IceCube-Gen2 Prototype Station Overview

• IceCube-Gen2 Prototype Station is a hybrid detector system combining scintillation panels, radio antennas, and a central DAQ to capture coincident air-shower signals.
• It employs a distributed prototyping approach that validates timing, calibration, and environmental resistance, informing future surface array designs.
• The station serves as a technology demonstrator, commissioning platform, and early physics instrument for the large-scale IceCube-Gen2 surface array.

Searching arXiv for the cited IceCube-Gen2 prototype-station papers and closely related instrumentation papers. The IceCube-Gen2 Prototype Station is a hybrid surface-detector station developed in the IceCube Surface Array Enhancement and explicitly framed as a pathway to the future IceCube-Gen2 surface array. In its South Pole realization, it combines scintillation detectors, radio antennas, and a central hybrid data-acquisition system within the IceTop environment, where it functions simultaneously as a technology demonstrator, a commissioning platform, and a first physics instrument for coincident air-shower measurements with IceTop. In the wider Gen2 program, however, the term does not denote a single observatory-wide prototype across all subsystems; overview papers instead describe a distributed prototyping strategy that includes the IceCube Upgrade, the IceTop enhancement, and radio pathfinders (Venugopal, 11 Jun 2025, Clark, 2021).

1. Historical and programmatic setting

The prototype-station program emerged from the limitations of IceTop, the existing surface array of the IceCube Neutrino Observatory. IceTop’s 162 ice-Cherenkov tanks have been scientifically successful for cosmic rays from PeV to EeV energies and as a veto for atmospheric backgrounds in the in-ice neutrino detector, but long-term snow accumulation attenuates and spatially distorts tank signals, degrading low-energy efficiency and increasing reconstruction uncertainties. The Surface Array Enhancement addresses this by adding detector types that are less affected by snow and by introducing radio sensitivity to the longitudinal shower development, especially $X_\mathrm{max}$, the depth of shower maximum (Venugopal, 11 Jun 2025).

The prototype lineage predates the mature South Pole station. Two scintillator R&D stations with different designs were deployed in January 2018, one of them was upgraded with two radio antennas in January 2019, and a new prototype station combining and improving the earlier iterations replaced them in January 2020. In parallel, the enhancement program was defined as a 32-station hybrid array within the current IceTop footprint, with each station centered on 8 scintillation detectors, 3 radio antennas, and one central hybrid DAQ. That station concept was already described as both the IceTop-enhancement baseline and the R&D path toward a future large-scale IceCube-Gen2 surface array (Haungs, 2021, Oehler et al., 2021).

Within the Gen2 design studies, the prototype is not treated as an isolated demonstrator but as the station archetype for a much larger surface system. The planned Gen2 Surface Array places one co-located station above each new optical string, with about 160 stations in total when IceTop enhancement and transition coverage are included. In that baseline, each station again contains eight scintillation detectors and three radio detectors, and most of the array follows the optical-string spacing of approximately $240\,\mathrm{m}$. The scientific motivation is dual: a surface veto for neutrino detection and a multi-component air-shower instrument whose coincidence aperture with the deep optical array is expected to increase by about a factor of 30 relative to current IceCube (Schröder, 2023).

A common misconception is that the literature documents one formally defined “IceCube-Gen2 Prototype Station” for the entire observatory. The Gen2 overview literature is more specific: it identifies early scintillator panels and radio antennas deployed with the IceTop enhancement as the nearest surface prototype program, while the IceCube Upgrade serves as “a platform for developing Gen2 technologies” for the optical array, and RNO-G is presented as a development site for radio-station technologies (Clark, 2021).

2. Station architecture and environmental design

The canonical South Pole prototype station is a compact hybrid unit. A fully functioning prototype station was deployed at the South Pole in 2020, its DAQ was improved in 2022, and the detectors were upgraded in 2023. In the South Pole surface-array papers, one station consists of 8 scintillation detectors, 3 radio antennas, and a central DAQ or fieldhub, with the station embedded in the IceTop environment so that IceTop reconstruction can be used as an external reference (Venugopal, 11 Jun 2025, Collaboration, 16 Oct 2025).

