Daniel Matthews

  • Alumnus


Title: Exploring the Distant Universe with Cross-Correlation Statistics Abstract: Future cosmological surveys will require distance information for an extremely large number of galaxies in order to gain insight into the structure and history of our Universe. Current methods of obtaining accurate distance information such as measuring the redshifts of galaxies via spectroscopy are not feasible for such enormous datasets, mainly due to the long exposure times required. Photometric redshifts, where the redshift is measured using broadband imaging through only a few lters, are a promising avenue of study, although there are inherent limitations to this method making them less understood than spectroscopic redshifts. Understanding these limitations and improving the calibration of photometric redshifts will be very important for future cosmological measurements. This thesis presents tests of a new technique for calibrating photometric redshifts that exploits the clustering of galaxies due to gravitational interaction. This cross-correlation technique uses the measured spatial clustering on the sky of a photometric sample that has only imaging information, with a spectroscopic sample that has secure and accurate redshifts. These tests shows that measurements of this clustering as a function of redshift can be used to accurately reconstruct the true redshift distribution of the photometric sample. In addition, this thesis shows how similar clustering measurements can be used to constrain the contamination of a high redshift candidate sample by low redshift interlopers. Finally it describes a new catalog that combines spectroscopic redshifts and deep photometry that can be used as a testbed for future photo-z studies.