Space & Cosmic

The Cosmic-Ray Knee: Origin of a Galactic Speed Limit

By The Unsolved Report Editorial Team · Last reviewed 2026-06-19

The cosmic ray knee origin is one of astrophysics' enduring puzzles: a kink at ~3.7 PeV where the galaxy's particle accelerators seem to hit their limit.

Plot the number of cosmic rays striking Earth against their energy, and you get one of the most reliable graphs in all of physics: a nearly straight, steeply falling line that holds for ten orders of magnitude. Then, at an energy of roughly three to four quadrillion electron-volts, the line bends. It does not break or vanish — it just tilts a little steeper and keeps falling. Astronomers call this subtle kink "the knee," and for more than sixty years it has functioned as a kind of cosmic milepost. Something happens out in the galaxy at that exact energy, and we are still arguing about precisely what.

The Documented Facts

Cosmic rays are not rays at all but particles — mostly bare protons and atomic nuclei — that slam into our atmosphere from space at nearly the speed of light. When one hits the upper air, it triggers a cascade of secondary particles called an extensive air shower, which ground arrays detect.

The knee was discovered this way. In 1958, Georgy Kulikov and German Khristiansen of Moscow State University, using an array of hodoscope counters, noticed a "kink" in the air-shower spectrum corresponding to primary energies of several PeV (a PeV is 10^15 electron-volts). The steepening they found is now universally known as the knee (CERN Courier; University of Siena, Early History of Cosmic Rays).

The feature has since been mapped with far greater precision. The Large High Altitude Air Shower Observatory (LHAASO) in Sichuan, China, published the most detailed measurement to date in Physical Review Letters in 2024. Using its KM2A array, LHAASO located the knee at 3.67 ± 0.05 ± 0.15 PeV. The spectral index — the steepness of the falling line — was measured at −2.7413 below the knee and −3.128 above it (LHAASO Collaboration, Phys. Rev. Lett. 132, 131002). The earlier KASCADE experiment in Germany had placed a similar knee-like feature near 4 PeV (arXiv:1308.2098).

To put that energy in perspective: 3.7 PeV is roughly a million times the energy the Large Hadron Collider gives a single proton. Nature is doing this somewhere in our galaxy for free, and the knee tells us that whatever does it begins to run out of steam right there.

The leading explanation rests on solid, testable physics. Charged particles are accelerated and confined by magnetic fields, and what matters physically is not raw energy but rigidity — roughly, energy divided by charge. The maximum energy a source can deliver therefore scales with a particle's charge number Z. If protons (Z=1) max out near a few PeV, then helium, carbon, and iron (Z=26) should reach their own knees at proportionally higher energies. This is the rigidity-dependent or "Peters cycle" picture, and it makes a sharp prediction: the cosmic-ray mix should grow heavier above the knee (IOPscience, Cosmic-Ray Physics).

That prediction has held up. LHAASO's 2024 data show the mean logarithmic mass of cosmic rays trending toward heavier elements above the knee — exactly what you would expect if light protons are dropping out first and heavier nuclei are filling in behind them (Phys. Rev. Lett. 132, 131002).

There is also a geometric ceiling. The Hillas criterion states that a source can only accelerate a particle as long as the particle's orbit fits inside the accelerating region — formally, maximum energy scales as the product of magnetic field strength and source size (E_max ≈ eBR). For most galactic objects, that math caps acceleration somewhere in the PeV range unless the magnetic field is dramatically amplified (Frontiers in Astronomy and Space Sciences).

Finally, we now have direct evidence that the galaxy contains accelerators reaching these energies. In 2021, LHAASO reported in Nature the detection of a dozen "PeVatrons" — sources emitting gamma rays above 100 TeV — including a photon of about 1.4 PeV, the highest-energy photon ever recorded, from the Cygnus star-forming region. The Crab Nebula was seen emitting photons above 1 PeV with no clear cutoff (Cao et al., Nature 594, 33, 2021).

The Genuine Open Question

Here is the puzzle that keeps the knee unsolved: we can see the galaxy accelerating particles to PeV energies, and we have a clean theory for why a knee should appear — but we have not conclusively matched a specific class of source to the job.

For decades, the prime suspects have been supernova remnants — the expanding blast waves of exploded stars, where diffusive shock acceleration is thought to boost particles to high energy (Astronomy & Astrophysics, arXiv:astro-ph/0303159). The theory is elegant. The trouble is the numbers. Detailed modeling of observed remnants suggests many of them soften or cut off their particle spectra around 100 TeV — roughly a factor of ten below the knee (LHAASO and Galactic Cosmic Rays, PMC). A standard middle-aged remnant, with its shock already slowed, may struggle to reach even 10 TeV.