The scintillation subsystem measures charged particles. Ionizing particles produce light in scintillation bars; the light is collected by optical fibers and routed to a silicon photomultiplier, and the resulting signal is sent to a board called the uDAQ, where it is digitized. Earlier instrumentation papers specify that each scintillation detector has an active area of $1.5~\mathrm{m}^2$, built from 16 organic plastic scintillator bars read out by wavelength-shifting fibers and a silicon photomultiplier, while later performance papers emphasize the three-gain readout architecture used in the mature station to extend dynamic range (Oehler et al., 2021, Collaboration, 16 Oct 2025).

The radio subsystem uses three SKALA v2 antennas. In the South Pole prototype description these are the second generation of the log-periodic antennas proposed for SKA, and for the enhancement application they operate in the frequency band $70\text{--}350\,\mathrm{MHz}$, although the main 2025 air-shower analysis filters the data to $100\text{--}230\,\mathrm{MHz}$. Each antenna measures two mutually perpendicular components of the electric field, and the two polarizations are handled independently in both hardware and analysis. In the broader Gen2 surface-array design study, the same antenna family is described in a $50\text{--}350\,\mathrm{MHz}$ operating concept for the final array (Venugopal, 11 Jun 2025, Schröder, 2023).

The radio chain is instrumented in substantially more detail than the scintillator chain in the first South Pole physics paper. The antenna signal passes through a two-board low-noise amplifier with $40\,\mathrm{dB}$ amplification while minimizing added noise. Power is supplied by a bias tee on the radioTAD board, which also applies high-pass and low-pass filtering. The signal is then split into four parts per polarization and sent to a DRS4 ring buffer with $(8+1)$ channels. Once triggered, the radio waveform is read out, sampled, and digitized with an ADC at $1\,\mathrm{GHz}$ (Venugopal, 11 Jun 2025).

The station DAQ is TAXI, the adapted “Transportable Array for eXtremely large area Instrumentation studies.” TAXI is housed in a shielded or elevated electronics box, distributes timing and communications to the scintillators, digitizes the radio waveforms, and interfaces the station to IceCube infrastructure. Timing and synchronization use a White Rabbit node synchronized to the central timing from the IceCube Laboratory, although the prototype papers also note that TAXI is currently independent of the IceCube and IceTop electronics, which leaves one integration step for a future large deployment (Oehler et al., 2021, Venugopal, 11 Jun 2025).

Environmental design is central rather than incidental. The scintillators and radio antennas are mounted on supports more than $1\,\mathrm{m}$ above the snow, and the stands are designed to be raised roughly every 5 years. The prototype has operated for years under all weather conditions, and the reported snow-height increase under and around the detectors is about $240\,\mathrm{m}$0 per year, consistent with the natural background rate rather than an enhanced local accumulation caused by the structures. That result is one of the key reasons the elevated hybrid station became the baseline design for the Gen2 surface array (Schröder, 2023).

3. Triggering, timing, calibration, and analysis workflow

The prototype’s defining systems feature is its hybrid trigger. In the South Pole implementation, the antennas do not self-trigger; radio readout is initiated by the scintillator trigger. The trigger condition is a scintillator multiplicity of at least 6 above-threshold hits within a $240\,\mathrm{m}$1 window. Until that trigger arrives, radio data are continuously stored in the DRS4 ring buffer. This architecture is central to the Gen2 relevance of the station, because it suppresses radio background with a particle trigger while retaining buffered radio information for impulsive air-shower signals (Venugopal, 11 Jun 2025).

