Hostname: page-component-54dcc4c588-42vt5 Total loading time: 0 Render date: 2025-09-24T01:55:42.991Z Has data issue: false hasContentIssue false
  • English
  • Français

Nowcasting of typhoon tracks based on LLM and RAG

Published online by Cambridge University Press:  19 September 2025

Baichuan Peng ,
Youchao Jiang  and
Li Yao Open the ORCID record for Li Yao [Opens in a new window]
Baichuan Peng
Affiliation:
Merchant Marine College, Shanghai Maritime University, Shanghai, China
Youchao Jiang
Affiliation:
Merchant Marine College, Shanghai Maritime University, Shanghai, China
Li Yao*
Affiliation:
Merchant Marine College, Shanghai Maritime University, Shanghai, China
*
Corresponding author: Li Yao; Email: yaoli@shmtu.edu.cn
Get access Rights & Permissions [Opens in a new window]

Abstract

Accurate typhoon track nowcasting is vital for navigation and coastal disaster prevention. This research integrates a Large Language Model (LLM) with Retrieval-Augmented Generation (RAG) technology for typhoon path prediction. Leveraging LLMs as the predictive foundation, the approach tailors forecasts to individual typhoon characteristics. The methodology involves collecting satellite imagery, standardizing data, and employing optical flow methods to track typhoons and derive path coordinates. These coordinates are preprocessed and embedded into the LLM. RAG enhances the LLM’s predictive performance, enabling effective forecasting. Increasing typhoon-specific embedded data further improves accuracy. Using the FY-4 dataset, the method achieved an average absolute error of 10.78 km in 12-hour predictions. The study demonstrates that LLM-RAG integration excels in nowcasting.

Keywords

LLMnowcasting of typhoon tracksRAG

Information

Type
Research Article
Information
The Journal of Navigation , First View , pp. 1 - 16
Check for updates
Copyright
© The Author(s), 2025. Published by Cambridge University Press on behalf of The Royal Institute of Navigation

