Comparison on quantitative estimation of precipitation using radar based on different calibration methods

Based on the observational data from the low-elevation angle radar and precipitation data from intensive automatic weather station,the local Z-I relationships were established by an optimization method according to precipitation feature and the different precipitation types.According to application test,the Z-I relationship of the cumulus clouds mixed precipitation type is more applicable.The results of precipitation estimated using six calibration methods were compared after the Z-I relationships have been modified specifically for Tianjin area.The results indicate that the errors from both the Z-I relationship calibration method and the maximum integrated method are large;they both overestimate rainfall.The estimation accuracy of the optimal interpolation method is the highest among the all methods,and its mean absolute error and root-mean-square error are the smallest.The correlation coefficient between the estimated rainfall from different methods and the observed rainfall are calculated,and it shows that the correlation between the estimated rainfall of variational calibration method and rain gauge rainfall is significant.To some extent,the rainfall is overestimated to light rain grade among the all methods.