January 23

Michael Kuhlen (UC Berkeley)

Computational Cosmology and Galaxy Formation

Fueled by continuing advances in numerical methods and computational capabilities, the future of galaxy formation theory is going to be driven by numerical simulations. Yet computational galaxy formation is extremely challenging, owing to the multitude of important physical processes and the wide range of scales over which they operate. Much of the galaxy formation simulation work to date has relied on simple, and often ad-hoc, subgrid models for star formation and feedback. In this talk I will describe my recent efforts to improve this situation by including more realistic and physics-driven treatments of some of the relevant processes. As one example, I will discuss cosmological adaptive mesh refinement simulations in which star formation is regulated by the local abundance of molecular hydrogen. These simulations reproduce much of the observational phenomenology of star formation rates as a function of atomic and molecular gas content and metallicity. At the same time this new piece of physics leads to a suppression of the stellar content of low mass dark matter halos, thereby helping to explain the vexing dwarf galaxy problem.