The Astronomical Journal, 146:81 (28pp), 2013 October doi:10.1088/0004-6256/146/4/81
C
2013. The American Astronomical Society. All rights reserved. Printed in the U.S.A.
TARGET SELECTION FOR THE APACHE POINT OBSERVATORY
GALACTIC EVOLUTION EXPERIMENT (APOGEE)
G. Zasowski
1,2,3,26
, Jennifer A. Johnson
1,2
, P. M. Frinchaboy
4
,S.R.Majewski
3
, D. L. Nidever
5
, H. J. Rocha Pinto
6,7
,
L. Girardi
6,8
, B. Andrews
1
, S. D. Chojnowski
3
, K. M. Cudworth
9
, K. Jackson
4
,J.Munn
10
, M. F. Skrutskie
3
,
R. L. Beaton
3
, C. H. Blake
11
, K. Covey
12
, R. Deshpande
13,14
, C. Epstein
1
, D. Fabbian
15,16
, S. W. Fleming
13,14
,
D. A. Garcia Hernandez
15,16
, A. Herrero
15,16
, S. Mahadevan
13,14
, Sz. M
´
esz
´
aros
15,16
, M. Schultheis
17
, K. Sellgren
1
,
R. Terrien
13,14
, J. van Saders
1
, C. Allende Prieto
15,16
, D. Bizyaev
18
, A. Burton
3
, K. Cunha
19,20
,L.N.daCosta
6,20
,
S. Hasselquist
3
, F. Hearty
3
, J. Holtzman
21
,A.E.Garc
´
ıa P
´
erez
3
,M.A.G.Maia
6,20
,R.W.O’Connell
3
, C. O’Donnell
3
,
M. Pinsonneault
1
, B. X. Santiago
6,22
, R. P. Schiavon
23
, M. Shetrone
24
,V.Smith
20,25
, and J. C. Wilson
3
1
Department of Astronomy, The Ohio State University, Columbus, OH 43210, USA; gail.zasowski@gmail.com
2
Center for Cosmology and Astro-Particle Physics, The Ohio State University, Columbus, OH 43210, USA
3
Department of Astronomy, University of Virginia, Charlottesville, VA 22904, USA
4
Department of Physics and Astronomy, Texas Christian University, Fort Worth, TX 76129, USA
5
Department of Astronomy, University of Michigan, Ann Arbor, MI 48109, USA
6
Laborat
´
orio Interinstitucional de e-Astronomia-LIneA, Rio de Janeiro, RJ 20921-400, Brazil
7
Observat
´
orio do Valongo, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ 20080-090, Brazil
8
Osservatorio Astronomico di Padova-INAF, I-35122 Padova, Italy
9
Yerkes Observatory, The University of Chicago, Williams Bay, WI 53191, USA
10
US Naval Observatory, Flagstaff Station, Flagstaff, AZ 86001, USA
11
Department of Astrophysical Sciences, Princeton University, Princeton, NJ 08544, USA
12
Lowell Observatory, Flagstaff, AZ 86001, USA
13
Department of Astronomy and Astrophysics, The Pennsylvania State University, University Park, PA 16802, USA
14
Center for Exoplanets and Habitable Worlds, The Pennsylvania State University, University Park, PA 16802, USA
15
Instituto de Astrof
´
ısica de Canarias, Calle V
´
ıa L
´
actea s/n, E-38205 La Laguna, Tenerife, Spain
16
Departamento de Astrof
´
ısica, Universidad de La Laguna, E-38206 La Laguna, Tenerife, Spain
17
Institut Utinam, CNRS UMR 6213, OSU THETA, Universit
´
e de Franche-Comt
´
e, F-25000 Besan¸con, France
18
Apache Point Observatory, Sunspot, NM 88349, USA
19
Steward Observatory, University of Arizona, Tucson, AZ 85721, USA
20
Observat
´
orio Nacional, Rio de Janeiro, RJ 20921-400, Brazil
21
Department of Astronomy, New Mexico State University, Las Cruces, NM 88003, USA
22
Instituto de F
´
ısica, UFRGS, RS 91501-970, Brazil
23
Astrophysics Research Institute, Liverpool John Moores University, Wirral, CH41 1LD, UK
24
McDonald Observatory, The University of Texas at Austin, Austin, TX 78712, USA
25
National Optical Astronomy Observatories, Tucson, AZ 85719, USA
Received 2013 May 4; accepted 2013 July 26; published 2013 August 22
ABSTRACT
The Apache Point Observatory Galactic Evolution Experiment (APOGEE) is a high-resolution infrared
spectroscopic survey spanning all Galactic environments (i.e., bulge, disk, and halo), with the principal goal
of constraining dynamical and chemical evolution models of the Milky Way. APOGEE takes advantage of the
reduced effects of extinction at infrared wavelengths to observe the inner Galaxy and bulge at an unprecedented
level of detail. The survey’s broad spatial and wavelength coverage enables users of APOGEE data to address
numerous Galactic structure and stellar populations issues. In this paper we describe the APOGEE targeting
scheme and document its various target classes to provide the necessary background and reference information to
analyze samples of APOGEE data with awareness of the imposed selection criteria and resulting sample properties.
