[Paper Review] Extragalactic neutrino factories

S. Buson, A. Tramacere+7 more|arXiv (Cornell University)|May 18, 2023

Astrophysics and Cosmic Phenomena18 citations

TL;DR

The paper tests for a spatial correlation between 5BZCat blazars and IceCube northern-hemisphere neutrino data, finding a significant all-sky correlation when combined with southern results, implying extragalactic PeVatron blazars as likely neutrino sources.

ABSTRACT

Identifying the astrophysical sources responsible for the high-energy cosmic neutrinos has been a longstanding challenge. In a previous work, we report evidence for a spatial correlation between blazars from the 5th Roma-BZCat catalog and neutrino data of the highest detectable energies, i.e. >0.1 PeV, collected by the IceCube Observatory in the southern celestial hemisphere. The statistical significance is found at the level of 2 x 10^{-6} post-trial. In this work we test whether a similar correlation exists in the northern hemisphere, were IceCube is mostly sensitive to <0.1 PeV energies. We find a consistent correlation between blazars and northern neutrino data at the pre-trial p-value of 5.12 x 10^{-4}, and a post-trial chance probability of 6.79 x 10^{-3}. Combining the post-trial probabilities observed for the southern and northern experiments yields a global post-trial chance probability of 2.59 x 10^{-7} for the genuineness of such correlation. This implies that the spatial correlation is highly unlikely to arise by chance. Our studies push forward an all-sky subset of 52 objects as highly likely PeVatron extragalactic accelerators.

Motivation & Objective

• Motivate the search for astrophysical sources of high-energy neutrinos beyond the diffuse flux.
• Test whether blazars are long-term neutrino emitters by analyzing northern IceCube data in an independent dataset.
• Extend the all-sky blazar–neutrino correlation analysis to the northern hemisphere to compare with southern results.
• Quantify the statistical significance of blazar–neutrino associations using a robust, post-trial framework.

Proposed method

• Use IceCube 9-year northern sky map data (muon-proxy energies ~TeV to <0.1 PeV) and a 5BZCat blazar catalog (δ ≥ -3°, |b| > 10°).
• Identify neutrino spots with high L-values (L = -log p) above L_min thresholds and within a chosen angular association radius r_assoc around blazars.
• Perform a positional cross-correlation allowing at most one-to-one associations, scanning over L_min and r_assoc to obtain a test statistic TS_astro.
• Estimate pre-trial p-values by Monte Carlo randomizations of blazar positions (1,000,000 catalogs) while preserving blazar counts and spatial distribution.
• Compute post-trial p-values by accounting for a-posteriori parameter choices (L_min, r_assoc) and compare with a northern control sample to assess robustness.
• Combine northern and southern results using Fisher’s method to obtain global pre-trial and post-trial significances.

Experimental results

Research questions

• RQ1 Do 5BZCat blazars spatially correlate with IceCube northern hemisphere neutrino hotspots?
• RQ2 Is the northern hemisphere correlation consistent with the southern hemisphere results, supporting a population of extragalactic PeVatron accelerators?
• RQ3 What is the statistical significance of the blazar–neutrino association after accounting for trials and multiple parameter choices?
• RQ4 How does the all-sky blazar/neutrino correlation inform the nature of PeVatron blazars and their neutrino emission across the sky?

Key findings

• Northern analysis finds a pre-trial p-value of 5.12×10^-4 and a post-trial p-value of 6.79×10^-3 for the best parameters (L_min = 3.0, r_assoc = 1.25°).
• The northern neutrino hotspot sub-sample with L ≥ 3.0 contains 66 spots, with 42 matches to 5BZCat blazars.
• Across the northern and southern hemispheres, the combined global post-trial p-value is 2.59×10^-7, corresponding to 5.02σ significance.
• Table shows 42 northern matches and 66 hotspots with 5BZCat associations, versus 10 southern matches and 19 hotspots with 3×10^-7 pre-trial and 2×10^-6 post-trial values (Paper I).
• The results support the existence of an all-sky subset of 52 objects as highly likely PeVatron extragalactic accelerators.