We would like to highlight an upcoming session at NAM 2026: Space Weather Forecasting in the Big Data, Large Models Era. NAM will take place at the University of Birmingham from the 20th-24th of July. We invite interested parties to submit abstracts via the link:
All submissions must be completed by 23:45 on the 2nd of April, 2026.
Session details:
Space Weather Forecasting in the Big Data, Large Models Era
Solar flares and Coronal Mass Ejections (CMEs) are explosions releasing vast amounts of energy and plasma, respectively in the Solar atmosphere. The most violent ones can not only disrupt the Solar atmosphere, but also Earth’s atmosphere, causing damage to satellites, disrupting communication and posing a threat to astronauts. Hence forecasting explosive solar events is of vital importance. This entails processing large volumes of data including images, videos and time-series to provide timely and precise forecasts.
Operational space weather forecasting demands principled geomagnetic models, with the efficient forecasting capabilities that artificial intelligence models can provide. A central challenge in this setting is that solar outbursts represent a complex, dynamically changing spatio-temporal feature-set (for eg. Evolving active regions) that might used to build models for space weather forecasting. Furthermore, there is a gap between a model tracing the sequence of physical processes leading to an outburst and one optimized to provide timely yet precise space weather forecasts. In this session, we will explore a range of problems from solar wind forecasting with data assimilation to large scale (foundation) models in forecasting flares and CMEs.
This necessitates experts ranging from solar physicists, space weather forecasting experts, and computer / data scientists to pool in their expertise and co-design such studies.
Session Organisers:
Nachiketa Chakraborty, Mathew Owens, Shaun Bloomfield, Malcolm Dunlop
Contact email – n.chakraborty@reading.ac.uk