A combined neural network‐ and physics‐based approach for modeling plasmasphere dynamics
- Abstract In recent years, feedforward neural networks (NNs) have been successfully applied to reconstruct global plasmasphere dynamics in the equatorial plane. These neural network‐based models capture the large‐scale dynamics of the plasmasphere, such as plume formation and erosion of the plasmasphere on the nightside. However, their performance depends strongly on the availability of training data. When the data coverage is limited or non‐existent, as occurs during geomagnetic storms, the performance of NNs significantly decreases, as networks inherently cannot learn from the limited number of examples. This limitation can be overcome by employing physics‐based modeling during strong geomagnetic storms. Physics‐based models show a stable performance during periods of disturbed geomagnetic activity if they are correctly initialized and configured. In this study, we illustrate how to combine the neural network‐ and physics‐based models of the plasmasphere in an optimal way by using data assimilation. The proposed approach utilizesAbstract In recent years, feedforward neural networks (NNs) have been successfully applied to reconstruct global plasmasphere dynamics in the equatorial plane. These neural network‐based models capture the large‐scale dynamics of the plasmasphere, such as plume formation and erosion of the plasmasphere on the nightside. However, their performance depends strongly on the availability of training data. When the data coverage is limited or non‐existent, as occurs during geomagnetic storms, the performance of NNs significantly decreases, as networks inherently cannot learn from the limited number of examples. This limitation can be overcome by employing physics‐based modeling during strong geomagnetic storms. Physics‐based models show a stable performance during periods of disturbed geomagnetic activity if they are correctly initialized and configured. In this study, we illustrate how to combine the neural network‐ and physics‐based models of the plasmasphere in an optimal way by using data assimilation. The proposed approach utilizes advantages of both neural network‐ and physics‐based modeling and produces global plasma density reconstructions for both quiet and disturbed geomagnetic activity, including extreme geomagnetic storms. We validate the models quantitatively by comparing their output to the in‐situ density measurements from RBSP‐A for an 18‐month out‐of‐sample period from June 30, 2016 to January 01, 2018 and computing performance metrics. To validate the global density reconstructions qualitatively, we compare them to the IMAGE EUV images of the He+ particle distribution in the Earth's plasmasphere for a number of events in the past, including the Halloween storm in 2003.…
Author details: | Irina ZhelavskayaORCiDGND, Nikita AseevORCiDGND, Yuri ShpritsORCiDGND |
---|---|
DOI: | https://doi.org/10.1029/2020JA028077 |
ISSN: | 2169-9380 |
ISSN: | 2169-9402 |
Title of parent work (English): | JGR / AGU, American Geographical Union. Space Physics |
Publisher: | Wiley |
Place of publishing: | Hoboken, NJ |
Publication type: | Article |
Language: | English |
Date of first publication: | 2021/03/16 |
Publication year: | 2021 |
Release date: | 2024/07/26 |
Tag: | Kalman filter; data assimilation; machine learning; neural networks; plasma density; plasmasphere |
Volume: | 126 |
Issue: | 3 |
Article number: | e2020JA028077 |
Number of pages: | 30 |
Funding institution: | NASA Heliophysics Guest Investigator Program under NASA |
Funding institution: | Geo.X, the Research Network for Geosciences in Berlin and Potsdam |
Funding institution: | pilot project "MAP", Helmholtz Pilot Projects Information & Data Science II - Initiative and Networking Fund of the Helmholtz Association |
Funding institution: | European Union's Horizon 2020 research and innovation program |
Funding institution: | Deutsche Forschungsgemeinschaft (DFG)German Research Foundation (DFG) |
Funding number: | NNX07AG48G |
Funding number: | SO_087_GeoX |
Funding number: | 870452 |
Funding number: | SFB 1294 |
Organizational units: | Mathematisch-Naturwissenschaftliche Fakultät / Institut für Physik und Astronomie |
DDC classification: | 5 Naturwissenschaften und Mathematik / 52 Astronomie / 520 Astronomie und zugeordnete Wissenschaften |
5 Naturwissenschaften und Mathematik / 55 Geowissenschaften, Geologie / 550 Geowissenschaften | |
Publishing method: | Open Access / Hybrid Open-Access |
License (German): | CC-BY-NC-ND - Namensnennung, nicht kommerziell, keine Bearbeitungen 4.0 International |