Speaker
Description
Abstract
Machine learning (ML) methods are increasingly utilized in space weather research; however, their performance is often limited by sparse observations during extreme events and a lack of physical constraints. Physics-based models, on the other hand, rely on empirical parameterizations and simplifying assumptions that can limit their
performance. Bridging these two approaches offers a pathway toward more reliable and generalizable models of the near-Earth plasma environment.
In this study, we investigate plasmaspheric electron density modeling using a physics-based model (VERB-CS) [1], a purely data-driven neural network model (PINE) [4], and three hybrid physics–machine learning configurations. The hybrid frameworks are driven by geomagnetic indices and their time history, together with spacecraft location and in situ electron density measurements. In each configuration, density outputs from the physics-based model are incorporated into the neural network at different stages of training, enabling varying degrees of physical guidance.
Model performance is evaluated against independent in situ observations [2] and global plasmaspheric imaging [3]. The hybrid approaches consistently outperform both the purely data-driven and purely physics-based models, achieving lower root-mean-square errors, reduced bias, and improved generalization across quiet and disturbed geomagnetic conditions. The results demonstrate that embedding physical information within machine learning frameworks enhances predictive capability, particularly during periods of elevated geomagnetic activity.
References
[1] Aseev, N., Shprits, Y., 2019. Reanalysis of ring current electron phase space densities using Van Allen Probe observations, convection model, and log-normal Kalman filter. Space weather 17, 619–638.
[2] Kletzing, C., Kurth, W., Acuna, M., MacDowall, R., Torbert, R., Averkamp, T., Bodet, D., Bounds, S., Chutter, M., Connerney, J., et al., 2013. The electric and magnetic field instrument suite and integrated science (EMFISIS) on RBSP. Space Science Reviews 179, 127–181.
[3] Sandel, B., Goldstein, J., Gallagher, D., Spasojevic, M., 2003. Extreme ultraviolet imager observations of the structure and dynamics of the plasmasphere. Magnetospheric imaging—The image prime mission, 25–46.
[4] Zhelavskaya, I.S., Shprits, Y.Y., Spasojević, M., 2017. Empirical modeling of the plasmasphere dynamics using neural networks. Journal of Geophysical Research: Space Physics 122, 11–227.
| Numerical model | PINE |
|---|
