Abstract: The conventional observation data,the European Centre for Medium-Range Weather Forecasting(ECMWF)numerical weather prediction products,the Weather Research and Forecasting Model(WRF)data,and radar detection products were used to study a local heavy rain process happening on June 30,2016,in Shenyang.The evolution of the precipitation system,short-term and very short-term predictability,and model production application were analyzed.The results show that this heavy rain has clear characteristics for short-term prediction,but it does not meet the threshold values obtained from the past rainfall forecasts in Liaoning province.The 90% percentile of accumulated precipitation and EFI (Extreme Forecast Index) forecasted by the ECMWF ensemble model indicate the extreme feature of this heavy rain process.The large vertical wind shear happening in the southern Liaoning results in sudden changing and recurring convective cloud clusters.All models analyzed in this study can predict the distribution characteristics of rainfall,but the location of precipitation needs improving.The ECMWF forecasts from different start times have relatively large uncertainties.The forecast result from the last starting time has a good reference value.The WRF model can be used to analyze the evolution characteristics of convective cloud clusters and therefore we should focus on the investigation of evolution processes of convective clouds using the WRF model.

Key words: Local heavy rainstorm, Convective cloud cluster, ECMWF model, WRF model