The Oh-My-God Particle

Summary

On 15 October 1991 the Fly's Eye detector in Utah recorded a single cosmic-ray particle with an energy of about 3.2 × 10²⁰ electronvolts, or 51 joules. It is still the most energetic cosmic ray on record, and no one knows what kind of particle it was or where it came from.

Physics does not forbid such a particle. It limits how far one can travel, so the source ought to be relatively near, and none has been found. In 2021 the Telescope Array, also in Utah, recorded a second particle almost as energetic, the Amaterasu particle, from another direction with no obvious source. The open question now is what the particles were, because magnetic fields can bend a heavy nucleus far off course.

What is documented

The detector. The Fly's Eye was a University of Utah experiment at the US Army's Dugway Proving Ground, in Utah's West Desert. On clear moonless nights its 880 photomultiplier tubes watched the whole visible sky for the faint fluorescent light of air showers, the cascades of secondary particles that a cosmic ray sets off in the atmosphere. The light recorded along a shower's track, added up, gives the energy of the particle that started it. A partial second eye, Fly's Eye II, stood 3.4 kilometres away (Bird and colleagues, 1995).

The event. At 07:34:16 Universal Time on 15 October 1991 (01:34 in Utah), the Fly's Eye recorded a shower whose axis passed 13.0 kilometres from the main eye. It fell on the blind side of Fly's Eye II, so it was seen from one site only. The shower reached its maximum at an atmospheric depth of 815 grams per square centimetre, and the particle arrived from right ascension 85.2°, declination 48.0°.

The measurement. D. J. Bird and sixteen colleagues published the analysis in The Astrophysical Journal in March 1995. They put the energy at 320 EeV (exa-electronvolts; one EeV is 10¹⁸ electronvolts), uncertain by 92 EeV upward and 94 EeV downward: 51 joules, "substantially greater than the energy of any previously reported cosmic ray." The shower's profile best suited a mid-sized nucleus, but the authors wrote that "it could have been a nucleon or a heavy nucleus. It might even have been a gamma ray."

The limit it seemed to cross. In 1966 Kenneth Greisen, and separately G. T. Zatsepin and V. A. Kuzmin, predicted an end to the cosmic-ray spectrum. As Bird and colleagues explain it, to a proton of such energy the cosmic microwave background looks like a beam of gamma rays, many of them energetic enough to collide with it and make a pion, so it cannot cross intergalactic space without losing energy. This is the Greisen-Zatsepin-Kuzmin, or GZK, limit. The 1991 particle's production site, the authors concluded, "should lie within a distance of about 30 Mpc", roughly 100 million light-years.

The limit is real. The Akeno Giant Air Shower Array (AGASA) in Japan recorded another event above 10²⁰ electronvolts, of 1.7 to 2.6 × 10²⁰ eV (170 to 260 EeV), on 3 December 1993 (Hayashida and colleagues, 1994). The cutoff itself was observed later. The High Resolution Fly's Eye (HiRes), the Fly's Eye's successor, reported a sharp suppression of the flux at about 6 × 10¹⁹ eV, at five standard deviations, and the Pierre Auger Observatory in Argentina reported the flux steepening above 4 × 10¹⁹ eV, "consistent with the prediction by Greisen and by Zatsepin and Kuz'min". Both results were published in 2008. Auger's catalogue of the 100 most energetic events it recorded from 2004 to 2020 runs from 78 to 166 EeV.

The Amaterasu particle. At 10:35:56 UTC on 27 May 2021 a shower triggered 23 of the 507 surface detectors of the Telescope Array, which covers about 700 square kilometres outside Delta, Utah, and is led by the University of Utah and the University of Tokyo. The collaboration put the energy at 244 EeV, with a statistical uncertainty of 29 EeV and a systematic one of 51 EeV upward and 76 EeV downward: about 40 joules, on a scale comparable with the 1991 measurement. The result was published in Science on 24 November 2023, and the researchers named the particle after Amaterasu, the sun goddess of Japanese mythology. Its arrival direction, the paper says, "points back to a void in the large-scale structure of the Universe."

What the discoverers said. John Matthews, co-spokesperson of the Telescope Array, told the University of Utah that for both particles "you trace its trajectory to its source and there's nothing high energy enough to have produced it." His colleague John Belz noted that the two events "seem like they're coming from completely different places in the sky."

Leading explanations

A nearby accelerator not yet identified (the working assumption; no source has been identified). Matthews has said that supernovae "are nowhere near energetic enough for this." Michael Unger and Glennys Farrar (2024), treating Amaterasu as an iron nucleus and modelling the Galactic magnetic field, placed its source within 6.6 per cent of the sky (2,726 square degrees) and at most 8 to 50 megaparsecs away, depending on its true energy. Finding no radio galaxy powerful enough in that volume, they judged the most straightforward explanation to be a transient, such as a tidal disruption event, a young magnetar or a neutron star merger, in an otherwise ordinary galaxy. Farrar (2025) has since proposed that the highest-energy cosmic rays come from binary neutron star mergers, and that those well beyond 100 EeV are heavy nuclei forged there. The idea predicts neutrinos above 10 PeV arriving together with gravitational waves. It is untested. Nadine Bourriche and Francesca Capel (2026), simulating Amaterasu's journey, found candidate sources extending beyond the Local Void, among them the starburst galaxy M82.

Heavy nuclei, bent by magnetic fields (supported by recent studies, not established for these two particles). Magnetic fields bend a heavy nucleus more than a proton of the same energy, so its arrival direction need not point back to its source. The Fly's Eye team could not rule out a heavy nucleus. In 2024 the Telescope Array collaboration reported that its events above 10²⁰ eV do not follow the nearby large-scale structure of galaxies, and inferred a very heavy composition above 100 EeV. In July 2025 the IceCube neutrino observatory, having found no extremely-high-energy neutrinos in 12.6 years of data, put the proton fraction of cosmic rays above about 30 EeV at less than 70 per cent, under an assumption about how their sources evolved. In 2023 Belz had doubted large bending, citing "other observations that show they're not strong enough to produce significant curvature" at these energies.

New physics (untested). The Science paper lists "an incomplete knowledge of particle physics" among its possibilities. Belz mentioned "defects in the structure of spacetime, colliding cosmic strings", and called these "spit-balling crazy ideas". They were offered as ideas, not results.

Measurement error (does not remove the problem). Both energies carry wide uncertainties, and the 1991 shower was seen from one site only. Even at the lower edge of their stated uncertainties both particles lie well above the 40 to 60 EeV at which HiRes and Auger found the flux suppressed.

Current status

Unsolved. Two particles, thirty years apart, each carrying tens of joules, are securely recorded, and the limit that made the first look impossible has been observed and turns out to be a limit on distance. What the particles were and where they came from are not known, and every proposed source is a hypothesis. The case would change with an identified source: a clustering of the highest-energy events around a known object, a neutrino or gravitational-wave signal arriving with one, or composition measurements precise enough to say how far such particles are bent. The expanding Telescope Array and the Pierre Auger Observatory are collecting the data.

Last reviewed: September 2026. Records are provisional. Where the evidence changes, the entry changes. Found an error? Tell us.