The Cosmic-ray Knee as a Local Signature of Nearby PeVatrons
Ke Fang, Francis Halzen · 2026
Read The Cosmic-ray Knee as a Local Signature of Nearby PeVatrons by Ke Fang, Francis Halzen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-05435.txt.
Introduction
Ke Fang and Francis Halzen ask, in a 2026 astrophysics paper, whether the knee in the cosmic-ray spectrum is a local feature. The knee is a sharp steepening near four PeV, where PeV means a very high particle energy. It could reflect the whole Galaxy, or just a few nearby accelerators called PeVatrons.
The authors simulate how cosmic rays travel through Galactic magnetic fields and gas. Comparing those simulations with cosmic-ray and gamma-ray observations favors nearby sources under certain birth rates. Sources about a kiloparsec away, aged a tenth to one million years, fit best. The Cygnus Cocoon stands out as a promising candidate for shaping the locally observed knee.
Catalog record
| CREATOR | Ke Fang, Francis Halzen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 32 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-05435
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Cosmic-ray Knee as a Local Signature of Nearby PeVatrons Ke Fang, Francis Halzen Abstract A “knee” in the cosmic-ray spectrum, characterized by a sudden steepening of the spectral shape at ∼ 4 \sim 4 PeV, may be interpreted either as a global feature of Galactic cosmic rays or as a local signature. In the former scenario, cosmic-ray spectra throughout the Galaxy would be similar to that observed in the solar neighborhood, and the knee would be a common feature of the cosmic-ray sea. In the latter scenario, the PeV cosmic-ray flux varies across the Galactic disk, and the knee is dominantly contributed by a small number of nearby sources. By simulating cosmic-ray propagation in the Galactic magnetic field and interstellar medium, we show that the two scenarios correspond to different regimes of the birth rate of PeV proton accelerators and depend on the presence of powerful nearby sources. By comparison with both cosmic-ray and gamma-ray observations, we find that a local knee would be best explained by sources located at distances of order ∼ 1 \sim 1 kpc and with ages in the range 0.1-1 Myr, with the Cygnus Cocoon being a particularly promising candidate. I introduction The knee of the cosmic-ray spectrum at around 4 PeV ( Hoerandel 2003 ) and the related sharp softening of the proton spectrum at 3 PeV ( Cao et al. 2025 ) indicates that certain Galactic sources, commonly referred to as PeVatrons, accelerate protons up to multi-PeV energies. The spectrum of the accelerated particles, dN/dE\propto E^{-s} , needs to be relatively hard, with s\approx 2.1 - 2.4 ( Blasi & Amato 2012a ) , so that after propagation the observed spectral index reaches s_{\rm obs}\sim 2.7 up to the knee. A direct search for PeVatrons in the ultrahigh-energy (UHE; >100 TeV) gamma-ray sky has recently become possible with gamma-ray air shower observatories. The 1LHAASO catalog reports 43 UHE gamma-ray sources ( Cao et al. 2024a ) . However, most of these sources exhibit softer spectra, and about half of them are associated with pulsars, whose TeV gamma-ray emission is accommodated by inverse Compton scattering by electrons. Only a handful of promising PeVatron candidates have been identified, including the Galactic center ( Abramowski et al. 2016 ) , star-forming region Cygnus Cocoon ( Abeysekara et al. 2021 ; Cao et al. 2024b ) , the supernova remnant G106.3+2.7 ( Amenomori et al. 2021 ; Fang et al. 2022 ) and the microquasar V4641 Sgr ( Alfaro et al. 2024 ; LHAASO Collaboration 2024 ) . The scarcity of PeVatrons in the UHE gamma-ray sky suggests that PeVatrons may be rare compared to the supernova rate, in which case PeV cosmic rays originate from old sources that are no longer active ( Cristofari et al. 2020 ) . Alternatively, they may be inefficient gamma-ray emitters so that the majority lie below the detection threshold; see example Fang et al. 2024 . While GeV–TeV cosmic rays form a smoothly distributed “cosmic-ray sea” in the Galactic disk, the distribution may be less homogeneous at energies near the knee. This can arise from two