Morphology and structure of BzK-selected galaxies at z~2 in the CANDELS-COSMOS field

1.1 MB·English

by  Guan-Wen Fang

ResearchinAstronomyandAstrophysicsmanuscriptno.
(LATEX: ms1971.tex; printedonJanuary7,2015; 1:22)
5 Morphology and structure of BzK-selected galaxies at z ∼ 2 in the
1
0
∗
2 CANDELS-COSMOS field

n
a
J Guan-WenFang1,2,Zhong-YangMa2,3,YangChen2,3,4,XuKong2,3

6

]
A 1 InstituteforAstronomyandHistoryofScienceandTechnology,DaliUniversity,Dali671003,China;
G
[email protected]
.
h 2 KeyLaboratoryforResearchinGalaxiesandCosmology,TheUniversityofScienceandTechnologyof
p
- China,ChineseAcademyofSciences,Hefei,Anhui,230026,China
o
r 3 CenterforAstrophysics,UniversityofScienceandTechnologyofChina,Hefei230026,China;
t
s
a [email protected]
[
4 SISSA,viaBonomea265,I-34136Trieste,Italy

1
v
7 Received2014??;accepted2014??
0
1
1
0
Abstract Utilizing a BzK-selected technique, we obtain 14550 star-forming galaxies
.
1
(sBzKs) and 1763 passive galaxies (pBzKs) at z ∼ 2 from the K-selected (K < 22.5)
0 AB
5
catalog in the COSMOS/UltraVISTA field. The differential number counts of sBzKs and
1
: pBzKsareconsistentwiththeresultsfromtheliterature.Comparedtotheobservedresults,
v
i
X semi-analyticmodelsofgalaxyformationandevolutionprovidetoofew(many)galaxiesat
r high(low)-massend.Moreover,we find thatthe star formationrate (SFR) andstellar mass
a
of sBzKs follow the relation of main sequence. Based on the HST/Wide Field Camera 3
(WFC3)F160Wimaging,wefindawiderangeofmorphologicaldiversitiesforsBzKs,from
diffusetoearly-typespiralstructures,withrelativelyhighM ,largesizeandlowG,while
20
pBzKsareelliptical-likecompactmorphologieswithlowerM ,smallersizeandhigherG,
20
indicatingthemoreconcentratedandsymmetricspatialextentofstellarpopulationdistribu-
tioninpBzKsthansBzKs.Furthermore,thesizesofpBzKs(sBzKs)atz ∼2areonaverage
twotothree(onetotwo)timessmallerthanthoseoflocalearly-type(late-type)galaxieswith
similarstellarmass.Ourfindingsimplythatthetwoclasseshavedifferentevolutionmodes
andmassassemblyhistories.
Keywords: galaxies:evolution—galaxies:fundamentalparameters—galaxies:structure
—galaxies:high-redshift
SupportedbytheNationalNaturalScienceFoundationofChina.
2 G.-W.Fangetal.
1 INTRODUCTION
Theformationandevolutionofmassivegalaxies(M∗ >1010M⊙)atz ∼2isahotissueofobservational
astronomy.Therearemanyreasons,forinstancethepopulationofHubblesequencegalaxiesisalreadyin
place at z ∼ 1.5–2 (Fang et al. 2012), the universe star formation rate density (SFRD) peaks at z ∼ 2
(Oeschetal.2012),thespecificSFR(sSFR)evolvesweaklyatz > 2(Gonza´lezetal.2014),thegalaxy’s
massgrowsquicklyat1 <z < 3(Ilbertetal.2013),luminousinfraredgalaxies(L8−1000µm > 1011L⊙)
aremorecommonatredshiftz ∼1–3(Murphyetal.2013),andthenumberdensityofquasi-stellarobjects
(QSOs)hasapeakatz ∼2(Richardsetal.2006).
