Currently, major challenges exist in the area of the quantitative precipitation estimations of convective short-term flooding caused by local heavy rainfall events. In this study, in order to improve the accuracy of the forecasting of the convective precipitation estimates, various currently used quantitative measurements methods for precipitation were summarized. Then, the limitations of these statistical methods for rainfall estimations were analyzed. It is known that different types of precipitation affect the accuracy of the precipitation estimates due to the variations in power, heat, and water vapor, which result in different micro-physical mechanisms and internal structures. Therefore, based on the physical mechanisms of the precipitation, along with the precipitation echo structures, the coefficient A and exponent b in the Z-I relationship become spatial functions rather than simple constants. Therefore, they can provide a theoretical basis for an adaptive method which can be effectively used to quantitatively estimate precipitation.
In this research study, the mechanisms of convective precipitation, mixed precipitation, and stratiform precipitation were examined, and the echo structural features, precipitation micro-physical characteristics, and echo extinction laws were analyzed. It is known that multiple parameters have certain impacts on precipitation, such as composite reflectivity factors (CR) and vertical liquid water content (VIL). Therefore, it was necessary to first distinguish the three different types of precipitation from the structure, and then use statistical methods for the fitting process. The lgR_C-lgI and R_C-I data pairs were respectively utilized. Then, depending on the type of rain event and the corresponding Z-I relationship, the precipitation measurement errors were reduced which had resulted from the unstable Z-I relationship caused by changes in the rain events, in order to solve the problem of the serious underestimations of heavy rainfall events. The results of this study showed that the lgR_C-lgI data pairs displayed better fitting results when fit by order, and R_C-I data pairs exhibited better results when fit by high-order. These findings indicated that the adaptive method which was used to quantitative estimate the precipitation was reasonable. Also, the instability of the Z-I relationship caused by the physical properties and structural characteristics of the precipitation was revealed.
This study also investigated the binary function relationships among the CR, VIL, and precipitation of convective and mixed precipitation. It was observed that there was a high correlation between the fitting results. The results confirmed that by taking the CR and VIL into account, the issue of significantly underestimating heavy rainfall events could be resolved.