An Enhanced Hydrometeor Detection Algorithm for Precipitation Identification in GPM

Precipitation is a fundamental component of the climate system, linking the water and energy cycles and regulating climate variability, feedbacks, and extremes. Accurate precipitation detection provides essential constraints on the global water cycles and serves as a key benchmark for evaluating climate model performance. The Global Precipitation Measurement (GPM) mission, equipped with the first spaceborne dual-frequency radar (DPR), offers unprecedented capability for observing the three-dimensional structure of global precipitation, including phase, intensity and microphysical properties. However, existing precipitation detection algorithm does not explicitly account for weak radar signals, limiting their ability to capture the early development and full evolution of precipitation systems. In this study, we propose an enhanced hydrometeor detection algorithm for GPM DPR observations that combines a bilateral filter technique and a power-probability weighting scheme. Through effectively suppressing background noise while preserving physically meaningful weak signals, our method can significantly improve the detection of small hydrometeors at high altitudes and along precipitation boundaries—regions that are frequently missed by the current GPM operational algorithm. Validation using collocated CloudSat observations demonstrates substantial improvements of our method, with increases of 63.6% the Ka band and a 37.5% in the Ku band. These advancements provide a more comprehensive depiction of precipitation systems structure and offer improved observational support for enhancing precipitation estimation and climate modeling.