ebook img

Morphologies and stellar populations of galaxies in the core of Abell 2218 PDF

26 Pages·2007·4.08 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Morphologies and stellar populations of galaxies in the core of Abell 2218

Mon.Not.R.Astron.Soc.376,125–150(2007) doi:10.1111/j.1365-2966.2007.11335.x Morphologies and stellar populations of galaxies in the core of Abell 2218 (cid:2) S. F. Sa´nchez,1 N. Cardiel,1,2 M. A. W. Verheijen,3 S. Pedraz1,2 and G. Covone4 1CentroAstrono´micoHispanoAlema´n,CalarAlto(CSIC-MPI),C/Jesu´sDurba´nRemo´n2-2,04004-Almer´ıa,Spain 2DepartamentodeAstrof´ısica,FacultaddeF´ısicas,UniversidadComplutensedeMadrid,28040Madrid,Spain 3KapteynAstronomicalInstitute,POBox800,9700AVGroningen,theNetherlands 4INAF–OsservatorioAstronomicodiCapodimonte,Naples,Italy Accepted2006November20.Received2006November14;inoriginalform2006June21 D o w n lo ABSTRACT ad e d Wepresentastudyofthestellarpopulationsandmorphologiesofgalaxiesinthecoreofthe fro galaxyclusterAbell2218.Integralfieldspectroscopy(IFS)observationswereperformedusing m PMASinthePPAKmodecoveringafieldofviewof∼74×64arcsec2centredonthecoreofthe http cluster,inordertoobtainspectroscopyofanunbiasedflux-limitedsampleofclustergalaxies. ://m n 43objectsweredetectedintheIFSdata,31ofthemwithenoughsignal-to-noiseratiotoderive ra s the redshift, all of them brighter than I < 21.5 mag. 28 are at the redshift of the cluster (17 .ox withpreviouslyunknownredshift).Individualspectraoftheclustermemberswereextracted ford and compared with single stellar population models to derive the luminosity-weighted ages jou rn and metallicities. In addition, deep Hubble Space Telescope (HST)/ACS F475W-, F555W-, a ls F625W- and F850LP-band images centred on the cluster core were obtained from the HST .org archive(z ∼28mag).Adetailedmorphologicalanalysisofallthegalaxieswithinthefield a/ ofviewoflitmheseimagesdowntozlim<22.5magwasperformedclassifyingthemaslate-type, t UN intermediateandearly-type,onthebasisoftheirSe´rsicindices.Theliteraturewasscannedto IV E R lookforspectroscopicallyconfirmedclustermemberslocatedwithinthefieldofviewofthe S ID ACSimage.Thefinalsampleof59galaxiescomprisesourreportedsampleof28galaxiesin A D the core, and 31 additional galaxies in the outer regions. In addition, multiband broad-band C O photometrywasextractedfromtheliteratureforallobjects. M P Themorphologicallysegregatedcolour–magnitudediagramshowsthattheearly-typegalax- L U iescovertherangeofbrighterandreddercolours(theso-called‘redsequence’).Alargefraction T E ofspiralgalaxies(∼50percent)isfound,mostofthemfainterthanthelimitofpreviousstudies. NS E Theycoverawiderangeincolours,frombluecolourstypicalofButcher–Oemlergalaxiesto D E redcolourssimilartothoseofearly-typegalaxies.Thisresultindicatesthatearly-typegalaxies M A aremoremassiveandhaveolderstellarpopulations,whilelate-typegalaxiesarelessmassive D R and have a wider range of stellar populations. The distribution of luminosity-weighted ages ID asafunctionofmetallicitiesandluminosity-weightedmasses,andthedistributionofSe´rsic on J indices as a function of the luminosity-weighted masses confirm these results. They in fact un e agreewithaproposedtwo-stepscenariofortheevolutionofgalaxiesinclusters,wherethestar 1 4 formationisquenchedfirstintheinfallingspirals,afterwhichamorphologicaltransformation , 2 0 1 followsthatrequireslargertime-scales.Thisscenarionaturallyexplainsthepopulationfaint 3 late-typegalaxieswitholdstellarpopulationsobservedinthiscluster.Inaddition,anextremely bluemerginggalaxysystemisfoundatthecore,withthenominalredshiftofthecluster. Keywords: techniques:spectroscopic–galaxies:clusters:individual:Abell2218–galaxies: ellipticalandlenticular,cD–galaxies:fundamentalparameters–galaxies:spiral–galaxies: stellarcontent. 1 INTRODUCTION Galaxyclustershavebeenusedfordecadestostudytheevolution ofgalaxies.Beingtracersofthelargestdensityenhancementsinthe (cid:2)E-mail:[email protected] Universe, they are considered the regions where galaxies formed (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS 126 S.F.Sa´nchezetal. first.Atlowredshifttheyaredominatedbylargeearly-typegalaxies trographmultislitsormasks.Inparticular,theblendingofdifferent (E/S0),whichcoloursareconsistentwithabulkformationoftheir components in the core of the clusters is a problem, impossible starsathighredshift.Classically,itwasthoughtthattheevolution tohandlewithclassicalslitspectroscopy.Duetothat,mostofthe ofthisdominantpopulationthroughdifferentcosmologicalepochs spectroscopicsurveysarebasedonpre-selectionoftheclustercan- waslargelypassivewithoutsignaturesofimportantdeviations(i.e. didates,mostofthembasedoncolourormorphologicalselections periodsofmorerecentstarformation).However,thissimplepicture (e.g.Baloghetal.1999;Dressleretal.1999;Kelsonetal.2000; haschangedinthelastdecades.Butcher,Wells&Oemler(1983) Ziegler et al. 2001; Fritz et al. 2005). Alternatively, they are fo- andButcher&Oemler(1984)showedanincreaseinthefraction cusedongalaxiesinthelesscrowdedouterregions.Theseselections ofbluegalaxiesinclusterswithredshift.High-resolutionground- mayaffectsubstantiallytheconclusions.Forthesereasonswehave basedandHubbleSpaceTelescope(HST)imageshaveshownthat startedanobservationalprogrammeaimedatobtainingintegralfield theycorrespondtogalaxieswithlate-typemorphologies,oftendis- spectroscopy(IFS)ofthecoreofsomeoftherichestgalaxyclusters turbedorwithclosecompanions(e.g.Thompson1986;Thompson inthelocaluniverse,includingAbell1689(ESO73.A-0605)and 1988;Couchetal.1994;Dressleretal.1994,1997).Oneobserves Abell2218.ForbothofthemdeepHST/ACSimagesareavailable. D amorphologicalevolutionofthegalaxiesinclusters,inthesense AsimilarapproachhasbeenusedbeforebyCovoneetal.