocean-Global Atmosphere Prowith high surf. The coastal erosion was
gram, or simply, TOGA. By the end of
all the more catastrophic for coastal the 1990s,hundreds of meteotologists,
communities due to the fact that mean
oceanographers,biologists, and engisea level had risen about a foot (30
neers will have been involved in TOGA.
centimetres) above normal levels (figure
t r y ing t o u n r a v e l t h e m a n y r e m a i n i n g
12). The offshore circulation ofthe normysteriesof ENSO:
mally southward California Current was
1. What are the mechanismsby which
reversed, and hundreds of marine spethe
oceanic relaxation Droducesunusual
cies were carried far north of their
surface warming over iarge areas of the
normal ranges [22]. Some fisheries prostropical Pacific?
pered; others collapsed;of these, a few
have not yet recovered. And the list goes 2. Do the tropospheric teleconnections
on: record drought in Australia, South occur as Bjerknes explained, or does the
Pacific hurricanes as far east as Tahiti,
atmosphere transmit anomalous variabi.unusual winter rainfall in the southern
lity through wavelike or other means?
USA and northern Caribbean. and so
3. Does the coastalwave propagationof
on. Certainly, the global weather disturbanceswere remarkable durine 1982-83 oceanic disturbances to high latitudes
really occur, or do the atmospheric tele, [2 3j. con firming the bas ic hy pot hes isof
connections produce similar effects and
Bjerknes. However, the pattern of
merely make them appearwavelike (or,
weather phenomena produced by the
more likely, do both occur)?
teleconnectionsturned out to be radically different from that on previous occa- 4. What first causesthe trade winds and
sions. We currently fear that the details Walker Circulation to decelerate. initiaof the atmospheric response to El Nino 'ting an ENSO sequence,and can this
warmings are hypersensitive to initial
event be predicted so as to produce
conditions in ways that are difficult or
reliable and useful forecasts?
impossible to predict, and we need to
The TOGA research stratesv to
know more about the teleconnection answer these and many other qu"iiions
process.
is to foster investigations along three

204

broad and interactive fronts [24]: long

term (decadal) monitoring of key environmental variables; shorter term (several months to several years) field experiments designedto gain an understanding
of the key processesthat occur during
ENSO; and theoretical and numerical
modelipg of ENSO with an aim to ultimately predicting its occurrence and
time evolution.
What is the greatest obstacle to further progress?This author's worst fear is
that the attention span of administrators
and politicians will be too short to keep
the effort funded for the full decade.
Will another disaster be reouired to
ensure final success?
References
[1] Walker, G. T. Mem. Ind. Met. Dept.,
24,75,1923.
enVerhandel.,
[2] Berlage.H.P. Mededel.
Kon. Ned. Met. Inst..No. 69, 152pp.,
1957.
[3] Murphy, R. C. GeographicReview,16,
26.1926.
[4] Schott,G. Ann. Hydr. Marit. Meteorolog.,59,162,1931.
[5] Quinn,W. H., NealV. T. andAntunez
de Mayolo S. E. "/. Geophys. Res.
(Oceans),in press,1987.
[6] Bjerknes,J. Bull. Inter-Am.Trop. Tuna
C o m m . ,1 2 , 1 , 1 9 6 6 .
[7] Wooster,W. S. and Guillen O. J. Mar.
Res..32.287.1974.
5,572,
[8] Wyrtki. K. l. Phys.Oceanogr.,
1975.
[9] Godfrey,J. J. Phys.Oceanogr.,5,399,
1975.
[10] McCreary,J. P. "r. Phys. Oceanogr.,6,
632.1976.
[11] Enfield, D. B. ResourceMana7ement
and EnvironmentalIJncertainry.M. H.
GlantzandJ. D. Thompson,
eds.,Wiley
Interscience,Wiley and Sons, New
York,213-254,1981.
[12] Barber,R. T. and ChavezF. P. Science,
222, 1203,1983.
[13] Bjerknes,J. Tellus,lS, 820, 1966.
[14] Enfield, D. B. and Allen J. S. l. phys.
Oceanogr.,
l0 557,1980.
[15] Chelton,D.B. CaICOFIReports,Yot.
x x i l . 3 4 - 4 8 ,t 9 8 1 .
[16] Mysak,L. A., HsiehW. H. andparsons
T . R. Biolog.Oceanogr.,
2 (1),63,1982.
[17] Namias,J. Mon. Wea. Rev., 106,279,
1978.
E. M. and CarpenterT. H.
[18] Rasmusson,
M o n . We a .R e y . ,l l 0 . 3 5 4 ,19 8 2 .
[19] Arntz, W., LandaA. and TarazonaJ.
'El Nino': Su Impacto en La Fauna
Marina,Bulletin,Institutodel Mar, Callao, Peru, 222pp., 1985.
[20] Schreiber,R. W. and SchreiberE. A.
Science,225,713,1984
[21] Anon.Piura'33,Charityraisingpublication fundedby the CatholicArchdiocese
of Piura/Tumbes,
Peru, 63 pp., 1983.
[22] Wooster,W. S. andFluhartyD. L., eds.
El Nino Norlr, WashingtonSea Grant
Program,Univ. Wash., Seattle,WA
98795,312
pp., 1985.
[23] Canby,T.Y . NationalGeographic,165,
144,1984.
[24] Anon. Ocean Researchfor Understanding ClimaticVariations.NationalAcademyPress,Washington,
D. C.,58 pp.,
1983.