ebook img

Distant early-type galaxies: tracers of the galaxy mass assembly evolution PDF

0.28 MB·English
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 Distant early-type galaxies: tracers of the galaxy mass assembly evolution

Mem.S.A.It.Vol.75,282 (cid:13)c SAIt 2004 Memoriedella Distant early-type galaxies: tra ers of the galaxy mass assembly evolution A. Cimatti1 Istituto Nazionale di Astrofisica – Osservatorio Astrofisicodi Arcetri, Largo E. Fermi 5, 6 I-50125Firenze,Italye-mail:[email protected] 0 0 2 Abstract. Wereviewthemostrecentobservationalresultsontheformationandevolution n ofearly-typegalaxiesandtheirmassassemblybyfocusingon:theexistence,propertiesand a roleofdistantold,massive,passivesystemstoz ∼ 2,thestellarmassfunctionevolution, J the“downsizing”scenario,andthehigh-zprecursorsofmassiveearly-typegalaxies. 5 2 Keywords.Galaxies:evolution–Galaxies:formation 2 v 7 1. Introduction evolution of the stellar population. However, 9 in the modern scenario of CDM cosmology, 4 Early-typegalaxies(ETGs)playacrucialrole theevolutionofthegravitationalperturbations 1 in cosmology. They represent the most mas- implies that galaxies assembled their mass 0 sive galaxies in the local Universe, contain more gradually through hierarchical merging 6 most of the stellar mass (M) and are the pri- of CDM halos and make the monolithic col- 0 mary probes to investigate the cosmic history lapseunlikelyorevenunphysical.Forinstance, h/ of galaxy mass assembly. As ETGs are the the most recent ΛCDM hierarchical models p most clustered galaxies, they are also funda- predictthatmostoftheETGstellarmassisas- - mental in tracing the evolution of the large o sembled at 0 < z < 1 (DeLuciaetal.2006). scalestructures.Becauseofthecorrelationbe- r The moderngenerationof galaxysurveyscan st tweentheblackholeandgalaxybulgemasses, testthedifferentpredictionsontheevolutionof ETGsarealsoimportanttoinvestigatetheco– a the number density, colors, merger types and : evolutionofnormalandactivegalaxies. v rates,luminosityandmassfunctions,andpro- Although ETGs at z ∼ 0 are rather “sim- i videacrucialfeedbacktoimprovethetheoret- X ple” and homogeneous systems in terms of icalsimulations. morphology,colors,stellar populationcontent r Despite the intense observational activity a andscalingrelations(Renzini2006),theirfor- onETGevolution,ageneralconsensushasnot mation and evolution is still a debated ques- been reached yet. For instance, recent results tion.Inthesocalled”monolithic”collapsesce- based on optically-selected, moderately deep nario(Eggenetal.1962;Tinsley1972),ETGs (R < 24) samples of “red sequence” ETGs converted most of the gas mass into stars at (e.g. COMBO-17, DEEP2) (Belletal.2004; high redshifts (e.g. z > 3) through an intense Faberetal.2005) suggest a strong evolution burst of star formation followed by passive of the number density with φ∗(z) displaying Sendoffprintrequeststo:A.Cimatti an increase by a factor of ∼6 from z ∼ 1 Cimatti:Early-typegalaxyevolution 283 to z ∼ 0, and a corresponding increase by a spectral energy distribution (SED) selections. factor of 2 of their stellar mass density since ETGs have strong k–corrections in the opti- z ∼ 1. Dissipationless ETG–ETG merging cal as they become rapidly very faint for in- (alsocalled”dry”merging)isadvocatedtoex- creasing redshiftsand can be easily missed in plain how ETGs assembled at 0 < z < 1 optically–selected samples (e.g. Maoz 1997), (Belletal.2005;vanDokkum2005).HSTop- while the k–corrections are much smaller in tical, moderately deep (I < 24) samples of thenear-infrared.Apossibleexampleisgiven morphologically–selected ETGs also