Search

[Abridged] We present the evolution of the stellar mass function (SMF) of galaxies from z=4.0 to z=1.3 measured from a sample constructed from the deep NIR MUSYC, the FIRES, and the GOODS-CDFS surveys, all having very high-quality optical to mid-infrared data. This sample, unique in that it combines data from surveys with a large range of depths and areas in a self-consistent way, allowed us to 1) minimize the uncertainty due to cosmic variance and empirically quantify its contribution to the total error budget; 2) simultaneously probe the high-mass end and the low-mass end (down to ~0.05 times the characteristic stellar mass) of the SMF with good statistics; and 3) empirically derive the redshift-dependent completeness limits in stellar mass. We provide, for the first time, a comprehensive analysis of random and systematic uncertainties affecting the derived SMFs. We find that the mass density evolves by a factor of ~17(+7,-10) since z=4.0, mostly driven by a change in the normalization Phi*. If only random errors are taken into account, we find evidence for mass-dependent evolution, with the low-mass end evolving more rapidly than the high-mass end. However, we show that this result is no longer robust when systematic uncertainties due to the SED-modeling assumptions are taken into account. Taking our results at face value, we find that they are in conflict with semi-analytic models of galaxy formation. The models predict SMFs that are in general too steep, with too many low-mass galaxies and too few high-mass galaxies. The discrepancy at the high-mass end is susceptible to uncertainties in the models and the data, but the discrepancy at the low-mass end may be more difficult to explain., Comment: 33 pages, 22 figures. Accepted for publication in the Astrophysical Journal. Version revised according to referee's comments; figures updated accordingly

We study the evolution of the star formation rate (SFR) of mid-infrared (IR) selected galaxies in the extended Chandra Deep Field South (E-CDFS). We use a combination of U-K GaBoDS and MUSYC data, deep IRAC observations from SIMPLE, and deep MIPS data from FIDEL. This unique multi-wavelength data set allows us to investigate the SFR history of massive galaxies out to redshift z ~ 1.8. We determine star formation rates using both the rest-frame ultraviolet luminosity from young, hot stars and the total IR luminosity of obscured star formation obtained from the MIPS 24 um flux. We find that at all redshifts the galaxies with higher masses have substantially lower specific star formation rates than lower mass galaxies. The average specific star formation rates increase with redshift, and the rate of incline is similar for all galaxies (roughly (1+z)^{n}, n = 5.0 +/- 0.4). It does not seem to be a strong function of galaxy mass. Using a subsample of galaxies with masses M_*> 10^11 M_sun, we measured the fraction of galaxies whose star formation is quenched. We consider a galaxy to be in quiescent mode when its specific star formation rate does not exceed 1/(3 x t_H), where t_H is the Hubble time. The fraction of quiescent galaxies defined as such decreases with redshift out to z ~ 1.8. We find that, at that redshift, 19 +/-9 % of the M_* > 10^11 M_sun sources would be considered quiescent according to our criterion., Comment: 7 pages, 6 figures, accepted for publication in ApJ

We present an analysis of galaxies in the CDF-South. We find a tight relation to z=3 between color and size at a given mass, with red galaxies being small, and blue galaxies being large. We show that the relation is driven by stellar surface density or inferred velocity dispersion: galaxies with high surface density are red and have low specific star formation rates, and galaxies with low surface density are blue and have high specific star formation rates. Surface density and inferred velocity dispersion are better correlated with specific star formation rate and color than stellar mass. Hence stellar mass by itself is not a good predictor of the star formation history of galaxies. In general, galaxies at a given surface density have higher specific star formation rates at higher redshift. Specifically, galaxies with a surface density of 1-3 10^9 Msun/kpc^2 are "red and dead" at low redshift, approximately 50% are forming stars at z=1, and almost all are forming stars by z=2. This provides direct additional evidence for the late evolution of galaxies onto the red sequence. The sizes of galaxies at a given mass evolve like 1/(1+z)^(0.59 +- 0.10). Hence galaxies undergo significant upsizing in their history. The size evolution is fastest for the highest mass galaxies, and quiescent galaxies. The persistence of the structural relations from z=0 to z=2.5, and the upsizing of galaxies imply that a relation analogous to the Hubble sequence exists out to z=2.5, and possibly beyond. The star forming galaxies at z >= 1.5 are quite different from star forming galaxies at z=0, as they have likely very high gas fractions, and star formation time scales comparable to the orbital time., Comment: 20 pages, accepted for publication in ApJ, 2008, 689

