Improving Ensemble Forecasts of Abnormally Deflecting Tropical Cyclones with Fused Atmosphere-Ocean-Terrain Data

2026-03-30

AI Summary

This study addresses the challenge of accurately forecasting anomalous recurving tracks of tropical cyclones, a task at which existing deep learning approaches often falter due to their reliance on single-trajectory assumptions or homogeneous meteorological variables. To overcome this limitation, the authors introduce AOT-TCs—the first multi-source, heterogeneous dataset integrating atmospheric, oceanic, and topographic data—and propose a novel deep learning model that explicitly couples physical processes across multiple Earth system spheres. Evaluated on all tropical cyclones in the western North Pacific from 2017 to 2024, the proposed method substantially improves forecast accuracy, achieving breakthrough performance particularly in predicting anomalous turning trajectories.

Technology Category

• Machine Learning: Multimodal Learning
• Search and Optimization: Learning to Search
• Multiagent Systems: Multiagent Learning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Deep learning-based tropical cyclone (TC) forecasting methods have demonstrated significant potential and application advantages, as they feature much lower computational cost and faster operation speed than numerical weather prediction models. However, existing deep learning methods still have key limitations: they can only process a single type of sequential trajectory data or homogeneous meteorological variables, and fail to achieve accurate forecasting of abnormal deflected TCs. To address these challenges, we present two groundbreaking contributions. First, we have constructed a multimodal and multi-source dataset named AOT-TCs for TC forecasting in the Northwest Pacific basin. As the first dataset of its kind, it innovatively integrates heterogeneous variables from the atmosphere, ocean, and land, thus obtaining a comprehensive and information-rich meteorological dataset. Second, based on the AOT-TCs dataset, we propose a forecasting model that can handle both normal and abnormally deflected TCs. This is the first TC forecasting model to adopt an explicit atmosphere-ocean-terrain coupling architecture, enabling it to effectively capture complex interactions across physical domains. Extensive experiments on all TC cases in the Northwest Pacific from 2017 to 2024 show that our model achieves state-of-the-art performance in TC forecasting: it not only significantly improves the forecasting accuracy of normal TCs but also breaks through the technical bottleneck in forecasting abnormally deflected TCs.

Problem

• tropical cyclone
• abnormal deflection
• ensemble forecast
• multimodal data
• atmosphere-ocean-terrain coupling

Innovation

• multimodal data fusion
• atmosphere-ocean-terrain coupling
• abnormally deflecting tropical cyclones
• deep learning forecasting
• AOT-TCs dataset