Prediction of Ionospheric TEC based on EEMD-Prophet-BiLSTM Model
Anjan Kumar Dasanjandasmssv@gmail.comDepartment of Computer ApplicationMahapurusha Srimanta Sankaradeva ViswavidyalayaNagaon, Assam, India0009-0009-9425-1719View full profile → , *Santanu KalitaCorresponding authorsantanu.kalita@rediffmail.comDepartment of Computer ApplicationMahapurusha Srimanta Sankaradeva ViswavidyalayaNagaon, Assam, India0000-0003-2715-9225View full profile → , Surajit Nathsurajitnath2@gmail.comDepartment of Computer ApplicationMahapurusha Srimanta Sankaradeva ViswavidyalayaNagaon, Assam, India0000-0002-7355-6013View full profile → , Bornali Chetiabornalichetia.physics@gmail.comDepartment of PhysicsRoyal Global UniversityGuwahati, Assam, India0009-0002-7565-8547View full profile →
* Corresponding author · click or hover a name for details
- Received:
- 01 Apr 2025
- Published Online:
- 31 Aug 2026
- Article type:
- Research Article
- Language:
- EN
- Article no.:
- JSMS-1553
- Pages:
- 1–16
Abstract
Predicting ionospheric GPS Total Electron Content (TEC) data is quite challenging due to the need for processing large datasets and handling variations across time and space. Machine learning techniques have increasingly been adopted in recent years for predicting complex time series characterized by nonlinear and non-stationary behavior. This work emphases on using machine learning techniques to predict GPS-derived ionospheric TEC. A unique hybrid version combining EEMD (Ensemble Empirical Mode Decomposition), facebook’s Prophet, and BiLSTM is proposed to forecast TEC values for distinctive GNSS stations in India. The model demonstrated reliable performance during periods of both minimal and heightened geomagnetic activity. Its effectiveness is evaluated using various performance metrics and compared with existing models.
Keywords
Subject Classifications
References
[1] A. J. Mannucci, B. D. Wilson, D. N. Yuan, C. M. Ho, U. J. Lindqwister, and T. F. Runge, “A global mapping technique for GPS-derived ionospheric total electron content measurements,” Radio Science, vol. 33, no. 3, pp. 565–582 (1998), doi: 10.1029/97RS02707.
[2] G. Sivavaraprasad, V. S. Deepika, D. Sreenivasa Rao, M. Ravi Kumar, and M. Sridhar, “Performance evaluation of neural network TEC forecasting models over equatorial low-latitude Indian GNSS station,” Geodesy and Geodynamics, vol. 11, no. 2, pp. 127–136 (2020), doi: 10.1016/j.geog.2019.11.002.
[3] Y. Tulunay, E. T. Senalp, E. Tulunay, and S. M. Radicella, “Forecasting total electron content maps by using neural network,” Radio Science, vol. 39, no. 1, Art. no. RS1010 (2004), doi: 10.1029/2002RS002854.
[4] I. L. Mallika, D. V. Ratnam, Y. Otsuka, G. Sivavaraprasad, and S. Raman, “Implementation of hybrid ionospheric TEC forecasting algorithm using PCA-NN method,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 12, no. 1, pp. 371–381 (2019), doi: 10.1109/JSTARS.2018.2881791.
[5] A. Boulch, N. Cherrier, and T. Castaings, “Ionospheric activity prediction using convolutional recurrent neural networks,” arXiv preprint arXiv:1810.13273 (2018). [Online]. Available: https://arxiv.org/abs/1810.13273.
[6] W. Sun, Y. Wang, B. Xu, Z. Wang, and Y. Yuan, “Forecasting of ionospheric vertical total electron content (TEC) using LSTM networks,”in Proc. Int. Conf. Machine Learning and Cybernetics (ICMLC), pp. 340–344 (2017).
[7] Y. Tan, Q. Hu, Z. Wang, and Q. Zhong, “Geomagnetic index Kp forecasting with LSTM,” Space Weather, vol. 16, no. 4, pp. 406–416 (2018), doi: 10.1029/2017SW001764.
[8] Z. Pala and R. Atici, “Forecasting sunspot time series using deep learning methods,” Solar Physics, vol. 294, no. 5, Art. no. 50 (2019), doi: 10.1007/s11207-019-1449-8.
[9] A. Boulch, N. Cherrier, and T. Castaings, “Ionospheric activity prediction using convolutional recurrent neural networks,” arXiv preprint arXiv:1810.13273 (2018). [Online]. Available: https://arxiv.org/abs/1810.13273.
[10] H. K. Yadav, Y. Pal, and M. M. Tripathi, “Short-term PV power forecasting using empirical mode decomposition in integration with back-propagation neural network,” Journal of Information and Optimization Sciences, vol. 41, no. 1, pp. 25–37 (2020), doi: 10.1080/02522667.2020.1714181.
[11] V. K. Bairagi and V. K. Harpale, “Improved epileptic seizure detection using singular spectrum empirical mode decomposition and machine learning approach,” Journal of Statistics and Management Systems, vol. 25, no. 1, pp. 103–123 (2022), doi: 10.1080/09720510.2020.1862958.
[12] R. Song, X. Zhang, C. Zhou, J. Liu, and J. He, “Predicting TEC in China based on the neural networks optimized by genetic algorithm,” Advances in Space Research, vol. 62, no. 4, pp. 745–759 (2018), doi: 10.1016/j.asr.2018.03.043.
[13] A. Tebabal, S. M. Radicella, M. Nigussie, B. Damtie, B. Nava, and E. Yizengaw, “Local TEC modeling and forecasting using neural networks,” Journal of Atmospheric and Solar-Terrestrial Physics, vol. 172, pp. 143–151 (2018), doi: 10.1016/j.jastp.2018.03.004.
[14] A. Tebabal, S. M. Radicella, B. Damtie, Y. Migoya-Orúe, M. Nigussie, and B. Nava, “Feed forward neural network based ionospheric model for the East African region,” Journal of Atmospheric and Solar-Terrestrial Physics, vol. 191, Art. no. 105051 (2019), doi: 10.1016/j.jastp.2019.05.016.
[15] R. Mukesh, V. Karthikeyan, P. Soma, and P. Sindhu, “Forecasting of ionospheric TEC for different latitudes, seasons and solar activity conditions based on OKSM,” Astrophysics and Space Science, vol. 365, no. 1, Art. no. 9 (2020), doi: 10.1007/s10509-020-3730-x.
[16] A. Ruwali, A. J. S. Kumar, K. B. Prakash, G. Sivavaraprasad, and D. V. Ratnam, “Implementation of hybrid deep learning model (LSTM-CNN) for ionospheric TEC forecasting using GPS data,” IEEE Geoscience and Remote Sensing Letters, vol. 18, no. 9, pp. 1605–1609 (2021), doi: 10.1109/LGRS.2020.3004378.
[17] Z. Wu and N. E. Huang, “Ensemble empirical mode decomposition: A noise-assisted data analysis method,” Advances in Adaptive Data Analysis, vol. 1, no. 1, pp. 1–41 (2009), doi: 10.1142/S1793536909000047.
[18] S. J. Taylor and B. Letham, “Forecasting at scale,” The American Statistician, vol. 72, no. 1, pp. 37–45 (2018), doi: 10.1080/00031305.2017.1380080.
