What happens when the second most powerful cosmic ray ever recorded seems to arrive from… nowhere?
Welcome to FreeAstroScience, where we break down complex scientific ideas into clear, human language — because we believe the sleep of reason breeds monsters, and your mind deserves to stay sharp. Whether you're a physics student, a curious space enthusiast, or someone who simply loves a good cosmic mystery, this article is for you.
In May 2021, a particle slammed into the Utah desert with roughly 40 million times the energy of anything the Large Hadron Collider has ever produced. Scientists named it the Amaterasu particle, after the Japanese sun goddess. And when they traced its path backward through space, they found something unsettling: it seemed to come from a vast, nearly empty region called the Local Void.
How does a particle carry that much energy if nothing out there could have launched it?
A new study published on January 28, 2026, in The Astrophysical Journal offers a fresh answer — and it points toward a real, active galaxy we can actually see. Stay with us. This story gets better the deeper we go.
📖 Table of Contents
What Exactly Is the Amaterasu Particle?
Let's start with the basics.
Cosmic rays aren't really "rays" at all. They're charged particles — mostly protons and stripped atomic nuclei — that race through the cosmos at nearly the speed of light . They slam into Earth's atmosphere millions of times every day. Our planet's magnetosphere deflects most of them, but some punch straight through to the surface .
Most cosmic rays carry modest energies. The rare ones, though, carry energies so extreme that physicists have given them a special label: ultra-high-energy cosmic rays (UHECRs), defined as particles with energies above 10¹⁸ electronvolts (eV) .
The Amaterasu particle sits at the very top of this already extreme category. Detected on May 27, 2021, by the surface detector of the Telescope Array in Utah's west desert, it arrived with a measured energy of approximately 244 ± 29 EeV (that's 244 × 10¹⁸ eV) . Its arrival direction was recorded at right ascension 255.9° and declination 16.1° .
Why Does Its Energy Matter So Much?
To give you a sense of scale: the Large Hadron Collider, the most powerful particle accelerator humanity has ever built, smashes protons together at about 6.5 TeV per beam. The Amaterasu particle carried roughly 40 million times that energy — packed into a single subatomic particle .
Only one cosmic ray has ever been detected with more energy: the legendary Oh-My-God particle, recorded in 1991 at about 320 EeV. That makes Amaterasu the second most energetic cosmic ray ever observed and the highest-energy particle ever recorded by a currently operating observatory .
Here's why that matters: at these extreme energies, particles can't travel very far. They lose energy through interactions with the cosmic microwave background — a process described by the GZK limit (named after physicists Greisen, Zatsepin, and Kuzmin) . So the source must be relatively nearby, cosmically speaking. That narrows the search.
And yet, when scientists traced Amaterasu's path backward, the trail pointed toward… almost nothing.
| Property | Value | Context |
|---|---|---|
| Detection date | May 27, 2021 | Telescope Array, Utah, USA |
| Nominal energy (Enom) | 244 ± 29 EeV | 2nd highest ever recorded |
| Lower systematic energy (Elow) | 168 ± 29 EeV | Accounts for heavier composition |
| Arrival direction (R.A., Dec.) | (255.9° ± 0.6°, 16.1° ± 0.5°) | Points near the Local Void |
| Energy comparison | ~40 million × LHC energy | Single subatomic particle |
| Assumed source composition | Iron nucleus (Z = 26) | Heavier nuclei accelerate more easily |
Data: Bourriche & Capel 2026, ApJ, 997:264
The Local Void Problem — Arriving from Nowhere?
When the Telescope Array team published their detection in November 2023 (Abbasi et al., Science), the results were puzzling .
They backtracked the particle's arrival direction through models of the Milky Way's magnetic field. The trail led to a region known as the Local Void — a vast, low-density bubble of space sitting right next to our Local Group of galaxies . The Local Void stretches across tens of megaparsecs and contains very few known galaxies.
In plain language: the particle seemed to come from a cosmic desert.
That's a problem. UHECRs at 244 EeV need an extraordinarily powerful engine to reach those energies. Think supermassive black holes, violent starburst galaxies, or cataclysmic astrophysical explosions. The Local Void doesn't host any obvious candidates .
