DEPARTMENT OF PHYSICS DISSERTATION DEFENSE: Mahlet Shiferaw

Public zoom link:  https://stanford.zoom.us/j/94610179379?pwd=zu1VWy628ZdHAQW5KWssX8sma3nd7x.1

Password: Email Physicsstudentservices [at] stanford.edu (physicsstudentservices[at]stanford[dot]edu) for password

Title:  

Exploring the Galaxy-Dark Matter Connection

Abstract:  

The Universe is shaped by a cosmic web of dark matter, with galaxies residing in its densest peaks, known as “halos”. Modeling the statistical relationship between these luminous galaxies and the invisible dark matter halos which host them—the “galaxy-halo connection”—allows us to infer cosmological parameters and address fundamental questions regarding the nature of dark matter and dark energy. The next decade offers an unprecedented opportunity to do just that, as current and upcoming surveys will map billions of galaxies across a wider range of scales than ever before. The primary limitation to maximally extracting cosmological information now lies in the accuracy of our models of the galaxy-dark matter connection. In my thesis, I present work that accomplishes this using both theory and observations. First, I adopt a relatively new approach known as Hybrid Effective Field Theory (HEFT), which combines the nonlinear displacement from dark matter N-body simulations with analytic perturbation theory to model galaxy clustering on weakly nonlinear scales. Using two distinct galaxy formation models with and without assembly bias, I quantify uncertainties in galaxy formation physics via the HEFT parameters, and set physically motivated priors for Bayesian cosmological analyses. Second, I utilize the Gaia-Unwise Quasar Catalog (Quaia): a large all-sky spectrophotometric catalog of quasars, which are extremely luminous galactic nuclei powered by accretion onto black holes. Understanding the co-evolution of quasars with galaxies and halos will ultimately contribute to building robust models of galaxy formation. I thus develop a pipeline for inferring the quasar-halo relationship from angular clustering in Quaia, and characterize the dependence of quasar clustering on redshift and luminosity. Finally, I discuss the outlook of these projects within the upcoming landscape of simulation-based inference, a technique which allows us to directly estimate the posterior distribution of cosmological parameters from forward models of structure formation. Together, this work advances our ability to extract cosmology from the next generation of galaxy surveys.