Extended probabilistic Typhoon forecast in South China Sea and operational weather forecast via satellite communication

Significant predictability is found in medium-range weather systems over East Asia, particularly in those convective hazard weather events bringing numerous obstacles to the transportation and infrastructures. Based on a most advanced global numerical model product, we developed a real-time warning system to provide medium-range (5-15 days) forecast on danger events. The forecast is probabilistic to help the decision makers do gain/lost estimation. In 2007 and 2008 three flooding events on the Brahmaputra River were well predicted 10 days in advance. The medium-range genesis forecast skill of typhoon, a relative small-scale but much destructive weather system, is explored on examples of Megi and Doksuri. For both typhoons the maximum winds and minimum center pressure are well predicted as well as the tracks. The forecast bias might be due to the global model's resolution limit or the missing of statistical rendering.