(2006), o w thatthefractionofE-typegalaxiesremainsconstantasafunctionof tostudythedynamicalmassoflowredshiftgalaxyclusters. nlo differentcosmologicalepochs,whilethefractionofS0’sincreases Abell2218isoneoftherichestclustersintheAbellcatalogue ad e a2t0t0h0e).expenseofL-typestowardslowerredshift(e.g.Fasanoetal. (aArbeedlslh,iCftoorwfizn∼&0O.1lo7w, ainnd19a8v9e),lowciittyhadirsipcehrnseiossncolfas1s3o7f04k.mIths−as1 d from The presence in clusters of a significant population of post- (Kristian,Sandage&Westphal1978;LeBorgne,Pello´&Sanahuja h starburstgalaxiesknownas‘E+A’or‘k+a’galaxies(Dressler 1992).Theclustershowsasetofarclenses.Detailedhigh-resolution ttp://m & Gunn 1983; Couch & Sharples 1987), indicates that star for- X-raymaps(McHardyetal.1990)andmass-concentrationstudies n mation was active in the recent past and sharply halted at some basedonthepropertiesofthegravitationallenseshaveshownthat ras .o point during the last 1–1.5 Gyr (Poggianti 2003; Poggianti et al. theclusterhastwodensitypeaks,thestrongestofthemdominatedby x fo 2004). Most of these galaxies are late-type, and they show a ra- alargecDgalaxy.Severalobservationalprogrammeshaveprovided rd dialdistributionwithintheclustersthatisintermediatebetweenthe uswithalargedatasetofwell-deblendedslitspectraforthegalaxies jou passiveandtheemission-linegalaxypopulations(Poggiantietal. intheouterpartofthecluster(Ziegleretal.2001),aswellasvery rna ls 1999).Theseresultsleadtothehypothesisofatwo-stepevolution goodmultibandground-basedandHSTimaging(LeBorgneetal. .o rg ofgalaxies.Aspiralgalaxyiscapturedbytheclusterand,during 1992;Smailetal.2001;Ziegleretal.2001).Wefocusedoursurvey a/ itsinfall,suffersfromgas-strippingduetothetidalforcesinvolved onthecentralarcminregionofthecluster,aroundthecDgalaxy. t U N (e.g.Verheijen1997).Provokedbythelackofgas,thestarforma- Theoutlineofthisarticleisasfollows.Section2describesthe IV tionishalted,andthegalaxyevolvespassively,beingobservedasa observationaldataset,includingadescriptionoftheIFSdataandthe E R ‘k+a’post-starburst.Aninitialshortstarburstprocess,previousto HST/ACSimages.Section3includestheanalysisofthedatasetand S ID thegas-stripping,maybealsopresent(Poggiantietal.2004).After presentstheresults.Here,Sections3.1and3.2describetheauto- A D thisthegalaxyevolvesmorphologicallyoverlargertime-scales.The maticmorphologicalanalysisthatweperformedasopposedtoclas- C discdimsduetothelackofanewgenerationofstars,thatotherwise sicalmethods.Section3.3showsthemethodusedtodeblendandex- O M willdominateitsluminosity,andthegalaxytransformsitselfinto tracttheintegratedspectraofthedifferentdetectedobjectsfromthe P L alenticulargalaxy.Drymergersbetweenlenticularandspheroidal IFSdata,withtheredshiftdeterminationexplainedinSection3.4. U T galaxiesbuildupnewlargespheroidals(e.g.Belletal.2006). Theanalysisofthemorphologicallysegregatedcolour–magnitude E N The proposed scenario suggests that a fraction of the observed (CM)diagramisincludedinSection3.5.Theluminosity-weighted SE S0’s were previously star-forming spirals. For these galaxies the ages and metallicities of the galaxies, and their relation to their D E starformationwastruncatedatsometimebetween2and5Gyrago massesandmorphologiesarediscussedinSection3.6and3.7.The M anditisexpectedthattheirluminosityfadedby∼0.5–1.5magand resultsarediscussedinSection4,andtheconclusionsarelistedin AD becomeredder(Poggiantietal.2001),populatingtheredsequence Section 5. Throughout this article we have used a (cid:3)-cosmology, RID atmagnitudesfainterthanthebrightest1or2mag(Poggianti2003). withH0=70kms−1Mpc−1,(cid:4)m=0.3and(cid:4)(cid:3)=0.7. on Thesameprocessmayaffectdwarfgalaxiescapturedbytheclus- J u n ter,whichwillsuffermorestronglytheeffectsoftidalinteraction e 2 OBSERVATIONAL DATA SET 1 andgalaxyharassment,whichinducesashortbutintensestarburst 4 , 2 (Mooreetal.1996;Rakos,Dominis&Steindling2001).Mostofthe 0 fielddwarfgalaxiesaregas-rich,bluegalaxieswithlate-typemor- 2.1 IFSobservationsanddatareduction 13 phologiesatanyredshift(Cairo´setal.2001;Barazzaetal.2006). Observations were carried out on 2005 June 30 and July 6 at the Oncecapturedbythecluster,consumedorstrippedofitsgas,itis 3.5-mtelescopeatCalarAltoObservatorywiththePotsdamMul- expected that their luminosity fades and they become redder too, tiAperture Spectrograph, PMAS, (Roth et al. 2005) in the PPAK populating the faint-end of the red sequence. In both cases ones mode(Verheijenetal.2004;Kelzetal.2006).TheV300grating expectstofindapopulationofredlate-typegalaxies,fainterthan wasused,coveringawavelengthrangebetween4650–8000Åwith theoriginalearly-typegalaxiesinthecluster.However,mostofthe aresolutionof∼10Åfullwidthathalf-maximum(FWHM)anda previousstudieshavefoundthatlate-typegalaxiesaremostlyblue samplingof3.2Åpixel−1.ThePPAKfibrebundleconsistsof382 (e.g.Thompson1986). fibresof2.7arcsecdiametereachone.Ofthem,331fibres(thesci- Totestthishypothesis,completespectroscopicsurveysofclus- encefibres)areconcentratedinasinglehexagonalbundlecovering tergalaxiesandhigh-resolutiondeepimagesforapropermorpho- afieldofviewof72×64arcsec2,withafillingfactorof∼65per logicalanalysisofthefaintestpossibletargetsarerequired.Com- cent.Theskyissampledby36additionalfibres,distributedinsix pletespectroscopicsurveysaredifficultandtimeconsuming,due bundlesofsixfibreseachone,locatedfollowingacirculardistribu- totherequirementofdeblendingandplacingobjectsinthespec- tionat∼90arcsecfromthecentreandattheedgesofthecentral (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS,MNRAS376,125–150 ThegalaxiesinthecoreofAbell2218 127 hexagon.Thesky-fibresaredistributedbetweenthescienceonesin apseudo-slit,inordertohaveagoodcharacterizationofthesky.The remaining15fibresareusedforcalibrationpurposes,asdescribed below. Foreachnight,threeditheredexposuresof1heachweretaken, followingapatternof(cid:5)RA=+1.56arcsecand(cid:5)Dec.=±0.78 arcsec, in order to have a complete coverage of the science field of view and to increase the spatial resolution. Each pointing was dividedintoshorterindividualframesforapropercosmicrayre- moval.Atotalof6hintegrationtimewasachievedaftercombining theexposuresofthetwonights(3hpernight). Data reduction was performed using R3D (Sa´nchez & Cardiel 2005;Sa´nchez2006),incombinationwiththeIRAFpackage(Tody D 1993)1andE3D(Sa´nchezetal.2004).Thereductionconsistsofthe ow standardstepsforfibre-basedintegral-fieldspectroscopy.Amaster nlo biasframewascreatedbyaveragingallthebiasframesobserved Figure1. Averagenight-skyemissionspectrumderivedbycombiningthe ad e dcautriionngothfethneigshptecatnrda isnubthtreacCteCdDfrwomastdheetesrcmieinnceedfurasimngesa. Tchoentlion-- snpigehcttr.aobtainedthroughthe36sky-fibresforthedifferentpointingsofeach d from uumilluminatedexposuretakenbeforethescienceexposures.Each h ttp spectrumwasextractedfromthescienceframesbyco-addingthe ://m flux in 5 pixels in the perpendicular direction of each fibre, and 2.2 HST/ACSimagesandliteraturedata n stored in a row-staked-spectrum file RSS (Sa´nchez et al. 2004). ras WeobtainedpubliclyavailableHST/ACSimagesoftheclusterin .o WavelengthcalibrationwasperformedusingaHelampexposure x theF475W,F555W,F625W,F775WandF850LPbandsfromthe fo taken at the beginning of the night, and corrected for distortions rd usingThArexposuresobtainedsimultaneouslywiththescienceex- HSTarchive.Table1summarizestheirmaincharacteristics,includ- jou posuresthroughthecalibrationfibres(indicatedabove).Thefinal ingthecentralwavelengthandwidthoftheirbands(obtainedfrom rna Sirianni et al. 2005). They basically correspond to images in the ls accuracyofthewavelengthcalibrationwasestimatedbycomparing .o classicalB,V,R,Iandzbands.Theimageswerecalibratedtothe rg thecentralwavelengthoftheskyemissionlineswiththeirnominal Vega system using the corresponding magnitude zero-points and a/ values,findinganagreementwithinarangeof∼0.3Å.Differences t U prescriptionslistedinSiriannietal.