TheAstrophysicalJournal,667:97Y116,2007September20 A #2007.TheAmericanAstronomicalSociety.Allrightsreserved.PrintedinU.S.A. BLACK HOLE MASSES AND EDDINGTON RATIOS OFAGNs AT z <1: EVIDENCE OFRETRIGGERING FORAREPRESENTATIVESAMPLEOFX-RAY-SELECTEDAGNs L. Ballo,1,2 S. Cristiani,3 G. Fasano,4 F. Fontanot,5 P. Monaco,5,3 M. Nonino,3 E. Pignatelli,4 P. Tozzi,3 E. Vanzella,3 A. Fontana,6 E. Giallongo,6 A. Grazian,6 and L. Danese1 Received2005December29;accepted2007May1 ABSTRACT WeestimatedblackholemassesandEddingtonratiosforasampleofX-ray-selectedactivegalacticnuclei(AGNs) inthefieldscoveredbytheGreatObservatoryOriginsDeepSurvey(GOODS).Thespannedrangesinredshift(0:4< z<1)andhardX-rayluminosity(1042 PL P4;1043ergss(cid:1)1)allowustostudyarepresentativesubsampleofthe X maincontributorstothe2Y10keVX-raybackground.Nuclearandbulgemagnitudesinfourbandshavebeenmea- suredviaatwo-dimensionaldecompositionappliedtoHSTACSimages.Usingtheblackholeversusbulgeluminos- ityrelationandthe intrinsicnuclearemission,wederivedtheblackholemassandtheAGNbolometricluminosity. Wefindinoursamplethat(1)theX-rayYtoYopticalindicesarelargerthaninopticallyselectedQSOs,asexpecteddue totheX-rayselection);(2)theX-raybolometriccorrectionsaregenerallysmall,suggestingadecreasewiththenu- clearluminosity;(3)theEddingtonratiosareaboutafactor10belowthevaluesfoundathigherredshiftandlumi- nosity;(4)thecentralblackholeshaveratherlargemasses;and(5)atleastforzP0:8,ascarcenessofblackholeswith massM (cid:2)106M andaccretionrateneartheEddingtonlimit:thisresultcouldbeascribedtoadeclineintheirnum- BH (cid:3) berdensity,oritcouldsuggestasubstantialaccretionathigherredshift(zk1)alsoforthesesmallerblackholes. Subject headingsg: black hole physics — galaxies: active — galaxies: nuclei — galaxies: photometry — X-rays: diffuse background — X-rays: galaxies Online material: color figures 1. INTRODUCTION michistoryofSMBHaccretion.Asanultimategoal,oneshould matchthemassfunctionoftheBHaccretedmasswiththelocal Recent Chandra and XMM-Newton deep surveys (Brandt massfunctionofthequiescentBHs(see,e.g.,Saluccietal.1999; etal.2001;Rosatietal.2002;Hasingeretal.2001)haveresolved (cid:4)90% ofthe 2Y10keV X-ray background(hereafter XRB;see Yu&Tremaine2002;Marconietal.2004;Shankaretal.2004). ThedetailedhistoryofSMBHaccretionisalsoarelevantclue Baueretal.2004).ThemaincontributiontotheXRBinthisen- tounderstandtheobservedrelationshipsbetweenthemassoflo- ergyrangeisduetoamixtureofobscuredandunobscuredactive calSMBHsandthemassandluminosity(Kormendy&Richstone galactic nuclei (AGNs), with a redshift distribution peaking at 1995;Magorrianetal.1998;McLure&Dunlop2002;Marconi& (cid:4)0.7(Szokolyetal.2004),andamajorcontributionintheinter- Hunt2003)orthevelocitydispersion(Ferrarese&Merritt2000; val0:4<z<1fromobjectswithX-rayluminositiesbetween 3;1041 and1043ergss(cid:1)1(Uedaetal.2003). Gebhardtetal. 2000; Merritt & Ferrarese 2001; Tremaine etal. 2002)ofthespheroidalcomponentoftheirhostgalaxies.Even- Themassaccretedontothecentralsupermassiveblackholes tually,thiswillcastlightonphysicalprocessesinvolvingboththe (SMBHs)ofthesez<1AGNsisestimatedtobe(cid:4)30%ofthe SMBH and the host galaxy (Silk & Rees 1998; Granato et al. totalmassdensityaccretedatanyredshift(Marconietal.2004; 2001,2004;Hopkinsetal.2005). Shankaretal.2004).Animportantquestioniswhetherthisac- In this paper we estimate the BH mass, the bolometric lu- cretedmassisonaveragefallingontoSMBHsalreadyhavingvery minosityandtheEddingtonratioforarepresentativesampleof largemass((cid:1)M /M T1),oriftheaccretionisoccurringon acc BH low-luminosity X-ray-selected AGNs in the redshift interval smallerSMBHs.ThefirstregimecorrespondstoEddingtonratios 0:4(cid:2)z(cid:2)1. k¼L /L T1andcanbeassociatedwithreactivationofpre- bol Edd Thedatademandsaresubstantialforastudysuchasours.High existingblackholes(BHs),whilethelattercorrespondstokP1 spatialresolutionandhighsignal-to-noiseratiodataintheX-ray andisassociatedwiththemainepisodeofgrowthoflowermass produced by the Chandra observations in the Chandra Deep BHs.Thisissueisrelevantinordertoreconstructthedetailedcos- FieldsSouthandNorth(Giacconietal.2002;Rosatietal.2002; Alexander et al. 2003) are offundamental importance to select 1 SISSA/ISAS,InternationalSchoolforAdvancedStudies,34014Trieste, low-luminosityAGNs.Wealsoneedopticalimageswithexcel- Italy;[email protected],[email protected]. lentspatialresolutionandveryhighqualityphotometry,inorder 2 EuropeanSpaceAstronomyCenterofESA,E-28080Madrid,Spain;lballo@ todisentangle the galacticand nuclear components, and toesti- sciops.esa.int. mate the AGN bolometric luminosity and the BH mass. In this 3 INAFYOsservatorioAstronomicodiTrieste,34131Trieste,Italy;cristiani@ respecttheGreatObservatoriesOriginsDeepSurvey(GOODS; oats.inaf.it,[email protected],[email protected],[email protected],vanzella@ oats.inaf.it. seeGiavaliscoetal.2004)isreallyunique;infact,inadditionto 4 INAFYOsservatorioAstronomicodiPadova,35122Padova,Italy;fasano@ deepX-rayandopticalimagesfromtheChandraX-RayObser- pd.astro.it,[email protected]. vatory andtheHubbleSpaceTelescope(HST),italsoexploitsex- 5 DAUT,34131Trieste,Italy;[email protected],[email protected]. tensivefollow-upworkwithground-basedtelescopes,extending 6 INAFYOsservatorioAstronomicodiRoma,00040MontePorzioCatone, thesamplingoftheelectromagneticspectrumofthesourcesupto Italy;[email protected],[email protected],[email protected] .it. theradiowavelength. 97 98 BALLO ETAL. Vol. 667 Thepaperisorganizedasfollows.Inx2(butseealsoAppen- dixA)wedescribethesampleselectionandlistthedifferentdata sets available. In x 3 we introduce the morphological analysis carriedout(detailsarereportedinAppendixB)anddiscussthe resultsofthedecomposition.Section4isdevotedtorecoverthe nuclear properties (i.e., BH masses and nuclear bolometric lu- minosities).Inx5ourfindingsarediscussedandcomparedwith resultsfromtheliterature.Finally,inx6wesummarizeourwork. If not otherwise stated, throughout this paper magnitudes are givenintheABsystem,andhardX-rayluminositiesareinthe 2Y8keVenergyrange(whennecessary,convertedfromL2Y10keV assuming a power-law X-ray spectrum with photon index (cid:2)¼ 1:8: L2Y8keV/L2Y10keV ’0:84). We assume a cosmology with (cid:3) ¼0:3,(cid:3) ¼0:7,andH ¼70kms(cid:1)1Mpc(cid:1)1(Spergeletal. M (cid:4) 0 2003,2007). 