Scholarlink

<p>Meso- and micro-scale weather processes—such as cold air pools (CAPs) and nocturnal warming (NW)—exert significant influence on near-surface air temperatures in mountainous regions. However, they are often overlooked in conventional statistical downscaling methods because they operate at scales finer than the resolution of typical reanalysis or model output. This study develops an interpretable XGBoost-based framework that explicitly incorporates these processes to improve hourly temperature downscaling of ERA5 reanalysis data. Using observational data from 16 meteorological stations across ~ 20&#xa0;km<sup>2</sup> in the Zhangjiakou Competition Zone of the 2022 Beijing Winter Olympic Games during the 2017–2022 snow seasons, we analyze the spatial heterogeneity of 2-meter temperatures and the corresponding ERA5 biases, identifying key factors contributing to these differences. Multiple downscaling schemes are evaluated, comparing direct temperature prediction with bias-based modeling and assessing the impact of predictors associated with CAP and NW formation (e.g., radiative cooling, wind shear, and temperature advection). Results indicate that CAPs and NW occur on approximately 70% of nights, co-occurring in 57% of cases. Bias-based models consistently outperform direct downscaling approaches, particularly when meso- and micro-scale predictors are included. The best-performing model (Scheme D) reduces RMSE by up to 70% on mountain peaks, 40% in valleys, and 49% on slopes compared to raw ERA5 data, and achieves 4–9% improvements over conventional bias-based XGBoost models during NW-CAP nights. SHAP analysis improves model interpretability by linking predictors to the underlying physical processes and quantifying both the direction and magnitude of each predictor’s contribution to the model output. It reveals that bias-based modeling reduces over-reliance on ERA5 2&#xa0;m temperature, whose influence differs markedly between daytime and nighttime. By leveraging high-density station observations, the model captures meso- and micro-scale processes explicitly; because it requires only ERA5-derived predictors for application, the framework is readily transferable to other mountainous regions. This study underscores the importance of incorporating meso- and micro-scale dynamics in machine learning-based downscaling and offers practical insights for improving high-resolution temperature forecasting in complex terrain.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Improving ERA5 temperature downscaling over complex terrain with interpretable ML: incorporating cold air pools and nocturnal warming

  • 期刊论文

  • 发表日期：

    2026年6月22日

查看全文
  • Tianyu Yue,
  • Shuiqing Yin,
  • Xin Wang,
  • Jing Liu,
  • Hao Wang,
  • Xia Chen,
  • Deliang Chen

摘要

Meso- and micro-scale weather processes—such as cold air pools (CAPs) and nocturnal warming (NW)—exert significant influence on near-surface air temperatures in mountainous regions. However, they are often overlooked in conventional statistical downscaling methods because they operate at scales finer than the resolution of typical reanalysis or model output. This study develops an interpretable XGBoost-based framework that explicitly incorporates these processes to improve hourly temperature downscaling of ERA5 reanalysis data. Using observational data from 16 meteorological stations across ~ 20 km2 in the Zhangjiakou Competition Zone of the 2022 Beijing Winter Olympic Games during the 2017–2022 snow seasons, we analyze the spatial heterogeneity of 2-meter temperatures and the corresponding ERA5 biases, identifying key factors contributing to these differences. Multiple downscaling schemes are evaluated, comparing direct temperature prediction with bias-based modeling and assessing the impact of predictors associated with CAP and NW formation (e.g., radiative cooling, wind shear, and temperature advection). Results indicate that CAPs and NW occur on approximately 70% of nights, co-occurring in 57% of cases. Bias-based models consistently outperform direct downscaling approaches, particularly when meso- and micro-scale predictors are included. The best-performing model (Scheme D) reduces RMSE by up to 70% on mountain peaks, 40% in valleys, and 49% on slopes compared to raw ERA5 data, and achieves 4–9% improvements over conventional bias-based XGBoost models during NW-CAP nights. SHAP analysis improves model interpretability by linking predictors to the underlying physical processes and quantifying both the direction and magnitude of each predictor’s contribution to the model output. It reveals that bias-based modeling reduces over-reliance on ERA5 2 m temperature, whose influence differs markedly between daytime and nighttime. By leveraging high-density station observations, the model captures meso- and micro-scale processes explicitly; because it requires only ERA5-derived predictors for application, the framework is readily transferable to other mountainous regions. This study underscores the importance of incorporating meso- and micro-scale dynamics in machine learning-based downscaling and offers practical insights for improving high-resolution temperature forecasting in complex terrain.