A Bayesian analysis pipeline for continuous GW sources in the PTA band

TL;DR

This paper presents a Bayesian data analysis pipeline for detecting and characterizing single gravitational wave sources from supermassive black hole binaries in pulsar timing array data, enhancing detection capabilities.

Contribution

It introduces a fully Bayesian pipeline utilizing adaptive metropolis and parallel tempering algorithms for efficient GW source detection and parameter estimation in PTA data.

Findings

Successfully tested on realistic simulated PTA data.

Efficiently locates global maxima in parameter space.

Provides robust evidence evaluation via Bayes Factor.

Abstract

The direct detection of Gravitational Waves (GWs) by Pulsar Timing Arrays (PTAs) is very likely within the next decade. While the stochastic GW background is a promising candidate for detection it is also possible that single resolvable sources may be detectable as well. In this work we will focus on the detection and characterization of single GW sources from supermassive black hole binaries (SMBHBs). We introduce a fully Bayesian data analysis pipeline that is meant to carry out a search, characterization, and evaluation phase. This will allow us to rapidly locate the global maxima in parameter space, map out the posterior, and finally weigh the evidence of a GW detection through a Bayes Factor. Here we will make use of an adaptive metropolis (AM) algorithm and parallel tempering. We test this algorithm on realistic simulated data that are representative of modern PTAs.