Star Formation Histories in a Cluster Environment at z ~ 0.84

Abstract

We present a spectrophotometric analysis of galaxies belonging to the dynamically young, massive cluster RX J0152.7–1357 at z ~ 0.84, aimed at understanding the effects of the cluster environment on the star formation history (SFH) of cluster galaxies and the assembly of the red sequence (RS). We use VLT/FORS spectroscopy, ACS/WFC optical, and NTT/SofI near-IR data to characterize SFHs as a function of color, luminosity, morphology, stellar mass, and local environment from a sample of 134 spectroscopic members. In order to increase the signal-to-noise ratio, individual galaxy spectra are stacked according to these properties. Moreover, the D4000, Balmer, CN3883, Fe4383, and C4668 indices are also quantified. The SFH analysis shows that galaxies in the blue faint-end of the RS have on average younger stars (Δt ~ 2 Gyr) than those in the red bright-end. We also found, for a given luminosity range, differences in age (Δt ~ 0.5–1.3 Gyr) as a function of color, indicating that the intrinsic scatter of the RS may be due to age variations. Passive galaxies in the blue faint-end of the RS are preferentially located in the low density areas of the cluster, likely being objects entering the RS from the "blue cloud." It is likely that the quenching of the star formation of these RS galaxies is due to interaction with the intracluster medium. Furthermore, the SFH of galaxies in the RS as a function of stellar mass reveals signatures of "downsizing" in the overall cluster.

Full text

Page 1

arXiv:1009.3986v1 [astro-ph.CO] 21 Sep 2010
Draft version September 22, 2010
Preprint typeset using LATEX style emulateapj v. 03/07/07
STAR FORMATION HISTORIES IN A CLUSTER ENVIRONMENT AT Z ∼ 0.84
R. Demarco1,2, R. Gobat3, P. Rosati4, C. Lidman5, A. Rettura6, M. Nonino7, A. van der Wel8, M. J. Jee9,
J. P. Blakeslee10, H. C. Ford2, M. Postman11
Draft version September 22, 2010
ABSTRACT
We present a spectrophotometric analysis of galaxies belonging to the dynamically young, massive
cluster RX J0152.7-1357 at z ∼ 0.84, aimed at understanding the effects of the cluster environment on
the star formation history (SFH) of cluster galaxies and the assembly of the red-sequence (RS). We
use VLT/FORS spectroscopy, ACS/WFC optical and NTT/SofI near-IR data to characterize SFHs as
a function of color, luminosity, morphology, stellar mass, and local environment from a sample of 134
spectroscopic members. In order to increase the signal-to-noise, individual galaxy spectra are stacked
according to these properties. Moreover, the D4000, Balmer, CN3883, Fe4383 and C4668 indices are
also quantified. The SFH analysis shows that galaxies in the blue faint-end of the RS have on average
younger stars (∆t ∼ 2 Gyr) than those in the red bright-end. We also found, for a given luminosity
range, differences in age (∆t ∼ 0.5 − 1.3 Gyr) as a function of color, indicating that the intrinsic
scatter of the RS may be due to age variations. Passive galaxies in the blue faint-end of the RS are
preferentially located in the low density areas of the cluster, likely being objects entering the RS from
the “blue cloud”. It is likely that the quenching of the star formation of these RS galaxies is due to
interaction with the intracluster medium. Furthermore, the SFH of galaxies in the RS as a function
of stellar mass reveals signatures of “downsizing” in the overall cluster.
Subject headings: galaxies: clusters: general — galaxies: clusters: individual (RX J0152.7-1357) —
galaxies: evolution — galaxies: formation
1. INTRODUCTION
There is no doubt that galaxy evolution is influenced
by the local environment. The observed properties of
galaxies (color, magnitude, morphology, metallicity) are
associated with their local neighborhood. The latter is
usually characterized in terms of the local number den-
sity of galaxies.
One of the most prominent connections between galaxy
properties and environment is the morphology-density
relation (Dressler 1980), by which early-type galaxies
dominate high-density environments in contrast to late-
type galaxies that dominate low-density ones. This rela-
tion has been quantified up to z ∼ 1, showing different
evolutionary patterns depending on whether the galaxy
sample is selected based on luminosity (Postman et al.
