The Relationship Between Stellar Populations and Lyman Alpha Emission in Lyman Break Galaxies

Abstract

We present the results of a photometric and spectroscopic survey of 321 Lyman break galaxies (LBGs) at z ~ 3 to investigate systematically the relationship between Lya emission and stellar populations. Lya equivalent widths (EW) were calculated from rest-frame UV spectroscopy and optical/near-infrared/Spitzer photometry was used in population synthesis modeling to derive the key properties of age, dust extinction, star formation rate (SFR), and stellar mass. We directly compare the stellar populations of LBGs with and without strong Lya emission, where we designate the former group (EW > 20 AA) as Lya emitters (LAEs) and the latter group (EW < 20 AA) as non-LAEs. This controlled method of comparing objects from the same UV luminosity distribution represents an improvement over previous studies in which the stellar populations of LBGs and narrowband-selected LAEs were contrasted, where the latter were often intrinsically fainter in broadband filters by an order of magnitude simply due to different selection criteria. Using a variety of statistical tests, we find that Lya equivalent width and age, SFR, and dust extinction, respectively, are significantly correlated in the sense that objects with strong Lya emission also tend to be older, lower in star formation rate, and less dusty than objects with weak Lya emission, or the line in absorption. We accordingly conclude that, within the LBG sample, objects with strong Lya emission represent a later stage of galaxy evolution in which supernovae-induced outflows have reduced the dust covering fraction. We also examined the hypothesis that the attenuation of Lya photons is lower than that of the continuum, as proposed by some, but found no evidence to support this picture. Comment: 23 pages, 13 figures, accepted to ApJ, replaced with accepted version

Full text

Page 1

arXiv:0911.2000v2 [astro-ph.CO] 19 Jan 2010
Draft version January 19, 2010
Preprint typeset using LATEX style emulateapj v. 11/26/04
THE RELATIONSHIP BETWEEN STELLAR POPULATIONS AND Lyα EMISSION IN LYMAN BREAK
GALAXIES1
Katherine A. Kornei and Alice E. Shapley2,3
Department of Physics and Astronomy, 430 Portola Plaza, University of California at Los Angeles, Los Angeles, CA 90025, USA
Dawn K. Erb4
Department of Physics, University of California at Santa Barbara, Santa Barbara, CA 93106, USA
Charles C. Steidel
California Institute of Technology, MS 105-24, Pasadena, CA 91125, USA
Naveen A. Reddy5
National Optical Astronomy Observatory, 950 N. Cherry Ave, Tucson, AZ 85719, USA
Max Pettini
Institute of Astronomy, Madingley Road, Cambridge, CB3 0HA, UK
Milan Bogosavljevi´ c
California Institute of Technology, MS 105-24, Pasadena, CA 91125, USA
Draft version January 19, 2010
ABSTRACT
We present the results of a photometric and spectroscopic survey of 321 Lyman break galaxies
(LBGs) at z ∼ 3 to investigate systematically the relationship between Lyα emission and stellar
populations. Lyα equivalent widths (WLyα) were calculated from rest-frame UV spectroscopy and
optical/near-infrared/Spitzer photometry was used in population synthesis modeling to derive the key
properties of age, dust extinction, star formation rate (SFR), and stellar mass. We directly compare
the stellar populations of LBGs with and without strong Lyα emission, where we designate the former
group (WLyα ≥ 20˚ A) as Lyα-emitters (LAEs) and the latter group (WLyα < 20˚ A) as non-LAEs.
This controlled method of comparing objects from the same UV luminosity distribution represents an
improvement over previous studies in which the stellar populations of LBGs and narrowband-selected
LAEs were contrasted, where the latter were often intrinsically fainter in broadband filters by an order
of magnitude simply due to different selection criteria. Using a variety of statistical tests, we find that
Lyα equivalent width and age, SFR, and dust extinction, respectively, are significantly correlated in
the sense that objects with strong Lyα emission also tend to be older, lower in star formation rate, and
less dusty than objects with weak Lyα emission, or the line in absorption. We accordingly conclude
that, within the LBG sample, objects with strong Lyα emission represent a later stage of galaxy
evolution in which supernovae-induced outflows have reduced the dust covering fraction. We also
examined the hypothesis that the attenuation of Lyα photons is lower than that of the continuum, as
proposed by some, but found no evidence to support this picture.
Subject headings: galaxies: high-redshift — galaxies: evolution — galaxies: starburst
1. INTRODUCTION
An increasing number of high-redshift galaxies have
been found in the last two decades using selection tech-
niques reliant on either color cuts around the Lyman
limit at 912˚ A in the rest frame (e.g., Steidel et al. 1996,
1999) or strong Lyα line emission (e.g., Cowie & Hu
1Based, in part, on data obtained at the W.M. Keck Observa-
tory, which is operated as a scientific partnership among the Cal-
ifornia Institute of Technology, the University of California, and
NASA, and was made possible by the generous financial support
of the W.M. Keck Foundation.
2Packard Fellow.
3Alfred P. Sloan Fellow.
4Spitzer Fellow.
5Hubble Fellow.
1998; Rhoads et al. 2000; Gawiser et al. 2006). These
two methods, which preferentially select Lyman break
galaxies (LBGs) and Lyα-emitters (LAEs), respectively,
have successfully isolated galaxies at redshifts up to z =
7 (Iye et al. 2006; Bouwens et al. 2008). Extensive data
sets of LBGs and LAEs have afforded detailed studies
of galactic clustering (e.g., Adelberger et al. 1998; Gi-
avalisco & Dickinson 2001), the universal star formation
history (e.g., Madau et al. 1996; Steidel et al. 1999), and
the galaxy luminosity function (e.g., Reddy et al. 2008;
McLure et al. 2009). While the nature of LBGs and
LAEs have been studied at a range of redshifts (e.g.,
Shapley et al. 2001; Gawiser et al. 2006; Verma et al.
2007; Pentericci et al. 2007; Nilsson et al. 2009b; Ouchi

