Call For Paper September 2026
Research Article | Open Access | Download PDF
Volume 13 | Issue 9 | Year 2026 | Article Id. IJEEE-V13I9P110 | DOI : https://doi.org/10.14445/23488379/IJEEE-V13I9P110

Multimodal Machine Learning for Cloud-to-Ground Lightning Classification and Polarity Localization in Tropical Thunderstorm

Xian Gie Cheah, Mawarni Mohamed Yunus, Mohd Riduan Ahmad, Nor Hadizah Mohd Khalid

Received Revised Accepted Published
30 Jul 2026 08 Sep 2026 16 Sep 2026 26 Sep 2026

Citation :

Xian Gie Cheah, Mawarni Mohamed Yunus, Mohd Riduan Ahmad, Nor Hadizah Mohd Khalid, "Multimodal Machine Learning for Cloud-to-Ground Lightning Classification and Polarity Localization in Tropical Thunderstorm," International Journal of Electrical and Electronics Engineering, vol. 13, no. 9, pp. 123-135, 2026. Crossref, https://doi.org/10.14445/23488379/IJEEE-V13I9P110

Abstract

Accurate classification and localization of Cloud-to-Ground (CG) lightning are essential for the development of reliable early warning systems in tropical regions and the improvement of thunderstorm monitoring. This study proposes a multimodal machine learning framework for CG lightning classification and polarity-based localization with synchronized atmospheric and electromagnetic measurements collected in Melaka, Malaysia, in 2024. The framework includes electric field waveform from PicoScope, CAPPI radar reflectivity images and Himawari infrared cloud images, which contain the electrical discharge characteristics and storm structure information. A dataset comprising 2,799 synchronized lightning events was analyzed using two independent learning approaches; a Convolutional Neural Network (CNN) and a feature-engineered XGBoost model. In addition to the lightning classification, a polarity-based localization approach is proposed. This approach is based on the Magnetic Direction Finder (MDF) where the prediction and interpretation of the North-South (NS) and East-West (EW) magnetic field polarities are used to find the location of the discharge quadrant. The CG classification accuracy was 99.24% for CNN and 99.91% for XGBoost with three-modality fusion. In polarity prediction, CNN achieved 88.25% accuracy, while XGBoost improved significantly from 26.29% with a single-modality input to 87.34% with multimodal integration. These findings underline the importance of a multimodal approach to lay a solid basis for the fusion of heterogeneous atmospheric and electromagnetic data and to improve the accuracy and robustness of lightning intelligence systems.

Keywords

Cloud-to-Ground lightning, Convolutional Neural Network, Magnetic direction finder, Multimodal machine learning, Polarity-based localization, XGBoost model.

