Radio antennas have become a standard tool for the detection of cosmic-ray air showers in the energy range above
This article reviews the status and recent progress regarding the radio detection of extensive air showers-atmospheric particle cascades initiated by high energy cosmic particles. It complements my other reviews on this topic [1][2][3] and provides an updated overview of current and planned experiments featuring radio antennas for air-shower detections. For a more complete view of the topic including theory and the physics behind simulations of the radio signal, I also recommend reading reviews by other authors, such as [4,5]. Particular emphasis in this review is given on recent experimental progress.
The radio signal of air showers is mostly generated by the electrons and positrons of the electromagnetic shower component and beamed in the forward direction. The emission is coherent, i.e., the amplitude of the radio signal scales approximately linearly and the power quadratically with the number of electromagnetic particles, which itself is approximately proportional to the energy of the primary particle [6]. The coherence conditions for the emission are optimal under the Cherenkov angle, which is of order 1 • in air, where the radio emission extends to the highest frequencies of a few GHz [7]. As the shower front has a thickness of only a few meters, the radio emission is a short pulse, few ns to few 100s of ns, and has a broad frequency spectrum with typical experiments operating in a band somewhere between 30 MHz and a few GHz. Because the unavoidable galactic radio noise decreases with frequency, the frequency band has impact on the signal-to-noise ratio [8,9]. Hence, the frequency band is an important aspect of the design of an antenna array and ideally is optimized for the science goals of an experiment.
The main mechanism of air-shower radio emission is the deflection of the electrons and positrons in the shower front by the Earth's magnetic field. This induces a time varying transverse current leading to radio emission linearly polarized in the v × B direction in the shower plane, where v is the shower axis and B the direction of the magnetic field. Therefore, the strength of the radio emission does not only depend on the zenith angle, but Fig. 1 Air showers with different depth of shower maximum (A-C, where A is closest to the ground) and the lateral distribution of the energy fluence of their radio emission illustrated in two ways: through the darkness of the color of the two-dimensional footprint and through the one-dimensional lateral distribution. The asymmetry is caused by interference of geomagnetic and Askaryan emission. The dip in the center is visible only for distant shower maxima and results from the emission being enhanced at the Cherenkov angle. Reprinted from [10] with permission from Elsevier Fig. 2 Amplitude of the Askaryan emission A relative to the geomagnetic G emission for the ground level at the geographic South Pole for two different zenith angles as function of the distance from the shower axis and the distance to the shower maximum (determined with CoREAS simulations of the radio emission of air showers). The geomagnetic amplitude is corrected for the size of the geomagnetic Lorentz force with the plotted Askaryan fraction defined as a rel = A G sin α, with α the angle between the Earth's magnetic field and the shower axis (Fig. modified from Ref. [24]) also on the azimuth angle of the air showers, which needs to be taken into account when modeling the aperture of radio arrays [11]. For very inclined showers and at high frequencies, the polarization pattern of geomagnetic emission becomes more synchrotron-like, which indicates that the curved tracks of the deflected electrons and positrons are important and not just the induction of the transverse current [12,13]. Moreover, Askaryan emission due to the net charge excess in the shower front plays a small, but non-negligible role in air showers (see Fig. 2). Although it is an order of magnitude weaker than geomagnetic emission for most shower geometries, it interferes with geomagnetic emission, leading to an asymmetric lateral distribution of the radio signal on ground. Figure 1 illustrates that asymmetric radio footprint on ground and its dependence on the distance to the shower maximum. The size of the radio footprint can be as small as 100 -200 m in diameter for vertical showers or extend over several 10 km for near-horizontal air showers [14].
When aiming at an accurate reconstruction of the air-shower direction, energy, and depth of maximum (X max ), several subtle effects need to be taken into account to achieve the best possible accuracy: for the direction, a planewave fit of the arrival times may achieve a resolution of O(1 • ), which is usually sufficient for cosmic-ray physics. Nonetheless, sub-degree resolution has been achieved when taking into account that the radio wavefront of air showers is approximately a hyperboloid [15,16], which in some cases can be further approximated by a spherical or conical wavefront. For an accurate reconstruction of the energy of the air shower, the azimuthal asymmetry of the radio footprint on ground is important [6,17,18], and one needs to be aware that the radiation energy is a not a direct measure for the total shower energy, but proportional to the size of the electromagnetic shower component. Depending on the method used, such as interferometry [19,20] or template fitting [21][22][23], the shape of the radio wavefront and the lateral distribution are also important for an accurate reconstruction of X max .
