We present the first detailed study of the properties (temperatures, gravities, and masses) of the NGC 6791 white dwarf population. This unique stellar system is both one of the oldest (8 Gyr) and most metal-rich ([Fe/H] similar to +0.4) open clusters in our Galaxy and has a color-magnitude diagram (CMD) that exhibits both a red giant clump and a much hotter extreme horizontal branch. Fitting the Balmer lines of the white dwarfs in the cluster using Keck/LRIS spectra suggests that most of these stars are undermassive, < M > = 0.43 +/- 0.06 M-circle dot, and therefore could not have formed from canonical stellar evolution involving the helium flash at the tip of the red giant branch. We show that at least 40% of NGC 6791's evolved stars must have lost enough mass on the red giant branch to avoid the flash and therefore did not convert helium into carbon-oxygen in their core. Such increased mass loss in the evolution of the progenitors of these stars is consistent with the presence of the extreme horizontal branch in the CMD. This unique stellar evolutionary channel also naturally explains the recent finding of a very young age (2.4 Gyr) for NGC 6791 from white dwarf cooling theory; helium-core white dwarfs in this cluster will cool similar to 3 times slower than carbon-oxygen-core stars, and therefore the corrected white dwarf cooling age is in fact greater than or similar to 7 Gyr, consistent with the well-measured main-sequence turnoff age. These results provide direct empirical evidence that mass loss is much more efficient in high-metallicity environments and therefore may be critical in interpreting the ultraviolet upturn in elliptical galaxies.

The initial-final mass relation represents a mapping between the mass of a white dwarf remnant and the mass that the hydrogen-burning main-sequence star that created it once had. The empirical relation thus far has been constrained using a sample of similar to 40 stars in young open clusters, ranging in initial mass from similar to 2.75 to 7 M-circle dot, and shows a general trend that connects higher mass main-sequence stars with higher mass white dwarfs. In this paper, we present CFHT CFH12K photometric and Keck LRIS multiobject spectroscopic observations of a sample of 22 white dwarfs in two older open clusters, NGC 7789 (t = 1.4 Gyr) and NGC 6819 (t = 2.5 Gyr). We measure masses for the highest signal-to-noise ratio spectra by fitting the Balmer lines to atmosphere models and place the first direct constraints on the low-mass end of the initial-final mass relation. Our results indicate that the observed general trend at higher masses continues down to low masses, with M-initial = 1.6 M-circle dot main-sequence stars forming M-final = 0.54 M-circle dot white dwarfs. When added to our new data from the very old cluster NGC 6791, the relation is extended down to M-initial = 1.16 M-circle dot (corresponding to M-final = 0.53 M-circle dot). This extension of the relation represents a fourfold increase in the total number of hydrogen-burning stars for which the integrated mass loss can now be calculated from empirical data, assuming a Salpeter initial mass function. The new leverage at the low-mass end is used to derive a purely empirical initial-final mass relation. The sample of white dwarfs in these clusters also shows several interesting systems that we discuss further: a DB (helium) white dwarf, a magnetic white dwarf, a DAB (mixed hydrogen/helium atmosphere or a double degenerate DA+DB) white dwarf(s), and two possible equal-mass DA double degenerate binary systems.

To determine the relative contributions of galactic and intracluster stars to the enrichment of the intracluster medium (ICM), we present X-ray surface brightness, temperature, and Fe abundance profiles for a set of 12 galaxy clusters(4) for which we have extensive optical photometry. Assuming a standard initial mass function and simple chemical evolution model scaled to match the present-day cluster early-type SN Ia rate, the stars in the brightest cluster galaxy (BCG) plus the intracluster stars (ICS) generate 31(-9)(+11)%, on average, of the observed ICM Fe within r(500) (similar to 0.6 times r(200), the virial radius). An alternate, two-component SN Ia model (including both prompt and delayed detonations) produces a similar BCG+ ICS contribution of 22(-9)(+9)%. Because the ICS typically contribute 80% of the BCG+ ICS Fe, we conclude that the ICS are significant, yet often neglected, contributors to the ICM Fe within r(500). However, the BCG+ICS fall short of producing all the Fe, so metal loss from stars in other cluster galaxies must also contribute. By combining the enrichment from intracluster and galactic stars, we can account for all the observed Fe. These models require a galactic metalloss fraction (0.84+(+0.11)(-0.14)) that, while large, is consistent with the metal mass not retained by galactic stars. The SN Ia rates, especially as a function of galaxy environment and redshift, remain a significant source of uncertainty in further constraining the metal-loss fraction. For example, increasing the SN Ia rate by a factor of 1.8-to just within the 2 sigma uncertainty for present-day cluster early-type galaxies-allows the combined BCG + ICS + cluster galaxy model to generate all the ICM Fe with a much lower galactic metal-loss fraction (similar to 0.35).

