GRAND@Auger: Autonomous Radio Prototype

• The paper demonstrates the first self-triggered cosmic ray event detection in coincidence with the Pierre Auger Observatory, proving the feasibility of autonomous radio detection.
• The prototype employs 10 densely arranged stations with horizon antennas and FPGA-based filtering to mitigate interference in a challenging radio environment.
• Galactic background calibration using LFmap simulations validates the detector response and reconstruction methods for inclined air showers.

GRAND@Auger is the Argentine prototype of the Giant Radio Array for Neutrino Detection, installed at the Pierre Auger Observatory in Malargüe to test, under realistic field conditions, the core GRAND concept of autonomous, self-triggered radio detection of extensive air showers, especially inclined ones, using sparse antenna arrays. Established through an agreement between the GRAND and Pierre Auger collaborations, it repurposes ten Auger Engineering Radio Array stations as GRAND detection units and, by virtue of its co-location with Auger, functions as a cross-calibration and validation instrument rather than simply a standalone prototype. Its defining feature is the possibility of event-by-event comparison between GRAND hardware and an established hybrid ultra-high-energy cosmic-ray observatory (Errico et al., 10 Jul 2025).

1. Programmatic role within GRAND

GRAND@Auger, often abbreviated G@A, is one of the three prototype arrays operated by the GRAND collaboration, alongside GRANDProto300 in China and GRAND@Nançay in France. Within this prototype ensemble, its role is specific: GRAND@Nançay serves mainly as a hardware and trigger testbench, GRANDProto300 is the mid-scale pathfinder for autonomous radio detection in a radio-quiet environment, and GRAND@Auger is the prototype dedicated to validating reconstruction and self-triggering against external reference data from the Pierre Auger Observatory (Guelfand, 3 Jan 2025).

This functional specialization is central to its scientific meaning. Earlier status papers described GRAND@Auger as an ideal testbench for evaluating reconstruction quality in terms of arrival direction, energy, and nature of the primary particle, with an expected rate around $1$ cosmic ray EAS/day in coincidence with Auger SD data. The later proceedings contribution that reports the first results, however, does not provide a measured event rate and explicitly states that the expected coincidence rate still requires further investigation. The continuity between these statements is that GRAND@Auger was conceived primarily as a validation array, not as a science-scale detector (Guelfand, 3 Jan 2025).

The broader GRAND program is designed around very inclined air showers and ultimately targets ultra-high-energy neutrinos, cosmic rays, and gamma rays through radio detection. In that context, GRAND@Auger provides a controlled environment for testing whether autonomous radio detection, reconstruction, and background handling perform as required when compared directly to an established observatory. This suggests that its importance is methodological and strategic: it anchors GRAND’s detector concept to external truth information before scaling to much larger arrays.

2. Array layout, detection units, and instrumentation

The deployed array consists of 10 stations arranged as two superimposed hexagons with about $250~\mathrm{m}$ spacing, covering roughly $0.5~\mathrm{km}^2$. This relatively dense geometry was chosen to increase the probability of observing lower-energy and less inclined showers during commissioning, rather than to mimic the final sparse GRAND scale immediately (Errico et al., 10 Jul 2025).

Feature	GRAND@Auger
Stations	10
Layout	Two superimposed hexagons
Spacing	About $250~\mathrm{m}$
Area	Roughly $0.5~\mathrm{km}^2$
Antenna	Horizon Antenna
Polarizations	NS, EW, V

Each station is a GRAND detection unit built around a Horizon Antenna mounted on a $3~\mathrm{m}$ pole and measuring three polarizations: North–South, East–West, and Vertical. Some AERA infrastructure was reused, notably the solar panels and communication antenna. GRAND-specific components were then added: the Horizon antenna at the top, supported by a structure containing the low-noise amplifier; GPS and wireless communication hardware mounted midway up the pole; and, at the base, a weather-sealed aluminum box containing the front-end electronics, battery, and charge controller. The electronics are enclosed in Faraday cages to reduce self-generated electromagnetic interference (Errico et al., 10 Jul 2025).

The electronics chain includes a $500$ MSPS ADC for digitization and a Xilinx SoC combining FPGA and CPU resources. The system applies a $30\text{--}200~\mathrm{MHz}$ bandpass filter and supports periodic or threshold-based triggers. A programmable digital notch filter is implemented online in the FPGA to suppress narrowband radio-frequency interference. Event data are acquired locally and sent, when requested, to the central DAQ via a wireless Ubiquiti Bullet-Rocket link. At the Central Radio Station, the central system assembles triggered events, writes them to disk, and transfers them daily to CCIN2P3 in Lyon for offline analysis (Errico et al., 10 Jul 2025).

