The Morphological Transformation of Red-Sequence Galaxies in the Distant Cluster XMMU J1229+0151

1.1 MB·

by  P. Cerulo

Mon.Not.R.Astron.Soc.000,1–25(2014) Printed21January2014 (MNLATEXstylefilev2.2)
The Morphological Transformation of Red-Sequence
Galaxies in the Distant Cluster XMMU J1229+0151
4
P. Cerulo1⋆, W. J. Couch1,2, C. Lidman2, L. Delaye3,6, R. Demarco4, M. Huertas-
1
0 Company3, S. Mei3, R. S´anchez-Janssen5
2
1Centre for Astrophysics and Supercomputing, Swinburne Universityof Technology, PO Box 218, Hawthorn, VIC 3122, Australia
n
2Australian Astronomical Observatory, PO Box 915, North Ryde, NSW 1670, Australia
a
J 3GEPI, Paris Observatory, 77 av. DenfertRochereau, 75014 Paris, France
4Department of Astronomy, Universidad de Concepcion, Casilla 160-C Concepcion, Chile
9
5NRC Herzberg Institute of Astrophysics, 5071 West Saanich Road, Victoria, BCV9E 2E7, Canada
1
6CEA-Saclay, DSM/IRFU/SAp, F-91191 Gif-sur-Yvette,France

]
A
21January2014
G
.
h
ABSTRACT
p
We present the results of a detailed analysis of galaxy properties along the red se-
-
o quence in XMMU J1229+0151,an X-ray selected cluster at z =0.98 drawn from the
r HAWK-I ClusterSurvey(HCS). Takingadvantageofthe broadphotometriccoverage
t
s and the availability of 77 spectra in the cluster field, we fit synthetic spectral energy
a
distributions, and estimate stellar masses and photometric redshifts, which we use to
[
determinethe clustermembership.Weinvestigatemorphologicalandstructuralprop-

1 erties of red sequence galaxies and find that elliptical galaxies populate the bright
v end, while S0 galaxies represent the predominant population at intermediate lumi-
1 nosities, with their fraction decreasing at fainter magnitudes. A comparison with the
4 low-redshift sample of the WINGS cluster survey reveals that at z ∼ 1 the bright
6
end of the red sequence of XMMU J1229+0151is richer in S0 galaxies.The faint end
4
of the red sequence in XMMUJ1229+0151 appears rich in disc-dominated galaxies,
.
1 which are rarer in the low redshift comparison sample at the same luminosities. De-
0 spite these differences between the morphological composition of the red sequence in
4 XMMUJ1229+0151andinlowredshiftsamples,wefindthattowithinthe uncertain-
1
ties, no such difference exists in the ratio of luminous to faint galaxies along the red
:
v sequence.
i
X Key words: galaxy, clusters, morphology,evolution
r
a
1 INTRODUCTION whereas blue and star-forming spiral and irregular galax-
iesdominatetheoutskirts(seeDressler1980;Dressler et al.
Clusters of galaxies are the most massive virialised large
1997; Postman et al. 2005; Hilton et al. 2009; Muzzin et al.
scale structures in the universe and, thanks to the broad
2012; Mei et al. 2012). However, the fraction of blue
range of densities available in these systems, they can
star forming galaxies in clusters increases with redshift
be used as laboratories for the study of the environmen-
(Butcher& Oemler 1978) and above z ∼ 1.5 there is evi-
tal drivers of galaxy evolution (DeLucia et al. 2007). Ac-
dence of star formation in cluster cores (Hilton et al. 2010;
cording to the hierarchical scenario of structure formation,
Tran et al. 2010; Hatch et al. 2011; Hayashiet al. 2012)
galaxyclustersformafterthecollapseofthehighestdensity
These results suggest that most of the processes that led
peaksintheprimordialmatterdistribution,accretingother
totheestablishmentofthestar-formation-andmorphology-
smaller dark matter haloes at later times. By the present
localdensityrelations,astheyareobservedinlocalclusters,
day, they have built up into systems whose total mass can
were active in the interval1.0<z<1.5.
reachupto1015 M⊙,withcharacteristicsizesofacoupleof
Mpc. A large number of mechanisms have been proposed to
Up to z ∼ 1.5, red and passive elliptical and S0 both trigger star formation and then quench it. Galaxy-
galaxies are the predominant population in cluster cores, galaxy merging within groups that are falling intotheclus-
ter for the first time, galaxy-galaxy harassment within the
clusters as galaxies pass each other at high velocities, and
⋆ E-mail:[email protected] collisional compression and ram pressure stripping of gas
2 P. Cerulo et al.
in galaxies by the hot intracluster medium are mecha- that, at fixed environment (i.e. cluster core and outskirts,
nisms that are all thought to be at play in the dense clus- and field), the star formation rate (SFR), specific star for-
terenvironment(Gunn & Gott1972;Lavery & Henry1988; mationrate(SSFR)andtheamplitudeofthe4000 ˚Abreak
Moore et al. 1996;Bekki1999;Bekki & Couch 2003). How- (D (4000)) were all correlated with stellar mass. However,
n
ever, from the current observations, it is not clear which of at fixedstellar mass, thesame quantitieswere not found to
these processes is driving galaxy evolution in clusters. Fur- correlatewiththeenvironment.Theyconcludedthatstellar
thermore, clusters are highly heterogeneous systems, and mass is theprimary driverfor the quenching of star forma-
core and outskirts constitute different environments, with tion, regardless of the environment, while the environment
different global physical properties. The interactions of the actsontheglobalgalaxypopulationquenchingthestarfor-
galaxies with their surroundings may therefore change sub- mationrateoverarelativelyshortperiodoftime,regardless
stantially according to their location within thecluster. of galaxy stellar mass. In other words, the effect of the en-
Regardless of the environment in which they reside, vironmentis toaccelerate star-formation quenchingleaving
galaxies have a bimodal colour distribution, such that the thecorrelationswithmassofSFR,SSFRandDn(4000)un-
colour-magnitudediagram ofagalaxypopulationischarac- changed. The conclusions of Muzzin et al. (2012) extended
terised byanarrow sequenceof red quiescent objects and a to higher galaxy densities what had previously been ob-
diffusecloudofbluestar-forminggalaxies.Asgalaxiesfinish served in the field up to z = 2 by Penget al. (2010, 2012)
to form stars, depleting their gas reservoirs, they migrate and Quadriet al. (2012).