We have compiled a sample of early-type cluster galaxies from 0 < z < 1.3 and measured the evolution of their ellipticity distributions. Our sample contains 487 galaxies in 17 z > 0.3 clusters with high-quality space-based imaging and a comparable sample of 210 galaxies in 10 clusters at z < 0.05. We select early-type galaxies (elliptical and S0 galaxies) that fall within the cluster R-200, and which lie on the red-sequence in the magnitude range -19.3 > M-B > -21, after correcting for luminosity evolution as measured by the fundamental plane. Our ellipticity measurements are made in a consistent manner over our whole sample. We perform extensive simulations to quantify the systematic and statistical errors, and find that it is crucial to use point-spread function (PSF)-corrected model fits; determinations of the ellipticity from Hubble Space Telescope image data that do not account for the PSF "blurring" are systematically and significantly biased to rounder ellipticities at redshifts z > 0.3. We find that neither the median ellipticity, nor the shape of the ellipticity distribution of cluster early-type galaxies evolves with redshift from z similar to 0 to z > 1 (i.e., over the last similar to 8 Gyr). The median ellipticity at z > 0.3 is statistically identical with that at z < 0.05, being higher by only 0.01 +/- 0.02 or 3 +/- 6%, while the distribution of ellipticities at z > 0.3 agrees with the shape of the z < 0.05 distribution at the 1-2% level (i.e., the probability that they are drawn from the same distribution is 98-99%). These results are strongly suggestive of an unchanging overall bulge-to-disk ratio distribution for cluster early-type galaxies over the last similar to 8 Gyr from z similar to 1 to z similar to 0. This result contrasts with that from visual classifications which show that the fraction of morphologically-selected disk-dominated early-type galaxies, or S0s, is significantly lower at z > 0.4 than at z similar to 0. We find that the median disk-dominated early-type, or S0, galaxy has a somewhat higher ellipticity at z > 0.3, suggesting that rounder S0s are being assigned as ellipticals. Taking the ellipticity measurements and assuming, as in all previous studies, that the intrinsic ellipticity distribution of both elliptical and S0 galaxies remains constant, then we conclude from the lack of evolution in the observed early-type ellipticity distribution that the relative fractions of ellipticals and S0s do not evolve from z similar to 1 to z = 0 for a red-sequence selected samples of galaxies in the cores of clusters of galaxies.

We use extensive new observations of the very rich z similar to 0.4 cluster of galaxies A851 to examine the nature and origin of starburst galaxies in intermediate-redshift clusters. New HST observations, 24 mu m Spitzer photometry and ground-based spectroscopy cover most of a region of the cluster about 10' across, corresponding to a cluster-centric radial distance of about 1.6 Mpc. This spatial coverage allows us to confirm the existence of a morphology-density relation within this cluster, and to identify several large, presumably infalling, subsystems. We confirm our previous conclusion that a very large fraction of the star-forming galaxies in A851 has recently undergone starbursts. We argue that starbursts are mostly confined to two kinds of sites: infalling groups and the cluster center. At the cluster center, it appears that infalling galaxies are undergoing major mergers, resulting in starbursts whose optical-emission lines are completely buried beneath dust. The aftermath of this process appears to be proto-S0 galaxies devoid of star formation. In contrast, major mergers do not appear to be the cause of most of the starbursts in infalling groups, and fewer of these events result in the transformation of the galaxy into an S0. Some recent theoretical work provides possible explanations for these two distinct processes, but it is not clear whether they can operate with the very high efficiency needed to account for the very large starburst rate observed.

We present the first results from the largest spectroscopic survey to date of an intermediate redshift galaxy cluster, the z = 0.834 cluster RX J0152.7-1357. We use the colors of galaxies, assembled from a D similar to 12 Mpc region centered on the cluster, to investigate the properties of the red sequence as a function of density and clustercentric radius. Our wide-field multislit survey with a low-dispersion prism in the Inamori Magellan Areal Camera and Spectrograph at the 6.5 m Baade telescope allowed us to identify 475 new members of the cluster and its surrounding large-scale structure with a redshift accuracy of sigma(z)/(1 + z) approximate to 1% and a contamination rate of similar to 2% for galaxies with i' < 23.75 mag. We combine these new members with the 279 previously known spectroscopic members to give a total of 754 galaxies from which we obtain a mass-limited sample of 300 galaxies with stellar masses M > 4 x 10(10) M-circle dot (log M/M-circle dot > 10.6). We find that the red galaxy fraction is 93 +/- 3% in the two merging cores of the cluster and declines to a level of 64 +/- 3% at projected clustercentric radii R greater than or similar to 3 Mpc. At these large projected distances, the correlation between clustercentric radius and local density is nonexistent. This allows an assessment of the influence of the local environment on galaxy evolution, as opposed to mechanisms that operate on cluster scales (e.g., harassment, ram-pressure stripping). Even beyond R > 3 Mpc we find an increasing fraction of red galaxies with increasing local density. The red galaxy fraction at the highest local densities in two large groups at R > 3 Mpc matches the red galaxy fraction found in the two cores. Strikingly, galaxies at intermediate densities at R > 3 Mpc, that are not obvious members of groups, also show signs of an enhanced red galaxy fraction. Our results point to such intermediate-density regions and the groups in the outskirts of the cluster, as sites where the local environment influences the transition of galaxies onto the red sequence.

