We’re excited to announce the launch of SolarStorm.Space! Keen AI’s interactive, end-to-end application for exploring how geomagnetic storms impact the power network. Currently focused on the UK, we are working to scale the platform to global grids!
Geomagnetically Induced Currents (GICs) pose a real and increasing risk to national power grids. During periods of strong solar activity, rapid changes in the Earth’s magnetic field can drive large, quasi-DC currents into transformers and substations. These currents can cause overheating, saturation, voltage instability and, in extreme cases, lead to equipment damage or widespread power outages. Understanding when and where these currents may occur is essential for improving the resilience of the UK electricity network.
Our tool provides clear, interactive visualisations that show the full sequence of processes that lead to GICs in the UK. The application maps:
- Magnetic field disturbances during geomagnetic storms
- Induced geoelectric fields, shaped by the UK’s underlying geology
- Resulting GICs flowing through substations in the high-voltage transmission network
Users can explore three major storm scenarios, from the well-known St Patrick’s Day storm to an extreme 1-in-100-year “Gannon” scenario. You can also switch between different field components (Bx or Ex and By or Ey) to see how these drivers influence the behaviour of the grid.
The tool highlights substations experiencing elevated induced current levels, giving a hands-on understanding of how severe space weather translates into electrical risk on the ground.
SolarStorm.Space’s pipeline integrates machine learning, geophysical modelling and UK-specific network data to recreate and forecast the impact of solar storms.
1. Predicting Magnetic Field Variations
A Random Forest model uses solar wind measurements from ACE and DSCOVR to predict magnetic field changes at UK magnetometer stations, followed by spatial interpolation across the whole country.
2. Calculating the Induced Electric Field
Using 1D Earth resistivity models (EURHOM), the predicted magnetic field is converted into the geoelectric field through magnetotelluric calculations involving FFTs and surface impedance modelling.
3. Computing Geomagnetically Induced Currents
The electric field is then applied to a detailed model of the UK power grid, using standard GIC calculation methods to estimate current flow at each substation.
We are also developing a related forecasting tool that uses the planetary K index, machine learning models and substation-specific quadratic fits to estimate the likelihood of significant GICs in real time. Please check out Ruby’s incredibly detailed blog post series if you’d like to understand more about the science and engineering behind this pipeline!
This cutting-edge project was created by our brilliant multi-disciplinary team at Keen AI:
- Ruby Pearce-Casey, PhD Researcher at the Open University leading the scientific and modelling pipeline
- Dr. Cristian Silva-Perez, Geospatial Engineer responsible for geophysical modelling and mapping
- Thomas Emment, Software Developer who built the interactive web application
Ruby also showcased this project at the DISCnet Showcase 2025 on 8 December at The Royal Society in London.
We would like to thank those whose expertise, research and data made this project possible:
- Andrew Renton (Transpower New Zealand) for insightful discussions
- Greg Lucas (University of Colorado Boulder) for support with electric field modelling
- Rachel Bailey (GeoSphere Austria) for foundational research and the GEOMAGICA examples used in our GIC calculations
- NOAA: National Oceanic and Atmospheric Administration for the solar wind and K index data used in our predictions
SolarStorm.Space is an exciting step in turning space weather science into something you can see and use for the UK power network. We can’t wait to keep building the tool and explore even more ways to support grid resilience, risk modelling and infrastructure monitoring.
If you’re curious, want to collaborate, or just fancy a chat about our work here, we’d love to hear from you!
