From Shapes to S8: Extracting precision cosmological constraints from the Hyper Suprime-Cam galaxy survey

Abstract

In this dissertation I present cosmological parameter constraints from the Hyper Suprime-Cam Subaru Strategic Program (HSC SSP), obtained using weak gravitational lensing, specifically cosmic shear angular power spectra. I begin with an overview of weak gravitational lensing and cosmic shear, as well as a description of the HSC survey. I then describe the measurement of cosmic shear power spectra from the HSC Y3 data, including the estimation of galaxy shears, our methodology to prevent confirmation bias, the Pseudo-Cl method used to obtain unbiased power spectra, the estimation of the covariance matrix and the validation and null tests carried out on our measured spectra. I then discuss in detail the contamination to the power spectrum measurement from point spread function (PSF) systematics, highlighting a novel methodology to measure and model the impact of both second and fourth moments of the PSF on the power spectrum and inferred cosmological parameters. I also describe systematics arising from the measurement of the source redshift distribution, discussing our estimation of individual galaxy photometric redshifts and the underlying source redshift distribution, as well as our modeling of the uncertainties in the latter in our cosmological analysis. I highlight a key finding from the HSC Y3 cosmology analyses, namely evidence of a systematic bias in the inferred source redshift distributions at z>1, and discuss how this is corrected for in our cosmological analysis, and the work that is necessary to mitigate this issue for future surveys and cosmological analyses. I then describe our fiducial cosmological analysis, and the resulting parameter constraints, focusing on the measured value of S8, which is 2-sigma lower than the constraint from primary CMB measurements, lending credence to the apparent S8 tension observed by other weak lensing surveys. I conclude by discussing the potential for baryonic feedback modeling to be the source of the S8 tension, and show how jointly optimizing the scale cuts and baryonic model complexity will enable unbiased, high-precision cosmological constraints from future data sets.