suggest by recent results (Yamadaetal.2005) based a substantial decrease of the number density on a large sample of ETGs at z ∼ 1 ex- withn ∝ (1+z)−2.5 (Ferrerasetal.2005),and tracted from a ∼ 1 square degree field in the spatially resolved color maps show that ∼30- Subaru/XMM-NewtonDeep Survey . Despite 40% of ETGs at 0 < z < 1.2 display varia- the field as large as that coveredin COMBO- tionsintheirinternalcolorproperties(e.g.blue 17,thenumberdensityofETGsatz ∼ 1turns coresandinversecolorgradients)suggestiveof outtobesubstantiallylargerthanthatderived recent star formation activity in a fraction of atz=0.95inCOMBO-17,probablyduetothe the ETG population (Menanteauetal.2004). deeperandredder–bandselection(z′ < 25vs. However, the optically-selected (I < 24) R<24). VIMOS VLT Deep Survey (VVDS) does not Also the “progenitor bias” confirm the above findings, and shows that (vanDokkum&Franx2001) plays a role the rest-frame B-band luminosity function of in the comparison between low and high ETGs (selected based on the SED) is con- redshiftETGsamples.Asthethemorphology sistent with passive evolution up to z ∼ ofgalaxiesareexpectedtoevolve,somelow-z 1.1, while the number of bright ETGs seems ETGswerespiralgalaxiesormergersathigher to decrease only by ∼40% from z ∼ 0.3 redshift. These young ETGs are included to z ∼ 1.1 (Zuccaetal.2005). The over- in low redshift samples, but drop out of the all picture is made more controversial by re- samples at high redshift. Therefore, the high cent results indicating a weak evolution of redshiftsamplesareabiasedsubsetofthelow the stellar mass function (Fontanaetal.2004; redshift samples, containing only the oldest Caputietal.2006;Bundyetal.2006)andthat progenitors of low-z ETGs, i.e. the ones with “drymergers”cannotrepresentthemajorcon- morphological and spectral properties similar tributorstotheassemblyofETGsat0<z<1 tothenearbyETGs. (Bundyetal.2006;Masjedietal.2005). 3. HighredshiftETGs 2. Theproblemofselectioneffects What is the highest redshift at which we still After the discovery of the strong clustering findETGssimilartothoseobservedatz ∼ 0? of ETGs at z ∼ 1 (Daddietal.2000), it be- After the isolated case of a spectroscopically came clear that a major source of uncertainty identified old ETG counterpart of a radio was the “cosmic variance” (Daddietal.2000; source at z = 1.55 (Dunlopetal.1996), Somervilleetal.2004), and that its effects studies based on the HDF-South and HDF- couldexplainthediscrepantresultsoftenfound North suggestedthe existenceof a substantial by narrow-field surveys, such as the dif- number of photometric/morphological ETG ferent content of luminous high-z ETGs in candidates up to z ∼ 2 (Treuetal.1998; the HDF-North and HDF-South (Zepf1997; Benitezetal.1999; Stanfordetal.2004). It Benitezetal.1999). was with the recent spectroscopic surveys of Anotherimportantsourceofuncertaintyis near-infrared galaxy samples like the K20 played by the heterogeneous criteria adopted (Cimattietal.2002b; Mignolietal.2005), to select and classify ETGs: deep vs. shal- GDDS (Abrahametal.2004), TESIS low surveys, optical vs. near-infrared, color (Saraccoetal.2003), GRAPES vs. morphological vs. spectrophotometric or (Daddietal.2005a), LCIRS 284 Cimatti:Early-typegalaxyevolution (Dohertyetal.2005), that high-z ETGs wereunveileduptoz∼2. The distant ETGs spectroscopically iden- tified at 1 < z < 2 are very red (R− K > 5, s I −H > 3inVegaphotometricsystem),show the spectral features of passively evolving old stars with ages of 1-4 Gyr (Fig.1), and have large stellar masses with M > 1011M ⊙ (Cimattietal.2002a; Cimattietal.2003; Cimattietal.2004; McCarthyetal.2004; Saraccoetal.2003; Saraccoetal.2005; Daddietal.2005a; YanL.etal.2004; Fuetal.2005). Their spectral properties ETGs suggest a weak evolution in the range of 0 < z < 1 (Mignolietal.2005) (Fig.2). The deepest optical images of ETGs at 0.5 < z < 1.3 in the Hubble Ultra Deep Field coupled with the GRAPES ACS slitless Fig.1. The observed average spectrum (top) of spectra show that their stellar populations four massive ETGsat 1.6 < z < 2compared with are rather homogeneous in age and metal- asynthetic spectrumof asimplestellar population licity and formed at redshifts z ∼ 2 − 5, (bottom)(Cimattietal.2004). f and that the isophotal structure of z ∼ 1 ETGsobeysthe correlationsalreadyobserved among nearby elliptical galaxies, implying no significant structural differences between z ∼ 0andz ∼ 1(Pasqualietal.2006).Distant ETGs are also strongly clustered, with a comoving r ∼ 10 Mpc at z ∼ 1 measured 0 for the Extremely Red Object (ERO) popu- lation (Daddietal.2002; Firthetal.2002; Rocheetal.2003; Brownetal.2004; Georgakakisetal.2005), and similar to thatofpresent-dayluminousETGs. Thestellarmassesofthesegalaxiesarees- timatedbyfittingtheirmulti–bandphotometric SEDswithstellarpopulationsynthesismodels Fig.2.CompositespectraoftheK20surveyETGs (Fontanaetal.2004; Saraccoetal.2005; divided in three redshift bins (arbitrary flux scale) Longhettietal.2005; Bundyetal.2006). (Mignolietal.2005).Frombottomtotop:0 < z < 0.6, 0.6 < z < 0.75, 0.75 < z < 1.25. The upper Such “photometric” stellar masses are in rea- and lower panels also show the difference spectra sonableagreementwiththedynamicalmasses betweentheintermediate-redshiftcompositeandthe estimated from the absorption line velocity high-zandlow-zcompositespectrarespectively. dispersion (diSeregoAlighierietal.2005; Retturaetal.2006). ETGs dominatethe high- luminosity andhigh-masstails of the totallu- data at 1 < z < 2 are consistent with a minosityandstellarmassfunctionsuptoz∼1 range from 10% to 100% of the local den- (Pozzettietal.2003; Fontanaetal.2004; sity of luminous ETGs (Cimattietal.2004; Glazebrooketal.2004;Bundyetal.2006). Daddietal.2005a; Saraccoetal.2005). Old, However, due to the strong cosmic vari- passive,massiveETGsmayexistevenatz>2 ance, the number density of high-z ETGs is (Labbe’etal.2005), but their fluxes are typi- still so poorly constrained that the current callybeyondthecurrentspectroscopiclimits. Cimatti:Early-typegalaxyevolution 285 4. Downsizing Distant ETGs have a higher stellar mass- to-optical light ratio (M/L ) than late-type R star-forming galaxies at the same redshift (Fontanaetal.2004) (Fig.3). Moreover, the M/L showsanoveralltrenddecreasingfrom R low to high redshifts, but with a substantial scatterthatcanbeinterpretedasduetoawide rangeofformationredshiftsand/orstarforma- tion histories (see also McCarthy et al. 2001; Fig.3. M /L ratio as a function of redshift stars R Cimattietal.2003).ThetypicalM/LRofmore forthespectroscopicallyidentifiedETGsinthethe luminous objects (M < −22) is significantly K20 sample. Large hollow and small filled circles R largerthanthatofthefainterones(MR >−22) indicate MR < −22 and MR > −22 respectively. atlowandintermediateredshifts(Fig.3).This Lines show Mstars/LR values predicted for a set of impliesthat, while the M/L of the luminous exponentially–decaying star formation rate models R withz = 3(thinlines)orz =20(thicklines)and populationis consistentwith either very short f f time-scalesofτ=0.1Gyr(dashedlines)andτ=3 star-formation time-scales (τ) or high forma- Gyr(solidlines)(SalpeterIMFandnodustextinc- tion redshifts (z ≥ 3) (and some objects ap- f