BEACON: Neutrino Observatory Design

• BEACON is a mountaintop neutrino observatory that uses phased radio arrays to detect upgoing air showers from Earth-skimming tau neutrinos.
• The design employs a scalable beamforming technique with a proven prototype and advanced simulation frameworks to optimize sensitivity.
• It targets both diffuse cosmic neutrino fluxes and transient events by strategically deploying stations and integrating local topography.

BEACON is the Beamforming Elevated Array for COsmic Neutrinos, a planned ultra-high-energy neutrino observatory that uses phased radio arrays on mountains to detect radio emission from upgoing air showers produced by Earth-skimming tau neutrinos. In its instrument concept, small elevated stations view the horizon, where long radio-propagation distances and large target mass in the Earth make the $\nu_\tau$ channel attractive. A prototype program established the basic instrument architecture and radio-frequency operating environment, while later sensitivity studies recast BEACON as a scalable mountaintop array for diffuse and transient neutrino searches (Southall et al., 2022, Zeolla et al., 17 Apr 2025).

1. Scientific scope and observational objective

BEACON is designed for neutrinos with energies $E_\nu \gtrsim 10^{17}\,\mathrm{eV}$, where fluxes are expected to be extremely low but the science case is unusually strong. The relevant signals include neutrinos produced inside extreme accelerators, cosmogenic neutrinos generated during ultra-high-energy cosmic-ray propagation, and transient emissions from source classes such as short gamma-ray bursts, newly born magnetars, and flat-spectrum radio quasars (Zeolla et al., 17 Apr 2025).

The instrument’s stated objective is to detect radio emission from upgoing extensive air showers initiated by Earth-skimming $\nu_\tau$. In this channel, a tau neutrino enters the Earth at a shallow angle, undergoes a charged-current interaction in rock, produces a tau lepton, and the tau exits the Earth before decaying in the atmosphere. The decay products then generate an upward-going air shower whose radio pulse can be observed from high-elevation stations. The radio emission is modeled as being dominated by the geomagnetic mechanism (Zeolla et al., 17 Apr 2025).

A central consequence of this design is that BEACON is optimized simultaneously for diffuse searches and for time-dependent point-source searches. The transient case is especially prominent in the sensitivity study, which treats both short events and long-duration episodes and derives point-source effective area as the basic observable for source-dependent sensitivity (Zeolla et al., 17 Apr 2025).

2. Detection principle and measurement formalism

The sensitivity study defines the point-source effective area in the ANITA-style form

$$A(t,E_\nu,\hat{r}) = \int_{A_g} dA_g \:(\hat{r} \cdot \hat{u}) \:\Theta(\hat{r} \cdot \hat{u}) \: P_\text{obs}(t, E_\nu,\hat{r}, \hat{u}),$$

where $A_g$ is the geometric area on the Earth surface containing potentially detectable tau exits, $\hat r$ is the source direction, $\hat u$ is the local surface normal, and $P_{\rm obs}$ is the detection probability (Zeolla et al., 17 Apr 2025).

Its Monte Carlo estimator is written as

$$A(E_\nu, \alpha, \delta) = \frac{A_g}{N} \sum_{i=1}^{N} (\hat{r}_i \cdot \hat{u}_{i})\:\Theta(\hat{r}_i \cdot \hat{u}_i) \: P_{i,\text{exit}} \: P_{i,\text{detect}},$$

with $P_{i,\text{exit}}$ the tau emergence probability and $E_\nu \gtrsim 10^{17}\,\mathrm{eV}$0 the probability that the resulting shower radio signal triggers at least one station (Zeolla et al., 17 Apr 2025).

The radio-electric-field scaling used in the simulation is

$E_\nu \gtrsim 10^{17}\,\mathrm{eV}$1

and the phased-array trigger metric is

$E_\nu \gtrsim 10^{17}\,\mathrm{eV}$2

so the signal-to-noise ratio grows as $E_\nu \gtrsim 10^{17}\,\mathrm{eV}$3 for $E_\nu \gtrsim 10^{17}\,\mathrm{eV}$4 phased antennas (Zeolla et al., 17 Apr 2025).

For diffuse sensitivity, the acceptance is

$E_\nu \gtrsim 10^{17}\,\mathrm{eV}$5

and the corresponding all-flavor $E_\nu \gtrsim 10^{17}\,\mathrm{eV}$6 diffuse-flux sensitivity is

$E_\nu \gtrsim 10^{17}\,\mathrm{eV}$7

For transient fluence, the analogous all-flavor $E_\nu \gtrsim 10^{17}\,\mathrm{eV}$8 point-source sensitivity is

$E_\nu \gtrsim 10^{17}\,\mathrm{eV}$9

These expressions make explicit that BEACON’s source reach is controlled by effective area for transients and by sky-integrated acceptance for diffuse searches (Zeolla et al., 17 Apr 2025).

