Bayesian analysis of the DES Supernova Survey

The Dark Energy Survey (DES; www.darkenergysurvet.org) will end in 2018 having completed a 5 year survey of the southern sky. A key part of DES is a dedicated tranisent survey which has already detected more Type Ia Supernovae (SNeIa) than any previous search. By the end of DES, we expect to have detected and classified over 4000 SNeIa allowing us to accurately map the expansion history of the Universe over half its age.

Key to harnessing the power of these SNeIa will be controlling the systematic uncertainties, including photometric calibration, host galaxy correlations and mis-classifications. The latter represents a new challenge for DES as many of its SNeIa will be photometrically-classified leading to additional, presently unknown, contamination from other transient events.

This thesis will build on the work of previous ICG students (e.g. Campbell et al. 2013, https://arxiv.org/pdf/1211.4480.pdf) to include such uncertainties in the analysis of DES SNeIa. In particular, we will exploit new Bayesian techniques, only developed in the last few years, to create unbiased SNeIa samples for cosmology (Jones et al. 2017, https://arxiv.org/pdf/1611.07042.pdf). We will then use this sample to measure cosmological parameters (including the equation of state of dark energy) and investigate new opportunities to detect the gravitational lensing of SNeIa by foreground large-scale structures (Scovacricchi et al. 2017, https://arxiv.org/pdf/1611.01315.pdf).

This project will require the development of new statistical methodology and the creation of large simulations of the DES supernova survey.