Hurricane Forecasting: A Novel Multimodal Machine Learning Framework

Léonard Boussioux, Cynthia Zeng, Théo Guénais, Dimitris Bertsimas

    Research output: Contribution to journalArticlepeer-review


    This paper describes a novel machine learning (ML) framework for tropical cyclone intensity and track forecasting, combining multiple ML techniques and utilizing diverse data sources. Our multimodal framework, called Hurricast, efficiently combines spatial–temporal data with statistical data by extracting features with deep learning encoder–decoder architectures and predicting with gradient-boosted trees. We evaluate our models in the North Atlantic and eastern Pacific basins in 2016–19 for 24-h lead-time track and intensity forecasts and show they achieve comparable mean absolute error and skill to current operational forecast models while computing in seconds. Furthermore, the inclusion of Hurricast into an operational forecast consensus model could improve upon the National Hurricane Center’s official forecast, thus highlighting the complementary properties with existing approaches. In summary, our work demonstrates that utilizing machine learning techniques to combine different data sources can lead to new opportunities in tropical cyclone forecasting.

    Original languageEnglish (US)
    Pages (from-to)817-831
    Number of pages15
    JournalWeather and Forecasting
    Issue number6
    StatePublished - Jun 2022


    • Artificial intelligence
    • Atmosphere
    • Data science
    • Decision trees
    • Deep learning
    • Dimensionality reduction
    • Ensembles
    • Forecasting
    • Machine learning
    • Neural networks
    • Operational forecasting
    • Optimization
    • Other artificial intelligence/machine learning
    • Reanalysis data
    • Regression
    • Statistical techniques
    • Superensembles
    • Time series
    • Tropical cyclones

    ASJC Scopus subject areas

    • Atmospheric Science


    Dive into the research topics of 'Hurricane Forecasting: A Novel Multimodal Machine Learning Framework'. Together they form a unique fingerprint.

    Cite this