2. SAMPLE SELECTION The starting point for the present work is the deep, high- resolution optical imaging performed in the F435W, F606W, F775W,andF850LPfilterswiththeAdvancedCameraforSur- veys(ACS)onboardHSTintheframeworkoftheGOODSpro- gram.Inthefollowing,werefertothesefourpassbandsasB,V, i,andz,respectively.GOODScoversatotalareaof 320arcmin2 intwofieldscenteredontheChandraDeepFieldYSouth(CDF-S) andtheChandraDeepFieldYNorth(CDF-N).Thesetworegions havebeentargetsofdeepX-raypointings(1Ms[seeGiacconi etal.2002;Rosatietal.2002;Alexanderetal.2003] and2Ms [seeAlexanderetal.2003],respectively)carriedoutbyChandra. Overall,80%ofobjectshaveredshiftinformation.Ifwerestrict ourselvestotheCDFareascoveredbyGOODS,(cid:4)98%ofthe X-rayobjectshaveopticalcounterparts. In theGOODSnorth field, 63% and 17% of the X-ray sources have spectroscopic andphotometric redshift,respectively;inthesouthareathese percentagesincreaseto68%and32%,respectively. We collect all the sources in the Chandra X-ray catalogs lyingintheGOODSfieldswithredshift(spectroscopicorphoto- metric)between0.4and1,andwithX-rayluminosityL2Y8keV > 1042ergss(cid:1)1inordertoselectbonafideAGNs(Zezasetal.1998; Moran et al. 1999). They amount to 66 sources, 88% of them havingspectroscopicredshifts.Theredshiftintervalischosento bracketasignificantfractionofthelow-luminositysourcescon- tributingtotheXRB.Theupperboundz¼1isimposedbythe needofasufficientsignal-to-noiseratiotomakeamorphological analysisformostofthesources.Thisupperlimitontheredshift alsoensuresthattheobservedz-bandisalwayssamplingtheop- ticalrestframe.Inordertocarryoutthemorphologicalanalysis described in the next section, the simultaneous availability of Fig.1.—Top:HardX-rayfluxesvs.i-bandmagnitudesfortheX-raysources high-quality,noncrowdedB,V,i,and zACSsubimages(cutouts) intheGOODSfieldswith(dots)andwithout(opencircles)redshiftinformation. isessential.Thisselectionreducesthesampleto25sourcesinthe Bottom:i-bandmagnitudesvs.redshiftforX-raysourcesintheGOODSfields CDF-Sand28intheCDF-N(sampleA). withredshiftinformation,withsuperimposedthei-bandmagnitudesexpected FulldetailsoftheselectionprocessaregiveninAppendixA. forbulgeswithblackholemassesofMBH¼105,106and107M(cid:3)(fromtopto bottom:solidlineandopentriangles,dottedlineandstars,anddashedlineand HereletusnotethattheX-raysourcesintheGOODSfieldswith filledsquares,respectively). a redshift determination cover the same region in the F -F X opt planeasthewholedatabaseofthe‘‘GOODSX-raysources’’(see Fig.1,toppanel),althoughmostoftheobjectswithoutredshift forthreeblackholemasses(fromtoptobottom,M ¼105,106, BH determinationhavefaintopticalcounterpart.Accordingtoabi- 107 M ) withthedistributionofthe optical magnitude ofX-ray (cid:3) dimensional Kolmogorov-Smirnov (K-S) test, there is no evi- sourcesintheGOODSfieldswitharedshiftdetermination.The dencethatthetwosamplesaredrawnfromdifferentdistributions selectedrangeinredshiftandtheavailablemagnitudesdonotim- (the K-S probability is (cid:4)60%). This makes us confident that plythatweareforcedtostudyonlyblackholewithlargemasses. selectingonlysourceswithredshiftdeterminationdoesnotinduce Figure2showshowthesampleAisdistributedinredshift(top a bias toward optically bright objects. Moreover, assuming the leftpanel),hardX-rayluminosity(toprightpanel),andoptical blackholemassYbulgeluminosityrelations,wecanestimatehow magnitude(bottompanel).Thespikesvisibleintheredshiftdis- brightabulgewithagivenblackholemasswouldbeatagiven tribution trace the large-scale structure identified in the whole redshift:inthebottompanelof Figure1,wecomparethetracks CDF-Satz¼0:67andz¼0:73,bothintheoptical(Vanzella No. 1, 2007 M AND EDDINGTON RATIOS OFAGNs AT 0.4 < z < 1 99 BH Fig.2.—Distributionsinredshift(topleftpanel),hardX-rayluminosity(toprightpanel),and i-bandmagnitude(bottompanel)forthesampleA(GOODSX-raysources withspectroscopicorphotometricredshiftsbetween0.4and1,observedhardX-rayluminosityL2Y8keV>1042ergss(cid:1)1,andgoodcutoutsinallbands;seex2);hatched areasshowthesamedistributionsfortheanalyzedsources(sampleB).RedshiftsarefromSzokolyetal.(2004),Grazianetal.(2006),andZhengetal.(2004)fortheCDF-S, andfromBargeretal.(2003)fortheCDF-N;luminositiesarefromAlexanderetal.(2003);i-bandmagnitudesarefromtheGOODSACScatalog(seefootnote12). etal. 2005)and in the X-rayranges (Gilli et al. 2003), and at redshiftdistributions(theK-Sprobabilitiesare99%and75%, z(cid:4)0:85intheCDF-N(seeBargeretal.2003;Gillietal.2005). respectively). We have removed from the sample A objects with a close ThecontributionstothehardXRB7ofthesampleAis16%. companion(withinaprojectedradiusof200,withoutdistinction InFigure3wecomparethetotalcontributiontothehardXRB between physical interaction or projection effect; see also Ap- computedfromtheCDF-SplusASCA sample(DellaCecaetal. pendixB);inthiswaytheanalyzedsample(sampleB)reducesto 2001; long-dashed line,withthethegrayarearepresentingthe 34 objects (19 in the CDF-S and 15 in the CDF-N), listed in estimateduncertainties;seeTozzi2001)withtheintegratedcon- Table1.AsisapparentinFigure2,theredshiftandluminosity tributionofthesampleA(solidline)andoftheselected34sources distributions of the sample B (hatched areas) match the distri- butionsofthesampleA,apartfromaslightdearthofobjectswith 7 Average total flux density in the 2-8 keV band: (1:79(cid:5)0:11) ; zk0:8.AccordingtoaK-Stest,wecannotprovethatthetwo 10(cid:1)11ergscm(cid:1)2s(cid:1)1deg(cid:1)2,fromDeLuca&Molendi(2004)convertedfrom samples are drawn from different hard X-ray luminosity or the2Y10keVbandassuming(cid:2)¼1:4. 100 BALLO ETAL. TABLE1 BasicInformationforX-Ray-selectedAGNs (ObservedL2Y8keV>1042 ergss(cid:1)1)intheGOODSFields with0:4<z<1AnalyzedinThisWork(SampleB) R.A. Decl. logL2Y8keV iAB (J2000.0) (J2000.0) ID Redshift (ergss(cid:1)1) (mag) (1) (2) (3) (4) (5) (6) CDF-N 123618.58......... +621115.0 114 0.679 43.20 20.46 123625.01......... +622115.7 141 0.747 42.29 21.87 123627.75......... +621158.4 150 0.762 42.34 23.10 123632.59......... +620759.8 170 0.680 42.16 21.83 123635.86......... +620707.7 194 0.555 42.37 21.99 123642.24......... +620612.8 222 0.857 42.60 21.38 123646.33......... +621404.7 243 0.961 43.62 21.57 123654.58......... +621110.6 285 0.955 42.43 22.47 Fig.3.—Contributiontothe2Y8keVX-rayfluxdensityasafunctionofthe 123659.09......... +622523.7 303 0.678 42.97 20.76 resolvedsources:comparisonbetween(1)thetotalresolvedcontribution(long- 123702.43......... +621926.1 323 0.514 43.27 19.70 dashedline,withgrayarearepresentingtheestimateduncertainties)computed 123710.07......... +620547.9 368 0.935 42.09 23.69 fromthe1MsCDF-SsampleplusthebrightsamplefromASCA (DellaCecaetal. 