A High-Fidelity Approach for Physics-Based Modeling of the Ionosphere-Thermosphere System

The current interest in Low-Earth Orbiting (LEO) satellite mega-constellations to provide navigation data, communications solutions or Earth observations devises the need for accurate trajectory prediction to minimize the risk of collisions and deorbit in an increasingly populated region. Nevertheless, the high variability caused by very low densities of neutral particles, unpredictable solar activity, the interaction of the megnetosphere with solar winds or the propagation of meteorological perturbations defines a challenging environment for the computation of atmospheric drag.

To this end, this work presents a high-fidelity numerical approach for the simulation of the ionosphere-thermosphere system relying on the high-order discontinuous Galerkin (DG) method, combined with implicit-in time Runge-Kutta schemes. The numerical scheme performs a matrix-free strategy, well-suited for high-performance computing on graphic processors (GPUs), and allows to deal with cube-sphere meshes to avoid pole singularities.

On the other hand, the physics-based model defines a non-hydrostatic atmosphere, where different external sources, such as the heating produced by solar extreme ultraviolet (EUV) radiation, drive the dynamics of the system. The model accounts for the multiscale variations along the radial direction by introducing a proper nonlinear scaling of variables, allowing to deal with the low densities at the upper atmosphere, and performs a non-segregated approach between the radial and angular components.Different numerical examples are presented for validation and to test the performance of the proposed methodology.