UC Santa Cruz

It is a grand challenge to develop a simple physical model that can faithfully emulate the results of high-resolution galaxy formation simulations and scale up their predictions to cosmologically representative volumes. Semi-analytic models (SAMs) are a promising means of tracking the physical processes associated with galaxy formation, but they are based on many approximations that have not been rigorously tested. In this dissertation, I demonstrate that SAMs can be re-tooled to roughly reproduce the overall baryon cycle of simulated galaxies.

I start by comparing predictions between the Santa Cruz SAM and the Feedback In Realistic Environments (FIRE) simulation suite. I show that both models predict remarkably similar stellar and interstellar medium (ISM) mass, but differ dramatically in terms of the underlying mass fluxes and circumgalactic medium (CGM) mass. The SAM predicts much higher gas accretion rates for dwarfs than FIRE-2, and compensates by requiring higher mass outflow rates. The SAM also predicts orders of magnitude lower CGM mass for dwarfs than FIRE-2. I argue that these model discrepancies are caused by the lack of preventative stellar feedback in the SAM and by its assumptions for halo gas cooling and recycling.

I then uniquely characterize the mass, momentum, energy and metallicity of multi-phase galactic winds in the same simulations. Among other results, I find that winds from dwarfs conserve nearly all of the available supernova energy, momentum and metal mass out to beyond the halo radius. I leverage these insights to prototype a new SAM in which supernova-driven winds can shock-heat gas within and beyond the CGM, thereby suppressing cooling and accretion. With this simple preventative feedback model, my SAM is able to simultaneously reproduce the bulk masses and mass fluxes of the simulated halos. Rapid inner halo recycling and the chemical enrichment of the CGM remain poorly understood, but extracting additional data from the simulations can pin down these and other uncertainties for SAMs. I discuss some observable consequences of my work and implications for the quest to develop a more complete, realistic and standard physical model of galaxy formation.