References

  1. Anisa Erma Putri et al., “A Comparative Analysis of External Lightning Strike Protection Area Determined by using Protection Angle and Rolling Sphere Methods,” International Journal on Advanced Science, Engineering and Information Technology, vol. 13, no. 1, pp. 203-210, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]
  2. Jussi Leinonen et al., “Thunderstorm Nowcasting with Deep Learning: A Multi-Hazard Data Fusion Model,” Geophysical Research Letters, vol. 50, no. 8, pp. 1-9, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]
  3. Yiming Han, Bin He, and Hongchun Shu, “Signal Positioning of Lightning Detection and Warning System Combining Direction of Arrival Algorithm and Capon Algorithm,” Processes, vol. 13, no. 2, pp. 1-17, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  4. Shunxing Zhu et al., “A Lightning Classification Method based on Convolutional Encoding Features,” Remote Sensing, vol. 16, no. 6, pp. 1-13, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
  5. Fei Wang et al., “Characteristics of Negative Cloud-To-Ground Lightning Flashes Associated with their Peak Currents of First Return Strokes,” Earth and Space Science, vol. 13, no. 3, pp. 1-17, 2026.
    [CrossRef] [Google Scholar] [Publisher Link]
  6. Xiaojie Liu et al., “Structural Characteristics of Thunderstorms Associated with Negative Triggered Lightning Flashes,” Advances in Atmospheric Sciences, vol. 42, no. 10, pp. 2053-2066, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  7. Marcos Antonio Alves et al., “Lightning Nowcasting using Dual-Polarization Weather Radar and Machine Learning Approaches: Evaluation of Feature Engineering Strategies and Operational Integration,” Climate, vol. 13, no. 8, pp. 1-21, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  8. Mingyue Lu et al., “Monitoring Lightning Location based on Deep Learning Combined with Multisource Spatial Data,” Remote Sensing, vol. 14, no. 9, pp. 1-18, 2022.
    [CrossRef] [Google Scholar] [Publisher Link]
  9. Kamyar Mehranzamir et al., “Implementation of Ground-Based Lightning Locating System using Particle Swarm Optimization Algorithm for Lightning Mapping and Monitoring,” Remote Sensing, vol. 15, vo. 9, pp. 1-30, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]
  10. Yueyang Wang et al., “Three-Dimensional Lightning Channel Structure Reconstruction: An Efficient Matching TOA Algorithm for VHF Pulses,” Journal of Atmospheric and Solar-Terrestrial Physics, vol. 278, pp. 1-14, 2026.
    [CrossRef] [Google Scholar] [Publisher Link]
  11. Ammar Alammari et al., “Lightning Mapping: Techniques, Challenges, and Opportunities,” IEEE Access, vol. 8, pp. 190064-190082, 2020.
    [CrossRef] [Google Scholar] [Publisher Link]
  12. Yiming Han, Bin He, and Hongchun Shu, “Location Distribution of Lightning Localization Monitoring Stations Integrating 3D Monitoring and PSO Algorithm,” Processes, vol. 13, no. 1, pp. 1-16, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  13. Jingyu Lu et al., “An Efficient Lightning Classifier using a Self-Supervised Learning Neural Network,” Geophysical Research Letters, vol. 52, no. 12, pp. 1-10, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  14. Shiye Huang et al., “Distinguishing Different Lightning Events based on Wavelet Packet Transform of Magnetic Field Signals,” Journal of Atmospheric and Solar-Terrestrial Physics, vol. 211, pp. 1-9, 2020.
    [CrossRef] [Google Scholar] [Publisher Link]
  15. Ting Wu, Daohong Wang, and Nobuyuki Takagi, “High-Accuracy Classification of Radiation Waveforms of Lightning Return Strokes,” Journal of Geophysical Research: Atmospheres, vol. 128, no. 14, pp. 1-15, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]
  16. Wenxing Sun et al., “A Novel Classification Framework for VLF/LF Lightning-Radiation Electric-Field Waveforms,” Atmosphere, vol. 16, no. 10, pp. 1-16, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  17. Mingyue Lu et al., “Lightning Identification based on Multiple Weather Radar Product Data,” International Journal of Digital Earth, vol. 18, no. 1, pp. 1-22, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  18. B. Yang et al., “A Thunderstorm Identification Method Combining the Area of Graupel Distribution Region and Weather Radar Reflectivity,” Earth and Space Science, vol. 7, no. 1, pp. 1-9, 2020.
    [CrossRef] [Google Scholar] [Publisher Link]
  19. Mingyue Lu et al., “Lightning Strike Location Identification based on 3D Weather Radar Data,” Frontiers in Environmental Science, vol. 9, pp. 1-10, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]
