Song et al. (2025) Indirect assimilation of radar reflectivity data with an adaptive hydrometer retrieval scheme for severe short-term weather forecasts

Identification

• Journal: Natural hazards and earth system sciences
• Year: 2025
• Date: 2025-10-13
• Authors: Lixin Song, Feifei Shen, Zhixin He, Yang Lu, Dongmei Xu, Aiqing Shu, Jiajun Chen
• DOI: 10.5194/nhess-25-3905-2025

Research Groups

• Key Laboratory of Meteorological Disaster, Ministry of Education (KLME)/Joint International Research Laboratory of Climate and Environment Change (ILCEC)/Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters (CIC-FEMD), Nanjing University of Information Science & Technology, Nanjing, China
• Key Laboratory of Urban Meteorology, China Meteorological Administration, Beijing, China
• China Meteorological Administration Tornado Key Laboratory, Guangzhou, China
• Anhui Meteorological Observatory, Hefei, China

Short Summary

This study develops an adaptive blending scheme for hydrometeor retrieval in indirect radar reflectivity assimilation, combining temperature-based and background hydrometeor-dependent methods. It demonstrates that this new scheme improves the accuracy of short-term severe weather forecasts by enhancing hydrometeor distributions and thermodynamic/dynamic structures.

Objective

• To design and evaluate an adaptive blending scheme that combines "temperature-based" and "background hydrometer-dependent" methods for hydrometeor retrieval in indirect assimilation of radar reflectivity data, aiming to improve severe short-term weather forecasts.

Study Configuration

• Spatial Scale:
  • Model domain for Case 1: 500 × 471 grid points with a horizontal resolution of 3 km and 50 vertical levels.
  • Model domain for Case 2: 723 × 691 grid points with a horizontal resolution of 3 km and 50 vertical levels.
  • Radar maximum coverage range: 230 km.
• Temporal Scale:
  • Case 1: 14 June 2020 – 15 June 2020, with forecasts up to 3 hours.
  • Case 2: 6 August 2018 – 7 August 2018, with forecasts up to 6 hours.

Methodology and Data

• Models used:
  • Weather Research and Forecasting (WRF) model v4.3
  • WRF Data Assimilation (WRFDA) system v4.3 (3DVar)
  • Microphysics: WRF Double-Moment 6-Class Microphysics (WDM6) scheme
  • Radiation: Rapid Radiative Transfer Model (RRTM) longwave, Dudhia shortwave
  • Boundary Layer: Yonsei University (YSU) scheme
  • Land Surface: Noah land surface model
• Data sources:
  • Radar observations: S-band Doppler radars (Nanjing radar for Case 1, Shenyang radar for Case 2). Radial velocity range resolution: 250 m, observation error: 2 m/s. Reflectivity range resolution: 1000 m, observation error: 5 dBZ.
  • Initial and lateral boundary conditions: NCEP Global Forecast System (GFS) data.
  • Precipitation observations: China Meteorological Administration (CLDAS-V2.0).
  • Hydrometeor classification algorithm (HCA) based on dual-polarization radar observations (for comparison).

Main Results

• The developed blending scheme (EXP_temp-bg) adaptively combines "temperature-based" and "background hydrometer-dependent" hydrometeor retrieval methods, effectively reducing errors from fixed relationships and background field uncertainties.
• For the 14 June 2020 case, EXP_temp-bg's retrieved wet snow and graupel distributions showed better consistency with dual-polarization radar HCA results compared to individual schemes.
• The blending scheme consistently improved radar reflectivity forecasts, leading to better organization of convective structures and more accurate predictions of precipitation intensity for both case studies.
• EXP_temp-bg produced more consistent thermal and dynamical structures, including a robust and more extensive/deeper updraft column, which are crucial for severe convective weather development.
• Quantitative assessment using Equitable Threat Scores (ETS) demonstrated that EXPtemp-bg consistently outperformed both "temperature-based" (EXPtemp) and "background hydrometer-dependent" (EXP_bg) schemes across various precipitation thresholds and forecast lead times.
• The blending method effectively suppressed spurious precipitation areas and showed superior skill in predicting heavy rainfall areas.

Contributions

• Development of a novel adaptive blending scheme for hydrometeor retrieval in indirect radar reflectivity assimilation, which intelligently combines the strengths of existing "temperature-based" and "background hydrometer-dependent" methods.
• Demonstrated significant improvements in the accuracy of hydrometeor distributions, thermodynamic and dynamic structures, and consequently, the short-term forecasts of radar reflectivity and precipitation intensity for severe convective events.
• Offers a more robust and less uncertain approach to radar reflectivity assimilation, addressing limitations of methods that rely solely on fixed empirical relationships or potentially inaccurate background fields.

Funding

• National Key R&D Program of China (grant no. 2024YFC2815702)
• Key Laboratory of Urban Meteorology, China Meteorological Administration, Beijing (grant no. LUM-2025-02)
• China Meteorological Administration Tornado Key Laboratory (grant no. TKL202306)
• Natural Science Fund of Anhui Province of China (grant no. 2308085MD127)
• Shanghai Typhoon Research Foundation (grant no. TFJJ202107)
• Chinese National Natural Science Foundation of China (grant no. G41805070)

Citation

@article{Song2025Indirect,
  author = {Song, Lixin and Shen, Feifei and He, Zhixin and Lu, Yang and Xu, Dongmei and Shu, Aiqing and Chen, Jiajun},
  title = {Indirect assimilation of radar reflectivity data with an adaptive hydrometer retrieval scheme for severe short-term weather forecasts},
  journal = {Natural hazards and earth system sciences},
  year = {2025},
  doi = {10.5194/nhess-25-3905-2025},
  url = {https://doi.org/10.5194/nhess-25-3905-2025}
}