123722.44......... +620536.1 404 0.978 42.61 21.87 2001)atfluxeslargerthan10(cid:1)13ergscm(cid:1)2s(cid:1)1;seeTozzi(2001);(2)thecontribu- tionofourinitialselection(sampleA:GOODSX-raysourceswithspectroscopic 123724.00......... +621304.3 412 0.474 42.33 24.25 orphotometricredshiftsbetween0.4and1,observedhardX-rayluminosity 123731.73......... +621703.7 439 0.839 42.61 20.89 123739.46......... +622239.2 451 0.838 42.27 21.39 L2Y8keV>1042ergss(cid:1)1andwithgoodcutoutsinallbands;continuousline); (3)thecontributionoftheanalyzedsources,asinTable1(sampleB;dashedline). ThetopdottedlinesrefertopreviousmeasuresoftheunresolvedhardX-ray CDF-S background; frombottomtotop:Marshalletal.(1980),Uedaetal.(1999), Ishisakietal.(2001),andVecchietal.(1999). 033233.02......... (cid:1)274547.4 34 0.839 42.93 22.19 033226.76......... (cid:1)274145.6 40 0.667 42.85 21.61 033227.00......... (cid:1)274105.1 42 0.734 44.17 19.12 constructmultiwavelengthspectralenergydistributions(SEDs) 033227.61......... (cid:1)274145.0 44 0.737 42.65 21.89 fortheanalyzedobjects,seex3: 033224.84......... (cid:1)275600.0 47 0.733 42.64 21.44 033224.84......... (cid:1)275600.0 48 0.534 42.48 20.61 CDF-S.—GOODS has imaged these fields at 3.6, 4.5, 5.8, 033220.07......... (cid:1)274447.0 51 0.670 42.35 23.76 and 8.0 (cid:1)m with the Infrared Array Camera (IRAC) onboard 033217.18......... (cid:1)275220.9 52 0.569 42.88 20.84 Spitzer.Foreachfield,observationshavebeendividedintotwo 033220.07......... (cid:1)274447.0 151 0.604 42.36 22.11 epochs,withameanexposuretimeperchannelperskypointing 033208.24......... (cid:1)274153.6 156 0.545 42.07 21.80 ofapproximately23hrperepoch(M.Dickinsonetal.2007in 033246.98......... (cid:1)274346.2 170 0.664 42.14 20.00 preparation).InthisworkwewillmakeuseoftheSpitzerdatafor 033222.51......... (cid:1)274804.8 189 0.734 42.04 22.64 the CDF-S as analyzed by Grazian et al. (2006). The GOODS 033235.23......... (cid:1)275317.8 192 0.733 42.48 21.76 field of the CDF-S has also been the target of a deep imaging 033239.73......... (cid:1)274611.2 201 0.679 42.39 25.38 campaigninthenear-infraredwiththeESOtelescopes.Alarge 033218.99......... (cid:1)274755.4 252 0.481 42.13 24.71 field (200;200) has been covered with the SOFI instrument at 033213.92......... (cid:1)275000.7 257 0.549 42.08 22.44 033213.83......... (cid:1)274525.6 266 0.735 42.01 21.23 theNewTechnologyTelescope(NTT)intheJandKsbandsas 033234.73......... (cid:1)275533.8 511 0.668 42.08 21.44 partoftheDeepPublicSurveycarriedoutbytheESOImaging 033223.88......... (cid:1)275842.4 613 0.910 42.29 22.54 Survey(EIS)program;thedataaredescribedinVandameetal. (2001).An H-bandsurveyoftheCDF-S,encompassingthespa- Notes.—Cols.(1)and(2):unitsofrightascensionarehours,minutes,and tialcoverageoftheEISobservations,hasbeencarriedoutwith seconds, and units of declination are degrees, arcminutes, and arcseconds; thesameinstrument;theresultsarepresentedinMoyetal.(2003). col.(3):sourcenumber;col.(4):redshiftfortheopticalcounterpart,fromBarger etal.(2003),Szokolyetal.(2004),Grazianetal.(2006),andZhengetal.(2004); TheGOODSfieldisbeingcoveredbydeeperobservationsinthe col.(5):observedX-rayluminosityinthe2(cid:1)8keVband;col.(6):ABmagnitude samenear-IRbandswiththeISAACinstrumentattheVLT.These intheibandfromtheGOODSACScatalog(seefootnote12). datahavebeenpartiallyreleasedbyESOandwillbeusedinthis work.8WealsomakeuseofU-banddatatakenwiththeESOWide FieldImager(WFI)atLaSilla(Chile),whicharepartoftheEIS (dashedline):thiscomparisonshowsthatwiththeanalyzedsources publicsurvey(Arnoutsetal.2001),aswellasrecent Uimages wearesamplinginanuniformwaythesamerangeoffluxofthe withVLT-VIMOSimager.TheWFIimageshavebeenobtainedin sampleA.Insummary,sampleBisconsideredrepresentativeof thewholepopulationofX-raysourceswithspectroscopicorpho- twofilters,theso-calledU35andU38filters,withanexposuretime tometric redshifts between 0.4 and 1 and luminosity L2Y8keV > of(cid:4)54and(cid:4)75ks,respectively.TheU-bandimageofVIMOS 1042ergss(cid:1)1,andonlyscaled(i.e.,sparselysampled)byafactor7. isbasedonaredderfilterandhasanexposuretimeof(cid:4)10ks. BoththesampleAandtheanalyzedobjectscovertheluminosity ThecoveragewithVIMOSoftheGOODSfieldcenteredonthe range1042 (cid:2)L2Y8keV (cid:2)3;1043ergss(cid:1)1,whileonlyfewsources CDF-Sispartial,sincetheobservingprogramhasnotbeencom- pletedyet. exhibitlargerluminosity.Takingintoaccounttheredshiftandlu- minositydistributions,wecanconcludethatthesampleBisrep- CDF-N.—In the CDF-N IR information has been collected matchingoursourceswiththeHK0catalogofCapaketal.(2004), resentativeoftheAGNscontributingtotheXRBatz(cid:2)1(see, e.g.,Fig.16inUedaetal.2003). Data from U-band to IR, available thanks to the remarkable 8 ThedetailsoftheongoingGOODSprogramatESOaregivenathttp:// multiwavelengthcoverageofthetwoGOODSfields,areusedto www.eso.org/science/goods/products.html. TABLE2 Two-dimensionalImageFittingMagnitudes Nucleus Disk Bulge ID z i V B z i V B z i V B (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) CDF-N 114............. 21.66(cid:5)0.20 21.81(cid:5)0.22 21.72(cid:5)0.21 22.02(cid:5)0.28 ... ... ... ... 20.52(cid:5)0.21 20.81(cid:5)0.22 22.32(cid:5)0.24 25.52(cid:5)0.25 141............. 25.57(cid:5)0.38 25.81(cid:5)0.41 26.00(cid:5)0.41 26.16(cid:5)0.43 22.84(cid:5)0.30 23.22(cid:5)0.31 24.42(cid:5)0.33 25.62(cid:5)0.34 21.71(cid:5)0.26 22.43(cid:5)0.27 23.81(cid:5)0.28 26.91(cid:5)0.36 150............. 26.66(cid:5)0.40 27.07(cid:5)0.42 27.83(cid:5)0.41 28.85(cid:5)0.50 23.90(cid:5)0.30 24.25(cid:5)0.31 25.32(cid:5)0.34 26.19(cid:5)0.36 23.15(cid:5)0.27 23.61(cid:5)0.28 25.63(cid:5)0.30 30.25(cid:5)0.35 170............. 25.82(cid:5)0.40 26.12(cid:5)0.41 26.87(cid:5)0.41 27.56(cid:5)0.43 22.90(cid:5)0.36 23.11(cid:5)0.37 23.93(cid:5)0.38 24.70(cid:5)0.41 21.72(cid:5)0.20 22.13(cid:5)0.22 23.67(cid:5)0.24 27.08(cid:5)0.24 194............. 27.67(cid:5)0.42 27.86(cid:5)0.45 28.31(cid:5)0.44 28.10(cid:5)0.48 22.25(cid:5)0.26 22.26(cid:5)0.26 23.25(cid:5)0.27 24.26(cid:5)0.28 22.63(cid:5)0.22 23.49(cid:5)0.23 24.76(cid:5)0.25 27.35(cid:5)0.29 222............. 24.59(cid:5)0.36 24.85(cid:5)0.36 25.56(cid:5)0.38 26.10(cid:5)0.41 21.78(cid:5)0.22 22.01(cid:5)0.23 23.03(cid:5)0.26 23.96(cid:5)0.26 21.58(cid:5)0.28 22.24(cid:5)0.29 24.27(cid:5)0.28 26.74(cid:5)0.31 243............. 24.43(cid:5)0.35 24.21(cid:5)0.37 24.53(cid:5)0.37 25.93(cid:5)0.39 21.55(cid:5)0.21 22.18(cid:5)0.24 23.17(cid:5)0.23 24.15(cid:5)0.25 21.66(cid:5)0.32 22.68(cid:5)0.35 24.50(cid:5)0.34 26.85(cid:5)0.37 285............. 