APOGEE’s primary sample consists of ∼10
5
red giant stars, selected to minimize observational biases in age and
metallicity. We present the methodology and considerations that drive the selection of this sample and evaluate
the accuracy, efficiency, and caveats of the selection and sampling algorithms. We also describe additional target
classes that contribute to the APOGEE sample, including numerous ancillary science programs, and we outline the
targeting data that will be included in the public data releases.
Key words: Galaxy: abundances – Galaxy: kinematics and dynamics – stars: general – surveys
Online-only material: color figures
1. INTRODUCTION
The Apache Point Observatory Galactic Evolution Experi-
ment (APOGEE) is a near-infrared (H-band; 1.51–1.70 μm),
high-resolution (R ∼ 22,500), spectroscopic survey targeting
primarily red giant (RG) stars across all Galactic environments
(Majewski 2012; S. R. Majewski 2013, in preparation). The
spectrograph’s capability to produce 300 simultaneous spectra
26
NSF Astronomy and Astrophysics Postdoctoral Fellow.
is facilitated by many new technologies, such as a system for
coupling “warm” and cryogenically embedded fiber optic ca-
bles, a 30.5 × 50.8 cm volume phase holographic grating, and
a six-element cryogenic camera focusing light onto three Tele-
dyne H2RG detectors. See Wilson et al. (2012) and J. C. Wilson
et al. (in preparation) for details of the APOGEE hardware de-
sign and construction. APOGEE is part of the Sloan Digital Sky
Survey III (SDSS-III; Eisenstein et al. 2011), observing during
bright time on the 2.5 m Sloan telescope (Gunn et al. 2006)atthe
Apache Point Observatory (APO) in Sunspot, NM, USA. After a
1
The Astronomical Journal, 146:81 (28pp), 2013 October Zasowski et al.
commissioning phase spanning 2011 May–September, the
APOGEE survey officially commenced during the 2011
September observing run, and observations are expected to
continue until the end of SDSS-III in 2014 June.
The primary observational goal of the APOGEE survey is to
obtain precise and accurate radial velocities (RVs) and chemical
abundances for ∼10
5
RG stars spanning nearly all Galactic en-
vironments and populations. APOGEE targets comprise mostly
first-ascent red giant branch (RGB) stars, red clump (RC) stars,
and asymptotic giant branch (AGB) stars. This unprecedented
dataset will fulfill several major objectives. In particular, it will
1. constrain models of the chemical evolution of the Galaxy;
2. constrain kinematicalmodels of the bulge(s), bar(s),disk(s),
and halo(s) and discriminate substructures within these
components;
3. characterize the chemistry of kinematical substructures in
all Galactic components;
4. infer properties of the first generations of Milky Way (MW)
stars, through either direct detection of these first stars or
measurement of the chemical compositions of the most
metal-poor stars currently accessible;
5. observe the dust-enshrouded inner Galaxy and bring our
understanding of its chemistry and kinematics on par with
what is currently available for the solar neighborhood and
unobscured halo regions; and
6. provide a statistically significant stellar sample for further
investigations into the properties of subpopulations or
specific Galactic regions.
To achieve these objectives, the survey’s target selection
procedures strive to produce a homogeneous, minimally biased
sample of RG targets that is easily correctable to represent
the total underlying giant population in terms of age, chemical
abundances, and kinematics.