Withinthepastdecade,manynoveltechniqueshavebeenappliedtoselectasampleofmassivegalaxies
at the epoch of z ∼ 2 and important investigation has been made in our understanding of high-redshift
galaxies(Chapmanetal.2003;Franxetal.2003;Daddietal.2004;Kongetal.2006;Deyetal.2008;Huang
etal.2009;Fangetal. 2012;Wang etal. 2012;Fangetal.2014).Suchassubmillimetergalaxies(SMGs
withF(850µm)>0.5mJy),distantredgalaxies(DRGswith(J−K) >2.3),ultraluminousinfrared
Vega
galaxies(ULIRGs;L8−1000µm > 1012L⊙),dusty-obscuredgalaxies(DOGswith(R−[24])Vega > 24),
etc.Basedonasimpletwo-color(B−z andz−K)approach,Daddietal.(2004)introducedthecriteria
of (z − K) = 2.5 and BzK = (z − K) −(B − z) = −0.2 to select the sample of z ∼ 2
AB AB AB
massivegalaxies.ObjectswithBzK >−0.2areclassifiedasstar-forminggalaxies(sBzKs).Sourceswith
BzK <−0.2and(z−K) >2.5aredefinedaspassivegalaxies(pBzKs).
AB
Following the BzK technique in Daddi et al. (2004), many groups selected large samples of BzKs
(includesBzKsandpBzKs)fromdifferentsurveys(Kongetal.2006;Laneetal.2007;Blancetal.2008;
Hartleyetal.2008;Hayashietal.2009;Cassata etal.2010;McCrackenetal.2010;Onoderaetal.2010;
Fang et al. 2012; Ryan et al. 2012). Furthermore, they also investigated the physical properties of these
galaxies,e.g.,surfacedensity,stellarmass,SFR,near-infrared(NIR)spectroscopy,morphology,clustering,
andsize.SofarthelargestsampleofBzKsisfromMcCrackenetal.(2010),theyfoundthattheclustering
ofBzKsismuchstrongerthanthatoffullK-limited(K <22)samplesoffieldgalaxies.Moreover,the
AB
comovingcorrelationlengthofpBzKs(r ∼7h−1Mpc)islargerthanthatofsBzKs(r ∼5h−1Mpc).
0 0
Forgalaxiesatz ∼2,HubbleSpaceTelescope(HST)/WideFieldCamera3(WFC3)NIRimagingcan
providecrucialcluestotherest-frameopticalmorphologies.AtsuchredshiftHST/WFC3NIRbandsmove
beyond the Balmer break (λ > 4000 A˚) to the redder wavelengthsand thus probe the light from the
rest
dominantstellarpopulationofgalaxy.Thiswillenableustostudytherest-frameopticalmorphologiesand
structuresofBzKsatz ∼ 2.By usingHST/WFC3 NIRimagesintheHubbleUltraDeepField(HUDF),
Cassata etal. (2010)reportedthestructuralfeaturesof 6pBzKs, thesegalaxiesin appearancehavea rel-
ativelyregularmorphologyandsmallersizethanlocalellipticalsofanalogousstellarmasses.Thesimilar
resultswerealsofoundinRyanetal.(2012),theirpBzHssampleincludes30passivegalaxiesatz ∼2(use
the H-band filter replaces the K-band filter). Fang et al. (2012) measured nonparametric morphological
parameters of BzKs (50 pBzKs and 173 sBzKs) at z ∼ 2 in the Extended Groth Strip (EGS) field, and
foundthatBzKshavebothearlyandlatetypes.
This paper will utilize HST/WFC3 F160W images to investigate the structural properties of BzKs.
Compared with previous works, we present the larger sample with high resolution NIR imaging,
MorphologyandstructureofBzK-selectedgalaxies 3
andforthefirsttimewestudythesizeevolutionofsBzKs.We introducethemulti-bandobservationsand
data reduction of the COSMOS field in Section 2. Section 3 describes the selection, redshift distribution
and numbercountsof BzKs. We show the SFR−M correlationof BzKs in Section 4. We presentthe
∗
structuralandmorphologicalpropertiesofBzKsinSection5and6,andsummarizeourresultsinSection7.
