Evolution of the Stellar-to-Dark Matter Relation: Separating Star-Forming and Passive Galaxies from z=1 to 0

We use measurements of the stellar mass function, galaxy clustering, and galaxy-galaxy lensing within the COSMOS survey to constrain the stellar-to-halo mass relation (SHMR) of star forming and quiescent galaxies over the redshift range z=[0.2,1.0]. For massive galaxies, M*>~10^10.6 Msol, our results indicate that star-forming galaxies grow proportionately as fast as their dark matter halos while quiescent galaxies are outpaced by dark matter growth. At lower masses, there is minimal difference in the SHMRs, implying that the majority low-mass quiescent galaxies have only recently been quenched of their star formation. Our analysis also affords a breakdown of all COSMOS galaxies into the relative numbers of central and satellite galaxies for both populations. At z=1, satellite galaxies dominate the red sequence below the knee in the stellar mass function. But the number of quiescent satellites exhibits minimal redshift evolution; all evolution in the red sequence is due to low-mass central galaxies being quenched of their star formation. At M*~10^10 Msol, the fraction of central galaxies on the red sequence increases by a factor of ten over our redshift baseline, while the fraction of quenched satellite galaxies at that mass is constant with redshift. We define a migration rate to the red sequence as the time derivative of the passive galaxy abundances. We find that the migration rate of central galaxies to the red sequence increases by nearly an order of magnitude from z=1 to z=0. These results imply that the efficiency of quenching star formation for centrals is increasing with cosmic time, while the mechanisms that quench the star formation of satellite galaxies in groups and clusters is losing efficiency.

The Correlated Formation Histories of Massive Galaxies and Their Dark Matter Halos

Using observations in the COSMOS field, we report an intriguing correlation between the star formation activity of massive (~10^{11.4}msol) central galaxies, their stellar masses, and the large-scale (~10 Mpc) environments of their group-mass (~10^{13.6}msol) dark matter halos. Probing the redshift range z=[0.2,1.0], our measurements come from two independent sources: an X-ray detected group catalog and constraints on the stellar-to-halo mass relation derived from a combination of clustering and weak lensing statistics. At z=1, we find that the stellar mass in star-forming centrals is a factor of two less than in passive centrals at the same halo mass. This implies that the presence or lack of star formation in group-scale centrals cannot be a stochastic process. By z=0, the offset reverses, probably as a result of the different growth rates of these objects. A similar but weaker trend is observed when dividing the sample by morphology rather than star formation. Remarkably, we find that star-forming centrals at z~1 live in groups that are significantly more clustered on 10 Mpc scales than similar mass groups hosting passive centrals. We discuss this signal in the context of halo assembly and recent simulations, suggesting that star-forming centrals prefer halos with higher angular momentum and/or formation histories with more recent growth; such halos are known to evolve in denser large-scale environments. If confirmed, this would be evidence of an early established link between the assembly history of halos on large scales and the future properties of the galaxies that form inside them.

The Clustering of MgII Absorption Systems at z=0.5 and Detection of Cold Gas in Massive Halos

We measure the large-scale clustering of MgII lambdalambda 2796,2803 absorbers with respect to a population of luminous red galaxies (LRGs) at z sim 0.5. From the cross-correlation measurements between MgII absorbers and LRGs, we calculate the mean bias of the dark matter halos in which the absorbers reside. We investigate systematic uncertainties in the clustering measurements due to the sample selection of LRGs and due to uncertainties in photometric redshifts. First, we compare the cross-correlation amplitudes determined using a it flux-limited LRG sample and a volume-limited one. The comparison shows that the relative halo bias of MgII absorbers using a flux-limited LRG sample can be overestimated by as much as approx 20%. Next, we assess the systematic uncertainty due to photometric redshift errors using a mock galaxy catalog with added redshift uncertainties comparable to the data. We show that the relative clustering amplitude measured without accounting for photometric redshift uncertainties is overestimated by approx 10%. After accounting for these two main uncertainties, we find a 1-sigma anti-correlation between mean halo bias and absorber strength that translates into a 1-sigma anti-correlation between mean galaxy mass and W_r(2796). The results indicate that a significant fraction of the MgII absorber population of W_r(2796)=1-1.5 AA are found in group-size dark matter halos of log M_h < 13.4, whereas absorbers of W_r(2796)>1.5 AA are seen in halos of log M_h <12.7. A larger dataset would improve the precision of the clustering measurements and the relationship between W_r and halo mass. Finally, the strong clustering of MgII absorbers down to sim 0.3 h^{-1} Mpc indicates the presence of cool gas inside the virial radii of the halos hosting the LRGs.

Interpreting the Clustering of Distant Red Galaxies

We analyze the angular clustering of z~2.3 distant red galaxies (DRGs) measured by Quadri et al 2008. We find that, with robust estimates of the measurement errors and realistic halo occupation distribution modeling, the measured clustering can be well fit within standard halo occupation models, in contrast to previous results. However, in order to fit the strong break in w(theta) at theta=10 arcsec, nearly all satellite galaxies in the DRG luminosity range are required to be DRGs. Within this luminosity-threshold sample, the fraction of galaxies that are DRGs is ~44%, implying that the formation of DRGs is more efficient for satellite galaxies than for central galaxies. Despite the evolved stellar populations contained within DRGs at z=2.3, 90% of satellite galaxies in the DRG luminosity range have been accreted within 500 Myr. Thus, satellite DRGs must have known they would become satellites well before the time of their accretion. This implies that the formation of DRGs correlates with large-scale environment at fixed halo mass, although the large-scale bias of DRGs can be well fit without such assumptions. Further data are required to resolve this issue. Using the observational estimate that ~30% of DRGs have no ongoing star formation, we infer a timescale for star formation quenching for satellite galaxies of 450 Myr, although the uncertainty on this number is large. However, unless all non-star forming satellite DRGs were quenched before accretion, the quenching timescale is significantly shorter than z~0 estimates. Down to the completeness limit of the Quadri et al sample, we find that the halo masses of central DRGs are ~50% higher than non-DRGs in the same luminosity range, but at the highest halo masses the central galaxies are DRGs only ~2/3 of the time.