Search for neutrino emission from blazar $γ$-ray flares accounting for possible neutrino time delays

TL;DR

This work tests whether high-energy neutrinos detected by IceCat-1 are temporally associated with gamma-ray flares from blazars when allowing for a jet-frame delay $t^{\prime}_{\mathrm{delay}}$ due to proton acceleration or $p\gamma$ energy losses. By cross-matching IceCat-1 alerts with MOJAVE-based H21+ and CGRaBS-based Rodrigues 2023 (R24+) blazar samples and leveraging Fermi-LAT light curves, the authors scan a wide range of $t^{\prime}_{\mathrm{delay}}$ and compute a global test statistic TS=$(\sum_{\nu} w_{\nu}(t^{\prime}_{\mathrm{delay}}))$ that aggregates per-alert weights incorporating spatial, temporal, and flux information. Across both source lists, the most significant pre-trial signals reach about $2\sigma$, but after trial corrections the post-trial $p$-values are 0.11–0.14, indicating no significant association between IceCat-1 neutrinos and the studied blazar samples. The results imply that either blazars contribute only a small fraction of IceCube’s $\gtrsim 100$ TeV neutrinos or that the assumed universal jet-frame delay and Doppler-factor uncertainties obscure any potential signal, highlighting the need for larger, more precise catalogs and next-generation neutrino observatories to resolve the role of AGNs in the high-energy neutrino flux.

Abstract

We report the results of the search for the high-energy neutrino emission associated with blazar flares, accounting for a possible lag of neutrinos with respect to the electromagnetic emission, either due to the slowness of the proton energy losses in $pγ$ collisions and/or proton acceleration. We perform two tests, cross-matching neutrinos with energies $E_ν \gtrsim 100$~TeV from the public catalogue of neutrino alerts IceCat-1 with active galactic nuclei from two source samples based on 1) the MOJAVE database and 2) the CGRaBS catalogue, and utilising Fermi-LAT light curves from the public light curve repository. We scan over a wide range of values of the jet-frame time delay $t^{\prime}_{\mathrm{delay}}$ between the neutrino arrival and the time of the prior major $γ$-ray flare and find a pre-trial $\sim 2σ$ correlation at $t^{\prime}_{\mathrm{delay}} \sim 10^{3}$ d, which is consistent ($p_{\mathrm{post-trial}} \sim 0.1$) with expectations under the null hypothesis after trial correction.

Figures (3)

• Figure 1: An example of the graph of the temporal part $G_{\nu a}$ of weights defined in Eq. (\ref{['eq:lagging_weight']}) as a function of $t^{\gamma \max}_{a}$ for a fixed $t_{\nu}$ (shown as the red solid vertical line) and a set of fixed $t_{a\mathrm{\, delay}}$ (see the legend).
• Figure 2: $p$-value (see Eq. \ref{['eq:p_value']}) as a function of the jet-frame time delay assumed to be universal for all AGNs for the test with the sources from H21+ (dotted curve) and sources from R24+ (dashed-dotted curve).
• Figure 3: Light curves of 3C 454.3 (upper panel), and 3C 279 (lower panel), which are among the most significant associations between IceCat-1 neutrinos and AGNs in both the tests (see Tables \ref{['tab:top_ten_Homan']} and \ref{['tab:top_ten_Rodrigues']}). The connected markers with error bars show monthly-binned Fermi-LAT photon fluxes in the energy range 100 MeV--100 GeV from the public repository Fermi-LAT:2023iml (ordinate axis). Dotted vertical lines indicate the time $t_{\nu}$ of the neutrino arrival. Time intervals between the alert and the global maximum of the light curve prior to $t_{\nu}$ are shown as double-pointed arrows. Vertical upper dashed risks indicate the anticipated position of the maximum with the Doppler factor from R24+ and $\hat{t}^{\prime}_{\mathrm{delay}} = 1.9 \times 10^{3}$ d, and vertical lower dashed-dotted risks --- with the Doppler factor from H21+ and $\tilde{t}^{\prime}_{\mathrm{delay}} = 7.7 \times 10^{3}$ d. The width of the Gaussian temporal factor $G_{\nu a}$ in weights is equal to the distance between the vertical risks and the line of $t_{\nu}$ (see Eq. \ref{['eq:lagging_weight']}).