HPC-approach to Ionospheric Situational Awareness
The ionised part of the upper atmosphere – the ionosphere – is a region where the neutral atmosphere and near-Earth space meet. Its dynamics are controlled both by magnetospheric forcing from above and by neutral atmospheric processes from below. Modern society is completely dependent on power networks as well as satellite, radar, and radio signals, which are all susceptible to disturbances in the ionosphere. Hence, ionospheric situational awareness, achieved through accurate observations and numerical models, is increasingly important. The HISSA consortium combines high-fidelity ionospheric observations from the upcoming EISCAT_3D radar system and the world’s most accurate space environment simulation, Vlasiator. We will develop cutting-edge computational methods to fully exploit Europe's fastest supercomputer in 2022, LUMI, installed at CSC in Finland. These methods enable us to:
1. Construct an accurate model for magnetosphere-ionosphere interactions and develop novel load-balancing and parallel algorithms facilitating high-resolution magnetospheric simulations
2. Develop a HPC approach to data-analysis for the upcoming EISCAT_3D incoherent scatter radar, also utilising an independent remote station, KAIRA, allowing for an unprecedented 3D volumetric imaging of the ionosphere
3. Port and optimise the software to LUMI and other HPC platforms, and investigate and evaluate HPC programming models in LUMI for their developer productivity, performance and portability.
These goals are all highly important for any HPC code development project, where the gained knowledge is urgently needed for the benefit of the Finnish computational science community in user support, training and projects. This project will cultivate and expand Finnish pre-exascale HPC skills and take them to the level required for exascale computing.
This project has received funding from the Research Council of Finland.