Page 2

2
et al. 2008; Finkelstein et al. 2009), the epoch around z ∼
3 is particularly well-suited to investigation of these ob-
jects’ detailed physical properties. At this redshift, the
prominent HI Lyα line (λrest = 1216˚ A), present in all
LAE spectra and a significant fraction of LBG spectra, is
shifted into the optical, where current imaging and spec-
troscopic instrumentation is optimized.
there are large existing data sets of spectroscopically-
confirmed z ∼ 3 LBGs (e.g., Steidel et al. 2003) and
LAEs (e.g., Lai et al. 2008), where extensive multiwave-
length surveys often complement the former and, less
frequently, the latter.
The mechanism responsible for LAEs’ large Lyα equiv-
alent widths is not fully understood, although several
physical pictures have been proposed (e.g., Dayal et al.
2009; Kobayashi et al. 2010). As Lyα emission is eas-
ily quenched by dust, one explanation for LAEs is that
they are young, chemically pristine galaxies experiencing
their initial bursts of star formation (e.g., Hu & McMa-
hon 1996; Nilsson et al. 2007). Conversely, LAEs have
also been proposed to be older, more evolved galaxies
with interstellar media in which dust is segregated to
lie in clumps of neutral hydrogen surrounded by a ten-
uous, ionized dust-free medium (Neufeld 1991; Hansen
& Oh 2006; Finkelstein et al. 2009).
Lyα photons are resonantly scattered near the surface
of these dusty clouds and rarely encounter dust grains.
Continuum photons, on the other hand, readily pene-
trate through the dusty clouds and are accordingly scat-
tered or absorbed. This scenario preferentially attenu-
ates continuum photons and enables resonant Lyα pho-
tons to escape relatively unimpeded, producing a larger
Lyα equivalent width than expected given the underly-
ing stellar population. To date, the distribution of dust
in the interstellar medium has only been investigated us-
ing relatively small samples (e.g., Verhamme et al. 2008;
Atek et al. 2009; Finkelstein et al. 2009).
Given the different selection techniques used to isolate
LBGs and LAEs, understanding the relationship between
the stellar populations of these objects has been an im-
portant goal of extragalactic research. Recent work by
Gawiser et al. (2006) has suggested that LAEs are less
massive and less dusty than LBGs, prompting these au-
thors to propose that LAEs may represent the beginning
of an evolutionary sequence in which galaxies increase in
mass and dust content through successive mergers and
star formation episodes (Gawiser et al. 2007). The high
specific star formation rate – defined as star formation
rate (SFR) per unit mass – of LAEs (∼ 7 × 10−9yr−1;
Lai et al. 2008) relative to LBGs (∼ 3 × 10−9yr−1; Shap-
ley et al. 2001) illustrates that LAEs are building up stel-
lar mass at a rate exceeding that of continuum-selected
galaxies at z ∼ 3. This rapid growth in mass is consis-
tent with the idea that LAEs represent the beginning of
an evolutionary sequence of galaxy formation. However,
results from Finkelstein et al. (2009) cast doubt on this
simple picture of LAEs as primordial objects, given that
these authors find a range of dust extinctions (A1200 =
0.30–4.50) in a sample of 14 LAEs at z ∼ 4.5. Nils-
son et al. (2009b) also find that z ∼ 2.25 LAEs occupy
a wide swath of color space, additional evidence that
not all LAEs are young, dust-free objects. Furthermore,
the assertion that LAEs are pristine galaxies undergoing
their first burst of star formation is called into question
