Bayesian Inference for Scattering Arches

In ground-based gravitational wave detectors, transient noise sources known as "glitches" pose a significant challenge for the detection and analysis of gravitational-wave sources. One common glitch mechanism is scattering of light from the main beam path, which later recombines to produce transients with characteristic morphology. We utilize a previously characterized model for this morphology and perform Bayesian inference to generate distributions on free parameters in this model, allowing for the subtraction of these glitches. Moreover, using standard tools for Bayesian inference - also used in the inference of compact binary coalescence (CBC) parameters - this method is suitable for analysis of scattered light glitches near or overlapping astrophysical signals, including by joint inference. Using this tool may improve the recovery of accurate CBC parameters, in cases where glitch power may otherwise bias the recovery of those parameters. This model also allows us to discriminate between the presence or absence of extra structure in the glitch morphology, which may not be distinguishable in previous methods of analysis.