Xu et al. (2025) Assessment of observation errors in AWS data assimilation: Application to a thunderstorm gale event forecast

Identification

• Journal: Atmospheric Research
• Year: 2025
• Date: 2025-11-24
• Authors: Dongmei Xu, Yi Liu, Jingyao Luo, Zhixin He, Yakai Guo, Feifei Shen
• DOI: 10.1016/j.atmosres.2025.108653

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 210044, China
• Key Laboratory of Urban Meteorology, China Meteorological Administration, Beijing 100089, China
• China Meteorological Administration Tornado Key Laboratory, Guangzhou 510641, China
• Shanghai Typhoon Institute, China Meteorological Administration, Shanghai 200030, China
• Anhui Meteorological Observatory, Hefei 230000, China
• China Meteorological Administration Henan Meteorological Bureau, Zhengzhou 450003, China
• Key Laboratory of Transportation Meteorology of China Meteorological Administration, Nanjing Joint Institute for Atmospheric Sciences, Nanjing 210041, China

Short Summary

This study assesses the impact of varying observation errors in Automatic Weather Station (AWS) data assimilation on the high-resolution simulation and forecast of a thunderstorm gale event in Beijing using the WRF model and 3DVAR system. It demonstrates that optimized observation error estimates significantly improve wind analysis and extreme wind speed forecasting, with a Desroziers-based method showing superior performance.

Objective

• To systematically assess and explore the impact of different observation errors in Automatic Weather Station (AWS) data assimilation on the performance of wind analysis and forecasting for a thunderstorm gale event.

Study Configuration

• Spatial Scale: High-resolution simulation of a thunderstorm gale event in Beijing.
• Temporal Scale: A specific thunderstorm gale event (duration not explicitly stated, but characteristic of short-lived severe convective weather).

Methodology and Data

• Models used: Weather Research and Forecasting (WRF) model, Three-dimensional Variational (3DVAR) data assimilation system.
• Data sources: Observations from Automatic Weather Stations (AWSs).

Main Results

• Varying observation errors significantly influence the performance of wind analysis and forecasts.
• Mitigating observation errors in cases with significant observation-minus-background (OMB) discrepancies enhances the reliability of observations, leading to greater analysis increments and improved agreement with assimilated observations.
• The Desroziers method, which computes observation error covariances through statistical analysis of OMB and observation-minus-analysis (OMA) residuals, assigns higher error estimates, allowing assimilation of more observations with reduced weights and enhancing spatial continuity.
• Experiments with modified observation error parameters (including the Desroziers method) demonstrated superior predictive accuracy compared to the default configuration, enabling more precise identification of strong wind regions and mitigating overestimation.
• Statistical metrics (BIAS, RMSE) and FSS scores confirmed the superior performance of optimized error estimates, particularly in extreme wind speed forecasting, with Exp_D05 outperforming all other experiments.

Contributions

• Systematically assesses the impact of observation error estimates from AWSs on high-resolution thunderstorm gale forecasts.
• Demonstrates that optimized observation error parameters, including those derived from the Desroziers method, significantly improve the accuracy of wind analysis and extreme wind speed forecasting.
• Provides evidence for the importance of refined observation error estimates in AWS data assimilation for enhancing numerical weather prediction of severe convective events.

Funding

• Not specified in the provided text.

Citation

@article{Xu2025Assessment,
  author = {Xu, Dongmei and Liu, Yi and Luo, Jingyao and He, Zhixin and Guo, Yakai and Shen, Feifei},
  title = {Assessment of observation errors in AWS data assimilation: Application to a thunderstorm gale event forecast},
  journal = {Atmospheric Research},
  year = {2025},
  doi = {10.1016/j.atmosres.2025.108653},
  url = {https://doi.org/10.1016/j.atmosres.2025.108653}
}