Consequently,
In this picture,
by the results of Lai et al. (2008). These authors present
a sample of 70 z ∼ 3.1 LAEs, ∼ 30% of which are de-
tected in the 3.6 µm band of the Spitzer Infrared Array
Camera (IRAC; Fazio et al. 2004). These IRAC-detected
LAEs are significantly older and more massive (?t⋆? ∼ 1.6
Gyr, ?M? ∼ 9 × 109) than the IRAC-undetected sam-
ple (?t⋆? ∼ 200 Myr, ?M? ∼ 3 × 108M⊙); Lai et al.
(2008) suggest that the IRAC-detected LAEs may there-
fore be a lower-mass extension of the LBG population.
Narrowband-selected LAEs are clearly marked by hetero-
geneity, and the relationship between these objects and
LBGs continues to motivate new studies.
When comparing the stellar populations of LBGs and
LAEs, it is important to take into account the selection
biases that result from isolating these objects with broad-
band color cuts and line flux/equivalent width require-
ments, respectively.By virtue of selection techniques
that rely on broadband fluxes and colors, LBGs generally
have brighter continua than LAEs. Spectroscopic sam-
ples of LBGs typically have an apparent magnitude limit
of R ≤ 25.5 (0.4L∗at z ∼ 3; Steidel et al. 2003) while
LAEs have a median apparent magnitude of R ∼ 27 (Ga-
wiser et al. 2006), where R and R magnitudes are compa-
rable. Even though the majority of LBGs studied to date
are an order of magnitude more luminous in the contin-
uum than typical LAEs, both populations have similar
rest-frame UV colors (Gronwall et al. 2007). Therefore,
LAEs fainter than R = 25.5 are excluded from LBG spec-
troscopic surveys not because of their colors, but rather
because of their continuum faintness. Given the signif-
icant discrepancy in absolute magnitude between LBGs
and LAEs, understanding the relationship between these
objects can be fraught with bias. An preferable approach
to comparing these populations is to investigate how the
strength of Lyα emission is correlated with galaxy pa-
rameters, for a controlled sample of objects at similar
redshifts drawn from the same parent UV luminosity dis-
tribution.
Several authors have looked at the question of the ori-
gin of Lyα emission in UV flux-limited samples (e.g.,
Shapley et al. 2001; Erb et al. 2006a; Reddy et al. 2008;
Pentericci et al. 2007; Verma et al. 2007). Shapley et al.
(2001) analyzed 74 LBGs at z ∼ 3 and constructed
rest-frame UV composite spectra from two samples of
“young” (t⋆ ≤ 35 Myr) and “old” (t⋆ ≥ 1 Gyr) galax-
ies, respectively. These authors found that younger ob-
jects exhibited weaker Lyα emission than older galax-
ies; Shapley et al. (2001) attributed the difference in
emission strength to younger LBGs being significantly
dustier than their more evolved counterparts. On the
other hand, Erb et al. (2006a) examined a sample of 87
star-forming galaxies at z ∼ 2 and found that objects
with lower stellar mass had stronger Lyα emission fea-
tures, on average, than more massive objects. In a sam-
ple of 139 UV-selected galaxies at z ∼ 2–3, Reddy et al.
(2008) isolated 14 objects with Lyα equivalent widths ≥
20˚ A and noted no significant difference in the stellar pop-
ulations of strong Lyα-emitters relative to the rest of the
sample. Pentericci et al. (2007) examined 47 LBGs at z ∼
4 and found that younger galaxies generally showed Lyα
in emission while Lyα in absorption was associated with
older galaxies (in contrast to the Shapley et al. (2003) re-
sults). Probing even earlier epochs, Verma et al. (2007)

