Journal article · 2026 · Open access

A Three-Stage Deep Learning Framework for Short-Term Tropical Cyclone Track Prediction

Haocheng Shi, Dan Song, Guijing Yang, Longyu Jiang, Xuezhu Wang, Shuangyan He

Journal of Marine Science and Engineering, 14(13), 1159 · Published 23 June 2026 · DOI 10.3390/jmse14131159

Open full textCite this paper

In brief

“A Three-Stage Deep Learning Framework for Short-Term Tropical Cyclone Track Prediction” is a 2026 journal article by Haocheng Shi and 5 co-authors, published in Journal of Marine Science and Engineering. OpenAlex counts 1 citations as of 20 November 2026. It is open access under a CC BY licence: the full text is free to read at Journal of Marine Science and Engineering.

Citations
1
Most cited in
2026 (1)
Field-weighted impact
1.40
References
44
Authors
6

Abstract

Accurate tropical cyclone (TC) track prediction remains challenging, as numerical models suffer from high computational cost, substantial storage requirements, and physical parameterization uncertainties, while data-driven large AI models depend heavily on training data volume and high-resolution inputs, resulting in prohibitive computational overhead. To address these issues, this paper proposes TCN-GAN-DM, a three-stage deep learning framework based on the China Meteorological Administration (CMA) Tropical Cyclone Best Track Dataset. Specifically, a dual-stream temporal convolutional network (TCN) first extracts temporal features from track and meteorological sequences, respectively. A generative adversarial network (GAN) then takes these features and produces multiple physically plausible candidate tracks via noise injection. Finally, a conditional diffusion model (DM) refines the predicted positions through progressive denoising. Experimental results for TCs in 2024 show that under the fair deterministic comparison using a single fixed candidate, the model achieves a 6 h track error of 49.10 km, which is comparable to CMA-GFS (49.75 km) and HWRF (44.34 km), and substantially lower than the large AI model FuXi (120.44 km). When evaluating the oracle metric (best-of-K, K = 6) as an upper bound of coverage, the model achieves the smallest errors among all models at 6 h (24.04 km) and 12 h (55.81 km). In addition, the proposed model has advantages over CMA-GFS, HWRF, and FuXi in terms of computational resource consumption and hardware deployment cost. However, its mean track error increases more rapidly beyond 12 h, and at lead times of 18 h and 24 h the model is outperformed by HWRF, FuXi, and CMA-GFS, indicating that its current strength lies primarily in short-term prediction. Consequently, the practical utility of TCN-GAN-DM is currently demonstrated for 6–12 h TC track prediction, offering a new solution for disaster prevention and mitigation that balances accuracy and deployment cost at these specific time scales.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Shi, H., Song, D., Yang, G., Jiang, L., Wang, X., & He, S. (2026). A Three-Stage Deep Learning Framework for Short-Term Tropical Cyclone Track Prediction. Journal of Marine Science and Engineering, 14(13), 1159. https://doi.org/10.3390/jmse14131159

MLA 9

Shi, Haocheng, et al. “A Three-Stage Deep Learning Framework for Short-Term Tropical Cyclone Track Prediction.” Journal of Marine Science and Engineering, vol. 14, no. 13, 2026, p. 1159. https://doi.org/10.3390/jmse14131159.

Chicago

Shi, Haocheng, Dan Song, Guijing Yang, Longyu Jiang, Xuezhu Wang, and Shuangyan He. 2026. “A Three-Stage Deep Learning Framework for Short-Term Tropical Cyclone Track Prediction.” Journal of Marine Science and Engineering 14 (13): 1159. https://doi.org/10.3390/jmse14131159.

Harvard

Shi, H. et al. (2026) ‘A Three-Stage Deep Learning Framework for Short-Term Tropical Cyclone Track Prediction’, Journal of Marine Science and Engineering, 14(13), p. 1159. doi:10.3390/jmse14131159.

IEEE

H. Shi, D. Song, G. Yang, L. Jiang, X. Wang, and S. He, “A Three-Stage Deep Learning Framework for Short-Term Tropical Cyclone Track Prediction,” Journal of Marine Science and Engineering, vol. 14, no. 13, p. 1159, 2026, doi: 10.3390/jmse14131159.

BibTeX

@article{shi2026three,
  title     = {{A Three-Stage Deep Learning Framework for Short-Term Tropical Cyclone Track Prediction}},
  author    = {Shi, Haocheng and Song, Dan and Yang, Guijing and Jiang, Longyu and Wang, Xuezhu and He, Shuangyan},
  journal   = {Journal of Marine Science and Engineering},
  year      = {2026},
  volume    = {14},
  number    = {13},
  pages     = {1159},
  doi       = {10.3390/jmse14131159},
  url       = {https://doi.org/10.3390/jmse14131159}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Earth and Planetary Sciences › Atmospheric Science › Tropical and Extratropical Cyclones Research

  • tropical cyclone track forecasting
  • short-term forecasting
  • temporal convolutional network
  • generative adversarial networks
  • conditional diffusion models
  • noise injection
  • best track data
  • computational complexity reduction

Work with this paper

Most-cited papers in Earth and Planetary Sciences

All highly cited Earth and Planetary Sciences papers

Questions about this paper

How do I cite this paper in APA style?

Shi, H., Song, D., Yang, G., Jiang, L., Wang, X., & He, S. (2026). A Three-Stage Deep Learning Framework for Short-Term Tropical Cyclone Track Prediction. Journal of Marine Science and Engineering, 14(13), 1159. https://doi.org/10.3390/jmse14131159

How many times has this paper been cited?

1 times, according to OpenAlex on 20 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at Journal of Marine Science and Engineering: sign in and open the full text from this page.

Who wrote this paper?

Haocheng Shi, Dan Song, Guijing Yang, Longyu Jiang, Xuezhu Wang, Shuangyan He.

What is the DOI of this paper?

10.3390/jmse14131159, permanently at https://doi.org/10.3390/jmse14131159. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 20 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.