Since its inception in 1980 by Guth [1], inflation has emerged as the dominant paradigm for describing the physics of the early universe. While inflation has matured theoretically over two decades, it has only recently begun to be rigorously tested observationally. Measurements of the cosmic microwave background (CMB) and large-scale structure surveys (LSS) have begun to unravel the mysteries of the inflationary epoch with exquisite and unprecedented accuracy. This thesis is a contribution to the effort of reconstructing the physics of inflation. This information is largely encoded in the potential energy function of the inflaton, the field that drives the inflationary expansion. With little theoretical guidance as to the probable form of this potential, reconstruction is a predominantly data-driven endeavor. This thesis presents an investigation of the constrainability of the inflaton potential given current CMB and LSS data. We develop a methodology based on the inflationary flow formalism that provides an assessment of our current ability to resolve the form of the inflaton potential in the face of experimental and statistical error. We find that there is uncertainty regarding the initial dynamics of the inflaton field, related to the poor constraints that can be drawn on the primordial power spectrum on large scales. We also investigate the future prospects of potential reconstruction, as might be expected when data from ESA's Planck Surveyor becomes available. We develop an approach that utilizes Markov chain Monte Carlo to analyze the statistical properties of the inflaton potential. Besides providing constraints on the parameters of the potential, this method makes it possible to perform model selection on the inflationary model space. While future data will likely determine the general features of the inflaton, there will likely be many different models that remain good fits to the data. Bayesian model selection will then be needed to draw comparisons between these different models in a statistically rigorous fashion.The software used to evaluate the power spectra, Flowcodel. ... code and further details about the calculation are available at http:/ /www. physics.buffalo.edu/ whkinney/flowcode. ... We work to 6^-order in the flow space, i.e. 1\h = 0 for i agt; 6.

Title | : | Recovering the Inflationary Potential: An Analysis Using Flow Methods and Markov Chain Monte Carlo |

Author | : | Brian A. Powell |

Publisher | : | ProQuest - 2008 |

Continue