We investigate the properties of quiescent and star-forming galaxy populations to z~2 with purely photometric data, employing a novel rest-frame color selection technique. From the UKIDSS Ultra-Deep Survey Data Release 1, with matched optical and mid-IR photometry taken from the Subaru XMM Deep Survey and Spitzer Wide-Area Infrared Extragalactic Survey respectively, we construct a K-selected galaxy catalog and calculate photometric redshifts. Excluding stars, objects with uncertain z_phot solutions, those that fall in bad or incomplete survey regions, and those for which reliable rest-frame colors could not be derived, 30108 galaxies with K<22.4 (AB) and z<2.5 remain. The galaxies in this sample are found to occupy two distinct populations in the rest-frame U-V vs. V-J color space: a clump of red, quiescent galaxies (analogous to the red sequence) and a track of star-forming galaxies extending from blue to red U-V colors. This bimodal behavior is seen up to z~2. Due to a combination of measurement errors and passive evolution, the color-color diagram is not suitable to distinguish the galaxy bimodality at z>2 for this sample, but we show that MIPS 24um data suggest that a significant population of quiescent galaxies exists even at these higher redshifts. At z=1-2, the most luminous objects in the sample are divided roughly equally between star-forming and quiescent galaxies, while at lower redshifts most of the brightest galaxies are quiescent. Moreover, quiescent galaxies at these redshifts are clustered more strongly than those actively forming stars, indicating that galaxies with early-quenched star formation may occupy more massive host dark matter halos. This suggests that the end of star formation is associated with, and perhaps brought about by, a mechanism related to halo mass., Comment: 17 pages, 16 figures, accepted to ApJ after minor revisions. The K-selected catalog described herein can be downloaded at http://www.strw.leidenuniv.nl/galaxyevolution/UDS

We present a Ks-selected catalog, dubbed FIREWORKS, for the Chandra Deep Field South (CDFS) containing photometry in U_38, B_435, B, V, V_606, R, i_775, I, z_850, J, H, Ks, [3.6 um], [4.5 um], [5.8 um], [8.0 um], and the MIPS [24 um] band. The imaging has a typical Ks limit of 24.3 mag (5 sigma, AB) and coverage over 113 arcmin^2 in all bands and 138 arcmin^2 in all bands but H. We cross-correlate our catalog with the 1 Ms X-ray catalog by Giacconi et al. (2002) and with all available spectroscopic redshifts to date. We find and explain systematic differences in a comparison with the 'z_850 + Ks'-selected GOODS-MUSIC catalog that covers ~90% of the field. We exploit the U38-to-24 micron photometry to determine which Ks-selected galaxies at 1.5

We present the mid-infrared colors of X-ray-detected AGN and explore mid-infrared selection criteria. Using a statistical matching technique, the likelihood ratio, over 900 IRAC counterparts were identified with a new MUSYC X-ray source catalog that includes ~1000 published X-ray sources in the Chandra Deep Field-South and Extended Chandra Deep Field-South. Most X-ray-selected AGN have IRAC spectral shapes consistent with power-law slopes, f_{nu} ~ nu^{alpha}, and display a wide range of colors, -2 < alpha < 2. Although X-ray sources typically fit to redder (more negative alpha) power-laws than non-X-ray detected galaxies, more than 50% do have flat or blue (galaxy-like) spectral shapes in the observed 3-8 micron band. Only a quarter of the X-ray selected AGN detected at 24 micron are well fit by featureless red power laws in the observed 3.6-24 micron, likely the subset of our sample whose infrared spectra are dominated by emission from the central AGN region. Most IRAC color-selection criteria fail to identify the majority of X-ray-selected AGN, finding only the more luminous AGN, the majority of which have broad emission lines. In deep surveys, these color-selection criteria select 10-20% of the entire galaxy population and miss many moderate luminosity AGN., Comment: 29 pages, including 7 figures. Accepted for publication in the Astrophysical Journal

Using the Gemini Near-InfraRed Spectrograph (GNIRS), we have completed a near-infrared spectroscopic survey for K-bright galaxies at z~2.3, selected from the MUSYC survey. We derived spectroscopic redshifts from emission lines or from continuum features and shapes for all 36 observed galaxies. The continuum redshifts are driven by the Balmer/4000 Angstrom break, and have an uncertainty in dz/(1+z) of <0.019. We use this unique sample to determine, for the first time, how accurately redshifts and other properties of massive high-redshift galaxies can be determined from broadband photometric data alone. We find that the photometric redshifts of the galaxies in our sample have a systematic error of 0.08 and a random error of 0.13 in dz/(1+z). The systematic error can be reduced by using optimal templates and deep photometry; the random error, however, will be hard to reduce below 5%. The spectra lead to significantly improved constraints for stellar population parameters. For most quantities this improvement is about equally driven by the higher spectral resolution and by the much reduced redshift uncertainty. Properties such as the age, A_V, current star formation rate, and the star formation history are generally very poorly constrained with broadband data alone. Interestingly stellar masses and mass-to-light ratios are among the most stable parameters from broadband data. Nevertheless, photometric studies may overestimate the number of massive galaxies at 2

