桂林地区一次暴雨过程ERA5/MERRA-2资料计算PWV的精度分析
Precision analysis of PWV calculated from ERA5/MERRA-2 data during a rainstorm process in Guilin
为了评估再分析资料在特定地区和特定天气条件下反演大气水汽(precipitable water vapor,PWV)的性能,利用2015-2017年桂林探空站观测数据建立桂林地区大气加权平均温度模型,并用该模型计算出的连续运行参考站(continuously operating reference stations,CORS)大气水汽(CORS-PWV)为参考值,与ERA5和MERRA-2这2种再分析资料提取的PWV作比较。结果表明:新建立的桂林地区大气加权平均温度模型的精度比Bevis模型提高了8.4%,比广西模型提高了13.4%,比中国东部模型提高了17.4%;在暴雨过程中,以2017年年积日为178-183 CORS站反演的PWV为参考值, ERA5提取PWV的平均bias和RMSE分别为1.55 mm和3.38 mm,而MERRA-2提取PWV则表现出较大的平均bias和RMSE,分别为4.60 mm,6.18 mm,并且ERA5提取的PWV与CORS站反演的PWV具有更好的一致性,与地面实际降水情况相符。
In order to evaluate the performance of precipitable water vapor (PWV) inversion from reanalysis data in specific areas and under specific weather conditions, an atmospheric weighted mean temperature model was established based on the observation data of Guilin radiosonde stations from 2015 to 2017. The continuous operating reference stations (CORS) atmospheric water vapor (CORS-PWV) calculated by the model was used as the reference value, and was compared with the PWV extracted from the two reanalysis data ERA5 and MERRA-2.. The results show that compared with Bevis model, Guangxi model and Eastern China model, the accuracy of the new model is improved by 8.4%, 13.4% and 17.4%, respectively. In the course of a rainstorm, taking the CORS-PWV of day of year from 178-183 in 2017 as reference value, the mean bias and RMSE of PWV extracted by ERA5 are 1.55 mm and 3.38 mm, respectively; while the mean bias and RMSE of PWV extracted by MERRA-2 are 4.60 mm and 6.18 mm. The PWV extracted by ERA5 has a better consistency with the PWV retrieved by CORS station, which is consistent with the actual ground precipitation situation.
曾印、吴勇锋、张继洪、谢劭峰
大气科学(气象学)
加权平均温度ERA5MERRA-2大气水汽暴雨
weighted mean temperatureERA5MERRA-2precipitable water vaporrainstorm
曾印,吴勇锋,张继洪,谢劭峰.桂林地区一次暴雨过程ERA5/MERRA-2资料计算PWV的精度分析[EB/OL].(2023-01-18)[2025-08-18].http://www.paper.edu.cn/releasepaper/content/202301-86.点此复制
评论