  20. Yang Zhao, Lixia Bi, and Xiangzhen Kong, “Identification and Feature Analysis of Thunderstorm based on AGRI and LMI of Fengyun-4A Satellite,” Atmospheric Research, vol. 318, pp. 1-13, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  21. Melina Y. Luque, Rodrigo E. Bürgesser, and Juan J. Ruiz, “Machine-Learning-based Estimation of Electrical Activity in the Weather Research and Forecasting Model,” Quarterly Journal of the Royal Meteorological Society, vol. 152, no. 776, 2026.
    [CrossRef] [Google Scholar] [Publisher Link]
  22. Fengquan Li et al., “Nowcasting of Cloud-To-Ground Lightning Location and Frequency based on a Deep Learning Technique,” Atmospheric and Oceanic Science Letters, vol. 19, no. 3, pp. 1-6, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  23. Yu Wang et al., “Extreme Climate Drivers and their Interactions in Lightning-Ignited Fires: Insights from Machine Learning Models,” Forests, vol. 16, no. 12, pp.  1-19, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  24. Paolo Francesco Duminuco et al., “A Convolutional Neural Network for Lightning Strikes Detection Over the Italian Territory using SEVIRI@MSG Data,” IEEE Transactions on Geoscience and Remote Sensing, vol. 63, pp. 1-13, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  25. Amirhossein Mostajabi et al., “Nowcasting Lightning Occurrence from Commonly Available Meteorological Parameters using Machine Learning Techniques,” npj Climate and Atmospheric Science, vol. 2, no. 1, pp. 1-15, 2019.
    [CrossRef] [Google Scholar] [Publisher Link]
  26. Gaurishankar K. Sawaisarje et al., “Improving Thunderstorm and Lightning Prediction using a Novel Hybrid Index and Machine Learning Techniques: A Case Study in Northeast India,” Meteorology and Atmospheric Physics, vol. 137, no. 4, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  27. Shamsul Ammar Shamsul Baharin et al., “Very High Frequency Radiation Emitted by Negative Narrow Bipolar Events Occurred over Malacca Strait,” Journal of Atmospheric and Solar-Terrestrial Physics, vol. 259, pp. 1-7, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
  28. Antony Sumardi et al., “The Roles of Rainfall Rate and Cloud Top Height in Lightning Activity during Northeastern Monsoon in Malay Peninsula,” 2025 13th Asia-Pacific International Conference on Lightning (APL), Bali, Indonesia, pp. 154-159, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  29. Harry Sucitra Roslan et al., “Relationship between Lightning Flashes and Three Flash Flood Events in Malacca, Malaysia,” ICLP 2024; 37th International Conference on Lightning Protection, Dresden, Germany, pp. 599-604, 2024.
    [Google Scholar] [Publisher Link]
  30. Humuntal Rumapea et al., “Improving Convective Cloud Classification with Deep Learning: The CC-UNet Model,” International Journal on Advanced Science, Engineering and Information Technology, vol. 14, no. 1, pp. 28-36, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
  31. Caixia Wang et al., “Application Research of Convolutional Neural Network and its Optimization in Lightning Electric Field Waveform Recognition,” Scientific Reports, vol. 15, no. 1, pp. 1-15, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  32. Asim Roy et al., “Explainable AI (XAI) for Object Detection and Application to Satellite Imagery,” IEEE Access, vol. 13, pp. 185597-185623, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  33. Nick Barua, and Masahito Hitosugi, “Advanced Multi-Modal Sensor Fusion System for Detecting Falling Humans: Quantitative Evaluation for Enhanced Vehicle Safety,” Vehicles, vol. 7, no. 149, pp. 1-17, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
  34. Alim Samat et al., “Meta-XGBoost for Hyperspectral Image Classification using Extended MSER-Guided Morphological Profiles,” Remote Sensing, vol. 12, no. 12, pp. 1-23, 2020.
    [CrossRef] [Google Scholar] [Publisher Link]
  35. S.A. Mohammad et al., “Azimuth and Elevation Factors Correction for Single Station Lightning Electromagnetic Field Sensor,” Geografia, vol. 17, no. 4, pp. 378-392, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]
  36. Qin Jiang et al., “Hybrid Multilayer Perceptron and Convolutional Neural Network Model to Predict Extreme Regional Precipitation Dominated by the Large-Scale Atmospheric Circulation,” Atmospheric Research, vol. 304, pp. 1-12, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
  37. Marcos Antonio Alves et al., “Lightning Nowcasting using Dual-Polarization Weather Radar and Machine Learning Approaches: Evaluation of Feature Engineering Strategies and Operational Integration,” Climate, vol. 13, no. 8, pp. 1-21, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]