Throughoutthispaper,weadoptastandardcosmologyH =70kms−1Mpc−1,Ω =0.7,andΩ =0.3.
0 Λ M
AllmagnitudesusetheABsystemunlessotherwisenoted.
2 OBSERVATIONSANDDATA
TheCosmicEvolutionSurvey(COSMOS)istargetedonaspecialareaofthesky,thathasbeenobserved
withsomeoftheworld’smostpowerfultelescopesonthegroundandinspace,inwavelengthrangefromX-
raysthroughultravioletandvisiblelight,downtoinfraredandradiowaves(Scovilleetal.2007).Inaddition,
italsoincludesoptical/infraredspectroscopysurveyusingtheKeckDEIMOSandLRIS,MagellanIMACS,
andVLTVIMOSspectrographs.MoredetailsoftheobservationanddatareductionintheCOSMOSfield
can be found in McCracken et al. (2012) and Muzzin et al. (2013). The multi-band photometry data we
useinourworkisfromtheK-limited(K < 23.4)catalogoftheCOSMOS/UltraVISTAfieldprovidedby
Muzzin et al. (2013), which is produced based on the NIR data from the UltraVISTA DR1 (McCracken
etal.2012).Moreover,thederivedphysicalparametersweuseinourstudyalsocomefromthecatalogof
Muzzinetal.(2013),suchasstellarmass(M∗),SFR(SFRUV,uncorr+SFRIR),andphotometricredshift
(z ,ifthereisnospectroscopicredshiftavailable).
p
In this paper, we will utilize the latest released data of HST/WFC3 F160W high-resolution images
in the CANDELS1-COSMOS field to investigatethe morphologicalfeaturesof BzKs in oursample. The
CANDELS/wideCOSMOSsurveycoversatotalof∼210arcmin2 atJ−andH−band.The5σ limiting
magnitudeis26.9intheF160Wfilter.HST/WFC3F160Wimagesweredrizzledto0′′.06pixel−1.Further
detailsareinGroginetal.(2011)forthesurveyandobservationaldesign,andKoekemoeretal.(2011)for
thedataproducts.
3 SELECTION,REDSHIFTDISTRIBUTIONANDNUMBERCOUNTSOFBZKS
To construct a sample of galaxies at z ∼ 2, we use the BzK color criteria from Daddi et al. (2004).
The optical B- and z-band data is taken with Subaru/SuprimeCam (B , z+), while the K data is from
j
theVISTA/VIRCAM.Inaddition,wealso applythecolorcorrectionusedbyMcCrackenetal. (2010)to
B −z+ tokeepunanimouswiththeBzK selectiontechnique.AsshowninFigure1,bluedotsrepresent
j
14550sBzKswithBzK > −0.2(solidline)andK < 22.5,andreddotscorrespondto1763pBzKswith
BzK <−0.2andz−K >2.5(dot-dashedline).Objectswithz−K <0.3(B−z)−0.5(dashedline),
areclassifiedasstars.
Figure2showstheredshiftdistributionforBzKsintheCOSMOSfield.Forasampleofgalaxieswith
K < 22.5, in Figure 2(a), we find that the BzK color technique successfully selects more than 80% of
galaxiesatredshift1.5 < z < 2.7(thisfractiongoestomorethan90%at1.6 < z < 2.6),indicatingthe
BzK criteria isquite effectivein selectinggalaxiesatz ∼ 2. Forthe massivegalaxies(M∗ > 1010M⊙)
1
4 G.-W.Fangetal.
Fig.1 BzKtwo-colordiagramforallobjectsintheCOSMOSfield.GalaxieswithBzK >−0.2
(solid line) and K < 22.5 are selected as sBzKs. Sources with BzK < −0.2 and z −K >
2.5(dot-dashedline)aredefinedaspBzKs. Thecolorcriterionofstar andgalaxyseparationis
z−K =0.3(B−z)−0.5(dashedline).
at 1.6 < z < 2.6, the percentage of objects selected as BzKs is > 90% (the dotted red lines in Figure
2(a)).ThemeanredshiftsofsBzKs(Figure2(b))andpBzKs(Figure2(c))are1.75±0.48and1.69±0.33,
respectively. In the meantime, we use the extrapolation of the red-blue separation method of Bell et al.
