Cat 2015 ApJS 220 19

The Astrophysical Journal Supplement Series, 220:19 (18pp), 2015 September doi:10.
1088/0067-0049/220/1/19
© 2015. The American Astronomical Society. All rights reserved.
LAMOST OBSERVATIONS IN THE KEPLER FIELD. I. DATABASE OF LOW-RESOLUTION SPECTRA*

P. De Cat2, J. N. Fu1, A. B. Ren2, X. H. Yang2, J. R. Shi3, A. L. Luo3, M. Yang3, J. L. Wang3, H. T. Zhang3, H. M. Shi3,
W. Zhang3, Subo Dong4, G. Catanzaro5, C. J. Corbally6, A. Frasca5, R. O. Gray7, J. Molenda- Żakowicz8,9,
K. Uytterhoeven10,11, M. Briquet12,13, H. Bruntt14, S. Frandsen14, L. Kiss15, D. W. Kurtz16, M. Marconi17,
E. Niemczura8, R. H. Østensen18, V. Ripepi17, B. Smalley19, J. Southworth19, R. Szabó15, J. H. Telting20, C. Karoff14,21,
V. Silva Aguirre14, Y. Wu3, Y. H. Hou22, G. Jin23, and X. L. Zhou24
1
Department of Astronomy, Beijing Normal University, 19 Avenue Xinjiekouwai, Beijing 100875, China; [email protected]
2
Royal observatory of Belgium, Ringlaan 3, B-1180 Brussel, Belgium
3
Key Lab for Optical Astronomy, National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100012, China
4
Kavli Institute for Astronomy and Astrophysics, Peking University, Yi He Yuan Road 5, Hai Dian District, Beijing, 100871, China
5
INAF—Osservatorio Astrofisico di Catania, Via S. Sofia 78, I-95123 Catania, Italy
6
Vatican Observatory Research Group, Steward Observatory, Tucson, AZ 85721-0065, USA
7
Department of Physics and Astronomy, Appalachian State University, Boone, NC 28608, USA
8
Astronomical Institute of the University of Wrocław, ul. Kopernika 11, 51-622 Wrocław, Poland
9
New Mexico State University, Department of Astronomy, P.O. Box 30001, MSC 4500, Las Cruces, NM 88003, USA
10
Instituto de Astrofísica de Canarias (IAC), E-38200 La Laguna, Tenerife, Spain
11
Universidad de La Laguna, Dept. Astrofísica, E-38206 La Laguna, Tenerife, Spain
12
Institut d’Astrophysique et de Géophysique, Université de Liège, Allée du 6 Août 19C, B-4000 Liège, Belgium
13
LESIA, Observatoire de Paris, PSL Research University, CNRS, Sorbonne Universités, UPMC Univ. Paris 06, Univ. Paris Diderot, Sorbonne Paris Cité, France
14
Stellar Astrophysics Center, Department of Physics and Astronomy, Aarhus University, Ny Munkegade 120, DK-8000 Aarhus C, Denmark
15
Konkoly Observatory, Research Center for Astronomy and Earth Sciences, Hungarian Academy of Sciences,
Konkoly Thege Miklós út 15-17. H-1121 Budapest, Hungary
16
Jeremiah Horrocks Institute, University of Central Lancashire, Preston PR1 2HE, UK
17
INAF—Osservatorio Astronomico di Capodimonte, via Moiariello 16, I-80131 Napoli, Italy
18
Instituut voor Sterrenkunde, KU Leuven, Celestijnenlaan 200D, B-3001 Leuven, Belgium
19
Astrophysics Group, Keele University, Staffordshire, ST5 5BG, UK
20
Nordic Optical Telescope, Rambla José Ana Fernández Pérez 7, E-38711 Breña Baja, Spain
21
Department of Geoscience, Aarhus University, Høegh-Guldbergs Gade 2, DK-8000, Aarhus C, Denmark
22
Nanjing Institute of Astronomical Optics & Technology, National Astronomical Observatories,
Chinese Academy of Sciences, Nanjing 210042, China
23
University of Science and Technology of China, Hefei 230026, China
24
Key Laboratory of Space Astronomy and Technology, National Astronomical Observatories,
Chinese Academy of Sciences, Beijing 100012, China
Received 2015 June 3; accepted 2015 August 18; published 2015 September 23
ABSTRACT
The nearly continuous light curves with micromagnitude precision provided by the space mission Kepler are
revolutionizing our view of pulsating stars. They have revealed a vast sea of low-amplitude pulsation modes that
were undetectable from Earth. The long time base of Kepler light curves allows for the accurate determination of
the frequencies and amplitudes of pulsation modes needed for in-depth asteroseismic modeling. However, for an
asteroseismic study to be successful, the first estimates of stellar parameters need to be known and they cannot be
derived from the Kepler photometry itself. The Kepler Input Catalog provides values for the effective temperature,
surface gravity, and metallicity, but not always with sufficient accuracy. Moreover, information on the chemical
composition and rotation rate is lacking. We are collecting low-resolution spectra for objects in the Kepler field of
view with the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST, Xinglong observatory, China).
All of the requested fields have now been observed at least once. In this paper, we describe those observations and
provide a useful database for the whole astronomical community.
Key words: astronomical databases: miscellaneous – stars: fundamental parameters – stars: general –
stars: statistics
Supporting material: machine-readable tables
1. INTRODUCTION Kepler spacecraft failed, preventing the telescope from

The space mission Kepler was designed to detect Earth-like precisely pointing toward the same FOV. Even though
planets around solar-type stars using the transit method (Koch ultrahigh precision photometry can no longer be collected in
et al. 2010). It was launched on 2009 March 7 and started to the original Kepler FOV, the Kepler photometry that is now
collect ultraprecise photometry within a spectral bandpass from available for about 200,000 stars is a pure goldmine for
400 to 850 nm for a fixed field of view (FOV) of 105 square asteroseismic studies of all types of pulsating stars, as well as
degrees in the constellations Lyra and Cygnus on 2009 May 2. for many other science cases.
On 2013 May 11, the second of four reaction wheels on the The success of asteroseismic studies has been shown to
depend crucially on the availability of first estimates of basic
*
Based on observations collected using the Large Sky Area Multi-Object Fiber stellar parameters, such as the effective temperature (Teff ),
Spectroscopic Telescope (LAMOST) located at the Xinglong observatory, China. surface gravity (log g), metallicity ([M H ]), and projected
1
The Astrophysical Journal Supplement Series, 220:19 (18pp), 2015 September De Cat et al.
the fiber slit was permanently fixed to a ×2/3 slit width for all
of the LAMOST projects, corresponding to a resolution of about
1800 for the low-resolution mode. The active optics technique
is used to control the reflecting corrector (Su et al. 1998).
LAMOST has a quasi-meridian transit configuration capable of
tracking the motion of celestial objects for about 4 hr while
they are passing the meridian. For more detailed information
about LAMOST, see Cui et al. (2012) and Zhao et al. (2012).
LAMOST is an excellent instrument to perform spectroscopic
follow-up for targets of the Kepler mission because it is capable
of collecting low-resolution spectra for thousands of objects
down to magnitudes of 17.8 simultaneously. In 2010, we
initiated the “LAMOST-Kepler project” (LK-project) to observe
as many objects in the Kepler FOV as possible from the start of
the test phase of LAMOST onward. This allows for a
homogeneous determination of both the stellar parameters
and the spectral classification of the observed objects. More-
over, with low-resolution spectra, it is possible to estimate the
radial velocity (vrad ) and, in the case of rapid rotation, the
projected rotational velocity (v sin i ) of the observed objects.
This paper is the first of a series dealing with the detailed
description and analysis of the spectra obtained for the LK-
project in the Kepler FOV during the first round of
Figure 1. Configuration of the 4000 fiber positioning units covering the observations, up to the end of the 2014 observation season.
5° focal plane. The fibers are homogeneously distributed on the focal plane, The structure of this paper is as follows. In Section 2, we
except for the position of the active optics wave front sensor (central hole) and
four Guiding CCD cameras (off-center holes). explain how we compiled a prioritized target list that was used
as input for the LAMOST observations. In Sections 3 and 4, we
justify our choice of the central positions of the different
stellar equatorial rotation velocity (v sin i ; Michel 2006; Cunha LAMOST fields that we requested to be observed and how the
et al. 2007). These parameters cannot always be directly selection of the fibers on the fields was done, respectively. A
derived from the Kepler data, as in the case of data with multi- detailed description of the observations themselves is given in
color photometry or spectroscopy. Before the launch of Kepler, Section 5. Section 6 provides an overview of the most
there was a large effort to derive stellar parameters from Sloan important steps of the reduction procedure that resulted in the
photometry for potential Kepler targets. These are available in database of LAMOST spectra described in Section 7. The full
the Kepler Input Catalog (KIC; Brown et al. 2011). Unfortu- results will be published in the subsequent papers of this series
nately, KIC stellar parameters are not available for all of the (A. Frasca et al. 2015, in preparation; R. O. Gray et al. 2015, in
stars of interest and the precision of Teff and log g in the KIC is preparation; A. B. Ren et al. 2015, in preparation). This paper
generally too low for asteroseismic modeling, especially for hot concludes with a discussion of the importance of the acquired
and peculiar stars (Molenda-Żakowicz et al. 2010b; McNamara LAMOST spectra for the community involved in Kepler research
et al. 2012). Also, detailed information on the stellar chemical in Section 8.
composition and rotation rate is lacking. Hence, to exploit the
Kepler data best, additional ground-based spectroscopic data 2. SELECTION OF THE TARGETS
are required (Uytterhoeven et al. 2010a, 2010b).
LAMOST (also called the Guo Shou Jing Telescope) is a
We constructed a prioritized list of targets for the LK-project
unique astronomical instrument located at the Xinglong that was used to prepare the LAMOST observations. We made use
observatory (China). It combines a large aperture (3.6–4.9 m) of our knowledge of objects within the Kepler FOV prior to the
with a wide FOV (circular with diameters of 5°; Wang start of our LK-project, i.e., their position (R.A. a2000 , decl.
et al. 1996). The focal surface is covered with 4000 optical d2000 ), their brightness (the KIC magnitude K p for most of the
fibers connected to 16 multi-object optical spectrometers with objects; see Section 2.2), the availability of stellar parameters
250 optical fibers each (Xing et al. 1998; Figure 1). Each in the KIC (the effective temperature Teff , the surface gravity
spectrometer has two CCD cameras: one for the blue arm log g, and the metallicity [M H ]), and their scientific
(optimized for 370–590 nm) and one for the red arm (optimized importance within the community involved in the Kepler
for 570–900 nm). Their spectral resolution at full slit is either research (see Section 2.1).
1000 or 5000 depending on the use of the low- or medium-
resolution grating mode and on the camera positions. During 2.1. Type of Targets
the initial LAMOST observations, it was possibile to enhance the The first level of prioritization was based on the type of
resolution to 2000 or 10,000 by limiting the fiber slit to a 1/2 object. The targets with a specific scientific interest are
slit width for the low or medium grating mode. However, at the presented in Figure 2. We made a distinction from high
time of our observations, only one set of medium-resolution priority to low priority among the following.
gratings was available and the wavelength coverage of the
medium-resolution mode is restricted to 510–540 nm in the 1. “Standard targets” (∼120): MK secondary standard stars.
blue and 830–890 nm in the red. This grating set was not used. These targets were originally introduced in the beginning
Moreover, at the start of the pilot survey on 2011 October 24, of the LK-project because LAMOST spectra of these objects
2
Figure 2. Targets of scientific interest in the FOV of the Kepler mission. The black dots refer to the centers of the 14 LK-fields that cover the Kepler FOV (see
Table 1). The following color coding is used (from high to low scientific importance): green for standard targets, blue for KASC targets, and orange for planet targets.
The scientific importance of the different types of targets within the LK-project is also reflected in the size of the symbols. The LK-fields observed in 2011–2014 are
indicated by the circles drawn with a full line going from thick to thin and from gray to black, respectively.
are needed to calibrate the results of the other objects. As nominal Kepler mission have ended, we know for which
the project evolved, other standard stars based on earlier objects Kepler photometry is available. For the next round of
LAMOST observations were selected automatically. observations, these 199,718 objects will be top priority.
2. “KASC targets” (∼6500): targets in the Kepler FOV To ensure optimal use of all of the available fibers, for each of
selected by the Kepler Asteroseismic Science Consortium the requested LAMOST fields (see Section 3) we provided a list of
(KASC).25 As this project was initiated by the chairs of the additional objects with V < 20 based on the USNO-B catalog
ground-based follow-up working groups for KASC targets, (Monet et al. 2003). These “field targets” are of no particular
these KASC targets were considered to be of the highest scientific interest for our project, but since they are observable
scientific interest. and used to fill all of the available fibers, we will obtain relevant
3. “Planet targets” (∼150,000): targets in the Kepler FOV information on them by homogeneously analyzing their spectra
selected by the Kepler planet search group (Batalha with the Kepler targets. Note that the field targets have not yet
et al. 2010). been cross-identified with the KIC when compiling the target
4. “Extra targets” (∼1,000,000): other targets in the Kepler list. This was only done afterward for all of the objects that were
FOV from the KIC (Brown et al. 2011). They have no observed with LAMOST (see Section 7.1). Of the 38,416 objects
specific scientific interest but were potential targets of the that were originally observed as field targets, only 79 could not
Kepler mission, except those that fall in the regions that be reclassified as a Kepler target after cross-identification (see
are not covered by the Kepler CCDs (white parts on the “Unique” row in Table 1).
Figure 2).
Henceforth, we will refer to all these targets as Kepler 2.2. Brightness
targets. Note that there is overlap with the targets of the LAMOST Ground-based spectroscopic data are also being collected
project of W. Wang et al. (2015, private communication) with 2 m class telescopes elsewhere (e.g., Catanzaro et al.
focusing on planet host star candidates in the Kepler FOV. The 2010; Molenda-Żakowicz et al. 2010a, 2011, 2013; Uytterhoe-
targets of common interest received the same priority as the ven et al. 2010a; Bruntt et al. 2011, 2012; Catanzaro
KASC targets from 2012 onward. It was already clear from the
et al. 2011; Lehmann et al. 2011; Thygesen et al. 2012;
beginning of our project that not all of the Kepler targets could Tkachenko et al. 2012, 2013; Karoff et al. 2013; Niemczura
be observed. Initially, the aim was to observe as many of the et al. 2015). As these observations are restricted to bright
KASC and planet targets as possible, as those targets are of a
targets (generally V < 12), the highest priority for observations
specific scientific interest. Now that the observations of the with LAMOST is given to fainter objects. However, several bright
25
stars studied in the aforementioned works were also targeted
Relevant information on KASC can be found on the Kepler Asteroseismic for comparison purposes.
Science Operations Center (KASOC) web page http://kasoc.phys.au.dk/,
maintained by Rasmus Handberg from the University of Birmingham in the The exposure times for LAMOST observations can only be
United Kingdom. optimized if the brightness range of the observed targets is not
3
Table 1
Statistical Overview of the Observations Obtained up to the End of the 2014 Observation Season for the Kepler FOV for the LK-project
LK-field R.A. (2000) decl. (2000) Cluster Date # Spectra S K P E F KO