This mismatch triggered a wave of creative — sometimes exotic — proposals. Researchers suggested everything from past transient astrophysical events that have since faded, to magnetic monopoles, Lorentz invariance violation, and even superheavy dark matter decay as potential explanations .
But Nadine Bourriche and Francesca Capel at the Max Planck Institute for Physics in Garching, Germany, asked a simpler question: What if the particle didn't actually come from the Local Void at all?
What if magnetic fields bent its path enough that its true origin lies somewhere else entirely?
How Simulations and Statistics Cracked the Case
Previous attempts to trace Amaterasu relied heavily on backtracking — essentially rewinding the particle's path through a model of the Galaxy's magnetic field. That approach is useful, but it treats the energy constraint and the directional constraint separately. It also tends to underestimate how much a particle's path can curve over millions of light-years .
Bourriche and Capel took a different road.
They built a simulation-based inference framework that combines two powerful tools :
CRPropa 3 — a sophisticated code that simulates cosmic ray propagation in full 3D, including all the ways particles lose energy (photopion production, photodisintegration, electron-pair production, nuclear decay) and all the ways magnetic fields bend their trajectories .
Approximate Bayesian Computation (ABC) — a statistical method that doesn't require writing down an explicit mathematical likelihood function. Instead, it runs thousands of simulations, compares each one to the actual observation, and keeps only the ones that match closely enough .
How Does This Work in Practice?
Imagine you're trying to figure out where a ball was thrown from, but you can only see where it landed and how fast it was moving when it hit the ground. Oh, and the ball bounced off invisible walls on the way.
That's essentially the challenge. Bourriche and Capel handled it like this :
- They picked random source locations, source energies, and magnetic field properties from broad prior distributions.
- For each combination, they launched one million simulated iron nuclei from that source.
- They checked whether any simulated particle arrived at Earth with an energy and direction matching Amaterasu (within 3 standard deviations).
- If yes, they kept those parameters. If no, they discarded them.
Out of all the proposed combinations, only about 0.04% produced a match — making the computation extremely demanding . The team ran their simulations on the Raven and Viper supercomputers at the Max Planck Computing and Data Facility .
The beauty of this approach? It handles energy and direction simultaneously. A particle from a distant source has more time to lose energy and get deflected. Those two effects are correlated, and the ABC method captures that naturally .
Why Magnetic Fields Change Everything
Here's something most people don't realize about cosmic rays: they don't travel in straight lines.
Because UHECRs are electrically charged, magnetic fields curve their paths — like a prism bending light, but in three dimensions and across millions of light-years. Two types of magnetic fields matter here :
The Galactic Magnetic Field (GMF): Inside our Milky Way, magnetic fields are strong — on the order of several microgauss (μG). They include a large-scale coherent component that bends all charged particles in a systematic way, plus turbulent components that scatter them randomly. Bourriche and Capel used the UF23 base model for the coherent field, one of the most up-to-date Galactic magnetic field models available .
The Extragalactic Magnetic Field (EGMF): Between galaxies, magnetic fields are much weaker — estimated below ~10 nanogauss (nG) in filaments and even less in voids . But over long distances, even these faint fields add up.
The estimated deflection angle from extragalactic fields follows this relationship :
Extragalactic Magnetic Field Deflection Estimate
Eq. (1), Bourriche & Capel 2026, adapted from Harari et al. 2002
For an iron nucleus (Z = 26) at 244 EeV traveling from a source 10 Mpc away through a 1 nG field, this formula gives a deflection of a few degrees . That might sound small — but when you're trying to pin down a source in a sky full of galaxies, a few degrees can shift you from an empty void to a bustling starburst galaxy.
And that's exactly what happened.
The researchers found that the GMF and EGMF contribute comparable deflections for their accepted parameter sets, with the Galactic field generally dominating but the extragalactic field occasionally producing even larger bends — especially for heavier particles in the lower-energy scenario .
What Did the Researchers Actually Find?
The results, published on January 28, 2026, paint a much richer picture than the original "Local Void" story .