(2005).Table1alsoliststhe N inthefibre-to-fibretransmissionthroughputwerecorrectedforby numberofindividualframesacquired,eachonewithadifferentori- IV comparingthewavelength-calibratedRSSscienceframeswiththe E entationandaslightlydifferentpointing,duetotheobservingstrat- R correspondingcontinuumilluminatedimages.Fluxcalibrationwas S egyoftheHST/ACScamera(inordertocoverthegapinbetween ID performedusingstandardcalibrationstarsobservedthroughoutthe A night (BD33 + 2642 and Hz44). The uncalibrated spectra of the tcheesstewdobyditfhfeerHenStT/CACCDSsr)e.dTuhcetioancqpuipireeldinfer,awmheischwienrceluadleresabdiyasparnod- D C starsextractedfromthereducedframeswerethencomparedwith O darkcurrentcorrection,flat-fielding,quantumefficiencycorrection, M publiclyavailablecalibratedspectra,determiningafluxcalibration P field-distortioncorrection,thederivationoftheastrometricsolution L correctionthatwasappliedtothescienceframes,oncecorrectedby U and flux calibration. We correct each individual from for cosmic T thedifferencesinexposuretime.Theaccuracyofthespectropho- E rays,realignedtothestandardnorth-eastandfinallycombinedinto N tometric calibration was ∼0.15 mag across the wavelength range SE coveredbyourdataset,asdeterminedbycomparingwithhighac- asingleimageperfilterafterasky-backgrounddeterminationand D subtraction.Thefinalexposuretimeandlimitingmagnitudeofeach E curatebroad-bandphotometry,aswewilldiscussbelow.Anight-sky M image,estimatedbylookingforthefaintestobjectdetectedat∼15σ A emissionspectrumwasobtainedforeachpointingbycombiningthe D overthebackground,arelistedinTable1.Alltheimagesaremuch R 36spectraobtainedthroughthesky-fibres,andsubtractedfromthe deeperthanourIFSdata,aswewillseebelowindetail.Inparticu- ID 331science-fibrespectra.Fig.1showstheaveragenight-skyspec- o larthisisthecasefortheF850LP-bandimage.Thisimageismuch n trumobtainedaftercombiningallofthem.Severalstrongnight-sky J deeperthantherest,beingasdeepasthecorrespondingimageof un emissionlinescontaminateourspectra,inparticularinthereddest e part(λ>7200Å). UDF (e.g. Coe et al. 2006), and therefore one of the deepest im- 14 agesevertakenofthiscluster.Theclustercoreiscentredonone , 2 The three dithered exposures were then combined, creating a 0 singleRSSframewith993sciencespectraandthecorresponding of the ACS chips for this particular image, and the field of view 13 of our IFS data is completely covered by it. This high-resolution combinationofthepositiontable.Afinaldatacubewith1arcsec ACSF850LP-bandimagewillbeusedforadetailedmorphological pixel−1 sampling was created for each night by interpolating the analysis. combined RSS frame using E3D (Sa´nchez et al. 2004). This cor- respondsto2.85kpcpixel−1 attheredshiftofourtarget.Finally, Table1. SummaryoftheHST/ACSimages. thetwodatacubesobtainedforeachnightwererecentredateach wavelength,correctingfordifferentialatmosphericrefraction,and Filter Pivotλ Width Number Exposure maglim combinedusingIRAFtasks.Thefinaldatacubecomprises4828in- (Å) (Å) offrames time(s) (15σ) dividual spectra, covering a field of view of ∼71 × 68 arcsec2. F475W 4744.356 418.80 3 4267.0 27.42 F555W 5359.547 375.19 4 5601.0 28.04 F625W 6310.454 415.31 5 7013.0 28.08 1IRAFisdistributedbytheNationalOpticalAstronomyObservatories,which F775W 7693.026 432.03 8 10732.0 27.46 areoperatedbytheAssociationofUniversitiesforResearchinAstronomy, F850LP 9054.768 538.95 12 11776.0 28.22 Inc.,undercooperativeagreementwiththeNationalScienceFoundation. (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS,MNRAS376,125–150 128 S.F.Sa´nchezetal. Wealsocollectedbroad-bandphotometryandredshiftinforma- Fig. 2 shows a grey-scale of the combined HST/ACS F850LP- tion of the cluster galaxies available in the literature. Le Borgne bandimage,togetherwithlabelsindicatingthelocationofthegalax- etal.(1992)performedaphotometricandspectroscopicsurveyof iesstudiedbyLeBorgneetal.(1992)(orangestars),Ziegleretal. thegalaxiesinafieldof4×4arcmin2centredonAbell2218.This (2001)(bluesquares)andSmailetal.(2001)(redcrosses).Theloca- surveycomprisesbroad-bandphotometryinfivebands(B,g,r,iand tionoftheconfirmedclustermemberssampledbyourIFSdataare zfilters)for729objects,andredshiftdeterminationfor66ofthem, alsoindicated(greencircles).Aswewillseebelow,mostofthem 51clustermembers.Theycalibratedtheirphotometricdatausing havenotbeenpreviouslytargetedbyanyspectroscopicsurvey. twodifferentsystems,theJohnsononefortheBband(Johnson& Morgan1953),andtheThuan–Gunnfortheg,r,iandzbands(Thuan 2.3 ThedepthofourIFSdata &Gunn1976).Wedidnotperformanyrecalibrationoftheirdata to a common photometric system. These data were initially used ToobtainaroughestimateofthedepthofourIFSdata,wecom- tocomparewiththecorrespondingintensityoftheindividualspec- paredthemwithavailabledeepopticalimages.Fig.3showsathree- traoftheobjectsdetectedinourIFUdata,andfordoingsowejust colourimagecreatedbyco-addingthefluxofthefinaldatacube D neededtousethecorrespondingJohnsonandThuan–Gunneffective throughthreebroad-bandscorrespondingapproximatelytoB,Rand o w flux(e.g.Fukugita,Shimasaku&Ichikawa1995).Theaccuracyof I(left-handpanel),togetherwithasimilarimagecreatedusingthe nlo theirphotometryrangesfrom0.1to0.5magfromthebrighter(B∼ previouslydescribedHST/ACSdataset.Itisinterestingtonotethe ad e 2la0ckmtahge)rteoqtuhierefadinqtuearl(itBy∼ne2e6demdatgo)poebrjfeocrtms.aTnheacscpuercattreosacnoaplyicsidsaotaf sciumbeilahraistieasfibneatlwFeWenHtMhefotwropoiimnta-gliekse,sdoeusrpcietesothfe∼f2ac.5tathrcasteocu,rmduactha d from theage–metallicityindicators,liketheequivalentwidthsofHβand largerthantheonefromtheACS.BasedontheI-bandmagnitudeof h ttp MspgeZcbit,ergoblusectrothpeeticyaolp.bro(s2ev0rivd0ae1t)ihoirngeshpooqrfutea4dl8ityocnorleomdusurh-listfietslldeitcettmeedromdeieanrraalyttie-ot-nyresps.eolmuteimon- qtfhoueraolfiautiyrnItsFepsSetcddtareattaefcIotlriemdth∼oeb2pj2ero.c4pt,mowsaeegdr.osHtuuogdwhyleyvfoeerrs,toiwmbejaeotcentsltyhbegriedgtehhtteeigcrthitohenannoliuImg<iht ://mnras bersofthecluster.Theyalsopresenthigh-qualitymultibandpho- 21.5mag(signal-to-noiseratio>5σ perspectralpixel),aswewill .ox fo tometry in four bands (U, B, V and I), derived from groud-based seebelow.Thislimitisfarbeyondpreviousspectroscopicstudies rd oiebssebravsaetdioonns,HanSdT/aWmFoPrCp2hoolbosgeicrvaaltaionnasly(sFis70o2fW19boafntdh,e∼4R8bgaanladx)-. onthiscluster,aroundIlim∼18.7mag(Ziegleretal.2001). journa ls Finally,Smailetal.