The paper does not provide a detailed timing-chain specification beyond GPS modules and station timestamps, and it does not quote an absolute synchronization accuracy. Even so, timing is structurally central, because both direction reconstruction and inter-experiment coincidence studies depend on inter-station and inter-observatory timing consistency.

3. Deployment and commissioning

Deployment occurred in two campaigns, in March and August 2023, with support from Auger local staff. Commissioning then continued through 2023 and early 2024. By March 2024, all ten detection units were operational, stable data taking had been achieved, and automatic hibernation protections against high temperature or low battery were working (Errico et al., 10 Jul 2025).

A representative periodogram from one station shows nearly 150 hours of continuously recorded periodically triggered data. In that visualization, persistent narrowband emitters appear as vertical yellow lines, while hibernation periods appear as white horizontal gaps where data taking stopped and later resumed automatically. This is an important commissioning result because it demonstrates long-duration autonomous operation while also revealing the spectral structure of the local radio environment, including both stable anthropogenic lines and clean spectral regions (Errico et al., 10 Jul 2025).

The site environment is not radio quiet in the same sense as GP300. Earlier status reports had already emphasized that the Auger site is scientifically valuable because of its existing detector infrastructure, not because it offers the cleanest RF conditions. GRAND@Auger therefore tests GRAND hardware and self-triggering in a comparatively demanding environment. A plausible implication is that successful performance at Malargüe strengthens confidence in the robustness of the GRAND detection concept under non-ideal field conditions.

4. Galactic background measurement and calibration strategy

A major early technical objective was calibration and radio-background characterization using diffuse Galactic emission. Because Galactic radio noise modulates with local sidereal time, it is a standard calibration source for low-frequency radio arrays. For GRAND@Auger, this study used data from March through June 2024, a period judged to have sufficiently stable hardware (Errico et al., 10 Jul 2025).

The initial data set included all ten stations, but extensive quality cuts were required. After excluding stations or periods affected by sporadic gain or noise problems, five units—58, 59, 69, 144, and 151—were rejected, and the Galactic analysis used stations 49, 60, 70, 83, and 84. Narrowband interference from video carriers, aviation communications, and satellite signals was removed through a two-stage procedure: online notch filtering in the FPGA and offline filtering during analysis. The paper illustrates the resulting spectrum with an average power spectral density for DU 83 derived from $16~\mu\mathrm{s}$-long traces with frequency resolution of about $62.5~\mathrm{kHz}$, showing both narrowband contaminants and usable broadband intervals (Errico et al., 10 Jul 2025).

After all cuts, about 50 days of data remained. Because this interval was too short to separate local-sidereal-time modulation unambiguously from 24-hour solar-day periodicity without simulation support, the collaboration modeled the expected Galactic signal with LFmap and propagated it through the GRAND RF chain within the GRANDlib framework, adding random Gaussian noise. The measured power in each LST bin was then fit with

$250~\mathrm{m}$0

where $250~\mathrm{m}$1 is the measured power, $250~\mathrm{m}$2 is the predicted Galactic contribution after detector response, $250~\mathrm{m}$3 is a fitted multiplicative calibration factor, and $250~\mathrm{m}$4 accounts for non-Galactic backgrounds and baseline system noise (Errico et al., 10 Jul 2025).

Three bands were analyzed: $250~\mathrm{m}$5, $250~\mathrm{m}$6, and $250~\mathrm{m}$7. The main result is that agreement with Galactic simulations is clearly better in the $250~\mathrm{m}$8 band. In that band, the gain calibration is understood at about $250~\mathrm{m}$9 for the horizontal polarizations and up to $0.5~\mathrm{km}^2$0 for the vertical polarization. Below $0.5~\mathrm{km}^2$1, and also across the full $0.5~\mathrm{km}^2$2 interval, the correlation is weaker, indicating remaining deficiencies in detector-response modeling and/or residual low-frequency interference or instrumental effects. The data and simulation both peak when the Galactic center passes Malargüe, around $0.5~\mathrm{km}^2$3 h LST. The paper interprets this as the first GRAND detection of the Galactic background at high frequencies and as a characterization of the Auger-site sky noise in the $0.5~\mathrm{km}^2$4 range (Errico et al., 10 Jul 2025).

