An arXiv version 1 preprint reports that fast radio bursts (FRBs) are distributed across the observed sky in a way compatible with statistical isotropy—the tested assumption of an intrinsically isotropic population after the surveys’ observational response is included—once the analysis accounts for a mask around the Galactic plane and survey-specific selection functions. That conclusion is conditional: the selection functions were reconstructed from the observed sky distributions, so they may absorb genuine anisotropy aligned with a survey footprint.
The correction changed the result
The adjustment mattered. In the pure comparison, the data remained strongly inconsistent with isotropy; applying only the Galactic-plane mask reduced the reported statistic by about a factor of three, but the masked scenario still remained strongly anisotropic.
Weighting the data with each survey’s empirically reconstructed selection function drove the reduced statistic down by nearly four orders of magnitude, to about 11.7. Even then, the selection-only scenario remained formally inconsistent with isotropy.
Only the combined, or fiducial, setup brought the main comparison into line with the isotropic mocks. The reduced statistic was 0.62, with a Gaussian-equivalent significance of −0.807, and the observed angular-correlation profile was reported as consistent with the mock distribution at every separation.
A separate absolute-anisotropy check pointed in the same direction: its score was 0.028, its empirical p-value was 0.984 and its Gaussian tension was 1.70. The authors reported this as consistent with the fiducial isotropic mock distribution.
How the comparison was built
The working catalogue contained 4,066 FRB events from multiple radio surveys. The fiducial cut kept events more than 20 degrees from the Galactic plane, leaving 2,518 for the masked analysis. The compilation excluded DSA-110 events, below-threshold CHIME events and records without an assignable reporting survey.
To compare the skies, the researchers used the Landy–Szalay two-point angular correlation function—a measure of how pairs are distributed across angular separations—alongside a separate absolute-anisotropy statistic. Jackknife and bootstrap resampling were used to estimate observational uncertainty.
The isotropic benchmarks came from mock ensembles that propagated uncertainty in the reconstructed selection functions and Poisson fluctuations in event counts. The covariance used for the comparison was stabilized before the resulting significances were calibrated empirically.
The result depends on the model
That modeling choice also sets the study’s main limit. The selection functions were reconstructed from the observed sky because a consistent exposure model across all surveys was unavailable; as a result, a real anisotropy aligned with a survey footprint could be absorbed into the modeled response. The sample’s single-survey-dominated, uneven effective sky coverage further limits its constraining power.
Across 13 independent configurations, none rejected statistical isotropy: the primary significance stayed below 3 and empirical p-values exceeded 0.01. The exact numerical significance varied across the robustness settings, so the result is best read as compatibility with the tested null rather than as a measurement independent of the modeling choice.
Taken together, the analysis supports compatibility with the modeled observational null, not an independent demonstration that the intrinsic FRB population is isotropic. The distinction matters because the selection functions may absorb anisotropy that follows a survey footprint.
Paper data and sources
Original title: Testing Statistical Isotropy in the FRB Sky Distribution: A Selection-Function-Aware Framework
Authors: Bruno W. N. Ribeiro, Thais Lemos, Carlos A. P. Bengaly et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-20
DOI: Not available
Original paper · Full text