Radio pulses from air showers were first measured in the 1960s with analog experiments [25]. While these measurements were sufficient to confirm the main features of geomagnetic radio emission, the accuracy obtained for important air-shower observables, such as the energy and depth of shower maximum could not compete with other techniques. This changed only with digital radio arrays combined with computational pipelines for data analysis [26][27][28], including modern methods, such as template fitting [21,22,29,30], and artificial neural networks [31][32][33]. Moreover, improved techniques for time [34,35] and amplitude calibration [36,37] of radio arrays have enabled a high data quality of current experiments. Finally, state-of-the-art Monte Carlo simulation codes, such as CoREAS [38] and ZHAireS [39], describe the radio emission in sufficient detail for an accurate interpretation of measured data.
LOPES [19] and CODALEMA [40] were a first generation of digital radio arrays, providing several proof-ofprinciples, such as for digital radio interferometry of air showers [41] and for the sensitivity of radio measurements to the longitudinal shower development [42]. AERA [30], LOFAR [44], TREND [45], and Tunka-Rex [18] belong to a second generation that demonstrated a precision for the reconstructed shower energy and X max comparable to the established air-fluorescence and air-Cherenkov techniques [6,21,22,30]. As enhancements of particledetector arrays, radio detectors can thus enhance the measurement accuracy for the cosmic-ray energy and mass composition.
Nowadays, a third generation of digital radio experiments has started aiming at cosmic-ray science in the energy range above 10 16 eV, with some experiments like the AugerPrime radio upgrade of the Pierre Auger Observatory under construction [46], and others planned, such as the IceCube-Gen2 surface array [47], with prototype stations in operation [23]. Following the example of LOFAR, also other antenna arrays built primarily for radio astronomy can also detect air showers in parallel to astronomical observations, in particular, OVRO [48] and the future SKA-low array [49]. Moreover, several experiments designed primarily to search for neutrino-induced radio signals also detected air showers, such as the ground-based radio arrays ARIANNA [50], RNO-G [51], TAROGE [52], BEACON [53], and GRAND (with its GRANDproto300 protoype array for cosmic-ray air showers) [54], and the balloon-borne radio probes ANITA [55], PUEO [56], and POEMMA Balloon Radio (PBR) [57].
Figure 3 and Table 1 provide an overview of the location and characteristics of several of these radio experiments for air-shower detection. A table with the abbreviated and full names of the experiments can be found at the beginning of this article (Caveat: In some cases, the full name does not reflect the later stages of the projects, e.g., because of additional locations or antenna types used.
). Many of these radio arrays are triggered by other air-shower detectors, such as arrays of particle detectors, and share a common infrastructure for data taking. Self-triggering, which is required for stand-alone radio detection, has been demonstrated by several experiments [45,48,55,58]. However, self-triggering increases the detection threshold compared to external air-shower triggers and is mostly important for radio detectors aiming at huge exposure for ultrahigh-energy neutrinos searches. For many cosmic-ray science goals, not only statistics, but also high accuracy is important, which is more easily achieved by hybrid detection of air showers with several techniques. Thus, an external trigger comes almost for free in air-shower arrays featuring coincident detection of air showers with particle detectors and radio antennas. The numbers of antennas, the approximate areas covered with antennas, and the nominal frequency bands are given. However, many experiments were used also in configurations different from the one stated in the table or feature additional antennas for other purposes than air-shower detection, and many analyses are based on subsets of antennas and smaller sub-bands Fig. 4 Figure of merit (FOM) for the event-by-event sensitivity of various air-shower observables and their combination to separate air showers initiated by protons and iron nuclei. The results are for CORSIKA simulations with Sibyll 2.3c and 2.3d in the zenith angle range of 40 • -60 • for the sites of the Pierre Auger Observatory and IceCube, respectively. Xmax and the muon number Nµ provide the highest separation power, where muons for the Auger site are muons at 800 m axis distance at the surface and for IceCube high-energy muons as detectable by the deep detector. R e/µ is the electron-muon ratio at ground. Xmax, R, and L are parameters of the Gaisser-Hillas function describing the longitudinal shower profile. No detector simulation has been done, but some measurement uncertainties have been included as stated in the legend (figure modified from Ref. [59])
Several future projects plan to use radio instrumentation, mostly to increase the accuracy of air-shower measurements. In particular, increasing the accuracy for the rigidity of cosmic-ray nuclei is important to search for the sources of the most energetic galactic and extragalactic cosmic rays through mass-sensitive anisotropy [60,61]. In addition to an accurate energy determination, which can be provided by radio detectors, this primarily requires event-by-event mass sensitivity. Provided that the shower energy is known, the two most mass sensitive shower observables are the the depth of shower maximum, X max , and the muon number N µ . Therefore, air-shower arrays featuring both radio and muon detection are ideal for this purpose [59,62].
