Improved Estimation of Backscattering Differential Phase in Rain and Its Utilization in Rainfall Estimation

In recent years, the rainfall estimator that combines the specific attenuation A and the specific differential phase ${K} _{\text {DP}}$ for X-band radar has been concerned and developed. However, the constraints of empirical coefficients and insufficient resolution of A and ${K} _{\text {DP}}$ estimates, as well as the uncertainties caused by the unknown shapes of raindrops, pose challenges to the estimator in maintaining accuracy of rainfall estimates. The high correlation between the differential reflectivity ${Z} _{\text {DR}}$ and the raindrop shape helps to mitigate the uncertainties associated with the variations of drop size distribution (DSD) and the unknown shapes of raindrops. However, as a power measurement, ${Z} _{\text {DR}}$ is inevitably affected by radar miscalibration, partial beam blockage (PBB), and bias from wet radome, which hinders its application for rainfall estimation. The backscattering differential phase $\delta _{\text {hv}}$ is also strongly dependent on raindrop shape and is not affected by the above negative factors, so it has the potential to be the substitute for ${Z} _{\text {DR}}$ . Unfortunately, reliable method for estimating $\delta _{\text {hv}}$ in rain is currently lacking. This article reviews an adaptive and high-resolution (HR) method for estimating A and ${K} _{\text {DP}}$ called adaptive and high-resolution empirical coefficient conditioning (AHRCC), and based on the outputs of AHRCC, proposes a method for estimating $\delta _{\text {hv}}$ accurately, which mainly reduces the cumulative bias caused by path integral. In addition, an algorithm for rainfall estimation based on A, ${K} _{\text {DP}}$ , and $\delta _{\text {hv}}$ is proposed to reduce the overestimation of rainfall caused by DSD variations and raindrop shape uncertainties, and the potential of retrieving characteristic raindrop sizes by $\delta _{\text {hv}}$ is also explored.