Page 3

3
Fig. 1.— Redshift distribution of the sample, where ?z? = 2.99
± 0.19.
examined a sample of 21 LBGs at z ∼ 5 and found no
correlation between Lyα equivalent width and age, stel-
lar mass, or SFR. These authors noted, however, that
only 6/21 of the brightest LBGs had corresponding spec-
troscopy from which equivalent widths were estimated.
Therefore, the lack of a correlation between Lyα equiva-
lent width and stellar populations may, in this case, have
been masked by a small sample that was biased towards
the brightest objects. These aforementioned investiga-
tions have shown that there does not yet exist a clear
picture relating stellar populations to Lyα emission.
In this paper, we present a precise, systematic inves-
tigation of the relationship between Lyα emission and
stellar populations using our large photometric and spec-
troscopic data set of z ∼ 3 observations. As an improve-
ment over previous studies, we approach the data analy-
sis from multiple aspects: we compare not only the stel-
lar population parameters derived from population syn-
thesis modeling, but also examine the objects’ best-fit
SEDs and photometry. Furthermore, all analysis is con-
ducted on objects drawn from the same parent sample of
continuum-bright (R ≤ 25.5) LBGs. By controlling for
continuum magnitude, we avoid the biases of comparing
objects with significantly different luminosities while still
retaining the ability to comment on the nature of strong
Lyα-emitting galaxies within the LBG sample. Our con-
clusions are applicable to both LBGs and bright (R ≤
25.5) narrowband-selected LAEs. While we are unable
to make inferences about the population of faint LAEs,
our study is complete with respect to bright LAEs given
these objects’ similar colors to LBGs in the rest-frame
UV.
We are motivated by the following questions: how do
the stellar populations of Lyα-emitting LBGs differ from
those of other LBGs at z ∼ 3 where the Lyα emission
line is weaker (or absent altogether)? To what degree
are galactic parameters such as dust extinction, SFR,
age, and stellar mass correlated with Lyα line strength?
What do the relative escape fractions of Lyα and contin-
uum photons reveal about the distributions of gas and
dust in these objects’ interstellar media?
This paper is organized as follows. In §2, we present
details of the observations and data reduction, including
a description of the systematic technique used to calcu-
late Lyα equivalent widths. Stellar population modeling
is discussed in §3. The properties of objects with and
without strong Lyα emission are presented in §4 and we
discuss how our data can be used to address several of the
outstanding questions pertaining to the physical nature
of LBGs and LAEs in §5. A summary and our conclu-
sions appear in §6. We assume a standard ΛCDM cos-
mology throughout with H0= 70 km s−1Mpc−1, ΩM=
0.3, and ΩΛ= 0.7. All magnitudes are based on the AB
system (Oke & Gunn 1983)6, with the exception of the
infrared passbands which are in the Vega system.
2. OBSERVATIONS AND DATA REDUCTION
2.1. Imaging and Spectroscopy
The data presented here are drawn from the LBG sur-
veys of Steidel and collaborators, with approximately
half of the observations described in Steidel et al. (2003,
2004) and half from subsequent programs by the same
authors. These surveys employed photometric preselec-
tion in the UnGR passbands in a variety of fields (Reddy
et al. 2008) to target galaxies in the redshift interval z
∼ 2–3.
these galaxies, paired with supplemental near and mid-
infrared photometry, has yielded an extensive data set
upon which multiple studies have been based (e.g., Shap-
ley et al. 2003; Adelberger et al. 2005a; Shapley et al.
2005; Erb et al. 2006b; Reddy et al. 2008).
Here, we introduce a spectroscopic and photometric
sample of z ∼ 3 LBGs. These data were photometrically
preselected with the following standard LBG UnGR flux
and color cuts:
Follow-up optical spectroscopy of a subset of
R ≤ 25.5,
where the Un, G, and R passbands sample λrest∼ 900,
1200, and 1700˚ A at z ∼ 3, respectively. Object detec-
tion, color cuts, and photometry are discussed in Steidel
et al. (2003). Multi-object optical spectroscopy was ob-
tained using the Low Resolution Imaging Spectrometer
(LRIS; Oke et al. 1995) on the Keck I 10m telescope. The
majority of the data (93%) were taken with the blue arm
of LRIS (LRIS-B; McCarthy et al. 1998; Steidel et al.
2004), and the remainder of the data were obtained with
LRIS prior to its blue arm upgrade in September 2000.
The LRIS-B data were collected using 300, 400, and 600
line mm−1grisms, which resulted in spectral resolutions
of λ/∆λ = 1000, 1200, and 2000, respectively. The 400
(600) line mm−1grism was used for 55% (39%) of the
observations, and the remaining ∼ 6% of the LRIS-B
spectra were obtained with the 300 line mm−1grism.
LRIS-B rest-frame wavelength coverage extended from ∼
900–1500˚ A and a typical integration time was 3 × 1800 s.
The data were reduced (flat-fielded, cosmic ray rejected,
background subtracted, extracted, wavelength and flux
calibrated, and transformed to the vacuum wavelength
frame) using IRAF scripts. Details of the data collection
G − R ≤ 1.2,Un− G ≥ G − R + 1 (1)
6AB magnitude and fν, the flux density in units of ergs s−1
cm−2Hz−1, are related by mAB = −2.5 log10 fν−48.6.
versions between AB and Vega magnitudes for the near-infrared
passbands are as follows: Ks(AB) = Ks(Vega) + 1.82; J(AB) =
J(Vega) + 0.90.
Con-