2005; Smith et al. 2005) or stellar mass (Holden et al.
1Department of Astronomy, Universidad de Concepci´ on. Casilla
160-C, Concepci´ on, Chile
rdemarco@astro-udec.cl
2Department of Physics & Astronomy, The Johns Hopkins Uni-
versity, 3400 N. Charles Street. Baltimore, MD 21218, USA
3CEA, Laboratoire AIM, Irfu/SAp, F-91191 Gif-sur-Yvette,
France
4European Southern Observatory, Karl-Schwarzschild-Strasse 2,
D-85748 Garching, Germany
5Anglo Australian Observatory, P.O. Box 296, Epping NSW
1710, Australia
6Department of Physics & Astronomy, University of California,
Riverside. 900 University Ave. Riverside, CA 92521, USA
7INAF-OAT, via G.B. Tiepolo 11, 40131 Trieste, Italy
8Max-Planck Institute for Astronomy, K¨ onigstuhl 17, D-69117,
Heidelberg, Germany
9Department of Physics, University of California, Davis, One
Shields Avenue, Davis, CA 95616, USA
10Herzberg Institute of Astrophysics, National Research Coun-
cil of Canada, Victoria, B.C. V9E 2E7, Canada
11Space Telescope Science Institute, Baltimore, MD 21218, USA
2007; van der Wel et al. 2007).
For mass selected samples, van der Wel et al. (2007)
show that galaxies have to evolve in mass, morphology,
and density such that the morphology-density relation
does not change since at least z ∼ 0.8.
of luminosity-selected samples, the morphology-density
relation observed at z ∼ 1 (Postman et al. 2005) is re-
ported to hold up to z ∼ 1.46 (Hilton et al. 2009).
Since early-type galaxies are among the reddest ob-
jects in any given sample at a given epoch and contain
the bulk of the stellar mass in the Universe, the above
morphology-density relation can be translated into two
other relations: color-density and stellar mass-density.
In particular, the color-density relation, characterized by
the tendency of red galaxies (mostly early-type ones) to
be found in the core of clusters, can be used as a tool to
identify high-redshift clusters. It takes advantage of one
of the most distinctive features in the color-magnitude
diagram of a cluster: the so-called red-sequence (RS;
de Vaucouleurs 1961; Visvanathan & Sandage 1977).
This RS, however, is not exclusive of clusters as it is
also found in the field. In fact, some of the observed
properties of the RS such as its color scatter and lu-
minosity coverage vary, at a given redshift, with local
galaxy density (e.g., Tanaka et al. 2005). This highlights
the influence that the local environment has on galaxy
properties such as colors.
It has been observed that the RS of clusters has a
small scatter in color and a slope which do not seem
to evolve over ∼ 9 Gyr of cosmic time since z ∼ 1.4 (e.g.,
Stanford et al. 1998; Blakeslee et al. 2003b; Mei et al.
2006b; Lidman et al. 2008). This lack of evolution was
shown to be better explained by a RS being primar-
ily a color-metallicity relation instead of a color-age one
In the case

Page 2

2Demarco et al.
(Kodama & Arimoto 1997). However, more recent evi-
dence gathered from local samples of galaxies shows that
variations of stellar age along the red-sequence are also
present (e.g., Gallazzi et al. 2006; Bernardi et al. 2006)
in addition to variations in metallicity, with some of the
intrinsic color scatter of the red-sequence being due to
stellar age differences (see also Tran et al. 2007).
A common procedure is to use the spectral energy
distribution (SED) fitting technique to estimate ages
and formation redshifts for cluster galaxies in the RS
(e.g., Blakeslee et al. 2003b; Lidman et al. 2004, 2008;
Blakeslee et al. 2006; Mei et al. 2006a,b; Gobat et al.
2008; Rettura et al. 2010). This procedure allow us to
constrain the epoch and mode of formation of mas-
sive early-type galaxies.When spectroscopic data al-
low it, spectral indices are also measured to constrain
the properties and evolution of cluster galaxies (e.g.,
Jørgensen et al. 2005; Tran et al. 2007; Braglia et al.