from the blue cloud to the red sequence. Thus, in princi- While Peng et al. (2010) and Muzzin et al. (2012) sug-
ple, the evolution of a galaxy population can be investi- gest that environment plays a role in accelerating theshut-
gatedbylookingatthegradualbuild-upoftheredsequence tingdownofstar-formation,neitherexplorethemechanisms
with redshift. However, bursts of star formation resulting by which environment does this. Demarco et al. (2010) in-
from merger events between a quiescent and a star-forming vestigated the properties of red sequence members in the
galaxy may push the galaxy back to the blue cloud (see e. clusterRXJ0152.7-1357,atz =0.84,findingthatearly-type
g. Faber et al. 2007). galaxies at the faint end of the red sequence are systemati-
Kodama & Arimoto (1997) explained the red sequence cally younger than galaxies at the bright end of the red se-
asamass-metallicity relation, thescatteraboutthebest-fit quenceandarepreferentiallylocatedintheclusteroutskirts.
straightlinebeingdrivenbyagedifferencesamonggalaxies. Theyconcludedthatstarformationinthesegalaxieshadre-
However, Gallazzi et al. (2006) demonstrated that metal- centlybeenquenchedeitherthroughrampressurestripping
licity contributes to the scatter too and that the age- ofgasbythehotintraclustermedium,orrapidexhaustionof
induced scatter is anti-correlated with stellar mass. As gasreservoirsfromanearlierepochofstarformationcaused
the local galaxy density increases, the red sequence be- bygalaxy-galaxymerging.Theseresultssupportonceagain
comes more pronounced with respect to the blue cloud the notion of a build-up of the red sequence at low masses,
(see e. g.: Hogg et al. 2004) and galaxy clusters are char- suggesting that the evolution of cluster galaxies fits in the
acterised by a tight and well defined red sequence, which downsizing scenario.
can be used to estimate the cluster redshift when no spec- WeextendtheworkofDemarco et al.(2010)bymaking
troscopic information is available (Gladders & Yee 2005; a comprehensive study of the morphological and structural
Andreon & Huertas-Company 2011). properties of red sequence galaxies in the cluster XMMU
Between z = 1 and z = 0, the number of red sequence J1229+0151 (hereafter XMM1229, Santos et al. 2009, Fig.
galaxies fainter than M = −20.0 is found to increase, ap- 1), at z = 0.98. The cluster is part of the HAWK-I Clus-
V
proximatelyhalvingtherelativeratiobetweenluminous(i.e. ter Survey (HCS, Lidman et al. 2013), comprising a sam-
M < −20.0) and faint galaxies (M > −20.0) (see e. ple of 9 galaxy clusters at 0.8 < z < 1.5. At the low
V V
g. Tanaka et al. 2005; De Lucia et al. 2007; Gilbank et al. redshift end in the HCS sample there is the cluster RX
2008;Capozzi et al.2010;Bildfell et al.2012;Lemaux et al. J0152.7-1357,whosegalaxypopulationwasstudiedindetail
2012).This deficitofgalaxies at thefaint endof theredse- byDemarco et al.(2010).XMM1229hasimagingandspec-
quencehasbeendetectedinclustersuptoredshift z=1.62 troscopic data from several space and ground based facili-
(Rudnicket al.2012).However,Andreon(2008)studiedthe ties and its X-ray and dark matter properties were studied
trendoftheluminoustofaintratioinasampleof28galaxy by Santos et al. (2009) and Jee et al. (2011), respectively.
clusters at 0.0 < z < 1.3, finding no deficit. The existence Furthermore,Santos et al. (2009) also analysed the proper-
of a deficit of galaxies at the faint end of the red sequence tiesofthespectroscopicallyconfirmedmembersofthisclus-
supports the notion of a build-up at low masses. Low-mass ter. The present work extends the analysis of Santos et al.
galaxies quenchtheirstarformation atlater times,with re- (2009) to other cluster members selected with photometric
spect to higher-mass systems, and therefore they join the redshifts, whose estimation has been possible thanksto the
red sequence at later times. This property, which is known additionalimaging databecomeavailableafterthepublica-
as downsizing (Cowie et al. 1996), is observed also in the tion of that work and described in §2.
field up to z ∼ 2 (see e.g. Tanaka et al. 2005; Ilbert et al. The position of the cluster in the redshift range of the
2010). HCS and the availability of a considerable quantity of data
Muzzin et al.(2012)investigatedtheseparateeffectsof makeXMM1229 suitable to develop a method for theanal-
massandenvironmentontheevolutionofgalaxiesinasam- ysis of the red sequence in the HCS clusters. The method
pleofclustersat0.8<z<1.4fromtheGeminiClusterAs- developedinthispaperwillbeextendedtotheotherclusters
trophysicsSpectroscopicSurvey(GCLASS).Theyobserved oftheHCSsampleinordertostudythebuild-upofthered
adecreaseofthefractionofstar-forminggalaxiesatincreas- sequence in galaxy clusters at 0.8 < z < 1.5. In particular,
ing galaxy stellar masses and local densities. They found ourinterestwillbefocused onthreepoints:thecharacteris-
Morphological Evolution in XMMU J1229+0151 3
2.1 HST imaging
2.1.1 Advanced Camera for Surveys (ACS)
XMM1229 was imaged in November 2006 in the F775W
(i ) and F850LP (z ) bands of the Wide Field Chan-
775 850
nel(WFC) of theAdvancedCamera for Surveys(ACS),on
boardtheHubbleSpaceTelescope(HST).Theobservations
werepartoftheHubbleSpaceTelescopeClusterSupernova
Survey (Dawson et al. 2009), aimed at the search of Type
Iasupernovae (SNeIa) in cluster galaxies at 0.9<z<1.5.
We summarise here the survey strategy and the data
reduction process for the ACS observations of XMM1229,
referring to Dawson et al. (2009) and Suzukiet al. (2012)
for a more detailed description. The HST Cluster Super-
Figure2.PhotometriccoverageoftheXMM1229field.Fromleft nova Survey collected i and z observations of 25 X-
775 850
to right: R SPECIAL (R), F775W (i), F850LP (z), F105W (Y), ray, optically or infrared (IR) detected galaxy clusters over
F125W (J), F160W (H), Ks. The F110W transmission curve is
the redshift range 0.9 < z < 1.5. The clusters were ob-
not plotted, as it covers the same spectral range of the F105W
served in multiple visits and, in each visit, at least one
and F125W bands. The SofI J band transmission curve is not
exposure in i and three or four exposures in z were
plottedbecauseitoverlapswithF125W.TheF775WandF850LP 775 850
collected. The images were calibrated using the standard
bands usedforthecolour-magnitudediagramarehighlightedby
the arrows. The solid black line represents the template SED of calibration pipeline provided by the Space Telescope Sci-
anellipticalgalaxyfromColemanetal.(1980)attheredshiftof ence Institute (STScI) and were registered and stacked us-
XMM1229. ingMultidrizzle(Fruchter& Hook2002;Koekemoer et al.