Event building is hybrid at the analysis level as well. A “coincident event” is formed offline by combining radio, scintillator, and IceTop information if their timestamps lie within a $240\,\mathrm{m}$2 window. Directional reconstruction of the radio component uses a plane-wave fit to the three antenna timing measurements, and the resulting direction is compared against IceTop because a single station with three antennas does not yet provide a fully independent radio reconstruction (Venugopal, 11 Jun 2025).

The radio signal-processing chain is explicitly documented. Although the DRS4 buffer can store $240\,\mathrm{m}$3, the analysis uses only $240\,\mathrm{m}$4 readout windows. For each antenna polarization, the median of the four redundant DRS4 readout channels is used. Readout artifacts are cleaned, the pedestal is removed, the baseline is restored to zero, ADC counts are converted to voltage, and a spike filter based on median background spectra suppresses frequency bins with unusually high noise. The waveform is then filtered to $240\,\mathrm{m}$5, and the electronics response of the LNA, cable, and DAQ is deconvolved. Background antenna data taken with a fixed-rate trigger determine the SNR threshold; the threshold is chosen such that $240\,\mathrm{m}$6 of background events are rejected in each antenna channel, and an air-shower candidate must have at least one channel above threshold in each antenna (Venugopal, 11 Jun 2025).

The mature scintillator-processing chain is described separately for Station 0 after the 2022 firmware update. Its stages are: check uDAQ firmware version, extract hitbuffer data, process monitoring data, COBS decoding, interpret binary data, process hitbuffer data, convert to MIP units, search for coincident events, scintillator reconstruction, and produce processed dataset and supplementary data. Signals are calibrated in units of MIP, the charge deposited by a single vertical MIP. The calibration includes temperature-dependent gain correction in the high-gain channel, MIP peak calibration in medium- and low-gain channels, and gain-scaling factors between channels to correct saturation. The MIP distribution in each gain channel is modeled by a Landau convolved with an exponential (Collaboration, 16 Oct 2025).

The three-gain architecture is one of the most quantitative aspects of the scintillator station. Using commissioning and calibration data from January–June 2023 for uDAQ version 4.1a, the reported gain-scaling factors are $240\,\mathrm{m}$7 for high gain / medium gain, $240\,\mathrm{m}$8 for medium gain / low gain, and $240\,\mathrm{m}$9 for high gain / low gain. The paper highlights the overall factor of $1.5~\mathrm{m}^2$0 as the operative dynamic-range extension, with reconstruction using high gain until saturation, then medium gain, then low gain (Collaboration, 16 Oct 2025).

Temperature calibration required additional handling after the 2022 firmware upgrade reduced onboard monitoring to one temperature readout per run. Station temperatures from earlier periods were correlated with South Pole ARO weather-station / IceCube Live temperature using RANSAC. The resulting seasonal fits led to a reported summer offset of approximately $1.5~\mathrm{m}^2$1, which is incorporated into the calibration chain (Collaboration, 16 Oct 2025).

The prototype literature also documents a methodological divergence between conventional and ML-based radio searches. The South Pole air-shower paper notes a traditional SNR-based search and a CNN-based event search, with the latter identifying a higher event rate, but it does not provide quantitative side-by-side performance metrics. That distinction is significant because the station is used not only to validate hardware but also to benchmark low-level classification strategies relevant to future large deployments (Venugopal, 11 Jun 2025).

4. South Pole observations and measured performance

The first South Pole physics result from the prototype was the successful identification of radio air showers in coincidence with IceTop. After requiring radio-direction agreement with IceTop at an opening angle of less than $1.5~\mathrm{m}^2$2, 102 events remain in the January–July 2022 data set used in the analysis. The accepted events show the expected geomagnetic dependence: the event yield increases with $1.5~\mathrm{m}^2$3, where $1.5~\mathrm{m}^2$4 is the geomagnetic angle, and the sky distribution shows fewer events near the geomagnetic-field direction. This behavior is the standard physical consistency check for cosmic-ray radio emission (Venugopal, 11 Jun 2025).

