European Center for Medium-range Weather Forecasting (ECMWF) high-resolution model and ECMWF Extreme Weather Forecast Index (EFI) are evaluated and compared for heavy rainfall during the flood season (May to September) in Yangtze River Delta. The studies show the following results. (1) The ETS score of ECMWF model and EFI of precipitation for heavy rainfall forecast are significantly reduced with the extension of forecast period. In a short time period, ECMWF model precipitation forecast is superior, but EFI tends to be better after 60 h. (2) In different forecast time periods, using different precipitation thresholds to predict the heavy rainfall is expected to achieve the best score. In the short term period, the threshold is gradually reduced from 55 mm to 35 mm with the extension of forecast period. (3) For the EFI of precipitation, when the EFI is 0.65~0.7 in 12-36 h, the ETS score of heavy rainfall forecast is the highest. The threshold of EFI also decreases with the prolongation of the prediction period. The ETS score is the highest when the EFI takes 0.55~0.6 at 60-84 h. (4)The combination of EC-thin precipitation and EFI of precipitation by reasonable methods and thresholds can obtain higher score of heavy rainfall forecast in different forecasting time periods.
.2018. Evaluation and application of ECMWF model precipitation and extreme weather forecast index of precipitation on heavy rainfall forecast[J].
Torrential Rain and Disasters, 37(6): 566-573.