Page 4

4
TABLE 1
Spectroscopic Survey Fields
Field Name
αa
δb
Field Size
(arcmin2)
NLBGc
(J2000.0)(J2000.0)
Q0100⋆
Q0142
Q0449
Q1009⋆
Q1217
GOODS-Nd⋆†
Q1307
Q1549⋆†
Q1623⋆†
Q1700⋆†
Q2206⋆
Q2343⋆†
Q2346
01 03 11
01 45 17
04 52 14
10 11 54
12 19 31
12 36 51
13 07 45
15 51 52
16 25 45
17 01 01
22 08 53
23 46 05
23 48 23
13 16 18
–09 45 09
–16 40 12
29 41 34
49 40 50
62 13 14
29 12 51
19 11 03
26 47 23
06 11 58
–19 44 10
12 49 12
00 27 15
42.9
40.1
32.1
38.3
35.3
155.3
258.7
37.3
290.0
235.3
40.5
212.8
280.3
22
20
13
30
13
54
8
48
24
39
23
26
1
TOTAL... ... 1698.9321
aRight ascension in hours, minutes, and seconds.
bDeclination in degrees, arcminutes, and arcseconds.
cNumber of spectroscopically-confirmed LBGs with redshifts
z ≥ 2.7, excluding QSOs and AGN.
dThis field is also referred to as “HDF”.
⋆Denotes a field with near-infrared imaging.
†Denotes a field with mid-infrared Spitzer IRAC imaging.
and reduction of both the preselection and spectroscopic
samples are presented in Steidel et al. (2003).
Approximately 3% of the spectroscopically-confirmed
z ∼ 3 LBGs were classified as either active galactic nu-
clei (AGN) or quasi-stellar objects (QSOs) on the basis
of broad lines and high-ionization emission features, re-
spectively (Reddy et al. 2008). These objects were ex-
cluded from the spectroscopic sample, as were galaxies at
redshifts z ≤ 2.7. The final sample, spanning 13 photo-
metric preselection fields totaling 1700 arcmin2, includes
321 objects with an average redshift of ?z? = 2.99 ± 0.19
(Table 1, Figure 1). We note that this sample is distinct
from previous studies of z ∼ 3 LBGs (e.g., Shapley et al.
2001, 2003) in that the majority of these objects have
corresponding near- and mid-infrared photometry.
Near-infrared photometry in the J (λc= 1.25 µm) and
Ks(λc = 2.15 µm) bands was obtained for a subset of
the sample (8/13 fields) using the Wide Field Infrared
Camera (Wilson et al. 2003) on the Palomar 5m tele-
scope. 102/321 objects (32%) were detected in Ksimag-
ing and an additional 69 objects fell on the Ks images
and were not detected. We assigned Ksupper limits cor-
responding to 3σ image depths (Ks∼ 22.2 (Vega); Erb
et al. 2006b) to these 69 galaxies. J band photometry
was also obtained for 57/102 objects (56%) detected in
the Kssample. Details of the data collection and reduc-
tion of the near-infrared sample are presented in Shapley
et al. (2005) and Erb et al. (2006b).
Mid-infrared imaging was obtained for 5/13 fields with
IRAC on Spitzer. Observations at 3.6, 4.5, 5.8, and 8.0
µm were obtained for the GOODS-N (Dickinson et al.
2003; Giavalisco et al. 2004; Reddy et al. 2006a), Q1700
(Shapley et al. 2005), and Q1549, Q1623, and Q2343
(Erb et al. in preparation) fields, where 3σ IRAC detec-
tion limits ranged from 25.1–24.8 (AB). The mid-infrared
data were reduced according to procedures described in
Shapley et al. (2005). 112/321 objects (35%) have detec-
tions in at least one IRAC passband, and 34/321 objects
(11%) have both Ksand IRAC detections.
2.2. Galaxy Systemic Redshifts
