SALT spectroscopy and classification of supernova spectra using Bayesian techniques

In this thesis, we present the Southern African Large Telescope spectroscopic follow-up programme for supernova candidates discovered by the international Dark Energy Survey, the goals of which are to measure the expansion history of the Universe and shed light on the mysterious nature of dark energy. In total, we took spectra for 36 supernova candidates. These were classified using a new Bayesian Supernova spectra classifier, SuperNovaMC, that we developed to address limitations with existing algorithms. SuperNovaMC simultaneously finds the best fitting supernova and host galaxy using Bayesian model selection, fitting the entire spectrum with Monte Carlo Markov Chain methods which allow estimation of the entire parameter posterior distributions, and hence principled statistical analysis even at low signal-to-noise. After extensive testing of SuperNovaMC against simulations and literature data, we use it to classify 20 of our Dark Energy Survey candidates as Type Ia supernovae. We further performed equivalent width measurements of two Type Ia supernova spectral features: Ca II H&K and Si II 4000, using a sub-sample of the 20 Type Ia supernovae. We compared our results to those of a similar study conducted on a low-redshift (z < 0:1) Type Ia supernova sample and found the two sets of results to be consistent, suggesting no redshift evolution in the equivalent widths of the two spectral features in the redshift range 0:1 < z < 0:3 that we conducted the study in.

Grantor dc:publisher.institution

Department of Mathematics and Applied Mathematics

Year dc:date.issued

2017

Author dc:creator

• Kasai, Eli Kunwiji

Advisors dc:contributor.advisor

• Bassett, Bruce A
• Crawford Steven, M

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/11427/27283

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/27283