Air temperature is an important indicator to analyze climate change in mountainous areas.ERA5 reanalysis air temperature data are important products that were widely used to analyze temperature change in mountainous a...Air temperature is an important indicator to analyze climate change in mountainous areas.ERA5 reanalysis air temperature data are important products that were widely used to analyze temperature change in mountainous areas.However,the reliability of ERA5 reanalysis air temperature over the Qilian Mountains(QLM)is unclear.In this study,we evaluated the reliability of ERA5 monthly averaged reanalysis 2 m air temperature data using the observations at 17 meteorological stations in the QLM from 1979 to 2017.The results showed that:ERA5 reanalysis monthly averaged air temperature data have a good applicability in the QLM in general(R2=0.99).ERA5 reanalysis temperature data overestimated the observed temperature in the QLM in general.Root mean square error(RMSE)increases with the increasing of elevation range,showing that the reliability of ERA5 reanalysis temperature data is worse in higher elevation than that in lower altitude.ERA5 reanalysis temperature can capture observational warming rates well.All the smallest warming rates of observational temperature and ERA5 reanalysis temperature are found in winter,with the warming rates of 0.393°C/10a and 0.360°C/10a,respectively.This study will provide a reference for the application of ERA5 reanalysis monthly averaged air temperature data at different elevation ranges in the Qilian Mountains.展开更多
提出一种高精度的ZWD模型(tianjin_zwd,TZ)。TZ基于2016-2018年逐小时气压分层的ERA5,欧洲中尺度气象预报中心第五代再分析产品数据,采用BP神经网络建立。然后,根据2019年的ERA5产品导出的ZWD对TZ模型进行了验证。结果表明:相比GPT3模型...提出一种高精度的ZWD模型(tianjin_zwd,TZ)。TZ基于2016-2018年逐小时气压分层的ERA5,欧洲中尺度气象预报中心第五代再分析产品数据,采用BP神经网络建立。然后,根据2019年的ERA5产品导出的ZWD对TZ模型进行了验证。结果表明:相比GPT3模型,TZ模型可提供更贴近真值的ZWD估值;并且,其RMSE由5.0 cm (GPT3)降至4.5 cm,表明10%的精度提升。上述结果表明TZ模型实现了更优的预测性能,该模型的构建策略可为全国其他地区的ZWD建模提供借鉴。展开更多
利用全国40个地面台站的观测资料对ERA5及ERA5-Land两种不同空间分辨率的再分析资料开展了地面风速误差评估研究,结果表明:ERA5和ERA5-Land资料多年平均风速偏差的平均值分别为0.08 m s^(−1)、-0.06 m s^(−1),偏差的最大值分别为0.46 m ...利用全国40个地面台站的观测资料对ERA5及ERA5-Land两种不同空间分辨率的再分析资料开展了地面风速误差评估研究,结果表明:ERA5和ERA5-Land资料多年平均风速偏差的平均值分别为0.08 m s^(−1)、-0.06 m s^(−1),偏差的最大值分别为0.46 m s^(−1)、-0.19 m s^(−1),相对偏差的平均值为4.4%、-2.0%,相对偏差的最大值分别为33.0%、-10.1%;月平均风速线性拟合方程的斜率分别为0.93、0.97,截距分别为0.29 m s^(−1)、0.02 m s^(−1),相关系数分别为0.98、0.99;月平均风速均方根误差的平均值分别为0.17 m s^(−1)、0.14 m s^(−1),均方根误差的最大值分别为0.49 m s^(−1)、0.22 m s^(−1),相对均方根误差的平均值为7.4%、5.7%,相对均方根误差的最大值分别为35.2%、13.3%。ERA5-Land高分辨率资料地面风速误差相对较低,有利于提高风能资源评估的准确性。展开更多
基金financially supported by the National Natural Science Foundation of China(No.41621001)。
文摘Air temperature is an important indicator to analyze climate change in mountainous areas.ERA5 reanalysis air temperature data are important products that were widely used to analyze temperature change in mountainous areas.However,the reliability of ERA5 reanalysis air temperature over the Qilian Mountains(QLM)is unclear.In this study,we evaluated the reliability of ERA5 monthly averaged reanalysis 2 m air temperature data using the observations at 17 meteorological stations in the QLM from 1979 to 2017.The results showed that:ERA5 reanalysis monthly averaged air temperature data have a good applicability in the QLM in general(R2=0.99).ERA5 reanalysis temperature data overestimated the observed temperature in the QLM in general.Root mean square error(RMSE)increases with the increasing of elevation range,showing that the reliability of ERA5 reanalysis temperature data is worse in higher elevation than that in lower altitude.ERA5 reanalysis temperature can capture observational warming rates well.All the smallest warming rates of observational temperature and ERA5 reanalysis temperature are found in winter,with the warming rates of 0.393°C/10a and 0.360°C/10a,respectively.This study will provide a reference for the application of ERA5 reanalysis monthly averaged air temperature data at different elevation ranges in the Qilian Mountains.
文摘提出一种高精度的ZWD模型(tianjin_zwd,TZ)。TZ基于2016-2018年逐小时气压分层的ERA5,欧洲中尺度气象预报中心第五代再分析产品数据,采用BP神经网络建立。然后,根据2019年的ERA5产品导出的ZWD对TZ模型进行了验证。结果表明:相比GPT3模型,TZ模型可提供更贴近真值的ZWD估值;并且,其RMSE由5.0 cm (GPT3)降至4.5 cm,表明10%的精度提升。上述结果表明TZ模型实现了更优的预测性能,该模型的构建策略可为全国其他地区的ZWD建模提供借鉴。
文摘利用全国40个地面台站的观测资料对ERA5及ERA5-Land两种不同空间分辨率的再分析资料开展了地面风速误差评估研究,结果表明:ERA5和ERA5-Land资料多年平均风速偏差的平均值分别为0.08 m s^(−1)、-0.06 m s^(−1),偏差的最大值分别为0.46 m s^(−1)、-0.19 m s^(−1),相对偏差的平均值为4.4%、-2.0%,相对偏差的最大值分别为33.0%、-10.1%;月平均风速线性拟合方程的斜率分别为0.93、0.97,截距分别为0.29 m s^(−1)、0.02 m s^(−1),相关系数分别为0.98、0.99;月平均风速均方根误差的平均值分别为0.17 m s^(−1)、0.14 m s^(−1),均方根误差的最大值分别为0.49 m s^(−1)、0.22 m s^(−1),相对均方根误差的平均值为7.4%、5.7%,相对均方根误差的最大值分别为35.2%、13.3%。ERA5-Land高分辨率资料地面风速误差相对较低,有利于提高风能资源评估的准确性。