In order to prepare the spectra for subsequent mea-
surement and analysis, we transformed each spectrum
into the stellar systemic frame where the galaxy’s center
of mass was at rest. To do so, we employed the procedure
of Adelberger et al. (2003) to infer the galaxy’s systemic
redshift from measurements of its redshifts of both Lyα
in emission and interstellar lines in absorption. For the
spectra that clearly exhibited a double-peaked Lyα emis-
sion feature (12/321 objects), we adopted the convention
of setting the Lyα emission redshift equal to the average
redshift of the two emission peaks. This technique of
inferring a zero-velocity center-of-mass redshift, as op-
posed to measuring it directly, was necessary due to the
fact that stellar lines arising from OB stars (assumed to
be at rest with respect to the galaxy) are too weak to
measure in individual spectra at z ∼ 3. Furthermore, a
systemic redshift could not be measured from prominent
LBG spectral signposts (e.g., Lyα or interstellar absorp-
tion lines) as these features trace outflowing gas which is
offset from the galaxy’s center-of-mass frame by several
hundred km s−1(Shapley et al. 2003).
2.3. Lyα Equivalent Width
HI Lyα, typically the strongest feature in LBG spectra,
is characterized by its equivalent width, WLyα, where we
use a negative equivalent width to correspond to the fea-
ture in absorption. We present here a systematic method
for estimating WLyα, taking into account the various
Lyα spectroscopic morphologies that were observed in
the sample. In particular, this method employs a more
robust technique than used previously to determine the
wavelength extent over which the Lyα feature should be
integrated to extract a line flux.
We first binned the 321 systemic-frame spectra into
one of four categories based on the morphology of Lyα:
“emission”, “absorption”, “combination”, and “noise”.
The spectra in the “emission” bin were clearly dominated
by a Lyα emission feature, and a small subset of this
sample exhibited two peaks in emission. The spectra in
the “absorption” bin were dominated by a trough around
Lyα, typically extending for tens of angstroms bluewards
of line center.The spectra deemed to be “combina-
tion” contained a Lyα emission feature superimposed on
a larger Lyα absorption trough and the “noise” spectra
were generally featureless around Lyα, save for a possi-
ble absorption signature whose secure identification was
hindered by low signal-to-noise. Four example spectra,
characterized as falling into each of these four bins, are
shown in Figure 2.
Each spectrum, regardless of its category classification,
was fit with two average continuum levels, one bluewards
(1120–1180˚ A; cblue) and one redwards (1225–1255˚ A;
cred) of Lyα; these wavelength ranges were chosen to
avoid the prominent Si III and Si II absorption features
at 1206 and 1260˚ A, respectively. We worked with both
the spectra and the adopted continua in fλ units (erg
s−1cm−2˚ A−1). Below, we briefly describe the procedure
for calculating WLyαfor each of the four morphological
classification bins.

