CoREAS: Monte Carlo Radio Emission Simulator

• CoREAS is a Monte Carlo simulation code that models radio emission from extensive air showers using a first-principles, microscopic approach.
• It integrates with CORSIKA to compute full time-dependent radiation fields from charged particle endpoints without relying on intermediate histogramming.
• CoREAS accurately reproduces both geomagnetic and Askaryan signals across varied frequencies, enabling precise reconstruction of shower energy and Xₘₐₓ.

to=arxiv_search 天天中彩票这个json code lower? to=arxiv_search ाठमाडौं code 《凤凰大参考json {"query":"CoREAS cosmic ray air shower radio emission", "max_results": 10} CoREAS is a Monte Carlo code for the simulation of radio emission from extensive air showers, implemented directly in CORSIKA and based on the endpoint formalism for radiation from moving charges (Huege et al., 2013). In this formulation, the radio signal is computed from the full time-dependent radiation field of the electrons and positrons tracked in the air shower, without assumptions on the emission mechanism and without intermediate histogramming of the shower particles (Huege et al., 2013). Across the literature, CoREAS is described as parameter-free, first-principles, and microscopic; it has been used to model MHz-to-GHz emission, to compare simulations with measurements from arrays such as LOPES and LOFAR, and to support radio-based reconstruction of shower energy and shower maximum $X_{\max}$ (Huege et al., 2013).

1. Origins, scope, and integration into CORSIKA

CoREAS was introduced as the successor to REAS 3.11 and as the end-point-formalism implementation of the radio-emission module integrated into CORSIKA (Apel et al., 2013). Its defining architectural feature is that the radio calculation is performed during the air-shower simulation itself rather than as an afterburner using histogrammed particle distributions (Huege et al., 2013). CORSIKA propagates charged particles in small steps, and CoREAS computes the radio contribution from the endpoints of each track segment and accumulates the electric field at user-specified observer positions (Huege et al., 2013).

This integration has two immediate consequences. First, CoREAS preserves correlations in the simulated particle distributions, including energy, angle, and lateral position, because no intermediate histogramming is performed (Huege et al., 2013). Second, it incorporates the full electromagnetic shower development alongside the radio calculation, including the effects of geomagnetic deflection, charge excess, and atmospheric refractivity (Apel et al., 2015). In the 2013 description, benchmarks indicated a speed-up of up to a factor of ten over the predecessor REAS3.1 (Huege et al., 2013).

Later work extended the CoREAS ecosystem beyond the original FORTRAN-based CORSIKA 7 implementation. In CORSIKA 8, the radio code was re-implemented as a modular observer in a C++17 framework designed to support both the CoREAS and ZHS formalisms (Alameddine et al., 2021). That re-write preserved the endpoint-based physics while introducing a plugin architecture, HDF5/ROOT output, automated validation, and a path toward GPU acceleration (Alameddine et al., 2021). This suggests that CoREAS has evolved from a specialized CORSIKA extension into a more general radio-emission framework embedded in modern air-shower simulation infrastructure.

2. Electrodynamic formulation and emission content

The central theoretical ingredient of CoREAS is the endpoint formalism, derived from the Liénard–Wiechert fields (Huege et al., 2013). In the time-domain presentation, the vector potential for a point charge $q$ with trajectory $\mathbf r(t)$ at observer position $\mathbf x$ is written as

$$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$

with the electric field obtained by

$$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$

After Fourier transformation and discretization into track endpoints, CoREAS evaluates a sum over segment endpoints rather than a continuous acceleration integral (Huege et al., 2013).

In practical terms, each charged-particle step is represented as a straight segment with a start and an end. The radiation is then computed from the velocity change at those endpoints, and the total field is the coherent superposition of all endpoint contributions from all shower particles (Huege et al., 2013). Because the method is based on the exact particle tracks provided by CORSIKA, CoREAS does not need to impose separate sub-models for different emission channels (Huege et al., 2013).

Within this framework, the two dominant radio-emission mechanisms emerge automatically. The geomagnetic contribution arises from the transverse acceleration of shower electrons and positrons in the geomagnetic field and is dominant (Apel et al., 2013). The Askaryan, or charge-excess, contribution arises from the net negative charge buildup in the shower front and is subdominant but non-negligible (Apel et al., 2013). For LOPES energies and geomagnetic field strength, the Askaryan contribution was summarized as approximately $5$–$10\%$ of the geomagnetic contribution (Apel et al., 2013). A South Pole simulation study expressed these components in shower-plane coordinates as

$$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$

along the $\mathbf v\times(\mathbf v\times\mathbf B)$ axis, and defined both the amplitude ratio

$q$0

and the Askaryan fraction

$q$1

for characterizing their relative importance (Paudel et al., 2021).

