Skip to main content Skip to navigation

Developing the capability to forecast extreme Space Weather events

Supervisors: Ravindra Desai, Jeremie Houssineau and David Jackson


Extreme Space Weather is driven by large-scale eruptions from the Sun called coronal mass ejections. Upon arrival at the Earth, these produce amazing auroral displays but also endanger satellites and astronauts, and disrupt communications and power grids. Forecasting these events is of primary importance to the UK MET Office, one of three centres world-wide providing round-the-clock space weather predictions. This project, in collaboration with the UK MET Office, will develop state-of-the-art plasma simulations to forecast extreme Space Weather and develop advanced statistical techniques to quantify the uncertainties in their prediction.


The successful applicant will have the opportunity to work with a first-of-its-kind particle model of the Earth’s radiation belts which is uniquely integrated within a dynamic model of the global magnetosphere and capable of capturing sub-second relativistic particle dynamics across multi-day Space Weather events.

The project will examine the most extreme events observed across the Space Age, as well as upcoming events within Solar Cycle 25, and quantify the uncertainties in their hindcasts and forecasts through Bayesian inference techniques such as the Ensemble Kalman filtering.

The novelty in combining advanced statistical and simulation techniques will contribute to a level of robustness capable of accurately detecting extreme physical regimes, hence triggering these intensive simulations only when warranted. The potential exists to implement results for real-time predictions at the UK MET Office.

This project requires a multi-disciplinary approach to quantify uncertainties in state-of-the-art plasma simulations. This has relevance to space weather forecasting and the involvement of the UK MET Office signifies the potential for its real-world application. The taught elements of the HetSys programme (PX911 multiscale simulations, PX917 computational plasma physics, and PX913 research software engineering) will provide the candidate with the perfect foundation for tackling this ambitious research project.

This project will span discipline boundaries through combining space plasma theory with advanced statistical techniques. The synergy of these two areas is envisaged to be greater than the sum of their parts and to lay the foundation for a world-first space weather forecasting prototype system for extreme events.

These plasma simulations are computationally intensive which renders ensemble modelling expensive due to the large uncertainties on the initial conditions. Uncertainty therefore needs to be approached from, and integrated at, a fundamental level to enable these state-of-the-art simulations to be useful in the real-world for forecasting.

The advanced numerical models will need to be developed and implemented on Warwick and the MET Office High-Performance-Computing systems and, due to their highly parallel nature, this requires a high level of research software engineering skills.