The History of Enrichment of the Intergalactic Medium Using Cosmological Simulations

Abstract

I investigate the chemical evolution of the Universe in a series of cosmological hydrodynamic simulations with the purpose of finding a self-consistent evolutionary scenario of galaxy formation able to reproduce key observables focusing on the enrichment of the intergalactic medium (IGM). The most successful models I run and analyze use the scalings of momentum-driven feedback whereby UV photons generated during the Main Sequence stage accelerate dust-driven winds while providing a significantly larger energy budget than from supernovae alone. The success of this outflow model relies on its ability to drive highly mass-loaded winds from small galactic haloes. These feedback relations, supported by observations of local starburst, are inserted into simulations at all epochs, reproducing observables including the C IV column density and linewidth distributions at z=6->1.5 and the O VI forest at z=0-0.5. Outflows at z>=5 propagate early nucleosynthetic products traced by C IV and lower ionization species into an otherwise metal-free IGM. Continual outflows at the peak ages of star formation (z=5->1.5) produce a non-evolving cosmic mass density of C IV despite continual enrichment increasing IGM metallicity by a factor of ten. The z=0-0.5 O VI forest is composed of weaker absorbers tracing photo-ionized diffuse IGM metals, sometimes injected by primordial galaxies, and stronger absorbers tracing recently injected metals, often unable to escape their parent haloes and sometimes collisionally ionized. Tracking the individual histories of metals in outflows shows the average outflow travels ~100 physical kpc and returns to galaxies on an average timescale of 1-2 Gyr; this result implies metals in superwinds do not remain in the IGM for a Hubble time and are more likely to rejoin galaxies. Metal absorbers aligned with Lyman-alpha are examined in detail, finding that the two often trace different phases of gas with the former tracing an inhomogeneous distribution of metals exhibiting turbulence imparted during the outflow phase dissipating on a Hubble timescale. I find this is the first model to self-consistently reproduce the wide range of IGM observables spanning the history of heavy metal production while being consistent with key galaxy observables. The link between star formation and galactic superwinds requires that a successful model of galaxy formation reproduces both the evolution of galaxies and the IGM.