2009).
At z = 1.2, Gobat et al. (2008) find that the bulk of
the stars in cluster early-type galaxies is formed ∼ 0.5
Gyr earlier than that in field early-type galaxies, and
RS galaxies were already in place ∼ 1 Gyr earlier. Al-
though the most massive (in stellar content) early-type
galaxies do not show such an age difference with envi-
ronment, this age divergence is most noticeable at stel-
lar masses ? 1011M⊙. A similar conclusion was reached
by Rettura et al. (2010), who show that the environment
regulates the timescale associated with the SFHs of early-
type galaxies, with a fraction of field system having a
more extended period of stellar mass assembly.
Braglia et al. (2009) find that the SFHs of galaxies in
two clusters at z ∼ 0.3 depend on local environment
which is also related to the cluster dynamical state. In
addition to the expected gradient of star formation with
clustercentric distance, both luminous (L ≥ L∗) and sub-
luminous members contribute to a sharp increase of the
star formation activity along filaments connected to the
dynamically young, merging system. The more relaxed
cluster, on the other hand, is mostly dominated by red,
passive galaxies or galaxies whose star formation is being
quenched.
The novel procedure used by Gobat et al. (2008) to
determine the SFH of cluster galaxies combines both
the SED fitting technique and, simultaneously, a fit to
the spectroscopic features (pseudo-continuum and ab-
sorption) of a galaxy spectrum. The SED covers a wide
range of wavelengths and provides information on mass
and current SFH while the spectrum, although on a
much more limited wavelength range, allows one to de-
termine the age of the stellar population of a galaxy with
greater precision. The combination of photometry and
spectroscopy therefore puts stronger constraints on the
SFH than either alone. This approach can also be com-
plemented by determining some relevant spectral indices
available at the observed wavelength range.
The galaxy cluster RX J0152.7-1357 (Della Ceca et al.
2000) at z ∼ 0.84, with its dynamically young and com-
plex structure, represents and ideal laboratory to study
the relation between galaxy SFH and environment. Here
we apply the above spectrophotometric fitting technique
to the spectroscopically confirmed population of cluster
members. Our goal is to deepen our understanding of
how galaxy evolution is driven by environmental pro-
cesses and, in particular, to better constrain and under-
stand the physical mechanisms that contribute to form
the RS and set its observed properties.
RX J0152.7-1357 is one of the most distant X-ray lu-
minous clusters discovered in the ROSAT Deep Clus-
ter Survey (Rosati et al. 1998).
epoch of greater cosmic activity in terms of stellar mass
buildup in galaxies (e.g., Doherty et al. 2006), with the
cluster itself being assembled by the merging of two sub-
clusters (Demarco et al. 2005; Girardi et al. 2005) and
by the accretion of groups from surrounding filaments
(Tanaka et al. 2006).
To date, RX J0152.7-1357 has been the subject of a
number of studies: ICM structure and X-ray properties
(Maughan et al. 2003), RS properties (Blakeslee et al.
2006; Patel et al. 2009), cluster dynamics and sub-
structure (Demarco et al. 2005; Girardi et al. 2005),
star-forming members (Homeier et al. 2005), Sunyaev-
Zel’dovich properties (Joy et al. 2001), weak lensing
mass structure (Jee et al. 2005), physical properties of
galaxy members (Jørgensen et al. 2005), infrared sources
in the cluster (Marcillac et al. 2007), and large-scale fil-
aments associated with it (Tanaka et al. 2006). It is one
of the clusters in the ACS intermediate-redshift cluster
program (Ford et al. 2004) with one of the most compre-
hensive datasets, from X-ray to infrared.
We improve over previous studies of this cluster by con-
sidering, simultaneously, a large enough (∼ 130) number
of spectroscopic members with high enough (< 15˚ A)
spectral resolution data and 5-band photometry (from
optical to near-infrared). In addition, we characterize the
local environment by the relative, projected dark mat-
ter density instead of the local projected number density
of galaxies. We estimate SFHs and spectral indices as
a function of color-magnitude, morphology, stellar mass
and location within the cluster, which allows us to es-
tablish a comprehensive view of galaxy properties and
their connection with the environment. In contrast to
other works at this redshift, given the available number
of members in the RS, we divide this one into a grid in
color-magnitude space where the average SFH of galaxies
can be studied at different colors and magnitudes. This
approach is aimed at providing a deeper insight into RS
variations in order to unveil the physics responsible of its
observed properties such as its intrinsic scatter.