tion)(Fontanaetal.2004). pear to requireboth),the less luminouspopu- lationexperiencedamorerecenthistoryofas- sembly,asindicatedbythelargerτandlower low-mass ETGs respectively. Although, the z required to reproduce the typical M/L f R relation betweenthe measured M/L evolution (Fig.3).Thus,themostluminousandmassive and mass is partially due to selection effects ETGs reach the completion of star formation because the galaxies are selected by luminos- first,whilelessmassiveoneshaveamorepro- ity, not mass, the M/L – mass relation still longed star formation activity till later times. holdsevenif thisselectioneffectistakeninto This downsizing scenario was first noted by account.Theseresultsindicatethatthefraction (Cowieetal.1996),anditisnowsupportedby of stellar mass formed at recent times ranges severalresults which suggestthat galaxyevo- from <1% for M > 1011.5M to 20%-40% ⊙ lution is strongly driven by the galaxy mass. below M ∼ 1011M , and that there is no ⊙ It is interesting to note that recent studies of significantdifferencebetweentheevolutionof ETGsatz∼0havereachedconsistentconclu- massivefieldandclusterETGs(Fig.4). sions(Thomasetal.2005). Recently it has been found that while the Additional clues on downsizing come bright end of the cluster colour-magnitude from the most recent results on the evo- relation is already built at z ∼ 0.8, the lution of the Fundamental Plane (FP) faint end is still in the process of build-up both in the field and in distant clusters (Tanakaetal.2005) (see also De Lucia et al. (vanderWeletal.2004; Treuetal.2005; 2004). In contrast to this, the bright end of diSeregoAlighierietal.2005; the field colour-magnitude relation has been Jorgensenetal.2006). The FP at z ∼ 1 built all the way down to the present-day,but showsalreadya remarkablesmallscatter and, the build-up at the faint end has not started with respect to the local FP, an offset and a yet. This suggests again the downsizing evo- possible rotation.The evolutionof the overall lutionary pattern: massive galaxies complete FP can be represented by a mean change in their star formation first and the truncation of the effective mass-to-light ratio, but with a starformationispropagatedtosmallerobjects rate depending on the dynamical mass, being as time progresses. This trend is likely to de- slower for larger masses. This differential pendonenvironmentsincethebuild-upofthe evolutionisconsistentwithstellarpopulations colour-magnituderelation is delayed in lower thatformedatz > 2andz ∼ 1forhigh-and densityenvironments.Theevolutionofgalax- f f 286 Cimatti:Early-typegalaxyevolution Fig.5. Stellar mass functions in the K20 sam- ple. Different symbols correspond to different methods adopted to estimate the stellar mass (Fontanaetal.2004). In the highest redshift bin, largehollowcirclescorrespondtoobjectswithonly Fig.4. ThedifferentialevolutionoftheM/L ratio B spectroscopicredshift.Thethicksolidlineisthelo- withrespect tothat ofoldclustermassivegalaxies calgalaxymassfunction(Coleetal.2001). atthesameredshiftforETGsatz∼1fromtheK20 survey(diSeregoAlighierietal.2005). iestookplaceearliestinmassivegalaxiesand inhigh-densityregions,anditisdelayedinless massivegalaxiesandinlowerdensityregions. Other evidences of the downsizing sce- nario come from the evolution of the stellar mass function (Fig.5-6). Several surveys consistentlyindicatethatthehigh-masstailof the stellar mass function (which is populated mostly by ETGs; Fig.6) evolves weakly from z∼0.8−1.0toz∼0comparedtothelocalstel- larmassfunctionandslowerthanthelow-mass tail (Fontanaetal.2004; Droryetal.2005; Caputietal.2006; Bundyetal.2006; Pannellaetal.2006;Franceschinietal.2006). Recently, Bundy