Hurricane Forecasting: A Novel Multimodal Machine Learning Framework

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 extr...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Weather and forecasting Jg. 37; H. 6; S. 817 - 831
Hauptverfasser: Boussioux, Léonard, Zeng, Cynthia, Guénais, Théo, Bertsimas, Dimitris
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Boston American Meteorological Society 01.06.2022
Schlagworte:
Coders
Cyclone forecasting
Cyclones
Data sources
Deep learning
Feature extraction
Forecasting
Frameworks
Hurricane forecasting
Hurricanes
Lead time
Learning algorithms
Machine learning
Mathematical models
Spatiotemporal data
Tropical climate
Tropical cyclone forecasting
Tropical cyclone intensities
Tropical cyclone movement
Tropical cyclones
ISSN:0882-8156, 1520-0434
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung: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.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0882-8156
1520-0434
DOI:10.1175/WAF-D-21-0091.1