The present work is structured in the following way. In
§2 we describe the observations and the dataset used in
our analyses. In §3 we focus on the technical details of
this investigation, such as the definition of regions in the
hyperspace of galaxy properties considered for this inves-
tigation, the available spectra, the spectrophotometric
fitting procedure used to characterize the average SFH
within those regions, and the spectral indices used to
complement such SFH characterization. In §4 we present
the results of our analyses followed, in §5 by a discussion
focused on the formation of the RS in the cluster. We
finally summarize our main conclusions in §6.
Throughout the paper, unless explicitly indicated, we
assume a ΛCDM cosmology with H0 = 70 km s−1
Mpc−1, ΩM= 0.3 and ΩΛ= 0.7.
It is observed at an
2. OBSERVATIONS
RX J0152.7-1357 has been observed with a number of
instruments from the ground and space. In this work

Page 3

Star formation histories in a cluster environment at z ∼ 0.843
we have used optical and near-IR imaging data obtained
with HST/ACS and NTT/SofI, respectively, and op-
tical spectroscopy obtained with VLT/FORS. Descrip-
tions of these observations and data reductions can be
found in the existing literature (Demarco et al. 2005;
Blakeslee et al. 2006), hence, we only provide a short
summary of them below.
2.1. Imaging and photometry
As reported in Blakeslee et al. (2006), RX J0152.7-
1357 was observed with ACS (Ford et al. 1998) on HST
in three bands: F625W, F775W and F850LP, hereafter
referred to as r625, i775and z850, respectively. The clus-
ter was imaged using a 2 × 2 overlapping pattern, pro-
ducing a mosaic of about 5.′8 on a side with a central
overlapping region of about 1′. Each pointing was ob-
served for two orbits per filter, giving a total orbit ex-
penditure of 24 orbits. The images were processed and
the final mosaic produced using the ACS GTO Apsis
pipeline (Blakeslee et al. 2003a). For more details, see
Blakeslee et al. (2006).
The near-IRdata were
(Moorwood et al.1998)on
Demarco et al. 2005). The cluster was imaged in
the J and Ks bands under sub-arcsecond seeing for 3.8
and 3 hours, respectively.
region of 4.′9 on a side and were reduced in a standard
manner.
In order to use the ACS and SofI data in a consistent
way, we produced a new multi-band photometric cata-
log, different from those used in Demarco et al. (2005)
(SofI) and Blakeslee et al. (2006) (ACS). Photometry
from the ACS data was obtained in dual mode using
the ACS z850 image for detections.
tractor (Bertin & Arnouts 1996) on the ACS mosaic, 41
point sources were selected to derive aperture correc-
tions. Magnitudes (in the AB system; Oke 1974) within
radii of 0.′′75 and 2.′′0 were compared, resulting in differ-
ences of 0.039 for both the r625and i775bands and 0.046
for the z850 filter. Zero points (in AB magnitudes) for
the r625-, i775- and z850-band data are 36.542, 36.321 and
35.520, respectively.
The near-IR J and Ksimages were registered onto the
ACS ones, with residuals of less than 1 pixel for both J
and Ks. The same point sources were used to derive cor-
rections between apertures of 1′′and 5′′in radius. These
corrections are 0.254 and 0.239 for J and Ks, respectively.
Additionally, extinction corrections (Schlegel et al. 1998)
of 0.014 in J and 0.009 in K were applied, resulting in a
total correction for the 1′′radius aperture of 0.268 in J
and 0.248 in Ks. Zero points (in the AB system) in J and
Ksare 27.260 and 26.745, respectively. Hereafter, unless
otherwise indicated, magnitudes are in the AB system.
To transform the near-IR photometry from the Vega sys-
tem to AB magnitudes, we used corrections of 0.960 and
1.895 for the J and Ksbands.
