Radio Detection of Ultra-High-Energy Neutrinos: From ARA to GRAND

Every neutrino detection technology is, at its core, a bet on a particular physical process. Liquid scintillator bets on secondary photon emission from recoiling electrons. Water Cherenkov detectors bet on the cone of light emitted by relativistic charged particles. The inverse-beta-decay technique bets on the coincidence signal from a positron and a thermalized neutron.

Radio detection bets on something different: that a neutrino-induced particle cascade in a dense, radio-transparent medium will emit a coherent burst of radio waves intense enough to detect with buried antennas, and that the medium will be transparent enough that the signal reaches the antenna from distances of hundreds of metres or several kilometres. At energies above about 10 PeV, this bet increasingly pays off — and below those energies it does not.

The technique is not new. Its physical basis was identified in 1962. But building the instruments capable of exploiting it at the required scale has taken decades, and the first generation of dedicated in-ice radio detectors is only now reaching operational maturity.

---

The Askaryan Effect

The physical mechanism underlying radio detection is the Askaryan effect, predicted by Soviet physicist Gurgen Askaryan in 1962 and confirmed experimentally at SLAC in 2001.

When a high-energy neutrino interacts in a dense medium — ice, salt, or lunar regolith — it initiates an electromagnetic and hadronic shower of secondary particles. As the shower develops, a charge asymmetry builds up: electrons are knocked out of the medium by Compton scattering and pair production adds equal numbers of electrons and positrons, but positrons annihilate and electrons do not. The shower front therefore carries a net negative charge excess of approximately 20% of the total charged particle number.

This charge excess propagates coherently through the medium at the shower’s speed. For wavelengths longer than the shower length, all parts of the shower contribute in phase, and the radiated power scales as $N^2$ where $N$ is the number of particles — coherent emission. For wavelengths shorter than the shower length, the emission is incoherent and scales as $N$.

The electromagnetic shower initiated by a $10^{18}$ eV neutrino develops over a length of roughly 1–2 metres (for the electromagnetic component) to several metres (for the hadronic component) in ice. The corresponding wavelength at which the transition from coherent to incoherent emission occurs is:

${\lambda }_c\approx L_{\text{shower}}\approx 1\text{–}5\text{m}\Rightarrow {\nu }_c\approx 60\text{–}300\text{MHz}$

Below a few hundred MHz — frequencies accessible to ordinary radio electronics — the emission is coherent and the signal amplitude scales linearly with shower energy. The electric field at the detector from a shower of energy $E_{\text{shower}}$ at distance $r$ is approximately:

$\mathcal{E}\propto \frac{E_{\text{shower}}}{r}\cdot \sin {\theta }_C\cdot f(\nu )$

where ${\theta }_C$ is the Cherenkov angle and $f(\nu )$ is the frequency-dependent form factor. The emission is maximum at the Cherenkov angle ${\theta }_C\approx 56°$ in ice (refractive index $n\approx 1.78$) and is strongly polarised in the plane containing the shower axis and the observation direction.

---

Why Radio for UHE Neutrinos?

The scientific target of radio neutrino detectors is the cosmogenic neutrino flux, also called the GZK flux after Greisen, Zatsepin, and Kuzmin, who predicted in 1966 that protons accelerated to ultra-high energies would lose energy by photomeson production on the cosmic microwave background:

$p+{\gamma }_{\text{CMB}}\to {\Delta }^{+}\to p+{\pi }^0/n+{\pi }^{+}$

The charged pions decay to muons and neutrinos; the neutral pions to photons. The result is a flux of neutrinos at energies of $10^{17}$ to $10^{20}$ eV. The flux is expected to be small — predictions range from one to a few events per square kilometre per year, depending on the composition and spectrum of the primary cosmic rays — but it is the highest-energy neutrino signal accessible by any process in the Standard Model, and detecting it would confirm the hadronic origin of ultra-high-energy cosmic rays.

IceCube, with its roughly 1 km³ instrumented volume, has set the world’s best upper bounds on the cosmogenic flux at energies above a few hundred TeV, but its optical photomultiplier array does not extend efficiently above a few PeV for contained events. At $10^{18}$ eV, even a kilometre of ice transmits a significant fraction of neutrinos, and the shower signal in the optical is too bright and too spatially diffuse to reconstruct well with PMTs spaced 125 m apart.

The Askaryan signal in ice scales as $E^2$ (coherent power), while optical Cherenkov scales as $E$. At $10^{18}$ eV, the radio signal is detectable from distances of 1–3 km with antennas of modest gain, allowing a single station of a few antennas to instrument an effective detection volume of several cubic kilometres. A hundred such stations, spaced a kilometre apart, instrument $\sim 100{\text{km}}^3$ — a scale that optical detectors cannot reach economically.

