Not just a phase: detecting nanohertz gravitational waves from phase alone

Abstract

Unambiguously detecting a gravitational-wave background (GWB) in pulsar-timing data requires a robust detection statistic. The pulsar timing community currently relies heavily on the cross-correlation-based "optimal statistic." Its design was motivated to avoid false-positive detections from unmodeled pulsar red noise. However, recent work by van Haasteren (2025) suggests that the optimal statistic is not, in fact, optimal. Moreover, the optimal statistic is still susceptible to false positives…