Saša Dujko

Sprites are large -scale transient luminous events occurring high above thunderstorm clouds, typically in the mesosphere at altitudes between 50 and 90 km. They are initiated by intense electric fields following lightning discharges and are characterized by rapidly propagating streamers. Streamers are thin filaments of weakly -ionised nonstationary plasma produced by an ionisation front that moves through non -ionised matter. Understanding the dynamics of sprite streamers is essential for modelling upper at mospheric electrical phenomena and their coupling to space weather. In this study, we present a classical fluid model for streamer inception and propagation, implemented using the AMReX software framework for adaptive mesh refinement and parallel computing (Simonovi ć et al. 2024 ). We specifically investigated the influence of background ionisation on the evolution of positive and negative streamers. Background ionisation in the upper atmosphere is strongly dependent on galactic cosmic ray (GCR) flux, which is modulated by solar a ctivity through the heliospheric magnetic field. Periods of low solar activity correspond to increased GCR penetration and higher ionisation levels, potentially altering streamer morphology, velocity, and branching behavior . Our simulations reveal distinct differences in streamer dynamics under varying ionisation conditions, highlighting the sensitivity of sprite formation to space weather parameters. These findings contribute to a deeper understanding of the electrodynamic coupling between the troposphere and ionosphere and underscore the importance of solar -terrestrial interactions in shaping transient luminous events.

Our axisymmetric fluid streamer model has been extended to enable simulations in domains with curvilinear boundaries by employing the AMReX inbuilt embedded boundary approach. We have performed benchmark simulations of positive streamers in air in the pin-to-plane electrode configuration. The results of these benchmark simulations have been verified by comparing them with the results of the open-source Afivo-streamer code. We have also performed simulations of negative streamers in SF ₆ and the mixtures of CO ₂ and C ₄ F ₇ N in the pin-to-plane electrode configuration. We have observed that the mixture of 80% CO ₂ and 20% C ₄ F ₇ N exhibits more potent dielectric properties than SF ₆ , while the mixture of of 85% CO ₂ and 15% C ₄ F ₇ N has dielectric properties that are comparable to those of SF ₆ . This suggests that mixtures of CO ₂ and C ₄ F ₇ N are promising alternatives to SF ₆ in high-voltage insulation technology.

Ionization is among the most important processes within the interstellar medium and planetary atmospheres since it defines both their physical and chemical nature. The elementary processes that lead to ionization events need to be understood and the cross sections which define the probability of different ionization pathways need to be defined in order to support the modelling of these regions. In this article we discuss the elementary processes leading to ionization events and their importance in planetary atmospheres. We focus on calculations of ionization and dissociative rate coefficients. Ionization rate coefficients derived from electron swarms as a function of the reduced electric field for planetary atmospheres in Solar system are presented as an example of ionization data required for this study. Dissociative rate coefficients for molecular hydrogen cation over a wide range of temperatures are discussed as an example of data needs. Our model is tailored for collisional reactions and may handle various atomic and molecular species, as shown for different compositions of planetary atmospheres. The model incorporates a significantly enriched set of collisional cross sections, including energy-dependent and temperature-dependent rates, as shown in calculations of dissociative rate coefficients. We also present several databases that may be used to source ionization data, with an emphasis on those developed in Belgrade.

In this work, we present a revised set of electron collision cross sections for 1,1,1,2 tetrafluoroethane (C₂H₂F₄). The cross-section set is the result of a standard swarm analysis of measured drift velocities (W) and effective ionization coefficients ((α–η)/N) under pulsed-Townsend conditions, where α is the ionization coefficient, and η is the attachment coefficient. Calculations and numerical analysis are performed for pure C₂H₂F₄ and its mixtures with argon over a wide range of reduced electric fields E/N (where E is the electric field and N is the gas number density). A special attention was given to the region of low and moderate electron energies, where electron attachment and vibration excitation are dominant inelastic channels of scattering. The cross-section set we derive provides a good balance of energy, momentum, and the number of charged particles. Using a Monte Carlo simulation technique and the derived cross-section set as input, we calculate other transport coefficients and properties, such as characteristic energy, diffusion coefficients, and rate coefficients for individual collisional processes. We expect this cross-section set and electron swarm transport coefficients to be useful in modelling non-equilibrium plasmas as well as in studying resistive plate chambers in high-energy physics.

We have implemented an axisymmetric fluid model of streamers in the AMReX open-source library. Our implementation is based on the first-order fluid model with a local field approximation. Photoionization is implemented by employing the Zhelenznyak’s model using the Helmholtz approach. We have verified our code in standard conditions by comparing our results for positive streamers in air with the existing benchmarks from the literature. To verify the performance of our code in strongly attaching gases, we compare the properties of negative streamers with those obtained from the Afivo-streamer open-source code. Calculations have been performed in mixtures of carbon dioxide and perfluoro-nitrile with a background number density of electrons and positive ions of 10¹³ m⁻³. We found an excellent agreement between the two sets of results, which indicates the numerical integrity of our code.