Abstract

We model satellite quenching at z similar to 1 by combining 14 massive (10(13.8) < M-halo/M-circle dot < 10(15)) clusters at 0.8 < z < 1.3 from the GOGREEN and GCLASS surveys with accretion histories of 56 redshift-matched analogues from the IllustrisTNG simulation. Our fiducial model, which is parametrized by the satellite quenching time-scale (tau(quench)), accounts for quenching in our simulated satellite population both at the time of infall by using the observed coeval field quenched fraction and after infall by tuning tau(quench) to reproduce the observed satellite quenched fraction versus stellar mass trend. This model successfully reproduces the observed satellite quenched fraction as a function of stellar mass (by construction), projected cluster-centric radius, and redshift and is consistent with the observed field and cluster stellar mass functions at z similar to 1. We find that the satellite quenching time-scale is mass dependent, in conflict with some previous studies at low and intermediate redshift. Over the stellar mass range probed (M-star > 10(10) M-circle dot), we find that the satellite quenching time-scale decreases with increasing satellite stellar mass from similar to 1.6 Gyr at 10(10) M-circle dot to similar to 0.6-1 Gyr at 10(11) M-circle dot and is roughly consistent with the total cold gas (HI + H-2) depletion time-scales at intermediate z, suggesting that starvation may be the dominant driver of environmental quenching at z < 2. Finally, while environmental mechanisms are relatively efficient at quenching massive satellites, we find that the majority (similar to 65-80 per cent) of ultra-massive satellites (M-star > 10(11) M-circle dot) are quenched prior to infall.