Bourriche and Capel ran their analysis under two energy assumptions :
- Case 1 (Nominal): E = 244 ± 29 EeV — the reported value assuming a proton or light nucleus
- Case 2 (Low): E = 168 ± 29 EeV — the lower end of the systematic range, accounting for a possible heavier nucleus
They also split their results by what the particle looked like when it arrived at Earth (not at the source — it may have broken apart during its journey) :
- Light nuclei: mass number A ≤ 4 (hydrogen, helium)
- Medium nuclei: 4 < A ≤ 28 (up to silicon)
- Heavy nuclei: A > 28 (iron-group elements)
The Big Picture
For the nominal energy case, the most probable source distance sits around ~3 Mpc (roughly 10 million light-years). At closer distances, the composition is diverse — light, medium, and heavy nuclei all appear. But beyond 6 Mpc, heavy arrival compositions dominate overwhelmingly .
For the low-energy case, the peak distance shifts to ~4 Mpc for light and medium groups, but heavy nuclei show a relatively flat distribution stretching out to 15 Mpc .
Here's the key finding: about 78% of accepted simulations in the nominal case arrived as heavy nuclei, and about 73% in the low-energy case . The data strongly prefer a heavier arrival composition.
And when the particle is heavy, its path bends more. The posterior distributions — the 3D probability maps of where the particle could have originated — stretch well beyond the Local Void .
As Bourriche herself stated: "Our results suggest that, rather than originating in a low-density region of space like the Local Void, the Amaterasu particle is more likely to have been produced in a nearby star-forming galaxy such as M82" .
M82 — Is the Cigar Galaxy the Smoking Gun?
Among all the candidate sources, one stands out: Messier 82, better known as the Cigar Galaxy .
M82 is a powerful starburst galaxy — a galaxy undergoing an intense burst of star formation. It sits just 3.6 Mpc (about 12 million light-years) from Earth . It lies only a few degrees from the Telescope Array hot spot, a known concentration of high-energy cosmic ray detections .
Why is M82 such an appealing candidate?
- It's the brightest starburst galaxy within the distance range considered, accounting for over 18% of the total 1.4 GHz radio flux from nearby starburst galaxies . Radio flux at that frequency is commonly used as a proxy for cosmic ray production.
- In the nominal energy case with a heavy arrival composition, M82 falls within the 70% highest posterior density contour — meaning it's in the sweet spot of probable source locations .
- Starburst galaxies like M82 are rich in young, massive stars. They drive powerful superwinds — outflows fueled by the combined action of stellar winds and supernovae — that could accelerate particles to extreme energies .
M82 has been a long-standing suspect in the UHECR mystery, but previous analyses of the Amaterasu particle specifically hadn't highlighted it . The reason? Those earlier studies didn't account for the full 3D deflection picture or the joint constraints from energy and direction working together.
There is a catch, though. M82 doesn't fall within the 90% contours for the low-energy case (168 EeV), regardless of the assumed arrival particle type . So if Amaterasu's true energy turns out to be at the lower end of the systematic range, M82 becomes harder to justify as the source.
Other Source Candidates on the Map
M82 isn't alone on the suspect list. Several other galaxies sit within the posterior probability volumes :
| Galaxy | Type | Distance | Energy Case | Notes |
|---|---|---|---|---|
| M82 | Starburst | 3.6 Mpc | Nominal (244 EeV) | Brightest SBG; 18%+ of 1.4 GHz flux; within 70% contour (heavy) |
| NGC 6946 | Starburst | ~7.7 Mpc | Both cases | Within 30% contour (heavy & medium); only 3% of radio flux |
| NGC 2403 | Starburst / AGN | ~3.2 Mpc | Nominal | Within 90% contour (heavy composition) |
| NGC 7331 | Starburst | 14.71 Mpc | Low (168 EeV) | Within 10% contour; moderate radio flux |
| NGC 891 | Starburst | ~9.8 Mpc | Low | Within 90% contour |
| Maffei 2 | Starburst | ~3.4 Mpc | Low | At the edge of 90% contour (medium & heavy) |
| NGC 6789 | Irregular dwarf | ~3.6 Mpc | Both cases | Quiescent; less likely as active source |
Source: Bourriche & Capel 2026, ApJ, 997:264 — Figures 3 & 4
A few things stand out from this list. NGC 6946 appears in both energy cases, making it a persistent candidate — though its relatively low radio luminosity (only 3% of the total starburst radio flux in this volume) works against it . NGC 7331, at nearly 15 Mpc, enters the picture only in the low-energy scenario .
