A Statistical Analysis of the Erosion and Refilling of the Plasmasphere Using Neural Network-Based DEN3D Model
Abstract
The density and composition of Earths Plasmasphere strongly influences wave growth and propagation, as well as energetic particle scattering. This affects Earth-to-space communication and killer electron behavior. The plasmaspheric dynamics are both time-dependent and history-dependent. Previous empirical plasma density models have been based on statistical averages and are limited in their capability to make accurate predictions of the dynamic state of Earths plasmasphere. With recent advances in machine learning techniques, we are able to more accurately quantify complex global processes and nonlinear responses to driving conditions, especially during geomagnetic storms. This project presents a three-dimensional dynamic electron density model based on an artificial neural network. This model uses a feedforward neural network which was generated using electron densities from the satellite missions of CRRES, ISEE, IMAGE, and POLAR. The three-dimensional electron density model takes spacecraft location as well as time series of solar wind and geomagnetic indices (AE and F10.7) obtained from NASAs OMNI database as inputs. The model can predict out-of-sample data with a correlation coefficient of 0.94, meaning over 90% of the variations are captured. The three-dimensional model was applied to a number of magnetic storms, and it successfully reconstructed the expected plasmaspheric dynamics. We carried out a statistical analysis of the plasmaspheric erosion rates and refilling processes during these geomagnetic storms using the ML-based reconstruction, which show the plasmaspheric dynamics that cannot be obtained using spacecraft observations. The statistical results are consistent with previous studies. This model demonstrates the potential for machine learning techniques to be utilized in understanding the physics and insight discovery, as well as advance the state-of-the-art space weather prediction.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFMSM42A..02L