Uncovering stochastic gravitational-wave backgrounds with LISA

TL;DR

This paper proposes a Bayesian model selection strategy to detect stochastic gravitational-wave backgrounds with LISA, addressing the challenge of unknown noise characteristics and exploring the detectable parameter space for various SGWB models.

Contribution

It introduces a novel Bayesian approach using flexible noise models and transfer functions to identify SGWBs with LISA despite unknown noise PSD.

Findings

Effective detection of SGWBs with unknown noise PSD

Probing LISA's sensitivity to power-law SGWB shapes

Demonstrating the method's robustness in simulated scenarios

Abstract

Finding a stochastic gravitational-wave background (SGWB) of astrophysical or primordial origin is one of the quests of current and future gravitational-wave observatories. While detector networks such as LIGO-Virgo-Kagra or pulsar timing arrays can use cross-correlations to tell instrumental noise and SGWB apart, LISA is likely to be the only flying detector of its kind in 2035. This particularity poses a challenge for data analysis. To tackle it, we present a strategy based on Bayesian model selection. We use a flexible noise power spectral density~(PSD) model and the knowledge of noise and signal transfer functions to allow SGWBs detection when the noise PSD is unknown. With this technique, we then probe the parameter space accessible by LISA for power-law SGWB shapes.