Frequency-Domain Analysis of Black-Hole Ringdowns

TL;DR

This paper introduces a frequency-domain method for analyzing black-hole ringdown signals that improves parameter estimation accuracy by avoiding spectral leakage and utilizing Bayesian inference.

Contribution

The novel approach models inspiral and merger with sine-Gaussian wavelets and applies Bayesian inference in the frequency domain for better black-hole parameter estimation.

Findings

Tighter constraints on black-hole mass and spin compared to time-domain methods

Effective analysis of simulated signals across different SNRs and detector networks

Ability to search over sky position and ringdown start time with Bayesian methods

Abstract

We propose a novel, frequency-domain approach to the analysis of the gravitational-wave ringdown signal of binary black holes and the identification of quasinormal mode frequencies of the remnant. Our approach avoids the issues of spectral leakage that would normally be expected (associated with the abrupt start of the ringdown) by modeling the inspiral and merger parts of the signal using a flexible sum of sine-Gaussian wavelets truncated at the onset of the ringdown. Performing the analysis in the frequency domain allows us to use standard (and by now well-established) Bayesian inference pipelines for gravitational wave data as well as giving us the ability to readily search over the sky position and the ringdown start time, although we find that it is necessary to use an informative prior for the latter. We test our method by using it to analyze several simulated signals with varying signal-to-noise ratios injected into two- and three-detector networks. We find that our frequency-domain approach is generally able to place tighter constraints on the remnant black-hole mass and spin than a standard time-domain analysis.