DONG Lin ZHAO Dajun XU Hongxiong WANG Hui QU Honghyu HUANG Yiwu NIE Gaozhen XIANG Chunyi WANG Qian LYU Xinyan

The use of artificial intelligence to advance scientific research (AI for Science) is transforming the traditional scientific research paradigm, and AI weather models have already been deployed in operational applications. Although AI weather models have made significant progress in typhoon track forecasting, there are still obvious limitations in capturing the complex physical processes inside typhoons and predicting the changes in typhoon intensity, especially the rapid intensification (RI) of typhoons and the evolution of their internal structures, mainly because AI weather models lack sufficient resolution support in distinguishing typhoon intensity and fine structure. The use of AI weather model-driven regional model technology can overcome this bottleneck, so as to more accurately forecast typhoon intensity and structural evolution. To this end, the Chinese Academy of Meteorological Sciences and the National Meteorological Center jointly developed the AI-driven Typhoon Rapid Analysis and Forecasting System (AI-TRANS) based on this technical route. In this paper, the forecast results of AI-TRANS are compared with other six operational models by using the data of typhoon forecast in 2024. The results show that the performance of AI-TRANS is basically the same as that of the existing operational model in typhoon track forecasting. The prediction of typhoon intensity and structure shows obvious advantages. The system integrates the advantages of AI weather model and numerical model, and significantly improves the accuracy of typhoon landfall forecast by improving the track forecast stability of AI weather model. At the same time, the AI weather model-driven regional model technology has effectively improved the forecasting effect of key indicators such as typhoon RI, lifetime maximum intensity and fine structure. Taking the severe typhoons Gaemi and Yagi in 2024 as an example, the AI-TRANS system successfully predicted the track rotation, RI process and concentric eyewalls structure of Gaemi in the east side of Taiwan Island, and accurately captured the RI, concentric eyewalls formation and eye wall replacement process after Typhoon Yagi entered the South China Sea. This is the first time that a model has successfully predicted the concentric eyewalls structure and the eyewall replacement process of typhoon in real-time typhoon forecasting, fully demonstrating the critical supporting role of the AI-TRANS system in improving the accuracy of typhoon intensity forecasts and enhancing disaster prevention and mitigation efforts.