Abstract
The field of gravitational wave astronomy has provided the first direct observations of gravitational waves. Using ground-based LASER interferometersgravitational wave signals from compact binary coalescence have been observed
including the observation of binary neutron stars and binary black holes. This
thesis focuses on modelled searches for compact binary coalescence and introduces new methods to search for signals in challenging areas of the parameter space, improving their sensitivity to currently unobserved systems.
We begin by reviewing the theory of gravitational waves, focusing on compact
binaries coalescences. We describe the basic setup of LASER interferometers
and the data that they produce. We then discuss the methods used within the
modelled search to analyse the detector data and to mitigate the effects of noise.
We introduce a new framework for the construction and tuning of χ2 signalconsistency tests using machine-learning methods. This framework is applied to a search targeting intermediate mass black hole binaries, and a new χ2 test is created to target short-duration noise transients that reduce the sensitivity of the search. We demonstrate the effect of the new χ2 test on the sensitivity of
the search, showing an increase in the sensitive volume of ∼ 11% for binaries
with masses > 300M⊙.
We develop a search for precessing binaries using a harmonic decomposition of
the precessing waveforms. We demonstrate a method for creating template banks
for use in this search and adapt the existing χ2 and coincident signal-consistency
tests in order to minimise the additional noise introduced by this model. We
show that for neutron star black hole binaries the sensitivity of the search can
be increased by ∼ 50% for the most extreme cases of precession, demonstrating
this method as a promising solution to the observation of precessing binaries.
We perform a targeted search for strongly lensed counterparts of known
gravitational wave events and the results are tested for consistency with the
noise hypothesis. A method is developed using nested sampling in order to test
if any new candidate is more consistent with a second strongly lensed image or
an independent astrophysical event. Simulated signals are then used to test the
sensitivity of the search and to investigate a proposed extreme lensing hypothesis, ruling out the hypothesis with high confidence.
Finally, we describe the Black Hole Hunter public outreach tool and discuss
updates to the game and its application to public outreach.
| Date of Award | 27 Feb 2023 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Ian Harry (Supervisor), Robert Nichol (Supervisor) & Coleman Krawczyk (Supervisor) |