Space weather refers to the study of the conditions in the near-earth space environment, including the magnetosphere, ionosphere, and thermosphere, and how the sun affects those regions. Space weather events are relevant for many technological systems, such as satellites and electric-power networks, which can be severely damaged in case of enhanced fluxes of energetic particles. The ability to forecast such harmful events represents the main scientific challenge of space weather. In order to be able to produce accurate forecasting, we need to understand the complex solar-terrestrial relationship and the underlying physical processes.
The Multiscale Dynamics group develops forecasting algorithms for space weather based on machine learning. The abundance of freely available satellite and ground-based data makes the use of modern machine learning techniques an ideal way to tackle the problem of space weather forecasting. Moreover, we aim at enhancing the current state-of-the-art simulations for space weather, following the ‘grey-box’ paradigm of using data and simulations synergically.
Contact person: Enrico Camporeale
Research group: Multiscale Dynamics (MD)
Research partners: Johns Hopkins University Applied Physics Laboratory (USA), GFZ German Research Center for Geosciences (Germany), University of California (USA), KU Leuven (Belgium)