An Evidence Based Search Method For Gravitational Waves From Neutron Star Ring-downs

TL;DR

This paper introduces a Bayesian search method for gravitational waves from neutron star ring-downs, effectively distinguishing signals from noise and instrumental artifacts, with promising detection efficiency at realistic astrophysical distances.

Contribution

It presents a novel Bayesian analysis technique that incorporates prior source information and models instrumental artifacts to improve gravitational wave detection from neutron star ring-downs.

Findings

Achieves 50% detection efficiency at SNR 5.2

Maintains 1% false alarm probability

Detects signals from sources up to 15 kpc with specific energy emission

Abstract

The excitation of quadrupolar quasi-normal modes in a neutron star leads to the emission of a short, distinctive, burst of gravitational radiation in the form of a decaying sinusoid or `ring-down'. We present a Bayesian analysis method which incorporates relevant prior information about the source and known instrumental artifacts to conduct a robust search for the gravitational wave emission associated with pulsar glitches and soft $\gamma$-ray repeater flares. Instrumental transients are modelled as sine-Gaussian and their evidence, or marginal likelihood, is compared with that of Gaussian white noise and ring-downs via the `odds-ratio'. Tests using simulated data with a noise spectral density similar to the LIGO interferometer around 1 kHz yield 50% detection efficiency and 1% false alarm probability for ring-down signals with signal-to-noise ratio $\rho=5.2$. For a source at 15 kpc this requires an energy of $1.3\times 10^{-5}M_{\astrosun}c^2$ to be emitted as gravitational waves.