effects. First, as implied by UHE gamma-ray observations, the population of PeVatrons could be small. Second, PeV cosmic rays have much shorter residence times in the Galaxy than GeV-TeV cosmic rays, making their spatial distribution more dependent on the recent history of sources ( Kaci & Giacinti 2025 ; Stall & Mertsch 2025 ) . As a consequence, localized “cosmic-ray islands” (sometimes referred to in the literature as local cosmic-ray bubbles) may form, characterized by enhanced cosmic-ray density and upward fluctuations in the cosmic-ray spectrum at PeV energies. The possibility that few nearby bright sources contribute to the cosmic-ray knee has been investigated in the literature based on both cosmic-ray spectral and anisotropy measurements. The rapid rise and steepening of the cosmic-ray spectrum at the knee was interpreted as the additional contribution from a single source ( Erlykin & Wolfendale 1997 ) or from several nearby sources ( Sveshnikova et al. 2013 ; Evoli et al. 2021 ) superimposed on a smoothly steepening background component. Consistent with this picture, the large-scale cosmic-ray anisotropy exhibits an amplitude and phase that are essentially independent of energy between 1 and 100 TeV, while at higher energies the amplitude increases monotonically up to \sim 30 PeV where the phase flips compared to lower energies ( Abbasi et al. 2025 ) . This trend is interpreted as the transition from an anisotropy controlled by the local magnetic field below 100 TeV to the dominance of nearby individual sources at PeV energies ( Erlykin & Wolfendale 2006 ; Ahlers 2016 ) . In this paper, we will show that the two phenomena, namely the presence of cosmic-ray islands and a significant contribution from nearby sources, are correlated and naturally arise when PeVatrons have a low birth rate (Section II ). In light of the latest air shower gamma-ray observations, we examine the properties of a potential knee-producing source in Section III . Based on these results, we speculate in Section IV that the Cygnus Cocoon, a promising PeVatron in the Northern sky, may be among the nearby sources that make an important contribution to the knee. II PeV particles in the Galaxy: islands in the cosmic-ray sea Figure 1: Counts of 1 PeV cosmic rays normalized to the maximum pixel (top row) and the maximum fraction originating from a single source (bottom row) for locations across the disk with |z|<0.17 kpc, sampled in spacial bins of 0.4 kpc in the x and y directions. Sources are assumed to follow a pulsar distribution in space and uniform distribution in time, and steadily emit cosmic rays during a lifetime of 100 kyr. The left, middle, and right panels correspond to source birth rates of 10%, 1%, and 0.1% of a baseline, supernova-like rate of 0.01\,\rm yr^{-1} , respectively. The black star marks the position of the Sun. We first investigate how the homogeneity of PeV cosmic-ray density in the Galactic disk depends on distribution and evolution history of the PeVatrons. Because their population remain largely unknown, we treat their birthrate as a free parameter, normalized to the supernova rate of 0.01\,\rm yr^{-1} ( Tammann et al. 1994 ) . The rate is parametrized as a fraction f_{\rm SN} of the supernova rate, with representative values f_{\rm SN}= 10\% , 1\% , and 0.1\% , corresponding to N_{s}=10^{5} , 10^{4} and 10^{3} sources over the past 100 Myr. The generic population of PeVatrons are assumed to inject 1 PeV protons with identical spectra and lifetimes of \Delta t= 100 kyr. These simplifying assumptions reduce the number of parameters which are however sufficient to demonstrate the important features of our proposal. In the simulation, we randomly generate N_{s} sources with positions in the Galactic disk following the pulsar distribution, which is known to correlate with star formation regions ( Faucher-Giguère & Kaspi 2006 ; Blasi & Amato 2012a ) , and with ages drawn uniformly between 0 and 100 Myr. We propagate the injected protons in the Galactic magnetic field using an anisotropic diffusion model, with a parallel diffusion coefficient D_{\parallel}=6.1\times 10^{28}\left({R}/{4\,\rm GV}\right)^{1/3}\,\rm cm^{2}\,s^{-1} ( Strong & Moskalenko 1998 ) , where R=E/Z