Abstract:

Based on original ensemble forecast data obtained from the European Center for Medium-range Weather Forecast (ECMWF) and a joint probability method, we produced two ensemble forecast operational products for severe weather occurring frequently in Beijing, including cold wave in winter and severe convection weather in summer.The joint probability forecast product for the cold wave warning was developed based on air temperature at 2 m and mean winds at 10 m, and that for severe convection was produced using the convective available potential energy (CAPE) and vertical wind shear within 0-6 km.We also evaluated the performance of the joint probability forecast products using the observations during cold wave and severe convection events in Beijing in recent years.The results indicated that when the forecast probability reaches 10% and above, the cold wave can probably reach the level of blue warning signals in real-time.The joint probability forecast product for cold wave performs the best in the northwestern region in Beijing, followed by the southeastern region.In areas with the blue warning signals, the cold wave product has a higher capture rate, while in areas with the yellow warning signals it has a high missing rate.The potential prediction of severe convection events using the joint probability method has a high false alarming ratio.Severe convection events can probably occur in real-time only when the forecast probability reaches 90%-100%.Compared with local severe convection events, areas with high forecast probability correspond to the city-scale severe convection events concentrated over the whole city.

Key words: Ensemble forecast, Joint probability, Cold wave, Severe convection