Project overview
This proposal contains the main research goals of the Southampton LIGO-Virgo-KAGRA (LVK) Gravitational Waves group for the period 20025-2028. It consists of two Themes, both central to the LVK science remit. Theme 1 concerns searches for continuous gravitational waves signals (CGWs) from spinning neutron stars (NSs). No such signal has been detected yet, but the first detection could come at any time. CGW searches typically assume a signal template, which is then compared against the detector data. The template?s phase needs to match the signal over the course of the observation. For known NSs this is not a problem, but for unknown NSs, or know ones with unknown spin frequency, this requires a search in the very large space of the NSs spin-down parameters and (if relevant) sky location. This makes the searches computationally very expensive. It also renders the searches vulnerable to sudden step changes known as glitches in the signal, a phenomenon known to be common in the electromagnetically observed NS population. The Project Lead has previously led a study confirming that this is a real issue for CGW searches, with the NSs that spin-down most rapidly, and are therefore potentially the strongest CGW emitters, glitching the most often and the most strongly. In Theme 1 we will devise modifications to the hierarchical search pipelines commonly used in CGW searches to see where the effects of unmodelled glitches first degrade detection efficiency, and how to then make the pipelines robust against such glitches, while being mindful of the computational cost. Theme 2 concerns systematics in waveform models for compact binary coalescence events. With nearly one hundred such events now reported, a statistically meaningful array of source parameters has now been assembled. However, the parameter estimation itself makes use of phenomenological waveforms that approximate accurate waveforms produced by full numerical relativity simulations. These approximations introduce systematic errors into the parameter estimation, on top of the random errors that the finite signal-to-noise introduces. As the sensitivity of the detectors improves for O5, the random errors due to noise will shrink, exposing systematic biases and potentially degrading all parameter estimation in future Observing Runs, and limiting our ability to carry out our science goals such as testing General Relativity and probing the high-density equation of state. Indeed, there is evidence that this was already a problem for a handful of detections in observing run O3. We plan to build on our existing expertise in investigating systematics to extend an in-preparation procedure for identifying systematics in waveform models to fully generic signals. At the moment, this procedure is prototyped for simple aligned-spin, quadrupolar signals, and the work outlined for this grant will extend the method to work with fully precessing models with higher multipoles. Another goal of this project is to increase the efficiency of the method, allowing for on-the-fly systematics estimation in Bayesian model selection schemes. We are asking for staff time for both Themes, and a PhD student for Theme 1, and a RIA (i.e. post-doc) for Theme 2.