(2004)toseparatetheredsequenceandbluecloudintherest-frameU −V vs.M diagram.About91%
V
ofpBzKscanberoughlydividedintoredsequence,andmostofsBzKs(84%)arelocatedinthebluecloud
region.
We calculateddifferentialK-bandnumbercountsforallsBzKs andpBzKsinoursample(see Figure
3).For comparisonwith the results frompreviousworks,we also plottedtheir data in Figure3 (Deep3a-
F and Daddi-Fof Kong et al. 2006;Lane et al. 2007;Blanc et al. 2008;Hartley et al. 2008;McCracken
et al. 2010;Fang et al. 2012).The dot-dashedlinesrepresentthe countsof quiescentgalaxies(QGs) and
star-forming galaxies (SFGs) from the semi-analytic model (Kitzbichler & White 2007). In general, our
countsagreewiththeresultsfromtheliterature.Owingtotheexistenceofphotometricoffsetsandcosmic
variance,thereisadiscrepancyamongdifferentworksforsBzKsandpBzKscounts.Combiningwiththe
datafromtheliterature,weconfirmthatthenumbercountsofpBzKshasabreakatK ∼ 21.0,apossible
explanationforthisliesinthesmallredshiftrangeofpBzKs(Kongetal.2006).Theredshiftdistributionof
ourpBzKssamplealsosupportsthefinding,comparedtosBzKs.Comparedwiththeobservedresults,the
semi-analyticmodelpredictstoofew(many)galaxiesathigh(low)-massend.
MorphologyandstructureofBzK-selectedgalaxies 5
Fig.2 (a)FractionofBzKs(includesBzKsandpBzKs)inthetotalgalaxysampleofK <22.5.
Thedottedredlinesshowthefractionofmassivegalaxies(M∗ >1010M⊙)selectedasBzKs.(b)
RedshiftdistributionforsBzKsinoursample.(c)RedshiftdistributionforpBzKsinoursample.
5 3.5
(a) sBzKs (b) pBzKs
ag) 4 Kitzbichler & White (2007) ag) 3.0
m m 2.5
5 5
0. 3 0. 2.0
2g/ 2g/
e e 1.5
d 2 d
N (/ N (/ 1.0 This paper
og 1 HThairstl epya peet ral.(2008) og 0.5 DFaandgd ie-Ft al.(2012)
l Lane et al.(2007) l Deep3a-F
0 Blanc et al.(2008) 0.0 McCracken et al.(2010)
19 20 21 22 23 24 19 20 21 22 23 24
K (mag) K (mag)
AB AB
Fig.3 DifferentialnumbercountsofsBzKsandpBzKsintheCOSMOS/UltraVISTAfield.The
resultsfromtheliteratureandthemodelarealsoshowninthisdiagram.Deep3a-FandDaddi-F
fromKongetal.(2006).