In this paper, we describe the motivation and technical aspects
behind the selection of APOGEE’s calibration and science tar-
get samples. Section 2 contains a summary of the overall survey
targeting philosophy, observing strategy, and target documen-
tation. Section 3 briefly describes the APOGEE field plan as it
pertains to target selection considerations, and Section 4 con-
tains the details of the base photometric catalog along with the
reddening corrections, color and magnitude limits, and magni-
tude sampling. In Section 5, we describe the calibration target
scheme adopted to aid in overcoming the challenges imposed by
telluric absorption and airglow on ground-based high-resolution
IR spectroscopy. In Section 6 we evaluate the accuracy and ef-
ficiency of our target selection algorithms based on data taken
during the survey’s first year. Sections 7 and 8 and Appendix C
contain descriptions of APOGEE’s “special” targets, such as
stellar clusters, stellar parameters calibrator targets, and ancil-
lary program targets. Finally, in Section 9, we list the targeting
and supplementary data that will be included along with the first
APOGEE data release in SDSS Data Release 10 (DR10). Read-
ers are strongly encouraged to refer to Appendix A, which con-
tains a glossary of SDSS- and APOGEE-specific terminology
that will be encountered in this paper, other APOGEE technical
and scientific papers, and the data releases.
2. SURVEY TARGETING AND
OBSERVATION STRATEGIES
RG stars are the most effective tracer population to target
for questions of large-scale Galactic structure, dynamics, and
chemistry because they are luminous, ubiquitous, and members
of stellar populations with a very wide range of age and
metallicity. Because they are luminous, they can be seen to
very great distances, allowing samples of populations far out in
the halo and across the disk, even beyond the bulge. Because
they are ubiquitous, we can observe large numbers of them
in all directions, allowing for statistically significant samples
even when divided into smaller subsamples by, e.g., Galactic
kinematical component or age. And because RG stars are
found in stellar populations of most ages and metallicities, we
can use them to measure quantitative differences across these
populations and trace their evolution in a Galactic context.
To minimize possible sample biases, the target selection must
be based as much as possible on the intrinsic property distribu-
tions of the stars selected. The observed photometry of stars is
determined by the intrinsic stellar properties (such as effective
temperature and metallicity) but is also affected by interstellar
extinction, which varies enormously within APOGEE’s foot-
print (spanning the Galactic Center to the North Galactic Cap;
Section 3). To mitigate these effects, the target selection in-
cludes reddening corrections. However, because APOGEE is
the first large survey of its type, and because we desire a sample
whose selection function is easy to determine, every effort has
been made to minimize the total number of selection criteria,
with particular attention to those that may potentially introduce
sample biases with respect to metallicity or age.
2.1. Overview of APOGEE Observations
In this section we present a brief description of APOGEE’s
observation scheme, as an introduction to some of the most
relevant SDSS-III/APOGEE-specific terminology. This discus-
sion will be considerably expanded in subsequent sections, and
Appendix A contains a glossary of terms for reference.
The survey uses standard SDSS plugplates, with holes for
300 APOGEE fibers; of these,
∼70 fibers are reserved for
telluric absorption calibrators and airglow emission calibration
positions (Sections 5.1–5.2), and the remaining ∼230 fibers
are placed on science targets. The patch of sky contained within
each plate’s field of view (FOV) is called a “field,” defined by its
central coordinates and angular diameter; the latter ranges from
1–3
◦
, depending on the field’s location in the sky (Section 3).
The base unit of observation for most purposes is a “visit,” which
corresponds to slightly more than 1 hr of detector integration
time.
27
The number of visits per field varies from one to ∼24, for
different types of fields (Section 3). Most APOGEE fields are
visited at least three times (excluding, e.g., the bulge fields;
Section 3.2) to permit detection of spectroscopic binaries in
the APOGEE sample. With typical RV variations of a few
km s
−1
or more, spectroscopic binaries can complicate the
interpretation of APOGEE’s kinematical results—for example,
by inflating velocity dispersions. In addition, given a bright
enough companion, the derived stellar parameters may be
influenced by the companion’s flux, so the detection of these
systems is very useful. Furthermore, fields with more visits can
have samples with fainter magnitude limits (Section 4.4) that
still meet the survey’s signal-to-noise (S/N) goal. Visits are
27
Visits comprise typically eight individual “exposures,” which are
approximately 8 minutes of integration each, taken at one of two ∼0.5 pixel
offset dither positions. Sub-pixel dithering in the spectral direction is required
because, at the native detector pixel size, the resolution element is
under-sampled in the bluer section of APOGEE spectra. These multiple
dithered exposures are combined by the data reduction pipeline to produce a
single “visit” spectrum (D. L. Nidever et al., in preparation).