We investigate the stellar populations of a sample of 162 Lyman-alpha emitting galaxies (LAEs) at z = 3.1 in the Extended Chandra Deep Field South, using deep Spitzer IRAC data available from the GOODS and SIMPLE surveys to derive reliable stellar population estimates. We divide the LAEs according to their rest-frame near-IR luminosities into IRAC-detected and IRAC-undetected samples. About 70% of the LAEs are undetected in 3.6 micron down to [3.6] = 25.2 AB. Stacking analysis reveals that the average stellar population of the IRAC-undetected sample has an age of ~ 200 Myr and a mass of ~ 3x10^8 solar masses, consistent with the expectation that LAEs are mostly young and low-mass galaxies. On the other hand, the IRAC-detected LAEs are on average significantly older and more massive, with an average age > 1 Gyr and mass ~ 10^10 solar masses. Comparing the IRAC colors and magnitudes of the LAEs to z ~ 3 Lyman break galaxies (LBGs) shows that the IRAC-detected LAEs lie at the faint blue end of the LBG color-magnitude distribution, suggesting that IRAC-detected LAEs may be the low mass extension of the LBG population. We also present tentative evidence for a small fraction (~ 5%) of obscured AGN within the LAE sample. Our results suggest that LAEs posses a wide range of ages and masses. Additionally, the presence of evolved stellar populations inside LAEs suggests that the Lyman-alpha luminous phase of galaxies may either be a long-lasting or recurring phenomenon., Comment: Accepted for publication in ApJ; 5 pages, 4 figures

IRAC excels at detecting distant objects. Due to a combination of the shapes of the spectral energy distributions of galaxies and the low background achieved from space, IRAC reaches greater depth in comparable exposure time at 3.6 and 4.5 micron than any ground- or space-based facility currently can at 2.2 micron. Furthermore, the longer wavelengths probed by IRAC enable studies of the rest-frame optical and near-infrared light of galaxies and AGN to much higher redshift than is possible from the ground. This white paper explores the merits of different survey strategies for studying the distant universe during the warm mission. A three-tiered approach serves a wide range of science goals and uses the spacecraft effectively: 1) an ultra-deep survey of ~0.04 square degrees to a depth of ~250 hrs (in conjunction with an HST/WFC3 program), to study the Universe at 76 and characterize the relation between the build-up of dark matter halos and their constituent galaxies at 2

Using the Gemini Near-InfraRed Spectrograph (GNIRS), we have assembled a complete sample of 20 K-selected galaxies at 2.0

We analyze very deep HST, VLT and Spitzer photometry of galaxies at 22 galaxies. The estimated distributions of these properties do not change significantly when adding IRAC data to the UBVIJHK photometry. However, for individual galaxies the addition of IRAC can improve the constraints on the stellar populations, especially for red galaxies: uncertainties in stellar mass decrease by a factor of 2.7 for red (U-V > 1) galaxies, but only by a factor of 1.3 for blue (U-V < 1) galaxies. We find a similar color-dependence of the improvement for estimates of age and dust extinction. In addition, the improvement from adding IRAC depends on the availability of full near-infrared JHK coverage; if only K-band were available, the mass uncertainties of blue galaxies would decrease by a more substantial factor 1.9. Finally, we find that a trend of galaxy color with stellar mass is already present at z>2. The most massive galaxies at high redshift have red rest-frame U-V colors compared to lower mass galaxies even when allowing for complex star formation histories., Comment: Accepted for publication in the Astrophysical Journal, 16 pages, 16 figures

We present first results of a spectroscopic survey targeting K-selected galaxies at z=2.0-2.7 using the GNIRS instrument on Gemini-South. We obtained near-infrared spectra with a wavelength coverage of 1.0-2.5 micron for 26 K-bright galaxies (K<19.7) selected from the MUSYC survey using photometric redshifts. We successfully derived spectroscopic redshifts for all 26 galaxies using rest-frame optical emission lines or the redshifted Balmer/4000 Angstrom break. Twenty galaxies have spectroscopic redshifts in the range 2.0

