Assessing the Credibility of Gamma-Ray QPO Candidates in 41 TeV-selected Blazars

Le résumé fourni par la source

Abstract We analyze quasi-periodic oscillation (QPO) candidates in the 0.1–100 GeV Fermi-LAT light curves of 41 TeV blazars with a conservative, multistage timing analysis combining FFT power spectra, red-noise Monte Carlo simulations, false-discovery-rate control, bootstrap period uncertainties, split-sample cross-validation, Lomb–Scargle and WWZ diagnostics, a detection-only sensitivity test, observing-window checks, and injection–recovery experiments. A first-pass FFT search flags 16 sources above a raw 95% level and six strong candidates after data-quality and split-sample filtering, none satisfying a fully automated multimethod tier. Source-level significance is quantified with forward simulations that preserve the observed sampling and non-Gaussian flux distribution, treating the red-noise slope as a nuisance parameter and never assuming independence of periodogram powers. At simulation-selected reference slopes, only two of the 41 sources reach a raw source-level p < 0.05, only J1555.7+1111 reaches Benjamini–Hochberg significance, and none survive Benjamini–Yekutieli correction. No source survives the conservative slope envelope. Even J1555.7+1111, which the first-pass search flagged as significant but did not rank among its strong candidates, has a global p -value that rises to 7.5 × 10 −3 under the least favorable slope (∼0.31 after the full-sample search), so it is not a robust detection. With substantial period uncertainties, a weak observing window, and limited injection–recovery sensitivity, we establish no robust gamma-ray QPO detection: a sampling-faithful red-noise treatment substantially contracts the candidate set, providing an explicit assessment of QPO credibility.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.