In this study, Gaussian mixture model clustering analysis was carried out to examine characteristics of Global Precipitation Measurement (GPM) Dual‐frequency Precipitation Radar (DPR)‐retrieved mass‐weighted mean diameter (D_m), and normalized intercept parameter (N_w) of the drop size distribution (DSD) for heavy rainfalls (>10 mm h⁻¹) for 6 years (2014–2019). Three objective DSD types – continental, oceanic deep, and oceanic shallow convective types – emerged. The means and standard deviations of D_m and N_w obtained for the three types are in good agreement with various ground‐based observations, indicating that global view of DSD characteristics can be obtained from DPR‐derived DSD parameters. Global distributions of occurrence and contribution of each DSD type to total heavy rainfall are produced for the first time, which will help examine the dominant DSD type, its contribution to total heavy rainfall, and composition of different convective types in the rainfall system at a given location.

中文翻译：

---

从GPM / DPR测量中代表暴雨的三种类型的液滴尺寸分布的全局分布

在本研究中，进行了高斯混合模型聚类分析，以检验全球降水测量（GPM）双频降水雷达（DPR）提取的质量加权平均直径（D _m）和归一化拦截参数（N _w）的特征。6年（2014-2019年）强降雨（> 10 mm h ^-1）的液滴尺寸分布（DSD ）的变化。出现了三种客观的DSD类型：大陆，海洋深层和海洋浅层对流类型。D _m和N _w的均值和标准偏差从这三种类型获得的DSD与各种地面观测值都非常吻合，表明可以从DPR派生的DSD参数获得DSD特性的全局视图。首次生成每种DSD类型在总强降雨中的发生和贡献的全球分布，这将有助于检查DSD的主要类型，其对总强降雨的贡献以及给定降雨系统中不同对流类型的组成位置。