Using Spitzer IRAC mid-infrared imaging from the Great Observatories Origins Deep Survey, we study z_850-dropout sources in the Hubble Ultra Deep Field. After carefully removing contaminating flux from foreground sources, we clearly detect two z_850-dropouts at 3.6 micron and 4.5 micron, while two others are marginally detected. The mid-infrared fluxes strongly support their interpretation as galaxies at z~7, seen when the Universe was only 750 Myr old. The IRAC observations allow us for the first time to constrain the rest-frame optical colors, stellar masses, and ages of the highest redshift galaxies. Fitting stellar population models to the spectral energy distributions, we find photometric redshifts in the range 6.7-7.4, rest-frame colors U-V=0.2-0.4, V-band luminosities L_V=0.6-3 x 10^10 L_sun, stellar masses 1-10 x 10^9 M_sun, stellar ages 50-200 Myr, star formation rates up to ~25 M_sun/yr, and low reddening A_V<0.4. Overall, the z=7 galaxies appear substantially less massive and evolved than Lyman break galaxies or Distant Red Galaxies at z=2-3, but fairly similar to recently identified systems at z=5-6. The stellar mass density inferred from our z=7 sample is rho* = 1.6^{+1.6}_{-0.8} x 10^6 M_sun Mpc^-3 (to 0.3 L*(z=3)), in apparent agreement with recent cosmological hydrodynamic simulations, but we note that incompleteness and sample variance may introduce larger uncertainties. The ages of the two most massive galaxies suggest they formed at z>8, during the era of cosmic reionization, but the star formation rate density derived from their stellar masses and ages is not nearly sufficient to reionize the universe. The simplest explanation for this deficiency is that lower-mass galaxies beyond our detection limit reionized the universe., Comment: 4 pages, 3 figures, emulateapj, Accepted for publication in ApJ Letters

(Abridged) We present the evolution of the volume averaged properties of the rest-frame optically luminous galaxy population to z~3, determined from four disjoint deep fields with optical to near-infrared wavelength coverage. We select galaxies above a rest-frame V-band luminosity of 3x10^10 Lsol and characterize their rest-frame UV through optical properties via the mean spectral energy distribution (SED). To measure evolution we apply the same selection criteria to a sample of galaxies from the Sloan Digital Sky Survey and COMBO-17. The mean rest-frame 2200Ang through V-band SED becomes steadily bluer with increasing redshift but at z<3 the mean SED falls within the range defined by ``normal'' galaxies in the nearby Universe. We measure stellar mass-to-light ratios (Mstar/L) by fitting models to the rest-frame UV-optical SEDs and derive the stellar mass density. The stellar mass density in luminous galaxies has increased by a factor of 3.5-7.9 from z=3 to z=0.1, including field-to-field variance uncertainties. After correcting to total, the measured mass densities at z<2 lie below the integral of the star formation rate (SFR) density as a function of redshift as derived from UV selected samples. This may indicate a systematic error in the mass densities or SFR(z) estimates. We find large discrepancies between recent model predictions for the evolution of the mass density and our results, even when our observational selection is applied to the models. Finally we determine that Distant Red Galaxies (selected to have J_s - K_s>2.3) in our LV selected samples contribute 30% and 64% of the stellar mass budget at z~2 and z~ 2.8 respectively. These galaxies are largely absent from UV surveys and this result highlights the need for mass selection of high redshift galaxies., Comment: Accepted for publication in the Astrophysical Journal, 24 pages, 16 figures

We study the clustering properties of K-selected galaxies at 22.3 distant red galaxies (DRGs) have r_0 ~ 11 h^-1 Mpc. Contrary to findings for optically-selected galaxies, K-selected galaxies that are faint in the R-band cluster more strongly than brighter galaxies. These results suggest that a color-density relation was in place at z>2. Our results indicate that K-bright blue galaxies and K-bright red galaxies are fundamentally different, as they have different clustering properties. Using a simple model of one galaxy per halo, we infer halo masses ~ 5 x 10^12 M_sun for K<21 galaxies and ~ 2 x 10^13 M_sun for DRGs. A comparison of the observed space density of DRGs to the density of their host halos suggests large halo occupation numbers; however, this result is at odds with the lack of a strong small-scale excess in the angular correlation function. We find that the z=0 descendants of the galaxies considered here reside primarily in groups and clusters. [abridged], Comment: 17 pages in emulateapj style, 14 figures. Resubmitted to ApJ after referee's report; minor clarifications and revisions. Update: accepted by ApJ