4 THESTELLARMASS−SFRCORRELATIONOFBZKS
For local star-forming galaxies, Brinchmann et al. (2004) found that there is a tight correlation between
M∗ and SFR (SFR∝M∗α), and called it main sequence (MS). At redshift 0.5 < z < 3, the MS is also
confirmed (Daddi et al. 2007; Elbaz et al. 2007; Rodighiero et al. 2011; Fang et al. 2012), but the slope
(α)rangesfrom0.6to1.0(relyingondifferentsampleandtheapproachesforcalculatingM andSFR).In
∗
Figure4,weshowtherelationofM andSFRofBzKsintheCOSMOSfield.ForsBzKs,abest-fitslope
∗
α = 0.67±0.06(blueline)isfound,inagreementwiththoseprovidedbyDaddietal.(2007)(α ∼ 0.9,
grayline)andRodighieroetal.(2011)(α∼0.79,cyanline).Thediscrepancyfordifferentslopesisdueto
thedifferentmethodsinderivingSFR(Rodighieroetal.2014).Ontheotherhand,we findthattheSFRs
andstellarmassesofpBzKsalsoshowacorrelation,butwithlowerSFRscomparedtosBzKsforagiven
stellarmass.GraysquaresinFigure4representsBzKsinPannellaetal.(2009).FromKarimetal.(2011),
SFGs(graytriangles)withdifferentmassandredshiftbinsarealsoplottedinthisfigure.
6 G.-W.Fangetal.
Fig.4 Relationshipofstellarmassvs. SFRforBzKsin theCOSMOSfield.Solidsquaresand
trianglesrepresentthestar-forminggalaxiesfromPannellaetal.(2009)andKarimetal.(2011),
respectively.GrayandcyanlinescorrespondtotheMSfromDaddietal.(2007)andRodighiero
etal.(2011),respectively.
5 STRUCTURESOFBZKS
In order to analyze the structural properties of BzKs in the CANDELS-COSMOS field, we employ the
latest catalog2 (version 1.0) providedby van der Wel et al. (2012). Galaxy sizes (r ) are measured from
e
theHST/WFC3F160Wimaging.Withinamatchedradiusof0′′.5,weobtainthestructuralparametersof
52pBzKsand378sBzKsfromvanderWeletal.(2012).TheM −r relationsareshowninFigure5of
∗ e
pBzKsandsBzKs,respectively.Shenetal.(2003)providedtherelationswith1σdispersionforlocallate-
andearly-typegalaxies(LTGsandETGs)asplottedinthisfigure.Andforcomparison,otherdatafromthe
literaturearealsoshowninFigure5(b)(Cassataetal.2010;Gobatetal.2012;Ryanetal.2012;Szomoruet
al.2012).FromFigure5,wefindthatthesizesofpBzKsandmassivesBzKsatz ∼2aresmallerthantheir
localcounterpartsatafixedstellarmass.Moreover,wealsoseeadiversityofstructuralpropertiesamong
BzKs,somesourcesaresimilartolocalgalaxies,butthereisalsotheexistenceofmassivecompactBzKs,
comparedto present-day counterparts.Generally, pBzKs have elliptical-like compactstructureswith low
r ,whilesBzKsarerelativelyextendandirregular,withhigherr .
e e
TofurtherinvestigatethesizeevolutionwithredshiftforourBzKssampleatz ∼2.Weshowthesizes
for pBzKs (1.85±1.09 kpc) and sBzKs (2.63±1.36 kpc) in Figure 6, respectively. The effective radii
of QGs and SFGs from the literature are also plotted in this figure (Shen et al. 2003;Cassata et al.2010;
Gobatetal.2012;Ryanetal.2012;Szomoruetal.2012;Fanetal.2013;Pateletal.2013;Morishitaetal.
2014).AsshowninFigure6,wefindthatthesizesofpBzKsareafactorof∼ 2−3smallerthanthoseof
localcounterparts.ForsBzKs, themeansize isonetotwotimessmallerthanthoseoftypicallocalLTGs
withcomparablemass.CombinedwiththedatapointsofPateletal.(2013)andMorishitaetal.(2014),the
differenceofsizes forpBzKsandsBzKs indicatesthatthetwo classes havedifferentevolutionprocesses
andassemblyhistories,suchasminormergerswithalowincreaseingalaxy’smassandsecularevolution
2
MorphologyandstructureofBzK-selectedgalaxies 7
Fig.5 Distribution of M vs. r for BzKs ((a) sBzKs and (b) pBzKs) in the CANDELS-
∗ e
COSMOSfield.TheresultsfromtheliteraturearealsoshowninFigure5(b)(Cassataetal.2010;
Gobatetal.2012;Ryanetal.2012;Szomoruetal.2012).