A key consequence of the endpoint treatment is that no explicit choice between “geomagnetic,” “geosynchrotron,” and “charge-excess” descriptions is required at the algorithmic level (Huege et al., 2013). The literature repeatedly emphasizes that both the dominant geomagnetic signal and the charge-excess signal are recovered from the same microscopic electrodynamics (Huege et al., 2013, Huege et al., 2013). At GHz frequencies, however, a 2013 study noted that the observed “clover-leaf” polarization structure might indicate that a direct geosynchrotron acceleration component becomes non-negligible at the shortest wavelengths, while also stressing that this point remained under study (Huege et al., 2013). That constitutes a clear example where the simulations reproduce a pattern but the physical interpretation is treated cautiously.

3. Atmospheric refractivity, coherence, and observable structure

A defining physical feature of CoREAS is its treatment of the atmospheric refractive index. In early descriptions, CoREAS used a simple exponential parameterization,

$q$2

with $q$3 at sea level and scale height $q$4 (Apel et al., 2013). In another presentation, this appears as

$q$5

with $q$6 and $q$7 (Huege et al., 2013). The refractive index enters both the retarded-time calculation and the coherence condition.

The Cherenkov-like enhancement follows from the denominator $q$8 in the endpoint expression approaching zero near the Cherenkov angle (Huege et al., 2013). In simplified form, the local Cherenkov condition is written as

$q$9

or, for $\mathbf r(t)$0,

$\mathbf r(t)$1

(Huege et al., 2013, Apel et al., 2013). This time compression is responsible for the well-known “bump” around an axis distance of approximately $\mathbf r(t)$2 in the lateral distribution and for rising lateral profiles in some LOPES events (Apel et al., 2013). The 2013 LOPES comparison stated explicitly that only simulations including the refractive index can reproduce rising lateral distributions (Apel et al., 2013).

At MHz frequencies, CoREAS predicts the familiar asymmetric footprint produced by the vector superposition of geomagnetic and Askaryan emission (Huege et al., 2013). In the 43–74 MHz band, individual showers were described by a two-parameter exponential lateral distribution,

$\mathbf r(t)$3

where $\mathbf r(t)$4 is the electric-field amplitude at $\mathbf r(t)$5 and $\mathbf r(t)$6 the slope parameter (Apel et al., 2013). In that same study, $\mathbf r(t)$7 was reported to scale nearly linearly with primary energy, $\mathbf r(t)$8 with $\mathbf r(t)$9, and the slope $\mathbf x$0 was found to correlate with zenith angle and $\mathbf x$1 (Apel et al., 2013).

At higher frequencies, the refractive index has even stronger consequences. In the 300–1200 MHz range, CoREAS predicts Cherenkov rings with field strengths $\mathbf x$2–$\mathbf x$3 times larger than off-ring values when the shower geometry places the Cherenkov cone on the ground (Huege et al., 2013). In the 3.4–4.2 GHz band, the Cherenkov ring persists but the polarization pattern changes qualitatively: the north–south component exhibits a characteristic “clover-leaf” quadrupole structure (Huege et al., 2013).

The same atmospheric refractivity that shapes coherence also contributes materially to reconstruction systematics. A dedicated LOFAR study found that a realistic $\mathbf x$4 variation in refractivity induces a systematic error in inferred $\mathbf x$5 between $\mathbf x$6 and $\mathbf x$7 in the 30–80 MHz band for proton showers with zenith angles from $\mathbf x$8 to $\mathbf x$9, and between $$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$0 and $$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$1 in the 120–250 MHz band (Corstanje et al., 2017). A later implementation of time-dependent event-specific atmospheres using GDAS in CORSIKA and CoREAS showed that under normal conditions the systematic shift in reconstructed $$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$2 was about $$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$3, but under extreme weather conditions it could reach $$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$4 (Mitra et al., 2020). For highly inclined showers, refractivity also causes a displacement of the radio-emission footprint with respect to the Monte Carlo shower impact point, reaching approximately $$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$5 in the ground plane for $$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$6 zenith (Schlüter et al., 2020).

