Skip to main content

Geomagnetic Index Forecast

Ensemble Kp forecast based on PAGER ensemble solar wind predictions

This forecast product is prepared at GFZ in the context of PAGER HORIZON 2020 project. A machine learning model is trained to predict Kp values using historical solar wind data from OMNIWeb from 01-01-1998 to 31-05-2019. The trained model is run every hour and it provides an ensemble forecast of Kp up to 72 hours ahead. This is achieved by feeding the model with the the real-time solar-wind measurements made available by NASA's Advanced Composition Explorer (ACE) spacecraft and with the ensemble forecast of solar wind at L1 produced by the SWIFT code. The figure below shows the latest forecast produced. The green bars indicate the ensemble median values for the Kp index, while the red lines indicate the maximum value among the ensemble members for each forecast horizon. Below the figure also more statistics of the latest forecast produced is available for download.


  • Kp Forecast Statistics in a common data exchange format  (CSV)
  • Kp Forecast Statistics in JavaScript Object Notation format  (JSON)
  • Kp Forecast data README File
  • Kp Forecast Figure (PNG)

The data/data products are provided “as-is” without warranty of any kind either expressed or implied, including but not limited to the implied warranties of merchantability, correctness and fitness for a particular purpose. The entire risk as to the quality and performance of the Data/data products is with the Licensee.
In no event will GFZ be liable for any damages direct, indirect, incidental, or consequential, including damages for any lost profits, lost savings, or other incidental or consequential damages arising out of the use or inability to use the data/data products.