[Paper Review] TeV-PeV neutrinos over the atmospheric background: originating from two groups of sources?

Hao-Ning He, Ruizhi Yang+2 more|arXiv (Cornell University)|Jul 4, 2013

Astrophysics and Cosmic Phenomena4 citations

TL;DR

This paper proposes that the observed TeV-PeV neutrinos detected by IceCube originate from two distinct astrophysical source populations: a flatter spectrum with a cutoff near 250 TeV, likely from proton-proton collisions in starburst galaxies, and a sharp peak at ~1 PeV, possibly from photomeson interactions in active galactic nuclei. The non-detection of neutrinos between 250 TeV and 1 PeV disfavors a single power-law spectrum at ~2σ significance, suggesting a two-component model is more consistent with current data.

ABSTRACT

In addition to the two ~1 PeV neutrinos, the IceCube Collaboration recently reported a detection of 26 neutrino candidates at energies from 30 TeV to 250 TeV, implying a confidence level of 4.3σover the atmospheric background. We suggest that these TeV-PeV non-atmospheric neutrinos may originate from two groups of sources, motivated by the non-detection of neutrinos in the energy range 250 TeV- 1 PeV in current data. If intrinsic, the non-detection of 250 TeV-1 PeV neutrinos disfavors the single power-law spectrum model for the TeV-PeV non-atmospheric neutrinos at a confidence level of ~ 2σ. We then interpret the current neutrino data with a two-component spectrum model. One has a flat spectrum with a cutoff at the energy ~ 250 TeV and the other has a sharp peak at ~1 PeV. The former is likely via pp collision while the latter may be generated by the photomeson interaction.

Motivation & Objective

• To investigate whether the observed TeV-PeV non-atmospheric neutrino excess in IceCube data can be explained by a single power-law spectrum.
• To assess the statistical significance of the non-detection of neutrinos in the 250 TeV–1 PeV energy range.
• To explore the possibility that the neutrino spectrum consists of two components with distinct spectral shapes and origins.
• To identify plausible astrophysical sources for each component based on spectral features and physical mechanisms.

Proposed method

• The authors analyze IceCube's 662-day high-energy contained vertex search data, focusing on 26 events between 30 TeV and 250 TeV and two 1 PeV events.
• They compare the observed event distribution to atmospheric background expectations (10.6 ± 3.9 events) to assess significance (4.3σ excess).
• A single power-law spectral model is fitted to the data, yielding a spectral index Γ ≈ 2.2, which predicts ~9 events in the 250 TeV–1 PeV range.
• The absence of observed events in the 250 TeV–1 PeV band leads to a statistical exclusion of the single power-law model at ~2σ confidence level.
• A two-component model is proposed: one with a flat spectrum and exponential cutoff at ~150 TeV (from pp collisions), and another with a sharp peak at ~1 PeV (from pγ interactions).
• The model is tested against observed fluxes: the 1 PeV component matches predictions from AGN-accelerated protons interacting with extragalactic background light (EBL), as in Kalashev et al. (2013).

Experimental results

Research questions

• RQ1 Is the observed TeV-PeV neutrino excess consistent with a single power-law spectrum, or does the data suggest spectral structure?
• RQ2 What is the statistical significance of the non-detection of neutrinos between 250 TeV and 1 PeV in the current IceCube data?
• RQ3 Can the observed neutrino flux and spectral shape be explained by two distinct astrophysical sources with different emission mechanisms?
• RQ4 What physical processes—pp collisions or pγ interactions—best explain the two components of the observed spectrum?

Key findings

• The non-detection of neutrinos in the 250 TeV–1 PeV energy range disfavors a single power-law spectrum for the TeV-PeV neutrinos at a confidence level of approximately 2σ.
• The data are better described by a two-component model: one with a flat spectrum and cutoff at ~150 TeV, and another with a sharp peak at ~1 PeV.
• The sub-PeV component (cutoff at ~150 TeV) is consistent with neutrino production via pp collisions in starburst galaxies with high Type II supernova rates and dense gas.
• The PeV neutrino excess (peak at ~1 PeV) is best explained by photomeson interactions (pγ process) in AGNs, where protons interact with extragalactic background light photons, as modeled by Kalashev et al. (2013).
• The predicted flux from the AGN pγ model (0.5–2 × 10⁻⁸ GeV cm⁻² s⁻¹ sr⁻¹ at 1 PeV) is consistent with the observed flux of (3.28 ± 2.28) × 10⁻⁸ GeV cm⁻² s⁻¹ sr⁻¹ in the 1–2 PeV range.
• The authors caution that the current evidence is based on a small sample of ~18 non-atmospheric neutrino events, and future data from IceCube will be critical to confirm or rule out the two-component hypothesis.