2002),withafinalpixelscaleof0.05′′/pixelforalltheclus-
ters.XMM1229 wasobservedduring8visits,with 9and30
ticsoftheredsequenceitself(itsshape,slope,andscatter), exposuresrespectivelycollected forthei775 andz850 bands.
theratiobetweenluminousandfaintgalaxies,andthemor- Theresulting image covers5′.1 × 5′.1, with an image qual-
phologicalpropertiesofgalaxiesalongtheredsequence.The ity1 of ∼ 0.09′′ in both bands. The details of the ACS ob-
availability of the Wide Field Nearby Galaxy-clusters Sur- servations are reported in Table 1.
vey(WINGS,Fasano et al.2006)andtheMORPHSsurvey
(Smail et al. 1997) allow us to build comparison samples of
clusters at z ∼0 and 0.3<z <0.6, respectively, which are 2.1.2 Wide Field Camera 3 (WFC3)
usedtocomparethepropertiesoftheredsequencemembers
XMM1229wasimagedintheIRchannelofWideFieldCam-
inXMM1229and,inaforthcomingpaper,oftheentireHCS
era3(WFC3),onboardHST,aspartoftheprogram12051
sample.
(P. I.: S. Perlmutter), aimed at the calibration of the sen-
The paper is organised as follows: we describe the ob-
sitivities of NICMOS and WFC3 for faint objects. The ob-
servations and data analysis in §§2. and 3. §4 presents the
servations were taken in 2010 May 24 in the F105W (Y),
resultsofourstudywhicharediscussedin§5,whilewesuma-
F110W, F125W (J) and F160W (H) filter bands, following
mariseourworkanddrawourconclusionsin§6.Throughout
a BOX-MIN dithering pattern. The images were reduced
the paper we use a ΛCDM cosmology with H = 71 km ·
0 with calwf3, using the most recent versions of the calibra-
s−1 · Mpc−1, Ω = 0.27, and Ω = 0.73. All magnitudes
M Λ tion frames available for WFC3. We combined the reduced
arequotedintheABsystem(Oke1978),unlessstatedoth-
images( fltfiles)withMultidrizzle, settingthedropsize
erwise. (pixfracparameter) to0.8andthepixelsizeto0.06′′.The
resultingimagescoveranareaof3′×3′,withimagequalities2
thatvarybetween0.11′′ and0.14′′.TheWFC3observations
of XMM1229 are summarised in Table 1.
2 OBSERVATIONS AND DATA REDUCTION
XMM1229 was first discovered as an X-ray overdensity
2.2 Ground Based Imaging
in the XMM-Newton Distant Cluster Project (XDCP)
(Fassbender et al. 2011), a survey that used XMM-Newton 2.2.1 FORS2
telescope data to detect distant galaxy clusters. The clus-
XMM1229 was observed with the FOcal Reducer/low dis-
ter has been later targeted by several space- and ground-
persion Spectrograph 2 (FORS2, Appenzelleret al. 1998),
based telescopes (HST, VLT, NTT), providing us with a
mounted on Yepun, the fourth unit of the 8 m ESO/VLT,
rich dataset covering the spectral region 0.65 <λ<2.2µm
(Fig.2),aswellasspectrafor26clustermembers.Withthe
datahavingbeenacquiredatdifferenttimesandondifferent
1 ThroughoutthispaperweusetheFWHMofstarsasameasure
telescopes, with strategies varying with each program, the
oftheimagequality.
resultingdatasetisverydiverse.Thissectiondescribesthis 2 The image quality reported in Table 1 corresponds to the
richmulti-wavelengthdatasetdetailinginparticularthere- FWHM of the intrinsicPSFresulting fromthe deconvolution of
duction methods we have employed to be able to utilise it the observed PSF and the pixel response function of the WFC3
as a complete ensemble. IRdetector (Koekemoer etal.2011).
4 P. Cerulo et al.
Figure 1.ColourimageofXMM1229obtainedbycombiningtheACSF775WandF850LPimages,andtheHAWK-IKsimage.White
circles are photometrically selected red sequence members and yellow squares are spectroscopically confirmed members (see §4.1 and
Table3forthedeterminationoftheclustermembershipandacompletelistofredsequence galaxies,respectively).
onCerroParanal(Chile).Theobservationsweretakendur- XMM1229 was observed for a total of 40 minutes with
ing program 073.A-0737(A) (P.I. A. Schwope), an optical the instrument operating in Large Field Mode, which has a
follow-up oftheXDCP,carriedoutwiththeR SPECIAL and 5 ′ × 5 ′ field of view, with a resolution of 0.290′′/pixel. In
Z GUNN filters. Weonly usedtheR SPECIAL (R)data,as the order to take into account the high variability of the NIR
quality of the ACS z image is significantly higher. The background, dithered exposures of the field were taken and
850
FORS2 camera comprises a detector array of two 2k × 4k reduced with the ESO/MVM software 5. The resulting im-
MIT CCDs, separated by a 7 pixel gap and, in order to agequalityis∼0.9′′.TheSofIJbandobservationsaresum-
correct for that, the XMM1229 field was observed in three marised in Table 2.
ditheredexposuresof380seach,thatwerereducedusingthe
FORS data reduction pipeline3. Each chip was separately
biassubtractedandflat-fieldcorrected,andastandardstar
2.2.3 HAWK-I
field was used to estimate the zero point. The Software for
Calibrating AstroMetry and Photometry (SCAMP, Bertin AsubsampleoftheHSTClusterSupernovaSurvey,oneclus-
2006)wasthenusedtofindanappropriateastrometricsolu- ter from SpARCS (Muzzin et al. 2012; Lidman et al. 2012)
tionfortheimages, thatwerefinallyco-addedusingSWarp and the cluster RX J0152.7-1357, from the ACS Intermedi-
v2.19.1 (Bertin et al. 2002)4. The field of view of the final ateRedshiftClusterSurvey(Ford et al.2004),allobservable
R band image of XMM1229 is 8′ × 9′ with an image qual- fromthesouthernhemisphere,wereobservedwiththeHigh
ity of ∼0.6′′ and a resolution of 0.252′′/pixel. The FORS2 Acuity Wide field K-band Imager (HAWK-I, Pirard et al.
observations are summarised in Table 2. 2004), mounted at the Nasmith A focus of Yepun, the
fourth unit of the 8 m ESO/VLT. These clusters are part
oftheHAWK-IClusterSurvey(HCS)(Lidman et al.2013),
2.2.2 SofI a near-infrared program providing deep imaging data nec-
essary for the study of the properties of old stellar pop-
XMM1229wasobservedintheJandKsfilterbandsofSofI
ulations in cluster early-type galaxies at high redshift. We
(Moorwood et al.1998),mountedontheESONewTechnol-
refertoLidman et al.(2013)foradetaileddescriptionofthe
ogy Telescope (NTT), at the La Silla Observatory (Chile).
HAWK-Iobservationsanddatareduction oftheXMM1229
ThedataweretakeninMarch2007aspartofanearinfrared
field.