(2001)presentaphotometricstudyoftheages .o 3 ANALYSIS AND RESULTS rg andmetallicitiesof81early-typeclustercandidates(33withcon- a/ firmedredshifts),basedonground-basedK-bandandHST/WFPC t U 3.1 Photometricanalysis N multibandimages,intheF450W,F606WandF814Wbands.Inboth IV cases,theirphotometricdatawerecalibratedintheVega-basedsys- WeusedthepreviouslydescribedHST/ACSimagestoobtainhomo- E R tem.Manyotherstudieshavetargetedthisclustersinceitsredshift geneousmultibandphotometryforallthedetectedobjectsinfieldof S ID confirmation by Kristian et al. (1978), and the early photometric Abell2218.Thishomogeneousphotometricdatasetisfundamental A D studiesbyButcheretal.(1983).However,inthispaperweonlyuse to estimate the accuracy of the spectrophotometric calibration of C thephotometricandspectroscopicdataofthearticleslistedabove. ourIFSdataandtostudythestellarpopulationsofthegalaxiesin O M thecluster.WeuseSEXTRACTOR(Bertin&Arnouts1996)todetect PL objects in the images up to 15σ over the sky (20 connected pix- U T elswithasignal-to-noiseratioofatleast2σ eachone).Thislimit E N is very relaxed, and most probably we are losing many real faint SE objectsbyusingit.However,wearenotinterestedinthefaintest D E targets,mostprobablyhigh-zobjectsorarclenses(seeFig.1),but M A onlyintheclustermembers.Foreachdetectedtargetweobtained D R aperturephotometryineachband,usingtheMAG BESTparameter ID derivedbySEXTRACTOR.ThelittlevariationsoftheHST/ACSpoint o n spreadfunction(PSF)comparedtothesizeofourtargetsofinter- J u n estguaranteeshomogeneousphotometricdataset.Table2liststhe e 1 resultsfromthisanalysisforthe59galaxiesconfirmedtobecluster 4 , 2 members,usingthepublisheddatalistedaboveortheresultsfrom 0 1 our spectroscopic survey (see below). For each target it includes 3 theSEXTRACTORID,thecoordinates(RAandDec.),andthepho- tometryineachbandpluserrors.Duetothechangeoforientation andslightlydifferentpointingoftheindividualHST/ACSframes, the F850LP band covers a slightly larger area than the images in theotherbands,andsomeoftheobjectswereonlydetectedinthis band. Figure2. Grey-scaleimageofthecombinedF850LP-bandimage,together 3.2 Morphologicalanalysis withlabelsindicatingthelocationofthegalaxiesstudiedbyLeBorgneetal. Oneofthegoalsofthepresentstudyistodeterminethestellarpopu- (1992)(orangestars),Ziegleretal.(2001)(bluesquares)andSmailetal. (2001)(redcrosses).Thelocationoftheconfirmedclustermemberssampled lationsofgalaxiesintheclusterasafunctionoftheirmorphologies, byourIFSdataarealsoindicated(greencircles).Northisupandeastis inordertounderstandtheiroriginandevolution.Toachievethisgoal towardsleft-handside. weperformedamorphologicalanalysisoftheHST/ACSF850LP (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS,MNRAS376,125–150 ThegalaxiesinthecoreofAbell2218 129 D o w n lo a d e d fro Figure3. Left-handpanel:three-colourimagecreatedbyco-addingthefluxofthefinaldatacubethroughthreebroad-bandscorrespondingapproximatelyto m B,RandI.Right-handpanel:similarimagecreatedusingHST/ACSdatadescribedinthetext.Anelectronicversionofthefullsizeimagecanbefoundinthe http webpagehttp://www.caha.es/sanchez/abell2218/BRIbig.jpg. ://m n ra s images,thedeepestofourdataset,followingtheprescriptionspre- ofdecreasingbrightness,modellingandsubtractingeverygalaxyin .ox fo sented in Coe et al. (2006), where similar images were analysed. theSEXTRACTORcatalogue.GALFITisableofsimultaneousfitseveral rd Thisanalysisrecovers,foreachgalaxy,theSe´rsicindex,theinte- targets,producingreasonableresults,inmostofthecases.However, jou gratedmagnitude,theeffectiveradius,theellipticity,theposition inthiscasethefittingprocessisdominatedbythebrightertargets, rna angleandtheasymmetryindex. andfluxofthefainteronesismorepoorlyrecovered.Weperformed ls.o ForundisturbedgalaxiestheSe´rsicindexncorrelatesfairlywell extensivetestsbeforeadoptingthedescribedprocedure.Inaddition, arg/ withmorphologicaltype(e.g.Andredakis,Peletier&Balcells1995; wedeterminetheasymmetryindex(A),followingtheprescriptions t U Bceenltlreatlaclo.n2c0e0n4tr)a,tiinonthse(is.ee.nbseultgheatdgoamlainxaietesdwgiathlahxiigehs)nthhaanvgeahliagxhieesr iinntChoeeFe8t50alL.P(2-b0a0n6d).imOfagtheebr8i0g2htoerbjtehcatnszde<te2c6te.5dmbyagS,EwXeTRaAppClTiOeRd NIVE R withlown(i.e.discdominatedgalaxies).Lesswellbehavedgalax- GALFITtothe241galaxiesbrighterthanz<22.5mag.Theerror S ies, including mergers and irregulars, generally do not have well inthederivedSe´rsicindexisrestrictedtoarangeof±0.25,based IDA definedSe´rsicindices.Fortunately,thesegalaxiescangenerallybe onthesimulationspresentedinCoeetal.(2006),andthereforea D C weededout(orselectedfor)bymeasuringtheirlargeasymmetries morphologicalsegregationbasedonthatparameterisreliable.The O M (e.g.Conseliceetal.2003). remaining galaxies are far beyond the depth of our spectroscopic P L Insummarythemorphologicalanalysisconsistsofthefollowing data, and therefore excluded. Table 3 lists the resulting fitted pa- U T steps.First,weuseSEXTRACTOR(Bertin&Arnouts1996)todetect rameters and their uncertainties for the 59 galaxies confirmed to EN galaxies in the combined F850LP-band image. For each detected be cluster members, using the published data listed above or the S E galaxywerecoverthephysicalandimagecoordinates(RA,Dec.,X resultsfromourspectroscopicsurvey(seebelow).Itincludes,for D E andY),theintegratedmagnitude(mag best),thescalelength(r50), each galaxy, the SEXTRACTOR identification, the χ2/ν of the fit- M the position angle and the ellipticity. Then, we create a postage tingprocedure,thez-bandmagnitude,theeffectiveradius(R ),the A e D stamp5×r onaside,forthebrightestgalaxy.Withinthatpostage axisratio(a/b),thepositionangle(θ),theSe´rsicindex(n)andthe R 50 s ID stamp,neighbouringgalaxiesaremaskedoutusingellipses,each asymmetryindex(A).Thelistederrorswereestimatedbasedonthe o ellipsegivenaminoraxislengthb=2×r50forthatgalaxy.Using resultsofthesimulationspresentedinCoeetal.