Page 5

5
Fig. 2.— The Lyα feature varies widely in its morphology. Four
spectra are plotted to show representative examples of objects clas-
sified in the “emission,” “combination,” “absorption,” and “noise”
bins, respectively. In order to systematically calculate Lyα equiv-
alent width, we adopted red- and blue-side continua (horizontal
lines from 1120–1180˚ A and from 1225–1255˚ A, respectively) and
inferred the extent of the Lyα feature (thick line below each spec-
trum) using the methodology described in §2.3. Note that in the
case of the “absorption” spectrum shown here, the extent of the
Lyα feature appears to extend redwards of the adopted red-side
continuum – this difference arises because the plotted spectrum is
unsmoothed while a smoothed spectrum was employed to calculate
the wavelength bounds of the Lyα feature.
Emission: 189/321 objects (59%): The wavelength of
the maximum flux value between 1213–1221˚ A was cal-
culated, as well as the wavelengths on either side of the
maximum where the flux level intersected credand cblue,
respectively. These latter two wavelengths were adopted
as the extremes of the emission feature. In a limited num-
ber of cases (12/188objects), double-peaked spectra were
individually examined to ensure that this methodology
counted both peaks as contained within the Lyα feature.
The IRAF routine SPLOT was next used to calculate the
enclosed flux between the two wavelength bounds. The
enclosed flux was then divided by the level of credto yield
a measurement of WLyαin˚ A. The level of cbluewas not
used in the calculation of WLyα due to its substantial
diminution by the intergalactic medium (IGM).
Absorption: 50/321 objects (16%): The boundaries of
the Lyα absorption feature were calculated in the same
manner as those of the “emission” spectra described
above, with the exception that the flux value between
1213 and 1221˚ A was isolated as a minimum and the
“absorption” spectra were initially smoothed with a box-
car function of width six pixels (∼ 2.5˚ A) in order to
minimize the possibility of noise spikes affecting the de-
rived wavelength boundaries of the Lyα feature. These
smoothed spectra were only used to define the extent of
the Lyα line; the original unsmoothed spectra were used
for the flux integration in IRAF and the enclosed flux was
divided by credto yield WLyα.
Combination: 31/321 objects (10%): Objects in the
“combination” bin were characterized by a Lyα emission
feature superposed on a larger absorption trough. The
boundaries of the Lyα feature were computed by begin-
ning at the base of the Lyα emission peak and moving
toward larger fluxes until the smoothed spectrum (see
above) intersected credand cblue, respectively (the same
technique used for the “absorption” spectra). Flux inte-
gration and division by cred were furthermore identical
to those objects discussed above.
Noise: 51/321 objects (16%): For these spectra domi-
nated by noise, we adopted set values for the endpoints
of the Lyα feature based on the average boundary val-
ues of the absorption and combination spectra – 1199.9
and 1228.8˚ A. (The boundaries of the “emission” spectra
were not included in this calculation, as the spectral mor-
phologies of the “emission” galaxies differed greatly from
those of the “noise” galaxies). As above, the integrated
flux was divided by the level of credto yield WLyα.
Rest frame equivalent widths ranged from −40˚ A ?
WLyα? 160˚ A, although one object (HDF–C41) had an
equivalent width of ∼ 740˚ A; we attributed this outlier
to a continuum level in the spectrum comparable with
zero and omitted this object from further analysis. The
median equivalent width of the sample was ∼ 4˚ A (Fig-
ure 3), consistent with values reported by Shapley et al.
(2001, 2003) for z ∼ 3 LBGs.
2.4. Composite Spectra
A composite spectrum offers the distinct advantage of
higher signal-to-noise over individual observations. We
accordingly constructed several composite spectra from
our sample, using, in each case, the same basic steps
discussed below.
The one-dimensional, flux-calibrated, rest-frame input
spectra of interest were stacked (mean-combined) using
the IRAF scombine routine. Each input spectrum was
scaled to a common mode over the wavelength range
1250–1380˚ A and a small number of positive and neg-
ative outliers (< 10% of the data) were rejected at each
pixel position to prevent poor sky-subtraction or cosmic
ray residuals from affecting the composite spectrum. The
final composite spectrum was then rebinned to a disper-
sion of 1˚ A pixel−1. A composite spectrum7of the entire
sample is shown in Figure 4, where several photospheric
7This composite spectrum is meant to represent only the average
of the objects in our sample, not the average of the entire z ∼ 3
LBG population. There are a variety of observational biases that
affect that relative proportions of Lyα-emitters and Lyα-absorbers
selected: large Lyα emission lines contaminating the G band result
in redder Un – G colors, scattering objects into the color section
window (Equation 1), while Lyα absorption limits the dynamic
range in continuum magnitude over which objects are selected. We