(NIR)follow-upoftheXDCP.InthispaperweonlyusetheJ
TheHAWK-Icameraconsistsofanarrayof4detectors,
band,astheKsdatafromHAWK-Iareconsiderablydeeper
each definingthe quadrantof a square surface, with atotal
(see §2.2.3 and Table 2). We summarise here the SofI ob- areacorrespondingtoa7.′5×7.′5fieldofview.Thefinalco-
servations of the XMM1229 field, referring to Santos et al. added mosaic of XMM1229 has a field of view of 10′ × 10′
(2009) for a more detailed description. withanimagequalityof0.34′′andaresolutionof0.1′′/pixel.
The HAWK-Iobservations are summarised in Table 2.
3 http://www.eso.org/sci/software/pipelines/.
4 TheSCAMP and SWarp binaryand sourcefiles can be found
athttp://www.astromatic.net. 5 http://archive.eso.org/cms/eso-data/data-packages/eso-mvm-software-package
Morphological Evolution in XMMU J1229+0151 5
2.3 Ground-based Spectroscopy likelytobetheresultofimproperdeblending(e.g.substruc-
tures in bright nearby galaxies or star formation clumps in
XMM1229 was observed with FORS2 in 2006 during the
distant late-type galaxies) and would contaminate the final
spectroscopic follow-up of SNe Ia in the HST Cluster Su-
catalogue as spurious detections.
pernova Survey. We summarise here the observations and
We used the PSF Extractor (PSFex) software (version
data reduction, referring the reader to Santos et al. (2009)
3.9), written by the Terapix group (Bertin 2011) to model
and Suzukiet al. (2012) for more detailed descriptions.
the PSF in each image7. PSFex selects point sources in the
XMM1229 was observed five times with FORS2 using
halflightradiusvsfluxplaneusingtheSExtractorparame-
the300IgrismandtheOG590ordersortingfilter.Thiscon-
tersFLUX RADIUSandFLUX APERasdiagnosticsofthosetwo
figuration produces aspectral resolution of 2.3 ˚A/pixeland
quantities. The PSF is modelled as a linear combination of
a wavelength coverage extending from 5900 to 10,000 ˚A.
basisvectorsthatcanbechoseneitherusingeachpixelasa
Several SNe were detected in this cluster, and for this rea-
free parameter (pixel basis), as done in this work, or using
son, and to allow the supernovae to be observed both near
actualpointsourceimages(Gauss-LaguerreandKarhunen-
theirmaximumlight and whentheirluminosity hadsignifi-
Lo`evebases).TheFWHMofthePSFsforeachphotometric
cantlyfaded,XMM1229wastargetedmultipletimes.Atotal
band are reported in thefifth row of Tables 1 and 2.
of 77 objects were observed, 26 of which were found to lie
within 3σ from the average cluster redshift (z ∼0.98). The
spectroscopically confirmed members of XMM1229 are rep-
3.2 Photometric Catalogue
resented as red dots in Fig. 4 and are listed in Table 3. Six
galaxies were only detected in the HAWK-I image and we
In order to study the properties of the members of
exclude them from the analysis of the red sequence, as no
XMM1229 across the whole spatial extension of the clus-
colour information is available. The galaxy XMM1229 316
terand given thedifferent areas covered in each band pass,
was also not considered, as SExtractor (Bertin & Arnouts
we decided to split our imaging database into two samples:
1996)failedtoreturnareliableestimateofthei andz
775 850 the centre and the outskirts of the cluster. We produced a
aperture magnitudes (SExtractor FLAGS = 16: data within
multiband catalogue for each sample.
theaperture incomplete or corrupted, see also Table 3).
Inthis process wedefinedthecentralregion astheone
delimitedbytheWFC3observedfield,whichhasthesmall-
est extent. This provided us with 8 photometric bands in
the cluster centre (R, i , z , F105W, F110W, F125W,
775 850
3 DATA ANALYSIS AND MEASUREMENTS
F160W, Ks) and 5 in the cluster outskirts (R, i , z , J,
775 850
Ks). We used the SofI J band only for the analysis of the
3.1 Object Detection and PSF modelling
clusteroutskirtsbecauseitsspectralcoverageoverlapswith
We group the images according to the observing program theoneoftheWFC3F125Wwhich,beingdeeper,wasused
and the instrument with which they were observed. This astheJbandofchoiceinthestudyoftheclustercentre.The
subdivision produces five groups: ACS, WFC3, FORS2, central region extends up to 600 kpc at the redshift of the
HAWK-IandSofI.WeuseamodifiedversionoftheGALA- cluster, which corresponds to 0.54×R (Jee et al. 2011),
200
PAGOSIDLpipeline6 (H¨außler et al.2007)torunSExtrac- whiletheoutskirtsregionapproximatelycorrespondstothe
tor in high dynamic range mode (HDR) on each set of ob- regionbetween0.6Mpcand1.04Mpc,theupperlimitbeing
servations and optimise the number of detections. In fact, imposed by thewidth of theACS field.
whenrunwith lowdetectionthresholds,smalldetectionar- The images of the cluster centre were degraded to the
eas and aggressive deblending, SExtractor can improperly PSF of the R-band image, while in the outskirts they were
split a bright object into a number of sub-components. On degraded to the PSF of the SofI J band, which had the
the other hand, when run with high detection thresholds broadest FWHM. We then ran GALAPAGOS on each of
and large detection areas, the software might fail in de- thePSFmatchedimages, usingtheunconvolvedimages for
blending two closely separated objects that are therefore detection. With the HST observations having been taken
classified as a single source. The HDR technique consists in more than one filter band, we used the z and F110W
850
of two distinct runs of SExtractor: one to detect only the images fordetection in theACSandWFC3 groups,respec-
brightest sources (COLD run) and the other to detect the tively.Inordertoremoveanybiasinducedbyintrinsiccolour
faintest sources and deblend objects with very close neigh- gradients, we measured aperture magnitudes within fixed
bours(HOTrun).Thisisachievedbyvaryingtheparameters circular apertures of 2′′, corresponding to a physical radius
DETECT MINAREA,DETECT THRESHandDEBLEND MINCONT,and R ∼ 8kpc at z = 0.98. With this choice the colour gra-
ap
settingsmallerdetectionareas,detectionthresholdsandde- dients for bright galaxies become negligible, while at low
blending contrasts in the HOTrun, when faint sources and luminosities galaxies are almost entirely contained within
closeneighbouringobjectsaredetected.Thisprocedurepro- theapertureradius. Magnitudes were corrected for galactic
duces two separate catalogues: one for the bright and the extinctionusingthedustmapsofSchlegel et al.(1998)and
other for the faint sources, that are eventually merged into applyingthe Cardelli et al. (1989) equations.
a single catalogue. GALAPAGOSimplements an algorithm In order to consistently compare galaxy colours in
in which the HOT sources that are within a certain distance the cluster centre and outskirts, following Meyers et al.
fromeachCOLDsourcearerejected.Infact,thesesourcesare
7 PSFexcanbedownloaded at:
6 http://astro-staff.uibk.ac.at/ m.barden/galapagos/ http://www.astromatic.net/software/psfex
6 P. Cerulo et al.
Table 1.SummaryoftheHSTobservations ofXMM1229.