(2006).Theerrorin n Ju GALFIT(Pengetal.2002),thebrightestgalaxyisfittedtoasingle theasymmetryindexisdifficulttoquantifyforindividualobjects, ne componentSe´rsicmodel(cid:9)(R)=(cid:9) ×exp{−κ [(R/R )1/n−1]}, however,basedonthequotedsimulations,weestimateittobelower 14 where Re is the effective or half-lieght radius, (cid:9)n e is thee effective than∼0.1forthegalaxiesofoursample.Ofallthesegalaxies,only , 201 surfacedensity,(cid:9)(R)isthesurfacedensityasafunctionofradius onehasastrongasymmetryindex(∼0.8).Thisgalaxyiscovered 3 andκ =κ(n)isanormalizationconstant.Thefitisconstrainedto bythefieldofviewofourIFSdata,anditwillbediscussedlaterin 0.2<n<8and0.3<R <500pixels,andthecentroidisconfined detail. e towithin2pixelsofthepositionderivedbySEXTRACTOR.Asinitial guessesfortheGALFITparameters,weusetheSEXTRACTORoutput 3.3 TheSe´rsicindicesandtheHubblesequence parameters.LackingestimatesfortheSe´rsicindexfromSEXTRAC- TOR,westartallfitswithn=1.5.GALFITconvolvesthemodelwith OurmorphologicalclassificationschemeisbasedontheSe´rsicin- therequiredPSF.Priortothefit,wecreatedaPSFimagebycom- dexofthegalaxyprofile.Weconsiderthisautomaticclassification bining postage stamps of unsaturated point-like objects extracted schememorestraightforward,easytoreproduce,andlesssubjective fromthecombinedF850LP-bandimage. thantheclassicalby-eyeclassification(readWolfetal.2005,foran Havingbeencalculatedforthebrightestgalaxy,theSe´rsicmodel illustrationoftheproblem).However,mostofthepreviousstudies issubtractedfromtheF850LP-bandimage.Thissubtractionben- onthemorphologiesofgalaxiesinclustersarebasedonby-eyeclas- efitsthesubsequentmodellingoffainternearbygalaxies.Wethen sificationsfollowingtheHubblescheme(Hubble1926),revisedby proceedtomodelthesecondbrightestgalaxy,andcontinueinorder deVaucouleurs,deVaucouleurs&Corwin(1976).Normally,they (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS,MNRAS376,125–150 130 S.F.Sa´nchezetal. Table2. Summaryofthephotometricanalysis. ID RA Dec. B475 V555 R625 I775 z850 1 248.9537297 +66.2117906 18.09±0.02 17.29±0.02 16.52±0.01 15.85±0.01 15.44±0.01 247 248.9853560 +66.1981066 18.73±0.03 17.94±0.02 17.14±0.02 16.44±0.01 16.08±0.02 169 248.9870850 +66.1850276 18.86±0.03 18.09±0.03 17.34±0.02 16.70±0.02 16.19±0.02 512 248.9569254 +66.2059389 18.92±0.03 18.32±0.03 17.51±0.02 16.80±0.02 16.44±0.02 554 248.9546845 +66.2094133 19.53±0.04 18.87±0.04 18.26±0.03 17.56±0.03 17.00±0.03 172 249.0078962 +66.2089431 19.22±0.04 18.46±0.03 17.67±0.02 16.96±0.02 16.47±0.02 501 248.9646342 +66.2088796 19.40±0.04 18.67±0.04 17.93±0.03 17.12±0.02 16.61±0.03 561 248.9762849 +66.2328158 19.33±0.04 18.62±0.04 17.92±0.02 17.17±0.02 16.65±0.03 189 248.9961745 +66.2011969 19.57±0.05 18.75±0.04 17.93±0.02 17.27±0.02 16.78±0.03 103 249.0154109 +66.1939177 19.50±0.05 18.72±0.04 17.91±0.02 17.25±0.02 16.68±0.03 132 249.0082842 +66.1973755 19.57±0.05 18.76±0.04 17.95±0.03 17.27±0.02 16.73±0.03 D o 555 248.9468602 +66.2111799 19.92±0.05 19.19±0.05 18.44±0.03 17.69±0.03 17.14±0.03 w n 201 249.0162635 +66.2230941 19.61±0.05 18.83±0.04 18.09±0.03 17.39±0.02 16.89±0.03 lo a 2 248.9453711 +66.2204800 19.89±0.06 19.10±0.05 18.35±0.03 17.65±0.03 17.12±0.03 de 255069 224488..99863032212262 ++6666..12906580612416 1290..8290±±00..0056 1199..2404±±00..0055 1188..4463±±00..0033 1177..8703±±00..0033 1177..2206±±00..0044 d from 349 248.9959951 +66.2142513 20.00±0.06 19.23±0.05 18.54±0.03 17.77±0.03 17.27±0.04 h 152760 224498..09265330610996 ++6666..22112964825146 1290..9101±±00..0066 1199..1325±±00..0055 1188..4612±±00..0034 1177..7933±±00..0033 1177..2461±±00..0044 ttp://m n 523 248.9871643 +66.2387374 – – – – 17.39±0.04 ra s 604 248.9528976 +66.2162360 19.41±0.04 18.95±0.04 18.35±0.03 17.94±0.03 17.49±0.04 .o x 599 248.9437771 +66.2056763 20.23±0.07 19.40±0.05 18.65±0.04 18.00±0.03 17.46±0.04 fo 658 248.9415886 +66.2137478 20.29±0.07 19.53±0.06 18.80±0.04 18.11±0.04 17.59±0.04 rdjo 337 248.9878980 +66.2037120 20.53±0.08 19.76±0.06 18.97±0.04 18.35±0.04 17.64±0.05 urn 596 248.9468067 +66.2083220 20.68±0.08 19.83±0.07 19.07±0.05 18.37±0.04 17.77±0.05 als 462 248.9709853 +66.2100529 20.30±0.07 19.49±0.06 18.85±0.04 18.21±0.04 17.86±0.05 .o 471 248.9564109 +66.1951060 20.64±0.08 19.83±0.07 19.03±0.04 18.46±0.04 17.86±0.05 arg/ 302 248.9999910 +66.2116528 20.68±0.08 19.91±0.07 19.18±0.05 18.53±0.04 17.79±0.05 t U 388 248.9849322 +66.2108143 20.41±0.07 19.74±0.06 19.08±0.05 18.48±0.04 17.97±0.05 N 449 248.9708088 +66.2082924 20.62±0.56 19.85±0.10 19.05±0.04 18.39±0.04 18.01±0.05 IVE 72 249.0309189 +66.1939739 – – – – 17.89±0.05 RS 408 248.9916711 +66.2220150 20.62±0.08 19.91±0.07 19.19±0.05 – 17.95±0.05 ID A 95 249.0389514 +66.2155899 20.96±0.09 20.24±0.08 19.53±0.06 18.87±0.05 18.33±0.06 D 798 249.0240772 +66.1820752 – – – – 18.41±0.07 C O 511 248.9579508 +66.2047618 20.35±0.11 19.72±0.10 19.25±0.06 18.68±0.06 18.51±0.07 M 493 248.9754372 +66.2174786 21.44±0.12 20.74±0.10 19.99±0.07 19.33±0.07 18.64±0.07 PL 643 248.9401830 +66.2084529 21.45±0.12 20.67±0.10 19.97±0.07 19.29±0.06 18.59±0.07 UT 329 249.0093949 +66.2250084 21.19±0.11 20.49±0.09 19.69±0.06 19.00±0.06 18.52±0.07 EN 549 248.9519752 +66.2020070 21.33±0.11 20.65±0.10 19.94±0.07 19.23±0.06 18.76±0.08 SE 632 248.9484012 +66.2157872 21.47±0.14 20.50±0.10 19.84±0.06 19.21±0.06 18.89±0.07 D E 357 248.9817003 +66.2002660 21.85±0.14 20.54±0.09 19.73±0.06 19.05±0.06 18.53±0.07 M 560 248.9510744 +66.2127998 22.45±0.19 21.81±0.17 21.06±0.12 20.49±0.12 20.00±0.10 A D 196 249.0143811 +66.2200223 21.71±0.13 20.99±0.12 20.26±0.08 19.57±0.07 19.01±0.09 R 384 248.9824979 +66.2061905 21.46±0.12 20.83±0.11 20.17±0.08 19.57±0.07 19.09±0.09 ID 411 248.9954237 +66.2254124 21.75±0.14 20.99±0.12 20.27±0.08 19.56±0.07 18.91±0.08 on 603 248.9619939 +66.2146043 21.83±0.14 21.10±0.12 20.37±0.09 19.73±0.08 19.13±0.09 Ju n 533 248.9615077 +66.2088133 22.07±0.15 21.39±0.14 20.60±0.10 19.78±0.08 19.30±0.09 e 1 641894 224488..99462274550703 ++6666..22006184049761 2211..9844±±00..1154 2211..2214±±00..1133 2200..5582±±00..0099 1199..9973±±00..0099 1199..5300±±00..1100 4, 20 1 153 249.0133589 +66.2067467 21.83±0.14 21.16±0.13 20.41±0.09 19.79±0.08 19.24±0.10 3 513 248.9540625 +66.2052672 22.31±0.23 21.58±0.15 20.64±0.18 19.64±0.18 19.04±0.09 507 248.9578870 +66.2008693 22.40±0.20 21.68±0.18 21.03±0.12 20.39±0.11 19.84±0.13 431 248.9752130 +66.2068778 20.89±0.09 20.60±0.10 20.27±0.08 19.97±0.09 19.78±0.13 659 248.9460977 +66.2174923 23.69±0.34 22.91±0.29 22.14±0.20 21.46±0.18 20.81±0.17 575 248.9613337 +66.2169292 22.94±0.24 22.05±0.20 21.43±0.15 20.79±0.13 20.36±0.17 574 248.9534579 +66.2081824 23.14±0.32 22.48±0.28 21.74±0.16 21.02±0.14 20.47±0.16 subdividethegalaxiesintheirdifferentHubbletypes,oratleastin Asweindicatedabove,Andredakisetal.(1995)showthatthere threegroups:purespheroidals(E-type),lenticulars(S0),andspirals isacorrelationbetweentheSe´rsicindexandmorphologicaltype. (L-type)(e.g.Fasanoetal.2000;Poggiantietal.2001;Smailetal. However,hereweneedtodeterminehowtheseparameterscorrelate 2001;Ziegleretal.2001).Inordertocomparewithpreviousresults forourdataset.Fordoingso,wefollowFasanoetal.