Flat Spectrum Radio Quasars as high-energy neutrino sources

TL;DR

The paper tests whether FSRQs are significant sources of high-energy IceCube neutrinos by cross-correlating IceCat-1 events with optically selected SDSS quasars, separating FSRQs from radio-quiet quasars using CLASS data. A modest 2.7σ correlation is found for the FSRQ subset, while radio-quiet quasars show no signal; a strong 4σ deviation emerges for high-declination neutrino events within 0.7° of bright FSRQs, supporting jet-based neutrino production in FSRQs. The authors estimate that >60% of the IceCube high-energy neutrino events could originate from FSRQs, reconciling the diffuse neutrino flux with AGN demographics and evolution, and suggesting leptohadronic jet mechanisms as the likely origin. These findings emphasize the jet environment as the primary site for neutrino production in luminous quasars and highlight the importance of declination-dependent detector sensitivity in such cross-correlation analyses.

Abstract

The astrophysical sources responsible for the production of high-energy neutrinos remain largely uncertain. The strongest associations suggest a correlation between neutrinos and active galactic nuclei. However, it is still unclear which specific regions and mechanisms of the accreting supermassive black hole are responsible for their production. In this paper we investigate the correlation between the positions of IceCat-1 neutrino events and a large, optically selected quasar catalogue extracted from the Sloan Digital Sky Survey. Within this sample, we distinguish radio-quiet quasars from flat-spectrum radio quasars (FSRQs) based on radio emission data from the Cosmic Lens All Sky Survey (CLASS) catalogue. While all the associations between neutrino events and radio-quiet quasars are consistent with being random matches, FSRQs exhibit a moderately significant correlation (2.7 sigma) with neutrino positions. Additionally, we observe that the distribution of minimum distances between neutrino events and FSRQs differs significantly for events at declinations above and below 20 deg. In particular, using the Kolmogorov-Smirnov test, we find that the high-declination event distribution deviates strongly (4 sigma) from a random distribution. We interpret all these results as an indication that a large fraction of the neutrino events (>60%) observed by IceCube could be produced by the FSRQs and that the emission mechanism is likely related to the relativistic jets rather than the radio-quiet component of these sources, such as the accretion disk or corona.

Figures (7)

• Figure 1: Sky distribution of the data. The small grey dots represent the positions of 105,783 quasars from the S11 catalogue. The small blue crosses indicate the 1,218 CLASH FSRQs. Black circles mark the 33 IceCAT-1 events in the northern sky with the required positional accuracy. Circles with a red centre represent the 15 neutrinos located within our area of interest uniformly covered by the CLASH objects. A red square around a circle indicates that a CLASH FSRQ is included in the nominal error region.
• Figure 2: Comparison of the properties of CLASH and RQ quasars. The distributions of the redshift, mass, bolometric luminosity, and Eddington ratio of the RQ sample (black) and CLASH blazars (blue) are shown. The values of the possible neutrino emitters are given in red (Table \ref{['tab:matches1']}).
• Figure 3: Energy distribution of the 15 selected events (red) compared to the energy distributions of all 267 Icecat-1 events (grey) and those observed in the northern sky (hatched grey).
• Figure 4: p-values for the neutrino--CLASH associations (red) and neutrino--S11-RQ quasar associations (grey). The solid red line, plotted for different flux density limits, represents the probability of randomly finding an equal or greater number of neutrino events with at least one CLASH FSRQ within the error box. The dotted and dashed lines indicate the probabilities that the distribution of minimum distances between neutrinos and CLASH sources arises from a random distribution (KS test) for all 15 events and for the high-declination sample, respectively. The p-values for 100 RQ samples, of the same size as the FSRQ sample at the corresponding radio flux limit, are displayed in grey. The median value, along with the 90th and 99th percentiles, is calculated from 10,000 samples, which are not shown for clarity. We plot the p-values for the most luminous and most massive quasars of the S11 sample in green and blue, respectively.
• Figure 5: Distances between the 15 neutrino events and the closest CLASH FSRQs as a function of the declination (red circles). A square around the circle indicates that the CLASH FSRQ is included in the nominal error region. For each event, in grey, we plot the mean and standard deviation of the distribution of the minimum distances of a quasar sample of the same size with randomized positions.