A study of galaxy formation across cosmic time
from cosmological hydrodynamical simulations

Abstract

The evolution of galaxies across cosmic time, from the first galaxies to the local Universe, are studied using cosmological hydrodynamical simulations. It is demonstrated
that, for the first time, a hydrodynamical simulation can reproduce the observed evolution of galaxy stellar masses and the trends in star formation rates. The success of the simulation in producing galaxies with similar histories to those observed increases the potential for using hydrodynamical simulations to explore galaxy formation physics. With this intention, we consider the effects of the environment and active galactic nuclei (AGN) quenching on the galaxy population and the shape of the galaxy stellar mass function (GSMF). We find environmental processes are effective at quenching galaxies in the simulation and operated on short timescales. AGN feedback, which produces a large passive fraction at high stellar
masses, drives the exponential break in the GSMF.

Specific star formation rates (SFRs) in simulations, both hydrodynamical and semi-analytical, have been shown to be discrepant with observations. We investigate proposed solutions to this problem using a suite of cosmological hydrodynamical simulations. The offset in the simulations, at the level of 0.2 to 0.4 dex, can only be resolved by employing an extreme model that does not recover any of
the observed trends in stellar mass or the cosmic star formation rate density. This study implies that the observed star formation rates across comic time are inconsistent with the growth of stellar mass.

Two galaxy populations are then explored in more detail. We examine the first galaxies and their potential to reionize the Universe. We find that low-mass galaxies, undergoing extreme star formation for their stellar mass, produce the majority of ionizing photons at redshifts 6 and above. The second study considers the most highly star forming galaxies in the simulation, which represent an extreme population. These galaxies have similar stellar and gas masses to the observed sub-mm population, however, their SFRs are lower. On further investigation, the selection of galaxies based on SFRs is not adequate to compare the
simulation to the sub-mm population, in particular at high redshift. These results highlight the importance of dust temperature in the selection.