F775W(i775) F850LP(z850) F105W(Y) F110W F125W(J) F160W(H)
telescope/instrument HST/ACS HST/ACS HST/WFC3 HST/WFC3 HST/WFC3 HST/WFC3
exposuretime(s) 4160 10940 1312 1112 1212 1112
imagedfield 5.1′ ×5.1′ 5.1′ ×5.1′ 3′ ×3′ 3′ ×3′ 3′ ×3′ 3′ ×3′
pixelscalea (′′/pixel) 0.05 0.05 0.06 0.06 0.06 0.06
imagequality(FWHM)b 0.08′′ 0.09′′ 0.11′′ 0.11′′ 0.13′′ 0.14′′
90%completeness (mag) 25.0 25.0 23.0 23.2 23.3 23.5
a Drizzledpixelscale,see§2.1.1and§2.1.2
b FWHMofthePSFmodelledbyPSFex,see§3.1
Table 2.Summaryofthegroundbasedobservations ofXMM1229.
R SPECIAL(R) J Ks
telescope/instrument VLT/FORS2 NTT/SofI VLT/HAWK-I
exposuretime(s) 1140 2280 11310
imagedfield 8′ ×9′ 5′ ×5′ 10′ ×10′
pixelscale(′′/pixel) 0.252 0.290 0.10
imagequality(FWHM)a 0.63′′ 0.94′′ 0.34′′
90%completeness (mag) 25.3 22.4 24.6
a FWHMofthePSFmodelledbyPSFex,see§3.1
(2012), we performed a cross-convolution of the (uncon- fractions of recovered simulated objects as a function of in-
volved) F775W and F850LP images, in which each image putmagnitude.Thisfractionisadirectmeasurementofthe
was convolved by thePSF of the other image. These are in incompleteness of the photometric catalogues extracted on
fact the two photometric bands that were used for colour eachimage.The90%magnitudecompletenesslimit,quoted
measurementsinthiswork,astheyalmostbracketthe4000 in the sixth row of Tables 1 and 2, is used in this work to
˚AbreakattheredshiftofXMM1229 andarealsothedeep- parametrise completeness in theXMM1229 sample.
est images of the sample (see Fig. 2). The choice of this
strategy for colour measurement allowed us to match the
image qualities of the ACS images without degrading them
to the ground-based level. In fact, after cross-convolution, 4 RESULTS
theresultingimagequalityis∼0.14′′,whichisconsiderably
4.1 Cluster Membership: Photometric Redshifts
narrower than in the ground-based images (see Table 2).
The resampling and co-addition steps of the data re- With the available FORS2 spectra, it is possible to study
duction,aswellasthePSFmatchingprocessintroducecor- the red sequence only down to z = 23.0 (see Fig. 4). In
850
relations between pixels, which are not taken into account order to estimate the distance of fainter galaxies and as-
bySExtractor(seee.g.Casertano et al.2000;Lidman et al. sess their membership to XMM1229, we could rely either
2008; Trentiet al. 2011). Following the method outlined in on statistical background subtraction or photometric red-
Labb´eet al. (2003),we took random regions of skyon each shifts. These two methods are shown to produce compa-
imageandmeasuredthefluxwithindifferentapertures.This rable luminosity functions for red sequence cluster mem-
provideduswithadirectmeasureofthevariationofthesky bers(Rudnicket al.2009).Withnineavailablephotometric
flux with the aperture radius and therefore allowed us to bandsintheXMM1229field,spanningtherange0.65<λ<
quantifythedeviationsfrom apurelypoissonian noise.The 2.2 µm, corresponding to the rest-frame wavelength range
advantage of this method, with respect to the application 0.33<λ<1.1µm,coveringfromthenear-ultraviolettothe
of analytical relations, as those outlined in Casertano et al. near-infrared regions of the spectrum, and given the avail-
(2000),is thatthenoiseisestimated directly ontheimages abilityof77spectra,wedecidedtousephotometricredshifts
andnoassumptionsaremadeaboutthepropertiesofthein- to determine the membership of the cluster. We were able
strumentortheco-additionandresamplingalgorithmsused to determine photo-z’s for objects as faint as z = 24.0
850
in datareduction. on the red sequence, going one magnitude fainter than the
Inordertoquantifythedepthoftheimages,weinserted limitimposedbytheFORS2spectroscopicobservations(see
simulated galaxy images, generated as described in §4.4.3, also Santos et al. 2009) and remaining within the magni-
in empty regions of each science image. We ran SExtractor tudelimitforareliablemorphologicalclassificationatz∼1
on each single image in single image mode with the same (Postman et al. 2005).
configurationusedfortheoriginal imagesandlookedatthe The program zpeg (LeBorgne & Rocca-Volmerange
Morphological Evolution in XMMU J1229+0151 7
Figure 3. (left): Calibration of the zpeg photometric redshift estimate for the cluster centre. On the x-axis it is plotted the value
zspec of the spectroscopic redshifts measured with FORS2 in the cluster centre, while on the y-axis it is plotted the discrepancy
∆z=(zspec−zphot)/(1+zspec).(right):PhotometricredshiftdistributioninthecentralregionofXMM1229.Theverticaldashedlines
arethetwolimitsusedtodefinetheclustermembership.Theredshiftdistributionoftheclustermembersandoftheclusterredsequence
members(hatched histogram)isshownintheinsetplot,wherewealsoreportthecorrespondingvaluesofthegalaxypeculiarvelocities
alongthetophorizontalaxis.
