Robust Detection and Characterization of Transient Gravitational Waves

University Of Alabama In Huntsville, Huntsville AL

Investigators

Abstract

Hardware upgrades to the Advanced LIGO observatories are complete and soon the most sensitive gravitational wave (GW) detectors the world has ever seen will begin collecting data. Sophisticated data analysis methods are needed to take full advantage of the improved sensitivity and make GW astronomy a reality. Previous analyses have used simplifying assumptions about the detector noise, data calibration, and in some cases the GW signal itself, hindering our ability to make confident detection claims or accurately infer the nature of a GW signal. The research enabled by this award will advance GW data analysis by improving our ability to detect and characterize transient astrophysical signals. The broad theme of the work supported by this award is to build a detailed model of LIGO data including instrument noise, calibration uncertainty, and GW signals. The resulting data analysis methods will improve detection efficiency and reduce systematic errors in parameter estimation. This award will also allow the PI to continue participating in impactful education and public outreach activities and to do so in a region of the United States which has little exposure to gravitational physics. Outreach activities related to this award will include interactions with area high schools, colleges, and universities to publicize the exciting science being done by Advanced LIGO. Research supported by this award will advance three main topics in LIGO data analysis: (i) Un-modeled gravitational wave signals present a challenge for detection and parameter estimation because the signals are rare while transient noise events, or 'glitches', are common. We will complete initial development and then continue to upgrade the BayesWave and BayesLine data analysis pipelines, which together are the state-of-the-art method for distinguishing between burst signals and glitches and drawing inferences about physical properties of the GW source. (ii) The breakthrough technology used by BayesWave and BayesLine is a realistic model for the LIGO noise. The deployment of this model is most mature in the burst analysis; however, all GW searches compete with the same instrument noise. We will adapt the BayesLine algorithm to work with the parameter estimation pipeline, LALInference, for compact binary coalescence signals. (iii) The LIGO data must be calibrated, and there are small statistical uncertainties in amplitude and phase as a result. Previous source characterization analyses have not incorporated uncertainty in the calibration in their data analysis method. The BayesLine algorithm provides the foundation for modeling calibration errors and will be modified to account for systematic phase and amplitude uncertainty in compact binary and burst parameter estimation. This award is supported by the LIGO Research Program of Physics Division in collaboration with the EPSCoR program of the Office of International and Integrative Activities.