Recently, the reporter learned from the University of Science and Technology of China that the Advanced Laboratory of Atmospheric Sciences (LACAR) of the university has successfully "predicted" the large-scale Meiyu event in Japan and surrounding areas in 2020 using the Global Convection Analysis Model (iAMAS) independently developed based on the domestically produced Sunway Ocean Light supercomputer. The predictions made by the laboratory actually refer to retrospective analysis and validation of events that have occurred through scientific research and model simulations, in order to demonstrate the effectiveness and accuracy of the model. The relevant paper was recently published in Environmental Research Letters. Meiyu is a long-term and large-scale rainfall phenomenon that often occurs in East Asia during summer, mainly distributed in the latitude zones of the middle and lower reaches of the Yangtze River, the Korean Peninsula, and Kyushu Island in Japan. In 2020, the record breaking "violent plum blossom" brought heavy flood control pressure and serious damage to the Jianghuai region of China, and also caused severe plum rain disasters in the Kyushu Island area of Japan, seriously affecting the lives and production activities of local residents. The sub seasonal forecasting of such rainfall events has always been a challenge in academic research and operational forecasting, making it difficult to meet the accuracy requirements of residents and emergency management departments for mid-term weather forecasting. The research team relies on the Sunway Ocean Light supercomputer to build iAMAS, optimizes algorithms based on its hardware features, improves computing speed, breaks through the bottleneck of massive data reading and writing, and makes monthly forecasting of global convective resolution scales possible. Researchers conducted multiple monthly forecasting experiments with different resolutions to explore the predictability of summer plum rain in 2020. Researchers have found that when using global low resolution forecasts, there is a significant northward shift deviation in the Meiyu Belt, and regional densification to convective resolution scale cannot solve the problem. At a coarser resolution, the expansion of the subtropical high pressure system in the northwest Pacific leads to the northward movement and weakening of the rainfall belt. Further research by the team found that iAMAS can effectively capture deep convection in the equatorial region and simulate rainfall processes and circulation patterns in the western Pacific equatorial region that are close to observational data. This reasonable simulation reproduced the intensity and location of the subtropical high pressure, which affects the circulation in mid latitude regions. By mastering the sinking and rising positions of the airflow, researchers successfully predicted the intensity and location of this plum rain event. (New Society)