Accurate identification of the nature of signals in ground-based gravitational-wave interferometer data

Bayesian parameter estimation and model selection methods have been essential to measure the properties of the binary-black-holes and of the binary-neutron-star detected with the gravitational-wave observations from the LIGO-Virgo network. Those techniques are a key component of the new field of gravitational-wave astrophysics. This project will improve existing methods to advance our understanding of the next observations.

It is not currently possible to solve exactly Einstein's equations and produce perfectly accurate gravitational-wave signal models. So current analysis techniques involve comparing observational results using different approximate models (which can be computed quickly enough to be usable). The difference between such waveform templates is then used to check the uncertainties associated with the models' approximations. And this source of error dominates the measurement of certain gravitational-wave parameters.

One of the most crucial challenges, and a principal direction of this project, involves moving beyond this single point-estimate of the model errors (by providing rigorous marginalisation methods for instance). This work will lead to enabling new constraints of alternative theories of gravity and measurements of neutron-star equation-of-state. This involves developing an using accurate, precise, and efficient methods for observational gravitational-wave astronomy on future observations from a growing worldwide network of detectors.