Improvement In The Quantitative Measurement Of Precipitation In Use Of Doppler Radar

In this paper, several questions of traditional statistical method for Z-I relationship are discussed. To virtually achieve the optimization in space-time, the design of realtime statistic is provided. In statistical process, the improved discriminant criterion, that the estimated precipitation by radar shouldn’t only reflect the distribution of the rain rate, but the mean of the rain in analized area, is given.Base on the given improved discriminant criterion, the improved optimization Z-I fitting method and the relative CTF is given, which insures the minimum of the fitting residuals with non-systerm erro theoretically. In the improved method, both of the parameter A and the CTF are the function of the parameter b, only the b need be adjusted during the statistical process, so that the calculated amount can be economized. The improved fitting method is much more sutable for realtime statistic.The improved optimization Z-I fitting method is adopted in the statistics of dynamic Z-I relationship, in comparison with non-linearization fitting method, offered by Matlab, logarithmic linearization fitting method, traditional optimation fitting method, and the fixed Z-I relationship. It concludes that the improved optimation Z-I fitting method is the best of all the method mentioned above.The improvement to the programme of rain estimation by radar:the estimating programme by the radra combined with rain gauge is adopted, the improved optimtion Z-I fitting method is used to calculate the fitting Z-I relationship. Based on the realtime Z-I relationship, the initial value field of the precipitation intensity is calculated through the precipitation echo of radar reflectivity. Joint with the realtime observational data of intensive rain-rate gauges, the correcting value field is calculated out by means of the distance weighing interpolation method of the freak function which is proposed by the author, so that the systematic error of the initial field is corrected. Finally, an accumulated rainfall field which is of much higher spatial and temporal resolution than the rain-rate gauges is obtained by the integration of certain temporal and spatial interval as we wanted towards the corrected field, so that the situation of the real-time rainfall distribution for certain researched area can be grasped more timely and precisely.In the end, several certain precipitation events are tested with the fixed and dynamic methods, the estimating errors of each methods compared with the real amount of the precipitation events are analyzed. It’s concluded that the dynamic Z-I relationship method is superior to the fixed Z-I relationship method.