科技科学

Tropical cyclones could be predicted with an extra day’s warning, thanks to an AI model

Nature正文已收录本站
  • NEWS AND VIEWS

The track and intensity of tropical cyclones can be predicted with high accuracy using an AI model, which has the potential to protect lives if shared responsibly worldwide.

By

  1. Tom Beucler

    1. Tom Beucler is in the Faculty of Geosciences and Environment and the Expertise Center for Climate Extremes at the University of Lausanne, Lausanne CH-1015, Switzerland.

  2. Milton Gomez

    1. Milton Gomez is in the Faculty of Geosciences and Environment and the Expertise Center for Climate Extremes at the University of Lausanne, Lausanne CH-1015, Switzerland.

Tropical cyclones — also called typhoons or hurricanes, depending on where they occur — are a major cause of deaths worldwide, and inflict considerable damage on property and infrastructure. The formation and evolution of these rapidly rotating weather systems are driven by effects on a range of scales, which makes it incredibly challenging to forecast them accurately using conventional models. Now, writing in Nature, Alet et al.1 introduce WeatherNext Cyclones (WN-C), an artificial-intelligence model that can produce two-week forecasts of unfolding tropical cyclones.

doi: https://doi.org/10.1038/d41586-026-02643-w

References

  1. Alet, F. et al. Nature https://doi.org/10.1038/s41586-026-10953-2 (2026).

    Article  Google Scholar 

  2. Mooney, K. R. et al. Earth Space Sci. 13, e2025EA004869 (2026).

    Article  Google Scholar 

  3. Zhang, Z. et al. Trop. Cyclone Res. Rev. 12, 30–49 (2023).

    Article  Google Scholar 

  4. Knapp, K. R., Kruk, M. C., Levinson, D. H., Diamond, H. J. & Neumann, C. J. Bull. Am. Meteorol. Soc. 91, 363–376 (2010).

    Article  Google Scholar 

  5. Gomez, M., Poulain-Auzéau, L., Berne, A. & Beucler, T. Artif. Intell. Earth Syst. 5, e250073 (2026).

    Article  Google Scholar 

  6. Alet, F. et al. Preprint at arXiv https://doi.org/10.48550/arXiv.2506.10772 (2025).

  7. Selz, T. & Craig, G. C. J. Geophys. Res. Mach. Learn. Comput. 3, e2025JH001180 (2026).

    Article  Google Scholar 

  8. Sun, Y. Q. et al. Proc. Natl Acad. Sci. USA 122, e2420914122 (2025).

    Article  PubMed  Google Scholar 

  9. Knutson, T. et al. Bull. Am. Meteorol. Soc. 101, E303–E322 (2020).

    Article  Google Scholar 

  10. McGovern, A. et al. Bull. Am. Meteorol. Soc. 105, E567–E583 (2024).

    Article  Google Scholar 

Download references

Competing Interests

The authors declare no competing interests.

  • Read the paper: Operational tropical-cyclone forecasting with AI

  • The Antarctic ice sheet has gained mass — but probably not for long

  • Do wet or dry soils trigger thunderstorms? It depends on how the wind blows

  • See all News & Views

Subjects

Latest on:

  • Climate sciences
  • Mathematics and computing

  • UN finally admits global warming will shoot past 1.5 ºC climate limit

    News

  • Atmospheric CH4 plateau sustained by NOx emission rise and southward shift

    Article

  • Satellite images before Nepal disaster showed warning signs

    News Q&A

  • Explainable deep learning improves human mental models of self-driving cars

    Article

  • Self-driving cars make mistakes, and now users can see why

    News & Views

  • Designing physics experiments with artificial intelligence

    Review Article