2
The Astronomical Journal, 146:81 (28pp), 2013 October Zasowski et al.
Figure 1. Organization of observed targets in plate designs and on physical plates, using the field 180+04 as an example. This field has 12 anticipated visits, which
are covered by four designs (indicated by blue, yellow, green, and orange). Each design has stars from one of four short cohorts (S1, S2, S3, S4), one of two medium
cohorts (M1, M2), and the long cohort (L); that is, stars in the long cohort appear in all four designs, and stars from the medium cohorts appear in two designs. At least
one plate is drilled for each design, and some designs (here, the first two) are drilled on multiple plates. Most frequently, this occurs when a field is to be observed at
different hour angles (HAs), as in this example.
(A color version of this figure is available in the online journal.)
separated by at least one night and may be separated by more
than a year, depending on the given field’s observability and
priority relative to others at similar right ascensions (R.A.s).
Different stars may be observed on different visits to a field.
Stars are grouped into sets called “cohorts,” based on their
H-band apparent magnitude, and each cohort is observed for
only as many visits (generally in multiples of three) as needed
for all stars in the cohort to achieve the final desired S/N. For
example, the brightest candidate targets in a given 12-visit field
may only need three visits to reach this goal, whereas stars 1 mag
fainter need all 12 visits to reach the same S/N. Observing
the bright stars for all 12 visits would be an inefficient use of
observing time, so a cohort composed of these stars is only
observed three times, and then replaced with another cohort of
different bright stars, while a cohort composed of the fainter
stars is observed on all 12 visits to the field. Thus, by grouping
together cohorts with different magnitude ranges on a series of
plates, we increase the number of total stars observed without
sacrificing starsat the faint endof the APOGEE magnitude range
(see additional details on the cohort scheme in Section 4.4).
A particular combination of cohorts (equivalently, a particular
combination of stars) defines a “design,” with a unique ID
number; a given cohort may appear on a single or on multiple
designs. See Figure 1 for an example. Each physically unique
aluminum “plate” is drilled with a single design, but a given
design may appear on multiple plates—for example, if a new
plate is drilled for observing the same stars at a different hour
angle. Thus a field (a location on the sky) may have multiple
designs (sets of targets), and each design may have multiple
plates, but a plate has only one design, and a design is associated
with only one field. We anticipate ∼650 designs to be made over
the course of the survey for the approximately 450 distinct fields
(Section 3).
2.2. Targeting Flags
Reconstruction of the target selection function, however sim-
ple it may be, is crucial for understanding how well the spec-
troscopic target sample represents the underlying population in
the field. To track the various factors considered in each target’s
selection and prioritization, APOGEE has defined two 32-bit
integers, apogee_target1 and apogee_target2, whose bits corre-
spond to specific target selection criteria (Table 1). Every target
in a given design is assigned one of each of these integers, also
called “targeting flags” (Appendix A), with one or more bits
“set” to indicate criteria that were applied to place a target on a
design.
These flags indicate selection criteria for a given design,or
particular set of stars (Appendix A), and thus may differ for the
same star on different designs and plates. For example, many
commissioning plates were observed without a dereddened-
color limit (Section 4.3), so a bit used to indicate that a target was
selected because of its dereddened color (e.g., apogee_target1 =
3, “dereddened with RJCE/IRAC”) would not be set for those
observations; however, if later designs drilled for that same field
do have a color limit, and the same stars are re-selected and
observed, that bit would be set for those later observations of
the same stars.
Throughout this paper, we will use the notation apogee_
target1 = X to indicate that bit “X” is set in the apogee_target1
flag (and likewise for apogee_target2), even though mathemati-
cally, that bit is set by assigning apogee_target1 = 2
X
. Because
a target may have multiple (N) bits set, its final integer flag value
is a summation of all set bits:
N
i=0
2
bit(i)
.
3
The Astronomical Journal, 146:81 (28pp), 2013 October Zasowski et al.