We use near-infrared (NIR) spectroscopy obtained with GNIRS on Gemini, NIRSPEC on KECK, and ISAAC on the VLT to study the rest-frame optical continua of three `Distant Red Galaxies' (having Js - Ks > 2.3) at z>2. All three galaxy spectra show the Balmer/4000 Angstrom break in the rest-frame optical. The spectra allow us to determine spectroscopic redshifts from the continuum with an estimated accuracy dz/(1+z) ~ 0.001-0.04. These redshifts agree well with the emission line redshifts for the 2 galaxies with Halpha emission. This technique is particularly important for galaxies that are faint in the rest-frame UV, as they are underrepresented in high redshift samples selected in optical surveys and are too faint for optical spectroscopy. Furthermore, we use the break, continuum shape, and equivalent width of Halpha together with evolutionary synthesis models to constrain the age, star formation timescale, dust content, stellar mass and star formation rate of the galaxies. Inclusion of the NIR spectra in the stellar population fits greatly reduces the range of possible solutions for stellar population properties. We find that the stellar populations differ greatly among the three galaxies, ranging from a young dusty starburst with a small break and strong emission lines to an evolved galaxy with a strong break and no detected line emission. The dusty starburst galaxy has an age of 0.3 Gyr and a stellar mass of 1*10^11 Msun. The spectra of the two most evolved galaxies imply ages of 1.3-1.4 Gyr and stellar masses of 4*10^11 Msun. The large range of properties seen in these galaxies strengthens our previous much more uncertain results from broadband photometry. Larger samples are required to determine the relative frequency of dusty starbursts and (nearly) passively evolving galaxies at z~2.5., Comment: Accepted for publication in the Astrophysical Journal. 12 pages, 6 figures

We present Spitzer 24micron imaging of 1.5 < z < 2.5 Distant Red Galaxies (DRGs) in the 10arcmin by 10arcmin Extended Hubble Deep Field South of the Multiwavelength Survey by Yale-Chile. We detect 65% of the DRGs with K_AB < 23.2 mag at S_24micron > 40uJy, and conclude that the bulk of the DRG population are dusty active galaxies. A mid-infrared (MIR) color analysis with IRAC data suggests that the MIR fluxes are not dominated by buried AGN, and we interpret the high detection rate as evidence for a high average star formation rate of = 130+/-30 M/yr. From this, we infer that DRGs are important contributors to the cosmic star formation rate density at z ~ 2, at a level of \~0.02 M/yr/Mpc^3 to our completeness limit of K_AB = 22.9 mag, Comment: 4 pages, 3 figures, accepted for publication in ApJL

We present the evolution of the luminosity-size and stellar mass-size relations of luminous (L_V>3.4x10^10h_70^-2L_sun) and of massive (M_*>3x10^10h_70^-2M_sun) galaxies in the last ~11 Gyr. We use very deep near-infrared images of the Hubble Deep Field-South and the MS1054-03 field in the J_s, H and K_s bands from FIRES to retrieve the sizes in the optical rest-frame for galaxies with z>1. We combine our results with those from GEMS at 0.2

We recently identified a substantial population of galaxies at z>2 with red rest-frame optical colors. These distant red galaxies (DRGs) are efficiently selected by the simple observed color criterion J-K>2.3. In this paper we present NIR spectroscopy with Keck/NIRSPEC of six DRGs at 2.4

We present the evolution of the rest-frame optical luminosity density, of the integrated rest-frame optical color, and of the stellar mass density for a sample of Ks-band selected galaxies in the HDF-S. We derived the luminosity density in the rest-frame U, B, and V-bands and found that the luminosity density increases by a factor of 1.9+-0.4, 2.9+-0.6, and 4.9+-1.0 in the V, B, and U rest-frame bands respectively between a redshift of 0.1 and 3.2. We derived the luminosity weighted mean cosmic (U-B)_rest and (B-V)_rest colors as a function of redshift. The colors bluen almost monotonically with increasing redshift; at z=0.1, the (U-B)_rest and (B-V)_rest colors are 0.16 and 0.75 respectively, while at z=2.8 they are -0.39 and 0.29 respectively. We derived the luminosity weighted mean M/LV using the correlation between (U-V)_rest and log_{10} M/LV which exists for a range in smooth SFHs and moderate extinctions. We have shown that the mean of individual M/LV estimates can overpredict the true value by ~70% while our method overpredicts the true values by only ~35%. We find that the universe at z~3 had ~10 times lower stellar mass density than it does today in galaxies with LV>1.4 \times 10^{10} h_{70}^-2 Lsol. 50% of the stellar mass of the universe was formed by $z~1-1.5. The rate of increase in the stellar mass density with decreasing redshift is similar to but above that for independent estimates from the HDF-N, but is slightly less than that predicted by the integral of the SFR(z) curve., Comment: 19 pages, 12 figures, Accepted for Publication in the Dec. 20, 2003 edition of the Astrophysical Journal. Minor changes made to match the accepted version including short discussions on the effects of clustering and on possible systematic effects resulting from photometric redshift errors