LK01 19:03:39.258 +39:54:39.24 L 110530 1 1435 1 131 444 859 L 607
110608 2 1108 4 137 336 629 2 525
140602 2 4944 2 237 2201 2500 4 2586
LK02 19:36:37.977 +44:41:41.77 NGC6811 120604 1 582 18 131 157 273 3 355
140913 2 6858 5 483 3504 2861 5 4171
LK03 19:24:09.919 +39:12:42.00 NGC6791 120615 3 9655 18 591 4825 4215 6 5833
LK04 19:37:08.862 +40:12:49.63 NGC6819 120617 3 9666 18 574 3785 5281 8 5282
LK05 19:49:18.139 +41:34:56.85 L 131005 2 6374 4 474 2704 3191 1 3487
140522 1 2422 4 278 1048 1092 L 1431
LK06 19:40:45.383 +48:30:45.10 L 130522 1 2488 4 366 1383 735 L 1770
130914 1 3037 3 237 1778 1015 4 2046
LK07 19:21:02.816 +42:41:13.07 L 130519 1 3025 12 416 1820 777 L 2303
130523 1 2304 4 152 1590 558 L 1793
130926 1 3165 8 275 1995 886 1 2357
LK08 19:59:20.425 +45:46:21.15 NGC6866 130925 2 5928 2 387 1957 3576 6 2386
131002 1 2795 L 20 19 2751 5 56
131017 1 2769 L 41 906 1820 2 990
131025 1 2970 L 50 1024 1896 L 1119
LK09 19:08:08.340 +44:02:10.88 L 131004 1 3141 4 266 1748 1122 1 2106
LK10 19:23:14.829 +47:11:44.80 L 140520 2 3781 4 222 2401 1154 L 2683
LK11 19:06:51.499 +48:55:31.77 L 140918 1 3070 4 340 1572 1149 5 1940
LK12 18:50:31.041 +42:54:43.72 L 131007 1 2922 5 229 1329 1354 5 1590
LK13 18:51:11.993 +46:44:17.52 L 140502 1 2501 2 199 1102 1196 2 1345
140529 2 4693 1 232 2091 2358 11 2434
LK14 19:23:23.787 +50:16:16.64 L 140917 1 3167 L 187 1455 1521 4 1685
140927 1 3160 1 107 1424 1625 3 1585
140929 1 3126 1 87 1291 1740 7 1441
Total 101,086 129 6849 45,889 48,134 85 55,906

Unique 80,447 93 4712 34,904 40,659 79 42,209
1× 63,333 65 3142 25,919 34,133 74 31,147
2× 14,186 20 1129 7286 5747 4 8859
3× 2483 8 355 1457 662 1 1870
4× 332 L 54 199 79 L 258
+5× 113 L 32 43 38 L 75
Note. The top lines give the specifications of the LK-fields that have been observed. For each LK-field, we give the right ascension (R.A. (2000)) and declination
(decl. (2000)) of the central bright star, the name of the open cluster that it contains (Cluster, if applicable), the date of observation (YYMMDD; Date), the number of
plates that were used to observe the LK-field (#), the total number of spectra that have been reduced successfully (Spectra), and the number of spectra for Kepler
targets—subdivided into standard (S), KASC (K), planet (P), extra (E), targets—field targets (F), and objects that were observed by the Kepler mission (KO). The
bottom lines give the summary of the observations of all of the LK-fields together. For each category of targets, we give the total number of spectra that have been
reduced successfully (Total), the number of different objects that it contains (Unique) and the number of targets that have been observed one (1×), two (2×), three
(3×), four (4×), and at least five (+5´) times.
larger than 5 mag. For the initial observations (2011 May 31, 2012 June 4, a distinction was made only between brighter
June 8, and 2012 June 4), we therefore made a distinction (9.0 < K p  14.0 ) and fainter (K p > 14.0 ) targets. The
between three different brightness intervals from high to low brighter and fainter targets were prioritized in the same way
priority: middle targets (10.5 < K p  15.5), bright targets as the middle and faint targets from the initial observations,
(K p  10.5), and faint targets (K p > 15.5). For the group of respectively.
middle targets, those without KIC stellar parameters are
considered as the most important ones within each type of
target (standard, KASC, planet, and extra). They were prioritized 3. SELECTION OF THE LAMOST FIELDS
according to brightness from bright to faint. Those for which We made use of a simplified version of the bright central star
KIC stellar parameters are already available were prioritized tiling method to cover the Kepler FOV (Cui et al. 2012). This
from high to low Teff because the determination of the KIC Teff method requires a star brighter than V = 8 in the center of the
for hot stars suffered from the lack of an ultraviolet filter in the LAMOST FOV (for the active optics wave front sensor; central
Sloan photometry. The objects in the other brightness groups hole on Figure 1) and four fainter stars with magnitude V < 17
were prioritized according to brightness regardless of the (for the guiding of the CCD cameras; off-center holes on
availability of KIC parameters. The bright targets were Figure 1). We therefore started by selecting 14 circular regions
prioritized from faint to bright to minimize saturation effects with a diameter of 5°, which is the minimum to (almost) fully
(too much flux) while the faint targets were prioritized from cover the Kepler FOV. We refer to them as LK-fields.
bright to faint to minimize the occurrence of underexposed In a second step, we searched for the optimized central
objects (not enough flux). For the observations obtained after coordinates (am , dm ) for the LK-fields that maximize the total
4
number of observable KASC targets. This was done by adding/ 5. OBSERVATIONS

