The many photomultiplier tubes of the HiRes detector captured tell-tale signs of cosmic rays.

An experiment to detect subatomic particles arriving from deep space has triumphantly announced … their absence.

The finding, a swansong from the now defunct High Resolution Fly’s Eye (HiRes) cosmic-ray observatory in Utah, is far from a disappointment. It is the long-awaited confirmation of a decades-old prediction that there is a critical threshold of energy beyond which these cosmic rays dwindle in number (R. U. Abbasi et al. Phys. Rev. Lett. 100, 101101; 2008 ). And it adds weight to initial measurements from the Pierre Auger Cosmic Ray Observatory in Argentina.

This energy ‘cut-off’ was predicted in 1966 by Kenneth Greisen of Cornell University in Ithaca, New York, and in the same year by Soviet physicists Georgiy Zatsepin and Vadim Kuzmin of the Lebedev Institute of Physics in Moscow. They predicted that there would be very few cosmic rays with energies greater than about 6×1019 electronvolts (eV) because of energy losses through interactions with the ubiquitous photons of the cosmic microwave background, the radiation that fills the Universe.

But scientists studying high-energy cosmic rays have failed to observe the so-called GZK cut-off. Indeed, ultra-high-energy cosmic rays with energies of up to 3×1020 eV have been detected by Earth-based instruments. Such cosmic rays are mainly protons, thought to be generated by awesomely energetic astrophysical phenomena such as supernovae or supermassive black holes.

Most perplexingly, researchers working at a Japanese cosmic-ray observatory called the Akeno Giant Air Shower Array (AGASA) near Tokyo have previously reported a cosmic-ray energy spectrum that shows no obvious sign of a cut-off. “This excited many theorists,” says Douglas Bergman of Rutgers University in Piscataway, New Jersey. There was speculation about whether the Japanese results revealed new physics beyond Einstein’s theory of special relativity, such as the existence of a ‘shortest possible length’ analogous to the fastest possible speed (the speed of light) imposed by relativity.

There could be some really exciting particle physics involved here.

The new HiRes results, reported by Bergman and colleagues, discount such speculation — for now, at least. The team describes a cosmic-ray spectrum that drops sharply at around the predicted GZK cut-off energy. “They’ve pretty clearly seen the effect,” says astrophysicist Alan Watson of the University of Leeds, UK.

One of the main difficulties in spotting the cut-off has been a poverty of statistics. Most cosmic rays have energies lower than the GZK limit, and so it is tricky to detect enough particles at high energies for a drop in the energy spectrum to become clear. Bergman’s study drew on almost a decade of data from the HiRes experiment, which ran at the US Army Dugway Proving Ground in Utah from the late1990s until early 2006, when it was shut down.

HiRes’s two telescopes searched the sky for the characteristic flashes of ultraviolet light produced when a cosmic ray collides with a molecule in Earth’s atmosphere and creates a shower of secondary particles. The two ‘eyes’ — hemispheres covered in photomultiplier tubes that look like a fly’s compound eyes — capture just about all the light in the shower, giving a good measure of the original particle’s energy. “To see the GZK cut-off, it is vitally important to have good energy resolution,” Bergman says.

So why hasn’t AGASA seen it? “The AGASA people are really good experimentalists, and you can’t doubt their measurements,” Watson says. But AGASA doesn’t measure the cosmic-ray energies directly, so Watson thinks there could be something wrong in the theoretical model used to calculate them — something that might, after all, point to unknown new physics at these high energies. “There could be some really exciting particle physics involved here,” he says.

Watson also cautions that the drop in the energy spectrum observed by the HiRes team does not by itself provide conclusive proof of the GZK effect. “It could just be the cosmic-ray sources running out of steam at high energies,” he says. However, the new Pierre Auger Observatory, operating while still under construction in the Argentinian pampas, has also seen an apparent dip in the spectrum that is consistent with the GZK cut-off (see Nature 448, 8–9; 2007 ).

This cut-off is not an absolute limit on the energy of a cosmic ray because the slowing down is cumulative. If a very-high-energy cosmic ray is produced close enough to Earth, it might not lose much energy before it hits the atmosphere. Clearly, this does seem to happen, although astrophysicists are still debating which cosmic events could be energetic enough to create these particles. “The sources of ultra-high-energy cosmic rays we observe must be within about 50 megaparsecs [160 million light years, a typical distance to nearby galaxies] of Earth,” says Bergman. There don’t seem to be many potential sources that close.

And yet, he says, the high-energy particles seem to come from all directions. This discrepancy remains unexplained. So, although Bergman says the apparent confirmation of the GZK cut-off is “reassuring”, he adds that “I would not say all is well with ultra-high-energy cosmic rays”.

figure b