25.83(cid:5)0.36 26.01(cid:5)0.35 26.54(cid:5)0.37 26.94(cid:5)0.38 23.00(cid:5)0.33 23.82(cid:5)0.34 25.41(cid:5)0.34 26.58(cid:5)0.35 22.01(cid:5)0.31 22.87(cid:5)0.32 24.94(cid:5)0.36 29.01(cid:5)0.37 303............. 23.93(cid:5)0.31 24.23(cid:5)0.33 23.98(cid:5)0.34 24.04(cid:5)0.36 21.47(cid:5)0.24 22.01(cid:5)0.25 22.93(cid:5)0.25 23.75(cid:5)0.26 20.98(cid:5)0.26 21.26(cid:5)0.26 22.85(cid:5)0.28 26.05(cid:5)0.30 323............. 21.58(cid:5)0.22 21.74(cid:5)0.23 21.32(cid:5)0.22 21.34(cid:5)0.25 21.86(cid:5)0.23 21.93(cid:5)0.24 22.59(cid:5)0.26 23.02(cid:5)0.28 19.80(cid:5)0.20 20.08(cid:5)0.21 21.29(cid:5)0.22 24.02(cid:5)0.24 368............. 24.61(cid:5)0.38 25.17(cid:5)0.38 26.89(cid:5)0.40 28.21(cid:5)0.43 ... ... ... ... 23.25(cid:5)0.21 23.98(cid:5)0.22 25.41(cid:5)0.24 30.75(cid:5)0.28 404............. 24.52(cid:5)0.38 24.55(cid:5)0.37 25.09(cid:5)0.39 26.01(cid:5)0.40 22.11(cid:5)0.28 22.73(cid:5)0.31 23.79(cid:5)0.33 24.09(cid:5)0.34 21.83(cid:5)0.28 22.67(cid:5)0.30 24.05(cid:5)0.35 26.70(cid:5)0.38 412............. 26.52(cid:5)0.40 26.08(cid:5)0.41 26.25(cid:5)0.41 26.14(cid:5)0.43 24.33(cid:5)0.20 24.81(cid:5)0.22 25.40(cid:5)0.24 25.50(cid:5)0.24 25.21(cid:5)0.36 25.80(cid:5)0.37 26.81(cid:5)0.38 29.30(cid:5)0.41 439............. 24.27(cid:5)0.33 23.70(cid:5)0.33 23.93(cid:5)0.32 24.65(cid:5)0.35 21.46(cid:5)0.24 21.93(cid:5)0.24 22.89(cid:5)0.25 23.30(cid:5)0.27 21.01(cid:5)0.26 21.66(cid:5)0.27 23.17(cid:5)0.32 26.83(cid:5)0.33 451............. 24.57(cid:5)0.38 24.88(cid:5)0.37 25.12(cid:5)0.39 25.43(cid:5)0.40 22.73(cid:5)0.31 22.97(cid:5)0.31 24.63(cid:5)0.33 25.83(cid:5)0.36 20.87(cid:5)0.28 21.74(cid:5)0.30 23.63(cid:5)0.31 26.31(cid:5)0.35 CDF-S 34............... 23.85(cid:5)0.35 24.89(cid:5)0.35 25.18(cid:5)0.36 25.20(cid:5)0.36 22.40(cid:5)0.33 22.68(cid:5)0.33 23.63(cid:5)0.34 24.17(cid:5)0.36 22.74(cid:5)0.32 23.51(cid:5)0.33 24.91(cid:5)0.36 27.87(cid:5)0.38 40............... 25.04(cid:5)0.40 25.34(cid:5)0.41 25.63(cid:5)0.43 25.84(cid:5)0.45 22.86(cid:5)0.34 23.19(cid:5)0.35 23.68(cid:5)0.35 24.58(cid:5)0.36 21.76(cid:5)0.21 21.98(cid:5)0.24 23.61(cid:5)0.28 25.90(cid:5)0.30 42............... 18.96(cid:5)0.22 19.10(cid:5)0.24 19.25(cid:5)0.26 19.36(cid:5)0.26 ... ... ... ... 23.18(cid:5)0.32 24.30(cid:5)0.32 26.02(cid:5)0.35 28.68(cid:5)0.35 44............... 25.29(cid:5)0.38 25.16(cid:5)0.40 25.51(cid:5)0.43 26.04(cid:5)0.46 22.11(cid:5)0.28 22.74(cid:5)0.29 23.66(cid:5)0.31 24.78(cid:5)0.31 22.31(cid:5)0.31 22.64(cid:5)0.32 24.47(cid:5)0.33 27.59(cid:5)0.35 47............... 23.45(cid:5)0.33 23.62(cid:5)0.34 23.88(cid:5)0.36 24.57(cid:5)0.36 21.93(cid:5)0.26 22.32(cid:5)0.27 23.48(cid:5)0.27 24.04(cid:5)0.28 21.86(cid:5)0.29 22.34(cid:5)0.29 23.76(cid:5)0.30 26.52(cid:5)0.32 48............... 23.47(cid:5)0.34 23.60(cid:5)0.35 24.18(cid:5)0.35 24.36(cid:5)0.37 21.81(cid:5)0.25 22.44(cid:5)0.25 23.32(cid:5)0.27 24.69(cid:5)0.28 20.52(cid:5)0.20 20.92(cid:5)0.21 22.23(cid:5)0.21 25.10(cid:5)0.23 51............... 25.82(cid:5)0.39 25.67(cid:5)0.39 26.19(cid:5)0.41 27.42(cid:5)0.43 24.56(cid:5)0.35 24.86(cid:5)0.36 26.05(cid:5)0.37 27.67(cid:5)0.39 24.02(cid:5)0.28 24.23(cid:5)0.30 26.10(cid:5)0.32 27.91(cid:5)0.35 52............... 23.02(cid:5)0.35 22.88(cid:5)0.34 23.59(cid:5)0.35 23.57(cid:5)0.38 20.90(cid:5)0.23 21.33(cid:5)0.24 22.08(cid:5)0.26 24.38(cid:5)0.28 22.21(cid:5)0.30 23.05(cid:5)0.32 24.31(cid:5)0.33 26.56(cid:5)0.33 151............. 25.41(cid:5)0.38 25.48(cid:5)0.38 26.14(cid:5)0.39 26.67(cid:5)0.42 22.59(cid:5)0.30 23.30(cid:5)0.30 24.18(cid:5)0.32 25.35(cid:5)0.34 22.16(cid:5)0.24 22.64(cid:5)0.24 23.99(cid:5)0.26 27.04(cid:5)0.28 156............. 25.54(cid:5)0.39 25.96(cid:5)0.38 25.66(cid:5)0.40 25.83(cid:5)0.43 23.26(cid:5)0.34 23.37(cid:5)0.35 24.61(cid:5)0.37 24.57(cid:5)0.38 21.72(cid:5)0.19 21.85(cid:5)0.20 22.97(cid:5)0.20 25.39(cid:5)0.22 170............. 25.97(cid:5)0.40 25.02(cid:5)0.41 26.11(cid:5)0.41 25.92(cid:5)0.44 21.18(cid:5)0.23 21.64(cid:5)0.24 22.71(cid:5)0.26 23.84(cid:5)0.27 19.85(cid:5)0.22 20.27(cid:5)0.24 21.60(cid:5)0.25 24.05(cid:5)0.27 189............. 26.01(cid:5)0.38 26.21(cid:5)0.39 26.47(cid:5)0.39 27.20(cid:5)0.41 24.02(cid:5)0.35 24.57(cid:5)0.38 25.09(cid:5)0.38 26.60(cid:5)0.39 22.25(cid:5)0.28 22.89(cid:5)0.30 24.84(cid:5)0.30 27.46(cid:5)0.32 192............. 25.56(cid:5)0.39 25.72(cid:5)0.40 25.91(cid:5)0.42 26.60(cid:5)0.44 22.20(cid:5)0.30 22.80(cid:5)0.30 23.76(cid:5)0.31 24.67(cid:5)0.32 21.81(cid:5)0.27 22.27(cid:5)0.28 24.01(cid:5)0.31 26.65(cid:5)0.33 201............. 25.77(cid:5)0.33 25.58(cid:5)0.35 25.52(cid:5)0.37 25.62(cid:5)0.38 ... ... ... ... 26.85(cid:5)0.37 27.34(cid:5)0.38 29.06(cid:5)0.38 30.58(cid:5)0.39 252............. 25.76(cid:5)0.35 26.29(cid:5)0.33 25.84(cid:5)0.34 26.12(cid:5)0.34 ... ... ... ... 24.62(cid:5)0.19 25.04(cid:5)0.20 26.49(cid:5)0.23 28.01(cid:5)0.26 257............. 26.68(cid:5)0.40 26.51(cid:5)0.40 27.01(cid:5)0.41 27.57(cid:5)0.42 23.27(cid:5)0.34 23.11(cid:5)0.34 23.37(cid:5)0.36 24.16(cid:5)0.36 22.81(cid:5)0.25 23.31(cid:5)0.26 25.06(cid:5)0.28 27.04(cid:5)0.30 266............. 24.92(cid:5)0.37 25.80(cid:5)0.38 25.94(cid:5)0.37 26.54(cid:5)0.40 21.35(cid:5)0.23 21.89(cid:5)0.23 22.64(cid:5)0.25 23.40(cid:5)0.27 21.91(cid:5)0.28 22.14(cid:5)0.28 23.94(cid:5)0.29 26.80(cid:5)0.30 511............. 24.73(cid:5)0.37 24.39(cid:5)0.37 24.61(cid:5)0.39 25.24(cid:5)0.40 22.17(cid:5)0.30 22.66(cid:5)0.30 23.99(cid:5)0.33 25.05(cid:5)0.34 21.58(cid:5)0.25 21.98(cid:5)0.26 23.41(cid:5)0.26 27.10(cid:5)0.29 613............. 24.43(cid:5)0.37 25.35(cid:5)0.37 26.01(cid:5)0.39 27.83(cid:5)0.40 22.50(cid:5)0.25 23.11(cid:5)0.26 24.02(cid:5)0.26 24.75(cid:5)0.27 22.91(cid:5)0.30 23.67(cid:5)0.30 25.77(cid:5)0.33 28.09(cid:5)0.34 Notes.—Col.(1):sourcenumber,asinTable1;cols.(2)Y(5):totalmagnitude(notcorrectedforGalacticextinction)ofthePSFcomponentintheHSTfiltersz, i,V,andB,respectively;cols.(6)Y(9):totalmagnitude(not correctedforGalacticextinction)ofthe exponentialcomponentintheHSTfiltersz,i, V,andB,respectively;cols.(10)Y(13):totalmagnitude(notcorrectedforGalacticextinction)ofthedeVaucouleurscomponentintheHST filtersz,i, V,andB,respectively.Errorsarequotedata68%ofdeclaredconfidencelevel(seeAppendixBfordetails). 102 BALLO ETAL. Vol. 667 covering an area of 0.1 deg2 centered on the CDF-N. The HK0 TABLE3 filtercoversboththeHandK0bandsinasinglefilter;itallowsa Two-dimensionalImageFittingMorphologicalParameters greater depth but involves some loss of color information. HK0 Disk Bulge datahavebeencollectedusingtheQUIRCcameraontheUniver- sityofHawaii2.2mtelescope,witha3:60;3:60fieldofviewand ID Redshift r P.A.