The luminosity-size and stellar mass-size distributions of galaxies out to z~3 is presented. We use very deep near-infrared images of the Hubble Deep Field South in the J_s, H, and K_s bands, taken as part FIRES at the VLT, to follow the evolution of the optical rest-frame sizes of galaxies. For a total of 168 galaxies with K_{s,AB}<23.5, we find that the rest-frame V-band sizes r_{e,V} of luminous galaxies (~2x10^{10}h^{-2}L_sun) at 2~2x10^{10}h^{-2}M_sun, has changed by 20(+-20)% since z~2.5. Both results can be reconciled by the fact that the stellar M/L ratio is lower in the luminous high z galaxies than in nearby ones because they have young stellar populations. The lower incidence of large galaxies at z~3 seems to reflect the rarity of galaxies with high stellar mass., Comment: 39 pages, 14 figures, ApJ in press, new changes made to match the accepted version

We confirm spectroscopically the existence of a population of galaxies at z>~2 with rest-frame optical colors similar to normal nearby galaxies. The galaxies were identified by their red near-infrared colors in deep images obtained with ISAAC on the Very Large Telescope. Redshifts of six galaxies with Js-Ks>2.3 were measured from optical spectra obtained with the Keck Telescope. Five out of six are in the range 2.43<=z<=3.52, demonstrating that the Js-Ks color selection is quite efficient. The rest-frame ultraviolet spectra of confirmed z>2 galaxies display a range of properties, with two galaxies showing emission lines characteristic of AGN, two having Ly-alpha in emission, and one showing interstellar absorption lines only. Their full spectral energy distributions are well described by constant star formation models with ages 1.4-2.6 Gyr, except for one galaxy whose colors indicate a dusty starburst. The confirmed z>2 galaxies are very luminous, with Ks=19.2-19.9. Assuming that our bright spectroscopic sample is representative for the general population of Js-Ks selected objects, we find that the surface density of red z>~2 galaxies is ~0.9/arcmin^2 to Ks=21. The surface density is comparable to that of Lyman-break selected galaxies with Ks<21, when corrections are made for the different redshift distributions of the two samples. Although there will be some overlap between the two populations, most 'optical-break' galaxies are too faint in the rest-frame ultraviolet to be selected as Lyman-break galaxies. The most straightforward interpretation is that star formation in typical optical-break galaxies started earlier than in typical Lyman-break galaxies. Optical-break galaxies may be the oldest and most massive galaxies yet identified at z>2, and could evolve into early-type galaxies and bulges., Comment: Accepted for publication in the Astrophysical Journal Letters. (see also papers by Franx et al. and Daddi et al. in this listing)

We use very deep near-infrared photometry of the Hubble Deep Field South taken with ISAAC/VLT to identify a population of high redshift galaxies with rest- frame optical colors similar to those of nearby galaxies. The galaxies are chosen by their infrared colors Js-Ks > 2.3, aimed at selecting galaxies with redshifts above 2. When applied to our dataset, we find 14 galaxies with Ks < 22.5, corresponding to a surface density of 3+-0.8 /arcmin**2. The photometric redshifts all lie above 1.9, with a median of 2.6 and a rms of 0.7. The spectral energy distributions of these galaxies show a wide range: one is very blue in the rest-frame UV, and satisfies the normal Lyman-break criteria for high redshift, star-forming galaxies. Others are quite red throughout the observed spectral range, and are extremely faint in the optical, with a median V = 26.6. Hence these galaxies would not be included in photometric samples based on optical ground-based data, and spectroscopic follow-up is difficult. The spectral energy distributions often show a prominent break, identified as the Balmer break or 4000 Ang. break. The median age is 1 Gyr when fit with a constant star formation model with dust, or 0.7 Gyr when fit with a single burst model. Although significantly younger ages cannot be excluded when a larger range of models is allowed, the results indicate that these galaxies are among the oldest at these redshifts. The volume density to Ks=22.5 is half that of Lyman-break galaxies at z = 3. Since the mass-to-light ratios of the red galaxies are likely to be higher, the stellar mass density is inferred to be comparable to that of Lyman-break galaxies. These red galaxies may be the descendants of galaxies which started to form stars at very high redshifts, and they may evolve into the most massive galaxies at low redshift., Comment: LaTex, 5 pages and 4 figures. Accepted for publication in ApJ Letters. See also related preprints on astroph today by Daddi et al and van Dokkum et al