4. Frequency coverage, wavefront structure, and geometry dependence

CoREAS has been used across a wide radio-frequency range, from tens of MHz to several GHz (Huege et al., 2013, Huege et al., 2013). In the tens-of-MHz regime relevant for LOPES, LOFAR, and AERA, the footprint is dominated by the interference of geomagnetic and Askaryan fields, with a pronounced east–west asymmetry (Huege et al., 2013). In this band, proton and iron primaries produce different lateral-distribution slopes because iron showers develop higher and therefore yield flatter footprints (Huege et al., 2013).

The wavefront structure of the radio pulse is another observable directly studied with CoREAS. For LOPES conditions, the wavefront was found to be approximately hyperbolic rather than spherical or conical (Schröder et al., 2015). In the notation of that study, the pulse-arrival time relative to the shower-core arrival time can be fitted by

$$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$7

where $$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$8 is the orthogonal distance to the shower axis, $$\mathbf A(\mathbf x,t) = \frac{\mu_0 q}{4\pi} \frac{\boldsymbol \beta(t_{\mathrm{ret}})} {R(t_{\mathrm{ret}})\left(1-\mathbf n\cdot \boldsymbol\beta(t_{\mathrm{ret}})\right)},$$9 is the orthogonal distance above or below the shower plane, $$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$0 is the cone-angle parameter, and $$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$1 is a time offset empirically found to be approximately $$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$2 (Schröder et al., 2015). The steepness of this hyperbolic wavefront was shown to be sensitive to the shower maximum.

After correcting for zenith-angle dependence, the same study obtained the empirical relation

$$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$3

with an $$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$4 resolution of approximately $$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$5 under ideal simulated conditions (Schröder et al., 2015). CoREAS also predicted a slight east–west asymmetry in the wavefront due to geomagnetic–Askaryan interference, but the arrival-time asymmetry was only at the $$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$6 level at $$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$7, much weaker than the amplitude asymmetry (Schröder et al., 2015). This suggests that symmetric hyperbolic fits are usually sufficient for practical reconstruction even though the fully microscopic calculation produces a weak asymmetry.

Shower geometry also strongly affects the size of the radio footprint. CoREAS simulations showed that the illuminated ground area increases strongly from near-vertical to very inclined showers (Huege et al., 2013). Reported characteristic scales include footprint radii of about $$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$8–$$\mathbf E(\mathbf x,t)=-\frac{\partial \mathbf A}{\partial t}.$$9 at $5$0 for $5$1, a major-axis radius of about $5$2 for $5$3, and extensions to approximately $5$4 at $5$5 and $5$6 at $5$7 along the major axis for $5$8 (Huege et al., 2013). These results underlie the frequent observation that very inclined air showers are especially promising for sparse radio arrays (Huege et al., 2013).

5. Comparison with measurements and resolution of the amplitude-scale tension

The most prominent experimental controversy associated with CoREAS concerned the absolute amplitude scale in comparisons with LOPES. In the 2013 LOPES comparison, CoREAS amplitudes were found to be lower than the measured $5$9 by roughly a factor of two, with a mean deviation of $10\%$0 even after allowing for the quoted systematic uncertainties of $10\%$1 in the LOPES calibration and $10\%$2 in the KASCADE-Grande energy scale (Apel et al., 2013). By contrast, REAS 3.11 agreed with the then-current LOPES amplitude scale to within about $10\%$3 on average (Apel et al., 2013). At that stage, the origin of the CoREAS-versus-REAS amplitude difference was explicitly described as not understood (Apel et al., 2013).

This tension was resolved in 2015 through a revised absolute amplitude calibration of LOPES. The original external reference source had been characterized under free-field conditions with a horizontal reflective ground plane, whereas cosmic-ray radio pulses required free-space calibration (Apel et al., 2015, Link et al., 2015). After the manufacturer re-measured the same reference source under free-space conditions, all previously published LOPES electric-field amplitudes had to be divided by a factor

$10\%$4

(Apel et al., 2015). The remaining one-sigma uncertainty on the absolute scale became about $10\%$5 (Apel et al., 2015).

