The Structural Characteristics of Precipitation Cases Detected by Dual-Frequency Radar of GPM Satellite

The GPM (Global Precipitation Measurement) satellite carrying the first Ka-and Ku-band precipitation radar was launched in February 2014.GPM realized the DPR (Dual-frequency Precipitation Radar) detection following TRMM's (Tropical Rainfall Measurement Mission) PR (Precipitation Radar).Based on the four precipitation products provided by DPR including KaHS (Ka-band high-sensitivity product), KuPR (Ku-band product), KaMS (Ka-band matched product), and DPR_MS (dual-frequency product), structural features of four selected precipitation cases occurred in 2014 are analyzed.The structural differences caused by detecting frequencies, scan modes, and retrieval methods are then compared and discussed.The results show that the four selected precipitation cases occurred in East China, the northwestern Pacific, the storm track area, and the United States, respectively.Although the background environment and precipitation characteristics of the four cases are significantly different, some consistent differences among these products are revealed by the study of these cases.KaHS presents the highest storm top height, which is 0.1 km higher than that from KuPR.KaHS shows great advantages in observing precipitations weaker than 0.5 mm h^-1.However, KaHS underestimates precipitation intensities larger than 10 mm h^-1.KuPR inherits the good performance of TRMM PR for intense precipitation observations, and presents the largest proportion of precipitation higher than 10 mm h^-1.But KuPR is not suitable for observing precipitations weaker than 0.5 mm h^-1 due to its operating frequency.KaMS's near surface rain rate is similar to that of KaMS except that KaMS omits large number of weak precipitation (limited by the minimum echo threshold).Thus KaMS is not good at observing both intense and weak precipitation and not appropriate to be used independently.The precipitation retrieval algorithm of DPR_MS is highly self-governed and performs well in retrieving both weak and intense precipitations, while DPR_MS's storm top height strongly relies on KuPR's with the correlation coefficient close to 1.In addition, the droplet size distribution retrieved by DPR_MS is most reasonable and can well reveal the nonuniformity on the two sides of the typhoon eye wall.