Toward the Formation of Realistic Galaxy Disks

In this review I demonstrate that a realistic model for the formation of galaxy disks depends on a proper treatment of the gas in galaxies. Historically, cosmological simulations of disk galaxy formation have suffered from a lack of resolution and a physically motivated feedback prescription. Recent computational progress has allowed for unprecedented resolution, which in turn allows for a more realistic treatment of feedback. These advances have led to a new examination of gas accretion, evolution, and loss in the formation of galaxy disks. Here I highlight the role that gas inflows, the regulation of gas by feedback, and gas outflows play in achieving simulated disk galaxies that better match observational results as a function of redshift. 1. The Role of Gas in the Formation of Disk Galaxies Stellar light provides us with most of our information on galaxy evolution over the history of the Universe. The ability to use this stellar light to trace the evolution of disk galaxies depends critically on our model of how gas gets into galaxies and when and where stars are subsequently allowed to form from that gas. Unfortunately, the evolution of gas within dark matter halos in our Cold Dark Matter (CDM) dominated Universe is complicated by our inability to accurately model complex gas processes that allow it to cool, to heat, and to be shocked, so that it is unlikely to follow the evolution of the dark matter. In the simplest analytic models for galaxy formation, the collapse of a galaxy is a two phase process, with the initial, rapid collapse of the dark matter shocking the gas to the virial temperature of the resulting halo (Lynden-Bell 1967). Any subsequently accreted gas is then also shock heated to the virial temperature of the halo. The second phase begins after halo collapse, with the baryons being able to cool and condense to the center of the dark matter halo after adopting some density profile (e.g., White & Frenk 1991; Kauffmann et al. 1993; Cole et al. 2000; Somerville et al. 2001). The angular momentum of the gas and dark matter is thought to be accumulated by tidal torques exerted by neighboring structures (e.g., White 1984; Barnes & Efstathiou 1987), (though see Maller et al. 2002; Vitvitska et al. 2002). Because mergers, supernova (SN) feedback, and gas accretion can be selfconsistently modeled in simulations, the results from simulations can be used to test predictions from analytic models and to fine tune the physical prescriptions that are used by modelers. However, before simulation results can be used to better model galaxy formation and evolution, simulations must be tested against existing scaling relations simultaneously to establish robust constraints. Historically, matching these relations has not been a trivial task. In particular,