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Article purchase

Temporarily unavailable

References

Chen, R., Zhang, W. and Wang, X. (2020). Machine Learning in Tropical Cyclone Forecast Modeling: A Review. Atmosphere, 11, 676–701.10.3390/atmos11070676CrossRefGoogle Scholar
Chen, Y. and Duan, Z. (2018). A Statistical Dynamics Track Model of Tropical Cyclones for Assessing Typhoon Wind Hazard in the Coast of Southeast China. Journal of Wind Engineering and Industrial Aerodynamics, 172, 325–340.10.1016/j.jweia.2017.11.014CrossRefGoogle Scholar
Chib, P.S.S. (2024). Pravendra, LG-Traj: LLM Guided Pedestrian Trajectory Prediction. Computer Vision and Pattern Recognition.Google Scholar
Fu, H., Gu, Z. and Wang, Y. (2022). Ship Pitch Prediction Based on Bi-ConvLSTM-CA Model. Journal of Marine Science and Engineering, 10, 840–859.10.3390/jmse10070840CrossRefGoogle Scholar
Gao, S., Zhao, P., Pan, B., Li, Y., Zhou, M., Xu, J., Zhong, S. and Shi, Z. (2018). A Nowcasting Model for the Prediction of Typhoon Tracks Based on a Long Short Term Memory Neural Network. Acta Oceanologica Sinica, 37, 8–12.10.1007/s13131-018-1219-zCrossRefGoogle Scholar
Hao, F. and Jin, J. (2022). A Broad Learning Ensemble System Using Bagging for Typhoon Trajectory Forecasting. In: 2022 2nd International Conference on Consumer Electronics and Computer Engineering (ICCECE), Institute of Electrical and Electronics Engineers (IEEE), Piscataway, NJ, USA, 729–733.10.1109/ICCECE54139.2022.9712832CrossRefGoogle Scholar
Hon, K.-K. (2020). Tropical Cyclone Track Prediction Using a Large-Area Wrf Model at the Hong Kong Observatory, Tropical Cyclone Research and Review, 9, 67–74.10.1016/j.tcrr.2020.03.002CrossRefGoogle Scholar
Jabbar, A., Li, X. and Omar, B. (2021). A Survey on Generative Adversarial Networks: Variants, Applications, and Training. ACM Computing Surveys, 54, 1–49.10.1145/3463475CrossRefGoogle Scholar
Jiang, D., Shi, G., Li, N., Ma, L., Li, W. and Shi, J. (2023). TRFM-LS: Transformer-Based Deep Learning Method for Vessel Trajectory Prediction. Journal of Marine Science and Engineering, 11, 880–901.10.3390/jmse11040880CrossRefGoogle Scholar
Jiang, Y., Yu, Y., Ma, Y., Zhuang, J., Ye, X., Yuan, Y. and Su, G.(2024). Typhoon Track Image Prediction Using a Comprehensive LSTM Attention Model in Deep Learning, In: 2024 4th International Conference on Neural Networks, Information and Communication (NNICE), Institute of Electrical and Electronics Engineers (IEEE), Piscataway, NJ, USA, 537–541.10.1109/NNICE61279.2024.10498809CrossRefGoogle Scholar
Jung, H., Baek, Y.-H., Moon, I.-J., Lee, J. and Sohn, E.-H. (2024). Tropical Cyclone Intensity Estimation Through Convolutional Neural Network Transfer Learning using Two Geostationary Satellite Datasets. Frontiers in Earth Science, 11, 104798.10.3389/feart.2023.1285138CrossRefGoogle Scholar
Lan, Z.X., Liu, L.S., Fan, B., Lv, Y.S., Ren, Y.L. and Cui, Z.Y. (2025). Traj-LLM: A New Exploration for Empowering Trajectory Prediction with Pre-trained Large Language Models. IEEE Transactions on Intelligent Vehicles, 10, 794–807.10.1109/TIV.2024.3418522CrossRefGoogle Scholar
Li, T., Lai, M., Nie, S., Liu, H., Liang, Z. and Lv, W. (2024) Tropical Cyclone Trajectory Based on Satellite Remote Sensing Prediction and Time Attention Mechanism ConvLSTM Model. Big Data Research, 36, 100439-100447.10.1016/j.bdr.2024.100439CrossRefGoogle Scholar
Lian, J., Dong, P., Zhang, Y., Pan, J. and Liu, K. (2020). A Novel Data-Driven Tropical Cyclone Track Prediction Model Based on CNN and GRU With Multi-Dimensional Feature Selection. IEEE Access, 8, 97114–97128.10.1109/ACCESS.2020.2992083CrossRefGoogle Scholar
Liang, Z., Yang, Y., Sun, C., He, W., Dai, Z., Su, G. and Yu, Y. (2023). Typhoon Track Prediction Based on Dual Attention LSTM Model. In: 2023 11th International Conference on Information Systems and Computing Technology (ISCTech), Institute of Electrical and Electronics Engineers (IEEE), Piscataway, NJ, USA,519–525.10.1109/ISCTech60480.2023.00098CrossRefGoogle Scholar
