The Z-Burst Hypothesis: An Idea That Almost Explained Ultra-High-Energy Cosmic Rays

In 1982, Yakir Wolfenstein noticed that neutrinos passing through dense matter are subject to coherent forward scattering — the origin of the MSW effect that would resolve the solar neutrino problem two decades later. In the late 1990s, Thomas Weiler took a quite different cue from the same idea: ultra-high-energy neutrinos travelling cosmological distances must occasionally pass through the cosmic neutrino background, and if their energy is just right, they can resonantly annihilate on a relic neutrino to produce a real Z boson.

The arithmetic that follows is striking. The required energy depends on the relic neutrino mass: $E_{\nu }\approx M_Z^2/(2m_{\nu })$. For $m_{\nu }\sim 0.1$ eV this puts the resonance energy at about $4\times 10^{22}$ eV — a zetta-electron-volt. That is roughly three orders of magnitude above the highest cosmic-ray events ever recorded. But unlike charged cosmic rays, neutrinos at any energy propagate cosmologically without losing energy or being deflected by magnetic fields. So if some distant cosmic accelerator produced enough zetta-electron-volt neutrinos, a small fraction would convert via the Z resonance into Standard-Model decay products — quarks, leptons, gluons — and the hadronic Z decays would produce a shower of secondary protons and photons that propagate to Earth as ultra-high-energy cosmic rays.

The Z-burst hypothesis was for a few years one of the most discussed explanations for the events above the GZK cutoff that the AGASA experiment was seeing in the late 1990s. By the early 2010s the hypothesis had been substantially ruled out by data from the Pierre Auger Observatory, by neutrino-flux limits from ANITA, and by improved composition measurements. This post is about why the idea once looked so compelling, what predictions it made, and how the field ultimately said no.

The original puzzle

The Greisen-Zatsepin-Kuzmin cutoff, predicted in 1966 and discussed in detail in our earlier post on cosmogenic neutrinos, is a steep suppression of the cosmic-ray flux above $5\times 10^{19}$ eV. Protons above this energy lose so much energy on the CMB through pion photoproduction that they cannot reach Earth from distances greater than about 50 megaparsecs. The cosmic-ray spectrum should therefore turn over sharply at the cutoff energy.

In the late 1990s, the AGASA experiment in Japan reported a handful of events with energies above $10^{20}$ eV that appeared to contradict this prediction. If the events were real and the cutoff was supposed to apply, then either the sources of these particles had to be relatively nearby — within the GZK horizon — or new physics had to deliver them around the suppression.

The Z-burst hypothesis was an elegant way to get around the cutoff. Its central idea: don’t try to deliver charged particles through the CMB. Deliver neutrinos, which interact only feebly and propagate freely; let them convert to hadrons locally, near Earth, through the Z resonance.

For this to work, the neutrino sources had to produce a sufficient flux at the resonant energy, and the resonant conversion rate had to be large enough to account for the observed event rate above the cutoff. Estimates at the time suggested both conditions could be satisfied for reasonable parameters — provided cosmological accelerators existed that could push particles to $10^{22}$ eV. No such accelerator was identified, but the mechanism was at least kinematically self-consistent.

The Z resonance

The cross-section for $\nu \bar{\nu }\to Z\to X$ is a textbook Standard-Model calculation. At the Z mass it reaches its peak value of approximately

${\sigma }_{\mathrm{peak}}\approx \frac{12\pi }{M_Z^2}\cdot \frac{{\Gamma }_{Z\to \nu \nu }}{{\Gamma }_Z}\cdot \frac{1}{1}$

which works out to about $4.2\times 10^{-32}$ cm². This is essentially the same as the cross-section that gave the original “neutrino counting” measurement at LEP. By cosmic-physics standards it is enormous — a typical weak-interaction cross-section at GeV energies is $10^{-38}$ cm² or less. But of course the path length through the relic neutrino background is also exquisite: the relic neutrino number density today is about 336 per cubic centimeter, giving an interaction length of order

${\lambda }_{\mathrm{res}}\sim \frac{1}{n_{\nu }{\sigma }_{\mathrm{peak}}}\approx 100\text{Mpc}$

If a cosmic neutrino with the resonant energy propagates over a Hubble distance of a few thousand megaparsecs, it has a few-per-cent probability of resonant absorption. The Z that emerges decays after a fleeting lifetime, producing roughly two quarks, one lepton pair, or one neutrino pair per decay channel; the hadronic branching ratio is about 70%. Each Z decay produces of order ten high-energy hadrons and photons through fragmentation and decay.