subtracting 0, 1, or 2 steps of 5 arcsec each to/from the central
As only one LK-field was observed per night, covered by, at
coordinates of the original LK-fields. Finally, we searched for maximum, three different plates (=unique configurations of the
the bright stars with V < 8 that are the closest to the resulting fibers), a strict minimum of 14 observation nights was required
am and dm values. These bright stars are indicated by the to observe all of the requested LK-fields of the LK-project.
numbers in Figure 2. The specifications of the corresponding This was due to both the limitations of LAMOST and the visibility
final LK-fields are given in Table 1. The LK-fields are given in of the Kepler FOV. The observation of a LK-field can last at
the suggested observation order, which was chosen to most 4 hr while the celestial objects are passing the meridian.
maximize the number of observable KASC targets. Note that However, as the observable period of the Kepler FOV
four of the LK-fields contain an open cluster (NGC 6791, coincides for a large part with the Monsoon (the Xinglong
NGC 6811, NGC 6819, and NGC 6866). Since the members of observatory is closed from mid-June to mid-September), the
open clusters have more constraints on the input parameters observations can only be done from late-May to mid-June (with
that can be used for asteroseismic modeling (such as a common a maximal observation window of 2 hr before meridian
age and metallicity), the priority of the LK-fields containing an passage) and from mid-September to late-October (with a
open cluster has been increased according to the asteroseismic maximal observation window of 2 hr after meridian passage).
potential of the open cluster. Unfortunately, because of the high About half an hour is needed to focus the system and CCD
density of the objects in these regions, it is impossible to readout takes 8 minutes. In general, 30 minutes (3 × 600 s) are
observe all of the targets of scientific interest in the LK-fields required to observe a “V-plate” (containing objects with a
containing an open cluster. magnitude range of 9 < r  14) and 75 minutes (3 × 1 500 s)
for a “B-plate” (containing objects with a magnitude range of
14 < r  16.3; see Table 2 from 2013 onward). For these
4. SELECTION OF THE FIBERS reasons, four observation seasons of the Kepler FOV were
needed to finish the first round of observations for the LK-
LAMOST is equipped with 4000 optical fibers. All of the fibers project.
have a diameter of 320 μm corresponding to 3.3 arcsec on the A total of 38 plates were observed on 27 different
sky (Cui et al. 2012). A new system for the positioning of the observation nights. The earliest observations (in 2011 May
fibers, called the “distributive parallel-controllable fiber and June) were performed during the test phase of LAMOST. The
positioning system,” has been developed (Xing et al. 1998). 2012 observations were obtained during the pilot survey of
The focal plane of the telescope with a diameter of 1.75 m (5°) LAMOST (from 2011 October 24 until 2012 June 24). The more
is covered by 4000 fiber positioning units whose configuration recent observations date from the general survey phase of
is shown in Figure 1. Each unit covers a circular area with a LAMOST. The LK-fields that were observed in the period
diameter of 33 mm (340 arcsec) on the focal surface and 2011–2014 are indicated in Figure 2 by the circles drawn with a
contains one fiber that can reach all of the places within this full line going from thick to thin and from gray to black,
area (Cui et al. 2012). There is some overlap between the areas respectively. In Table 2, we provide an overview of the
to ensure that there is no blind space on the focal surface. The specifications of the observed plates for the LK-project so far.
4000 fibers can all move separately and accurate positioning of In 2011, the resolution of the observations was either
the fibers only takes about 10 minutes. The total preparation R = 1000 (slit position “×1”) or 2000 (slit position
time for an exposure, including the focus procedure, is about “×1/2”), while the resolution from 2012 onward was fixed to
half an hour. R ~ 1800 (slit position “×2/3”). The observation of a LAMOST
The code “Survey Strategy System” was used to prepare the plate can last at maximum 4 hr as celestial objects can be
observation plans of the LK-project to optimize the effective observed from about 2 hr before to 2 hr after the passage at
use of the fibers (Cui et al. 2012). The prioritized list of targets meridian. The observation times of the sub-exposures of the
(see Section 2) and the equatorial coordinates of the selected plates were adjusted according to the brightness of the targets
bright central stars of the LK-fields (see Section 3) were used as and to the weather conditions. From 2013 onward, the observed
input for the distribution of the fibers. During the allocation of plates are called a “V-plate,” “B-plate,” “M-plate,” and “F-
the fibers, two basic rules apply: the number of observed plate” for targets with 9 < r  14, 14 < r  16.3,
16.3 < r  17.8, and 17.8 < r  18.5, respectively. Under
objects should be maximized and objects that can be observed
normal circumstances, their observation times are 3 ×
by more than one unit should be allocated to the unit that is the
10 minutes, 3 × 25 minutes, 3 × 30 minutes, and 3 ×
nearest to them to reduce the probability of mechanical 30 minutes, respectively (ΔT sub in Table 2).
collisions during the observations. For each spectrograph, at Before the Monsoon break of 2011, there were system errors
least 5 fibers should be allocated to standard targets whose flux between the astronomical coordinate system and the physical
is known to allow a flux calibration of the LAMOST spectra, and coordinates on the focal plane, leading to many low-quality
20 fibers should be pointing toward an object-free region to spectra. Thanks to a calibration of the system coordinates using
allow an accurate correction for the contribution from the sky. the stellar objects on the sky, the whole system error could be
The displacement of a fiber from an object is determined by reduced to less than 0.5 arcsec. Currently, the positioning errors
measuring the flux changes of objects passing through the can be compensated for in real time, ensuring a positioning
fibers when the telescope points to different positions around accuracy of less than 40 μm, i.e., about 0.4 arcsec (Cui
the center of a sky field. The displacements of the 4000 fibers et al. 2012).
are used to build a coordinate relation between the focal surface Before the pilot survey, there were occasional failures of the
and the celestial coordinates. For more details about these dewar system because of an exhaustion of the liquid nitrogen.
procedures, we refer interested readers to Cui et al. (2012). Images taken under these circumstances are dominated by
5
Table 2
Overview of the Specifications of the Observed Plates for the LK-project from 2011 to 2014
Plate LK-field ΔT tot ΔT sub Slit Seeing Transparency Observer

(s) (s) (arcsec)
110530_1 LK01 2400 2 × 1200 ×1 ∼2.75 mostly clear Wei Zhang
110608_1 LK01 900 1 × 600 + 1 × 300 ×1/2 ∼3.60 clear Jianrong Shi
110608_2 LK01 1200 2 × 600 ×1/2 ∼3.60 clear Jianrong Shi
120604_1 LK02 600 2 × 300 ×2/3 ∼3.10 poor Huoming Shi
120615_1 LK03 1600 2 × 600 + 1 × 400 ×2/3 ∼3.75 mostly cloudy Jianrong Shi
120615_2 LK03 1300 2 × 400 + 1 × 500 ×2/3 ∼3.25 mostly cloudy Jianrong Shi
120615_3 LK03 2400 2 × 1200 ×2/3 ∼3.75 mostly cloudy Jianrong Shi
120617_1 LK04 900 1 × 400 + 1 × 500 ×2/3 ∼4.45 cloudy Jianrong Shi
120617_2 LK04 900 1 × 400 + 1 × 500 ×2/3 ∼3.90 cloudy Jianrong Shi
120617_3 LK04 1200 1 × 1200 ×2/3 ∼4.00 cloudy Jianrong Shi
130519_1 LK07 3000 5 × 600 ×2/3 ∼3.50 clear Jianling Wang
130522_1 LK06 2100 3 × 700 ×2/3 ∼3.25 cloudy Jianling Wang
130914_1 LK06 1800 3 × 600 ×2/3 ∼3.10 clear Jianling Wang
130925_1 LK08 1800 3 × 600 ×2/3 ∼3.90 clear Ming Yang
131002_1 LK08 4930 2 × 1800 + 1 × 1330 ×2/3 ∼3.90 clear Wei Zhang
140502_1 LK13 2400 4 × 600 ×2/3 ∼3.50 mostly clear Jianling Wang
140520_1 LK10 1500 1 × 1500 ×2/3 ∼3.20 mostly clear Jianrong Shi
140520_2 LK10 2400 4 × 600 ×2/3 ∼3.10 mostly clear Jianrong Shi
140529_1 LK13 1800 3 × 600 ×2/3 ∼3.70 clear Wei Zhang
140529_2 LK13 1800 3 × 600 ×2/3 ∼3.40 clear Wei Zhang
140913_1 LK02 1800 3 × 600 ×2/3 ∼3.00 mostly clear Ming Yang
140913_2 LK02 1800 3 × 600 ×2/3 ∼3.20 mostly clear Ming Yang
140927_1 LK14 1800 3 × 600 ×2/3 ∼3.80 partly cloudy Jianrong Shi
Note. For each plate, we give the date with the sequence number of the plate (YYMMDD_N; Plate), the reference of the LK-field that was observed (LK-field, see
Table 1), the total integration time (ΔT tot ), and the integration times of the sub-exposures (ΔT sub ) in seconds, the slit width (Slit), the estimated value of the seeing
(Seeing), an indication of the transparency of the sky during the observations (Transparency), and the name of the observer (Observer).
thermal noise due to the rising temperature of the CCD chip, pipeline procedures, we refer interested readers to Luo et al.
making data obtained under these conditions useless. More- (2012) and Luo et al. (2015).
over, there are approximately 110 known bad fibers because
either they were broken or their fiber positioners did not work
properly. Those fibers were excluded from the reduction 6.1. Pipeline
pipeline. The data reduction pipeline consists of 2D and 1D
procedures. The LAMOST 2D procedure is used to reduce
separately the data from each of the 32 CCD chips (2 for each
6. REDUCTIONS
spectrograph) and of each exposure. Apart from the science
The main products of LAMOST are wavelength and flux- frames themselves, the 2D pipeline also makes use of the series
calibrated spectra which are provided to astronomers. They are of bias, dark, arc lamp, and sky flat-field frames that are taken
processed by a data reduction pipeline and an analysis pipeline. at the beginning of each observation night. The reduction
In the following subsections, we provide short descriptions of procedure consists of several steps. The bias and dark frames
the most important reduction steps which are relevant to our are combined into a master bias and a master dark frame each
project. The analysis pipeline includes the classification and night and they are both subtracted from each raw image. From
identification of the spectra and the measurement of the spectral the general survey onward, the dark subtraction is not
parameters of the observed objects. For more details about necessary any longer because the problem with an inside
6
CCD light source that affected the images has been eliminated. wavelength samples of the sub-spectra are not exactly the
For each CCD chip, 250 1D spectra are extracted from the raw same, the combined spectrum is over-sampled. In a second
data by using the row positions derived from the flat-field step, this combined spectrum is smoothed and re-sampled by a
frames. For each region of the sky covered by one spectro- spline function. Note that if the same field was observed on
graph, a super sky based on the sky measurements of at least 20 different nights, the spectra are not summed because the
fibers is created to model and subtract the contribution of the calibration is different for each night and the observed object
background. This step is important because the telluric might be a variable.
continuum and lines, if not properly subtracted, affect the
following analysis, especially for the faintest targets. The 7. DATABASE
centroids of the lines of the arc lamp spectra are measured and a
Legendre polynomial is fit to their wavelengths as a function of A total of 38 plates have been observed so far for the LK-
the pixels. A vacuum wavelength scale is applied for the project. Unfortunately, due to the pointing problems and/or the
wavelength calibration and the resulting wavelength calibra- malfunctioning of some of the spectrographs and/or fibers
tions are accurate to 10 km s−1 (or better). Also, the helio- (Section 5), the raw spectra for a fraction of the observed
centric corrections are applied. The difference in the efficiency objects could not be reduced by the reduction pipeline for some
of the different fibers is corrected for by dividing with a super of the nights, especially during the test phase and the pilot
flat that is constructed by combining the extracted, wavelength- survey in 2011 and 2012 (Section 6). This is reflected both in
calibrated, and re-scaled sky flat-field frames. The telluric the amount of missing spectrograph numbers in Table 3 and in
absorption lines are removed in four wavelength regions the evolution of the success rate of the LAMOST reduction
(685–696 nm; 715–735 nm; 756–772 nm; 810–824 nm). The pipeline for the LK-project, as shown by the full, light gray line
spectra are flux calibrated by matching the instrumental fluxes in Figure 8. In the LK-project, only LAMOST observations for
in the LAMOST spectra of the observed standard targets with which both the 2D and 1D reduction procedures could be
tabulated values in energetic units. The spectra from the applied successfully are analyzed. We did not include in the
red and blue arms of each spectrograph are combined by database those spectra for which the 1D reduction procedure
stacking the points with corresponding wavelengths using a classified the object as either “Unused” (fiber is not used),
B-spline function with inverse variance weighting. Outliers “Unassigned” (fiber is not assigned), “Dead” (fiber is not
due to cosmic rays are rejected and masked and the errors in working), or “Sky” (fiber is used to measure the flux of the
flux are estimated. The combined spectra are re-sampled to sky). This results in 101,086 entries in the database for a total
constant velocity pixels with a pixel scale of 69 km s−1, which of 80,447 different targets. In Table 1, we provide the
corresponds to a wavelength difference of D log (l ) = 0.0001. distribution of these targets for each of the observed LK-fields.
In the LAMOST 1D procedure, the reduced spectra from the In Table 3, we give an overview of the name structure of the
LAMOST 2D procedure are used to estimate the spectral class and files containing the LAMOST spectra and of the spectrograph
vrad of the observed targets. These characteristics are based on identification numbers for which LAMOST data are available for
template matching since, for low signal-to-noise ratio (S/N) the different plates of different nights.
spectra, the fitting and measurement of spectral lines is not We have constructed a database of these 101,086 LAMOST
precise. The vrad determination is performed with a generalized spectra available from the LK-project so far (Tables 4 and 5).
version of the cross-correlation technique. The analysis code The LAMOST 1D fits files of the LAMOST spectra, including those
estimates the spectral type (SpT) of the observed targets by for the LK-project, that have already been released to the public
making c 2 fits in wavelength space of their LAMOST spectra to can be downloaded from LAMOST’s official website.26 The
templates that were constructed from a set of Sloan Digital Sky LAMOST 1D fits files of the unreleased spectra that were made
Survey (SDSS) spectra. The results are checked through a available to us by the LAMOST team can be obtained upon
comparison of the lines of the observed spectrum with those request27 after becoming an external collaborator of the LK-
from the best-fit template. If they are very different, then it is project.28 In column 8 of Table 4, we give the most recent
decided by a visual check whether or not the classification is version of the pipelines for which the results are available. The
correct. The 1D pipeline only gives results for spectra with a first two numbers refer to the reduction pipeline and the last
S/N > 5. After a comparison of the results with the SpTs number to the analysis pipeline. Note that v2.7.5 refers to the
and radial velocities of the observed SDSS stars, it turns most recent versions that are available at this moment. The
out that 96% of the spectra with S/N > 10 in the r band spectra emerging from an older version of the reduction
are classified correctly and an accuracy of 13 km s−1 can be pipeline can be used at one’s own risk. The structure of the
reached for vrad . names of the LAMOST 1D fits files is as follows: spec-
MMMMM-YYYY_spXX-FFF.fit where MMMMM gives the
LAMOST modified Julian day number (LJDN; see note of Table 3
6.2. Combination of the Sub-exposures for the definition), YYYY denotes the plan identification
For LAMOST, the observation of each plate generally consists number, and XX and FFF are the spectrograph and fiber
of more than one sub-exposure (see Table 2). This observation identification number, respectively. The LJDN and plan
strategy is preferred because it makes it possible both to identification number (planID) of the observed plates are given
subtract the cosmic rays from the CCDs and to improve the in Table 3.
S/N by combining the flux of the sequential sub-exposures. 26
The wavelength and flux calibration are performed for each The LAMOST 1D fits files of the released spectra of the LK-project can be
downloaded from LAMOST’s official website: http://www.lamost.org/
sub-exposure individually to overcome the unstable effects of 27
The LAMOST 1D fits files of the unreleased spectra of the LK-project can be
the instrument (such as tracking and pointing) and the temporal requested by e-mail to Peter De Cat: [email protected]
variations of the weather. In the first step, these sub-spectra are 28
Requests to become an external collaborator of the LK-project should be
scaled to the same level before putting them together. As the sent by e-mail to Jianning Fu: [email protected]
7
Table 3
Overview of the LAMOST 1D Fits Files Available from Observations for the LK-project from 2011 to 2014
Plate LJDN planID Version Spectrographs

