Improvement on the Method for Estimating Precipitable Water from Ground-based GPS
Improvement on the Method for Estimating Precipitable Water from Ground-based GPS
复制标题
DOI：
--
复制
发表时间：
2009
期刊：
影响因子：
--
通讯作者：
Feng Guang-liu
中科院分区：
文献类型：
--
作者：
Feng Guang-liu
在地基GPS估算可降水量(PW)的方法中,GPS PW的精度取决于天顶静水延迟(ZHD)模型和参数П。通过对Saastamoinen、Hopfield和Black模型计算的ZHD进行误差分析,发现模型化的ZHD与探空仪剖面计算的ZHD存在系统偏差,相当于GPS上的偏差为4～10 mm。 PW.Saastamoinen、Hopfield和Black模型的系统偏差可以通过回归方法修正模型系数来减小，对系统均方根误差影响很小。参数П是大气加权平均温度(Tm)的函数，Tm与地表温度(Ts)具有很强的线性相关性。根据Tm和Ts之间的高度相关性，基于9°无线电探空仪观测剖面，通过线性回归构建了局部Tm模型。观测站均方根误差为2.8 K，对应相对误差为1.0%。采用修正的ZHD模型和本地Tm模型，减小了GPS PW的系统偏差。与探空仪剖面计算的PW相比，修正Hopfield模型和本地Tm模型估算的GPS PW存在-1.6 mm的偏差，与辐射计剖面计算的PW相比，GPS PW的偏差较大为-1.2 毫米。
In the method for estimating precipitable water(PW) from ground-based GPS,the accuracy of GPS PW depends on the zenith hydrostatic delay(ZHD) model and the parameter П.Through the error analysis on the ZHD calculated by Saastamoinen,Hopfield,and Black models,it is found that the modeled ZHD has a system bias from the ZHD calculated from radiosonde profiles,corresponding to a bias of 4～10 mm on GPS PW.The system bias of Saastamoinen,Hopfield,and Black models can be reduced by correcting the model coefficients with a regression method,with few effects on the system root-mean-square error.The parameter П is a function of atmospheric weighted mean temperature(Tm),and Tm has a strong linear correlation with surface temperature(Ts).According to the high correlation between Tm and Ts,a local Tm model is constructed by a linear regression based on radiosonde observing profiles at 9 observatories,with a root-mean-square error of 2.8 K,corresponding to a relative error of 1.0%.Adopting the modified ZHD model and the local Tm model,the system bias of GPS PW is reduced.Compared with PW calculated from radiosonde profiles,the GPS PW estimated with the modified Hopfield model and the local Tm model has a bias of-1.6 mm,while compared with PW calculated from radiometer profiles,the bias of GPS PW is-1.2 mm.