(2000)and it is necessary to determine to which type our galaxies, based on createasetofsimulatedgalaxies.Theirmagnitudes,scalelengths, theirSe´rsicindices,correspond. position angles and ellipticities are extracted from the previously (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS,MNRAS376,125–150 ThegalaxiesinthecoreofAbell2218 131 Table3. Summaryofthemorphologicalanalysis. ID χ2/ν z850 Re(arcsec) a/b θ ns A 1 0.017 16.10±0.02 12.06±0.05 0.49±0.01 37±4 1.63±0.06 0.093 247 0.023 16.21±0.02 5.18±0.05 0.85±0.01 51±4 1.88±0.06 0.175 169 0.024 17.07±0.02 1.66±0.06 0.49±0.02 89±4 1.48±0.07 0.043 512 0.033 17.20±0.02 1.51±0.06 0.77±0.02 −37±8 2.03±0.07 0.156 554 0.034 17.37±0.02 0.63±0.06 0.80±0.02 −10±5 1.80±0.07 0.321 172 0.016 16.84±0.02 1.64±0.06 0.82±0.01 −46±9 2.95±0.06 0.056 501 0.026 15.92±0.02 0.81±0.05 0.84±0.01 50±5 2.86±0.05 0.045 561 0.020 17.25±0.02 1.52±0.06 0.69±0.02 −81±12 1.47±0.07 0.081 189 0.019 17.14±0.02 2.69±0.06 0.90±0.02 −59±10 3.93±0.07 0.088 103 0.011 16.93±0.02 1.17±0.05 0.94±0.01 −26±5 2.69±0.06 0.042 132 0.018 17.28±0.02 1.20±0.06 0.46±0.02 −6±5 2.78±0.07 0.058 D o 201 0.027 17.49±0.02 0.89±0.06 0.50±0.02 −61±10 1.91±0.07 0.132 w n 2 0.013 17.43±0.02 0.96±0.06 0.80±0.02 −16±5 2.63±0.07 0.050 lo a 256 0.022 17.64±0.02 0.51±0.06 0.85±0.02 75±5 1.74±0.07 0.078 de 530499 00..002186 1177..4968±±00..0022 10..0896±±00..0067 00..6760±±00..0022 −3731±±74 32..1293±±00..0077 00..102797 d from 126 0.033 17.44±0.02 0.68±0.06 0.95±0.02 −83±12 2.00±0.07 0.159 h 557203 00..002108 1187..3728±±00..0032 10..1750±±00..0087 00..3577±±00..0032 −−6838±±1103 11..8714±±00..0087 00..005447 ttp://m n 73 0.018 17.90±0.02 0.83±0.07 0.81±0.02 38±5 2.02±0.07 0.077 ra s 604 0.021 18.09±0.03 1.13±0.08 0.44±0.03 84±4 0.72±0.08 0.129 .o x 599 0.019 17.76±0.02 1.04±0.07 0.91±0.02 20±5 3.94±0.07 0.051 fo 658 0.016 17.87±0.02 1.01±0.07 0.88±0.02 51±6 2.69±0.07 0.081 rdjo 337 0.015 17.96±0.02 0.59±0.08 0.78±0.02 5±4 2.49±0.07 0.050 urn 596 0.015 18.96±0.03 0.65±0.09 0.64±0.05 58±6 2.13±0.08 0.067 als 210 0.041 18.19±0.03 0.47±0.08 0.59±0.03 −21±6 2.03±0.08 0.823 .o 462 0.021 18.21±0.03 1.35±0.08 0.84±0.03 −21±7 2.35±0.08 0.102 arg/ 471 0.015 18.12±0.03 0.97±0.08 0.89±0.03 3±5 2.11±0.08 0.118 t U 302 0.014 18.14±0.03 0.48±0.08 0.93±0.03 5±6 2.01±0.08 0.048 N 388 0.015 18.25±0.03 0.71±0.08 0.76±0.03 45±5 1.68±0.08 0.065 IVE 449 0.021 18.57±0.03 1.60±0.08 0.87±0.04 89±6 1.03±0.08 0.316 RS 72 0.019 18.64±0.03 0.73±0.09 0.44±0.04 72±5 2.13±0.08 0.057 ID A 408 0.017 18.59±0.03 0.72±0.08 0.74±0.04 −12±6 1.46±0.08 0.152 D 95 0.015 18.41±0.03 0.63±0.08 0.85±0.03 80±6 1.45±0.08 0.062 C O 798 0.020 19.02±0.04 0.88±0.10 0.45±0.06 9±5 0.81±0.09 0.068 M 511 0.029 18.74±0.03 0.57±0.09 0.53±0.04 −67±12 2.35±0.08 0.169 PL 493 0.018 18.99±0.03 0.45±0.09 0.72±0.06 −42±9 1.56±0.08 0.058 UT 643 0.014 19.09±0.04 0.45±0.10 0.99±0.06 84±13 1.74±0.09 0.094 EN 329 0.015 18.71±0.03 0.76±0.09 0.77±0.04 41±6 1.46±0.08 0.071 SE 549 0.012 19.04±0.04 0.63±0.10 0.95±0.06 −35±9 1.77±0.09 0.067 D E 632 0.017 19.03±0.04 0.37±0.10 0.73±0.06 −82±13 1.49±0.09 0.102 M 357 0.018 18.70±0.03 0.60±0.09 0.67±0.04 −9±6 1.07±0.08 0.184 A D 560 0.049 18.39±0.03 0.17±0.08 0.87±0.03 −74±12 0.53±0.08 0.469 R 503 0.013 19.33±0.04 0.83±0.10 0.80±0.07 13±6 0.97±0.09 0.133 ID 196 0.016 19.14±0.04 0.51±0.10 0.67±0.06 −8±6 1.38±0.09 0.079 on 384 0.017 19.53±0.06 0.77±0.12 0.43±0.08 −45±10 1.29±0.09 0.067 Ju n 411 0.010 18.86±0.03 0.39±0.09 0.96±0.04 −81±13 1.88±0.08 0.080 e 1 650333 00..001125 1198..5897±±00..0063 00..4297±±00..1029 00..9964±±00..0084 4635±±99 11..1494±±00..0098 00..005495 4, 20 1 619 0.015 19.55±0.06 0.41±0.12 0.74±0.08 5±6 1.29±0.09 0.069 3 484 0.017 19.71±0.06 0.75±0.13 0.51±0.08 36±5 1.18±0.09 0.078 153 0.013 19.69±0.06 0.69±0.13 0.72±0.08 −76±13 1.34±0.09 0.062 513 0.024 20.04±0.08 1.15±0.17 0.50±0.10 39±10 2.31±0.10 0.194 507 0.013 20.27±0.09 0.38±0.18 0.85±0.11 74±10 1.31±0.10 0.061 431 0.014 20.20±0.09 1.10±0.18 0.85±0.11 32±11 0.89±0.10 0.218 803 0.017 21.21±0.16 0.67±0.26 0.86±0.15 64±17 0.40±0.15 0.240 659 0.019 20.66±0.11 0.25±0.22 0.81±0.12 −83±17 1.68±0.12 0.085 575 0.010 20.86±0.09 0.75±0.24 0.85±0.12 −40±13 1.39±0.13 0.119 574 0.011 20.49±0.06 0.27±0.22 0.84±0.11 −89±18 1.15±0.12 0.073 (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS,MNRAS376,125–150 132 S.F.Sa´nchezetal. describedmorphologicalcatalogueof241galaxies.Foreachentry nationsfrombothextremesubgroups.Anothersourceofcriticism inthemorphologicalcatalogue20galaxiesweresimulated,eachone abouttheadoptedschemeisthefactthatsomedwarfellipticalgalax- withadifferentbulge-to-discratio,rangingrandomlyfrom0.01(al- iesshownlowSe´rsicindices(e.g.Barazza,Binggeli&Jerjen2003). mostapuredisc)to1000(almostapurespheroidal).Eachgalaxy Althoughingalaxysurveyswithoutmorphologicalpre-selections image was created using GALFIT, with a two components Se´rsic mostofthegalaxieswithlowSe´rsicindicesarespiralgalaxiesAn- model,onewithn =4forthebulgeandanotherwithn =1for dredakisetal.(1995),thecontaminationfromdwarfellipticalscould s s thedisc,convolvingthemwiththesamePSFusedinthemorpho- beapotentialprobleminthestudyofgalaxyclusters.Toaddress logicalanalysis.Finally,noisewasaddedfollowingtheprescription thisproblemweperformavisualinspectionofthe29galaxieswith presentedinCoeetal.(2006)andSa´nchezetal.(2004a)forsim- n <1.75listedinTable2and3(i.e.theonesclassifiedasspirals) s ulatingACSimages,toresembleasmuchaspossibletheoriginal usingthepreviousdescribedHST/ACSimages.Onlytwoofthese29 F850LP-bandimage. galaxies(∼7percent)couldbeclassifiedasdwarfspheroidalsbased Thesamemorphologicalanalysisdescribedintheprevioussec- onthisvisualinspection.However,bothofthemarenearbybrighter tionwasperformedforeachofthesimulatedgalaxyimages,fitting galaxies,andtheirlowsurfacebrightnessareas(likediscstructures) D themtoasinglecomponentSe´rsicmodel.Finally,weendupwith arenotclearlydistinguished.Inanycase,thepossiblecontamina- o w ∼5000simulationscomprisingtherecoveredSe´rsicindexforacer- tionbydwarfspheroidalgalaxiesisexpectedtobemarginal.The nlo tainbulge-to-discratio.Simien&deVaucouleurs(1986)showthe remaininggalaxiesshowclearevidenceofdiscstructures.Taking ad e ttyhpeirceavlibseudlgHe-utob-bdliescserqautieonscfeo.rItnhepadritfifceurleanrt,fthameyilpiersesoefngtathlaeximeseainn isncthoeamcecoisunvtatlhidesweictahvineatthsewceocnotenxstidoefrtthhiaststthuedayd.optedclassification d from fractionalluminosityofthespheroidalcomponentofthegalaxies h ttp anditscorrespondingmagnitudedifferencefor eachgalaxytype. 