Understanding the Limitations of System Models for Geomagnetic Index Prediction
Abstract
The WINDMI physics based system model is an "instant" prediction tool that takes solar wind input data and predicts geomagnetic indices such as Dst. It does this by modeling the solar wind-magnetosphere-ionosphere system as a connected set of circuit elements. These components take into account the average amount of energy contained in a region and estimate an effective resistance, capacitance, or inductance. The Space Weather Modeling Framework (SWMF) couples several models to create a self-consistent system mostly derived from first principles. By using SWMF to reconstruct and analyze the energy contained in the various circuit elements of the WINDMI model allows us to make a direct comparison to understand under what conditions the circuit assumption performs well, and what features are missing. First, geomagnetic indices are compared for the two models along with OMNI data for a real series of events beginning on Feb 18, 2014. Then, several circuit elements are recreated within SWMF and compared over the simulation time. We discuss implications and limitations for the use of system models as prediction tools, along with suggestions for a possible new application of machine learning.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMSA0210021B
- Keywords:
-
- 2427 Ionosphere/atmosphere interactions;
- IONOSPHERE;
- 2431 Ionosphere/magnetosphere interactions;
- IONOSPHERE;
- 2704 Auroral phenomena;
- MAGNETOSPHERIC PHYSICS;
- 2736 Magnetosphere/ionosphere interactions;
- MAGNETOSPHERIC PHYSICS