Investigation of the effects of non-Gaussian noise transients and their mitigation in parameterized gravitational-wave tests of general relativity

TL;DR

This paper investigates how non-Gaussian noise transients affect gravitational-wave tests of general relativity and evaluates mitigation methods to reduce false deviations caused by instrumental glitches.

Contribution

It demonstrates that specific glitch mitigation techniques effectively eliminate false GR violations caused by transient noise in gravitational-wave data.

Findings

Glitches can cause false violations of GR in tests.

Inpainting filter and glitch model subtraction mitigate false positives.

Mitigation methods do not introduce additional biases.

Abstract

The detection of gravitational waves from compact binary coalescence by Advanced LIGO and Advanced Virgo provides an opportunity to study the strong-field, highly relativistic regime of gravity. Gravitational-wave tests of general relativity (GR) typically assume Gaussian and stationary detector noise and, thus, do not account for non-Gaussian, transient noise features (glitches). We present the results obtained by performing parametrized gravitational-wave tests on simulated signals from binary-black-hole coalescence overlapped with three classes of frequently occurring instrumental glitches with distinctly different morphologies. We then review and apply three glitch mitigation methods and evaluate their effects on reducing false deviations from GR. By considering nine cases of glitches overlapping with simulated signals, we show that the short-duration, broadband blip and tomte glitches under consideration introduce false violations of GR, and using an inpainting filter and glitch model subtraction can consistently eliminate such false violations without introducing additional effects.