| Project Reference: | ITP/047/23LP |
| Project Title: | AI-Assisted Production of Ocean Surface Winds from Remote Sensing Satellites for Typhoon Forecast Improvement |
| Hosting Institution: | The Hong Kong University of Science and Technology (HKUST) |
| Abstract: | Accurately forecasting the track and intensity of typhoons is crucial to reducing their associated economic losses. In Hong Kong, the issuance of No. 8 gale or storm force wind signal results in an average economic loss of approximately 3.7 billion HKD per day of a city-wide shutdown. The Hong Kong Observatory's decision to issue the No. 8 warning relies heavily on numerical model simulations of ocean surface wind (OSW) structures, which are highly uncertain. Although several datasets of satellite OSW observations exist, their varying resolution, coverage, and accuracy make them impractical for operational use. We propose utilizing advanced artificial intelligence techniques to perform satellite OSW data fusion and generate uniformly distributed gapfree high-accuracy OSW in the western Pacific Ocean and South China Sea. A physicsbased data assimilation model will incorporate the integrated OSW for typhoon forecast. Our project will deliver an integrated OSW dataset with 1-km and 3-hour resolution and 1- m/s accuracy, an optimized typhoon forecast model representing 20% or greater improvement in forecast accuracy over existing models, and a user-friendly data service platform. Our timely and accurate weather forecast system will benefit many industries including logistics and supply chain that rely on accurate typhoon forecasts for efficient operations. |
| Project Coordinator: | Prof Hui Su |
| Approved Funding Amount: | HK$ 9.26M |
| Project Period: | 30 Dec 2023 - 29 Dec 2025 |
