This paper presents the first study that uses the GEometry ANd Tracking 4 (GEANT4) toolkit to do quantitative comparisons with other modeling results related to the production of terrestrial gamma ray flashes and high-energy particle emission from thunderstorms. We will study the relativistic runaway electron avalanche (RREA) and the relativistic feedback process, as well as the production of bremsstrahlung photons from runaway electrons. The Monte Carlo simulations take into account the effects of electron ionization, electron by electron (Møller), and electron by positron (Bhabha) scattering as well as the bremsstrahlung process and pair production, in the 250 eV to 100 GeV energy range. Our results indicate that the multiplication of electrons during the development of RREAs and under the influence of feedback are consistent with previous estimates. This is important to validate GEANT4 as a tool to model RREAs and feedback in homogeneous electric fields. We also determine the ratio of bremsstrahlung photons to energetic electrons Nγ/Ne. We then show that the ratio has a dependence on the electric field, which can be expressed by the avalanche time τ(E) and the bremsstrahlung coefficient α(ε). In addition, we present comparisons of GEANT4 simulations performed with a “standard” and a “low-energy” physics list both validated in the 1 keV to 100 GeV energy range. This comparison shows that the choice of physics list used in GEANT4 simulations has a significant effect on the results.Key PointsTesting the feedback mechanism with GEANT4Validating the GEANT4 programming toolkitStudy the ratio of bremsstrahlung photons to electrons at TGF source altitude

Abstract. The emerging field of high energy atmospheric physics (HEAP) includes terrestrial gamma-ray flashes, electron–positron beams and gamma-ray glows from thunderstorms. Similar emissions of high energy particles occur in pulsed high voltage discharges. Understanding these phenomena requires appropriate models for the interaction of electrons, positrons and photons of up to 40 MeV energy with atmospheric air. In this paper, we benchmark the performance of the Monte Carlo codes Geant4, EGS5 and FLUKA developed in other fields of physics and of the custom-made codes GRRR and MC-PEPTITA against each other within the parameter regime relevant for high energy atmospheric physics. We focus on basic tests, namely on the evolution of monoenergetic and directed beams of electrons, positrons and photons with kinetic energies between 100 keV and 40 MeV through homogeneous air in the absence of electric and magnetic fields, using a low energy cutoff of 50 keV. We discuss important differences between the results of the different codes and provide plausible explanations. We also test the computational performance of the codes. The Supplement contains all results, providing a first benchmark for present and future custom-made codes that are more flexible in including electrodynamic interactions.

Several computer models exist to explain the observation of terrestrial gamma‐ray flashes (TGFs). Some of these models estimate the electric field ahead of lightning leaders and its effects on electron acceleration and multiplication. In this paper, we derive a new set of constraints to do more realistic modeling. We determine initial conditions based on in situ measurements of electric field and vertical separation between the main charge layers of thunderclouds. A maximum electric field strength of 50 kV/cm at sea level is introduced as the upper constraint for the leader electric field. The threshold for electron avalanches to develop of 2.86 kV/cm at sea level is introduced as the lower value. With these constraints, we determine a region where acceleration and multiplication of electrons occur. The maximum potential difference in this region is found to be ∼52 MV, and the corresponding number of avalanche multiplication lengths is ∼3.5. We then quantify the effect of the ambient electric field compared to the leader field at the upper altitude of the negative tip. Finally, we argue that only leaders with the highest potential difference between its tips (∼600 MV) can be candidates for the production of TGFs. However, with the assumptions we have used, these cannot explain the observed maximum energies of at least 40 MeV. Open questions with regard to the temporal development of the streamer zone and its effect on the shape of the electric field remain.

The emerging field of high-energy atmospheric physics studies how high-energy particles are produced in thunderstorms, in the form of terrestrial γ -ray flashes and γ -ray glows (also referred to as thunderstorm ground enhancements). Understanding these phenomena requires appropriate models of the interaction of electrons, positrons and photons with air molecules and electric fields. We investigated the results of three codes used in the community -Geant4, GRanada Relativistic Runaway simulator (GRRR) and Runaway Electron Avalanche Model (REAM) -to simulate relativistic runaway electron avalanches (RREAs). This work continues the study of Rutjes et al. (2016), now also including the effects of uniform electric fields, up to the classical breakdown field, which is about 3.0 MV m −1 at standard temperature and pressure.We first present our theoretical description of the RREA process, which is based on and incremented over previous published works. This analysis confirmed that the avalanche is mainly driven by electric fields and the ionisation and scattering processes determining the minimum energy of electrons that can run away, which was found to be above ≈ 10 keV for any fields up to the classical breakdown field.To investigate this point further, we then evaluated the probability to produce a RREA as a function of the initial electron energy and of the magnitude of the electric field. We found that the stepping methodology in the particle simulation has to be set up very carefully in Geant4. For ex-ample, a too-large step size can lead to an avalanche probability reduced by a factor of 10 or to a 40 % overestimation of the average electron energy. When properly set up, both Geant4 models show an overall good agreement (within ≈ 10 %) with REAM and GRRR. Furthermore, the probability that particles below 10 keV accelerate and participate in the high-energy radiation is found to be negligible for electric fields below the classical breakdown value. The added value of accurately tracking low-energy particles (< 10 keV) is minor and mainly visible for fields above 2 MV m −1 .In a second simulation set-up, we compared the physical characteristics of the avalanches produced by the four models: avalanche (time and length) scales, convergence time to a self-similar state and energy spectra of photons and electrons. The two Geant4 models and REAM showed good agreement on all parameters we tested. GRRR was also found to be consistent with the other codes, except for the electron energy spectra. That is probably because GRRR does not include straggling for the radiative and ionisation energy losses; hence, implementing these two processes is of primary importance to produce accurate RREA spectra. Including precise modelling of the interactions of particles below 10 keV (e.g. by taking into account molecular binding energy of secondary electrons for impact ionisation) also produced only small differences in the recorded spectra.

Discharge experiments were carried out at the Eindhoven University of Technology in 2013.The experimental setup was designed to search for electrons produced in meter-scale sparks using a 1 MV Marx generator. Negative voltage was applied to the high voltage (HV) electrode. Five thin (1 mm) plastic detectors (5 cm 2 each) were distributed in various configurations close to the spark gap. Earlier studies have shown (for HV negative) that X-rays are produced when a cloud of streamers is developed 30-60 cm from the negative electrode. This indicates that the electrons producing the X-rays are also accelerated at this location, that could be in the strong electric field from counterstreamers of opposite polarity. Comparing our measurements with modeling results, we find that ∼300 keV electrons produced about 30-60 cm from the negative electrode are the most likely source of our measurements. A statistical analysis of expected detection of photon bursts by these fiber detectors indicates that only 20%-45% of the detected bursts could be from soft (∼10 keV) photons, which further supports that the majority of detected bursts are produced by relativistic electrons.