Close pulsar pairs in globular clusters could isolate 10^{-6} solar-mass dark matter subhalos by suppressing nanohertz GWB noise

The authors model timing residuals from a nanohertz gravitational-wave background that is spatially correlated across Earth-pulsar baselines. For pairs closer than the GWB wavelength, the common-mode signal subtracts in the differential residual while a nearby subhalo produces a differential Shapiro delay that survives. They simulate an array of such pairs drawn from dense globular clusters and compare detection thresholds against an otherwise identical set of isolated pulsars. The calculation assumes a cold-dark-matter subhalo mass function normalized to Milky Way satellite counts and includes realistic pulsar noise spectra.

Standard pulsar timing arrays such as NANOGrav and the European PTA already report a stochastic GWB consistent with supermassive black-hole binaries. That background sets a sensitivity floor for substructure searches because its power exceeds the expected dark-matter-induced timing variance at nanohertz frequencies. The close-pair technique exploits the fact that the GWB coherence length exceeds 0.1 pc while subhalo-induced perturbations do not, providing a spatial filter absent from single-pulsar or wide-baseline analyses.

A key limitation is that globular-cluster baryonic dynamics, including stellar encounters and intra-cluster gas, can generate timing noise at similar amplitudes. The paper notes that multi-frequency observations and proper-motion constraints will be required to separate these contributions. Confirmation would need an end-to-end simulation that jointly fits cluster dynamics and a population of subhalos before claiming a dark-matter detection.

Next steps include targeted timing campaigns on known close pairs in 47 Tucanae and M15 with next-generation facilities such as the Square Kilometre Array, which can reach the required microsecond precision at 0.1 pc separations.