The characteristics of rain drop size distribution (DSD) in super typhoons are critical for the application of radar-retrieved precipitation to typhoon short-term forecasting. Based on disdrometer data collected at 12 national surface stations in Hainan, this study comprehensively analyzes the characteristics of DSD within the eyewall, inner rainbands, and outer rainbands of super typhoon "Yagi", and further investigates their application potential in quantitative precipitation estimation (QPE) using weather radar. The results indicate that the normalized Γ distribution can effectively characterize DSD features across distinct structural regions of the typhoon, while the mean precipitation spectrum in each region exhibits a distinct unimodal pattern. The precipitation formation mechanism of super typhoon "Yagi" is dominated by warm-cloud and ice-phase mixed microphysical processes, with medium-size raindrops making the dominant contribution to the rainfall intensity across all typhoon structural regions. Compared with typhoons of similar intensity, super typhoon "Yagi" exhibits smaller raindrop particles in its eyewall region. These raindrops make a greater contribution to the total rainfall intensity and lower the coefficient A in the corresponding Z-R relationship, which further promotes the formation of intense precipitation. Applying the fitted local parameters to QPE using the new-generation weather radar in Haikou yields better performance than the operational default scheme: The correlation coefficient reaches 0.9264, the normalized mean absolute error (NMAE) and root mean square error (RMSE) decrease by 0.1024 and 0.9 mm, respectively, and the retrieval results are more consistent with actual observations after incorporating surface rain gauge correction.

Attenuation effect for spaceborne millimeter wave radar is severe. The microphysical properties of cloud and precipit- ation particles are quite different, and the attenuation effects caused by these different particles are different as well. Based on different cloud and precipitation scenarios simulated by the Weather Research and Forecasting model (WRF), the radar echoes after attenuation are obtained by a spaceborne millimeter wave radar simulator, and a sensitivity analysis of the attenuation characteristics of the particles is carried out to explain important factors that cause attenuation uncertainty. Results indicate that the simulated radar echoes after attenuation are in good agreement with the Cloud Profiling Radar (CPR) echoes. The echo structures and intensities are similar, and the differences in vertical profiles of the average reflectivity are within 20%. Through the sensitivity analysis of the attenuation coefficient (K), it is found that the attenuation caused by cloud and precipitation particles is most sensitive to water content (W) in the Particle Size Distribution (PSD), followed by the median volume diameter (D₀). The change of D₀ may cause the average relative error of the attenuation of rain to be around 25%. Snow and graupel are more affected by the density parameters (a and b) than by the slope parameter (Λ), and the uncertainty caused by a is approximately 25%. Under the same PSD parameters, the difference in attenuation coefficient caused by ice state and liquid state is approximately 90%. Considering the high sensitivity of attenuation to phase states, the K-W relationship in different phases is fitted for a cloud case measured by CPR. The attenuation correction result for the CPR is compared with the band correction result for the Dual-frequency Ka-band Precipitation Radar (DPR). It is found that the echo intensities of the two are quite close, and the uncertainty of the attenuation correction mainly comes from the difference between the inversion results of the liquid water content and the actual situation.