DR. SHASVATH KAPADIA

Orbiting compact binaries are among the most promising sources of Gravitational Waves (GWs) observable by kilometre-long ground-based interferometers. Detecting these waves involves sophisticated data analysis techniques centered around matched filtering, requiring large banks of accurate GW templates, as well as methods to distinguish between candidate events triggered by true GW signals, and those caused by non-Gaussian noise (“glitches”). Focusing on searches for inspiralling Neutron-Star Black-Hole binaries, we present here a method to improve the accuracy of analytical (Post-Newtonian (PN)) waveforms by the inclusion of “pseudo” Post Newtonian terms determined by calibrating the PN waveforms to more accurate (though computationally expensive) fiducial waveforms. We further introduce a machine-learning based approach to segregate candidate events, and show that, by identifying and exploiting a number of their key properties, we can improve on existing classification methods. And finally, we briefly explore the possibility of detecting a permanent non-linear DC imprint left in the arms of interferometric detectors (the “Christodoulou memory”) by the recently discovered gravitational waves.