Analogue Models, Filters and Neural Networks in Solar-Terrestrial Physics

The Solar-Terrestrial environment (Sun, solar wind, geomagnetic activity) can be studied as a non-linear dynamical system.
 
Knowledge can be coded into neural networks and with the availablity today of real-time Space Weather data, it is possible to produce real-time predictions of the Space Weather and of Space Weather effects (e.g.: satellite systems, communication and power systems). The Lund Space Weather Model is an interactive intelligent hybrid system.

For more information:

• A guide to forecasting methods Lund, Sweden
• Neural Network Information - Eindhoven University, NL.
• Local-Linear Analogue Models of the Nonlinear Magnetospheric Dynamics

[Back to Models & Data Page]