Localization of Compact Binary Sources with Second Generation Gravitational-wave Interferometer Networks

TL;DR

This paper analyzes how second-generation gravitational-wave detector networks in the mid-2020s will localize compact binary sources, showing that increased network size improves sky localization and discussing expected localization areas for different binary systems.

Contribution

It provides realistic localization estimates for binary neutron star and neutron star-black hole mergers with upcoming detector networks, considering duty cycles and selection effects.

Findings

Median localization of 28 sq. deg. for binary neutron stars

50-120 sq. deg. localization for neutron star-black hole binaries

Full five-instrument network reduces localization to a few tens of degrees

Abstract

GW170817 began gravitational-wave multimessenger astronomy. However, GW170817 will not be representative of detections in the coming years -- typical gravitational-wave sources will be closer the detection horizon, have larger localization regions, and (when present) will have correspondingly weaker electromagnetic emission. In its design state, the gravitational-wave detector network in the mid-2020s will consist of up to five similar-sensitivity second-generation interferometers. The instantaneous sky-coverage by the full network is nearly isotropic, in contrast to the configuration during the first \change{three} observing runs. Along with the coverage of the sky, there are also commensurate increases in the average horizon for a given binary mass. We present a realistic set of localizations for binary neutron stars and neutron star--black hole binaries, incorporating intra-network duty cycles and selection effects on the astrophysical distributions. Based on the assumption of an $80\%$ duty cycle, and that two instruments observe a signal above the detection threshold, we anticipate a median of $28$ sq.\ deg.\ for binary neutron stars, and $50$--$120$ sq.\ deg.\ for neutron star--black hole (depending on the population assumed). These distributions have a wide spread, and the best localizations, even for networks with fewer instruments, will have localizations of $1$--$10$ sq.\ deg.\ range. The full five instrument network reduces localization regions to a few tens of degrees at worst.