LIN et al. (2026) Enhanced Retrieval of Raindrop Size Distribution by Integrating Disdrometer Observations and Atmospheric Variables With Machine Learning

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Geophysical Research Letters
• Year: 2026
• Date: 2026-10-06
• Authors: LiMin LIN, Huiling Yuan, Xuwei Bao, Shangshang Yang, Lin Deng, Shuai Zhang, Yuanming Zhao
• DOI: 10.1029/2026gl123664

Research Groups

• Shanghai Meteorological Service Center
• Department of Atmospheric Sciences, Fudan University
• National Key Laboratory of Atmospheric Science and Geophysical Fluid Dynamics

Short Summary

This study develops a machine learning framework that integrates disdrometer observations with environmental atmospheric variables to retrieve raindrop size distribution (DSD) parameters, outperforming conventional methods in capturing DSD variability.

Objective

• Develop a data-driven approach for retrieving DSD parameters using machine learning techniques.

Study Configuration

• Spatial Scale: Urban and regional scale over Shanghai, China.
• Temporal Scale: 2019–2025, including training (2019–2021) and validation periods (2022–2025).

Methodology and Data

• Models used: Three machine learning models: Random Forest, Support Vector Machine, and Artificial Neural Network.
• Data sources: Disdrometer observations from 11 stations in Shanghai during 2019–2021, and additional data for independent validation.

Main Results

• The developed ML framework effectively reduces biases and captures observed DSD variability, outperforming conventional methods.
• Reliable DSD and rain rate estimates are provided by the trained models, even under heavy-rainfall and tropical cyclone conditions.

Contributions

• This study offers a data-driven pathway to improve DSD parameterizations in numerical models, leveraging environmental factors for more accurate retrievals.

Funding

• Shanghai Meteorological Service Center (project code: SMSC-2020-01)
• National Natural Science Foundation of China (grant number: 41975055)

Citation

@article{LIN2026Enhanced,
  author = {LIN, LiMin and Yuan, Huiling and Bao, Xuwei and Yang, Shangshang and Deng, Lin and Zhang, Shuai and Zhao, Yuanming},
  title = {Enhanced Retrieval of Raindrop Size Distribution by Integrating Disdrometer Observations and Atmospheric Variables With Machine Learning},
  journal = {Geophysical Research Letters},
  year = {2026},
  doi = {10.1029/2026gl123664},
  url = {https://doi.org/10.1029/2026gl123664}
}