中国地区高时空分辨率地表变量再分析产品风速误差分析
Wind Speed Assessment of Surface Reanalysis Product with High Resolution in China

伴随全球气候变化研究的不断拓展，当前风速研究普遍采用气候再分析数据作为基础工具，本文以2009年中国大陆2000余个气象站点的逐小时风速观测值为基础数据，对1948~2018年间东亚区域地表气象要素的70年再分析产品(Eastern Asian Reanalysis for surface meteorological variables, EAR70)和ERA5数据进行了对比评估，考察了其在年、月、日和小时尺度上的风速模拟性能与误差分布。年尺度分析表明，2009年中国风速呈现“中部低、南北高”的格局，西北、东北及沿海地区风速较高，部分站点风速超过6 m/s，具备较大风能开发潜力，而四川盆地等区域风速偏小；月尺度分析表明，EAR70在中国东部存在系统性高估偏差，偏差集中于河北、河南、江苏一带，尤以冬季最显著，夏季偏差相对较小；日统计结果显示，EAR70的日最小风速模拟结果明显优于ERA5，而在日最大风速上，ERA5模拟值离散程度小，误差波动较小；小时尺度分析表明，EAR70在夜间至清晨风速模拟较为准确，日间风速模拟的稳定性以ERA5为优，ERA5覆盖多数区域，尤其是中东部地区，风速模拟精度较高，较EAR70误差更低。
With the continuous expansion of global climate change research, climate reanalysis data have become a fundamental tool in wind speed studies. This study evaluates the wind speed simulation performance and associated error distributions of the 70-year reanalysis product for surface meteorological variables over East Asia (EAR70, 1948~2018) and the ERA5 dataset, using hourly wind speed observations from over 2000 meteorological stations across mainland China in 2009 as reference data. At the annual scale, wind speeds in China exhibit a spatial pattern of “low in the central region and high in the north and south”, with relatively higher wind speeds in the northwest, northeast, and coastal regions. Some stations recorded wind speeds exceeding 6 m/s, indicating significant wind energy potential, whereas regions such as the Sichuan Basin experienced lower wind speeds. Monthly analysis shows a systematic overestimation by EAR70 in eastern China, with the largest positive biases observed in Hebei, Henan, and Jiangsu, especially in winter, while summer exhibits relatively smaller errors. Daily statistics indicate that EAR70 outperforms ERA5 in simulating daily minimum wind speeds, whereas ERA5 provides more consistent results with lower variability in daily maximum wind speed simulations. At the hourly scale, EAR70 demonstrates higher accuracy during nighttime and early morning hours, while ERA5 shows better stability and lower errors in simulating daytime wind speeds. Overall, ERA5 exhibits superior simulation accuracy across most regions, especially in central and eastern China, when compared to EAR70.