Data analysis for space-based gravitational wave detectors
Crowder, Jefferson Osborn.
MetadataShow full item record
With the launch of the Laser Interferometer Space Antenna (LISA) expected for the next decade, the nascent field of gravitational wave astronomy will be taking a giant leap forward. The data that will be gathered from space-borne gravitational wave detectors such as LISA will provide an expansive look through a new window on the Universe. This dissertation is presented to help open that window by exploring some of the techniques and methods that will be needed to understand the data from these detectors. The first original work presented here investigates the resolution of LISA and follow-on space-based gravitational wave missions. This work presents the methods of measuring the precision of these detectors and gives results for their resolving power for a large class of expected gravitational wave sources. The second original investigation involves the effect that multiple gravitational wave sources will have on the resolution of LISA. Previous results concerning detector resolution were limited to isolated sources of gravitational waves. As LISA is an allsky detector, it is necessary to understand the role played by concurrent detection of numerous sources. This work derives an extension of the Fisher Information Matrix approach for determining parameter resolution, and applies it to multiple sources for LISA. The next original work is an exploration of the method of genetic algorithms on the problem of extracting the binary parameters of gravitational wave sources from the LISA data stream. These are global algorithms providing a means to cover the entire search space of parameter values. This work describes the basics of and provides the results for such genetic algorithm-based searches, with a focus on improving algorithm efficiency. The last original work included is a study of Markov Chain Monte Carlo (MCMC) methods applied to parameter extraction of gravitational wave sources in the LISA data stream. This work shows how an MCMC approach provides a global means of both searching for and characterizing the distributions of the source parameters. Results also show that distributions found by this global method match with previous approaches that were limited to regions local to the source parameters.