Unbiased estimation of gravitational-wave anisotropies from noisy data

Abstract

One of the most exciting targets of current and future gravitational-wave observations is the angular power spectrum of the astrophysical GW background. This cumulative signal encodes information about the large-scale structure of the Universe, as well as the formation and evolution of compact binaries throughout cosmic time. However, the finite rate of compact binary mergers gives rise to , which introduces a significant bias in measurements of the angular power spectrum if not explicitly accounted for. Previous work showed that this bias can be removed by cross-correlating GW sky maps constructed from different observing times. However, this work considered an idealized measurement scenario, ignoring detector specifics and in particular noise contributions. Here we extend this temporal cross-correlation method to account for these difficulties, allowing us to implement the first unbiased anisotropic search pipeline for LIGO-Virgo-KAGRA data. In doing so, we show that the existing pipeline is biased , due to previously neglected subleading contributions to the noise covariance. We apply our pipeline to mock LIGO data, and find that our improved analysis will be crucial for stochastic searches from the current observing run (O4) onwards. Published by the American Physical Society 2024