(deg) b/a r P.A.(deg) b/a withanABmagnitudelimitacrossthefieldof22.1(ina90;90 d e (1) (2) (3) (4) (5) (6) (7) (8) fieldaroundtheCDF-Nthebandreachesalimitof22.8mag). CDF-N 3. DATA ANALYSIS 114........ 0.679 ... ... ... 1.57 (cid:1)6.62 0.87 Inordertodisentanglethemaingalacticcomponents(nucleus, 141........ 0.747 6.38 (cid:1)20.54 0.99 8.61 (cid:1)19.36 0.95 bulge, and disk) for the AGNs in our sample, we adopt a two- 150........ 0.762 3.59 (cid:1)16.31 0.65 5.26 (cid:1)21.31 0.62 dimensional fitting approach, applied to the GOODS HSTACS 170........ 0.680 2.77 (cid:1)9.47 0.32 7.09 (cid:1)8.47 0.91 images.Inparticular,toseparateatypicallyfaintnucleusfromthe 194........ 0.555 0.39 (cid:1)41.96 0.93 1 (cid:1)43.96 0.99 surroundingbrightbulge,weneedtodetermineaccurateandquan- 222........ 0.857 70.09 49.34 0.52 11.51 48.85 0.57 titativemorphologicalinformation.Ithasbeenshownthatmod- 243........ 0.961 1.38 58.78 0.77 4.01 60.30 0.80 elingintwodimensionsallowsabetterestimateoftheparameter 285........ 0.955 1.05 42.38 0.84 6 41.38 0.86 303........ 0.678 2.08 58.11 0.43 3.46 59.00 0.53 values in a bulge-disk decomposition (e.g., Byun & Freeman 323........ 0.514 3.19 (cid:1)86.18 0.81 4.09 (cid:1)88.77 0.76 1995;Wadadekaretal.1999).Toperformthefitweusethetwo- 368........ 0.935 ... ... ... 1.86 ...79.85 0.68 dimensionalimagedecompositionprogramGALFIT(Pengetal. 404........ 0.978 10.27 (cid:1)0.79 0.40 10.60 (cid:1)2.79 0.50 2002).Agalaxymodelisapplied,composedbyabulge,adisk, 412........ 0.474 0.51 (cid:1)54.11 0.35 0.81 (cid:1)56.33 0.31 andanuclearsource.Thedetailsoftheprocedure,aswellashow 439........ 0.839 3.01 84.41 0.79 3.90 83.90 0.86 weestimatedtheerrorsassociatedwiththemagnitudesofthedif- 451........ 0.838 1.54 20.62 0.48 2.36 19 0.52 ferentcomponents,arepresentedinAppendixB.Herewewantto stressthatinoursurfacephotometrydecompositionwetakead- CDF-S vantage of imagesin four different bands,in which the compo- 34.......... 0.839 1.28 (cid:1)7.22 0.48 4.59 (cid:1)6.22 0.46 nents of the host galaxy have a different relative importance. 40.......... 0.668 0.24 (cid:1)43.84 0.85 0.97 (cid:1)56.30 0.87 Results for the 34 sources analyzed are reported in Table 2 42.......... 0.734 ... ... ... 1.40 (cid:1)20.30 0.55 (magnitudesofthenucleusincols.[2]Y[5];magnitudesofthe 44.......... 0.734 1.37 (cid:1)2.07 0.68 1.32 (cid:1)2.70 0.67 diskincols.[6]Y[9];andmagnitudesofthebulgecomponentin 47.......... 0.733 0.62 25.96 0.86 1.27 36.17 0.96 cols. [10]Y[13]) and Table 3 (the host galaxy morphological 48.......... 0.534 10.00 48.40 0.82 10.23 62.10 0.82 parameters, i.e., scale length, position angle, and ratio of the 51.......... 0.670 1.21 (cid:1)57.39 0.80 0.64 (cid:1)51.41 0.66 semiminoraxistothesemimajoraxisforthediskaregivenin 52.......... 0.569 0.81 (cid:1)55.89 0.80 2.40 (cid:1)53.23 0.77 cols.[3]Y[5]andforthebulgeincols.[6]Y[8]). 151........ 0.604 1 (cid:1)69.09 0.28 2.37 (cid:1)70.40 0.26 156........ 0.545 4.95 (cid:1)16.24 0.79 5.41 (cid:1)14.24 0.76 InFigure4aweshowasanexampletheresultofthedecom- 170........ 0.664 1.58 33.21 0.52 4.56 32.91 0.53 position in a typical case (ID 34, CDF-S; z¼0:839). Similar 189........ 0.734 4.08 (cid:1)56.78 0.30 3.27 (cid:1)54.78 0.32 imagesforalltheanalyzedsourcesarepresentedintheelectronic 192........ 0.735 1.28 (cid:1)19.12 0.50 2.50 17.88 0.52 edition. The residuals after subtracting the final model (bottom 201........ 0.679 ... ... ... 1.20 21.36 0.33 panels) from the original images (top panels) in the four bands 252........ 0.481 ... ... ... 0.45 (cid:1)27.36 0.60 showthattheobjectiswellmodeled,apartfromtheresidualstrac- 257........ 0.549 2.21 3.38 0.31 2.12 3.27 0.27 ingthespiralarms(neveraccountedforintheappliedgalaxy 266........ 0.735 3.79 (cid:1)21.12 0.26 2.95 (cid:1)20.80 0.27 model)andsomeparticularlyintensespotsofemission. 511........ 0.668 4.78 16.97 0.93 5.14 18.97 0.90 Sincetheanalysisiscarriedoutseparatelyinthefourbands, 613........ 0.910 4.29 43.99 0.44 6.39 44.59 0.57 noaprioriconstraintisimposedonthemeasuredspectralenergy Notes.—Col.(1):sourcenumber,asinTable1;col.(2):redshiftfortheop- distributionsfortheindividualcomponents:bulge,disk,andnu- ticalcounterpart,asinTable1;cols.(3)Y(5):scalelength,positionangle,andsemi- cleus.Itisworthrecallingthatreddeningmaysignificantlyaffect minoraxistosemimajoraxisratiooftheexponentialdisk;cols.(6)Y(8):scalelength, theobservedSEDs.IntrinsicnuclearemissionintheUVandop- positionangle,andsemiminoraxistosemimajoraxisratioofthedeVaucouleurs law. ticalrangesmaybealteredbydustabsorptionfromcircumnuclear and/ormorediffusecomponent(s).Thelatterisalsorelevantfor thegalacticcomponents,particularlyforthedisk,whichiscom- elements in the interstellar gas and Yis the fraction by mass of posedbyyoungerandbluerstars. helium(thesolarvaluesareZ ¼0:02andY ¼0:21).Forboth (cid:3) (cid:3) Bearingthisinmind,asafirsttestofthegoodnessofouranal- metal contents, integrated spectra of SSP of different ages have ysiswecomparetheresultingSEDswithsuitabletemplates.SED beencomparedwiththemeanempiricalspectrum,findingtheage templates for the bulge have been derived using a set of single thatprovidesthebestagreement:10GyrforZ ¼0:05,and11Gyr stellarpopulation(SSP)SEDs.9FromthissetofSEDsweselecta forZ ¼0:02.Forthetwometallicities,havingfixedtheepochof templaterepresentingthe‘‘typical’’ellipticalgalaxyatz(cid:4)0,by formation,wecomputetheintegratedSSPspectrumattheredshift comparison with a mean local spectrum (obtained from the ofeachanalyzedsource.Whileforthebulgecomponentweas- observedonesof NGC1399andNGC1404;A.Bressan2005, sumedthatstarsformedinasingleburst,fortheSEDsofthedisk privatecommunication).Totakeintoaccountthewell-known componentweassumedacontinuousstarformationduringitslife. metallicity-agedegeneracy,weassumetwoinitialchemicalcom- TheSSPsusedarethesameasforthebulge.Finally,tocheckthe positions of the evolutionary sequences: [Z ¼0:02, Y ¼0:28], nuclear optical SED, we adopt the QSO template spectrum of and[Z ¼0:05,Y ¼0:352],whereZisthemassfractionofheavy Cristiani&Vio(1990)downto5388,modifiedasdescribedby Cristiani et al. (2004) and Monaco & Fontanot (2005) and ex- 9 Availableathttp://web.pd.astro.it/granato/grasil/SSP/ssp.html;seealsoSilva trapolatedto3008using f(cid:2) /(cid:2)(cid:1)1:75(followingRisaliti&Elvis etal.(1998). 2004). No. 1, 2007 M AND EDDINGTON RATIOS OFAGNs AT 0.4 < z < 1 103 BH Fig.4.—Resultoftheanalysisforatypicalcase[ID34,CDF-S;z¼0:839].