---

The Confirmation at SLAC: Experiment T-510 and ANITA

Askaryan’s prediction waited thirty-nine years for experimental confirmation. In 2001, a collaboration led by David Saltzberg bombarded silica sand targets at the SLAC National Accelerator Laboratory with an electron beam, observing coherent radio emission from electromagnetic showers with the predicted spectral shape, angular distribution, and polarisation. The result established the Askaryan effect as a well-understood and quantitatively predictable phenomenon.

The ANITA (Antarctic Impulsive Transient Antenna) experiment provided the first detection of Askaryan-channel events from a natural source. ANITA is a balloon-borne radio interferometer that has flown in several campaigns above Antarctica, detecting radio emission from cosmic-ray showers in the atmosphere (which produces coherent emission by a different but related mechanism) and searching for Askaryan emission from neutrino-induced showers in the ice below. ANITA established the first experimental upper bounds on the UHE neutrino flux from a radio instrument and identified several anomalous upward-going events that remain unexplained.

---

ARA: Askaryan Radio Array

The Askaryan Radio Array (ARA) at the South Pole is the first large-scale in-ice radio detector. ARA stations consist of clusters of radio antennas deployed at depths of 150–200 m in boreholes — deep enough to avoid the firn layer near the surface where the ice has variable density and attenuation length is reduced, but accessible with the South Pole drilling infrastructure.

Each ARA station instruments a cylinder of ice roughly 200 m in diameter and 100 m in depth, using a combination of vertically polarised dipole antennas (VPol) and horizontally polarised bicone antennas (HPol). The antennas operate in the 150–850 MHz band, corresponding to a shower-length coherence range appropriate for energies above $\sim 10^{16}$ eV.

ARA began deploying stations in 2012 and has operated five fully instrumented stations. The science result from the first five years of data set upper limits on the diffuse neutrino flux above $10^{16}$ eV, approaching but not yet surpassing the Auger Observatory’s bounds at the highest energies. The chief result of ARA has been to characterise the in-ice radio background at the South Pole and to develop the reconstruction and veto techniques needed for larger arrays.

---

RNO-G: Radio Neutrino Observatory Greenland

RNO-G, deployed at Summit Station on the Greenland ice sheet beginning in 2021, is the currently operational in-ice radio neutrino telescope that most directly succeeds ARA in scientific scope. Summit Station sits atop more than 3 km of ice at an altitude of roughly 3,200 m, offering an exceptionally radio-quiet environment and ice properties well-characterised by deep core drilling campaigns.

RNO-G’s design philosophy differs from ARA in several respects. Each of the planned 35 stations occupies a footprint of approximately 100 m, with antennas deployed at three depths (down to 100 m) and surface antennas for cosmic-ray air shower vetoing. The station spacing of roughly 1.25 km provides an effective volume of $\sim 50{\text{km}}^3$ at full deployment.

Crucially, RNO-G incorporates a phased array trigger — a technique borrowed from radar astronomy — that coherently sums the signals from multiple antennas to achieve a trigger threshold roughly three times lower in amplitude than a single-antenna trigger. This dramatically improves sensitivity to near-threshold events and extends the detectable energy range downward toward the IceCube overlap region.

Initial RNO-G results with the first nine deployed stations have confirmed the radio background environment at Summit Station and have demonstrated end-to-end event reconstruction for cosmic-ray signals detected by the surface antennas. The full 35-station array, expected by the late 2020s, should achieve sensitivity to the optimistic cosmogenic flux predictions within a few years of operation.

---

The Radio Extension of IceCube: IceCube-Gen2

The most ambitious near-term in-ice radio project is the radio component of IceCube-Gen2, the proposed upgrade to the IceCube Neutrino Observatory at the South Pole. IceCube-Gen2 envisions instrumenting roughly 8 km³ of ice with a next-generation optical array plus a radio array of approximately 200 stations covering $\sim 500{\text{km}}^3$ at the surface and in shallow ice.

The IceCube-Gen2 radio component is designed with sensitivity to the cosmogenic neutrino flux as its primary goal, but it also targets:

• Measurement of the neutrino spectrum from $10^{16}$ to $10^{21}$ eV in combination with the optical component
• Detection of the Glashow resonance at $6.3\times 10^{15}$ eV in ${\bar{\nu }}_e+e^{-}$ scattering, extending IceCube’s current single-event detection
• Searches for tau-neutrino-induced double-bang events at PeV energies

The radio and optical components of IceCube-Gen2 are designed to overlap in energy near $10^{16}$ eV, allowing cross-calibration of the two detection channels and providing a systematic handle on the radio analysis.