We present the first results from the Faint Infra-Red Extragalactic Survey (FIRES) of the Hubble Deep Field South. Using a combination of deep near infrared data obtained with ISAAC at the VLT with the WFPC2 HST data, we construct a K-band selected sample of 136 galaxies with K(AB)<=23.5. We derive zphot's using a newly developed technique which models the observed spectral energy distribution with a linear combination of empirical galaxy templates. Testing our zphot technique against spectroscopic redshifts in the HDF-N yields dz/(1+z)~0.07 for z<6. We show that we can derive realistic error estimates in zphot by combining the systematic uncertainties derived from the HDF-N with errors in zphot which depend on the observed flux errors. The redshift histogram of galaxies in the HDF-S shows distinct structure with a sharp peak at z~0.5 and a broad enhancement at z~1-1.4. We find that 12% of our galaxies with K(vega)<21 lie at z>=2. While this is higher than the fraction predicted in Omega_M=1 hierarchical models of galaxy formation we find that published predictions using pure luminosity evolution models produce too many bright galaxies at redshifts greater than unity. Finally, we use our broad wavelength coverage to measure the rest-frame UBV luminosities Lrest for z<=3. There is a paucity of galaxies brighter than Lrest_V>=1.4e10 h^{-2}Lsun at z~1.5-2, however, at z>2 we find very luminous galaxies with Lrest_V>=5e10 h^{-2}Lsun (for Omega_M=0.3,Lambda=0.7, H_o=100 h km s^{-1}Mpc^{-1}). Local B-band luminosity functions predict 0.1 galaxies in the redshift range 2<=z<=3.5 and with Lrest_B>= 5e10 h^{-2}Lsun,B but we find 9. The discrepancy can be explained if L*_B increases by a factor of 2.4-3.2 with respect to locally determined values (abridged)., Comment: 46 pages, 15 figures, 3 Tables, Accepted for publication in Astronomical Journal, paper and high resolution figures available at http://www.strw.leidenuniv.nl/~fires/, minor changes to comply with referee's comments

Can the submillimeter counts and background be produced by applying a locally derived extinction correction to the population of Lyman break galaxies? We investigate the submillimeter emission of two strongly lensed Lyman break galaxies (MS1512+36-cB58 and MS1358+62-G1) and find that the procedure that is used to predict the submillimeter emission of the Lyman break galaxy population overpredicts the observed 850micron fluxes by up to a factor of 14. This result calls for caution in applying local correlations to distant galaxies. It also shows that large extinction corrections on Lyman break galaxies should be viewed with skepticism. It is concluded that the Lyman break galaxies may contribute to the submillimeter background at the 25 to 50% level. The brighter submillimeter galaxies making up the rest of the background are either not detected in optical surveys, or if they are detected, their submillimeter emission cannot be reliably estimated from their rest-frame ultraviolet properties., Comment: 4 pages, 1 embedded postscript figure; to appear in proceedings of the UMass/INAOE conference on Deep Millimeter Surveys, eds. J. Lowenthal and D. Hughes; revised version corrects small numerical errors

We examine how the total X-ray luminosity correlates with stellar mass, stellar population, and redshift for a K-band limited sample of approximately 3500 galaxies at 0.5 < z < 2.0 from the NEWFIRM Medium Band Survey in the COSMOS field. The galaxy sample is divided into 32 different galaxy types, based on similarities between the spectral energy distributions. For each galaxy type, we further divide the sample into bins of redshift and stellar mass, and perform an X-ray stacking analysis using the Chandra COSMOS data. We find that full band X-ray luminosity is primarily increasing with stellar mass, and at similar mass and spectral type is higher at larger redshifts. When comparing at the same stellar mass, we find that the X-ray luminosity is slightly higher for younger galaxies (i.e., weaker 4000 angstrom breaks), but the scatter in this relation is large. We compare the observed X-ray luminosities to those expected from low- and high-mass X-ray binaries (XRBs). For blue galaxies, XRBs can almost fully account for the observed emission, while for older galaxies with larger 4000 angstrom breaks, active galactic nuclei (AGN) or hot gas dominate the measured X-ray flux. After correcting for XRBs, the X-ray luminosity is still slightly higher in younger galaxies, although this correlation is not significant. AGN appear to be a larger component of galaxy X-ray luminosity at earlier times, as the hardness ratio increases with redshift. Together with the slight increase in X-ray luminosity this may indicate more obscured AGNs or higher accretion rates at earlier times.