Fig.6 Evolutionofthe effectiveradius(r )with redshiftforpBzKsandsBzKsin oursample.
e
TheeffectiveradiiofQGsandSFGsfromtheliteraturearealsoplottedinthisfigure.Themean
sizeofpBzKsis1.85±1.09kpc(redsolidcircle),whilesBzKsis2.63±1.36kpc(bluesolid
circle). Left: Green and cyan lines correspond to r ∝ (1 + z)−1.16 (Patel et al. 2013) and
e
r ∝ (1+z)−1.06 (Morishitaetal.2014),respectively.Right:Greenandcyanlinesrepresent
e
r ∝(1+z)−0.63(Pateletal.2013)andr ∝(1+z)−0.56(Morishitaetal.2014),respectively.
e e
withoutmergers(ormonolithiccollapsemode).ForthesizeevolutionofpBzKs,ourobservationssupport
thatthepredictionsfromminormergers.AsregardsthesizegrowthofsBzKs,apossibleexplanationlies
ininternalevolutionthroughsomequenchedmechanisms(AGNfeedbackorstellarwinds).
6 MORPHOLOGIESOFBZKS
Using the WFC3 on board the HST, CANDELS provided high resolution F160W imaging data
( ′′ −1).Inourwork,weutilizeHST/WFC3F160Wimagestostudythemorphologicaldiversities
8 G.-W.Fangetal.
0.8
0.8
pBzKs
sBzKs
0.7
0.6
0.7
0.5
0.4
0.3
0.6
-0.5 -1.0 -1.5 -2.0 -2.5
ni
Gi
0.5
0.4
0.3 HST/WFC3 F160W
All images are 3" square
-0.5 -1.0 -1.5 -2.0 -2.5
M
20
Fig.7 Distribution of BzKs in the M vs. Gini coefficient plane. The red and blue circles
20
representpBzKsandsBzKs,respectively.
of z ∼ 2 BzKs. To clearly analyze their structural features, we calculated nonparametricmorphological
parametersofgalaxy,suchasGinicoefficient(G;therelativedistributionofthegalaxypixelfluxvalues)
andM (thesecond-ordermomentofthebrightest20%ofthegalaxy’sflux)(Abrahametal.1996;Lotz
20
etal.2004).
InFigure7,theredandbluecirclesrepresentpBzKsandsBzKs,respectively.Meanwhile,basedonthe
valuesofGandM of52pBzKsand378sBzKs,theirstampimages(3′′×3′′)arealsoplottedinthisfig-
20
ure.AsshowninFigure7,wefindthatpBzKsinappearancehaveregularandcompact(likespheroid),and
thatshowlowM andhighGinrest-frameopticalmorphology.ForsBzKs,thereisawiderangeofmor-
20
phologicaldiversities, includingclumpy,irregular,extended,and early-typespiral-like morphologies,but
mostofthemshowdiffusestructures,withhighM andlowG.Thatindicatesthemoreconcentratedand
20
symmetricspatialextentofstellarpopulationdistributioninpBzKsthansBzKs. Furthermore,we derived
the meanvaluesofG and M forpBzKs andsBzKs, correspondingto (0.63,−1.70)and (0.51,−1.49),
20
respectively.Ourfindingsfurtherimplythatpassivegalaxiesandstar-forminggalaxieshavedifferentevo-
lutionmodesandmassassemblyhistories.
