Tang et al. (2026) A profiled hydrometeor classification algorithm for dual-frequency precipitation radar onboard GPM core observatory

Identification

• Journal: Atmospheric Research
• Year: 2026
• Date: 2026-01-17
• Authors: Jiajia Tang, Xiong Hu, Weihua Ai, Shensen Hu, X. B. Zhao, Li Wang, Zhonghui Tan
• DOI: 10.1016/j.atmosres.2026.108785

Research Groups

• College of Meteorology and Oceanography, National University of Defense Technology, Changsha, China
• College of Basic Education, National University of Defense Technology, Changsha, China

Short Summary

This study developed and evaluated a profiled hydrometeor classification algorithm (HCA) for the GPM Dual-frequency Precipitation Radar (DPR) using ground-based radar retrievals. The GPM DPR-HCA demonstrated good agreement with ground-based radar classifications, achieving high Probability of Detection (POD) for liquid and mixed-phase hydrometeors, and showed promising regional application for spaceborne precipitation radar products.

Objective

• To develop and evaluate a profiled hydrometeor classification algorithm (HCA) for the Global Precipitation Measurement (GPM) mission's Dual-frequency Precipitation Radar (DPR) using collocated ground-based dual-polarization radar data.

Study Configuration

• Spatial Scale: Guangzhou (GZ) for ground-based radar data; regional application over the Yangtze-Huai River Valley Region and South China; GPM DPR provides quasi-global coverage.
• Temporal Scale: Event-based for algorithm evaluation (a precipitation case); regional application for hydrometeor distribution analysis. Specific dates for data acquisition are not provided.

Methodology and Data

• Models used: Centroid-based classification algorithm using Euclidean distance.
• Data sources:
  • GPM DPR: Ku-band reflectivity, Ka-band reflectivity, dual-frequency ratio, and air temperature.
  • Ground-based dual-polarization radar in Guangzhou (GZ): Used to generate a labeled dataset for training and for evaluation of the GPM DPR-HCA.

Main Results

• The retrieved hydrometeor profiles from the GPM DPR-HCA showed good agreement with those from the ground-based GZ radar.
• Probability of Detection (POD) exceeded 0.75 for liquid- and mixed-phase hydrometeors.
• PODs were 0.66 for hail and 0.43 for low-density graupel among solid-phase classes, attributed to substantial feature overlaps.
• Regional application revealed that wet snow, graupel, and ice crystals predominantly occurred in the mid- and upper levels during stratiform events.
• Graupel and hail were more frequently observed in convective events.

Contributions

• Developed a novel profiled hydrometeor classification algorithm (HCA) specifically for the GPM DPR, which lacks dual-polarimetric capabilities.
• Utilized ground-based dual-polarization radar data to train and validate a spaceborne radar HCA, bridging the gap between ground-based and spaceborne observations.
• Demonstrated the accuracy and reliability of the GPM DPR-HCA, positioning it as a strong candidate for the development of future hydrometeor classification products from spaceborne precipitation radars.
• Provided new insights into the vertical distribution and types of hydrometeors during stratiform and convective precipitation events in the Yangtze-Huai River Valley Region and South China.

Funding

• Not specified in the provided text.

Citation

@article{Tang2026profiled,
  author = {Tang, Jiajia and Hu, Xiong and Ai, Weihua and Hu, Shensen and Zhao, X. B. and Wang, Li and Tan, Zhonghui},
  title = {A profiled hydrometeor classification algorithm for dual-frequency precipitation radar onboard GPM core observatory},
  journal = {Atmospheric Research},
  year = {2026},
  doi = {10.1016/j.atmosres.2026.108785},
  url = {https://doi.org/10.1016/j.atmosres.2026.108785}
}