The knowledge of the background plasma density and the Kp index is of crucial importance for the modelling of radiation belt and ring current electron dynamics. The Kp index is used in many input parameterizations such as wave models, electric, and magnetic fields. Therefore, predictions of the Kp index are a necessary step in forecasting electron fluxes. The plasmapause location separates different regions of hiss and chorus wave-induced scattering of electrons and is a critical parameter for the modelling of electron dynamics.
In this work package, we will develop tools that allow forecasting the Kp index and plasma density several days ahead. We will use machine learning techniques to construct empirical models of the Kp index driven by solar wind and IMF parameters at L1. For modelling the plasmasphere dynamics, we will implement and test different neural network and physics-based approaches. We will test different electric field models and explore the possibility of real-time coupling of the physics-based plasmasphere models with the electric field from the BATS-R-US component of SWMF provided by WP5. Finally, we will estimate uncertainty levels by using ensemble forecast of solar wind parameters at L1 provided by WP2.
Lead: GFZ (Dr. Stefano Bianco, Dr Ruggero Vasile)