7 SUMMARY
Based on a BzK-selected technique, we present 14550 star-forming galaxies (sBzKs) and 1763 passive
galaxies (pBzKs) at ∼ from the -selected ( ) catalog of the COSMOS/UltraVISTA
MorphologyandstructureofBzK-selectedgalaxies 9
field. Utilizing data from HST WFC3/F160W imaging in the CANDELS-EGS field, we investigate the
morphologicalandstructuraldiversitiesofthesegalaxies.Ourfindingsareasfollows:
(1)WefindthattheBzKcolortechniquesuccessfullyselects>80%galaxiesatredshift1.4<z <2.7
(thisfraction> 90%at 1.6 < z < 2.6),indicatingthe BzK criteria is a quite effectivegalaxyselection
methodatz ∼2.Moreover,forthemassivegalaxies(M∗ >1010M⊙)at1.6<z <2.6,thepercentageof
sourcesselectedasBzKsis>90%.
(2) The differential number counts of sBzKs and pBzKs agree with the results from the literature.
Compared with the observedresults, models for galaxy formationand evolution providetoo few (many)
galaxiesathigh(low)-massend.
(3) We find that the star formation rate (SFR) and stellar mass of sBzKs follow the relation of main
sequence(SFR∝M0.67±0.06).pBzksalsoshowscorrelationbetweenSFRsandstellarmassesbutwithtoo
∗
lowSFRsforpBzKs.Moreover,about91%ofpBzkscanberoughlydividedintoredsequence.
(4)WefindthatthesizesofpBzKs(1.85±1.09kpc)andsBzKs(2.63±1.36kpc)atz ∼2aresmaller
thantheirlocalcounterpartsatafixedstellarmass.Moreover,wealsoseeadiversityofstructuralproperties
amongBzKs,somesourcesaresimilartolocalgalaxies,butthereisalsotheexistenceofmassivecompact
BzKs,comparedtopresent-daycounterparts.ThesizesofsBzKsarelargerthanpBzKsingeneral,evenin
high-masssystems,butsomehaveverycompactstructures,withr <1kpc.
e
(5)WefindawiderangeofmorphologicaldiversitiesforsBzKs,fromextendedordiffusetoearly-type
spiral-likestructures,whilepBzKsarerelativelyregularandcompact(likespheroid).Moreover,wecalcu-
latethemeanvaluesofGandM forpBzKsandsBzKs,correspondingto(0.63,−1.70)and(0.51,−1.49),
20
respectively.ThesBzKsshowhighM andlowG,whichindicateslessconcentratedandsymmetricspa-
20
tialdistributionofthestellarmassofsBzKsatz ∼2,comparingtopBzKs.Ourfindingsimplythatthetwo
classeshavedifferentevolutionmodesandmassassemblyhistories.
Acknowledgements This work is based on observations taken by the CANDELS Multi-Cycle Treasury
Program with the NASA/ESA HST, which is operated by the Association of Universities for Research
in Astronomy, Inc., under the NASA contract NAS5-26555. This work is supported by the National
Natural Science Foundation of China (NSFC, Nos. 11303002, 11225315, 1320101002, 11433005, and
11421303), the Specialized Research Fund for the Doctoral Program of Higher Education (SRFDP, No.
20123402110037),theStrategicPriorityResearchProgram“TheEmergenceofCosmologicalStructures”
oftheChineseAcademyofSciences(No.XDB09000000),theChineseNational973FundamentalScience
Programs(973program)(2015CB857004),theYunnanAppliedBasicResearchProjects(2014FB155)and
theOpenResearchProgramofKeyLaboratoryforResearchinGalaxiesandCosmology,CAS.
