Advancement to Candidacy: Cosmic Calibration - Statistical Modeling for Dark Energy and the Cosmological Constants

Tracy Holsclaw
Monday, October 19, 2009, 1:30 pm, Engineering 2 Room 280
Hosted by Herbie Lee
Applied Mathematics & Statistics

Abstract

The fact that the Universe is expanding has been known since the 1920's. If the Universe was filled with ordinary matter, the expansion should be decelerating. Beginning in 1998, however, observational evidence has been accumulating in favor of an accelerating expansion of the Universe.
The unknown driver of the acceleration has been termed dark energy. The nature of dark energy can be investigated by studying its equation of state, that is the relationship of its pressure to its density. The equation of state can be measured via a study of the luminosity distance-redshift relation for supernovae. In this study, we employ supernovae data, including measurement errors, to determine whether the equation of state is constant or not. Our method is based on Bayesian analysis of a differential equation and modeling w(z) directly, where w(z) is the equation of state parameter.  Bayesian parametric models with cosmological significance, Gaussian process models that take full advantage of their properties, and other Markov chain Monte Carlo simulations are used because of the complexities in dealing with the highly non-linear differential equation.