The colors of satellite galaxies in the Illustris Simulation

TL;DR

This study uses the Illustris simulation to successfully reproduce observed trends in satellite galaxy colors and distributions, highlighting the importance of gas content at infall for star formation and quenching timescales.

Contribution

First to demonstrate that the Illustris simulation can accurately match observed satellite galaxy properties, emphasizing the role of gas content at infall.

Findings

Good agreement between simulation and observations of satellite properties.

Large gas reservoirs at infall enable prolonged star formation in satellites.

Quenching timescales are ~2 Gyr in groups and >5 Gyr around Milky Way-like primaries.

Abstract

Observationally, the fraction of blue satellite galaxies decreases steeply with host halo mass, and their radial distribution around central galaxies is significantly shallower in massive (M_* >10e11M_sun) than in Milky Way like systems. Theoretical models, based primarily on semi-analytical techniques, have had a long-standing problem with reproducing these trends, instead predicting too few blue satellites in general but also estimating a radial distribution that is too shallow, regardless of primary mass. In this Letter, we use the Illustris cosmological simulation to study the properties of satellite galaxies around isolated primaries. For the first time, we find good agreement between theory and observations. We identify the main source of this success relative to earlier work to be a consequence of the large gas contents of satellites at infall, a factor ~5-10 times larger than in semi-analytical models. Because of their relatively large gas reservoirs, satellites can continue to form stars long after infall, with a typical timescale for star-formation to be quenched ~2 Gyr in groups but more than ~5 Gyr for satellites around Milky Way like primaries. The gas contents we infer are consistent with z=0 observations of HI gas in galaxies, although we find large discrepancies among reported values in the literature. A testable prediction of our model is that the gas-to-stellar mass ratio of satellite progenitors should vary only weakly with cosmic time.