References
Abraham,R.G.,Tanvir,N.R.,Santiago,B.X.,etal.1996,MNRAS,279,L47
Bell,E.F.,Wolf,C.,Meisenheimer,K.,etal.2004,ApJ,608,752
Blanc,G.A.,Lira,P.,Barrientos,L.,etal.2008,ApJ,681,1099
Cassata,P.,Giavalisco,M.,Guo,Y.,etal.2010,ApJ,714,L79
Chapman,S.C.,Blain,A.W.,Ivison,R.J.,etal.2003,Nature,422,695
10 G.-W.Fangetal.
Daddi,E.,Cimatti,A.,Renzini,A.,etal.2004,ApJ,617,746
Daddi,E.,Dickinson,M.,Morrison,G.,etal.2007,ApJ,670,156
Dey,A.,Soifer,B.T.,Desai,V.,etal.2008,ApJ,677,943
Daddi,E.,Cimatti,A.,Renzini,A.,etal.2004,ApJ,617,746
Dey,A.,Soifer,B.T.,Desai,V.,etal.2008,ApJ,677,943
Elbaz,D.,Daddi,E.,LeBorgne,D.,etal.2007,A&A,468,33
Fan,L.,Fang,G.,Chen,Y.,etal.2013,ApJ,771,L40
Fang,G.,Kong,X.,Chen,Y.,etal.2012,ApJ,751,109
Fang,G.,Huang,J.-S.,Willner,S.P.,etal.2014,ApJ,781,63
Franx,M.,Labbe´,I.,Rudnick,G.,etal.2003,ApJ,587,L79
Gobat,R.,Strazzullo,V.,Daddi,E.,etal.2012,ApJ,759,L44
Gonza´lez,V.,Bouwens,R.,Illingworth,G.,etal.2014,ApJ,781,34
Grogin,N.A.,Kocevski,D.D.,Faber,S.M.,etal.2011,ApJS,197,35
Hartley,W.G.,Lane,K.P.,Almaini,O.,etal.2008,MNRAS,391,1301
Hayashi,M.,Motohara,K.,Shimasaku,K.,etal.2009,ApJ,691,140
Huang,J.-S.,Faber,S.M.,Daddi,E.,etal.2009,ApJ,700,183
Ilbert,O.,McCracken,H.J.,LeFe`vre,O.,etal.2013,A&A,556,A55
Karim,A.,Schinnerer,E.,Mart´ınez-Sansigre,A.,etal.2011,ApJ,730,61
Kitzbichler,M.G.,&White,S.D.M.2007,MNRAS,376,2
Koekemoer,A.M.,Faber,S.M.,Ferguson,H.C.,etal.2011,ApJS,197,36
Kong,X.,Daddi,E.,Arimoto,N.,etal.2006,ApJ,638,72
Lane,K.P.,Almaini,O.,Foucaud,S.,etal.2007,MNRAS,379,L25
Lotz,J.M.,Primack,J.,&Madau,P.2004,AJ,128,163
McCracken,H.J.,Capak,P.,Salvato,M.,etal.2010,ApJ,708,202
McCracken,H.J.,Milvang-Jensen,B.,Dunlop,J.,etal.2012,A&A,544,A156
Morishita,T.,Ichikawa,T.,&Kajisawa,M.2014,ApJ,785,18
Murphy,E.J.,Stierwalt,S.,Armus,L.,etal.2013,ApJ,768,2
Muzzin,A.,Marchesini,D.,Stefanon,M.,etal.2013,ApJS,206,8
Oesch,P.A.,Bouwens,R.J.,Illingworth,G.D.,etal.2012,ApJ,759,135
Onodera,M.,Arimoto,N.,Daddi,E.,etal.2010,ApJ,715,385
Pannella,M.,Carilli,C.L.,Daddi,E.,etal.2009,ApJ,698,116
Patel,S.G.,vanDokkum,P.G.,Franx,M.,etal.2013,ApJ,766,15
Richards,G.T.,Strauss,M.A.,Fan,X.,etal.2006,AJ,131,2766
Rodighiero,G.,Daddi,E.,Baronchelli,I.,etal.2011,ApJ,739,40
Rodighiero,G.,Renzini,A.,Daddi,E.,etal.2014,MNRAS,443,19
Ryan,R.E.,Jr.,McCarthy,P.J.,Cohen,S.H.,etal.2012,ApJ,749,53
Scoville,N.,Aussel,H.,Brusa,M.,etal.2007,ApJS,172,1
Shen,S.,Mo,H.J.,White,S.D.M.,etal.2003,MNRAS,343,978
Szomoru,D.,Franx,M.,&vanDokkum,P.G.2012,ApJ,749,121