Utilizing aLIGO glitch classifications to validate gravitational-wave candidates
Advanced LIGO data contains numerous noise transients, or 'glitches', that have been shown to reduce the sensitivity of matched filter searches for gravitational waves from compact binaries. These glitches increase the rate at which random coincidences occur, which reduces the significance of identified gravitational-wave events. The presence of these transients has precipitated extensive work to establish that observed gravitational wave events are astrophysical in nature. We discuss the response of the PyCBC search for gravitational waves from stellar mass binaries to various common glitches that were observed during advanced LIGO's first and second observing runs. We show how these transients can mimic waveforms from compact binary coalescences and quantify the likelihood that a given class of glitches will create a trigger in the search pipeline. We explore the specific waveform parameters that are most similar to different glitch classes and demonstrate how knowledge of these similarities can be used when evaluating the significance of gravitational-wave candidates.
© 2020 IOP Publishing Ltd. Received 26 February 2020, revised 23 April 2020. Accepted for publication 11 May 2020. Published 25 June 2020. We would like to thank the Gravity Spy team and the detector characterization group for identifying and characterizing the noise sources addressed in this work, as well as the PyCBC search group for development of the pipeline. We also thank Scott Coughlin, Joshua Smith, and Thomas Massinger for helpful discussions. Computing support for this project was provided by the LDAS computing cluster at the California Institute of Technology. DD, LW, and PRS acknowledge support from NSF award PHY-1607169. LIGO was constructed by the California Institute of Technology and Massachusetts Institute of Technology with funding from the National Science Foundation, and operates under cooperative agreement PHY-0757058. This work carries LIGO document number P-1900372.
Submitted - 2002.09429.pdf