Accelerated Bayesian model-selection and parameter-estimation in continuous gravitational-wave searches with pulsar-timing arrays

TL;DR

This paper introduces new techniques to accelerate Bayesian searches for continuous gravitational waves from supermassive black-hole binaries using pulsar timing arrays, making the analysis more efficient and scalable.

Contribution

The paper presents novel methods that reduce the computational complexity of Bayesian gravitational-wave searches, enabling the inclusion of pulsar-term phases efficiently.

Findings

Techniques significantly speed up analysis at low to moderate SNRs (>100x)

Good agreement with full signal template methods in parameter estimation

Facilitates systematic injection and recovery studies for detection criteria

Abstract

We describe several new techniques which accelerate Bayesian searches for continuous gravitational-wave emission from supermassive black-hole binaries using pulsar timing arrays. These techniques mitigate the problematic increase of search-dimensionality with the size of the pulsar array which arises from having to include an extra parameter per pulsar as the array is expanded. This extra parameter corresponds to searching over the phase of the gravitational-wave as it propagates past each pulsar so that we can coherently include the pulsar-term in our search strategies. Our techniques make the analysis tractable with powerful evidence-evaluation packages like MultiNest. We find good agreement of our techniques with the parameter-estimation and Bayes factor evaluation performed with full signal templates, and conclude that these techniques make excellent first-cut tools for detection and characterisation of continuous gravitational-wave signals with pulsar timing arrays. Crucially, at low to moderate signal-to-noise ratios the factor by which the analysis is sped up can be > 100, permitting rigorous programs of systematic injection and recovery of signals to establish robust detection criteria within a Bayesian formalism.