(a)Originalimagesinthefourbands(toppanels)andresidualsaftersubtractingthefinal modelfromtheoriginalimages(bottompanels;seex3andAppendixB).(b)Leftpanel:Fitofthetemplates(bulge,dashedline;disk,dottedline;nucleus,solidline)tothe decomposedopticalmagnitudesforthedifferentcomponents(bulge,squares;disk,stars;andnucleus,circles),andcomparisonofthesumofthetemplates(dot-dashed line)withthefluxesobservedfromthewholegalaxyinalltheavailablebands(diamonds);acontributionfromacircumnucleartorus(long-dashedline),adaptedfromthe meanSEDofatypicalSeyfert1galaxyreportedinGranato&Danese[1994]),isassumedtoaccountforSpitzerdata.Rightpanel:FitoftheQSOtemplate(solidline)to thenuclearSED(opticalmagnitudes,circles;andX-rayflux,squares;grayareasrepresenttheuncertaintyintheopticalemission,estimatedasdescribedinx4.2,andthe consequentuncertaintyintheUVrange).[SeetheelectroniceditionoftheJournalforacolorversionofthisfigure.] InFigure 4b(leftpanel) we showthecomparisonofthede- remaining10casestheemissionintheVandespeciallyBbands composed SEDs with the three templates and the magnitudes islower thanexpectedfromtheSEDnormalizedtotheiandz forthewholegalaxyinalltheavailablebandsforagenericcase bands.Thepresenceofsignificantobscurationaroundthesenuclei (ID34,CDF-S;z¼0:839).Thesamecomparisonforthewhole isconfirmedbytheirhighcolumndensity,N (cid:6)1022 cm(cid:1)2,in- H sampleisshownintheelectronicedition.Forthebulgecompo- ferredfromtheirX-rayemission. nent the observed SEDs are in good agreement with the tem- Fluxesfromthedecompositionandtemplateshavealsobeen plate((cid:3)2perdegreeoffreedombetween0.8Y0.9and1.3Y1.5). compared with themagnitudes for thewhole galaxy recovered We stress that the bulge component is expected to be less af- fromtheGOODScatalog,aswellaswiththemagnitudesinthe fectedbyabsorption. Even forthediskcomponent verygood IR(includingdatafromSpitzer)andUband,whenavailable.For agreementisgenerally found,withoutinvokinglargecorrection the U-band and Spitzer data the angular resolution is not good forabsorption. enoughtoletustomakeamorphologicaldecomposition.Never- Concerning the nuclei, the agreement of the observed SEDs theless,wecanobtainimportantindicationsbycomparingtheob- withthetemplateisgoodfor70%oftheanalyzedobjects;inthe servedtotalfluxeswiththesumofthetemplatesofbulge,disk, 104 BALLO ETAL. Vol. 667 3.1.Host-dominated and Nucleus-dominated Sources Whentheopticalimagesaredominatedatallwavelengthsby oneofthegalaxycomponents(nucleus,bulge,ordisk),theother componentscannotbetightlyconstrained. InsourcesdominatedbytheAGN(twoobjectsinthepresent sample),wecanonlyputupperlimitstothehostcomponentby minimizingthenuclearcontribution(whichiswellconstrained). Weputathresholdonthenucleus-to-totalluminosityratioN/T (reportedinTable4)inthezbandofN/T(z)¼0:2asafiducial valuetoidentifythegalaxiesaffectedbythisproblem.Forthetwo analyzedsourcessatisfyingthiscriterion(ID42,CDF-S,N/T(z)¼ 0:329; and ID 201, CDF-S, N/T(z)¼0:293),we proceed as follows. 1. WeassumefortheQSOtemplateanormalization2(cid:4)lower thanthebestfit(wheretheuncertaintiesareevaluatedasdescribed inAppendixB),determiningaminimumvalueforthenuclear magnitudesinthefourbands. 2. Assumingthesenuclearmagnitudes,werescalethebulge component10untiltheobservedvalueofthetotalfluxesisreached. 3. Finally, we compare the new galaxy reconstruction (i.e., thesumofthetemplatesof bulgeandnucleusrescaled)withthe emissionintheIRbands,checkingthatthenewestimatedoesnot Fig.5.—AveragesurfacebrightnesswithintheeffectiveradiusintheV-band overpredict(atmorethan2(cid:4)level)theemissionatwavelengths (Johnsonmagnitude)vs.theeffectiveradius.Thesurfacebrightnessforeach greaterthan1(cid:1)m. bulgehostinganX-ray-selectedAGNs(triangles)hasbeenderivedfromthe parametersreportedinTable2andTable3(withaK-correctionappliedtoconvert Forbothsourceswefindthatthebulgeluminositycanincrease theobservedi-bandABmagnitudesinrest-frameV-bandJohnsonmagnitudes). uptoafactorof10withrespecttotheresultofthedecomposition. Thedistributionfoundfortheanalyzedsourcesisingoodagreementwiththatof Weconsiderthesevaluesrobustupperlimits. localellipticalgalaxiesandbulgesstudiedbyBenderetal.(1992)convertedin Ontheotherhand,whentheopticalemissionisdominatedby theVbandassumingB(cid:1)V ¼0:96andcorrectedforthedifferentcosmology adopted(crossesinthefigure).ThesameregionintheSB-r planeisoccupiedby thehostgalaxycomponents(bulgeand/ordisk,sixobjects),the e e the18bulgesofSeyfert1andSeyfert1.5galaxiesobservedintheV-bandby nuclearcontributionprovidedbyGALFITcouldbealowerlimit Granatoetal.(1993),correctedforthedifferentcosmologyadopted(dots).We totheactualone.Wereanalyzedseparatelythesixsourceswith alsofindagoodagreementwiththedistributionreportedbyFasanoetal.(1998)for nucleus-to-totalluminosityratiointheVbandN/T(V)<0:05 asampleof23early-typegalaxieswithspectroscopicredshiftzP3:4,withK-and evolutionarycorrectionsapplied(seetheirFig.6)andcorrectedforthedifferent (seeTable4). cosmologyadopted(opensquares). 1. We assume for bulge and disk templates a normalization 2(cid:4)lowerthanthebestfit(wheretheuncertaintiesareevaluated andnucleus,normalizedtothedecomposedmagnitudes.Inpar- asdescribedinAppendixB),determiningaminimumvaluefor ticular, we are abletoreproduce the Spitzer data byassuming a theirmagnitudesinthefourbands. contributionfromacircumnucleartorus,asobservedinatypical 2. Upperlimitstothenuclearmagnitudesarethencalculated Seyfert1galaxy(seeGranato&Danese1994). imposingthatthetotal(hostplusnucleus)valuesareequaltothe TheconstraintsderivedfromtheobservedIRemissionforthe observedones. fluxexpectedonthebasisofthisreconstructionareparticularly 3. WecheckthatthenewestimateofthenucleusintheUband importantwhenthenuclearcomponentisdominantandtendsto isnothigher(atmorethan2(cid:4)level)thantheobservedflux. overwhelmtheemissionfromthehostgalaxy:inthesecasesthe Again,theresultsofthisprocedureareconsideredrobustupper magnitudesdeterminedforbulgeanddiskarenotcompletelyre- limitstothenuclearluminosity. liable,andthismayaffecttheBHmassdetermination(seex4.1). Asdescribedinmoredetailinx3.1,insuchcaseswecanonly 4. RECOVERING THE NUCLEAR PROPERTIES provide upper limits for the bulge magnitudes; constraints im- posedbytheobservationsintheIRrangepreventusfromover- Disentanglingthedifferentgalacticcomponentsasdescribed estimatingthisfaintercomponent. in x 3 and in Appendix B provides detailed information on the The bulge component is relevant not only in the estimate of bulgeandnuclearluminosity.Thisinformationisrelevantinorder thenuclearcontribution,butalsoindeterminingtheBHmasses tostudythenuclearactivityofthegalaxiesinoursample.Inx4.1 (seex4).Therefore,wecheckedthattherelationbetweentheef- wepresentourderivationoftheBHmass,whilethebolometric fectiveradiusr andthesurfacebrightnesswithinit,SB ,iscon- luminosityisevaluatedfromopticalandX-raynuclearluminos- e e sistent with that derived for local samples. In particular, from itiesasdescribedinx4.2. Figure5itisapparentthatthebulgesderivedthroughouranal- 4.1. Black Hole Masses ysisaredistributedinther-SB planeasthesampleofnearby e e ellipticalgalaxiesandbulgesstudiedbyBenderetal.