Figure 2. Map of the APOGEE field plan. The map is in Galactic coordinates, with the Galactic Center in the middle, the anti-center (l = 180
◦
) on the left and right,
and the North/South Galactic Caps at the top and bottom, respectively. The lines of Galactic latitude are labeled on the left, and the solid gray lines indicate Galactic
longitudes (from left to right) l = 180
◦
, 120
◦
, 60
◦
, 0
◦
, 300
◦
, 240
◦
, and 180
◦
. The gray shaded area indicates those regions of the Galaxy that are never visible with an
airmass 2.3 from APO. See text for description of the field types.
(A color version of this figure is available in the online journal.)
Tab le 1
Targeting Flags
apogee_target1 apogee_target2
Selection Criterion Bit Selection Criterion Bit
—0—0
— 1 Flux standard 1
— 2 Abundance/parameters standard 2
Dereddened with RJCE/IRAC 3 Stellar RV standard 3
Dereddened with RJCE/WISE 4Skytarget 4
Dereddened with SFD E(B − V )5— 5
No dereddening 6 — 6
Washington+DDO51 giant 7 — 7
Washington+DDO51 dwarf 8 — 8
Probable (open) cluster member 9 Telluric calibrator 9
Extended object 10 Calibration cluster member 10
Short cohort (1–3 visits) 11 Galactic Center giant 11
Medium cohort (3–6 visits) 12 Galactic Center supergiant 12
Long cohort (12–24 visits) 13 — Young Embedded Clusters 13
—14—MW Long Bar 14
—15—B[e] Stars 15
“First Light” cluster target 16 — Cool Kepler Dwarfs 16
Ancillary program target 17 — Outer Disk Clusters 17
— M31 Globular Clusters 18 — 18
— MDwarfs 19 — 19
— Stars with High-R Optical Spectra 20 — 20
— Oldest Stars 21 — 21
— Kepler and CoRoT Ages 22 — 22
— Eclipsing Binaries 23 — 23
— Pal 1 GC 24 — 24
— Massive Stars 25 — 25
Sgr dSph member 26 — 26
Kepler asteroseismology target 27 — 27
Kepler planet-host target 28 — 28
“Faint” target 29 — 29
SEGUE sample overlap 30 — 30
Notes. Bits 13–17 in apogee_target2 also refer to ancillary programs. Bits with “—”
as their criterion have either yet to be defined or were reserved for criteria never
applied to released data.
In keeping with earlier SDSS conventions, if any bit in
apogee_target1 or apogee_target2 is set, bit 31 for that flag
is also set. For example, a well-studied star that is targeted as a
stellar chemical abundance standard (apogee_target2 = 2) and
Tab le 2
Field Plan Summary
Type Definition Approx. Survey Target Fraction
Disk 24
◦
l 240
◦
, |b| 16
◦
50%
Bulge 357
◦
l 22
◦
, |b| 8
◦
10%
Halo |b| > 16
◦
25%
Special Placed on calibration/ancillary sources 15%
also as a member of a calibration cluster (apogee_target2 =
10; see Section 7.1) would have a final 32-bit integer flag of
apogee_target2 = 2
2
+2
10
+2
31
=−2147482620 (the negative
sign is a result of the fact that these are signed integers).
3. APOGEE FIELD PLAN
We provide here a summary of the current APOGEE field
locations as they pertain to target selection considerations and
procedures; see S. R. Majewski et al. (2013, in preparation)
for a full discussion of the plan’s motivation and details. The
APOGEE survey footprint spans as wide a range of the Galaxyas
is visible from the Apache Point Observatory (latitude = 32.8
◦
N), and samples all major Galactic components. Figure 2 shows
the current complement of chosen field centers (summarized in
Table 2). “Disk” fields (Section 3.1) are in dark blue circles,
“bulge” fields (Section 3.2) are in light blue point-up triangles,
and “halo” fields (Section 3.3) are in green point-down triangles.
In addition to these primary classifications, the field plan
includes pointings covering the footprint of NASA’s Kepler
mission (yellow diamonds), well-studied open and globular
clusters (orange squares), and the Sagittarius dwarf galaxy core
and tails (red quartered squares), as described in Sections 7–8.