(å)
110530_1 55712 IF10M v2.6.5 01 02 03 04 05 06 13
110608_1 55721 IF10B v2.6.5 01 03 04 05 06 07 15
110608_2 55721 IF10M v2.4.4 06 10 13 15
120604_1 56083 IF04_B56083 v2.7.5 01 03 07 08 09 11 12 13 14 15 16
120615_1 56094 kepler05B56094 v2.7.5 01 02 03 04 05 06 07 09 10 11 12 13 14 15 16
120615_2 56094 kepler05B56094_2 v2.7.5 01 02 03 04 05 06 07 09 10 11 12 13 14 15 16
120615_3 56094 kepler05F56094 v2.7.5 01 02 03 05 06 07 09 10 11 12 13 14 15 16
120617_1 56096 kepler08B56096_1 v2.7.5 01 02 03 05 06 07 08 09 11 12 13 14 15 16
120617_2 56096 kepler08B56096_2 v2.7.5 01 02 03 05 06 07 08 09 10 11 12 13 14 15 16
120617_3 56096 kepler08F56096 v2.7.5 01 02 03 05 06 07 08 09 10 12 13 14 15 16
130519_1 56432 KP192102N424113V01 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
130522_1 56435 KP194045N483045V01 v2.7.5 01 03 04 05 06 07 08 09 11 12 13 14 15 16
130523_1 56436 kepler02_1 v2.7.5 01 02 03 05 06 07 08 09 10 11 12 13 14 15 16
130914_1 56550 KP194045N483045V02 v2.7.5 02 03 04 05 07 08 09 10 11 12 13 14 15 16
130925_1 56561 KP195920N454621V01 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
130925_2 56561 KP195920N454621V02 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
130926_1 56562 KP192102N424113V02 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
131002_1 56568 KP195920N454621M01 v2.7.5 01 02 03 04 05 07 08 09 10 11 12 14 15 16
131004_1 56570 KP190808N440210V01 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
131005_1 56571 KP194918N413456V01 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
131005_2 56571 KP194918N413456V02 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
131007_1 56573 KP185031N425443V01 v2.7.5 01 02 03 04 05 07 08 09 10 11 12 13 14 15 16
131017_1 56583 KP195920N454621V03 v2.7.5 02 03 04 05 07 08 09 10 11 12 13 14 15 16
131025_1 56591 KP195920N454621V3 v2.7.5 01 02 03 04 05 07 08 09 10 11 12 13 14 15 16
140502_1 56780 KP185111N464417V01 v2.7.5 01 02 03 04 05 06 07 08 09 10 12 14 15 16
140520_1 56798 KP192314N471144B01 v2.7.5 01 03 04 05 06 10 11 12 15
140520_2 56798 KP192314N471144V01 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
140522_1 56800 KP194918N413456V v2.7.5 01 02 03 04 05 06 07 08 09 10 11 13 14 15 16
140529_1 56807 KP185111N464417V02 v2.7.5 01 02 03 04 05 06 07 08 09 10 12 13 14 15
140529_2 56807 KP185111N464417V03 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15
140602_1 56811 KP190339N395439V01 v2.7.5 01 02 03 04 05 06 07 09 12 13 14 15 16
140602_2 56811 KP190339N395439V02 v2.7.5 01 02 03 04 05 06 08 09 10 11 12 13 14 15 16
140913_1 56914 KP193637N444141V01 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
140913_2 56914 KP193637N444141V02 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
140917_1 56918 KP192323N501616V v2.7.5 01 02 03 04 05 06 07 09 10 11 12 13 14 15 16
140918_1 56919 KP190651N485531V01 v2.7.5 01 02 03 04 05 06 07 09 10 11 12 13 14 15 16
140927_1 56928 KP192323N501616V02 v2.7.5 01 02 03 04 06 07 08 09 10 11 12 13 14 15 16
140929_1 56930 KP192323N501616V03 v2.7.5 01 02 03 04 05 06 07 08 09 10 11 12 13 14 16
Note. For each plate, we give the date with the sequence number of the plate (YYMMDD_N; Plate), the LAMOST modified Julian day number (LJDN; see note (å)
below for the definition) and the plan identification number (planID) as used in the name structure of the LAMOST 1D fits files, the latest available version of the
reduction pipeline (first two numbers) and analysis pipeline (last number) for which the results are available (Version), and spectrograph identification numbers for
which data is available in the remaining columns (Spectrographs).
(å) The LAMOST modified Julian day number (LJDN) is the integer part of the LAMOST modified Julian date (LJD), which is defined as the modified Julian date (=JD—
2400000.5) after adding 20 hr (5/6). This places the start of the LJD timescale at 1858 November 16 at noon in local Beijing time. Hence, all of the LAMOST data
obtained during one observation night will have the same LJDN.
7.1. Cross-identification between the equatorial coordinates of the position of the fibers
and the coordinates of the observed object as given in the target
For all of the available LAMOST 1D fits files, we cross-
list. The objects that were flagged as having a large offset are
checked whether or not the observed objects could be identified
marked with “N1” in column 14 of Table 4.
with a Kepler object from KIC. The R.A. and decl. of the
We initially searched for objects within the default search
positions of the fibers as recorded in the headers in the LAMOST radius of 0.02 arcmin of these coordinates. If no Kepler object
1D fits files (see columns 9 and 10 in Table 4) were used as was found in this region, then we enlarged the search region to
input for the KIC search tool.29 These coordinates generally a maximum radius of 0.16 arcmin (=10 arcsec). Objects for
correspond to the equatorial coordinates of the objects as given which the cross-identification was only found with such a
in the target list. For a fraction of those objects brighter than larger search radius are marked with “N2” or “N3” in
magnitude 11, an offset is added to the pointing of the fibers to column 14 of Table 4 and column 16 of Table 5. The few
avoid saturation. The term “offset” refers here to the difference observations for which another KIC object better matches the
equatorial coordinates of the fiber than the Kepler target
29
The Kepler Input Catalog can be consulted with the KIC search tool: http:// mentioned in the headers of the LAMOST 1D fits files (column 3
archive.stsci.edu/kepler/kic10/search.php of Table 4) are marked with “N4” to “N15” in column 14 of
8
Table 4
Database of the LAMOST Spectra Obtained for the LK-project: Information Extracted from the Headers of the LAMOST 1D Fits files
(1) (2) (3) (4) (5) (6) (7) (8)