The same analysis ties the prototype sample to standard IceTop observables through $1.5~\mathrm{m}^2$5, the IceTop energy proxy defined as the expected signal strength at a reference distance of $1.5~\mathrm{m}^2$6 from the shower axis. The identified radio events are shown as a function of $1.5~\mathrm{m}^2$7, and pure background data processed through the same pipeline imply about two false positive events after reconstruction and the $1.5~\mathrm{m}^2$8 opening-angle cut, consistent with the presence of two low-$1.5~\mathrm{m}^2$9 events below $70\text{--}350\,\mathrm{MHz}$0. The paper presents this as an early purity estimate rather than a full detector-performance characterization (Venugopal, 11 Jun 2025).

The prototype also demonstrated true threefold coincidences among scintillators, radio antennas, and IceTop. In the example-event displays, scintillator charges and times, radio pulses with Hilbert envelopes, reconstructed directions, and reconstructed core positions from different detector components are shown together. These examples are important because they make explicit the complementary observables provided by the hybrid station: particle densities and times from scintillators, radio pulse amplitudes and times from antennas, and the wider-array shower geometry from IceTop (Venugopal, 11 Jun 2025).

A later performance paper focuses on the post-commissioning scintillator subarray of Station 0, which has been taking data in its final design since 2023. The major operational improvement after the August 2022 firmware update is an up-time of up to 23 hours/day, with the remaining $70\text{--}350\,\mathrm{MHz}$1 hour/day used to transfer data from the central DAQ to the ICL machine. The station recorded 279,280 events identified with six or more scintillators, and a high-quality benchmark sample of 27,582 events was then selected by requiring successful reconstructions in both the scintillator station and IceTop, scintillator multiplicity exactly 8, IceTop zenith $70\text{--}350\,\mathrm{MHz}$2, IceTop core within $70\text{--}350\,\mathrm{MHz}$3 of the fieldhub, $70\text{--}350\,\mathrm{MHz}$4, and scintillator zenith $70\text{--}350\,\mathrm{MHz}$5 (Collaboration, 16 Oct 2025).

For that benchmark sample, the reported 68th-percentile angular differences between the scintillator reconstruction and IceTop are $70\text{--}350\,\mathrm{MHz}$6 for zenith with mean $70\text{--}350\,\mathrm{MHz}$7, $70\text{--}350\,\mathrm{MHz}$8 for azimuth with mean $70\text{--}350\,\mathrm{MHz}$9, and $100\text{--}230\,\mathrm{MHz}$0 for the solid angular distance $100\text{--}230\,\mathrm{MHz}$1. The same paper emphasizes that the solid-angle metric is the more meaningful summary because azimuth becomes ill-conditioned for near-vertical events. As a function of shower size, the resolution is roughly $100\text{--}230\,\mathrm{MHz}$2 for $100\text{--}230\,\mathrm{MHz}$3, then gradually worsens and plateaus around $100\text{--}230\,\mathrm{MHz}$4 at larger $100\text{--}230\,\mathrm{MHz}$5, a behavior attributed to the compact footprint of Station 0 relative to larger shower footprints (Collaboration, 16 Oct 2025).

The same performance study also introduces a data-driven signal-fluctuation model based on neighboring detector pairs separated by $100\text{--}230\,\mathrm{MHz}$6. Excluding detectors within $100\text{--}230\,\mathrm{MHz}$7 of the reconstructed core, the signal relation is fitted as $100\text{--}230\,\mathrm{MHz}$8, and the width of the pairwise signal-difference distribution above $100\text{--}230\,\mathrm{MHz}$9 follows a power law with slope $50\text{--}350\,\mathrm{MHz}$0. Dedicated single-station simulations reproduce this with slope $50\text{--}350\,\mathrm{MHz}$1. The paper interprets this as approximately $50\text{--}350\,\mathrm{MHz}$2, i.e. Poisson-like fluctuations (Collaboration, 16 Oct 2025).