The predictions and the data

For Z-bursts to explain a substantial fraction of the observed UHECR flux above $10^{20}$ eV, several specific predictions had to hold.

First, the composition of the observed cosmic rays at the highest energies should be dominantly protons and photons, because Z decays produce light-quark fragmentation products and prompt photons, not heavy nuclei. Heavy elements such as iron would not be produced by Z-bursts in any significant fraction.

Second, the arrival directions of the observed events should be isotropic, because the resonant absorption can happen anywhere along the line of sight that hosts relic neutrinos, which is essentially uniform on the sky. No correlation with the local matter distribution would be expected.

Third, the ultra-high-energy neutrino flux at $\sim 10^{22}$ eV had to be large enough to deliver the observed cosmic-ray rate after accounting for the few-per-cent resonant conversion probability. This translated to a parent-neutrino flux comparable to or larger than what cosmological accelerators were thought capable of producing.

All three predictions came under increasing pressure as data improved.

The Pierre Auger Observatory began full operation in 2008 and measured both the cosmic-ray spectrum and the average shower composition at the highest energies. The composition data showed an increasing fraction of heavier nuclei above $10^{19}$ eV — exactly the opposite of what the Z-burst hypothesis predicted. The spectrum showed a clear cutoff consistent with the GZK suppression, eliminating the original motivation for invoking new physics. And the arrival-direction data showed weak but detectable correlations with the distribution of nearby active galactic nuclei, suggesting the highest-energy particles were charged nuclei deflected slightly by intergalactic magnetic fields — again, exactly what one would expect from conventional acceleration in nearby sources.

The ANITA balloon experiment and the ARA ground array searched for the parent neutrinos at the predicted zetta-eV scale via the Askaryan radio-pulse signature in Antarctic ice. Neither found anything. The resulting upper limits on the neutrino flux are well below what the Z-burst hypothesis requires.

By the early 2010s, the consensus had shifted decisively away from Z-bursts as a primary UHECR mechanism. The model remains kinematically self-consistent — the Z resonance is a real Standard Model process, the relic neutrinos are presumably there — but the observational requirements have not been met, and conventional astrophysical accelerators provide an adequate description of what is seen.

What remains useful

Although the Z-burst hypothesis as an explanation for UHECRs did not survive, the underlying physics retains interest for a few reasons.

The annihilation cross-section at the Z peak is one of the cleanest measurements of how neutrinos interact, and the LEP measurement of the invisible Z width remains the most direct empirical input on the number of light neutrino species. The same physics would, if neutrinos at the resonant energy ever turn up, provide a unique probe of the cosmic neutrino background — converting an undetected sea of millimeter-wavelength neutrinos into observable cosmic-ray bursts.

The cosmic neutrino background detection problem itself is one of the long-term goals of the field. The PTOLEMY proposal targets it through tritium beta decay with a relic neutrino capture. The Z-burst mechanism is a wildly different approach that, if ever realised, would constitute a detection of the relic neutrino population — by counting cosmic-ray events tagged as Z-burst products. The required neutrino flux is currently far out of reach, but it is not impossible to imagine future cosmic-ray observatories sensitive enough to look for the specific spectral feature a Z-burst component would imprint on the UHECR flux.

The hypothesis also stands as a clean illustration of a recurring pattern in physics: an elegant idea that solves a real problem within current data, makes definite predictions, and is conclusively ruled out by improved measurements. The Z-burst story is one of the better examples of how a hypothesis can be tested out of existence by the accumulation of independent constraints — composition, spectrum, anisotropy, neutrino flux — rather than by a single decisive test.

Summary

The Z-burst hypothesis proposed that ultra-high-energy cosmic neutrinos annihilate on relic neutrinos through resonant Z-boson production, with the decay products providing the cosmic rays observed above the GZK cutoff. The required parent-neutrino energy of about $4\times 10^{22}$ eV is set by the Z mass and the relic-neutrino mass, and the resonant interaction probability over a Hubble distance is a few per cent. The hypothesis was widely discussed in the late 1990s as a possible explanation for events above the GZK cutoff, but composition measurements from the Pierre Auger Observatory, neutrino-flux limits from ANITA and ARA, and confirmation of the GZK cutoff itself have decisively disfavored it. The underlying physics nonetheless remains interesting as a hypothetical probe of the cosmic neutrino background, and the rise and fall of the Z-burst proposal is one of the clearer recent examples of a clean idea being tested out of existence by data.