For completeness, we note that flanking fields sur-
rounding the central r625, i775and z850mosaic were ob-
tained with ACS in F606W (broad V) and F814W (broad
I), however, we did no attempt to use those data in this
work because of their shallower integration, reduced pho-
tometric coverage, and lack of overlap with the existing
near-IR data.
obtained
the
with
NTT
SofI
(see ESO
The final images cover a
By running SEx-
2.2. Spectroscopy
RXJ0152.7-1357
ofa
anditsoutskirts
extensive
have
spectroscopic
been
subject
surveys (Demarco et al. 2005; Jørgensen et al. 2005;
Tanaka et al. 2006; Patel et al. 2009).
vey by Tanaka et al. (2006) concentrated on the large-
scale structures surrounding the main cluster, only
8 members in Jørgensen’s survey were not included
in Demarco’s work.In the present study, we use
the spectra of the 102 cluster members confirmed by
Demarco et al. (2005), complemented with the spectra
of 32 new cluster members obtained from subsequent
FORS2 (Appenzeller & Rupprecht 1992) spectroscopy
on the ESO VLT. These 134 sources are listed in table
1. IDs are in the same system of those in Demarco et al.
(2005), and the last column corresponds to their emis-
sion line flag: a value of 0 is given to passive galaxies, a
value of 1 is given to emission line galaxies, and a value
of 2 is given to AGN.
The new 32 spectroscopic members were obtained after
targeting clusters candidates with 4 multi-object masks,
using the Mask Exchange Unit (MXU) on FORS2, and
exposing each of them until reaching integration times
between 3 and 4 hours. The data were collected be-
tween November 4th and November 7th, 2005, in Vis-
itor Mode (ESO program ID 076.A-0889(A)) under an
average seeing of ∼ 0.′′9. A total of 152 galaxies were
observed with slits of 1′′in width using the 300V grism.
The data were binned by 2 pixels which resulted in a
dispersion of ∼ 3.3˚ A/pixel and a spectral resolution of
∼ 13˚ A. One mask was observed using no order-separation
filter while the other three masks were exposed using the
GG375+80 filter. Although these observations were de-
signed to target gravitational arc candidates, the new
members resulted from putting the slits on fillers that
were likely cluster members based on their photometric
redshift (0.7 < zphot< 0.95; see Demarco et al. 2005).
The observations were prepared and the data reduced
in a similar way and using the same dedicated soft-
ware as described in Demarco et al. (2005). Redshifts
were obtained by cross-correlating (Tonry & Davis 1979;
Kurtz et al. 1992) the observed spectra with template
galaxy spectra from Kinney et al. (1996). Out of 113
redshifts, 32 were securely confirmed within the range
defining cluster membership (0.81 < z < 0.87; see
Demarco et al. 2005).Observational errors (obtained
from observing several sources more than once) are
of the same order of magnitude as those reported in
Demarco et al. (2005, 2007), i.e., δz ∼ 8 − 12 × 10−4∼
10−3.
numberof
While the sur-
3. ANALYSIS
In an effort to better understand how galaxy proper-
ties, such as age and stellar mass, contribute to estab-
lishing the observed characteristics of the RS, we focus
the present analysis on the star formation history (SFH)
of cluster galaxies in the RS. As the luminosity coverage
and scatter of the RS are observed to vary between clus-
ter and field environment (e.g., Tanaka et al. 2005), we
also investigate the impact that the intracluster environ-
ment may have on the above SFHs.
Traditionally, SFHs are determined by fitting syn-
thetic galaxy spectra to the available photometry (e.g.,

Page 4

4 Demarco et al.
Rettura et al. 2006, 2010), however, in the present anal-
ysis, we additionally perform a simultaneous fit to the
available spectra as performed by Gobat et al. (2008).
In order to increase signal-to-noise (S/N) for the spec-
trophotometric fitting, galaxy spectra are co-added as
explained in §3.1. Before stacking, galaxy spectra are
grouped according to color, magnitude, location within
a given subcluster and with respect to the projected DM
distribution, stellar mass, and visual morphology. All
these grouping regions are defined in §3.2 and the details
of the spectrophotometric fitting procedure are given in
§3.4.