With this recalibration, the event-by-event comparison to CoREAS moved from a factor-of-two discrepancy to consistency at the absolute amplitude level. One report quoted

$10\%$6

for proton primaries and $10\%$7 for iron primaries (Apel et al., 2015), while another quoted

$10\%$8

with $10\%$9 (Link et al., 2015). In the latter analysis, a histogram of $$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$0 had a Gaussian width of about unity and the $$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$1 distributions matched nearly perfectly, leaving only a small zenith-angle dependence at large $$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$2, plausibly due to imperfect antenna directivity modeling rather than a flaw in CoREAS (Link et al., 2015).

The recalibration therefore changed the interpretation of CoREAS fundamentally. What had appeared to be a significant amplitude deficit became, after calibration correction, evidence that the absolute field strengths predicted by the current simulations agree with experimental data (Apel et al., 2015). The same recalibrated comparison placed REAS 3.11 in tension with the measurements, requiring an ad hoc scaling factor $$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$3 to match the data (Apel et al., 2015). This episode is central to the historiography of CoREAS because it transformed a perceived simulation deficiency into a calibration correction and reinforced the status of CoREAS as the more accurate of the two codes.

6. Parametrizations, computational surrogates, and calibration applications

Because full CoREAS simulations are computationally demanding, substantial work has focused on compressed parameterizations and surrogate models derived from CoREAS outputs. For LOFAR, a two-dimensional pulse-power footprint was parameterized with a double Gaussian requiring five fit parameters (Nelles et al., 2014). Those parameters were found to correlate strongly with shower energy, arrival direction, and $$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$4, and the parameterization reproduced CoREAS power footprints at the $$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$5–$$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$6 level while also fitting LOFAR measurements with reduced $$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$7–$$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$8 (Nelles et al., 2014).

A more explicitly simulation-accelerating approach is the template-synthesis method based on the universality of emission from small shower slices (Butler et al., 2019). In this framework, a high-statistics CoREAS shower is decomposed into slant-depth slices, and the synthesized field for a target shower is formed as

$$G \equiv |E_x| \approx |E_{\rm geo}|,\qquad A \equiv |E_y| \approx |E_{\rm Ask}|,$$9

with an extended frequency-domain formulation that also rescales the slice spectrum as a function of $\mathbf v\times(\mathbf v\times\mathbf B)$0 (Butler et al., 2019). The advanced synthesis reduced residuals to $\mathbf v\times(\mathbf v\times\mathbf B)$1 at all distances and both polarizations for the ensemble studied, while reducing run times from $\mathbf v\times(\mathbf v\times\mathbf B)$2 per shower to $\mathbf v\times(\mathbf v\times\mathbf B)$3 (Butler et al., 2019). This suggests that CoREAS has not only served as a forward model, but also as the training ground for fast emulators usable in dense-array reconstruction.

CoREAS has also been central to radio calorimetry. A dedicated study of the radiation energy in the 30–80 MHz band derived an efficient one-axis integration method and found that, after corrections for geomagnetic geometry and atmospheric density, the corrected radiation energy scales nearly quadratically with the electromagnetic shower energy (Glaser et al., 2016). The fitted relation

$\mathbf v\times(\mathbf v\times\mathbf B)$4

yielded $\mathbf v\times(\mathbf v\times\mathbf B)$5 and $\mathbf v\times(\mathbf v\times\mathbf B)$6, with an intrinsic uncertainty of $\mathbf v\times(\mathbf v\times\mathbf B)$7 in determining the electromagnetic energy when $\mathbf v\times(\mathbf v\times\mathbf B)$8 is not accessible (Glaser et al., 2016). In a later Auger Engineering Radio Array analysis using 844 events and per-event CoREAS simulations, the ratio of radio to fluorescence-detector energy scales was measured as

$\mathbf v\times(\mathbf v\times\mathbf B)$9

corresponding to radio energies $q$00 higher than the fluorescence-detector scale but consistent within uncertainties (Huege, 11 Jul 2025).

These developments indicate that CoREAS has become more than a simulation code for waveform prediction. It underpins energy-scale determinations, atmosphere-aware $q$01 reconstruction, parameterized footprint models, and fast synthesis schemes (Glaser et al., 2016, Mitra et al., 2020, Butler et al., 2019, Huege, 11 Jul 2025). A plausible implication is that the scientific role of CoREAS has broadened from validating the physics of radio emission to providing a reference model for precision calibration and high-throughput inference in radio detection of cosmic rays.