2002) was used to fit synthetic spectral energy dis- tity produced by zpeg is the stellar mass, which is defined
tributions (SED), grouped in seven galaxy types, as the mass locked into stars and is obtained with a me-
and built with the PE´GASE spectral evolution code dianfractionaluncertaintyof∼24%forredsequencecluster
(Fioc & Rocca-Volmerange 1997) assuming a Kroupa members in the central region8. The analysis of the stellar
(2001) initial mass function (IMF). The template types masses of red sequence galaxies in the cluster outskirts will
cover a wide range of spectral classes, going from passive bepresented in a forthcoming paper.
to actively star-forming galaxies. zpeg implements a χ2 Delayeetal.(2013,submitted)studiedthestellarmass
minimisation procedure, in which the best fitting SED is vs size relation in the HCS clusters. They estimated stel-
the one which minimises the χ2 in a three-dimensional larmasses usingthelephare software (Arnoutset al.1999;
parameter space of age, redshift and template type. The Ilbert et al. 2006) on a set of synthetic SEDs from the
metallicity of the synthetic SEDs is assumed to evolve Bruzual & Charlot(2003)librarywiththreedifferentmetal-
with time according to the star formation history of each licities(0.2Z⊙,0.4Z⊙,Z⊙),exponentiallydecliningstarfor-
template and with the stars forming at the metallicity of mation histories and a Chabrier (2003) IMF. As a consis-
theinterstellar medium. Nodust extinction is assumed. tencycheck,weestimatedthestellarmassesofredsequence
galaxiesinXMM1229usinglephareonthesamesetoftem-
We used the available spectroscopic redshifts to cal-
platesofDelayeetal(2013)andadoptingtheircosmology9.
ibrate the photo-z’s obtaining a median ∆z = (z −
spec
As in Delaye et al. (2013), we also fixed the redshift of red
z )/(1 + z ) = 0.0 ± 0.05 in the cluster centre and
phot spec
sequence galaxies at z = 0.98 and we found that the so
∆z = 0.02±0.11 in the outskirts, where the uncertainty
obtained stellar masses were a factor of 1.2 smaller. This
is computed as the normalised median absolute deviation
small differencecan be attributed to both thedifferent sets
(NMAD, see also Fig. 3, left panel). In both samples we
of photometric bands used in the two works10 and the dif-
fixed z =z for the spectroscopically confirmed clus-
phot spec
ferentaperturesusedforgalaxyphotometry,asDelayeetal.
termembersandletzpegperformtheSEDfittingwithonly
(2013) usedMAG AUTOmagnitudesinsteadoffixedaper-
age and template typeas free parameters.
turemagnitudes.Whencomparingwithourzpegestimates,
Clustermembercandidatesweredefinedasthosegalax-
wefoundamedianratioof1.35betweenDelayeetal.(2013)
iesintherange0.8<z <1.2,thecutbeingchosenfrom
phot and this work. Since this difference did not affect the con-
thewidthofthephoto-zdistribution(seeFig.3).Although
clusionsofthispaperandinordertobeconsistentwiththe
the median fractional photo-z error for all galaxies in the
range0.8<z <1.2is12%,wedecidedtofocusonlyon
phot
the red sequence, as the photo-z estimate of blue galaxies
8 ThereadercanrefertoBernardietal.(2010)forconversionsto
is expected to be significantly uncertain with the available
stellar masses obtained with the most commonly used Chabrier
spectral coverage. In fact, the R band, which is the bluest
andSalpeterIMFs.
for the XMM1229 samples, does not cover the blue side of 9 Delayeetal.(2013)useaΛCDMcosmologywithH0=70km
the 4000 ˚A break at z < 0.6 and foreground galaxies may ·s−1 ·Mpc−1,ΩM =0.30,andΩΛ=0.70
be misclassified as blue cluster members. A second quan- 10 Delayeetal.(2013)usei775,z850,J (fromSofI)andKs.
8 P. Cerulo et al.
cosmology chosen for this work, we kept our stellar mass and 22.5 6 z < 24.0 with the same colour width 0.7 6
850
estimates. (i −z ) < 1.1. This choice was motivated by the fact
775 850
that the cells sample the observed red sequence with suf-
ficient resolution without falling into a regime of excessive
4.2 Contamination from Field Interlopers
low-numberstatistics (seeFig. 4). Furthermore,themagni-
tude limits are the same adopted for the estimation of the
In order to estimate the contamination of the XMM1229
luminoustofaint ratio in§5. Theaverage P fortheob-
red sequence, we used the HST/ACS F775W and F850LP field
servedXMM1229 redsequenceisP ∼6%inthecluster
images of the two fields of the Great Observatories Ori- field
centreand P ∼32% in the outskirts.
gins Deep Survey (GOODS, Giavalisco et al. 2004, ver- field
With the definition of field contamination given in
sion 2.0). These images are very similar to those of
Equation (1), the number of cluster members N in
XMM1229, the only remarkable difference being their cluster
resolution (0.03′′/pixel for GOODS vs 0.05′′/pixel for each colour-magnitude cell is definedas:
XMM1229). N =(1−P )∗(N −N ).(2)
cluster field XMM1229 XMM1229,cont
GOODS is a deep astronomical survey centred on
We use this equation to correct for outliers in the esti-
two fields: the Hubble Deep Field North (GOODS North)
mation of the luminous-to-faint ratio (see §5).
and the Chandra Deep Field South (GOODS South). The
projectwasaimedatcollectingdeepmultibandphotometry
from various space- and ground-based facilities (e.g. HST, 4.3 Colour-Magnitude Diagram and Red
Spitzer, Chandra and XMM-Newton, VLT, KPNO, Sub-
Sequence
aru), in order to accurately study the properties of distant
galaxies.Spectroscopicfollow-upobservationswerealsocar- The colour-magnitude diagrams for the cluster centre and
riedoutatKeckandVLT(Wirth et al.2004;Vanzella et al. outskirtsarepresentedinfigure4.Inordertomodelthered
2005,2006,2008; Popesso et al. 2009; Balestra et al. 2010), sequence, we applied a robust linear fit, implementing the
resulting in an extensive ensemble of imaging and spectro- Tukey’sbi-squareweight function,and restricting thefit to
scopic data covering ∼300 square arcminutes thephotometrically confirmed membersin thecolour range
AsforXMM1229,wefollowedthemethodoutlinedin§3 0.75 < (i775 −z850) < 1.5 at z850 < 24.0. The model red
toprocesstheGOODSACSimagesandmodelthePSF.The sequencewas definedas:
90% magnitudecompletenesslimits, estimatedasdescribed
(i −z ) =a+b×(z −21.0) (3)
775 850 RS 850
in §3.2, are i = 27.3 mag and z = 26.7 mag,
775,lim 850,lim
about two magnitudes deeper than in the XMM1229 field. where b is the slope and a represents the colour of a
The i775 −z850 colours were again measured on the cross- galaxy on the red sequence at z850 = 21.0. We obtained
convolved images within 2′′ fixedapertures. b=−0.044±0.017 and a=0.94±0.03 for the cluster cen-
We adopted the method outlined in Pimbblet et al. tre, where the uncertainties on a and b were estimated by
(2002) to estimate the fraction of galaxies contaminating generating1000bootstrapsamplesfromthephotometrically
theXMM1229 redsequence,modifyingtheequationin Ap- confirmed members used to fit the red sequence. Following
pendix A of their paper to take into account the galaxies Lidman et al.(2004)andMei et al.(2009),weestimatedthe
with assigned spectroscopic redshift in the XMM1229 field. intrinsic scatter σc of the red sequence as the scatter that
This method assigns to each galaxy within a certain range needed to be added to the colour error to have χe2 = 1.0,
of magnitude and colour a probability of belonging to the where χe2 is the reduced χ2. We found σc = 0.026±0.012,
field. As a result, this probability quantifies the amount of wheretheuncertaintywasagainestimatedbycreating1000
contaminationoftheobservedcolour-magnitudediagramin bootstrap samples from the sample of photometrically con-
theclusterfield.Wedividedthecolour-magnitudeplanesof firmedmembersusedintheredsequencefit.Wediscussthe
the GOODS and XMM1229 fields into cells of equal width implications of theseresults in §5.4.