3.4 Spectraextraction ://m Fig.4showstheSe´rsicindicesofthesimulatedgalaxiesasafunc- n tionofthesemagnitudedifferences((cid:5)mag).Thedifferencesfor To deblend and extract the integrated spectra of each individual ras .o various families of galaxies along the Hubble sequence are indi- galaxyinthefieldweusedatechniquefor3Dcrowdedfieldspec- x fo cated (extracted from Simien & de Vaucouleurs 1986). Based on troscopy developed by ourselves (Sa´nchez et al. 2004b; Garc´ıa- rd this figure we adopted the following scheme to morphologically Lorenzo et al. 2005; Sa´nchez et al. 2006a). The technique is an jou acrlaescsoifnysgidaelarexdiesspbhaesreodidoanlsth(Ee-Stye´rpsei)c,ignadliacxeise:sgwaliathxi1e.s7w5i<thnnss<>22..55 eFIxTt3eDns.iIotnctoonIsFisStsooffGaALdFeITbl(ePnedninggetoaflt.h2e00sp2e),cttrhuamtwoefneaamchedobGjAeLc-t rnals.org areconsideredintermediatetype(mostlyS0’s,butstronglycontam- in the data cube by fitting of analytical models. IFS data may be a/ inatedbyotherfamiliesofbulge-dominatedspirals,likeSaandSb), understoodasasetofadjacentnarrow-bandimages,withthewidth t U andgalaxieswithns<1.75areconsideredspirals(L-type). ofthespectralpixel.Foreachnarrow-bandimageitispossibleto NIV This classification scheme is based on the fair correlation be- applymodellingtechniquesdevelopedfor2Dimaging,likeGALFIT, E R tweenthebulge-to-totalluminosityratioandtheHubbletypeofa andextractthemorphologicalandfluxinformationforeachobject S ID galaxySimien&deVaucouleurs(1986).However,wemayadmit inthefieldateachwavelength.Thespectraofalltheobjectsareex- A D that some degree of contamination is expected among the differ- tractedafterrepeatingthetechniquethroughthedatacube.Wehave C entmorphologicalsubgroupsduetothedispersioninthiscorrela- already shown that the use of additional information to constrain O M tion.Althoughwecanclearlydistinguishbetweenpurespheroidals the morphological parameters increases the quality of the recov- P L (E-type)andspirals(L-type),thesubgroupofintermediatetypeis eredspectra(Sa´nchezetal.2006a).Withthatpurposeweusedthe U T mostprobablylessaccuratelydefined,withconsiderablecontami- morphological parameters previously derived from the HST/ACS E N F850LP-bandimage. SE Wecreatea2Dimagebyco-addingthefluxthroughtheuseful D E wavelengthrange(4650–8000Å)coveredbyourdatacube.After M A this,wecross-checkedbyeyethisimage,lookingforthosegalaxies D R detectedintheF850LPbandforwhichwehadperformedamor- ID phologicalanalysis.Wedetect41objects(40galaxiesandonestar) o n in the collapsed 2D image, out to z < 23 mag. The gravitational J u n arclenseswereexcludedsincetheirstudyisoutthescopeofthe e 1 currentwork.Fig.5showsagrey-scaleoftheco-added2Dimage 4 , 2 derivedfromthedatacube,togetherwithanidentificationlabelat 0 1 thelocationofeachofthedetectedobjects. 3 Aswementionedbefore,weusethepreviouslyderivedmorpho- logicalparametersforeachgalaxytoextractitsintegratedspectrum bymodellingeachgalaxyinthedatacubewithGALFIT3D.Weuse thesamemodelthatwasusedtofittheF850LP-bandimage,withall themorphologicalparametersfixed,andfittingonlythefluxateach wavelength.Foreachwavelength,wecreatedaPSFusingpostage stampimagesoftheonlyfield-starinthefieldofviewofourIFSdata. Asinthepreviousmorphologicalanalysis,the3Dfitwasperformed inasequentialway.Wefirstextractedthespectrumofthebrightest galaxy, the central cD, masking the neighbouring galaxies. Once Figure4. Resultsfromthesimulations.SoliddotsshowtherecoveredSe´rsic indicesofthesimulatedgalaxiesasafunctionofthemagnitudedifference extracted, its 3D model was subtracted to the original data cube, betweenthebulgeandtotalcomponents((cid:5)mag).Overplottedcontoursshow andthespectrumofthesecondbrightestgalaxywasextracted.The thedensitydistributionoftheplottedpoints. processcontinuediterativelyuntilthefaintestgalaxyspectrumwas (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS,MNRAS376,125–150 ThegalaxiesinthecoreofAbell2218 133 & Charlot (2003). We will describe later in detail which models wereused.Wemanuallytunedtheredshiftofthemodelmatching thedifferentabsorptionfeatures,includingtheBalmerlines(Hβ, Hγ, Hδ), the different Fe and Mg lines covered by our spectral range,aswellasspectralfeatures,likethe4000-Åbreak.Dueto thespectralresolutionofourIFSdata,andthesignal-to-noiseratio of our targets, we do not expect to have an accuracy better than 0.001intheredshiftdetermination.Wederivedrealisableredshifts for31galaxies,28ofthemattheredshiftofthecluster(z=0.17 ±0.2),oneisaforegroundgalaxy(z=0.104),andothertwoare backgroundgalaxies(bothatz∼0.426).Table4liststhederived redshiftstogetherwiththeidentificationsshowninFig.5(thatwe willusethroughoutthispaper)andtheSEXTRACTORidentification D listedinTable3toallowfordirectcomparison. o w 11 of the 28 galaxies were studied before by Le Borgne et al. nlo (1992).Theredshiftsderivedfromourdataarecoincidentwiththe ad e rdeepriovretendeiwntrheidssahritfitcslefowrit1h7ingtahlaexeixeps,ecatleldofertrhoerms(σcozm∼p0at.0ib0l1e).wWithe d from theredshiftofthecluster.Thesenewgalaxieshavebeenincluded h ttp in Table 3. In total, the final sample of cluster members (28) is ://m comparablewiththatoneofprevioussimilarstudies,butwiththe n Figure5. Grey-scalerepresentationofthe2Dimageobtainedbyco-adding advantagethatnosamplehasbeenselected:oursampleiscomplete ras thefluxfromthedatacubethroughallthecoveredwavelengthrange(4650– .o withinthefieldofviewofourIFSdatadowntoourdetectionlimit. x 8000Å).Thecirclesindicatethelocationofthedetectedgalaxiesinthedata fo cube,togetherwithanumberindicatingthelabelcodeusedtoidentifythem. Oqufatlhiteysespgeaclatrxoiescs,oopnylytosidxehriavveetbheeenprporpeevritoieusssotfudthieedirwstiethllaernopuogph- rdjou ulations(Ziegleretal.2001).Therefore,oursamplewillincrease rna ls extracted.TheindividualspectraareshowninAppendixA.Allof significantlyourknowledgeofthecoreofthecluster. .o rg themarebackgroundlimited.Thus,thesignal-to-noiseratiocanbe a/ estimatedfromthebackground-noiselevel.Weusedthevariations t U N intheskyspectrumineachsky-fibretodeterminethelevelofback- 3.6 TheCMdiagram IV groundnoise.Inthisway,weestimatedthatthesignal-to-noiseratio E oftheextractedspectrarangebetween∼1000and∼5perspectral Fig.6showstherest-frameB−VversusMV CMdiagramforthose RS pixelforthebrightestandfaintesttargets.Thesignal-to-noiseratio clustermemberslistedinTable2,derivedfromtheHST/ACSbased IDA isaffectedstronglybytheskyemissionlines(seeFig.1),whereit photometricdatadescribedabove.Theabsolutemagnitudeswere D C dropstolessthan∼3σ evenforthebrightesttargets,respectively. corrected for cosmological effects and we applied a k-correction O M Theeffectismoresevereinthewavelengthrangeλ>7200Åwhere based on the same SSP synthetic model. For each synthetic SSP P manyskyemissionlineshappentobeblended. model (described below) we derived the expected B − V colours LU T Theprocedurereliesontheassumptionthatthemorphologicalpa- attheaverageredshiftoftheclusterandintherestframe,deriving E N rametersdonotchangesubstantiallyasafunctionofthewavelength, atthesametimetheV-bandk-correction.Anaveragerelationbe- SE somethingwhichisnotcompletelytrue,sincecolourgradientsare tweentheobservedandrest-framecoloursandthek-correctionwas D E expectedinallthegalaxies.But,iftheychange,weexpectthisef- derived,andappliedtothedatashowninFig.6.Redsolidcirclesin- M