3.1. Co-added spectra
The individual spectra of the 134 cluster members of
RX J0152.7-1357 used in the present analysis have S/N
ratios in the range 1 to 33 with a mean value of 7.6.
The S/N ratios were obtained from the ratio between
the mean flux and the r.m.s. flux calculated within the
wavelength intervals defining the continuum windows for
the HδAfeature (Worthey & Ottaviani 1997). Since the
redshift survey presented in Demarco et al. (2005) was
designed to mainly provide redshifts, the quality of the
individual spectra is not good enough to perform a mean-
ingful fit to the different spectral features sensitive to the
SFH. Therefore, in order to increase the S/N ratio to ob-
tain a satisfactory SFH characterization, we decided to
use average spectra obtained from co-adding individual
spectra that were grouped according to various criteria
(see §3.2).
This stacking technique has already been success-
fully used in previous studies about the stellar popu-
lations in cluster galaxies at intermediate-redshift (e.g.,
Dressler et al. 2004), and the algorithm employed here
is the same as in Demarco et al. (2007) and Gobat et al.
(2008). The individual spectra can be weighted by their
S/N, and only those with a S/N > 3 were selected for
stacking. The S/N ratios of the final, co-added spectra
vary between ∼6 and ∼42 for the unweighted stacking,
and between ∼6 and ∼62 for the weighted stacking.
3.2. Grouping cluster members
SFHs are determined from co-added spectra grouped
according to relevant observables. In this work we study
their dependence on galaxy color and magnitude, pro-
jected angular distribution, visual morphology, stellar
mass, and projected DM density. These observables can
be considered as forming an hyperspace in which galaxies
are located. The regions within this hyperspace used for
stacking spectra are defined in what follows and table 2
summarizes all these definitions.
3.2.1. Galaxy colors and luminosities
Since we are interested in understanding the physi-
cal origin of the scatter of the RS in clusters, and ul-
timately the details of the color-luminosity evolution of
cluster galaxies into the RS, we need first to separate
cluster members into blue and red galaxies. We follow
the traditional way of using two filters that straddle the
4000˚ A-break at the cluster redshift to achieve this. Fig.
1 shows a simple stellar population (SSP), 12 Gyr old,
solar metallicity spectral energy distribution from the
Bruzual-Charlot (Bruzual & Charlot 2003) library, red-
shifted to the cluster redshift (z = 0.837; Demarco et al.
2005). On top of it, our choice of filters (r625 and Ks)
to straddle the 4000˚ A-break is laid out, which allows
us to separate in an efficient way red (early-type) from
blue (late-type) galaxies. In contrast to Blakeslee et al.
(2006), we prefer to use the Ks-band as a the “red” filter
because of its ability to trace the rest-frame near-IR light
coming from the bulk of the stellar content of galaxies at
z ∼ 0.8, unaffected by biases due to recent star formation
(Stanford et al. 1998).
Fig. 2 shows the Color-Magnitude Diagram (CMD) of
RX J0152.7-1357 (for a detailed analysis of the CMD of
this cluster, see Blakeslee et al. 2006). Red circles corre-
spond to passive (no detectable emission features) clus-
ter members, black triangles correspond to star-forming
(with detectable [OII]) members, and the two blue
squaresare the confirmed AGN members (Demarco et al.
2005). The black dots are sources within the ACS mo-
saic for which photometric information is available, and
the cross at the lower-left of the plot indicates typical er-
ror bars in magnitude and color. The horizontal dashed
line has arbitrarily been set at r625− Ks= 2.3 to sepa-
rate blue from red galaxies. We consider galaxies with a
r625− Ks> 2.3 color as belonging to the cluster RS.
In order to better explore SFH variations as a function
of color and magnitude within the RS, we have subdi-
vided the latter into a number of regions12. A first sub-
division is defined, consisting of three bins in Ks mag-
nitude, 1/BRS, 2/MRS and 3/FRS, as defined in table
2. In addition, a second and finer subdivision, both in
color and magnitude, is established, as shown in Fig.