et al. (2006) confirmed Fig.6.StellarmassfunctionsintheK20samplefor this downsizing evolution of the stellar mass differentspectraltypes.Emptypointscorrespondto functionpergalaxytype uptoz ∼ 1 byusing latespectraltypes,filledtoETGspectraltypes.The the DEEP2 spectroscopicsample.In addition, solidlinesshowtheSchechterfitstothetotalmass theyidentifiedamasslimit,M ,abovewhich Q function of the K20 sample at the corresponding starformationappearstobe“quenched”.This redshifts(Fontanaetal.2004). threshold mass evolves as ∝ (1+ z)4.5, i.e. it decreases with time by a factor of 5 across the redshift range of 0.4 < z < 1.4, while wherethedownsizingseemstobeaccelerated. no significant correlation has been found be- AlsothespectralpropertiesofETGssupporta tween downsizing and environment, with the downsizingscenario:Fig.7showsthat,atfixed exception of the most extreme environments redshift,theD4000continuumbreakofETGs (i.e.theoneshostinghighernumbersofETGs), (taken as an age indicator of the stellar popu- Cimatti:Early-typegalaxyevolution 287 5. Comparisonwithmodel predictions These high-z old and massive ETGs are very rare objects in most N-body + semi- analytic simulations published to date. The predicted number densities are much lower than the observed ones (up to a factor of 10 in some cases), with this discrep- ancy becoming larger for increasing redshift (Somervilleetal.2004).However,asthereare enough dark matter halos to host these unex- pected massive galaxies (Fontanaetal.2004), what becomes crucial is to understand how the physics and evolution of baryons within the halos can explain the properties of high-z ETGs,andhowitispossibletobuild-upmas- siveETGsathigherredshiftsthanexpected. Fig.7. The 4000 Å break (D4000) in K20 ETG The models incorporating a new treat- composite spectra(Mignolietal.2005).TheETGs ment of the feedback processes (includ- ineachredshiftbinhavebeendividedintwoequally populated groups, according to their luminosity. ing AGN), or the ones based on hydro- Circlesandtrianglesindicatethebrighterandfainter dynamic simulations, seem to obtain a subsamples respectively. The dashed line indicates better agreement with the observations theD4000valuemeasuredintheaveragespectrum (Granatoetal.2004; Mencietal.2004; of all ETGs, whereas the dashed-dotted line is the Nagamineetal.2005; Silvaetal.2005). The D4000 measured in the SDSS composite of ETGs most recent simulations (Springeletal.2005) (Eisensteinetal.2003). are also capable to predict a downsizing evolutionarypatternforthestarformation,but mostofthemassassemblystilloccursatmod- eratelylowredshifts(0<z<1),andaspecific lation) is stronger in more luminous systems, comparison with the observed properties and suggesting a higher formation redshift of the evolutionof theETGshasnotbeenpublished mostmassiveETGs(Mignolietal.2005). yet (DeLuciaetal.2006; Boweretal.2005). Additionalevidence supportingthe down- It is interesting to mention that very recent sizing scenario comes from the results on the hydrodynamic simulations showed that it is star formation history properties. Feulner et possibletobuildarealisticmassive(M∼1011 al. (2005) found that, at all redshifts, lower M ) giant elliptical galaxy with a plausible mass galaxies show higher specific star for- ⊙ formation history without requiring neither mation rates (SSFR = SFR/M) than higher recent major mergers nor feedback from massgalaxies,andthatthehighestmassgalax- supernovae or AGN, but simply starting ies containthe oldeststellar populationsat all from appropriate ΛCDM initial conditions redshifts. With increasing