fecttohaveaminimimpactinthephotometryderivedwithGALFIT. dicategalaxieswithSe´rsicindexn>2.5(E-type),whilegreenstars AD Thus,ifweusethemorphologicalparametersderivedbyapplying indicatethosewithn<1.75(L-type).Orangesolidsquaresindicate RID GALFITtotheF850LPbandandweforcethefittingproceduretouse galaxieswithSe´rsicindicesinbetweenthosevalues,corresponding o themforderivingthephotometryatanyotherband,therecovered tointermediategalaxies.ThecDgalaxyisindicatedwithanorange n J u pflaurxamsheoteurlsdfbreeeblya.siAcaltlhlyouthgehsoaumrepraesviwouhsenexwpeerliemteGnAtLsFsIhTofiwtnthtehsaet oitpwenithcraosdsi,ffbeerienngtsthyembbroiglhsitnescteoibtjiescwteinlltkhneofiwenldt.hWatecpDregfaelratxoielasbdeol ne 14 thisisthecase(e.g.Garc´ıa-Lorenzoetal.2005),wehavealsotested notfollowaSe´rsicprofileofanyindex,showingaflatterandmore , 20 1 thishypothesishere,asdescribedinAppendixB.Inanycase,the extendedlightprofileintheouterregions(e.g.Oemler1976).Open 3 method is not strongly sensitive to this assumption, as described circlesindicatethegalaxieswithinthefieldofviewofourIFS,with inSa´nchezetal.(2006a).Inprinciple,itispossibletoperformthe redshiftslistedinTable4.Thebigopensquareindicatestheonly extractionprocedureusingtheparametersderivedfromthemorpho- galaxywithahighasymmetryindex. logicalanalysisasaninitialguess,andlettingGALFITtoderivethe correctonesforeachwavelength.Inthiscase,therecoveredspectra arenoisierandintheresidualdatacubethereisclearevidenceof 3.6.1 Morphologicalsegregation notcorrectlyextractedstructure. Toourknowledgethisisthefirsttimethatamorphologicallyseg- regated CM diagram is shown for confirmed cluster members. In 3.5 Redshiftdetermination previous works (e.g. Smail et al. 2001) only cluster galaxy can- We determined the redshift of each detected galaxy in the field didateshavebeenshown.Themorphologies,basedontheSe´rsic ofviewofourIFSdatabycomparingitsspectrumwithsynthetic indiceslistedinTable3,breakdownas:11E-type,19intermediate modelsforsinglestellarpopulations(SSP)obtainedfromBruzual and29L-type.IfweconsiderasE-typegalaxiestobealltheobjects (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS,MNRAS376,125–150 134 S.F.Sa´nchezetal. Table4. Summaryoftheparametersderivedfromthespectra. IFU SEXTRACTOR z Hβ Mgb Age Metallicity Mass ID ID (Gyr) Z (1010M(cid:4)) 1 1 0.174 1.6±0.1 3.0±0.1 9.8±2.8 0.019±0.011 102.1±2.6 2 512 0.179 2.1±0.1 3.4±0.1 12.0±3.4 0.025±0.014 74.6±2.8 3 501 0.163 1.1±0.1 4.2±0.2 12.9±3.6 0.046±0.026 39.5±2.0 4 554 0.175 2.8±0.1 3.6±0.1 12.1±3.4 0.025±0.014 67.9±3.3 5 555 0.181 1.4±0.1 4.1±0.2 11.6±3.2 0.032±0.018 36.2±2.2 6 2 0.181 1.5±0.6 6.1±0.9 12.7±3.6 0.045±0.025 75.7±4.6 7 509 0.184 1.0±0.1 3.0±0.1 9.9±2.8 0.019±0.011 17.2±1.2 8 570 0.184 1.5±0.2 2.3±0.7 12.7±3.6 0.045±0.025 39.9±2.7 9 599 0.163 1.5±0.1 4.4±0.2 9.6±2.7 0.019±0.011 10.5±0.8 11 658 0.172 2.2±0.1 3.9±0.1 9.5±2.7 0.018±0.010 11.5±0.8 D o 12 596 0.174 1.4±0.1 4.9±0.1 12.4±3.5 0.024±0.013 13.6±1.2 w n 13 462 0.179 3.2±0.1 2.7±0.2 4.5±1.3 0.018±0.010 7.8±0.6 lo a 14 449 0.178 2.7±0.1 2.8±0.2 5.2±1.4 0.009±0.005 4.5±2.3 de 1165 654131 00..116672? 11..20±±00..12 02..75±±00..32 99..75±±22..77 00..001091±±00..001011 42..12±±00..53 d from 17 632 0.179 2.8±0.3 3.0±0.2 3.5±1.0 0.009±0.005 2.1±0.3 h 2108 556409 00..118702 2.2±0.2– 03..75±±00..34 59..27±±12..47 00..000199±±00..000151 21..88±±00..43 ttp://m n 21 533 0.171 1.2±0.3 1.8±0.3 11.5±3.2 0.031±0.018 4.5±0.7 ra s 22 603 0.171 – 2.7±0.3 5.8±1.6 0.003±0.002 1.0±0.2 .o x 24 619 0.159 3.1±0.2 2.8±0.5 11.4±3.2 0.006±0.003 1.5±0.2 fo 25 484 0.174 1.2±0.5 0.9±0.3 4.2±1.7 0.007±0.004 1.0±0.2 rdjo 26 513 0.180 3.8±0.4 – 7.1±2.0 0.019±0.011 2.1±0.5 urn 27 659 0.168 – – 12.8±3.6 0.045±0.025 1.8±0.6 als 28 575 0.168 – – 12.7±3.6 0.044±0.025 3.6±0.9 .o 29 574 0.168 2.3±0.7 – 9.7±2.7 0.008±0.004 0.6±0.2 arg/ 38 431 0.170 – 1.2±0.4 0.3±0.1 0.007±0.004 0.1±0.1 t U 41 507 0.168 – – 5.8±1.6 0.007±0.004 0.6±0.1 N IV E Uncertainvaluesarewritteninitalic. R S ID A withn>1.75,theyrepresentonlya∼50percentofthetotal.Forthe D C completesubsampleofgalaxiescoveredwithintheIFUfieldofview, O M themorphologiesbreakdownas:7E-type,8intermediateand13 P L L-type.Contrarytopreviousresults(e.g.Smailetal.2001;Ziegler U T etal.2001)thisindicatesthattheclusterpopulationisnotdominated E N byearly-typegalaxies,andthereisanalmostparitybetweenearly- S E andlate-typegalaxies. D E Ourapproachtoderivethemorphologiesissubstantiallydiffer- M enttoinpreviousattempts,basedonvisualclassifications.There AD isaconsiderabledebateoverthereliabilityofvisualclassifications RID of early-type galaxies (e.g. Dressler et al. 1997; Andreon 1998; o n Fabricant,Franx&vanDokkum2000),andtheabilitytocompare J u themwithclassificationsbasedonprofileanalysis,likeconcentra- ne 1 tion and/or Se´rsic indices. However, since there are no previous 4 similar studies, we must compare with the existing ones. Ziegler , 20 1 etal.(2001)showvisualmorphologicalclassificationsof19oftheir 3 48clustermembers,basedonWFPC/HSTimaging.Theirclassifi- cation was performed by W. Couch, who classified them as 8E, 1E/S0,5S0,3SB/0,1Saand1Sab.TakingtogetheralltheE-type galaxies(E,E/S0andS0),theyaccountfora∼73percentofthe Figure6. Rest-frameB−Vcolourdistributionasafunctionofabsolute totalsample.Smailetal.(2001)presentedmorphologicalclassifi- mcleasgninitduidceateMgValfaoxriethsowseithclSue´srtesircminedmebxenrs>lis2t.e5d(Ein-tTyapbel)e,w2.hiRleedgrseoelnidstcairrs- cationfor81clustercandidates(33withconfirmedredshifts),based correspondtoobjectswithn<1.75(L-type).Orangesolidsquaresindi- onthesameanalysisasZiegleretal.(2001).Theirmorphologies catelenticulargalaxies,withSe´rsicindicesinbetweenbothvalues.ThecD breakdownas:31E,3E/S0,28S0,5Saand14late-typespirals. galaxyisindicatedwithanorangeopencross,beingthebrightestobjectin TakingtogetheralltheE-typegalaxies,theyaccountfora∼76per thefield.Opencirclesindicatethe28galaxieswithinthefieldofviewofour centofthetotalsample.Eventakingintoaccountthedifferentmeth- IFS.Thebigopensquareindicatestheonlygalaxywithahighasymmetry odsusedtoderivethemorphologies,andthedifficultiestocompare index. them,itisclearthatbothZiegleretal.(2001)andSmailetal.(2001) foundfewerL-typegalaxiesthanwhatwearefinding. (cid:3)C 2007TheAuthors.Journalcompilation(cid:3)C 2007RAS,MNRAS376,125–150

Description:
cluster, in order to obtain spectroscopy of an unbiased flux-limited sample of cluster galaxies. 43 objects were Key words: techniques: spectroscopic – galaxies: clusters: individual: Abell 2218 – galaxies: elliptical and We thank Dr R. Gredel, former director of Calar Alto, for support- ing
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.