in colour and magnitude and in each cell we computed the AsshowninTable2,theSofIJband90%completeness
probability for each galaxy of being in thefield: limitisJ =22.4.Thisresultsintoalossofredsequenceob-
jectsat magnitudes z >22.5 in thephoto-zselected out-
N ×A−N 850
P = field XMM1229,cont (1) skirts sample. Therefore, in order to study the cluster out-
field N −N
XMM1229 XMM1229,cont skirts.wefirstmodelledtheobservedredsequenceandthen
where N is the number of galaxies in each cell of the used Equation (2) to statistically subtract field interloper
field
GOODScolour-magnitudediagram,N isthenum- galaxies. We obtained the following result: a=0.88±0.05,
XMM1229
berofgalaxiesineachcelloftheXMM1229observedcolour- b=−0.01±0.03 andσ =0.052±0.015. Itcanbeseenthat
c
magnitudediagram,N isthenumberofgalax- thered sequenceis shallower than in thecluster centre and
XMM1229,cont
iesineachcelloftheXMM1229observedcolour-magnitude has a larger intrinsic scatter, although the slopes and the
diagram which are known not to be in the cluster from intrinsic scatters for thetwo samples are still consistent.
their spectroscopic redshift, and A is the ratio between In the cluster centre the red sequence population was
the areas of the XMM1229 field and the GOODS field. definedasthosegalaxieswith−4σ <(i −z )−(i −
c 775 850 775
The main drawback of this method is that one can have z ) <+7σ ,wherethelimitswerechosenaftervisually
850 RS c
P > 1 or P < 0. As suggested by Pimbblet et al. inspecting thecolour-magnitude diagram, as themost suit-
filed filed
(2002), in these cases the width of the cell is adjusted until able to bracket the red sequence. For the cluster outskirts
0.0<P <1.0. we used theboundaries −3σ and +4σ .
field c c
Forthepurposeofthispaperweonlyfocusedonthered We defined the total galaxy magnitude as the
sequenceand wesplit it intotwocells at 21.06z <22.5 extinction-corrected SExtractor MAG AUTO, although we are
850
Morphological Evolution in XMMU J1229+0151 9
awarethatthisquantityunderestimatesthetotalflux,espe- z∼1clustersanditallowsustoinvestigatethemainstruc-
ciallyforellipticalandlenticulargalaxies.Graham & Driver tural features of cluster members.
(2005) proposed general aperture corrections based on MostS0sfallintotheclassofbulge-dominatedgalaxies,
galaxy light profiles. However, in order to apply them to while the early and late disc classes comprise Hubble types
systems composed by a bulge and a disc, like S0 galaxies, goingfrom SatoSbcandSctoScd,respectively.Inthefol-
one should first perform a bulge-disc decomposition which lowingwewillusetheterms‘bulge-dominated’andS0inter-
at z∼1becomes highly uncertain dueto thehigh sky con- changeably,althoughweareawarethatwithourschemethe
tamination at low fluxes. bulge-dominatedsamplemaybecontaminatedbySagalax-
The photometric selection of the red sequence in the ies (see e. g. Mei et al. 2012) and the disc-dominated sam-
cluster centre produced a sample of 45 galaxies to which ples may be contaminated by low (B/T) S0 galaxies (S0
c
we added the bright spectroscopically confirmed member galaxies, Laurikainen et al. 2011). In fact, spiral arms be-
XMM1229 145 (z850 = 21.9), falling just below the red se- come fainter in the gas-poor galaxies which are typical of
quence. This brought the final number of objects analysed thered sequence population.
intheclustercentreto46.Theredsequencemembersofthe Thefourmorphologicalclassificationsoftheclustercen-
central region are listed in Table 3. We did not assign clus- treagreedonlyfor4galaxies,while11galaxieshadthesame
ter membership to the single galaxies in the outskirts, as a type assigned in the three visual classifications. In particu-
detailedstudyoftheclusteroutskirtsintheHCSwillbepre- lar, we note that while W. J. C. and C. L. agreed on 33 of
sentedinaforthcomingpaper.ForthisreasoninTable3we the46red sequencegalaxies, P.C.agreed with eachauthor
only report the 3 spectroscopically confirmed red sequence only on 16 and 17 galaxies, respectively. The main points
membersinthissubsample.Oursamplealmost doublesthe ofdisagreementweretheE/S0distinctionandthefactthat
number of red sequence galaxies analysed in Santos et al. faint elliptical galaxies tended to be classified by P. C. as
(2009),wholimitedthemselvestothespectroscopicallycon- disc-dominated systems with a very compact bulge and a
firmedclustermembers.Thisallowsustostudygalaxyprop- faint disc. This underlines the challenges in morphological
erties along the red sequence down to z850 =24.0 mag, i.e. classifications ofhigh redshiftpassivegalaxies. Forthisrea-
1 mag fainter than in that work. son we decided to adopt a majority rule for theassignment
of the morphological types and thus the final morphologi-
cal type was defined as the mode of the four independent
4.4 Galaxy Morphology and Structure classifications.Therewerefourgalaxiesforwhichtwoofthe
classifiers assigned the type E and the other two the type
Wehaveshownin§4.2thattheoutskirtsofXMM1229have
BD, while in two other cases two of the classifiers assigned
a considerably large outlier contamination and therefore in
the type BD and the other two the type EDD. In these
thispaperwerestrictthemorphologicalandstructuralanal-
six cases we assigned type E to the 4 galaxies with no ma-
ysesofredsequencegalaxiestotheclustercentre.Themor-
jority between type E and type BD and type BD to those
phologicalandstructuralanalysesoftheclusteroutskirtsin
with nomajority between typeBD and typeEDD.For one
the entire HCS sample will be the subject of a forthcoming
galaxy (XMM1229 322) two classifiers assigned type E and
paper.
theothertwo typeEDD.Inthiscase we decidedtoclassify
thegalaxy as bulge-dominated. Given the verylow number
of objects with late-typedisc and irregular morphologies in
4.4.1 Morphological Classification
both the XMM1229 and the two lower-redshift comparison
We classified red sequence galaxies in the centre of samples(see§4.5),wedecidedtomergethetwoclassesinto
XMM1229usingtheACSF850LPimage,onwhichwewere one “late disc-dominated / irregular” morphological class.
able to detect morphological features down to z = 24.0. However, for completeness, in Tables 3, 4 and 5 we report
850
TheF850LPfiltercorrespondsapproximatelytoarestframe theoriginal morphological schemewith fiveclasses.
