Application of hidden Markov model tracking to the search for long-duration transient gravitational waves from the remnant of the binary neutron star merger GW170817

TL;DR

This paper presents a hidden Markov model-based frequency tracking method to search for long-duration gravitational wave signals from neutron star remnants, validated through simulations on GW170817 data.

Contribution

It introduces an efficient, fully described hidden Markov model approach for detecting long-duration gravitational waves from neutron star remnants, enhancing search capabilities.

Findings

Method validated with Monte-Carlo simulations

Applied to GW170817 data set

Provides a reproducible analysis framework

Abstract

The nature of the post-merger remnant of the first binary neutron star coalescence observed by the Advanced Laser Interferometer Gravitational Wave Observatory (Advanced LIGO) and Advanced Virgo, GW170817, is unknown. Searches have been carried out for short ($\lesssim 1$ s), intermediate ($\lesssim 500$ s), and long ($\sim$ days) signals using various algorithms without yielding a detection. We describe an efficient frequency tracking scheme based on a hidden Markov model to search for long-duration transient signals from a neutron star remnant with spin-down time-scale in the range $\sim 10^2$s-$10^4$s. It was one of the methods used in the search for signals from a long-lived remnant of GW170817. We validate the method and estimate its sensitivity through Monte-Carlo simulations on the same data set as used in the GW170817 search. We describe the search configuration, procedure, and follow-up step by step. The methodology of the hidden Markov model is described fully to ensure that future analyses of this kind can be reproduced by an independent party.