Most fields are named using the Galactic longitude l and
latitude b of their center (i.e., “lll±bb”), though we note
that these centers are approximate in many cases, and the
exact coordinates should be obtained from the database if
field position accuracy 0.
◦
5 is required. A subset of fields,
particularly in the halo, are named for an important object or
objects they contain, such as specific stellar clusters or stellar
streams (e.g., the Sagittarius tidal streams; Section 8.2). In these
cases, the fields are deliberately not centered on the object,
because the SDSS plates have a 5 arcmin hole in the center (used
to attach the plate to the fiber cartridges) that precludes any fiber
4
The Astronomical Journal, 146:81 (28pp), 2013 October Zasowski et al.
holes being placed there. Throughout this paper, we will use
italics when referring to all field names, to remove ambiguity
between general discussion of targeting in a field named after
a specific object and targeting in the object itself (e.g., the
APOGEE field pointing M13 versus the globular cluster M13).
3.1. Disk
The subset of APOGEE fields termed “disk” fields form a
semi-regular grid spanning 24
◦
l 240
◦
, with |b| 16
◦
.
Each of these fields will be visited from 3 to 24 times, meaning
that their nominal faint magnitude limits range from H = 12.2 to
13.8 (Section 4.4). For the 3-visit fields, all stars in the selected
sample will be observed on all three visits, while the fields with
>3 visits employ the cohort scheme described in Sections 2.1
and 4.4 to balance the desires for dynamic range, survey depth,
and good statistics.
All stars in the disk grid fields are selected based on their
dereddened (J − K
s
)
0
colors (Section 4.3). Simulations of the
survey estimate that approximately 50% of the final survey
stellar sample will come from the disk fields (Table 2). In
addition to the normal APOGEE sample and a variety of
ancillary targets (Appendix C), the disk fields contain open
clusters (Section 7.2) falling serendipitously in the survey
footprint.
3.2. Bulge
The set of fields considered “bulge” fields are those spanning
357
◦
l 22
◦
and |b| 8
◦
(plus fields centered on the
Sagittarius dwarf galaxy, Section 8.2). Due to the low altitude
of these fields at APO,
28
and the strong differential atmospheric
refraction that results from observing at such high airmasses, the
bulge fields are restricted to a 1
◦
–2
◦
diameter FOV, compared
to the full 3
◦
diameter for the majority of the survey fields.
The density of target candidates meeting APOGEE’s selection
criteria is so high (up to ∼7500 deg
−2
), however, that even with
the restricted FOV, there are ample stars from which to choose in
these fields. Stars in the bulge fields are selected based on their
dereddened (J −K
s
)
0
color, and approximately 10% of the final
survey sample is projected to come from the bulge fields.
The R.A. range of the bulge also includes many of the closely
packed inner disk fields. Because of this R.A. oversubscription
and the small window during which the low-declination bulge
can be observed on any given night, the majority of the bulge
fields are only visited once, instead of the 3 visits anticipated
for all other fields. The few multi-visit exceptions include high-
priority calibration fields, such as the Galactic Center, Baade’s
Window, and those that overlap fields from other surveys
(such as BRAVA; Rich et al. 2007). While APOGEE cannot
distinguish single-lined spectroscopic binaries in the 1-visit
fields, it is worth noting that the magnitude limit for these fields
(H 11.0) is still faint enough to include RGB stars in the
bulge behind A(V ) 25 mag of extinction.
Special targets in the bulge fields include nearly 200 bulge
giants and supergiants, already studied with high-resolution
optical or IR spectroscopy (Section 8.1). These targets are useful
for calibrating APOGEE’s stellar abundance and parameters
pipeline, particularly at high metallicity.
3.3. Halo
APOGEE’s “halo” fields are defined as those with |b| > 16
◦
,
and in practice all have |b| 18
◦
. The stellar population
28
For example, the Galactic Center transits the meridian at an altitude of 28
◦
.
distribution in these fields is often substantially different from
those of the disk and bulge. For example, the dwarf-to-giant
ratio within APOGEE’s nominal color and magnitude range
is much higher in the halo fields, due to the overall lower
density of distant giants (see Section 4.3). To improve the
selection efficiency of giants, we have acquired additional
photometry in the optical Washington M & T
2
and DDO51
filters (hereafter, “Washington+DDO51”; Canterna 1976;Clark
&McClure1979;Majewskietal.2000)for∼90% of the halo
fields, to assist with identifying and prioritizing giant and dwarf
candidates. See Section 4.2 for details on the acquisition and
reduction of these data.