File Name Target ID yyyy-mm-ddThh:mm:ss.ss Type KO Slit Version
(9) (10) (11) (12) (13) (14)
R.A.deg decl.deg MAGtype MAG S/Nr Notes
(°) (°) (mag)
spec-55712-IF10M_sp01-001 KIC02140382 ZHT17_Std_0000110 2011-05-30T19:01:19.98 E N x1 v2.6.5
285.914888 37.548585 SSA_BR2I 14.95 49.81
spec-55712-IF10M_sp01-002 KIC01995622 extra00011944 2011-05-30T19:01:19.98 P Y x1 v2.6.5
286.184208 37.482444 Kp 13.52 112.71
spec-55712-IF10M_sp01-003 KIC01994759 ZHT17_Std_0000095 2011-05-30T19:01:19.98 E N x1 v2.6.5
285.926985 37.494571 SSA_BR2I 16.73 9.25
spec-55712-IF10M_sp01-004 KIC02141025 extra00014786 2011-05-30T19:01:19.98 E N x1 v2.6.5
286.134417 37.560833 Kp 14.83 41.10
spec-55712-IF10M_sp01-005 KIC01995904 extra00012008 2011-05-30T19:01:19.98 E N x1 v2.6.5
286.261583 37.482583 Kp 14.94 47.37
286.034917 37.664778 Kp 12.62 202.07
286.104708 37.604194 Kp 15.37 33.25
285.862583 37.682472 Kp 15.01 51.25
286.155583 37.667000 Kp 15.29 42.23
spec-55712-IF10M_sp01-011 KIC02422998 kplr002422998 2011-05-30T19:01:19.98 K Y x1 v2.6.5
286.192417 37.740083 Kp 10.63 531.39
Note. Only the first 10 entries are shown. For a description of the columns, we refer to the text (see Section 7.2). For clarity, columns (9)–(14) are given on a second
line for each entry.
(This table is available in its entirety in machine-readable form.)
Table 4 and column 16 of Table 5. All of the marked cross- that the cross-identifications of targets with a large angular
identifications should be treated with caution. In the case where separation should be treated with caution.
an object could be cross-identified with a KIC object, we
adopted the KIC number as the final identification of the object 7.2. Description
and retrieved the KIC values of the stellar parameters (Teff ,
For all of the LAMOST spectra in the database, Table 4 lists
log g, and [M H ]), the reddening (E (B - V )), the contamina-
information contained in or extracted from the headers of their
tion factors, and proper motions and listed them in the database
1D fits file. This table contains the following columns.
if available (see Table 5).
An overview of the “angular separation” between the 1. Column 1: the file name of the LAMOST 1D fits file
equatorial coordinates of these cross-identified objects and (File Name).
the equatorial coordinates of the position of their fiber 2. Column 2: the final identification of the target after cross-
(column 5 of Table 5) as a function of their Kepler magnitude identification with the KIC catalog (Target).
K p (column 7 of Table 5) is given in Figure 3. The higher the 3. Column 3: the identification of the target as given in
angular separation, the more doubtful is the cross-identifica- headers of the LAMOST 1D fits file (ID).
tion. The horizontal dashed line indicates the standard search 4. Column 4: the date and time (Obs) of mid-exposure
radius of 0.02 arcmin around the equatorial coordinates of the (yyyy-mm-ddThh:mm:ss.ss).
position of the fibers. The top panel shows the observed objects 5. Column 5: the type of target (Type). The standard targets
that were listed as KIC objects in the target list. It clearly shows are indicated with “S,” KASC targets with “K,” planet
that the angular separations of these objects are minimal, except targets with “P,” extra targets with “E,” and field targets
for a fraction of the targets with magnitude K p > 11. The with “F”.
amount of offset that has been added to the pointings of the 6. Column 6: the availability of data from the Kepler
fibers is related to the brightness of these objects. The bottom mission for this object (KO). “Y” means that the object
panel shows those objects that were originally observed as a has been observed by the Kepler mission, and hence that
field target but which were cross-identified with a KIC object. Kepler data is available. “N” means no Kepler data is
For these targets, the “offset” and “angular separation” are not available.
expected to be identical as their coordinates given in the target 7. Column 7: the position of the slit (Slit). “×1” means that
list originate from the USNO-B Catalog (Monet et al. 2003) the fiber slit uses the full slit width (R = 1000). In case of
and not from KIC (Brown et al. 2011). Indeed, for these “×2/3” or “×1/2,” the fiber slit is limited to 2/3 (R ∼
objects, large angular separations are found for objects of all 1800) or half (R = 2000) of the slit width, respectively.
magnitudes and no clear relation is found between the angular 8. Column 8: the most recent version number of the
separation and the brightness of the observed objects. We recall pipelines for which the results are available (Version).
9
Table 5
Database of the LAMOST Spectra Obtained for the LK-project: Additional Information About the Cross-identified Kepler
Targets Extracted from the KIC (Brown et al. 2011)
(1) (2) (3) (4) (5) (6) (7) (8)

File Name Target R.A. decl. Ang Con Kp MagDiff
(hh:mm:ss) (dd:mm:ss) (arcmin) (mag) (mag)
(9) (10) (11) (12) (13) (14) (15) (16)
Teff log g [M H ] E(B − V) pm pmR.A. pmdecl. Notes
(K) (dex) (dex) (mag) (arcsec yr−1) (arcsec yr−1) (arcsec yr−1)
spec-55712-IF10M_sp01-001 KIC02140382 19:03:39.559 37:32:54.92 0.003 L 14.734 0.216

6116 4.614 −0.319 0.127 0.017 0.015 0.007
spec-55712-IF10M_sp01-002 KIC01995622 19:04:44.213 37:28:56.78 0.0006 0.034(#4) 13.517 0.003
5675 4.223 −0.748 0.108 0.004 0.003 −0.003
spec-55712-IF10M_sp01-003 KIC01994759 19:03:42.471 37:29:40.58 0.002 L 16.212 0.518
0.000 0.000 0.000
spec-55712-IF10M_sp01-004 KIC02141025 19:04:32.263 37:33:38.95 0.001 0.066(#4) 14.827 0.003
5916 4.294 −0.096 0.151 0.010 −0.003 −0.010
spec-55712-IF10M_sp01-005 KIC01995904 19:05:02.777 37:28:57.32 0.0007 0.116(#4) 14.944 −0.004
6048 4.281 −0.166 0.159 0.000 0.000 0.000
spec-55712-IF10M_sp01-006 KIC02283124 19:04:08.376 37:39:53.24 0.001 0.023(#4) 12.622 −0.002
6821 4.232 0.045 0.114 0.011 −0.002 −0.011
spec-55712-IF10M_sp01-008 KIC02283334 19:04:25.126 37:36:15.12 0.0009 0.099(#4) 15.372 −0.002
6067 4.557 −0.170 0.151 0.004 0.0002 −0.004
spec-55712-IF10M_sp01-009 KIC02282652 19:03:27.017 37:40:56.93 0.0007 0.068(#4) 15.005 0.005
4910 4.336 0.018 0.109 0.005 −0.005 0.0003
spec-55712-IF10M_sp01-010 KIC02283496 19:04:37.339 37:40:01.20 0.0001 0.118(#4) 15.286 0.004
5397 4.438 −0.170 0.133 0.010 0.009 −0.005
spec-55712-IF10M_sp01-011 KIC02422998 19:04:46.178 37:44:24.32 0.0005 0.006(#4) 10.630 0.000
5185 4.644 −0.728 0.019 0.0005 −0.0001 −0.0003
Note. Only the first 10 entries are shown. For a description of the columns, we refer the reader to the text (see Section 7.2). For clarity, columns (9)–(16) are given on a
second line for each entry.
(This table is available in its entirety in machine-readable form.)
The first two numbers refer to the reduction pipeline and

the last number to the analysis pipeline.
9. Column 9: the R.A. of the fiber in degrees (R.A.deg). In
general, it corresponds to the R.A. of the object as given
in the target list (epoch 2000.0), except for a fraction of
the targets brighter than magnitude 11, for which an
offset is added to the pointings of the fibers to avoid
saturation.
10. Column 10: same as column 9, but for the declination of
the fiber in degrees (decl.deg).
11. Column 11: the type of magnitude as given in headers of
the LAMOST 1D fits file (MAGtype).
12. Column 12: the brightness of the target in magnitude as
given in headers of the LAMOST 1D fits file (MAG).
Targets for which the brightness is not known have no
entry in this column.
13. Column 13: the S/N in the r band (S/Nr). The entry
Figure 3. Overview of the angular separation (in arcmin) between the
“+99.99” means that S/Nr > 100 but that the exact value
equatorial coordinates of the fibers as given in the headers of the LAMOST 1D fits is not given.
files and the equatorial coordinates of the KIC objects they were cross- 14. Column 14: note on the cross-identification as described
identified with (column 5 of Table 5) as a function of their Kepler magnitude above in the text (Notes).
K p (column 7 of Table 5). The observed objects that were listed as a KIC
object in the target list are given in the top panel (black dots). Those objects For those targets identified with an object from the KIC,
that were originally observed as field targets are given in the bottom panel (gray
squares). The horizontal dashed lines in light gray indicate the standard search
extra information extracted from the KIC is given in Table 5.
radius of 0.02 arcmin around the equatorial coordinates of the fibers during the There are 17,241 observed targets for which there are no KIC
cross-identification process. parameters available. Apart from the first two columns, which
10
Figure 4. Examples of high-quality LAMOST spectra of KIC 4373195 (left), KIC 5188234 (middle), and KIC 5357380 (right) obtained during the LK-project. For each
of these objects, we plot the approximately flux-calibrated spectra in the full observed wavelength range (panels (a)) and the continuum-normalized fluxes in three
different wavelength regions: 380–430 nm (panels (b)), 640–690 nm (panels (c)), and 840–890 nm (panels (d)). The heliocentric radial velocity corrections were
applied during the pipeline reductions. The sampling resolution in pixels varies with wavelength and averages 2.5 sampling pixels per resolution element (Luo
et al. 2015).
are identical to those given in Table 4, Table 5 contains the 8. Column 10: the KIC surface gravity in dex (log g).
following extra columns. 9. Column 11: the KIC metallicity in dex ([M H ]).
10. Column 12: the KIC reddening in mag (E (B - V )).
1. Column 3: the KIC right ascension in hh:mm:ss.sss (R.A.). 11. Column 13: the total proper motion in arcsec yr−1 (pm).
2. Column 4: the KIC declination in dd:mm:ss.ss (decl.). 12. Column 14: the angular changes per year in right
3. Column 5: the angular separation in arcminutes between ascension in arcsec yr−1 (pmR.A.).
the equatorial coordinates of the fiber (columns 9 and 10 13. Column 15: the angular changes per year in declination in
of Table 4) and the equatorial coordinates of the KIC arcsec yr−1 (pmdecl.).
object it was cross-identified with (columns 3 and 4 of 14. Column 16: note on the cross-identification as described
this table; Ang) above in the text (Notes).
4. Column 6: the contamination factor of those objects for
which Kepler photometry is available (Con). The Kepler These parameters are not available for every target in the KIC.
light contamination can be found for different seasons Tables 4 and 5 serve as input catalogs for the comparison of
(season 0, 1, 2, and/or 3) with the Kepler Target search the results obtained in the other papers of this series.
tool30, where it is defined as the fraction of light in the
aperture that is not due to the target star. We list the
7.3. Examples
average value of the available seasons. The number of
seasons used is given between brackets (see the As an example of the output of the LAMOST reduction
discussion in Section 7.4.4). pipeline, in Figure 4, we show the spectra of three stars with a
5. Column 7: the Kepler K p magnitude in mag (K p). very different SpT. The spectra are those of KIC 3756031 (left
6. Column 8: the difference between the magnitude as given panels), KIC 8235498 (middle panels), and KIC 5357380 (right
in headers of the LAMOST 1D fits file (column 12 of Table 4) panels) classified by the 1D pipeline as A0-, G0-, and M1-type
and the Kepler K p magnitude of the KIC object it was stars, respectively. The top panels (Figure 4 (a)) show the full
cross-identified with (column 5 of this table; MagDiff). 1D “flux-calibrated” spectra, while the lower panels display
7. Column 9: the KIC effective temperature in K (Teff ). three portions of those spectra automatically normalized to the
local continuum or pseudo-continuum. The most relevant lines,
30
Parameters for Kepler targets can be found with the Kepler Target search such as the high members of the Balmer series for the hot stars
tool: http://archive.stsci.edu/kepler/kepler_fov/search.php or the Ca II H & K lines for the cool ones, are evident in panels
11
(b) of Figure 4. The other two spectral segments contain Hα depend on the accuracy of the flux calibration of the LAMOST
and the near-IR region (panels (c) and (d) of Figure 4, spectra. Based on tests of spectra classified by humans, the
respectively). For hot stars, the near-IR region clearly shows systematic error and standard deviation of the spectral and
several Paschen lines while the Ca II infrared triplet lines are luminosity classes are 0.1 and 0.5 spectral sub-classes (where a
visible for cool stars. All of these spectral features are unit spectral subclass is the difference between, for instance, F5
fundamental for classification purposes. and F6) and 0.02 and 0.5 luminosity classes (where a unit
There are three teams characterizing all of the observed luminosity class is the difference between, for instance, a dwarf
targets by analyzing all of the available LAMOST spectra with (V) and sub-giant (IV) classification), respectively. Thus, the
sufficient quality. They all use a different approach and so the accuracy of MKCLASS is similar to the level of agreement
results of the different teams are independent. obtained by two independently working skilled human
The “Asian team” (A. B. Ren et al. 2015, in preparation) is classifiers.
performing a statistical analysis of the stellar parameters In this paper, the results obtained by the different teams for a
resulting from the LAMOST stellar parameter pipeline (LASP; few example files are given only to illustrate their compat-
Luo et al. 2015). It determines the stellar parameters (Teff , log g, ibility. Table 6 lists the stellar parameters as determined with
[Fe H]), vrad , and the spectral subclass. For the initial guess of LASP, ROTFIT, and MKCLASS for a variety of SpTs (late B/early A,
the stellar parameters, the Correlation Function Interpolation late A/early F, late F/early G, late G/early K, and late K/early
method is used (Du et al. 2012). It searches for the best fit in a M). For each of these SpTs, we show the results for a spectrum
grid of 8903 synthetic spectra that was built by using the with S/Nr of >100, ∼50, and ∼20 from top to bottom. In
Kurucz spectrum synthesis code based on the ATLAS9 stellar general, the results for the stellar parameters Teff , log g, and
atmosphere models provided by Castelli & Kurucz (2003). The [Fe H] are fully compatible with each other. The few that do
final values are determined with a version of the Université de not agree within the formal errors are given in italics. They are
Lyon Spectroscopic analysis Software (Koleva et al. 2009) found both among high and low S/Nr LAMOST spectra. The
adapted to the LAMOST data. This is done by minimizing the agreement in the results for the spectral classification is less
squared difference between the observations and models. For obvious. In those cases where the spectral sub-classes show a
the construction of the model corresponding to a set of values range of more than two, the results are given in italics. These
of the stellar parameters, an interpolator is used that runs over differences are partly due to the bad coverage of certain SpTs in
the whole ELODIE library (Prugniel & Soubiran 2001, 2004; the grids of template spectra that were used for the LASP and
Prugniel et al. 2007). For more details about the methods of the ROTFIT calculations. For this reason, the results of MKCLASS
Asian team, we refer interested readers to Wu et al. (2011a, should be considered as the most trustworthy. However, given
2011b). the discrepancies in the standard templates, the agreement in
The “European team” (A. Frasca et al. 2015, in preparation) SpTs, especially between ROTFIT and MKCLASS, is rather good,
is determining the stellar parameters and the spectral even for the lowest S/Nr.
classification with an adapted version of the code ROTFIT In Figure 5, we show a statistical comparison of the stellar
(Frasca et al. 2003, 2006). The observed spectra are fit to those parameters obtained by the Asian team (LASP) and the European
available in a grid of spectra for a selection of more than 1000 team (ROTFIT) based on the LAMOST spectra observed in 2011.
comparison stars with known stellar parameters from the Indo- The results for Teff , log g, and [Fe H] derived from the 1035
U.S. Library of Coudé Feed Stellar Spectra (Valdes LAMOST spectra that have been analyzed by both groups are
et al. 2004). Because these spectra have a higher spectral given in panels (a)–(c), respectively. The overall agreement of
resolution compared to LAMOST, they had to be degraded to the results is very good as the mean value and the standard
match the low-resolution LAMOST spectra. As the library spectra deviation of the residuals amounts to −34 K and 221 K for Teff ,
are in the laboratory rest frame and are corrected for their −0.016 dex and 0.35 dex for log g, and −0.019 dex and
heliocentric vrad , they also serve as templates to derive vrad 0.26 dex for [Fe H], respectively. The intensity of the colors
using the cross-correlation technique. Moreover, the analysis that are used are linked to the S/Nr of the LAMOST spectrum that
method is also capable of providing a rough estimate of v sin i is analyzed: black corresponds to S/Nr = 0 while the lightest
for rapidly rotating stars. colors correspond to S/Nr 100. Figure 5 shows that the
The “American team” (R. A. Gray et al. 2015, in results which deviate most significantly tend to originate from
preparation) has developed the code MKCLASS for automatically LAMOST spectra with a lower S/Nr value. For Teff , the agreement
classifying stars on the MK spectral classification system of the results from both methods is excellent with no apparent
independent of the stellar parameter determination (Gray & systematic behavior, except for the hottest stars. This is likely
Corbally 2014). This method requires a library of spectral due to the fact that the grids of the model spectra used in the
standards and is designed to classify stellar spectra by direct analysis methods are sparse in this part of the Hertzsprung–
comparison with MK standards using the same criteria as Russel (H–R) diagram. For log g, the values provided by the
human classifiers. MKCLASS is capable of recognizing many of two analysis codes agree quite well with each other, but the
the common spectral peculiarities, including barium stars and ROTFIT results appear to be clustered around 2.5 and 4.0 dex.
carbon-rich giants, Ap and Am stars, λ Bootis stars, helium- This behavior may be due to the non-uniform density of the
weak and helium-strong B-type stars, etc. For the LAMOST templates in the grid used by ROTFIT and in the ELODIE library
classifications, the flux-calibrated standards library with used for interpolating in the LASP code. Indeed, these log g
0.36 nm-resolution spectra obtained using the GM values are those typical for main-sequence and giant stars,
spectrograph at the Dark Sky Observatory of Appalachian respectively, which are much more frequent in these spectral
State University is employed. Therefore, the LAMOST spectral libraries than stars with an intermediate log g or outside these
resolution is slightly degraded to match that of the spectral ranges. For [Fe H], the overall comparison is satisfactory but
library. The accuracy of the resulting classifications does not there seems to be a linear trend in the residuals that could be
12
Table 6
Representative Results Obtained by the Asian Team (LASP), the European Team (ROTFIT), and the American Team (MKCLASS)
File Name/Object Version S/Nr SpT Teff log g [Fe H] Method