We investigate star formation rates of quiescent galaxies at high redshift (0.3 < z < 2.5) using 3D-HST WFC3 grism spectroscopy and Spitzer mid-infrared data. We select quiescent galaxies on the basis of the widely used UVJ color-color criteria. Spectral energy distribution fitting (rest frame optical and near-IR) indicates very low star formation rates for quiescent galaxies (sSFR approx. 10(exp −12)/yr. However, SED fitting can miss star formation if it is hidden behind high dust obscuration and ionizing radiation is re-emitted in the mid-infrared. It is therefore fundamental to measure the dust-obscured SFRs with a mid-IR indicator. We stack the MIPS-24 micron images of quiescent objects in five redshift bins centered on z = 0.5, 0.9, 1.2, 1.7, 2.2 and perform aperture photometry. Including direct 24 micron detections, we find sSFR approx. 10(exp −11.9) × (1 + z)(sup 4)/yr. These values are higher than those indicated by SED fitting, but at each redshift they are 20-40 times lower than those of typical star forming galaxies. The true SFRs of quiescent galaxies might be even lower, as we show that the mid-IR fluxes can be due to processes unrelated to ongoing star formation, such as cirrus dust heated by old stellar populations and circumstellar dust. Our measurements show that star formation quenching is very efficient at every redshift. The measured SFR values are at z > 1.5 marginally consistent with the ones expected from gas recycling (assuming that mass loss from evolved stars refuels star formation) and well above that at lower redshifts.

Quiescent galaxies at z approx. 2 have been identified in large numbers based on rest-frame colors, but only a small number of these galaxies have been spectroscopically confirmed to show that their rest-frame optical spectra show either strong Balmer or metal absorption lines. Here, we median stack the rest-frame optical spectra for 171 photometrically quiescent galaxies at 1.4 < z < 2.2 from the 3D-HST grism survey. In addition to H (4861 ),we unambiguously identify metal absorption lines in the stacked spectrum, including the G band (4304 ),Mgi (5175 ), and Na i (5894 ). This finding demonstrates that galaxies with relatively old stellar populations already existed when the universe was approx. 3 Gyr old, and that rest-frame color selection techniques can efficiently select them. We find an average age of 1.3+0.10.3 Gyr when fitting a simple stellar population to the entire stack. We confirm our previous result from medium-band photometry that the stellar age varies with the colors of quiescent galaxies: the reddest 80 of galaxies are dominated by metal lines and have a relatively old mean age of 1.6+0.50.4 Gyr, whereas the bluest (and brightest) galaxies have strong Balmer lines and a spectroscopic age of 0.9+0.20.1 Gyr. Although the spectrum is dominated by an evolved stellar population, we also find [O iii] and H emission. Interestingly, this emission is more centrally concentrated than the continuum with LOiii = 1.7+/- 0.3 x 10(exp 40) erg/s, indicating residual central star formation or nuclear activity.

The assembly of galaxies can be described by the distribution of their star formation as a function of cosmic time. Thanks to the WFC3 grism on the Hubble Space Telescope (HST) it is now possible to measure this beyond the local Universe. Here we present the spatial distribution of H emission for a sample of 54 strongly star-forming galaxies at z 1 in the 3D-HST Treasury survey. By stacking the H emission, we find that star formation occurred in approximately exponential distributions at z approximately 1, with a median Sersic index of n = 1.0 +/- 0.2. The stacks are elongated with median axis ratios of b/a = 0.58 +/- 0.09 in H consistent with (possibly thick) disks at random orientation angles. Keck spectra obtained for a subset of eight of the galaxies show clear evidence for rotation, with inclination corrected velocities of 90.330 km s(exp 1-). The most straightforward interpretation of our results is that star formation in strongly star-forming galaxies at z approximately 1 generally occurred in disks. The disks appear to be scaled-up versions of nearby spiral galaxies: they have EW(H alpha) at approximately 100 A out to the solar orbit and they have star formation surface densities above the threshold for driving galactic scale winds.