(1992).Itis Startingfromtheabsolutemagnitudesofthebulgecomponent, interestingtonotethatthesameregionoftheplaneisoccupied weobtainanestimateofthemassofthecentralcompactobject, bythebulgesoflocalSeyfertgalaxies,asdeterminedbyGranato et al. (1993), as well as by a sample of early-type galaxies at 10 Wenotethatforboththesourcesreanalyzedthediskcomponentisnot higherredshift(seeFasanoetal.1998). requestedinthefit. No. 1, 2007 M AND EDDINGTON RATIOS OFAGNs AT 0.4 < z < 1 105 BH TABLE4 RatiooftheDifferentComponentsMagnitudes,asObtainedfromtheTwo-dimensionalImageFitting Bulge/Galaxy Nucleus/Bulge Nucleus/Total ID Redshift z i V B z i V B z i V B (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) CDF-N 114................ 0.679 1.000 1.000 1.000 1.000 0.349 0.396 1.739 25.26 0.115 0.124 0.241 0.325 141................ 0.747 0.739 0.675 0.636 0.234 0.029 0.045 0.133 1.986 0.020 0.028 0.068 0.194 150................ 0.762 0.666 0.643 0.428 0.022 0.040 0.041 0.132 3.631 0.024 0.025 0.048 0.065 170................ 0.680 0.747 0.712 0.560 0.101 0.023 0.025 0.052 0.643 0.016 0.017 0.027 0.054 194................ 0.555 0.413 0.245 0.200 0.055 0.010 0.018 0.038 0.501 0.004 0.004 0.007 0.025 222................ 0.857 0.546 0.447 0.242 0.072 0.063 0.091 0.305 1.809 0.031 0.036 0.060 0.093 243................ 0.961 0.476 0.386 0.227 0.077 0.078 0.245 0.972 2.323 0.033 0.074 0.133 0.116 285................ 0.955 0.714 0.705 0.605 0.097 0.030 0.055 0.229 6.742 0.020 0.035 0.098 0.221 303................ 0.678 0.610 0.667 0.518 0.107 0.066 0.065 0.354 6.362 0.036 0.038 0.118 0.224 323................ 0.514 0.870 0.846 0.768 0.285 0.194 0.218 0.967 11.805 0.112 0.119 0.230 0.303 368................ 0.935 1.000 1.000 1.000 1.000 0.286 0.336 0.257 10.365 0.100 0.112 0.093 0.313 404................ 0.978 0.565 0.515 0.441 0.083 0.084 0.177 0.383 1.905 0.041 0.072 0.112 0.107 412................ 0.474 0.309 0.287 0.214 0.029 0.298 0.774 1.676 18.345 0.072 0.133 0.173 0.205 439................ 0.839 0.602 0.560 0.436 0.037 0.050 0.153 0.495 7.399 0.027 0.068 0.131 0.151 451................ 0.838 0.847 0.756 0.715 0.391 0.033 0.056 0.254 2.254 0.026 0.038 0.118 0.242 CDF-S 34.................. 0.839 0.422 0.318 0.235 0.032 0.360 0.280 0.777 11.665 0.104 0.070 0.118 0.176 40.................. 0.668 0.734 0.753 0.516 0.229 0.049 0.045 0.156 1.057 0.032 0.031 0.065 0.140 42.................. 0.734 1.000 1.000 1.000 1.000 48.591 119.994 511.588 5323.043 0.329 0.331 0.333 0.333 44.................. 0.734 0.454 0.523 0.322 0.070 0.064 0.098 0.384 4.169 0.027 0.044 0.090 0.155 47.................. 0.733 0.516 0.495 0.436 0.092 0.231 0.308 0.895 6.026 0.088 0.105 0.180 0.209 48.................. 0.534 0.766 0.802 0.732 0.407 0.066 0.085 0.166 1.977 0.044 0.056 0.089 0.236 51.................. 0.670 0.622 0.641 0.488 0.445 0.191 0.265 0.920 1.570 0.087 0.113 0.191 0.226 52.................. 0.569 0.231 0.170 0.114 0.118 0.474 1.170 1.941 15.785 0.083 0.125 0.133 0.283 151................ 0.604 0.599 0.649 0.544 0.174 0.050 0.073 0.138 1.404 0.027 0.041 0.061 0.141 156................ 0.545 0.805 0.802 0.818 0.318 0.030 0.023 0.085 0.668 0.022 0.017 0.057 0.130 170................ 0.664 0.773 0.780 0.736 0.450 0.004 0.013 0.016 0.181 0.003 0.010 0.011 0.065 189................ 0.734 0.836 0.825 0.557 0.310 0.031 0.047 0.223 1.273 0.024 0.035 0.091 0.181 192................ 0.735 0.588 0.619 0.441 0.139 0.032 0.042 0.175 1.047 0.018 0.024 0.063 0.101 201................ 0.679 1.000 1.000 1.000 1.000 2.723 5.054 26.065 96.235 0.268 0.294 0.325 0.331 252................ 0.481 1.000 1.000 1.000 1.000 0.348 0.318 1.813 5.721 0.114 0.108 0.244 0.299 257................ 0.549 0.605 0.454 0.175 0.066 0.028 0.052 0.165 0.616 0.016 0.022 0.027 0.036 266................ 0.735 0.374 0.443 0.233 0.042 0.063 0.034 0.158 1.267 0.022 0.015 0.033 0.046 511................ 0.668 0.634 0.652 0.630 0.132 0.055 0.107 0.331 5.516 0.031 0.058 0.128 0.229 613................ 0.910 0.407 0.375 0.166 0.044 0.247 0.212 0.801 1.275 0.077 0.064 0.095 0.048 Notes.—Col.(1):sourcenumber,asinTable1;col.(2):redshiftfortheopticalcounterpart,asinTable1;cols.(3)Y(6):bulge-to-galaxyratiointhefourbands; cols.(7)Y(10):nucleus-to-bulgeratiointhefourbands;cols.(11)Y(14):nucleus-to-totalratiointhefourbands. exploitingtheobservedlocalrelationshipbetweenBHmassand thatof localquiescentgalaxies.Thereareclaimsthatthecorre- bulge component luminosity. Several versions of this relation lationbetweenBHmassandbulgevelocitydispersioncouldbe havebeenproposedwiththebulgeluminosityevaluatedatelec- subjecttoacosmicevolution(Wooetal.2006);however,sucha tromagnetic bandsrangingfromBtoK(see,e.g., Kormendy& behaviorwouldproduceBHmassestimatesslightlyhigherthan Gebhardt2001;McLure&Dunlop2002;Marconi&Hunt2003). thatpresentedinthefollowing.Ontheotherhand,itishardto InordertocomputetheBHmass,weassumethat(1)inthis imagine how a significant independent evolution can lead to a relationtheabsolutemagnitudeofthebulgestrictlymirrorsthe M -M relationwithascatterassmallasthatreportede.g., bulge BH massinoldstars,M ,whichisthequantityprimarilyrelated byFerrarese&Ford(2005). bulge totheBHmass;and(2)theM -M relationisimprintedat The relation with the total R-band magnitude of the bulge bulge BH highredshift,drivingthemainepisodeofaccretion. reportedinMcLure&Dunlop(2002)forasampleof20inac- The latter hypothesis is supported by the findings of Peng tiveE-typegalaxiesatz(cid:4)0andconvertedtoH ¼70kms(cid:1)1 0 etal.(2006),whoshowthattheM -M relationsforfiveAGNs Mpc(cid:1)1reads BH R atz(cid:2)1arecompatiblewiththelocalrelation,oncepassiveevo- (cid:1) (cid:2) lutionofthestellarpopulationisallowedfor,andbyMcLure& log MBH ¼(cid:1)0:50((cid:5)0:03)M (cid:1)2:69((cid:5)0:72); ð1Þ Dunlop(2002),who,analyzingwiththesametechniqueasam- M R (cid:3) pleof 72activegalaxies(SeyfertgalaxiesandQSOs)atzP0:5, demonstrate thatAGNhostgalaxies attheseredshifts followa withascatterof(cid:1)log(M )¼0:33.TheM -M relation BH bulge BH relationbetweenBHmassandbulgeluminosityconsistentwith reportedbyBettonietal.