Title: Preparing for next-generation of galaxy surveys
In the 1920s, Edwin Hubble discovered that the Milky Way is merely one of countless galaxies in the Universe, and the relative velocities of these galaxies indicate that the Universe is expanding. Developments in the ensuing decades have solidly established galaxy surveys as a crucial probe of cosmology. In this thesis, I describe my work related to preparations for the next generation of galaxy surveys.
In the first part of the thesis, we present catalogs of calibrated photometry and spectroscopic redshifts in the Extended Groth Strip, intended for studies of photometric redshifts (photo-z's). The data includes ugriz photometry from CFHTLS and Y-band photometry from the Subaru Suprime camera, as well as spectroscopic redshifts from the DEEP2, DEEP3 and 3D-HST surveys. These catalogs incorporate corrections to produce effectively matched-aperture photometry across all bands, based upon object size information available in the catalog and Moffat profile point spread function fits. We test this catalog with a simple machine learning-based photometric redshift algorithm based upon Random Forest regression, and find that the corrected aperture photometry leads to significant improvement in photo-z accuracy compared to the original SExtractor catalogs from CFHTLS and Subaru. The deep ugrizY photometry and spectroscopic redshifts are well-suited for empirical tests of photometric redshift algorithms for LSST.
In the second part of the thesis, we model the redshift-dependent clustering of a DESI-like LRG sample in the halo occupation distribution (HOD) framework using photometric redshifts (photo-z's). The LRG sample contains 2.7 million objects over a 5655 square degree area and spans the redshift range of 0.4 < z < 0.9. The photo-z's are computed using DECam and WISE photometry, and they are highly accurate for the LRG sample with NMAD = 0.02. We measure the galaxy clustering with the projected correlation function, rather than the angular correlation function, to make optimal use of the photo-z's. We find that there is little evolution in the host halo properties except at the high-redshift end. The bias evolution is mostly consistent with the simple prescription of constant clustering amplitude. We discuss a number of methodological improvements developed as part of this work and how they can be applied to future surveys.
Tuesday, July 2 at 2:00 p.m.