In addition to further efforts to increase the accuracy of radio measurements through better calibration and analysis techniques, a number of new projects are planned for the next decade relying on radio detection, some of them making use of coincident shower measurements of radio and muon detectors:
IceCube-Gen2 [47] will continue a long tradition of cosmic-ray physics at the South Pole [63], as cosmic rays in the energy range of the galactic-to-extragalactic transition will complement the primary science case of neutrino astronomy. The surface array will feature scintillation panels and radio antennas for air showers [64], which in combination with high-energy muons measured by an optical array deep in the ice will provide high per-event mass sensitivity.
GCOS will consist of air-shower arrays at several sites, together increasing the exposure of current ultrahighenergy cosmic-ray observatories by an order of magnitude and at the same time featuring per-event mass sensitivity [65]. At least one site will likely feature radio antennas, possibly for calibration or as enhancement of water-Cherenkov detectors following the concept of AugerPrime, the upgrade of the Pierre Auger Observatory. Among several other improvements of the Pierre Auger Observatory, such as new electronics and scintillation panels, a radio antenna has been added to each surface detector to provide a radio measurement of highly inclined air showers [46].
There are also science cases in cosmic-ray physics for future radio arrays without muon detectors. Examples are the huge exposure that can be provided by GRAND [54]. As GCOS, GRAND will feature several sites, with a total area even larger than GCOS. Although GRAND will be optimized for neutrino detection, it will also collect a huge statistics of ultrahigh-energy cosmic rays, and common sites between GRAND and GCOS can leverage synergies.
For the PeV-to-EeV energy range, SKA-low will provide the option to observe air showers in parallel to radio astronomy. The huge number of several 10, 000s of antennas will enable measuring the radio emission of air showers in unprecedented detail, including the shape parameters L and R for the width and asymmetry of the longitudinal shower profile, i.e., the number of particles as a function of atmospheric depth X :
exp -X -X max LR . While L R do not feature significant event-by-event mass separation power (cf. Fig. 4), they provide a novel approach to test hadronic interaction models and to statistically determine the proton and helium fractions among the primary cosmic-ray particles [66]. Last but not least, there are significant synergies among these experiments, e.g., the SKALA antennas developed for SKA-low [67] are also the reference design for the IceCube-Gen2 surface array, and prototype stations with SKALA v2 antennas have been deployed not only at the South Pole, but also at the Telescope Array and at the Pierre Auger Observatory (see Fig. 5).
Digital antenna arrays have been shown to feature an accuracy for the arrival direction, energy, and X max competitive with established detection techniques for air showers. This achievement has been enabled by progress in the instrumentation and its calibration, an improved understanding of the radio emission of air showers, and by sophisticated computational analysis techniques. While progress on all of these aspects continues, the level achieved is already sufficient to make radio detection an essential technique for cosmic-ray physics of the coming decade. We are therefore seeing a number of experiments targeting both particle physics and astrophysics questions related to the most energetic galactic and extragalactic cosmic rays with the help of radio antennas.