(2003)intheirequation(1)forasample 106 BALLO ETAL. Vol. 667 of nearby inactive ellipticals is in good agreement with equa- AGNsatlowredshiftspanthewholerangeinmass,whileatthe tion(1),oncethedifferentcosmologyistakenintoaccount(even upperendoftheredshiftdistributiononlylarge(M >107M ) BH (cid:3) ifinthissecondcasethescatterislarger,0.39dex).Itisworthnot- BHmassesarefound(seeFig.6,toppanels).Noobviousselec- ingthatequation(1)hasbeenderivedusingB-andV-bandmag- tioneffects,inthisregard,havebeenidentified.Suchabehavior nitudes,translatedtoRbandassumingaveragecolors(B(cid:1)R)¼ seemstobeanintrinsicpropertyofthesample.Wedonotsee 1:57and(V (cid:1)R)¼0:61.The relation reported by Marconi& anycorrelationbetweentheestimatedbulgeluminosityandthe Hunt(2003)exploitstheK-bandmagnitudesandonaveragepre- X-rayluminosity.Asaconsequence,nocorrelationisfoundalso dictsBHmasseshigherbyabout0.3dexatfixedluminosity.The betweenBHmassandX-rayluminosity. difference could be ascribed to the uncertainty in evaluating thebulgecomponentofspiralgalaxiesoftheobservedsample 4.2.Nuclear Bolometric Luminosities (Shankaretal.2004).TherelationshipestimatedbyKormendy Inordertogaininsightintotheaccretionratespoweringthese &Gebhardt(2001)using B-bandbulgeluminositiesyieldsBH sources,weneedtoinvestigatetheirnuclearbolometricluminos- masseslargeratmostbyabout0.2dexthanthosepredictedon ity.Wecomparethenuclearemission(i.e.,theX-rayflux,totally the basis of equation (1). On the other hand, Bernardi et al. ascribedtotheAGN,andtheopticalnuclearmagnitudes)toSEDs (2007)andTundoetal.(2006)havesuggestedthattheM -L BH of active nuclei with different X-rayYtoYoptical ratios. In these relationisbiasedtowardpredictingmoremassiveblackholes SEDs,thehigh-energyemissionisdescribedasapowerlawwith foragivenluminosity.The intrinsic M -Lrelation proposed BH photonindex(cid:2).FortheopticalbandsweadopttheQSOtemplate byBernardietal.(2007)intherangeofluminositiesofinterest spectrumdescribedinx3. forthepresentworkyieldsBHmassesthatareveryclosetothe TheseSEDs(describingtheemissionofatype1AGN)must correspondingonesofequation(1)((cid:1)logM (cid:2)0:2). BH becomparedwiththeintrinsicnuclearemission,whichincase Inthefollowingweconservativelyadopttherelationproposed of absorbed sources could be very different from the observed byMcLure&Dunlop(2002),whichyieldsthelowestmassesti- one.TheproblemistackledstartingfromtheX-rayspectralanal- mates.TheBHmassesquotedinthefollowingcouldbeafactorof 1.5Y2higher(and,correspondingly,theEddingtonratioslowerby ysisandassumingthepictureproposedbytheunifiedmodel: the same factor), if the relationships proposed by Kormendy & 1. X-rayemission: Gebhardt(2001)orbyMarconi&Hunt(2003)wereused. (a) We de-absorb the X-ray flux, adopting for the CDF-S Inordertoestimatetherest-frameR-bandmagnitudeatz(cid:4)0 sourcestheintrinsicN quotedbyTozzietal.(2006);sincethe forthebulgesinoursample,westartedfromtheSSPSEDsnor- H analysisofCDF-NX-rayemission(Alexanderetal.2003)does malizedasdescribedinx3(K-correction)andcomputedtheirpas- notinclude an intrinsic absorption,we derive it from thehard- siveevolutionuptothepresentepoch.11TheR-bandmagnitudes nessratiosandredshifts,assumingameanphotonindex(cid:2)¼1:8. obtainedbyusingthetwometallicitiesintroducedinx3arevery AsshowninTable5,onlyoneobjectofthesampleexhibitsahy- similar,within0.1magforallthesources.ThemeanoftheR-band drogencolumndensityN ’1:5;1024 cm(cid:1)2(Compton-thick magnitudes,aswellastheBHmassesobtainedfromequation(1), H candidate). arereportedinTable5.Forsourcesdominatedbythenuclear (b) WerecovertheintrinsicX-rayluminositynormalizinga componentorbythehostgalaxy,thesamefithasbeenalsocar- power-lawspectrumhavingthesameX-rayphotonindexofthe riedoutassumingforthebulgethemagnitudesrecoveredasde- source(fortheCDF-S)orwith(cid:2)fixedto1.8(fortheCDF-N)so scribedinx3.1.Thecorrespondingupperorlowerlimittothe thatitmatchestheintrinsicX-rayflux. massofthecentralBHisreportedasasecondlineinTable5. ErrorsfortheR-bandmagnitudeofthebulgecomponentare 2. Opticalbands: evaluatedbyquadraticallysummingtwodifferentcontributions: (a) WecomparetheQSOtemplatespectrumwiththeoptical 1. The uncertainties associated with the decomposed mag- nuclearmagnitudes,computing nitudespropagateontothebest-fitnormalizationoftheelliptical steomurpclea,tewaedroeppteeadtitnhecofimtpchuatinngginthgetRhe-bnaonrdmmalaigzantiitoundeosf:tfhoereteamch- (cid:3)2 ¼NXbands(cid:3)Fobs;j(cid:1)b;Ftempl;j(cid:4)2; (cid:4) plate until (cid:1)(cid:3)2 ¼1, and we compute the R-band magnitudes j¼1 j corresponding to the minimum and maximum values of the normalization. wherebisanormalizationconstant,whileF ,F ,and(cid:4) obs;j templ;j j 2. The difference between the R-band magnitudes found by aretheobservedandtemplatefluxes,andtheuncertaintyofthe usingfortheSSPstwodifferentmetallicities(seex3)isassumed former,intheopticalbandj,respectively. asanestimateoftheerrorinducedbyselectingthatparticulartem- (b) NormalizingtheQSOtemplatewiththevalueofbcorre- plates to represent a local inactive galaxy (i.e., by choosing the spondingtotheminimumvaluefor(cid:3)2,weobtaintheopticalcon- twoabove-mentionedpairsage/metallicity). tributiontothebolometricluminosity. (c) IfN >1022cm(cid:1)2(thisisthecaseofabouttwo-thirdsof WhenpropagatedtoyieldtheuncertaintiesoftheBHmasses, H theanalyzedsources),thepresenceofdustinthecentralregions thecontributionduetotheerrorinM isalwaysnegligiblewith R couldaffectseriouslyourestimateofnuclearmagnitudes.Inthis respecttothescatterintherelationdescribedbyequation(1). case,weprefertoignorethefluxesinBandVband(wherethe Themassesaredistributedovertwodecades,2;106(cid:2)M (cid:2) BH effectsoftheabsorptionaremoresevere),ratherthantoassumea 2:5;108 M ; only 3 out of 34 objects have M (cid:2)106 M . (cid:3) BH (cid:3) quitearbitrarycorrectionfactor.So,onlytheiandzmagnitudes areconsideredinthefit. 11 Asacheck,weevaluatedalsotheBHmassesobtainedfromtherelationfor 3. Bolometric properties: Connecting the power-law spec- AGNsathigherredshift(seeTable3inMcLure&Dunlop2002)usingtherest- trumandtheQSOtemplatenormalizedasdescribedbefore,we frameR-bandmagnitudeattheredshiftofeachsource(i.e.,withoutevolution); recovertheX-rayYtoYopticalratio,thebolometricluminosityof wefoundasystematicshiftoftheorderof+0.18dexwithrespecttothevalues reportedinTable5. theAGNandthehardX-raybolometriccorrection.
Description: