Rapid Localization of Gravitational Wave Sources from Compact Binary Coalescences Using Deep Learning
Abstract
The mergers of neutron star-neutron star and neutron star-black hole binaries (NSBHs) are the most promising gravitational wave (GW) events with electromagnetic (EM) counterparts. The rapid detection, localization, and simultaneous multimessenger follow-up of these sources are of primary importance in the upcoming science runs of the LIGO-Virgo-KAGRA Collaboration. While prompt EM counterparts during binary mergers can last less than 2 s, the timescales of existing localization methods that use Bayesian techniques, vary from seconds to days. In this paper, we propose the first deep learning-based approach for rapid and accurate sky localization of all types of binary coalescences, including neutron star-neutron star and NSBHs for the first time. Specifically, we train and test a normalizing flow model on matched-filtering output from GW searches to obtain sky direction posteriors in around 1 s using a single P100 GPU, which is several orders of magnitude faster than full Bayesian techniques.
- Publication:
-
The Astrophysical Journal
- Pub Date:
- December 2023
- DOI:
- 10.3847/1538-4357/ad08b7
- arXiv:
- arXiv:2207.14522
- Bibcode:
- 2023ApJ...959...42C
- Keywords:
-
- Gravitational waves;
- Gravitational wave sources;
- Convolutional neural networks;
- Neural networks;
- 678;
- 677;
- 1938;
- 1933;
- General Relativity and Quantum Cosmology
- E-Print:
- 18 pages, 8 figures