Wednesday, November 12, 2014

Quenching Star Formation in Cluster Galaxies

Left panel: the velocity vs. clustercentric radius phase space of galaxies in the nine GCLASS clusters. The velocities are in units relative to the individual cluster velocity dispersions and the radii are relative to the position of the brightest cluster galaxy scaled by the R200 of the cluster. The shaded regions are arbitrarily defined but are indicative of increasing time since infall (see text). Quiescent galaxies (red triangles), star forming galaxies (blue triangles), and poststarburst galaxies (green stars) all occupy distinct locations in phase space. Right panels: the ratio of quiescent and poststarburst galaxies compared to star-forming galaxies separated into the three radial bins marked by the dotted lines (top panel), and the three phase space bins marked by the shaded regions (bottom panel). The error bars are 1σ Poisson errors. Poststarburst galaxies are distributed fairly uniformly in the cluster by radius (top panel), with a peak in the middle bin; however, in phase space they are most prevalent in the middle bin and completely absent in the inner bin (bottom panel).

Result in a Nutshell: Understanding the behaviors of galaxies in clusters is a large and complex problem that has not daunted Adam Muzzin of the Leiden Observatory at Leiden University in The Netherlands. Muzzin led an international team using data from the Gemini Cluster Astrophysics Spectroscopic Survey (GCLASS) in order to explore galaxies that have recently stopped (quenched) the formation of stars. Their findings reveal that these galaxies are very different from other cluster galaxies, and for the first time show that these quenched galaxies tend to be closer to the cluster’s center and moving especially fast. As a critical part of their results, the team established unprecedented constraints on how long this quenching takes, and where it happens. It’s quick, by astrophysical timescales – between 100-500 million years, and happens roughly halfway out from the center of the cluster. 

 
For Scientists:

The paper is accepted for publication in the The Astrophysical Journal and can be accessed at http://iopscience.iop.org/0004-637X/796/1/65/ (subscription required) or at astro-ph

Scientific Abstract (from the paper):

We investigate the velocity versus position phase space of z ∼ 1 cluster galaxies using a set of 424 spectroscopic redshifts in nine clusters drawn from the GCLASS survey. Dividing the galaxy population into three categories: quiescent, star-forming, and poststarburst, we find that these populations have distinct distributions in phase space. Most striking are the poststarburst galaxies, which are commonly found at small clustercentric radii with high clustercentric velocities, and appear to trace a coherent “ring” in phase space. 

Using several zoom simulations of clusters we show that the coherent distribution of the poststarbursts can be reasonably well-reproduced using a simple quenching scenario. Specifically, the phase space is best reproduced if these galaxies are quenched with a rapid timescale (0.1 < τQ < 0.5 Gyr) after they make their first passage of R ∼ 0.5 R200 , a process that takes a total time of ∼ 1 Gyr after first infall. The poststarburst phase space is not well-reproduced using long quenching timescales (τQ > 0.5 Gyr), or by quenching galaxies at larger radii (R∼R200 ).We compare this quenching timescale to the timescale implied by the stellar populations of the poststarburst galaxies and find that the poststarburst spectra are well-fit by a rapid quenching (τQ = 0.4+0.3−0.4 Gyr) of a typical star-forming galaxy. The similarity between the quenching timescales derived from these independent indicators is a strong consistency check of the quenching model. Given that the model implies satellite quenching is rapid, and occurs well within R200 , this would suggest that ram-pressure stripping of either the hot or cold gas component of galaxies are the most plausible candidates for the physical mechanism. The high cold gas consumption rates at z ∼ 1 make it difficult to determine if hot or cold gas stripping is dominant; however, measurements of the redshift evolution of the satellite quenching timescale and location may be capable of distinguishing between the two.