3. Instrument architecture and prototype implementation

The prototype paper describes an 8-channel prototype instrument installed at high elevation at Barcroft Field Station, operating since 2018. It consists of 4 dual-polarized antennas sensitive between 30–80 MHz; the signals are filtered, amplified, digitized, and saved to disk using a custom data acquisition system. The prototype site is high elevation “to maximize effective volume” and uses a directional beamforming trigger to improve rejection of anthropogenic background noise at the trigger level. The prototype report discusses design, construction, calibration, the radio-frequency environment, event categories, and includes a likely cosmic ray candidate event (Southall et al., 2022).

The later sensitivity study elevates this prototype concept into a phased-station architecture. Its benchmark station contains 10 phased antennas per station, implemented as low-cost short dipoles whose signals are digitally delayed and summed into multiple trigger beams. The benchmark detector assumes 30–80 MHz operation, a trigger threshold of SNR = 5, and a one-station trigger as sufficient for event detection (Zeolla et al., 17 Apr 2025).

The benchmark 100-station array is placed at 3 km altitude, arranged along a single line of longitude, centered near the prototype region at latitude $\nu_\tau$0 N and longitude $\nu_\tau$1 W, with 3 km spacing between stations. The stations face East and each has a $\nu_\tau$2 azimuthal field of view. A 1000-station array is estimated by scaling the 100-station result by a factor of 10; the study argues that, for the assumed 3 km spacing, overlap effects are sufficiently linear between 100 and 1000 stations to justify that extrapolation (Zeolla et al., 17 Apr 2025).

The modeled antenna/noise chain gives a single-antenna RMS noise voltage

$\nu_\tau$3

for the prototype antenna model with $\nu_\tau$4, where $\nu_\tau$5 is the sky fraction in the field of view (Zeolla et al., 17 Apr 2025).

4. Simulation framework, environment modeling, and array scaling

The sensitivity calculations are performed with MARMOTS (“Multiple Antenna Arrays on Mountains Tau Simulation”), a point-source effective-area framework based conceptually on the ANITA code Tapioca but adapted to many stationary arrays. The simulation chain includes NuTauSim for neutrino propagation and tau exit, PYTHIA for tau decay and shower-energy fraction, ZHAireS-RASPASS lookup tables for radio emission from up-going showers on a spherical Earth, IGRF for the geomagnetic field, XFdtd for the prototype antenna model, Dulk parameterization for galactic and extragalactic radio noise, and PREM plus SRTM for grammage and topographic modeling (Zeolla et al., 17 Apr 2025).

The geometric setup is horizon-centered. At any instant, BEACON’s effective area is concentrated in a narrow band just below the horizon, where shallow Earth-skimming trajectories combine large visible area with favorable tau-exit probability. Above the horizon the effective area is zero, and farther below the horizon both geometric area and exit probability decrease rapidly. As the Earth rotates, this band sweeps in right ascension, yielding day-averaged sky coverage of about $\nu_\tau$6 (Zeolla et al., 17 Apr 2025).

Station overlap is energy dependent. At $\nu_\tau$7, stations are nearly independent for spacing of about 2 km, whereas at higher energies spacing greater than 5 km would be needed for true independence. Even so, for the benchmark geometry, the study concludes that 1000 stations have $\nu_\tau$8 the effective area of 100 stations despite overlap (Zeolla et al., 17 Apr 2025).

Topography is treated separately from the smooth-Earth baseline. For a single station at the prototype site, local topography increases the maximum instantaneous effective area at 1 EeV by about 15%, but narrows sky coverage because nearby terrain blocks steeper lines of sight. Below $\nu_\tau$9 eV, the same topographic treatment can improve maximum instantaneous effective area by more than a factor of 3. This indicates that site selection and local terrain are not perturbative details but part of the array design space (Zeolla et al., 17 Apr 2025).

5. Transient and diffuse sensitivity

For short transients, the study uses the maximum instantaneous effective area. Its headline claim is that with just 100 stations, BEACON achieves sensitivity to short-duration transients such as nearby short gamma-ray bursts, while with 1000 stations it reaches much deeper transient sensitivity (Zeolla et al., 17 Apr 2025).

A representative short-burst benchmark is an on-axis short gamma-ray burst at 40 Mpc. For the 1000-station array, the maximum expected detections for the extended emission model are 35 neutrinos, and the maximum distance for at least one detected neutrino is 237 Mpc. The same study emphasizes that such favorable geometry is rare instantaneously: only about $$A(t,E_\nu,\hat{r}) = \int_{A_g} dA_g \:(\hat{r} \cdot \hat{u}) \:\Theta(\hat{r} \cdot \hat{u}) \: P_\text{obs}(t, E_\nu,\hat{r}, \hat{u}),$$0 of the sky contains a detectable 40 Mpc short gamma-ray burst at a given instant for the benchmark setup (Zeolla et al., 17 Apr 2025).

