Uncovering gravitational-wave backgrounds from noises of unknown shape with LISA

TL;DR

This paper proposes a robust method for detecting stochastic gravitational-wave backgrounds with LISA, even when the instrumental noise spectrum is unknown and arbitrary, by using flexible modeling and realistic simulations.

Contribution

It introduces a novel approach that models unknown instrumental noise spectra with splines, enabling detection of GW backgrounds without assuming known noise shapes.

Findings

Detects GW backgrounds comparable to fixed-shape models.

Probes new regions of GW background parameter space.

Demonstrates robustness against unknown instrumental noise spectra.

Abstract

Detecting stochastic background radiation of cosmological origin is an exciting possibility for current and future gravitational-wave (GW) detectors. However, distinguishing it from other stochastic processes, such as instrumental noise and astrophysical backgrounds, is challenging. It is even more delicate for the space-based GW observatory LISA since it cannot correlate its observations with other detectors, unlike today's terrestrial network. Nonetheless, with multiple measurements across the constellation and high accuracy in the noise level, detection is still possible. In the context of GW background detection, previous studies have assumed that instrumental noise has a known, possibly parameterized, spectral shape. To make our analysis robust against imperfect knowledge of the instrumental noise, we challenge this crucial assumption and assume that the single-link interferometric noises have an arbitrary and unknown spectrum. We investigate possible ways of separating instrumental and GW contributions by using realistic LISA data simulations with time-varying arms and second-generation time-delay interferometry. By fitting a generic spline model to the interferometer noise and a power-law template to the signal, we can detect GW stochastic backgrounds up to energy density levels comparable with fixed-shape models. We also demonstrate that we can probe a region of the GW background parameter space that today's detectors cannot access.