The Evolving Structure and Dynamics of Quiescent Galaxies Since z~2
Sample spectra of quiescent galaxies at z=1-1.6 (Newman et al 2010)
Advances in infrared instrumentation have enabled major progress in the study of red galaxies at z > 1 over the last decade. The interval of roughly z=1–2.5 emerged as a key period when the abundance of quiescent galaxies increases rapidly. At the same time, these galaxies continually grow in size after the truncation of star formation — by a factor of 3-4 since z = 2 — and morphologies may change as well. My research involves several approaches to understand the build-up of the red sequence during this epoch.
At the same time as old galaxies are growing in size, galaxies are being quenched and added to the red sequence. Disentangling the evolution in scaling relations (such as mass–radius) arising from these channels is a key challenge. The stellar velocity dispersion σ offers a useful "tag," since it is thought to be relatively stable under mergers, and it also yields measures of the dynamical mass that form an important check on stellar population synthesis models at high redshifts. In Newman et al 2010, we used the enhanced red sensitivity of fully-depleted CCDs installed in Keck/LRIS to measure σ in 17 quiescent galaxies at z=1–1.6, the largest sample to that date. This survey was expanded to 56 systems in Belli, Newman & Ellis (2014a). This work provided a dynamical confirmation of the rate of size growth inferred photometrically and pointed to a relative change of size and mass (d log R/d log M) consistent with simulations of minor mergers. The spectra also provide age-sensitive absorption features that can be used to identify "new arrivals" on the red sequence and directly assess their contribution to the growth rate (Belli, Newman & Ellis 2014b, in preparation).
Montage of quiescent galaxies at z=0.4–2.5 as seen by WFC3/HST. Redshift increases toward the bottom and the physical size of the cutout boxes is constant.
The leading candidate for the source of the size growth seen in "dead" galaxies is the accretion of stars in low-mass satellites. It is not clear, however, that the rate of these "minor" mergers is enough to explain the rate of size growth observed. In Newman et al 2012 we used data from the CANDELS survey using the WFC3 camera onboard Hubble to provide a benchmark measurement of the size growth rate using a uniformly-selected, large sample of galaxies at z=0.4–2 with HST rest-optical images. The sample size allowed us to characterize the distribution of galaxy sizes, and thus the decline in the abundance of the most compact galaxies over time, which can only be attributed to their physical growth. We then searched for satellite galaxies around the same systems in order to estimate the rate of size growth that would result from their accretion, based on merger simulations. We found that the rate of size growth at z < 1 could possibly be explained by mergers of observed close galaxy pairs. At higher redshifts, however, it is much more difficult to account for the more rapid rate of size growth that is observed.
Measurements of size growth in different environments may help disentangle its physical origins, since merger histories are expected to vary with environment. I am also interested in galaxy evolution in the earliest clusters. Through a 17 orbit allocation in Cycle 20 (P.I. ABN), we used HST/WFC3 imaging and grism spectroscopy to study the galaxy population in JKCS041 at z=1.80, one of the most distant clusters known (Newman et al 2014). The grism spectra have sufficient signal to measure absorption lines in detail — a first at this redshift — which were used to determine membership and constrain the stellar populations of the massive cluster galaxies. Comparing the ages, sizes, and morphologies of the members with coeval counterparts in the field, we found minimal differences in age and size, with a hint of fewer passive disks in the cluster. This paper also introduced a new Bayesian code to infer stellar population properties from both broadband photometry and high-resolution spectra, which is being used in other ongoing programs.
Collaborators: Sirio Belli, Richard Ellis, Tommaso Treu, Kevin Bundy, Carlo Nipoti, Stefano Andreon, Anand Raichoor, Ginevra Trinchieri, Nick Konidaris
The Dark Matter Density Profile on Cluster Scales
Four views of a galaxy cluster at z=0.19 on different scales, from left to right: Weak lensing reconstruction of the mass distribution (image ~5 Mpc across), contours of X-ray emission overlaid on an optical image (~1 Mpc across), view of strongly-lensed background galaxies in the cluster core (~150 kpc across), and spatially-resolved stellar kinematics within the central giant galaxy.
A key feature of cold dark matter (CDM) halos is the presence of a central density cusp. Observations have challenged the reality of a suitably steep density profile in systems ranging from dwarf galaxies to galaxy clusters. Clusters are excellent locations to map dark matter halos in detail: they are dark matter-dominated outside the very central regions, and they provide several precise observational probes. In a series of papers with Tommaso Treu, Richard Ellis, and Dave Sand, we combined strong and weak gravitational lensing with X-ray observations and resolved stellar kinematics of the brightest cluster galaxy (BCG) to measure precise mass profiles over scales of ~3 kpc – 3 Mpc in a sample of 7 relaxed clusters, providing the detailed constraints on the small-scale density profile and disentangling the dark and baryonic contributions.
This work showed that the inner dark matter profile is shallower than expected in a halo composed of collisionless CDM. However, the dark matter distribution appears to be correlated with the properties of the BCG such that the total density profile (baryonic and dark matter) nearly matches the universal CDM profile. This suggests a connection between dark matter in cluster cores and the assembly of stars in the BCG, and provides important input into simulations of clusters, which have not yet converged on the expected effect of baryons on the small-scale dark matter profile. This work also poses puzzles for models that account for shallow cusps or cores as a result of dark matter particle physics.
Finally, matching the stellar mass-to-light ratio of the BCGs, measured from lensing and dynamics, to estimates from stellar population synthesis models requires a Salpeter (or equivalently "heavy") initial mass function, in agreement with other independent probes in massive ellipticals (e.g., Conroy & van Dokkum 2012; Cappellari et al 2012, Treu et al 2010).
Collaborators: Tommaso Treu, Richard Ellis, Dave Sand, Carlo Nipoti, Johan Richard, Eric Jullo