These South Pole results are intentionally uneven in maturity. Radio performance is demonstrated through successful detection and physical consistency checks, but the 2025 radio paper does not yet provide threshold curves, trigger efficiencies, timing precision, angular resolution, or $50\text{--}350\,\mathrm{MHz}$3 performance. Conversely, the later Station 0 scintillator paper provides explicit angular-resolution and uptime metrics for the particle subarray, while deferring radio performance to other contributions. The prototype therefore occupies an intermediate stage between technology demonstrator and fully characterized observatory subsystem.

5. External deployments and cross-observatory prototyping

The prototype program is not confined to the South Pole. Status papers on the Surface Array Enhancement state that one station each has been installed at the Pierre Auger Observatory and the Telescope Array for further R&D in different environmental conditions. The Auger deployment became the most fully documented external prototype because it provided an immediate cross-check against an established air-shower observatory (Shefali et al., 2024).

At Auger, the IceCube-Gen2 surface prototype station closely follows the planned South Pole station concept but is deployed inside the dense SD433 region and within AERA infrastructure. The station consists of 8 scintillator panels, 3 SKALA-v2 radio antennas, and a central DAQ system referred to as TAXI. Its geometry is a three-arm star-like configuration with three arms of about $50\text{--}350\,\mathrm{MHz}$4 length, a central pair of scintillators at the station center, one pair at the end of each arm, and one radio antenna in the middle of each arm. Power and data collection are provided through the Central Radio Station of AERA, and timing is distributed via a WhiteRabbit switch at the CRS (Verpoest, 2024, Verpoest et al., 11 Jul 2025).

The Auger radio program first established that the station sees real sky noise. Fixed-rate-trigger data showed strong anthropogenic RFI near $50\text{--}350\,\mathrm{MHz}$5 and $50\text{--}350\,\mathrm{MHz}$6, but after filtering to $50\text{--}350\,\mathrm{MHz}$7 the waveform RMS displayed predominantly sidereal modulation, the expected signature of Galactic radio noise. For air-shower searches, traces were filtered to $50\text{--}350\,\mathrm{MHz}$8, and candidate events were required to have at least one polarization in each antenna above the 95th-percentile background SNR threshold, followed by a planar-wavefront fit to the three antenna times (Verpoest, 2024).

The first Auger analysis reported 50 coincident events in approximately 3 months of stable data, using May 2023 data taken at $50\text{--}350\,\mathrm{MHz}$9 and August 2023 plus January 2024 data taken at $40\,\mathrm{dB}$0. The opening-angle distribution between the prototype radio reconstruction and Auger SD433 peaks around $40\,\mathrm{dB}$1, and the selected events have energies starting at several tens of PeV with an energy distribution peaking around $40\,\mathrm{dB}$2. A later analysis over the period from 2023-01-01 to 2024-02-12 found 135 matched radio–SD433 events initially, of which 118 retained a successful SD433 energy reconstruction, corresponding to $40\,\mathrm{dB}$3 events/day over 241.2 days of usable runtime. In that longer analysis the opening-angle distribution peaks near $40\,\mathrm{dB}$4, core positions are mostly within $40\,\mathrm{dB}$5 of the station, and the energy distribution peaks around $40\,\mathrm{dB}$6 (Verpoest, 2024, Verpoest et al., 11 Jul 2025).

The Auger prototype also exposed operational limitations with direct relevance to Gen2 engineering. The first analysis required daily statistical determination of the global time offset between the prototype DAQ and the Auger SD433 array because of a time synchronization issue in the prototype station. The later paper still used 12-hour recalibration of the inter-system offset and a $40\,\mathrm{dB}$7 matching window. Both analyses also report DAQ artifacts that can mimic near-vertical air showers, forcing a restricted zenith acceptance in the current event selection (Verpoest, 2024, Verpoest et al., 11 Jul 2025).