---

GRAND: Giant Radio Array for Neutrino Detection

On the opposite end of the instrumentation philosophy sits GRAND (Giant Radio Array for Neutrino Detection), a proposed surface radio array targeting Earth-skimming tau neutrinos at energies above $10^{17}$ eV.

The GRAND concept exploits a different physical scenario from the Askaryan technique. An ultra-high-energy tau neutrino can skim the Earth at a shallow angle, interact in the rock or ice near the surface, produce a tau lepton that exits the Earth, and then decay in the atmosphere, initiating an upward-going extensive air shower. The radio emission from this air shower — produced by the geomagnetic and Askaryan mechanisms — is detectable at the surface.

The advantage of the air-shower approach is geometric: the radio footprint of an inclined air shower at $10^{18}$ eV extends over several square kilometres of surface, allowing a sparse array of antennas with kilometre-scale spacing to instrument enormous effective volumes. GRAND’s eventual target is $2\times 10^5$ radio antennas deployed across $2\times 10^5{\text{km}}^2$ in radio-quiet mountainous terrain, achieving an effective aperture orders of magnitude larger than any in-ice array.

GRAND is proceeding in stages. The prototype phase, GRANDProto300 (300 antennas in China), is characterising the radio background and shower detection efficiency. A later stage, GRAND10k, would demonstrate the tau neutrino detection technique at sufficient scale to observe the first cosmogenic neutrino events if the optimistic flux predictions are correct.

The GRAND approach is sensitive to a different set of systematic uncertainties from the in-ice Askaryan technique: air shower simulations, tau lepton propagation in rock, and atmospheric refraction of the radio beam all require careful characterisation. But the scale is unmatched, and the complementarity with in-ice detectors in terms of systematics makes the combination scientifically powerful.

---

The Auger Bound and What It Implies

The Pierre Auger Observatory in Argentina, primarily a cosmic-ray detector, achieves sensitivity to UHE neutrinos through inclined air showers and Earth-skimming events detected by its surface water-Cherenkov array and radio antennas (AERA). Auger has set the current world-best bound on the diffuse UHE neutrino flux above $10^{17}$ eV and does not yet see a signal.

The Auger upper limit already constrains the most optimistic cosmogenic flux predictions — those based on pure-proton primary composition and a hard injection spectrum. Models with heavier composition (as suggested by Auger’s own fluorescence measurements) predict lower neutrino fluxes that remain consistent with non-detection.

This means that next-generation radio detectors are not guaranteed to see a signal. RNO-G at full deployment should reach sensitivity to the moderate cosmogenic flux predictions within roughly five years. If it does not detect events, that will itself be a significant constraint — implying either heavy cosmic-ray composition or a lower maximum acceleration energy than current models assume. If it does, the spectrum and arrival direction distribution of the detected neutrinos will constrain the sources and composition of ultra-high-energy cosmic rays in ways no existing instrument can achieve.

---

Calibration, Backgrounds, and the Path Forward

The principal experimental challenges for radio neutrino detectors are threefold.

Ice properties. Radio attenuation lengths in Antarctic and Greenlandic ice are well-measured but vary with depth and temperature. The birefringence of glacial ice — its anisotropic crystal structure — can rotate the polarisation of radio pulses, complicating reconstruction. These effects are being characterised by dedicated calibration pulsers deployed alongside detector stations.

Anthropogenic RFI. Radio-quiet sites reduce but do not eliminate human-generated radio signals. Aircraft, satellites, snowmobiles, and station electronics all produce impulsive signals that can mimic neutrino events. Coincidence requirements between multiple antennas, polarisation analysis, and directional vetoes suppress these backgrounds, but the rejection factor needed is $\sim 10^6$.

Thermal noise. At frequencies of 100–1000 MHz, the thermal noise power in a bandwidth $\Delta \nu$ is $k_{B}T\Delta \nu$ where $T\approx 250$ K for deep South Pole ice. The phased-array trigger technique aggregates coherently, giving a noise reduction proportional to $\sqrt{N}$ for $N$ antennas. With 10–20 antennas per station, trigger thresholds equivalent to 3× the single-antenna thermal noise level are achievable.

The technology has matured substantially since ARA’s deployment. RNO-G demonstrates the current state of the art. IceCube-Gen2 and GRAND represent the decade-scale targets that will either detect or definitively constrain the cosmogenic neutrino flux.

If the cosmogenic flux is there — and the physics of cosmic-ray propagation above the GZK energy strongly suggests it must be — radio will be the technique that finds it.

---

Related reading: the cosmogenic neutrino flux and GZK mechanism article explains the production physics. IceCube’s neutrino astronomy programme covers the optical Cherenkov approach at lower energies. The IceCube-Gen2 detailed architecture discusses the combined optical-radio upgrade.