5. First self-triggered candidate event in coincidence with Auger

The other primary result reported for GRAND@Auger is the first candidate cosmic-ray event detected by GRAND in coincidence with the Pierre Auger Observatory. The event occurred on 16 September 2024 at 20:06:08 UTC and was recorded as a self-triggered three-station coincidence involving DUs 49, 144, and 151. Raw voltage traces in the NS and EW channels show a coincident impulsive signal. DU 49 also recorded a coincident signal in its vertical channel, whereas the vertical channels of DUs 144 and 151 were not operational (Errico et al., 10 Jul 2025).

Using the three GRAND@Auger timestamps, the collaboration reconstructed the arrival direction with an analytic plane-wave fit: $0.5~\mathrm{km}^2$5 where $0.5~\mathrm{km}^2$6 is the azimuth measured from east and $0.5~\mathrm{km}^2$7 the zenith angle. The corresponding Auger event was a T5 surface-detector event with 22 SD stations participating. Auger’s own reconstruction, using a spherical-wave fit, gave

$0.5~\mathrm{km}^2$8

and the Auger surface-detector lateral-distribution fit yielded

$0.5~\mathrm{km}^2$9

The azimuth difference is about $250~\mathrm{m}$0, while the zenith differs by about $250~\mathrm{m}$1. The paper does not quote a formal GRAND@Auger directional uncertainty for this event, so the comparison is qualitative. It nevertheless treats the spatial and temporal coincidence, together with the directional consistency, as strong evidence that both instruments observed the same cosmic ray. A time offset of approximately $250~\mathrm{m}$2 was observed between GRAND@Auger and Auger, with GRAND@Auger triggering slightly earlier (Errico et al., 10 Jul 2025).

This event is presented both as the first high-energy cosmic-ray event observed in coincidence by GRAND and Auger and, more specifically, as the first autonomous radio detection of a cosmic ray by GRAND. Within the prototype program, that claim is significant because it demonstrates self-triggering without reliance on an external particle trigger from Auger.

6. Reconstruction context, limitations, and significance

GRAND@Auger should be interpreted against the background of the broader GRAND reconstruction program for highly inclined showers. Within the collaboration, current reconstruction development relies on a plane-wave initialization, refinement with a curved wavefront, and amplitude modeling through an Angular Distribution Function tailored to very inclined geometries. Those methods are being validated mainly on realistic simulations and GP300 data, and the collaboration explicitly links them to direction, energy, and composition reconstruction for inclined and potentially upward-going showers (Macias et al., 2024, Guelfand et al., 6 Jul 2025).

Against that backdrop, the first GRAND@Auger coincidence result is a commissioning benchmark rather than a full statement of final GRAND reconstruction performance. The GRAND@Auger paper does not report an event rate, trigger efficiency curve, energy threshold, livetime-integrated exposure, quantitative coincidence efficiency, false-trigger rate, signal-to-noise ratio, angular resolution, energy resolution, or reconstruction efficiency. It also does not provide detailed formulas for antenna effective height, noise temperature, waveform filtering, or direction-fit covariance; the only explicit equation in the paper is the Galactic calibration fit. This omission is deliberate in the sense that the proceedings contribution is framed as an overview of commissioning and preliminary results rather than a final detector-performance paper (Errico et al., 10 Jul 2025).

Even with those limitations, the prototype already clarifies several points. First, it shows that an autonomous radio array can operate stably for long periods in the field, including automatic hibernation management under environmental constraints. Second, it shows that the instrumental response in the $250~\mathrm{m}$3 band is already sufficiently understood to recover sidereal modulation of the Galactic background. Third, it shows that a self-triggered GRAND array can detect an ultra-high-energy air shower and reconstruct a direction broadly consistent with the Pierre Auger Observatory on an event-by-event basis (Errico et al., 10 Jul 2025).

Accordingly, GRAND@Auger is best understood as a small but strategically important reconstruction-validation and cross-calibration pathfinder. It does not yet provide the robust statistics required for a mature performance characterization, but it directly tests the central GRAND proposition: that sparse, autonomous radio stations can self-trigger on extensive air showers and that their reconstructions can be benchmarked against a reference observatory before the method is scaled to future large-area GRAND deployments.