The external prototypes therefore serve two roles. Scientifically, they demonstrate that the hybrid station concept can recover radio air showers in coincidence with an established external array. Technically, they provide a more accessible environment than the South Pole for debugging timing, noise handling, and reconstruction, while also opening the possibility of a future cross-calibration of the Auger and IceCube cosmic-ray energy scales.

6. Role in the Gen2 roadmap and remaining limitations

For the surface program, the prototype station has already validated the core architectural choices that define the Gen2 surface-array baseline. First, scintillators plus radio antennas can operate together at the South Pole and deliver coincident air-shower observations. Second, a scintillator multiplicity trigger is a practical method for radio readout in a low-background design. Third, the South Pole radio response shows the expected geomagnetic scaling, supporting the site and hardware choice. Fourth, the elevated detector concept avoids the long-term snow-burial problem that afflicts IceTop tanks and can be maintained by periodic raising (Venugopal, 11 Jun 2025, Schröder, 2023).

The broader Gen2 design studies translate these prototype lessons into an observatory-scale surface array. The planned array uses one surface station per optical string, approximately $40\,\mathrm{dB}$8 station spacing over most of the footprint, and the same basic station content of eight scintillation detectors and three radio antennas. In simulation studies, the scintillator system reaches a full-efficiency threshold of about $40\,\mathrm{dB}$9 for vertical proton showers, while the radio component is expected to become fully efficient for a significant fraction of the sky starting at a few $(8+1)$0, especially for zenith angles between $(8+1)$1 and $(8+1)$2. The South Pole prototype’s current radio threshold is described as around $(8+1)$3, with future reduction expected from improved DAQ electronics and neural-network pulse detection (Schröder, 2023).

Several limitations are equally explicit. With only one South Pole station and three antennas, radio reconstruction is not yet robustly stand-alone and must be cross-checked against IceTop. The main radio paper therefore imposes a $(8+1)$4 opening-angle agreement cut with IceTop, reports only modest event statistics, and does not provide full threshold, efficiency, angular-resolution, core-resolution, or timing-precision curves. Its $(8+1)$5 discussion is limited to motivation and references to earlier conference contributions rather than a developed reconstruction procedure (Venugopal, 11 Jun 2025).

The later Station 0 performance paper resolves some of these gaps for the scintillator subsystem but not for the full hybrid station. It demonstrates extended dynamic range, higher data-taking up-time, and quantitative angular performance of the scintillator station, yet radio performance and array-level hybrid reconstruction are still deferred. This division of labor in the literature is itself informative: the prototype has moved from proof of operation to subsystem-level characterization, but not yet to a complete end-to-end Gen2 performance envelope (Collaboration, 16 Oct 2025).

A second misconception concerns scope. Separate papers on the optical sensor for IceCube-Gen2 describe prototype optical modules being tested in the IceCube Upgrade, including modules intended to serve as prototypes for the planned mass production of about 10,000 OMs and twelve Gen2-DOM prototypes to be deployed in the 2025/2026 austral summer. These are prototype sensors rather than prototype surface stations. The Gen2 program therefore remains best described as a distributed prototyping ecosystem: the hybrid surface station for the surface array, the IceCube Upgrade for optical and calibration technologies, and external radio pathfinders for station-level radio development (Kappes, 11 Jul 2025, Butterfield et al., 9 Apr 2026).

Taken together, the prototype-station literature shows that the Gen2 surface concept has passed its basic feasibility stage. The station design has been operated under South Pole conditions, has produced hybrid air-shower data with IceTop, has demonstrated parallel deployment at Auger for cross-checks and R&D, and has progressed to quantitative scintillator performance validation in its final 2023 design. What remains before full Gen2 deployment is not proof that the station works, but completion of the array-scale steps: multi-station independent reconstruction, tighter DAQ integration where desired, and full characterization of efficiency, resolution, and composition sensitivity.