(K) (dex) (dex)
spec-56094-kepler05B56094_2_sp13-034 v2.6.4 581 A2 V 9380(56) 3.5(11) −0.35(5) LASP
KIC 4373195 v2.4.3 B9 IV 9707(589) 3.83(11) −0.16(14) ROTFIT
* v2.7.5 A0 IV–V MKCLASS
spec-56096-kepler08B56096_1_sp07-250 v2.6.4 66 A2 V 11,223(113) 3.79(18) −0.18(8) LASP
KIC 3137033 v2.4.3 B9 III 10,839(342) 4.00(11) −0.08(13) ROTFIT
v2.7.5 B9 V MKCLASS
KIC 3654218 v2.4.3 B9 IV 10,336(248) 3.88(10) −0.26(13) ROTFIT
v2.7.5 A1 IV MKCLASS

KIC 3446837 v2.4.3 F0 V 6,989(95) 4.08(11) −0.13(13) ROTFIT
v2.7.5 F2 IV–V Fe-0.6 MKCLASS
spec-56094-kepler05B56094_sp16-201 v2.6.4 65 F5 6593(30) 4.14(5) −0.18(6) LASP
KIC 5609930 v2.4.3 F0 V 6592(215) 4.10(13) −0.16(15) ROTFIT
v2.7.5 F5 V MKCLASS
spec-56094-kepler05B56094_sp12-098 v2.6.4 25 F5 6733(36) 4.10(6) −0.22(7) LASP
KIC 5449646 v2.4.3 F0 V 6820(281) 4.12(11) −0.14(14) ROTFIT
v2.7.5 F3 V MKCLASS
spec-56094-kepler05B56094_2_sp09-061 v2.6.4 279 G0 5619(12) 3.63(4) −0.15(3) LASP

KIC 5188234 v2.4.3 G0 V 5774(97) 4.13(13) −0.11(17) ROTFIT
* v2.7.5 G0 IV–V MKCLASS
spec-56094-kepler05F56094_sp15-020 v2.6.4 50 G0 5741(41) 4.30(12) −0.33(10) LASP
KIC 5095915 v2.4.3 G0 V 5731(91) 4.22(14) −0.41(16) ROTFIT
v2.7.5 G1 V MKCLASS
spec-56096-kepler08B56096_1_sp05-091 v2.6.4 17 F9 5765(76) 3.88(26) −0.18(16) LASP
KIC 3655915 v2.4.3 G0 V 5647(216) 4.05(25) −0.43(20) ROTFIT
v2.7.5 G2 III–IV MKCLASS
spec-56094-kepler05B56094_sp07-059 v2.6.4 114 G5 5325(15) 4.65(4) −0.00(2) LASP

KIC 2308241 v2.4.3 K0 V 5286(141) 4.54(15) 0.02(18) ROTFIT
v2.7.5 K2 V MKCLASS
spec-56094-kepler05B56094_2_sp15-172 v2.6.4 46 K1 5336(28) 4.80(8) 0.17(5) LASP
KIC 5442808 v2.4.3 K0 V 5162(157) 4.57(15) 0.05(12) ROTFIT
v2.7.5 K0 V MKCLASS
spec-56094-kepler05F56094_sp13-172 v2.6.4 24 G5 5285(66) 4.41(16) 0.21(10) LASP
KIC 4839014 v2.4.3 K0 V 4968(191) 3.52(40) 0.03(14) ROTFIT
v2.7.5 K0 V MKCLASS
spec-56094-kepler05B56094_2_sp15-166 v2.6.4 219 M1 3717(7) 1.49(9) −0.05(10) LASP

