Comparison of Dual-Polarization Radar Estimators of Rain

Abstract

Several polarimetric radar estimators of rain rate R and rainwater content M are examined. The accuracy of the estimators is analyzed using a gamma drop size distribution (DSD) simulation and a radar wavelength of 11 cm. The estimators that use combined measurements of specific differential phase KDP and differential reflectivity ZDR are superior to the estimators of R and M obtained from reflectivity factors Zh and Zv at orthogonal polarizations or KDP only. The standard deviation of the R(KDP, ZDR) estimate is 2?3 times less than for the best of the R(Zh, Zv) estimators. The statistical accuracy of the M(KDP, ZDR) estimator is at least 1.5?2 times better than for the M(Zh, Zv) estimator. Cumulative rainfalls obtained with the estimators are compared with accumulations recorded with 42 rain gauges in the Little Washita River basin. Biases, errors, and reasons for the superior performance of the R(KDP, ZDR) estimator are explained in terms of the microphysical processes in the squall line that contributed to the rainfall.