A modeling study of ≥2 MeV electron fluxes in GEO at different prediction time scales based on LSTM and transformer networks
Authors:
- Xiaojing SunXSXiaojing Sun
- Dedong WangDWDedong Wang
- Alexander DrozdovADAlexander Drozdov
- Ruilin LinRLRuilin Lin
- Artem SmirnovASArtem Smirnov
- Yuri ShpritsYSYuri Shprits
- Siqing LiuSLSiqing Liu
- Bingxian LuoBLBingxian Luo
- Xi LuoXLXi Luo
Open Access
- 9 September 2024
- journal article
- review article
- Published by EDP Sciences in Journal of Space Weather and Space Climate
- Vol. 14, 25
Abstract
In this study, we develop models to predict the log10 of ≥2 MeV electron fluxes with 5-minute resolution at the geostationary orbit using the Long Short-Term Memory (LSTM) and transformer neural networks for the next 1-hour, 3-hour, 6-hour, 12-hour, and 1-day predictions. The data of the GOES-10 satellite from 2002 to 2003 are the training set, the data in 2004 are the validation set, and the data in 2005 are the test set. For different prediction time scales, different input combinations with 4 days as best offset time are tested and it is found that the transformer models perform better than the LSTM models, especially for higher flux values. The best combinations for the transformer models for next 1-hour, 3-hour, 6-hour, 12-hour, 1-day predictions are (log10 Flux, MLT), (log10 Flux, Bt, AE, SYM-H), (log10 Flux, N), (log10 Flux, N, Dst, Lm), and (log10 Flux, Pd, AE) with PE values of 0.940, 0.886, 0.828, 0.747, and 0.660 in 2005, respectively. When the low flux outliers of the ≥2 MeV electron fluxes are excluded, the prediction efficiency (PE) values for the 1-hour and 3-hour predictions increase to 0.958 and 0.900. By evaluating the prediction of ≥2 MeV electron daily and hourly fluences, the PE values of our transformer models are 0.857 and 0.961, respectively, higher than those of previous models. In addition, our models can be used to fill the data gaps of ≥2 MeV electron fluxes.Keywords
Funding Information
- National Natural Science Foundation of China (Project U2106201)
This publication has 111 references indexed in Scilit:
- Acceleration and loss of relativistic electrons during small geomagnetic stormsGeophysical Research Letters, 2015
- Combined convective and diffusive simulations: VERB‐4D comparison with 17 March 2013 Van Allen Probes observationsGeophysical Research Letters, 2015
- Response of geostationary communications satellite solid-state power amplifiers to high-energy electron fluenceSpace Weather, 2015
- Nowcast model for low‐energy electrons in the inner magnetosphereSpace Weather, 2014
- Electron losses from the radiation belts caused by EMIC wavesJournal of Geophysical Research: Space Physics, 2014
- Simulation of high‐energy radiation belt electron fluxes using NARMAX‐VERB coupled codesJournal of Geophysical Research: Space Physics, 2014
- Simulating the Earth's radiation belts: Internal acceleration and continuous losses to the magnetopauseJournal of Geophysical Research: Space Physics, 2014
- Reconstruction of gaps in the past history of solar wind parametersGeophysical Research Letters, 2014
- Three‐dimensional electron radiation belt simulations using the BAS Radiation Belt Model with new diffusion models for chorus, plasmaspheric hiss, and lightning‐generated whistlersJournal of Geophysical Research: Space Physics, 2013
- Low‐energy electrons (5–50 keV) in the inner magnetosphereJournal of Geophysical Research: Space Physics, 2013