Ram pressure stripping in the EAGLE simulation

Ram pressure stripping of satellite galaxies is thought to be a ubiquitous process in galaxy clusters, and a growing number of observations reveal satellites at different stages of stripping. However, in order to determine the fate of any individual galaxy, we turn to predictions from either simulations or analytic models. It is not well-determined whether simulations and analytic models agree in their predictions, nor the causes of disagreement. Here we investigate ram pressure stripping in the reference EAGLE hydrodynamical cosmological simulation, and compare the results to predictions from analytic models. We track the evolution of galaxies with stellar mass $M_{*} > 10^{9} \, \mathrm{M_{\odot}}$ and initial bound gas mass $M_{\mathrm{gas}} > 10^{9} \, \mathrm{M_{\odot}}$ that fall into galaxy clusters ($M_{\mathrm{200c}} > 10^{14} \, \mathrm{M_{\odot}}$) between $z = 0.27$ and $z = 0$. We divide each galaxy into its neutral gas disk and hot ionized gas halo and compare the evolution of the stripped gas fraction in the simulation to that predicted by analytic formulations for the two gas phases, as well as to a toy model that computes the motions of gas particles under the combined effects of gravity and a spatially uniform ram pressure. We find that the analytic models generally underpredict the stripping rate of neutral gas and overpredict that of ionized gas, with significant scatter between the model and simulation stripping timescales. This is due to opposing physical effects: the enhancement of ram pressure stripping by stellar feedback, and the suppression of stripping by the compaction of galactic gas.