We examine the star formation rates (SFRs) of galaxies in a redshift slice encompassing the z = 0.834 cluster RX J0152.7-1357. We used a low-dispersion prism in the Inamori Magellan Areal Camera and Spectrograph to identify galaxies with z(AB) < 23.3 mag in diverse environments around the cluster out to projected distances of similar to 8 Mpc from the cluster center. We utilize a mass-limited sample (M > 2 x 10(10) M(circle dot)) of 330 galaxies that were imaged by Spitzer MIPS at 24 mu m to derive SFRs and study the dependence of specific SFR (SSFR) on stellar mass and environment. We find that the SFR and SSFR show a strong decrease with increasing local density, similar to the relation at z similar to 0. Our result contrasts with other work at z similar to 1 that finds the SFR-density trend to reverse for luminosity-limited samples. These other results appear to be driven by star formation (SF) in lower mass systems (M similar to 10(10) M(circle dot)). Our results imply that the processes that shut down SF are present in groups and other dense regions in the field. Our data also suggest that the lower SFRs of galaxies in higher density environments may reflect a change in the ratio of star-forming to non-star-forming galaxies, rather than a change in SFRs. As a consequence, the SFRs of star-forming galaxies, in environments ranging from small groups to clusters, appear to be similar and largely unaffected by the local processes that truncate SF at z similar to 0.8.

Modern population synthesis models estimate that 50% of the rest-frame K-band light is produced by thermally pulsing asymptotic giant branch (TP-AGB) stars during the first Gyr of a stellar population, with a substantial fraction continuing to be produced by the TP-AGB over a Hubble time. Between 0.2 and 1.5 Gyr, intermediate-mass stars evolve into TP-AGB C stars which, due to significant amounts of circumstellar dust, emit half their energy in the mid-IR. We combine these results using published mid-IR colors of Galactic TP-AGB M and C stars to construct simple models for exploring the contribution of the TP-AGB to 24 mu m data as a function of stellar population age. We compare these empirical models with an ensemble of galaxies in the Chandra Deep Field South from z = 0 to z = 2, and with high-quality imaging in M81. Within the uncertainties, the TP-AGB appears responsible for a substantial fraction of the mid-IR luminosities of galaxies from z = 0 to z = 2, the maximum redshift to which we can test our hypothesis, while, at the same time, our models reproduce much of the detailed structure observed in mid-IR imaging of M81. The mid-IR is a good diagnostic of star formation over timescales of similar to 1.5 Gyr, but this implies that ongoing star formation rates at z = 1 may be overestimated by factors of similar to 1.5-6, depending on the nature of star formation events. Our results, if confirmed through subsequent work, have strong implications for the star formation rate density of the universe and the growth of stellar mass over time.

We have measured velocity dispersions (sigma) for a sample of 36 galaxies with J < 21.2 or M-r < -20.6 mag in MS 1054-03, a massive cluster of galaxies at z = 0.83. Our data are of uniformly high quality down to our selection limit, our 16 hr exposures typically yielding errors of only delta(sigma) similar to 10% for L* and fainter galaxies. By combining our measurements with data from the literature, we have 53 cluster galaxies with measured dispersions, and HST/ACS-derived sizes, colors and surface brightness. This sample is complete for the typical L-star galaxy at z similar to 1, unlike most previous z similar to 1 cluster samples which are complete only for the massive cluster members (>10(11) M-circle dot). We find no evidence for a change in the tilt of the fundamental plane (FP). Nor do we find evidence for evolution in the slope of the color-sigma relation and M/L-B-sigma relations; measuring evolution at a fixed sigma should minimize the impact of structural evolution found in other work. The M/L-B at fixed sigma evolves by Delta log(10) M/L-B = -0.50 +/- 0.03 between z = 0.83 and z = 0.02 or d log(10) M/L-B = -0.60 +/- 0.04 dz, and we find Delta(U-V)(z) = -0.24 +/- 0.02 mag at fixed sigma in the rest frame, matching the expected evolution in M/LB within 2.25 standard deviations. The implied formation redshift from both the color and M/L-B evolution is z(star) = 2.0+/-0.2+/-0.3(sys), during the epoch in which the cosmic star formation activity peaked, with the systematic uncertainty showing the dependence of z(star) on the assumptions we make about the stellar populations. The lack of evolution in either the tilt of the FP or in the M/L-sigma and color-sigma relations imply that the formation epoch depends weakly on mass, ranging from z(star) = 2.3(-0.3)(+1.3) at sigma = 300 km s(-1) to z(star) = 1.7(-0.2)(+0.3) at sigma = 160 km s(-1) and implies that the initial mass function similarly varies slowly with galaxy mass.