| This thesis investigates the consequences for large-scale structure statistics of including realistic models for the formation of galaxies in simulations of cosmological structure formation. Recent improvements in computing power and computational algorithms has made possible realistic simulations of the interaction between dark matter, the dominant gravitational force in the universe, and baryonic gas, the component which can form stars and which we can directly observe. Here, I use such simulations to investigate the large-scale distribution of galaxies in the universe and how it compares to the distribution of mass. Since observations of the clustering properties of galaxies are often used to constrain the mass distribution, the relationship between galaxies and dark matter is a fundamental, yet poorly understood, element in the interpretation of galaxy surveys. The simulations here suggest that this relationship is more complicated than previously believed; the density of galaxies in the universe is a complicated function of mass density, gas temperature, galaxy type, and time. In particular, the physics and gas cooling appears to play a crucial role in determining where galaxies form, and when. I test predictions of these simulations using the Las Campanas Redshift Survey (LCRS), which has mapped the location of about 25,000 galaxies out to several billion light-years in distance, in a narrow band on the sky. Future observations will be able to place stronger constraints on this and other models of galaxy formation. |