redshift, the SSFR (Naabetal.2005). for massive galaxies increases by a factor of 10, reaching the era of their formation at z ∼ 2 and beyond. These findings can be inter- 6. Searchingfortheprogenitors preted as evidence for an early epoch of star formation in the most massive galaxies and The spectral and color properties of the for ongoing star formation activity in lower distant, old ETGs discussed in Section massgalaxies.Similarresultshavebeenfound 3 imply unambiguously a star formation also inotherrecentworks(Juneauetal.2005; history with: strong (> 100 M yr−1) and ⊙ Perez-Gonzalezetal.2005). short-lived (τ ∼0.1-0.3 Gyr) starburst (where 288 Cimatti:Early-typegalaxyevolution SFR∝ exp(t/τ)), the onset of star formation the same redshift range of 1.4 < z < 2.5 occurring at high redshift (z > 2 − 3), and (Daddietal.2005b; McCarthyetal.2004; f a passive–like evolution after the major star- Kongetal.2005). burst (Cimattietal.2004; Daddietal.2005a; McCarthyetal.2004; Longhettietal.2005). In addition, the ETG precursors should also 6.2.Otherprecursorcandidates have the strong clustering expected in the ΛCDM models for the populations located Other massive galaxy progenitor candi- dates at z > 2 characterized by old stellar in massive dark matter halos and strongly populations and/or strong star formation, biased environments. Recent observations large stellar and/or dynamical masses and have indeed uncovered galaxies matching the high metallicity include the submm/mm- aboverequirements. selected galaxies (Chapmanetal.2005; Dannerbaueretal.2004; 6.1.BzK–selectedgalaxies Swinbanketal.2004; Greveetal.2005), the “Distant Red Galaxies” se- The K20 survey unveiled a population of lected with J − K > 2.3 star-forming galaxies at z ∼ 2 which have s (Franxetal.2003; vanDokkumetal.2004; strong star formation rates of ∼ 200 − 300 Papovichetal.2006), the optically-selected M yr−1, substantial dust extinction with ⊙ BM/BX/LBG systems with bright K- E(B − V) ∼ 0.3 − 0.6, stellar masses up band fluxes (Adelbergeretal.2005; to 1011 M , merging-like rest-frame UV ⊙ Reddyetal.2005; Shapleyetal.2004), the morphology becoming more compact in the IRAC Extremely Red Objects (IEROs) rest-frame optical, strong clustering with (YanH.etal.2004), the HyperEROs r ∼8–12 Mpc depending on the adopted 0 (Totanietal.2001; Saraccoetal.2004), redshift distribution, and possibly high metal- andafractionofdustyEROs(Yanetal.2005). licity (Daddietal.2004a; Daddietal.2004b; Daddietal.2005b; deMelloetal.2004; Kongetal.2005). These starburst galaxies 6.3.TheGMASSproject can be efficiently selected at 1.4 < z < 2.5 with a “reddening-free” color–color dia- GMASS (“Galaxy Mass Assembly ultra-deep gram (z-K vs. B−z; the BzK selection) SpectroscopicSurvey”)isanewongoingspec- s (Daddietal.2004b). The BzK–selected star- troscopic project based on an ESO Large bursts with K < 20 are on average more Program (PI A. Cimatti) aimed at doing ul- s reddened, massive and star-forming than tradeep spectroscopy with VLT+FORS2 of the optically-selected star-forming galaxies, galaxiesselectedat4.5µmwithSpitzer+IRAC but they show a significant overlap with adoptingtwosimplecriteria:m(4.5µm)< 23.0 other star-forming galaxy populations in the (AB) and z > 1.4. The integration times phot same redshift range selected based on other are very deep (15-30 hours per spectroscopic optical/near-IR criteria (Reddyetal.2005) mask). The main scientific driver of GMASS or on mid–infrared - to - radio selections isthespectroscopicidentificationandstudyof (Daddietal.2005b; Dannerbaueretal.2006). thehigh-zprogenitorsofthemassiveETGsina The properties of