资源环境科技发展态势分析平台

Attenuation correction of measured reflectivity (Z(H)) and differential reflectivity (Z(DR)) is critical in short-wavelength radar applications, especially during extreme precipitation events such as heavy rain. This paper develops an improved self-consistent approach (improved ZPHI method) for attenuation correction in practical environment. In particular, a non-negative constraint is imposed on the specific attenuation, which is inferred from the monotonic increasing characteristic of differential propagation phase (Phi(DP)). The copolar correlation coefficient (rho(HV)) is used to partition the Phi(DP) profiles into independent range segments, which are featured by different hydrometeor phases such as liquid rain or mixed-phase precipitation. Additional minimization constraint is imposed on the cost function of the difference between preprocessed Phi(DP) and reconstructed Phi(DP )to ensure its appropriate convergence. In addition, an exponential Z(DR) - Z(H) relation derived from local raindrop size distribution (DSD) data is applied in the procedure of Z(DR) correction. Polarimetric measurements from a C-band radar (CPOL) in Hangzhou of China during two extreme precipitation events are used to demonstrate the improved attenuation correction method. DSD observations from disdrometers and radar data from a nearby system operating at non-attenuated frequency (i.e., S-band) at Huangshan Mountain are used to evaluate the attenuation correction performance. This paper also studies the impact of attenuation correction on the radar-derived quantitative precipitation estimation (QPE) product, through cross-comparison with rain gauge observations. Results show that the polarimetric observations from CPOL radar are effectively enhanced and more consistent with collocated S-band measurements and the simulated radar moments based on DSD data. Hourly rainfall products derived from R(Z(H)) and R(Z(H), Z(DR)) are significantly improved, which are comparable to R(K-DP) after attenuation correction.