So a gap remains between what supernova remnants seem to deliver and where the knee sits. The PeVatrons LHAASO found are mostly identified by their gamma rays, which can be produced by both protons (the cosmic rays we want) and electrons (which do not contribute to the knee). Disentangling the two — finding firm proof of hadronic PeV acceleration — is genuinely hard, and as of the most recent reviews, no single galactic source has been nailed down as a confirmed proton PeVatron beyond reasonable doubt (arXiv:2306.01484).

In short: the knee almost certainly marks the maximum energy of the galaxy's accelerators. We just cannot yet point to the machine — or machines — and say that one, working exactly like this, draws the line at 3.7 PeV.

Theories and Interpretations (Labeled Speculation)

The acceleration-limit view (mainstream). The knee is where galactic accelerators top out for protons, with heavier nuclei extending the spectrum to higher energies via the rigidity sequence. This is the consensus reading and is best supported by the composition data (IOPscience review).

The confinement / leakage view (plausible alternative). An overlapping idea holds that the knee partly reflects when the galaxy can no longer contain its cosmic rays — above a certain rigidity, particles leak out of the galactic magnetic field rather than staying trapped. Recent work argues that anisotropy measurements (the slight directional unevenness of arriving particles) may help test whether propagation, not acceleration, drives the bend (Astrophysical Journal study). This remains a credible competitor.

The single-source or exotic proposals (more speculative). A minority of models invoke contributions from one nearby dominant source, or novel particle-physics effects, to shape the knee. These are interesting but carry far less observational backing and should be treated as conjecture.

What makes the cosmic-ray knee such a satisfying mystery is that it is not vague. It is a precise number, measured to a few percent, written into the sky. We know roughly why it should be there. We are simply still hunting for the galactic engines that put it there — and every PeV photon LHAASO catches narrows the search.

Sources & Further Reading

LHAASO Collaboration, "Measurements of All-Particle Energy Spectrum and Mean Logarithmic Mass of Cosmic Rays from 0.3 to 30 PeV with LHAASO-KM2A," Physical Review Letters 132, 131002 (2024). ADS
Cao et al., LHAASO Collaboration, "Ultrahigh-energy photons up to 1.4 petaelectronvolts from 12 γ-ray Galactic sources," Nature 594 (2021). IHEP/LHAASO release
"The origin of cosmic rays," CERN Courier. cerncourier.com
"Early History of Cosmic Ray Research," University of Siena. PDF
"LHAASO and Galactic cosmic rays," PMC review. ncbi.nlm.nih.gov
"Chapter 4: Cosmic-Ray Physics," IOPscience. iopscience.iop.org
"Open Questions in Cosmic-Ray Research at Ultrahigh Energies," Frontiers in Astronomy and Space Sciences. frontiersin.org
"The knee in galactic cosmic ray spectrum and variety in supernovae," arXiv:astro-ph/0303159. arxiv.org
"Search for the Galactic accelerators of cosmic rays up to the knee with the Pevatron Test Statistic," arXiv:2306.01484 (preprint). arxiv.org

Sources & further reading

LHAASO Collaboration, Phys. Rev. Lett. 132, 131002 (2024) — https://ui.adsabs.harvard.edu/abs/2024PhRvL.132m1002C/abstract
Cao et al., LHAASO/Nature 594 (2021), IHEP release — http://english.ihep.cas.cn/lhaaso/News/202110/t20211026_286767.html
The origin of cosmic rays, CERN Courier — https://cerncourier.com/a/the-origin-of-cosmic-rays/
Early History of Cosmic Ray Research, University of Siena — https://galileo.dsfta.unisi.it/images/PSMPDFiles/Early-history-of-CR.pdf
LHAASO and Galactic cosmic rays, PMC — https://pmc.ncbi.nlm.nih.gov/articles/PMC9157250/
Cosmic-Ray Physics review, IOPscience — https://iopscience.iop.org/article/10.1088/1674-1137/ac3faa
Open Questions in Cosmic-Ray Research at Ultrahigh Energies, Frontiers — https://www.frontiersin.org/journals/astronomy-and-space-sciences/articles/10.3389/fspas.2019.00023/full
The knee in galactic cosmic ray spectrum and variety in supernovae, arXiv:astro-ph/0303159 — https://arxiv.org/pdf/astro-ph/0303159
Pevatron Test Statistic search (preprint), arXiv:2306.01484 — https://arxiv.org/pdf/2306.01484
KASCADE-Grande elemental spectra, arXiv:1308.2098 — https://arxiv.org/pdf/1308.2098
Joint constraint on propagation origin of the knee, ApJ — https://iopscience.iop.org/article/10.3847/1538-4357/ae3d2d