2. The vertical dashed line has arbitrarily been set at
Ks = 20.75 in order to divide the RS into “bright”
and “faint” bins. For comparison, K∗∼ 19.7 (AB) at
z ∼ 0.84 (see Ellis & Jones 2004), therefore, this separa-
tion corresponds to ∼ K∗+ 1.
The locus of the RS in Fig. 2 is obtained by a lin-
ear least-squares fit to the data in color-magnitude space
with r625− Ks> 2.3 and 18 < Ks< 24. This fit to the
RS (r625−Ks= (−0.219±0.001)×Ks+(7.751±0.026)),
indicated by the solid black line, is used to define the
“blue”, “green” and “red” areas parallel to it in color-
magnitude space, further separated into 3 bright (re-
gions 4/BBRS, 5/BGRS and 6/BRRS, respectively) and
3 faint (regions 7/FBRS, 8/FGRS and 9/FRRS, respec-
tively) bins (see table 2 for their definitions).
Only passive galaxies (red circles) within these 9 re-
gions were considered for stacking, because of our inter-
est in focusing on the quiescent, early-type galaxy pop-
ulation. In general red, star-forming sources are dust-
enshrouded systems (e.g., Smail et al. 1999; Wolf et al.
2005). However, we note that a few of the apparently
passive, RS galaxies could in fact be star-forming sys-
tems with the [OII] feature totally suppressed by a large
amount of dust (e.g., Smail et al. 1999).
The above color separation for galaxies in the RS does
not follow the slope of the fit to the RS. Selecting galaxies
following this slope would tend to include more galaxies
into the blue faint-end of the RS that may belong to the
12Throughout the text we use the pair N/Acronym to identify
the different stacking regions. N corresponds to the region ID listed
in the first column of table 2 and the Acronym is given in the
comment column of the same table. The goal of this is to make
figures easier to read while giving a physical meaning to the IDs at
the same time.

Page 5

Star formation histories in a cluster environment at z ∼ 0.845
so-called “green valley” or to the “blue cloud”. Since we
are interested in passive galaxies in the RS, the flat color
separation we have adopted produces no different result
from a slope-driven color separation.
We have 76 spectra corresponding to non [OII] emis-
sion line galaxies in the RS available for our analyses.
In order to have a reasonable number of sources to be
co-added, we set the width of the two (bright and faint)
central regions (green hatched areas) of the finer parti-
tion to be 0.2 mag in r625−Ks (±0.1 in r625−Ks from the
fit), which is about 1.1 times the observed r.m.s. color
scatter.
3.2.2. Projected angular distribution
To investigate the relation between stellar content and
local environment, we study the SFH of cluster galax-
ies as a function of angular distribution on the sky.
Due to the complex matter (DM, gas and galaxies)
distribution of RX J0152.7-1357 (Maughan et al. 2003;
Demarco et al. 2005; Jee et al. 2005; Girardi et al. 2005),
it is very difficult to determine the center of the cluster.
Instead of stacking galaxies in concentric rings with a
common center, we try the following approach that con-
siders the known main subclusters (see Demarco et al.
2005; Girardi et al. 2005).
We separate the cluster field of view (FoV) in
two halves at a fiducial center (R.A.= 01h52m41.80s,
DEC.=-13o57′′52.′′5) located at the mid-point between
the centers of the northern and southern clumps defined
by Demarco et al. (2005), as indicated by the dashed,
horizontal line in Fig. 3. We then consider concentric
(semi-)annuli or radial sectors centered at each clump,
however, truncated at the separation half-way between
them (solid contours in Fig. 3). Areas or sectors asso-
ciated with the northern clump (20/N0 through 22/N2)
are labeled as “North”, while those associated with the
southern clump (23/S0 through 25/S2) are labeled as
“South”. In table 2 we give the corresponding defini-
tions of the above regions. The center of each clump or
subcluster is taken from Demarco et al. (2005).
All galaxies, passive and those showing emission lines,
were considered for stacking within these regions. We
only excluded the known AGN members.
3.2.3. Galaxy morphology
We use the morphological classification given by
Postman et al. (2005) for galaxies in RX J0152.7-1357.
