Impact of Correlated Test-Mass Noise on Stochastic Gravitational-Wave Background Searches with Space-Based Interferometers

Abstract

The stochastic gravitational-wave background (SGWB) is a major science target of future space-based interferometers because it encodes information about unresolved astrophysical populations and physical processes in the early Universe. Its stochastic nature, however, means that its detection and parameter inference rely on distinguishing its statistical contribution to the data from that of instrumental noise. In particular, both the SGWB and instrumental noise contribute to the data covariance, so an incomplete noise model can be misinterpreted as part of the SGWB and thereby bias the inferred spectrum and its parameters. Here we investigate this effect in the presence of correlations between the acceleration (ACC) noises of adjacent test masses (TMs) in each spacecraft. We simulate the correlated noise at the single-link level, propagate it through the moving-orbit response into the time-delay interferometry (TDI) observables, and analyze the resulting TDI auto spectra with two Bayesian frequency-domain templates: the Corr. Template includes the correlated-noise covariance, whereas the Non-Corr. Template neglects it. We find that, when correlated ACC noise is present in the mock data, the Non-Corr. Template exhibits systematic parameter offsets that generally become more pronounced with increasing correlation amplitude. In contrast, the Corr. Template recovers the injected SGWB and instrumental-noise parameters substantially more accurately. These results demonstrate that correlations in test-mass ACC noise can constitute a relevant systematic for space-based SGWB searches and motivate their explicit inclusion in mission-realistic noise models and inference pipelines.