Constraining galaxy bias and cosmology using galaxy clustering data

Galaxy bias, which is the difference in distributions of galaxies and the underlying dark matter, has long been regarded as an obstacle to deriving cosmological parameters with galaxy clustering data from galaxy redshift surveys. On the other hand, the existence of galaxy bias can be viewed as a great advantage to test theories of galaxy formation. The observational side of my study is to constrain galaxy bias from observed galaxy clustering data and make comparisons with predictions from galaxy formation models. On the theoretical side, I perform detailed investigations to see whether we can overcome the obstacle so that we simultaneously constrain cosmology and infer galaxy bias from galaxy clustering data.

Galaxy bias can be completely described at the level of individual dark matter halos in the framework of the Halo Occupation Distribution (HOD). I apply HOD modeling to galaxy clustering data in galaxy redshift surveys, especially, in the Sloan Digital Sky Survey (SDSS). I show that the observed departures from a power law in the galaxy two-point correlation function can be physically explained within the HOD framework. With HOD modeling, I interpret the luminosity and color dependence of galaxy clustering in the SDSS and find that there are general agreements between HODs inferred from the observation and those predicted by galaxy formation models.

Armed with the HOD framework, I investigate the possibility of deriving cosmology and inferring galaxy bias simultaneously using galaxy clustering data. I first study the influence of cosmological parameters on statistical properties of halo population using N-body simulations. My study shows that halo populations produced by distinct cosmological models can always be told apart by choosing the right combination of statistics. I then study whether the HOD could mask the differences in the halo populations from different cosmological models so that we obtain the same galaxy clustering. From investigating a variety of galaxy clustering statistics that in principle can be measured from the data, I find that galaxy bias cannot mask substantial changes in cosmology and that it is possible to simultaneously infer the HOD and constrain cosmological parameters using galaxy clustering data alone.