The morphological T-types used by Postman and collab-
orators are those defined in de Vaucouleurs et al. (1976).
T values ranging from -5 to -3 correspond to Elliptical
(E) galaxies, while a value of -2 corresponds to S0 galax-
ies. Values between -1 and 1 are assigned to morpholo-
gies between S0 and Sa, with values > 1 given to later
type spiral (Sp) galaxies. Type 6 is associated with a Sd
morphology, while irregular (Irr) galaxies have T values
in the range 6 < T < 9. Taking all this under con-
sideration, we established 4 groups in morphology space
as defined in table 2. Namely, group 10/E contains el-
liptical galaxies (T < −2); group 11/(S0/Sa), lenticular
galaxies (−2 ≥ T ≥ −1); group 12/Sp, spiral galax-
ies (1 < T ≤ 6); and group 13/Irr, irregular galaxies
(T > 6).
3.2.4. Stellar mass
In addition to morphology, we also grouped galax-
ies based on their stellar mass content. The work by
Holden et al. (2007) provides stellar mass estimates for
cluster galaxies in RX J0152.7-1357 based on the mass-
to-light, M/LB, ratio and rest-frame (B−V ) color linear
relation derived by Bell et al. (2003). Those masses, in
despite of being based on a single color, are proven to
be consistent with other estimates as shown by a com-
parison with dynamical measurements for some of the
same galaxies (see Holden et al. (2007), and references
therein).
However, in the analyses shown here, we decided to
recompute stellar masses by using a SED fitting proce-
dure (Rettura et al. 2006) including all the 5 bands avail-
able (see §2.1). This information allowed us to establish
3 bins in stellar mass (14/RSHM through 16/RSLM)
as defined in table 2. Stellar masses span the range
4.8 × 109< M∗≤ 3.9 × 1011M⊙, and the mass interval
for each bin has been adjusted in order to have roughly
the same number of spectra to co-add per bin. A more
detailed explanation about the way these stellar masses
were obtained is presented in §3.4. In order to study SFH
variations with stellar mass for the same sample of RS
galaxies in §3.2.1, we also restrict ourselves to quiescent,
RS galaxies, i.e., those with colors 2.3 < (r625−Ks) < 4.5
and no visible emission line features.
3.2.5. Local dark matter density
RX J0152.7-1357 has been the subject of a detailed
weak lensing analysis by Jee et al. (2005). By using the
available r625, i775and z850ACS data together with pho-
tometric and spectroscopic (from Demarco et al. 2005)
redshifts, they are able to measure the shear signal of
the cluster and reconstruct its dimensionless mass den-
sity, κ. The smoothing scale of the map is ∼ 20′′, while
its accuracy is about 20%.
As an alternative way of characterizing the cluster en-
vironment to that presented in §3.2.2, here we use the
κ map from Jee et al. (2005) to identify environments
of different projected mass density in the ACS FoV of
RX J0152.7-1357. Because of the so-called sheet-mass
degeneracy, i.e., the invariance of the shear under trans-
formations of the kind κ → λκ + (1 − λ), we use this κ
map in a relative sense, only, during the interpretation
of the results.
We thus arbitrarily define 3 different environments
based on their local, projected (total) mass density, as
shown in Fig. 4. The first of them is characterized by
mass densities > 20 × σDM (solid contours in Fig. 4),
with σDM= 0.0057 × Σc, being Σc∼ 3650M⊙pc−2the
critical mass density of the cluster (Blakeslee et al. 2006).
The κ value around the two brightest central galaxies
of the northern clump (the cluster center adopted in
Jee et al. 2005) is ∼0.3. The second one is that contain-
ing mass densities between 5 (dashed contours in Fig. 4)
and 20 times σDM, while the last of the three encom-
passes mass densities < 5 × σDM, reaching negative val-
ues in some areas. These three environments correspond
to regions 17/HDMD, 18/MDMD, and 19/LDMD, re-
spectively, as presented in table 2. Also in Fig. 4, the
distribution of spectroscopic members is indicated by the
symbols. Members in the highest density regions are in-
dicated as squares; members in the intermediate density
regions, as triangles; and members in the lowest density