Recently, the Fengqing AI Global Short-to-Medium-Range Ensemble Weather Forecasting System V1.0 successfully cleared operational evaluation and has been officially implemented in national meteorological operations. Leveraging the AI4S supercomputing cluster at the National Supercomputing Center in Tianjin, the system undergoes comprehensive training and iterative refinement throughout its entire process. It marks China’s first operational-level meteorological AI ensemble forecasting model, boasting fully autonomous and controllable data, models, and computing capabilities. This breakthrough fills a technological void in China’s global medium-range ensemble intelligent forecasting landscape.

Developed collaboratively by the National Meteorological Center and multiple partner organizations, the system harnesses a blend of cutting-edge technologies to more precisely capture uncertainties in atmospheric dynamics and bolster probabilistic forecasting capabilities for severe weather events. It provides effective global forecasting lead times of up to 11 days, with several key performance metrics outperforming those of the latest AI ensemble forecasting system from the European Centre for Medium-Range Weather Forecasts. In contrast to traditional numerical models, this system can swiftly compute global-scale meteorological element derivations within minutes, striking an optimal balance between forecast speed, accuracy, and stability.

Following its operational deployment, the system is set to serve multiple sectors. Meanwhile, the development team will persist in refining the model to further enhance China’s domestic intelligent meteorological forecasting capabilities. This system stands as a testament to Tianjin’s supercomputing prowess in bolstering national-level meteorological technological innovation. It is poised to propel independent innovation within China’s meteorological science and technology sector, as well as foster advancements in global intelligent weather forecasting technologies.