We should also mention a wilder possibility: Amaterasu might have been launched by a past transient event — something that happened millions of years ago and has since disappeared. The average travel time for accepted particles in this study is about 10 million years for the nominal case and roughly 50 million years for the low-energy case . Long-duration gamma-ray bursts, tidal disruption events, and binary neutron star mergers are all on the table as possible one-time cosmic accelerators .
What Comes Next for UHECR Science?
This research opens doors, but it also reveals how much we still don't know.
The team was limited to source distances of 12–15 Mpc due to computational costs — extending the search volume would likely reveal even more candidates . They also used a single Galactic magnetic field model (UF23 base), though at least eight different models exist, each giving slightly different deflection patterns . Future work will need to test all of them.
Perhaps most pressing: we still can't measure what a UHECR is made of when it arrives. Was it a proton? A nitrogen nucleus? An iron nucleus? The answer changes everything — from how much the path bends to how far the source could be .
That's where upcoming observatory upgrades come in :
- TAx4 — an expansion of the Telescope Array that will quadruple its detection area
- AugerPrime — an upgrade to the Pierre Auger Observatory in Argentina that will add new detectors to better determine the mass composition of incoming cosmic rays
- POEMMA and GRAND — next-generation facilities designed to observe UHECRs from space and with massive ground arrays
These instruments, paired with neural network-based data analysis methods already showing results, will sharpen our picture of the cosmic ray spectrum and its composition at the highest energies .
As Francesca Capel, who leads the "Astrophysical Messengers" group at the Max Planck Institute for Physics, explained: "Exploring ultra-high-energy cosmic rays helps us to better understand how the Universe can accelerate matter to such energies, and also to identify environments where we can study the behavior of matter in such extreme conditions" .
Final Reflections
Let's step back and see the full picture.
For three years, the Amaterasu particle has been one of astrophysics' most tantalizing puzzles. A particle of almost incomprehensible energy, seemingly arriving from a cosmic void where nothing should be powerful enough to create it. The mystery sparked proposals ranging from exotic new physics to superheavy dark matter.
Now, Nadine Bourriche and Francesca Capel have shown us something remarkable: when we model cosmic ray propagation more carefully — in full 3D, with realistic magnetic fields, and with energy and direction constraints working together — the void isn't the only answer . Real galaxies with real astrophysical engines sit in the probability volumes. M82, one of the most vigorous star-forming galaxies in our neighborhood, stands as a prime candidate .
That doesn't mean the case is closed. We still can't rule out the Local Void. We still don't know Amaterasu's true composition. And larger search volumes may yet reveal sources we haven't considered. Science doesn't hand us neat endings — it hands us better questions.
But this study represents a genuine step forward. It gives us a reusable framework — a kind of GPS for tracing extreme cosmic rays back to their birthplaces. And as new observatories come online over the next decade, this method will only grow sharper.
Here at FreeAstroScience.com, we believe that understanding the universe isn't a luxury — it's a necessity. Every time we look up and ask "where did that come from?", we exercise the very thing that makes us human: curiosity. We explain complex science in clear terms not to simplify your world, but to expand it.
Never turn off your mind. Keep it active, always — because the sleep of reason breeds monsters.
Come back soon. There's always more cosmos to explore.
Sources
Bourriche, N. & Capel, F. (2026). "Beyond the Local Void: A Data-driven Search for the Origins of the Amaterasu Particle." The Astrophysical Journal, 997:264. Published January 28, 2026. DOI: 10.3847/1538-4357/ae2c89
Williams, M. (2026). "Scientists Continue to Trace the Origin of the Mysterious 'Amaterasu' Cosmic Ray Particle." Universe Today. Published February 13, 2026. universetoday.com
Written for FreeAstroScience.com by Gerd Dani — President and curator of Free AstroScience, a science and cultural group committed to making knowledge accessible to everyone.
Post a Comment