Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Duncan Brown


Black holes, Gravitational waves, LIGO, Numerical relativity

Subject Categories



Gravitational waves are a consequence of the general theory of relativity. Direct detection of such waves will provide a wealth of information about physics, astronomy, and cosmology. A worldwide effort is currently underway to make the first direct detection of gravitational waves. The global network of detectors includes the Laser Interferometer Gravitational-wave Observatory (LIGO), which recently completed its sixth science run.

A particularly promising source of gravitational waves is a binary system consisting of two neutron stars and/or black holes. As the objects orbit each other they emit gravitational radiation, lose energy, and spiral inwards. This produces a characteristic ``chirp'' signal for which we can search in the LIGO data. Currently this is done using matched-filter techniques, which correlate the detector data against analytic models of the emitted gravitational waves. Several choices must be made in constructing a search for signals from such binary coalescences.

Any discrepancy between the signals and the models used will reduce the effectiveness of the matched filter. However, the analytic models are based on approximations which are not valid through the entire evolution of the binary. In recent years numerical relativity has had impressive success in simulating the final phases of the coalescence of binary black holes. While numerical relativity is too computationally expensive to use directly in the search, this progress has made it possible to perform realistic tests of the LIGO searches. The results of such tests can be used to improve the efficiency of searches.

Conversely, noise in the LIGO and Virgo detectors can reduce the efficiency. This must be addressed by characterizing the quality of the data from the detectors, and removing from the analysis times that will be detrimental to the search.

In this thesis we utilize recent results from numerical relativity to study both the degree to which analytic models match realistic waveforms and the ability of LIGO searches to make detections. We also apply the matched-filter search to the problem of removing times of excess noise from the search.


Open Access

Included in

Physics Commons