Scientists studied subtle, non-random patterns in a primordial gravitational-wave background. They focused on stochastic gravitational-wave backgrounds (SGWBs) that can be produced early in the Universe. Instead of just looking at the usual two-point signal, or power spectrum, the team calculated four-point correlations — called the trispectrum — that capture non-Gaussian statistics, meaning departures from purely random noise.

The paper examines a specific, well-motivated source: gravitational waves produced at second order by vector fluctuations. These vector fluctuations can come from primordial magnetic fields or from dark-sector vector fields created during cosmic inflation. The authors developed tools to compute and describe higher-order gravitational-wave correlators in this setup. Working with vector-induced signals makes some of the integrals factorize, so key time integrals can be done analytically and the structure of the non-Gaussian signal becomes clearer.

Their calculations show two main features. First, the trispectrum amplitude scales like the square of the usual GW power spectrum. Second, the trispectrum peaks for characteristic “folded” momentum shapes — roughly speaking, configurations in which the wavevectors form a folded quadrilateral. Those features reflect the nonlinear way the vector sources generate gravitational waves at second order.

The authors then explored observational consequences. They show that the connected trispectrum adds to the variance of standard two-point correlation measures. In particular, it contributes to the variance of the Hellings–Downs curve, which is the expected angular correlation used by pulsar timing array experiments to search for low-frequency gravitational-wave backgrounds. They also computed four-point overlap reduction functions for ground-based interferometers and constructed an optimal statistical estimator that could be used to search for the connected trispectrum with such detectors.