Obtaining statistical significance of gravitational wave signals in hierarchical search
Kanchan Soni,Sanjeev Dhurandhar,Sanjit Mitra
DOI: https://doi.org/10.1103/physrevd.109.024046
IF: 5.407
2024-01-27
Physical Review D
Abstract:Gravitational wave (GW) astronomy has experienced remarkable growth in recent years, driven by advancements in ground-based detectors. While detecting compact binary coalescences (CBCs) has become routine, searching for more complex ones, such as mergers involving eccentric and precessing binaries and subsolar mass binaries, has presented persistent challenges. These challenges arise from using the standard matched filtering algorithm, whose computational cost increases with the dimensionality and size of the template bank. This urges the pressing need for faster search pipelines to efficiently identify GW signals, leading to the emergence of the hierarchical search strategy that reduces the computational cost of matched filtering in the search. This method looks for potential candidate events using a coarse bank of templates (with reduced density and sampling rate) in the first stage, which are then followed up in the second stage with the usual template bank (with optimal density and sampling rate) but only in the neighborhood of the parameter space of the candidate events identified in the first stage. Although the hierarchical search was demonstrated to speed up the standard p y cbc analysis by more than a factor of 20 in a previous work outlined in Soni et al. [ Phys. Rev. D 105 , 064005 (2022)], assigning statistical significance to detected signals, especially in the presence of non-Gaussian noise, was done in a heuristic way. This article introduces a method for background estimation in a two-stage hierarchical search. Our method models the distribution of background triggers obtained from time-shifted triggers in a two-detector network. This modeling precisely aligns with the background distribution across critical signal-to-noise ratios (SNRs), where distinguishing between signal and noise is particularly challenging. It incorporates a fitting procedure to extrapolate to higher detection statistic values. Using this background estimation approach, we assessed the sensitivity of the hierarchical search compared to the standard p y cbc search. Our findings indicate that the hierarchical search attains a sensitive volume-time product comparable to the standard two-detector p y cbc search. This equivalence holds for an inverse false alarm rate of 10 years by a factor ranging from 0.99±0.144 and 1.02±0.141 within the chirp mass range of approximately 1.4 to 20M⊙ . Our methodology accomplishes this while substantially reducing the overall computational cost of the analysis. Specifically, our pipeline exhibits a remarkable speed-up, nearly 13 times faster than p y cbc analysis, including background estimation. https://doi.org/10.1103/PhysRevD.109.024046 © 2024 American Physical Society
astronomy & astrophysics,physics, particles & fields