The paucity of targets in certain halo fields (compared with
APOGEE’s capability to observe 230 simultaneous science
targets) requires some special accommodations when selecting
targets. One of these is the deliberate targeting of dwarf stars in
fields lacking sufficient bright giants (Section 4.2), and another
is the inclusion of targets with H magnitudes up to 0.8 mag
fainter than the nominal limits for the fields. These “faint”
targets, which are not expected to attain a final S/N 100,
have bit apogee_target1 = 29 set and are described more
fully in Section 7.1. In addition, many of the halo fields are
placed on open or globular clusters with well-knownabundances
(Section 7.1), and members of these clusters can comprise up to
75% of all targets in their field.
Approximately 25% of the final survey sample is estimated
to come from the halo fields. These survey sample percentages
from the different field types do not include the ∼15% coming
from the “calibration” or other special fields, which include
the 3-visit bulge fields, the long 12–24-visit halo cluster fields
(Section 7.1), and the “APOGEE–Kepler” fields (Section 8.3).
4. PHOTOMETRIC TARGET SELECTION
CRITERIA AND PROCEDURES
4.1. Base Photometric Catalogs and Quality Requirements
The Two Micron All Sky Survey (2MASS) Point Source
Catalog (PSC; Skrutskie et al. 2006) forms the base catalog
for the targeted sample. The use of 2MASS confers several
advantages. (1) The need to construct a photometric pre-
selection catalog of our own is eliminated. (2) The all-sky
coverage allows us to draw potential targets from a well-tested,
homogeneous catalog for every field in the survey. (3) Even
in the most crowded bulge fields, where, due to confusion, the
magnitude limit of the PSC is brighter than in other parts of
the Galaxy,
29
the PSC is deep enough for APOGEE’s nominal
magnitude limits. (4) The wavelength coverage is well-matched
to APOGEE, and we can select targets based directly on
their H-band (λ
eff
= 1.66 μm) magnitude. (5) The astrometric
calibration for stars within APOGEE’s magnitude range is
sufficiently accurate (on the order of ∼75 mas
29
) for positioning
fiber holes in the APOGEE plugplates, even in closely packed
cluster fields. Furthermore, the PSC contains merged multi-
wavelength photometry (the J- and K
s
-bands, with λ
eff
= 1.24
and 2.16 μm, respectively) useful for characterizing stars (e.g.,
with photometric temperatures), as well as detailed data and
reduction quality flags for each band.
We combine the 2MASS photometry with mid-IR
data to calculate the extinction for each potential stellar
target (Section 4.3). Where available, we use data from
the Spitzer-IRAC Galactic Legacy Infrared Mid-Plane Survey
29
http://www.ipac.caltech.edu/2mass/releases/allsky/doc/sec2_2.html
5
![Figure 10. Comparison between NIR colors and ASPCAP Teff values for stars observed in 13 bulge, disk, and halo fields during APOGEE commissioning and Year 1. (a) Uncorrected (J − Ks ) colors of the stars. Because this sample comprises fields probing both the heavily reddened bulge (down to l, b = 0◦, 0◦) and the low-reddening halo (up to |b| = 56◦), the range of observed colors is broad. The solid red line is the locus of color–temperature spanned by giant star isochrones (−1.5 [Fe/H] +0.2; Girardi et al. 2002). The typical color and Teff uncertainties are shown by the errorbars in the upper right corner. (b) Almost identical to (a), except for RJCE-corrected (J − Ks )0 colors instead of observed (J − Ks ). The representative color uncertainty now includes uncertainties from dereddening. The dashed red lines to either side of the solid line represent the ∼1σ range for which we consider stars to be in good agreement with the theoretical color–temperature relationship (see text). (c) Distribution of [Fe/H] values for stars not in the range of good agreement indicated in (b), scaled bin-by-bin by the distribution of [Fe/H] for all stars in this sample. (d) Similar to (c) but for stellar log g values. (e) Similar to (c) but for stellar Teff values.](/figures/figure-10-comparison-between-nir-colors-and-aspcap-teff-bwpbo501.webp)