Modeling China's Sichuan-Yunnan's Ionosphere Based on Multichannel WOA-CNN-LSTM Algorithm

• Published in: IEEE transactions on geoscience and remote sensing 2024, Vol.62, p.1-18
• Main Authors: Li, Wang, Zhu, Haoze, Shi, Shuangshuang, Zhao, Dongsheng, Shen, Yi, He, Changyong
• Format: Article
• Language: English
• Subjects: Accuracy; Aerospace environments; Algorithms; Artificial neural networks; Convolutional neural networks; Correlation coefficient; Correlation coefficients; Early warning systems; Electrons; Emergency communications systems; Emergency warning programs; Equatorial regions; Geomagnetism; Global navigation satellite system; Ionization; Ionosphere; Ionospheric model; Latitude; Long short term memory; low-latitude region; Magnetic storms; Marine mammals; Mathematical models; Navigation; Navigational satellites; neural network; Neural networks; Predictive models; Recurrent neural networks; Root-mean-square errors; Seasonal variation; Seasonal variations; Storms; Temporal variations; Total Electron Content; total electron content (TEC); Warning systems; whale optimization algorithm (WOA)-convolutional neural network (CNN)-long short-term memory (LSTM)
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2024.3403684

The total electron content (TEC) of the ionosphere at low latitudes is significantly influenced by solar-geomagnetic activity and seasonal variations. Traditional ionospheric models often struggle to accurately forecast TEC in low latitudes, which limits the improvement of positioning accuracy for single-frequency global navigation satellite system (GNSS) receivers. This study focuses on the Sichuan and Yunnan areas of China located on the northern crest of the equatorial ionization anomaly (EIA), utilizing data from 48 stations of the Chinese GNSS network. It employs a convolutional long short-term memory (LSTM) network with multichannel characteristics, combined with the whale optimization algorithm (WOA), to construct a WOA-convolutional neural network (CNN)-LSTM model for predicting TEC variations. The results demonstrate that the WOA-CNN-LSTM model outperforms CNN-gated recurrent unit (CNN-GRU), bidirectional LSTM (BiLSTM), and recurrent neural network (RNN) models in spatial morphology. In 2015 (a year with geomagnetic storms), the root mean square error (RMSE) values for all four seasons were at or below 1.96 TECu, with mean absolute error (MAE) values at or below 1.42 TECu, and Pearson correlation coefficients at or above 0.98. In 2019 (a calm year), the RMSE values are all below 0.74 TECu, MAE values are all below 0.54 TECu, and Pearson correlation coefficients remain at or above 0.95. In terms of temporal variation, the RMSE prediction results for the four observation stations are all at or below 2.75 TECu for each of the four seasons in 2015, improving to 1.56 TECu in 2019. Therefore, this model significantly enhances ionospheric prediction accuracy in low-latitude regions, benefiting navigation positioning, space environment forecasting, and disaster early warning systems.