基于ECMWF再分析地表资料计算中国区域ZTD的精度分析

利用ECMWF再分析地表资料，结合GPT2w模型提供的水汽递减率和温度递减率计算中国区域对流层延迟值的精度。首先，以中国地区75个探空站2015年地表实测气象参数为参考值，利用ECMWF地表资料得到的气象参数（P，T，e）的精度分别为1.76 hPa、1.96 K、1.98 hPa。然后，以相同测站2010~2015年探空站分层数据算得的ZTD为参考值，对ECMWF地表资料计算的ZTD的精度进行分析，并与利用探空仪地面观测数据为输入参数计算的ZTD的精度进行对比。结果显示，利用ECMWF地表资料计算的ZTD的平均bias为0.07 cm，平均RMS为3.72 cm，在低纬度地区优于利用探空仪地面观测数据为输入参数计算的ZTD的结果。以陆态网237个GNSS测站2015年的ZTD作为参考值，比对利用ECMWF地表资料计算的ZTD的精度，结果为3.41 cm。由此可知，ECMWF地面资料计算的ZTD的精度能满足普通用户对流层延迟的计算需求，可用于缺少气象参数的测站进行对流层延迟值的计算及其他相关应用。

Analysis of the Accuracy of Zenith Tropospheric Delay Calculated by ECMWF Surface Data over China Area

We analyze the accuracy of the tropospheric delay in China calculated using ECMWF surface data with the water vapor decline rate and the temperature decline rate provided by the GPT2w model. At first,we analyze the accuracy of meteorological parameters(P,T,e) derived from ECMWF. The results show that the mean RMS are 1.76 hPa, 1.96 K, 1.98 hPa, respectively. Then, we compare the ZTD calculated from ECMWF surface data with bench mark values derived from 6 years(2010 to 2015) of radiosonde data recorded at 75 radiosonde stations in China. To make the result more credible, we also analyze the accuracy of ZTD calculated from radiosonde surface data. The results show that the mean bias value of ZTD derived from ECMWF is 0.07 cm and the mean RMS is 3.72 cm, which is even smaller than radiosonde in the low latitude region. We also assess the accuracy of the ZTD calculated from ECMWF surface data using ZTD time series derived from GNSS stations of CMONOC in 2015. The result shows an average accuracy of 3.41 cm. Through the research, we can draw a conclusion that the accuracy of ECMWF surface data is high enough to be used in the calculating ZTD of stations in China where the meteorological parameters are unavailable and for related GNSS applications.