Lu, P., Sun, A., Xu, M., Wang, Z., Zheng, Z., Xie, Y. and Wang, W. (2022a). A Time Series Image Prediction Method Combining a Cnn and Lstm and its Application in Typhoon Track Prediction. Mathematical Biosciences and Engineering, 19, 12260–12278.10.3934/mbe.2022571CrossRefGoogle ScholarPubMed
Lu, P., Xu, M., Chen, M., Wang, Z., Zheng, Z. and Yin, Y. (2023). Multi-Step Prediction of Typhoon Tracks Combining Reanalysis Image Fusion Using Laplacian Pyramid and Discrete Wavelet Transform with ConvLSTM. Axioms, 12, 107036.10.3390/axioms12090874CrossRefGoogle Scholar
Lu, P., Xu, M., Sun, A., Wang, Z. and Zheng, Z. (2022b). Typhoon Tracks Prediction with ConvLSTM Fused Reanalysis Data. Electronics, 11, 3279.10.3390/electronics11203279CrossRefGoogle Scholar
Mahmoud, H. and Akkari, N. (2016). Shortest Path Calculation: A Comparative Study for Location-Based Recommender System. In: 2016 World Symposium on Computer Applications & Research (WSCAR), Institute of Electrical and Electronics Engineers (IEEE), Piscataway, NJ, USA, 1–5.Google Scholar
Munir, K. and Sheraz Anjum, M. (2018). The Use of Ontologies for Effective Knowledge Modelling and Information Retrieval. Applied Computing and Informatics, 14, 116–126.10.1016/j.aci.2017.07.003CrossRefGoogle Scholar
Pang, Y., Zhao, X., Yan, H. and Liu, Y. (2021). Data-Driven Trajectory Prediction with Weather Uncertainties: A Bayesian Deep Learning Approach. Transportation Research Part C: Emerging Technologies, 130, 103326.10.1016/j.trc.2021.103326CrossRefGoogle Scholar
Peng, M.G., Chen, X., Zhu, X., Chen, M., Yang, K., Wang, H., Wang, Y.X. (2024). LC-LLM explainable lan-change intention and trajectory predictions with large language models.10.1016/j.commtr.2025.100170CrossRefGoogle Scholar
Qin, W., Tang, J. and Lao, S. (2022). DeepFR: A Trajectory Prediction Model Based on Deep Feature Representation. Information Sciences, 604, 226–248.10.1016/j.ins.2022.05.019CrossRefGoogle Scholar
Qin, W., Tang, J., Lu, C. and Lao, S. (2022). A Typhoon Trajectory Prediction Model Based on Multimodal and Multitask Learning. Applied Soft Computing, 122, 108804.10.1016/j.asoc.2022.108804CrossRefGoogle Scholar
Ren, J., Xu, N. and Cui, Y. (2022). Typhoon Track Prediction Based on Deep Learning. Applied Sciences, 12, 80–107.10.3390/app12168028CrossRefGoogle Scholar
Ruttgers, M., Jeon, S., Lee, S. and You, D. (2022). Prediction of Typhoon Track and Intensity Using a Generative Adversarial Network With Observational and Meteorological Data. IEEE Access, 10, 48434–48446.10.1109/ACCESS.2022.3172301CrossRefGoogle Scholar
Song, T., Li, Y., Meng, F., Xie, P. and Xu, D. (2022). A Novel Deep Learning Model by BiGRU with Attention Mechanism for Tropical Cyclone Track Prediction in the Northwest Pacific, Journal of Applied Meteorology and Climatology, 61, 3–12.10.1175/JAMC-D-20-0291.1CrossRefGoogle Scholar
Suo, Y., Chen, W., Claramunt, C. and Yang, S. (2020). A Ship Trajectory Prediction Framework Based on a Recurrent Neural Network. Sensors (Basel), 20, 5133.10.3390/s20185133CrossRefGoogle ScholarPubMed
Yang, Z.W.X. (2018). Moving Target Detection and Tracking Based on Pyramid Lucas-Kanade Optical Flow. In: 2018 IEEE 3rd International Conference on Image, Vision and Computing (ICIVC), Institute of Electrical and Electronics Engineers (IEEE), Piscataway, NJ, USA.Google Scholar
Zhan, R., Lu, J., Jiang, Y., Li, T., Zhong, G. and Lv, W. (2023). A Multimodal Deep Learning Approach for Typhoon Track Forecast by Fusing CNN and Transformer Structures. In: 2023 4th International Conference on Computer Vision, Image and Deep Learning (CVIDL), Institute of Electrical and Electronics Engineers (IEEE), Piscataway, NJ, USA, 391–394.10.1109/CVIDL58838.2023.10166556CrossRefGoogle Scholar