Analysis of Bayesian Transfer Function Fitting Method - A Potential Tool for Estimating Interferometer Uncertainty

Caden Swain

Research output: ThesisHonors Thesis

Abstract

The Response Function of the LIGO Interferometer is central to reconstructing the strain produced by incoming gravitational waves. As a function of the interferometer's response to external stimuli, the Response Function is both analytically modeled and experimentally measured using excitations from the photon calibrator system at discrete frequencies. The uncertainty in each data point is propagated to the residual between the model and measurements, with both the uncertainty and residual being interpolated over a broadband frequency range. This project explores an alternative method to estimating the uncertainty and systematic error present in the Response function model through the utilization of Bayesian Statistics as opposed to data interpolation. We fit a distribution of transfer functions directly to the residual of the Response Function, bypassing the inherent error in data interpolation. Using data gathered from an electronic whitening chassis at discrete frequency points as a substitute for Response Function data and varying the Signal-to-Noise Ratio as a proxy for varying the uncertainty in the measurements, we analyze the results of this method and compare its precision and accuracy as compared to the standard transfer-function-fitting method.

Original languageEnglish
Awarding Institution
  • Loyola Marymount University
Supervisors/Advisors
  • Mureika, Jonas, Advisor
  • Berube, David, Advisor
  • Varieschi, Gabriele U, Advisor
StatePublished - May 3 2024
Externally publishedYes

Cite this