BayesWave++: Implementing a Bayesian Inference Package for Gravitational Waves in C++

Other Versions:

Abstract:: Gravitational waves (GWs) are minute ripples in spacetime detectable only by a global network of interferometer observatories that sense length changes on the scale of a proton. Despite the careful engineering of these observatories, GW signals are notoriously difficult to extract and characterize, often mimicked or masked by the significant detector noise caused by terrestrial events, meteorological interference, and instrument glitches. BayesWave++ is an improved C++ implementation of BayesWave, a Bayesian inference software written to enable signal modeling in this noisy environment. Using a transdimensional Reversible Jump Markov Chain Monte Carlo (RJMCMC) algorithm, BayesWave++ fits sine-exponential wavelets to burst signals of both GW events and instrument glitches, and specific waveform templates from compact binary coalescence (CBC) simulations to GW events. Here, pipeline support for automated analysis of real GW data is added to BayesWave++, and a thorough review is conducted of the GW/glitch wavelet and CBC models to ensure proper algorithm performance. From this review, new issues throughout the BayesWave++ software are identified and corrected, and bounds are determined on the subset of parameter space in which the algorithm works well.

Files in Document:

Other Files:

LIGO Document T2400303-v2