KIC 5357380 v2.4.3 M1 III 3744(85) 1.47(15) −0.07(13) ROTFIT
* v2.7.5 M0.5 II–III MKCLASS
spec-56096-kepler08B56096_2_sp15-026 v2.6.4 44 M1 3754(21) 1.45(25) −0.31(28) LASP
KIC 6292615 v2.4.3 M1 III 3787(96) 1.52(17) −0.09(13) ROTFIT
v2.7.5 M0.5 II MKCLASS
spec-56096-kepler08B56096_2_sp13-010 v2.6.4 21 M0 3738(18) 1.31(19) 0.04(22) LASP
KIC 5725850 v2.4.3 M1 III 3761(80) 1.50(15) −0.05(11) ROTFIT
v2.7.5 M2 III MKCLASS
Note. The results derived from the LAMOST spectra for examples of late B-/early A-, late A-/early F-, late F-/early G-, late G-/early K-, and late K-/early M-type stars
are given from top to bottom. For each type, we give an example of a LAMOST spectrum with S/Nr > 100, S/Nr ∼ 50, and S/Nr ∼ 20 (from top to bottom). For each
spectrum, we give the name of the analyzed LAMOST spectrum and the KIC number of the observed target (File Name/Object; the LAMOST spectra indicated with ∗ are
shown in Figure 4), the version of the pipelines that was used for each method (Version), the spectral classification (SpT), the effective temperature in K (Teff ), the
surface gravity in dex (log g ), the metallicity in dex ([Fe H]), and the analysis method (Method; see Section 7.3 for more details about the methods used). The formal
errors are given between brackets in units of the last decimal. The stellar parameters that do not agree within the formal errors and the spectral sub-classes with a range
of more than two are given in italics.
due to similar contamination problems. Indeed, the metal-poor only shows results derived from the LAMOST spectra observed in
stars, as well as the very metal-rich ones, are far less frequent in 2011 during the test phase, and so the observation and pipeline
these libraries. procedures were not yet fully optimized. Hence, the discussion
Note that all of the LAMOST spectra shown in Table 6 were provided here can be considered as the worst case scenario. The
gathered during the pilot survey in 2012 June and that Figure 5 full results obtained by the different teams will be given in the
13
Figure 5. Statistical comparison of the stellar parameters obtained by the Asian team (LASP) and the European team (ROTFIT) from the LAMOST spectra observed in 2011.
The results for Teff , log g , and [Fe H] are given in red, green, and blue in panels (a), (b), and (c), respectively. The intensity of the color is linked to the S/Nr value of
the LAMOST spectrum from which the results were derived, going from black for S/Nr = 0 to light for S/Nr 100. In the top panels, we give the derived values as
colored dots, the errors as light gray lines, and the bisector as a full dark gray line. In the bottom panels, we show the residuals as colored dots, the mean μ (given in the
top right corner) as a dark gray dashed line, and the mean ± the standard deviation σ (given in the bottom left corner) as dark gray dashed–dotted lines.
following papers of this series. A more accurate and detailed four occasions (K p > 14.0 for plates 120615_3 and 120617_3,
statistical comparison of the results can only be done after the and 140520_1, and K p > 16.8 for plate 131002_1 in Table 3).
publication of the full results obtained by the different analysis Figure 6 (c)–(e) give the distributions of the KIC stellar
teams. parameters of the observed objects: [M H ] (column 11 of
Table 5), Teff (column 6 of Table 5), and log g (column 10 of
Table 5), respectively. The latter two distributions are
7.4. Characterization
visualized in a different way in Figure 6 (f) where the objects
In this section, we discuss a few figures to characterize the are placed in a Kiel diagram (log g versus Teff ), which is an
database of the LK-project. alternative version of the H–R diagram. The distribution of the
KIC [M H ] suggests that the majority of the objects are
7.4.1. Objects
slightly metal-poor with a median value around −0.4 dex. A
subset of 12,000 LAMOST spectra of stars from the LK-project
In Figure 6, we provide an overview of the objects in our have been used by Dong et al. (2014) to study the reliability of
database. The different types of targets are indicated with the metallicity values as given in the KIC catalog. They found
different colors. We use green for standard targets, blue for that the true metallicity and the dynamic range of the Kepler
KASC targets, orange for planet targets, dark gray for extra sample are both systematically underestimated by the KIC.
targets, and light gray for field targets. The scientific Hence, the KIC [M H ] values should be treated with caution.
importance of the different types of targets within the LK- The majority of our objects have a KIC Teff between 4000 and
project is also reflected in the size of the symbols. Note that 7000 K. The Kepler field contains only a few hot stars for
only a few standard targets and field targets were observed, and
which the parameter determination based on Sloan photometry
so they are invisible on some of the panels in Figure 6.
suffers most from the lack of an ultraviolet filter. The double-
The projection in equatorial coordinates is shown in Figure 6
peak structure of the Teff and log g distribution reflects the
(a). From this figure, it can be easily seen that the Kepler FOV
is well covered by observations with LAMOST. Only a few parts superposition of the main sequence (centered around Teff ∼
are not covered: the center and off-center holes of the LK-fields 6000 K and log g ∼ 4.5 dex) and the red giant branch (centered
(containing the central bright star and the guide stars, around Teff ∼ 4800 K and log g ∼ 2.7 dex).
respectively; see Section 3), and the edge of one Kepler There are no significant differences between the distributions
CCD (near R.A. ∼296° and decl. ∼51°). It also shows that of the different groups of targets, except for the spatial and
there is overlap between the observed LK-fields. brightness distributions: the standard targets are very bright, the
Figure 6 (b) shows the magnitude distribution of the objects KASC targets are mainly restricted to magnitude 12, while the
(column 12 of Table 4). The largest concentration of targets is other types of targets extend to fainter targets. Obviously, the
found between magnitudes 12.5 and 14, which is a reflection of gaps between the Kepler CCDs are filled by extra (dark gray
our observation strategy. Indeed, the observations only focused dots) and field (light gray) targets (Figure 6(a)). The currently
on very bright objects (K p < 10.5) on two occasions (plates available LAMOST spectra of each type of target in itself (except
110608_1 and 120604_1 in Table 3) and on faint objects on the standard and field targets that are few) are appropriate to
14
Figure 6. Global characterization of the objects in the database. The projection in equatorial coordinates (columns 9 and 10 of Table 4) of the available targets is
shown in panel (a). We give the distribution of their magnitude (column 12 of Table 4; bin size of 0.2 mag), their KIC metallicity (column 11 of Table 5; bin size of
0.05 dex), their KIC effective temperature (column 9 of Table 5; bin size of 100 K), and their KIC surface gravity (column 10 of Table 5; bin size of 0.05 dex) in
panels (b)–(e), respectively. The latter two distributions are represented in an alternative way by showing the position of the objects in a Kiel diagram (log g vs. Teff ).
The following color coding is used (from high to low scientific importance; from bottom to top within the distributions): green for standard targets, blue for KASC
targets, orange for planet targets, dark gray for extra targets, and light gray for field targets. The scientific importance of the different types of targets within the LK-
project is also reflected in the size of the symbols in panels (a) and (f).
study their stellar parameters, thus tracing the target properties of the parameters for the sub-sample of stars with multiple
across the H–R diagram. exposures are consistent with the general distributions shown in
Figure 6. Hence, these targets can be used to check the
7.4.2. Multiple Observations consistency of the results across the H–R diagram.
As there is considerable overlap between some of the
7.4.3. Signal-to-noise Ratio
observed LK-fields (Figures 2 and 6(a)), there are many objects
that have been observed more than once: 14,186 objects have Figure 7 gives the distribution of the S/N in the r band
been observed two times, 2483 objects three times, 332 objects (S/Nr) of the available LAMOST spectra (column 13 of Table 4).
four times, and 113 objects at least five times (see the bottom It gives an overview of the overall quality of the data. In this
rows of Table 1). Many of these objects are indeed found in the figure, we provide a histogram of the S/Nr values for two data
overlapping regions. Those within the non-overlapping regions sets: the full set of 101,086 LAMOST spectra that have been
are either objects that are observed on different plates of the reduced (in gray) and the full data set restricted to the 42,209
same field or objects that are observed as a field star that turned LAMOST spectra with the highest S/Nr for the targets that have
out to be an already observed Kepler star (or vice versa). Even been observed by the Kepler mission (in black). The top and
though the main aim of our LK-project is to determine the bottom panels give a global overview for S/Nr up to 200 and a
stellar parameters of as many objects within the Kepler FOV as detailed overview for the S/Nr range [0, 20], respectively. The
possible, it is extremely valuable to have a significant number small fraction of LAMOST spectra with S/Nr 200 is not
of objects with multiple LAMOST spectra as they can be used to shown. We consider S/Nr = 10 as the strict minimum
check the consistency of the results of the analysis methods requirement for a LAMOST spectrum to be analyzed and
used and to discover stars with a variable vrad . The distributions S/Nr = 20 as the minimum requirement for an object to be
15
Figure 8. Overview of the evolution of the success rate of LAMOST during the
LK-project over time. We show the percentage of the fibers leading to a
successfully reduced LAMOST spectrum (light gray full line) and to a LAMOST
spectrum with S/Nr 10 (gray circle) and S/Nr 20 (black square). The
observations are subdivided into semesters: May–June and September–October
of each year.
Figure 7. Histogram of the S/N values in the r band (S/Nr) of the LAMOST
spectra (see column 13 of Table 4) for the regions [200, 0] (bin size 20; top)
and [20,0] (bin size 1; bottom panel). The gray and black colors refer to the full
data set (101,086 spectra) and to the full data set restricted to the spectrum with
the highest S/Nr for the targets that have been observed by the Kepler mission
(42,209 spectra), respectively. The full lines give the cumulative percentage of
LAMOST spectra for each data set as a function of decreasing S/Nr value.
removed from our target list for future observations. The full
lines plotted on top of the histograms in each panel of Figure 7
give the cumulative percentage of LAMOST spectra for each data Figure 9. Overview of the Kepler contamination factors for objects in the
set as a function of decreasing S/Nr value. About 67.3% of the database (column 6 of Table 5). A histogram with a bin size of 0.01 is given.
reduced LAMOST spectra have a S/Nr 20 and about 77.5% The full lines give the cumulative percentage of targets as a function of
have a S/Nr 10. So far, we have LAMOST data for 21.1% of increasing contamination factor.
the 199,718 objects that have been observed during the Kepler
mission, but not all of these LAMOST spectra are of a sufficient Section 2). In Figure 9, we give the distribution of the Kepler
quality. For 17.9% of these Kepler mission targets, we have a contamination factor for the 50,053 observed objects for which
LAMOST spectrum with a S/Nr above 10. For 16.3% of them, the
a value can be found with the Kepler Target search tool. We
S/Nr of their best LAMOST spectrum exceeds 20. These latter used a bin size of 0.01 for the construction of the histogram.
targets can be removed from the target list for future LAMOST The full lines give the cumulative percentage of targets as a
observations. function of increasing contamination factor. The distribution
The S/Nr can also be used to estimate the percentage of peaks around a Kepler contamination factor of 0.025 and
LAMOST fibers leading to high-quality data (S/Nr 20) as a 89.77% of the objects have a value below 0.1, which means
function of time. This is illustrated in Figure 8. During the test that the contamination effects for the vast majority of the
phase, high-quality LAMOST spectra emerged from about 15% of objects are small. For 365 objects (0.73%), the Kepler
the fibers only while this percentage reached about 60% during contamination factor is higher than 0.5, which means that
the last semester. This is mainly due to an improvement in the more than 50% of the counts in the light curves of these objects
fiber positioning and the better calibration of the LAMOST comes from neighboring objects. The LAMOST data for such
spectra. Also, the meteorological circumstances during the objects should be interpreted with caution. We cannot rule out
observations can have an effect on the overall quality of the that in some cases that the wrong star was observed.
LAMOST spectra. They can cause seasonal fluctuations but
cannot be responsible for the clear upward trend seen in
Figure 8. 8. DISCUSSION
The LAMOST spectra described in the database of this paper
7.4.4. Contamination have been proven to be of high scientific value, especially for
the large community involved in the Kepler research. The LK-
The Kepler contamination factor of an object tells us how project has led to the first spectroscopic observation for many,
much the Kepler light curves are affected by the presence of generally relatively faint, Kepler targets. To maximize the
neighboring objects. The higher the Kepler contamination visibility of the LAMOST spectra within this community, the
factor, the more overlap with neighboring objects, and hence observations are listed in the overview table with ground-based
the less reliable the results that can be derived from the Kepler follow-up observations for Kepler targets on the web page of
data. As the LAMOST fiber size and typical seeing match the KASOC.
31
Kepler pixel size, the Kepler contamination factor can be used The LAMOST spectra have already been used by several
as an indication of how much the ground-based measurements groups of astronomers. Apart from the three teams that are
of LAMOST can be expected to be affected by contamination
effects. The Kepler contamination factor was not taken into 31
The KASOC web page (http://kasoc.phys.au.dk/) is a private web page that
account for the construction of the prioritized list of targets (see can be accessed by registered KASOC members only.
16
deriving stellar parameters from all of the good-quality LAMOST object Fiber Spectroscopic Telescope LAMOST) is a National
spectra, the data of the LK-project have been useful for Major Scientific Project built by the Chinese Academy of
selecting interesting targets for other types of studies. The Sciences. Funding for the project has been provided by the
wavelength coverage of the LAMOST observations includes the National Development and Reform Commission. LAMOST is
Lithium line at ∼670 nm, making this survey ideal for operated and managed by the National Astronomical Observa-
identifying candidates with unusual Li abundances. Of tories, Chinese Academy of Sciences. The research is
particular interest are those KASC targets where oscillations supported by the ASTERISK project (ASTERoseismic Inves-
have been detected, which allow us to place stringent tigations with SONG and Kepler) funded by the European
constraints on the evolutionary phase and physical properties Research Council (grant agreement No. 267864), the “Lendü-
of the star. Thanks to our LAMOST survey, Silva Aguirre et al. let-2009 Young Researchers” Program, ESA PECS Contract
(2014) have detected the first confirmed Li-rich core-helium- No. 4000110889/14/NL/NDe, and the National Science
burning giant. The LAMOST observations also include the Ca II H Foundation under grant No. NSF PHY05-51164. Part of the
& K lines at 396.85 and 393.37 nm, respectively (see research leading to these results received funding from the
Figure 4(b)), making it possible to measure the chromospheric European Research Council under the European Community’s
activity of the observed stars. The most commonly used Seventh Framework Programme (FP7 20072013) under grant
expression for stellar chromospheric activity is the dimension- agreements No. 227224 (ERC/PROSPERITY), No. 269194
less S index (Baliunas et al. 1995) and the excess flux (Hall (ASK), and No. 312844 (SPACEINN). Funding for the Stellar
et al. 2007). C. Karoff et al. (2015, in preparation) have Astrophysics Center (Aarhus, Denmark) is provided by The
measured both parameters for 7700 main-sequence solar-like Danish National Research Foundation (grant agreement No.
stars with Teff between 5000 and 6000 K in the Kepler field DNRF106). J.N.F., A.B.R., and X.H.Y. acknowledge the
with LAMOST spectra using the same method that was used by support from the Joint Fund of Astronomy of National Natural
Karoff et al. (2013). Bostancı et al. (2015) used the LAMOST Science Foundation of China (NSFC) and Chinese Academy of
spectra of members of NGC 6866 to derive the metallicity and Sciences through grant U1231202, and support from the
the vrad of this open cluster. Some individual LAMOST spectra National Basic Research Program of China (973 Program
from the LK-project have also been used in case studies: e.g., 2014CB845700 and 2013CB834900). M.B. is an F.R.S.-FNRS
the red giant star KIC 5689820 (Deheuvels et al. 2014). Postdoctoral Researcher, Belgium. S.D. is supported by “the
Even though the LAMOST spectra obtained during the test Strategic Priority Research Program—The Emergence of
Cosmological Structures” of the Chinese Academy of Sciences
phase and the pilot survey for the LK-project have been proven
(grant No. XDB09000000). J.M.Ż. acknowledges the Polish
to be of a sufficient quality for several scientific applications,
MNiSW grant NN203 405139 and the European Community’s
the LAMOST observations and procedures are still being
Seventh Framework Programme (FP7/2007–2013) ASK grant
subjected to continuous improvements. For the general survey,
No. 269194. K.U. acknowledges financial support by the
the observation plan is better and resulted in a rapid increase in
Spanish National Plan of R&D for 2010, project AYA2010-
the data quantity, although more bright stars (<14 mag) are
17803. E.N. acknowledges support from NCN grant No. 2014/
observed on bright nights. Also, the quality control has been 13/B/ST9/00902. R.Sz. acknowledges the University of
improved, which makes the success rate of the observations Sydney IRCA grant and was supported by the Hungarian
reach 60%. As more of the acquired spectra reach a S/Nr level Scientific Researc Fund—OTKA K-115709. We thank N.
of 10, more data can be collected. There have already been Batalha and D. W. Latham for providing the initial lists of
several updates in the LAMOST reduction and analysis pipelines. objects that were used to select the planet and extra targets,
For the faint objects, the sky subtraction now reaches the SDSS respectively.
level while there are still some sky continuum problems for
bright stars on bright moon nights. This issue can be solved by
a better sky model. The difficulty of looking for standard stars REFERENCES
in some spectrographs has been solved by using the estimates
Baliunas, S. L., Donahue, R. A., Soon, W. H., et al. 1995, ApJ, 438, 269
of the stellar parameters obtained before the flux calibration. Batalha, N. M., Borucki, W. J., Koch, D. G., et al. 2010, ApJL, 713, L109
Based on these parameter estimates, late F-type stars are Bostanci, Z. F., Ak, T., Yontan, T., et al. 2015, MNRAS, 453, 1095
selected as standard stars and the flux calibration for one Brown, T. M., Latham, D. W., Everett, M. E., & Esquerdo, G. A. 2011, AJ,
spectrograph is performed by using the curve obtained by 142, 112
Bruntt, H., Basu, S., Smalley, B., et al. 2012, MNRAS, 423, 122
comparing the observed spectra with model spectra of standard Bruntt, H., Frandsen, S., & Thygesen, A. O. 2011, A&A, 528, A121
stars. This makes it possible to connect almost all of the blue Castelli, F., & Kurucz, R. L. 2003, in IAU Symp. 210, Modelling of Stellar
and red arm spectra and to improve the relative flux calibration. Atmospheres, ed. N. Piskunov, W. W. Weiss & D. F. Gray (San Francisco,
A more detailed description of the recent improvements and of CA: ASP), 20P
Catanzaro, G., Frasca, A., Molenda-Żakowicz, J., & Marilli, E. 2010, A&A,
the data process itself are given in Luo et al. (2015). 517, A3
Astronomers who need help processing LAMOST spectra they Catanzaro, G., Ripepi, V., Bernabei, S., et al. 2011, MNRAS, 411, 1167
are interested in can contact the members of the LAMOST Cui, X.-Q., Zhao, Y.-H., Chu, Y.-Q., et al. 2012, RAA, 12, 1197
consortium.32 The prospect for LAMOST is bright, and the Cunha, M. S., Aerts, C., Christensen-Dalsgaard, J., et al. 2007, A&ARv,
14, 217
observations for the LK-project will continue. Deheuvels, S., Doğan, G., Goupil, M. J., et al. 2014, A&A, 564, A27
Dong, S., Zheng, Z., Zhu, Z., et al. 2014, ApJL, 789, L3
We are thankful for the useful comments and suggestions of Du, B., Luo, A., Zhang, J., Wu, Y., & Wang, F. 2012, Proc. SPIE, 8451, 37
the anonymous referee which improved the quality of our Frasca, A., Alcalá, J. M., Covino, E., et al. 2003, A&A, 405, 149
Frasca, A., Guillout, P., Marilli, E., et al. 2006, A&A, 454, 301
paper. The Guoshoujing Telescope (the Large Sky Area Multi- Gray, R. O., & Corbally, C. J. 2014, AJ, 147, 80
Hall, J. C., Henry, G. W., & Lockwood, G. W. 2007, AJ, 133, 2206
32
Send requests to Dr. Ali Luo ([email protected]). Karoff, C., Metcalfe, T. S., Chaplin, W. J., et al. 2013, MNRAS, 433, 3227
17
Koch, D. G., Borucki, W. J., Basri, G., et al. 2010, ApJL, 713, L79 Silva Aguirre, V., Ruchti, G. R., Hekker, S., et al. 2014, ApJL, 784, L16
Koleva, M., Prugniel, P., Bouchard, A., & Wu, Y. 2009, A&A, 501, 1269 Su, D. Q., Cui, X., Wang, Y., & Yao, Z. 1998, Proc. SPIE, 3352, 76
Lehmann, H., Tkachenko, A., Semaan, T., et al. 2011, A&A, 526, A124 Thygesen, A. O., Frandsen, S., Bruntt, H., et al. 2012, A&A, 543, A160
Luo, A.-L., Zhang, H.-T., Zhao, Y.-H., et al. 2012, RAA, 12, 1243 Tkachenko, A., Lehmann, H., Smalley, B., Debosscher, J., & Aerts, C. 2012,
Luo, A.-L., Zhao, Y.-H., Zhao, G., et al. 2015, RAA, 15, 1095 MNRAS, 422, 2960
McNamara, B. J., Jackiewicz, J., & McKeever, J. 2012, AJ, 143, 101 Tkachenko, A., Lehmann, H., Smalley, B., & Uytterhoeven, K. 2013,
Michel, E. 2006, CoAst, 147, 40 MNRAS, 431, 3685
Molenda-Żakowicz, J., Bruntt, H., Sousa, S., et al. 2010a, AN, 331, 981 Uytterhoeven, K., Briquet, M., Bruntt, H., et al. 2010a, AN, 331, 993
Molenda-Żakowicz, J., Jerzykiewicz, M., Frasca, A., et al. 2010b, Uytterhoeven, K., Szabo, R., Southworth, J., et al. 2010b, arXiv:1003.6089
arXiv:1005.0985 Valdes, F., Gupta, R., Rose, J. A., Singh, H. P., & Bell, D. J. 2004, ApJS,
Molenda-Żakowicz, J., Latham, D. W., Catanzaro, G., Frasca, A., & 152, 251
Quinn, S. N. 2011, MNRAS, 412, 1210 Wang, S.-G., Su, D.-Q., Chu, Y.-Q., Cui, X., & Wang, Y.-N. 1996, ApOpt,
Molenda-Żakowicz, J., Sousa, S. G., Frasca, A., et al. 2013, MNRAS, 434, 1422 35, 5155
Monet, D. G., Levine, S. E., Canzian, B., et al. 2003, AJ, 125, 984 Wu, Y., Luo, A.-L., Li, H.-N., et al. 2011a, RAA, 11, 924
Niemczura, E., Murphy, S. J., Smalley, B., et al. 2015, MNRAS, 450, 2764 Wu, Y., Singh, H. P., Prugniel, P., Gupta, R., & Koleva, M. 2011b, A&A,
Prugniel, P., & Soubiran, C. 2001, A&A, 369, 1048 525, A71
Prugniel, P., & Soubiran, C. 2004, arXiv:astro-ph/0409214 Xing, X., Zhai, C., Du, H., et al. 1998, Proc. SPIE, 3352, 839
Prugniel, P., Soubiran, C., Koleva, M., & Le Borgne, D. 2007, arXiv:astro-ph/ Zhao, G., Zhao, Y.-H., Chu, Y.-Q., Jing, Y.-P., & Deng, L.-C. 2012, RAA,
0703658 12, 723
18

