Abstract:Using percentile method and the 1981-2010 daily precipitation data, this paper proposed extreme rainstorm threshold and regional extreme rainstorm standard of Henan Province. Aiming at the selected 13 extreme rainstorm cases in 1981-2016, this paper analyzed the environment parameters of these cases using the ECMWF reanalysis datasets. The results show that the average values of environment parameters such as specific humidity and vorticity at 850 hPa, vertical speed and V wind component at 700 hPa, divergence at 200 hPa, whole-layer precipitatle water, K index and 0-3 km vertical wind shear, which represent the water vapor, dynamics and instability conditions, deviate far from their climatic ave-rage during the extreme rainstorms. This means these environment parameters are instructive to extreme rainstorm forecasting. The standard deviation multiple of environment parameters has good correlations with rainfall, and their quantile values are distributed along the diagonal line. When quantile value of standard deviation multiple of environment parameters is greater than or equal to 80%, it is positively correlated to abnormal precipitation (quantile values ≥90%). However there are more light precipitations when standard deviation multiple of environment parameters is at lower quantile value (<40%). The anomaly of single environment parameter cannot reflect the degree of precipitation anomaly in general. Thus we build the extreme rainstorm index (ERI) based on the eight environment parameters discussed above and also test it with the 13 extreme rainstorm cases. The test results show that the TS, missing rate and false rate are 35%, 27%, 49% respectively when ERI is greater than or equal to 0.7. These values can be used as the reference thresholds for the forecasting of extreme rainstorm. The practical application of ERI during “Rumbia” typhoon extreme rainstorm process indicated that ERI has good reference meaning for the forecast of extreme rainstorm location.