The source-time dependence is also explicit. For a source at $$A(t,E_\nu,\hat{r}) = \int_{A_g} dA_g \:(\hat{r} \cdot \hat{u}) \:\Theta(\hat{r} \cdot \hat{u}) \: P_\text{obs}(t, E_\nu,\hat{r}, \hat{u}),$$1, sensitivity peaks when it is viewed at roughly $$A(t,E_\nu,\hat{r}) = \int_{A_g} dA_g \:(\hat{r} \cdot \hat{u}) \:\Theta(\hat{r} \cdot \hat{u}) \: P_\text{obs}(t, E_\nu,\hat{r}, \hat{u}),$$2 elevation. Under that geometry, a short gamma-ray burst is detectable for about 30 minutes with BEACON-100 and roughly 1 hour with BEACON-1000. Lowering the trigger threshold materially improves performance: for the 40 Mpc short-burst benchmark, the maximum expected neutrino counts change from 3, 35 at SNR 5 to 5, 56 at SNR 4 and 9, 98 at SNR 3 for the 100-station and 1000-station arrays, respectively (Zeolla et al., 17 Apr 2025).

For long-duration transients, the study uses the day-averaged effective area, which is lower than the instantaneous peak by about two orders of magnitude but applies over large sky fractions. A 2-day newly born magnetar at 1 Mpc yields a maximum of 11 neutrinos in BEACON-1000, and the maximum distance for at least one detected neutrino is 3.4 Mpc. The declination band used to summarize the most favorable day-averaged sensitivity is

$$A(t,E_\nu,\hat{r}) = \int_{A_g} dA_g \:(\hat{r} \cdot \hat{u}) \:\Theta(\hat{r} \cdot \hat{u}) \: P_\text{obs}(t, E_\nu,\hat{r}, \hat{u}),$$3

The same long-duration comparison includes stacked all-flavor fluence from 10 FSRQs over 10 years (Zeolla et al., 17 Apr 2025).

For diffuse sensitivity, the study adopts a 5-year exposure. Its principal conclusion is that BEACON-100 would roughly reach existing IceCube and Auger limits near 1 EeV, whereas BEACON-1000 would improve on existing experiments by about a factor of 10 at 1 EeV and begin probing a broad class of cosmogenic models. This factor-of-ten statement is the central diffuse-performance claim of the paper (Zeolla et al., 17 Apr 2025).

6. Development trajectory, design tradeoffs, and limitations

The two BEACON papers together outline a staged program. The prototype establishes that a mountaintop, low-frequency, beamforming radio instrument can be built, calibrated, and operated in a real radio environment, and that its event stream can be classified at least to the point of identifying a likely cosmic-ray candidate (Southall et al., 2022). The sensitivity study then generalizes that prototype into a distributed array concept in which phased stations on mountains become a horizon-viewing neutrino observatory (Zeolla et al., 17 Apr 2025). This suggests a development path from local trigger and environment characterization to a geographically extended neutrino array.

The concept involves an explicit geometric tradeoff. A single mountain chain yields deep, narrow instantaneous sensitivity, especially for short transients, whereas distributing stations across multiple ranges would produce wider, shallower coverage. The study also notes that final station locations have not yet been chosen, that the 1000-station performance is largely an extrapolation from the 100-station simulation, and that local topography and geomagnetic conditions can change the sensitivity by factors of a few (Zeolla et al., 17 Apr 2025).

Another limitation is methodological rather than instrumental. The baseline benchmark uses a smooth Earth and a highly regular station layout, while the topographic branch shows that real terrain can substantially reshape both instantaneous effective area and sky visibility. The exact balance between optimized source discovery, cosmogenic-flux reach, and practical siting therefore remains open (Zeolla et al., 17 Apr 2025).

Within those constraints, BEACON is best understood as a mountaintop phased-radio implementation of the Earth-skimming $$A(t,E_\nu,\hat{r}) = \int_{A_g} dA_g \:(\hat{r} \cdot \hat{u}) \:\Theta(\hat{r} \cdot \hat{u}) \: P_\text{obs}(t, E_\nu,\hat{r}, \hat{u}),$$4 technique: a system whose defining elements are high-elevation deployment, directional beamforming, low-frequency radio triggering, and a science program that links nearby short transients, longer-duration source classes, and the diffuse cosmogenic frontier in a single architecture (Southall et al., 2022, Zeolla et al., 17 Apr 2025).