Single-source-class interpretation of the diffuse astrophysical neutrino flux

Abstract

We explore the interpretation that the diffuse astrophysical neutrino flux is dominated by a single standard candle-like source class. Since recent observations favor a broken power law with a spectral break around 30 TeV, we postulate that the $pγ$ channel is the dominant neutrino production process creating a peak at these energies. We use a SOPHIA-based photo-pion interaction model with a thermal target including high-energy processes, such as multi-pion production, which turns out to be relevant for the interpretation. We demonstrate that target photon temperatures 0.1 to 1 keV are preferred in a multi-parameter fit, whereas the maximal neutrino energies can be limited by A) soft injection spectra, B) a maximal proton energy in the PeV range, or C) magnetic field effects on the secondary muons, pions, and kaons with B in the few 10 kG range. We predict that future measurements, such as of the neutrino flavor composition or neutrino-antineutrino ratio (Glashow resonance), can discriminate scenarios. We also point out that the parameters obtained in our generic approach, such as in the strong magnetic field values, might be indicative for an AGN core origin as a driver of the diffuse flux.

Figures (9)

• Figure 1: Diffuse neutrino flux spectrum for the models with negligible (left, red) and large (right, blue) magnetic field. Solid curves refer to the overall best fit for each of the models (marked as stars in Figs. \ref{['fig:corner_noB']} and \ref{['fig:corner_B']}). Dashed curves refer to alternative fit models with different physical explanations for the break (denoted as white dots in Figs. \ref{['fig:corner_noB']} and \ref{['fig:corner_B']}): on the left, we show a model with fixed proton spectral index $\alpha_p=2$ (small $B$ and proton energy cutoff); on the right, we show a model with extreme magnetic field $B=10^7\,\mathrm{G}$. The collected parameters for all four highlighted scenarios are listed in Table \ref{['tab:models']}. The colored bands indicate the allowed model fluxes at 68% and 95% confidence levels respectively (using 1 d.o.f., corresponding to bin-wise uncertainties). Data points are extracted from Fig. 3 of IceCube:2025tgp (CF).
• Figure 2: Corner plot for the parameter space of the no-magnetic-field model. The contour plots show the marginalized TS in each pair of parameters, while the corner one-dimensional plots show the marginalized TS for each parameter. The threshold values for TS chosen for two and one degrees of freedom correspondingly. Scenario A (B) is marked by a white star (dot); the best-fit case is scenario A.
• Figure 3: Corner plot for the parameter space of the magnetic-field model. The contour plots show the marginalized TS in each pair of parameters, while the corner one-dimensional plots show the marginalized TS for each parameter. The threshold values for TS chosen for two and one degrees of freedom correspondingly. Scenario C (D) is marked by a white star (dot); the best-fit case is scenario C.
• Figure 4: Flavor and neutrino-antineutrino composition at the source (top row) and at detector (middle row), and different parent contributions (bottom row) of the diffuse neutrino flux, for all four benchmark scenarios (in columns). The flavor fraction at the source (top row) is shown in terms of the $\nu_\mu+\overline{\nu}_\mu$ fraction of the whole flux; we also display the reconstruction intervals for pion-beam and muon-damped compositions projected for IceCube-Gen2 IceCube-Gen2:2020qha, and the reference values for pion beam, muon-damped, muon beam, and neutron beam compositions. For the flavor fractions at the detector (middle row), we separate the fractions of $\nu_e$, $\nu_\mu$, $\nu_\tau$, $\overline{\nu}_e$, $\overline{\nu}_\mu$, and $\overline{\nu}_\tau$ at Earth (of the total flux), using the flavor mixing parameters $\theta_{12}=33.7^\circ$, $\theta_{13}=8.5^\circ$, $\theta_{23}=48.5^\circ$, $\delta_{\mathrm{CP}}=177^\circ$ to compute neutrino mixings (global fit, normal ordering, from Esteban:2024eli). Flavor fractions are shown in the energy range where the flux is at most five orders of magnitude below the maximum only. We also highlight in red shadings the energy range where $\nu_\tau$ detection (corresponding to the red curves for the $\nu_\tau$ fractions) from double-bang events is potentially accessible Learned:1994wg, and by a dashed purple line the energy of the Glashow resonance at 6.3 PeV (corresponding to the purple dashed curve for the $\bar{\nu}_e$ fraction), allowing a determination of the $\overline{\nu}_e$ fraction. In the bottom panels, we separate the neutrino fluxes in terms of their parents (muons, pions, kaons, neutrons), which are differently affected by magnetic field effects.
• Figure 5: Predicted flux for tracks and cascades separately, for our scenario C. Track and cascade fluxes are obtained assuming that electron and tau neutrinos and antineutrinos always produce cascades, while muon neutrinos and antineutrinos produce tracks with a probability $p_T=0.8$Palladino:2015zua and cascades with a probability $1-p_T=0.2$. The flux is shown per flavor, i.e., divided by a factor of three in consistency with previous figures. We also show as dashed lines a power-law fit for the cascades ($\Phi_{\nu+\overline{\nu}}^{\rm per\; flavor}\propto E_\nu^{-2.7}$) and the tracks ($\Phi_{\nu+\overline{\nu}}^{\rm per\; flavor}\propto E_\nu^{-2.5}$).