Winners announced for the Space Weather Competition. Visit for details.

Forecasting local GICs from solar wind data using a combination of geophysical and machine learning methods 

Rachel Bailey - Zentralanstalt für Meteorologie und Geodynamik, Conrad Observatory; Roman Leonhardt - Zentralanstalt für Meteorologie und Geodynamik, Conrad Observatory, Austria; Christian Möstl - Space Research Institute, Austrian Academy of Sciences, Austria 

Session: Ground-Level Geomagnetically Induced Currents


A multiyear collection of geomagnetically induced current (GIC) measurements at multiple locations in Austria provides the basis for this study, which aims to forecast the local maximum GIC from in situ solar wind measurements at the L1 point between the Sun and Earth. The prediction of GIC from solar wind will be carried out using a deep learning method, specifically a Long-Term-Short-Memory (LSTM) neural network. Measurements of GICs used as training datasets are available for six substations in the Austrian power grid, but the forecast will be extended to the whole grid of 129 substations by predicting the local geoelectric field and calculating the GICs per substation from there. For those substations with GIC measurements, a deep learning model will be trained on each station and compared to the results from the geoelectric field predictions to determine any differences in performance.

Approaching deadlines:

Registration opens:

16 July 2020

Abstract submission opens:

16 July 2020

European Space Weather Medals:

6 September 2020

Registration deadline:

25 September 2020

Registration deadline: [extended]

10 October 2020

Abstract submission deadline:

4 September 2020