Cat 2015 ApJS 220 19

Uploaded by

Copyright:

Available Formats

Cat 2015 ApJS 220 19

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cat 2015 ApJS 220 19

Uploaded by

Copyright:

Available Formats

The Astrophysical Journal Supplement Series, 220:19 (18pp), 2015 September doi:10.

LAMOST OBSERVATIONS IN THE KEPLER FIELD. I. DATABASE OF LOW-RESOLUTION SPECTRA*

1. INTRODUCTION Kepler spacecraft failed, preventing the telescope from

LK-ﬁeld R.A. (2000) decl. (2000) Cluster Date # Spectra S K P E F KO

Total 101,086 129 6849 45,889 48,134 85 55,906

number of observable KASC targets. This was done by adding/ 5. OBSERVATIONS

Plate LK-ﬁeld ΔT tot ΔT sub Slit Seeing Transparency Observer

Plate LJDN planID Version Spectrographs

(1) (2) (3) (4) (5) (6) (7) (8)

(1) (2) (3) (4) (5) (6) (7) (8)

spec-55712-IF10M_sp01-001 KIC02140382 19:03:39.559 37:32:54.92 0.003 L 14.734 0.216

The ﬁrst two numbers refer to the reduction pipeline and

File Name/Object Version S/Nr SpT Teff log g [Fe H] Method

spec-56096-kepler08B56096_1_sp02-050 v2.6.4 122 A9 V 7,168(13) 4.16(2) −0.42(3) LASP

spec-56094-kepler05B56094_2_sp09-061 v2.6.4 279 G0 5619(12) 3.63(4) −0.15(3) LASP

spec-56094-kepler05B56094_sp07-059 v2.6.4 114 G5 5325(15) 4.65(4) −0.00(2) LASP

spec-56094-kepler05B56094_2_sp15-166 v2.6.4 219 M1 3717(7) 1.49(9) −0.05(10) LASP

You might also like