IRO Journals
Journal of Trends in Computer Science and Smart Technology

Comparative Benchmarking Study of Deep Learning Models for Geomagnetic Storms Prediction

Open Access
Published: 29 August, 2026
Pages: 884-902
  • Prajwal Bhandari , Prajwal Bhandari
    School of Mathematical Science, Tribhuvan University, Kathmandu, Nepal
    School of Mathematical Science, Tribhuvan University, Kathmandu, Nepal
  • Subarna Shakya Subarna Shakya
    Institute of Engineering, Tribhuvan University, Kathmandu, Nepal
    Institute of Engineering, Tribhuvan University, Kathmandu, Nepal
view PDF
view PDF

How to Cite

Bhandari, Prajwal, and Subarna Shakya. 2026. “Comparative Benchmarking Study of Deep Learning Models for Geomagnetic Storms Prediction”. Journal of Trends in Computer Science and Smart Technology 8 (3): 884-902. https://doi.org/10.36548/jtcsst.2026.3.022.
ACM ACS APA ABNT Chicago Harvard IEEE MLA Turabian Vancouver AMA

Download Citation

Endnote/Zotero/Mendeley (RIS) BibTeX

Keywords

Geomagnetic Storm Prediction
Disturbance Storm Time (Dst) Index
Space Weather Forecasting
Time-Series Forecasting
Deep Learning
Long Short-Term Memory (LSTM)
Bidirectional LSTM (BiLSTM)

Abstract

Geomagnetic storms have been known to have a substantial impact on the performance of satellites, communication systems, navigation systems, and other space weather-based systems. This research is concerned with developing a data-driven forecast of geomagnetic activity using deep learning and traditional machine learning models. This study focuses on comparing deep learning and machine learning models for predicting the Disturbance Storm Time (Dst) index through solar wind and geomagnetic parameters. For the analysis, an hourly dataset for 2000-2024 has been utilized resulting 219,168 observations. From these, eight parameters including magnetic field components, solar wind speed, geomagnetic indices, and past Dst values have been selected. Additionally, Random Forest, XGBoost, LightGBM, and persistence were employed as traditional baseline models. The performance measures considered included RMSE, MAE, R², and Pearson correlation, alongside training time, inference time, number of parameters, and a feature ablation study. Among the evaluated deep-learning architectures, LSTM achieved the best average performance across the five independent seeds. The RMSE of LightGBM and persistence baselines was 3.77459 nT and 4.63338 nT, respectively. The ablation analysis shows that AE, Kp10, and solar-wind variables substantially influence forecasting performance, whereas historical Dst and a longer 48-hour window do not necessarily improve the reported error. Overall, the results have shown that it is possible to model nonlinear temporal dependencies in geomagnetic observations using recurrent architectures. However, the use of traditional baselines along with repeated-seeds and ablation tests makes the evaluation more thorough.

References

  1. Bhandari, Prajwal, Kiran Bagale, and Bidur Nepal. ”Seasonal Variation of Solar Wind–Geomagnetic Coupling: A Correlation Heatmap Approach.” SS Multidisciplinary Research Journal 2, no. 1 (2025): 117-126.
  2. Boteler, D. H. ”A New Versatile Method for Modelling Geomagnetic Induction in Pipelines.” Geophysical Journal International 193, no. 1 (2013): 98-109.
  3. Silwal, Ashok, Sujan Prasad Gautam, Prakash Poudel, Monika Karki, Narayan P. Chapagain, and Binod Adhikari. ”Variation of Total Electron Content Over Nepal During Geomagnetic Storms: GPS Observations.” Russian Journal of Earth Sciences 23, no. 3 (2023): 1-19.
  4. Abduallah, Yasser, Jason TL Wang, Prianka Bose, Genwei Zhang, Firas Gerges, and Haimin Wang. ”A Deep Learning Approach to Dst Index Prediction.” arXiv preprint arXiv:2205.02447 (2022).
  5. Echer, E., and W. D. Gonzalez. ”Geoeffectiveness of Interplanetary Shocks, Magnetic Clouds, Sector Boundary Crossings and Their Combined Occurrence.” Geophysical research letters 31, no. 9 (2004).
  6. M. Cristoforetti M Battiston Roberto and Gobbi A and Iuppa Roberto and Piersanti, ”Prominence of the Training Data Preparation in Geomagnetic Storm Prediction Using Deep Neural Networks,” Scientific Reports, vol. 12, 2022, 7631.
  7. Burton, Rande K., R. L. McPherron, and C. T. Russell. ”An Empirical Relationship between Interplanetary Conditions and Dst.” Journal of geophysical research 80, no. 31 (1975): 4204-4214.
  8. Akasofu, S-I. ”Energy Coupling Between the Solar Wind and the Magnetosphere.” Space Science Reviews 28, no. 2 (1981): 121-190.
  9. Newell, P. T., T. Sotirelis, K. Liou, C‐I. Meng, and F. J. Rich. ”A Nearly Universal Solar Wind‐Magnetosphere Coupling Function Inferred from 10 Magnetospheric State Variables.” Journal of Geophysical Research: Space Physics 112, no. A1 (2007).
  10. Kelley, Michael C. The Earth's Ionosphere: Plasma Physics and Electrodynamics. Vol. 96. Academic press, 2009.
  11. LI, Shaowen, Jun NIU, Bing MEI, Lizhu YAO, and Yanbin LI. ”Dst Index Prediction Method Based on LSTM Neural Network.” Chinese Journal of Space Science 45, no. 3 (2025): 641-652.
  12. Jang, Jun, Lang Xu, Chaoqian Xu, and Liang Zhang. ”Forecasting Single-Station Ionospheric TEC Over China Using a Combined DBO-LSTM Model During Geomagnetic Storms.” Advances in Space Research 75, no. 10 (2025): 7684-7695.
  13. Tang, Rongxin, Fantao Zeng, Zhou Chen, Jing-Song Wang, Chun-Ming Huang, and Zhiping Wu. ”The Comparison of Predicting Storm-Time Ionospheric TEC by Three Methods: ARIMA, LSTM, and Seq2Seq.” Atmosphere 11, no. 4 (2020): 316.
  14. S. J. Hochreiter, ”Long Short-Term Memory,” Neural Computation, vol. 9, 1997., 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735
  15. Subramanian, Dhanalakshmi, Poongothai Marimuthu, and Nithyalakshmi Ramadoss. 2026. “A Multifaceted Framework for Predicting Ambient Air Pollution in the National Capital Region”. Journal of Trends in Computer Science and Smart Technology 8 (3): 558-78. https://doi.org/10.36548/jtcsst.2026.3.007.
  16. Sarkar, Abhijeet, Ashmita Saha, Bhaskar Bhattacharjee, Bishnupada Sahoo, Rounak Bakshi, and Soumik Podder. 2023. “Nanozymes Induced Air Purification- A State of the Art Review”. Recent Research Reviews Journal 2 (1): 11-26. https://doi.org/10.36548/rrrj.2023.1.02