Background Filter: A method for removing signal contamination during significance estimation of a GstLAL anaysis

Understand

To evaluate the probability of a gravitational-wave candidate originating from noise, GstLAL collects noise statistics from the data it analyzes.

• Gravitational-wave signals of astrophysical origin get added to the noise statistics, harming the sensitivity of the search.
• We present the Background Filter, a novel tool to prevent this by removing noise statistics that were collected from gravitational-wave candidates.
• To demonstrate its efficacy, we analyze one week of LIGO and Virgo O3 data, and show that it improves the sensitivity of the analysis by 20-40% in the high mass region, in the presence of 868 simulated gravitational-wave signals.