the BzK–selected starbursts waycomplementarytoothersurveysandwith suggest that these galaxiesare massive galax- a selection more sensitive to the rest-frame ies caughtduring their major activity of mass near-IRemission(i.e.stellarmass)uptoz∼3. assembly and star-formation, possibly being theprogenitorsofthepresent-daymassiveand metal-richETGs.Atthestellarmassthreshold Acknowledgements. I am grateful to Emanuele of M > 1011M⊙, the numberdensity of these DaddiforhiscommentsandtotheK20andGMASS BzK–selected starbursts is comparable with collaborators for their invaluable contributions to that of old passively evolving galaxies in theprojects. Cimatti:Early-typegalaxyevolution 289 References JuneauS.etal2005,ApJ,619,L135 KongX.etal2005,ApJ,inpress AbrahamR.etal2004,AJ,127,2455 Labbe’I.etal2005,ApJ,624,L81 AdelbergerK.etal2005,ApJ,620,L75 LonghettiM.etal2005,MNRAS,361,897 BellE.etal2004,ApJ,608,752 MaozD.1997,ApJ,490,L135 BellE.etal2005,ApJ,inpress. MasjediM.etal2005,ApJ,submitted BenitezN.etal1999,ApJ,515,L65 McCarthyP.J.etal2001,ApJ,560,L131 BowerR.G.etal2005,MNRAS,submitted McCarthyP.J.etal2004,ApJ,614,L9 BrownM.J.I.etal2005,ApJ,621,41 MenanteauF.etal2004,ApJ,612,202 BundyK.etal2006,ApJ,submitted MenciN.etal2004,ApJ,604,12 CaputiK.etal2006,MNRAS,inpress MignoliM.etal2005,A&A,437,883 ChapmanS.etal2005,ApJ,622,772 NaabT.etal2005,ApJL,submitted CimattiA.etal2002a,A&A,381,L68 NagamineK.etal2005,ApJ,627,608 CimattiA.etal2002b,A&A,392,395 PannellaM.etal2006,ApJ,inpress CimattiA.etal2003,A&A,412,L1 PapovichC.etal2006,ApJ,inpress CimattiA.etal2004,Nature,430,184 PasqualiA.etal2006,ApJ,636,115 ColeS.,etal,2001,MNRAS,326,255 Perez-GonzalezP.G.etal2005,ApJ,630,82 CowieL.etal2006,AJ,112,839 PozzettiL.etal2003,A&A,402,837 DaddiE.etal2000,A&A,361,535 ReddyN.etal.2005,ApJ,633,748 DaddiE.etal2002,A&A,384,L1 RenziniA.2006,ARA&A,inpress DaddiE.etal2004a,ApJ,600,L127 RetturaA.etal.2006,A&A,submitted DaddiE.etal2004b,ApJ,617,746 RocheN.etal.2003,MNRAS,346,803 DaddiE.etal2005a,ApJ,626,680 SaraccoP.etal2003,A&A,398,127 DaddiE.etal2005b,ApJL,631,L13 SaraccoP.etal2004,A&A,420,125 DannerbauerH.etal2004,ApJ,606,664 SaraccoP.etal2005,MNRAS,372,L40 DannerbauerH.etal2006,ApJL,inpress ShapleyA.etal2004,ApJ,612,108 DeLuciaG.etal2004,ApJ,610,L77 SilvaL.etal2005,MNRAS,357,1295 DeLuciaG.etal2006,MNRAS,inpress SomervilleR.etal2004a,ApJ,600,L171 deMelloD.etal2004,ApJ,608,L29 SomervilleR.etal2004b,ApJ,600,L135 diSeregoAlighieriS.etal2005,A&A,442, SpringelV.etal2005,Nature,435,629 125 StanfordS.A.etal2004,AJ,127,131 DohertyM.etal2005,MNRAS,361,525 SwinbankA.M.etal2004,ApJ,617,64 DroryN.etal2005,ApJ,619,L131 TanakaM.etal2005,MNRAS,362,268 DunlopJ.etal1996,Nature,381,581 ThomasD.etal2005,ApJ,621,673 Eggen O.J., Lynden-Bell D., Sandage A. TinsleyB.M.1972,ApJ,178,319 1962,ApJ,136,748 TotaniT.etal2001,ApJ,558,L87 EisensteinD.etal2003,ApJ,585,694 Treu,T.etal1998,A&A,340,L10 FaberS.etal2005,ApJ,submitted Treu,T.etal2005,ApJ,633,174 FerrerasI.etal2005,ApJ,635,243 vanderWeletal2004,ApJ,601,L5 FeulnerG.etal2005,ApJ,633,L9 vanDokkumP. &FranxM. 2001,ApJ,553, FirthA.E.etal2002,MNRAS,332,617 90 FontanaA.etal2004,A&A,424,23 vanDokkumP.etal2004,ApJ,611,703 FranceschiniA.etal vanDokkumP.2005,AJ,130,2647 FranxM.etal2003,ApJ,587,L79 YamadaT.etal2005,ApJ,634,861 FuH.etal2005,ApJ,632,831 YanH.etal2004,ApJ,616,63 GeorgakakisA.etal2005,ApJ,620,584 YanL.etal2004,AJ,127,1274 GlazebrookK.etal2004,Nature,430,181 YanL.etal2005,ApJS,154,75 GranatoG.etal2004,ApJ,600,580 ZepfS.E.1997,Nature,390,377 GreveT.R.etal2005,MNRAS,359,1165 ZuccaE.etal2005,A&A,inpress JorgensenI.etal2006,ApJL,inpress

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.