SDSS g band, allowing us to compare directly with lower- ThetoppanelsofFig. B1showthat,asexpected,disc-
redshiftclassificationsperformedeitherintheBorVbands dominated galaxies tend to have lower values of concentra-
(e.g. Fasano et al. 2012). Galaxies were classified by three tion and Gini coefficient with respect to elliptical and S0
of the authors independently (P. C., W. J. C. and C. L.) galaxies(seee.g.Lotz et al.2004).Wealsonotethattheval-
on image cutouts whose size varied according to the SEx- ues of M20 (lower left panel of Fig. B1) for disc-dominated
tractor Kron radius of each object. We also ran the galSVM galaxies are comparable with those of early-type galaxies.
software(seeHuertas-Company et al.2008,2009a,2011and We interpret this result as a consequence of the fading of
Appendix B) on the entire F850LP image, which provided spiral arms in gas-poor spiral galaxies.
uswithafourthindependentandquantitatively-basedclas- The morphological classifications for the cluster cen-
sification. tre and outskirts are reported in Table 3 (column 6), while
Becauseatz=1manyhigher-ordermorphological and thumbnailimagesforthegalaxieslistedinthattablecanbe
structural features, such as spiral arms, bars and lenses are foundinFig.A1-A4.Themorphologyquotedforthecluster
not resolved, we classified galaxies according to their ob- outskirts corresponds only tothe outputof galSVM.
servedbulge-to-totalratio(B/T) andsplittheredsequence Santos et al. (2009) classified visually red sequence
sampleintofivebroadmorphologicalfamilies:elliptical(E), galaxiesinXMM1229usingaschemesimilar toours:ofthe
bulge-dominated (BD), early disc-dominated (EDD), late 15galaxiesincommonwithourclustercentresample,there
disc-dominated(LDD)andirregulargalaxies(Irr).Asimilar are 9 galaxies with the same assigned morphological type.
classification scheme was adopted by Postman et al. (2005) Whencomparingonlyearly-andlate-typegalaxies, wefind
and Mei et al. (2012) to classify galaxies in ACS images of that 13 of the15 galaxies havethesame assigned type.De-
10 P. Cerulo et al.
Figure4.(left):Colour-magnitudediagramofthecentralregionofXMM1229.ThecoloursaremeasuredontheF775WandF850LPPSF
cross-convolvedimages,adoptingfixedcircularapertureswith1′′radius(∼8kpcatz=0.98).Blackdotsrepresentallthephotometrically
selectedclustermembersandgreendotsaretheredsequence members.Reddotsarespectroscopicallyconfirmedclustermembers.The
verticaldashedlinerepresentsthefluxlimitofvisualmorphology(see§4.4)andtheslopingdottedlineisthe90%completenesslimitin
the F775W andF850LP images. Theblackdashed lineis thelinearfit tothe redsequence andthe twodotted linesrepresentthe -4σc
and+7σc envelopesdelimitingtheredsequence.Theerrorbarsrepresentthemediancolourerrorsalongtheredsequenceinbinsof0.5
magnitudes. (right): Observed colour-magnitude diagram in the outskirts of XMM1229. No cut in photometric redshifts is applied for
thissample.Greydotsareallthegalaxiesobservedintheclusteroutskirts,reddotsarespectroscopicallyconfirmedclustermembersand
greendotsareredsequence galaxies selectedasdescribedin§4.3.Errorbarsrepresentthemediancolourerrorsalongthe redsequence
inbinsof0.5magnitudes.Themeaningofthelinesisthesameoftheleftpanel.
layeetal(2013)usedgalSVMtoclassifygalaxiesintheHCS out ambiguity the extreme cases F = 1 and F = 0. In
T T
clusters, dividing them into early- and late-type. The com- fact,agaussian approximationofthebinomialdistribution,
parison with ourclassification showsthat ofthe45 galaxies suitableforlargesamples,wouldproducenullerrors,mean-
in common to bothsamples, 38(i.e. 84%) were classified as ing certain estimates of thetruevalue of the morphological
early or late-typein both works. fraction. We dealt with these two extreme cases defining as
bestestimateofF themedianoftheaposterioriprobabil-
T
itydistributionandweestimatedtheerrorsasthedifference
4.4.2 The Morphological Composition of the Red between this value and the upper and lower bounds of the
Sequence. binomial confidence interval.
Themorphologicalfractionsasafunctionofmagnitude
We divided the red sequence of the central region into bins
along the red sequence in XMM1229 are illustrated in the
of 0.5 magnitudes in the range 21.0 < z < 24.0 and, in
850
top panels of Fig. 5 and will be discussed in §5.
eachbin,wecomputedthefractionofthegalaxypopulation
of a certain morphological type,F , as:
T
N 4.4.3 Light Profile Fitting and Structural Parameters
F = T,i (4)
T,i N
tot,i In order to investigate the connection between morphologi-
where F , is the fraction of galaxies of type T in the ith calandstructuralpropertiesinredsequencegalaxies,wefit
T,i
bin,N is thenumberof galaxies of typeT in theith bin, aS´ersicfunctiontothelightdistributionoftheredsequence
T,i
and N is thetotal numberof galaxies in theith bin. members.WeusedGALFIT(Penget al.2002,2010),imple-
tot,i
The error bars were computed following the method mentedaspartof theGALAPAGOSpipeline,onthewhole
outlinedinCameron(2011)toestimatetheconfidenceinter- F850LPimage.TheS´ersicLawisparametrisedbytheequa-
vals for a binomial probability distribution. More precisely, tion:
ineachmagnitudebin,theconfidenceintervalwasestimated Σ(r)=Σ e−κ[(r/re)(1/n)−1] (5)
using a Bayesian approach in which the binomial probabil- e
itymassfunctionwastreatedastheaposterioriprobability whereΣ(r) is thegalaxy surface brightnessas afunction of
distribution, given a uniform prior over the expected num- projectedradius,r isthehalflightradius,Σ isthesurface
e e
ber of successes. The errors on the morphological fractions brightnessatthehalflight radius, nistheS´ersic indexand
werethereforeestimatedasthedifferencebetweenthemea- κisaparameterwhichiscoupledwithninsuchawaythat
sured valueof thefraction and theupperandlower bounds halfofthetotallightisalwaysenclosedwithinr .TheS´ersic
e
of theconfidenceinterval.This method is shown togive re- indexis related togalaxy light concentration: higher values
liableconfidenceintervalsevenwithsmallsamples,asinthe of n correspond to more concentrated light profiles. There-
case of this paper. Furthermore, it allows us to treat with- fore, spheroidal galaxies are expected to have higher S´ersic