UC Berkeley

The study of large-scale structure aims to describe the distribution of matter in the universe. This thesis examines two important topics in this field: intrinsic alignments and anisotropic galaxy clustering. Intrinsic alignments of galaxies are correlations between galaxy shapes and the surrounding density field and are a significant systematic uncertainty in weak gravitational lensing measurements. We examine intrinsic alignments from both theoretical and observational perspectives. First, the most commonly used model for intrinsic alignments, the tidal (linear) alignment model, is tested and developed in the context of recent measurements. The model is found to provide a reasonably accurate description of observations on large scales, particularly for luminous red galaxies. We raise unresolved issues with the model and explore several ways in which it could be improved and expanded. Second, we develop a technique to separate intrinsic alignments from galaxy-galaxy lensing measurements. This technique allows the removal of contamination from the desired lensing signal while also providing a probe of intrinsic alignments in different galaxy populations. Using data from the Sloan Digital Sky Survey (SDSS), we use this method to place the tightest current constraints on intrinsic alignment strength for the SDSS lensing sample, finding that it is a subdominant source of uncertainty at the current level of statistical precision.

The measured galaxy clustering signal in redshift surveys is anisotropic due to both geometric and dynamic distortions. We discuss the nature of these distortions and examine their information content in the context of future redshift surveys. We employ a model for the galaxy power spectrum in redshift space, based on an expansion in velocity moments of the phase-space distribution function of matter. This model includes nonlinear treatment of matter clustering, galaxy bias, shot noise, and redshift-space distortions. Comparing with numerical simulations, we find that this model is able to describe halo clustering with sufficient accuracy to allow the inclusion of broadband information on small scales, thus significantly improving constraints on both geometry and growth of structure in the universe.