Research On Key Parameters Of Water Vapor Detection Based On Ground-based GNSS

Ground-based GNSS is relatively new technique for retrieving atmospheric water vapor in the field of GNSS remote sensing applications and meteorology.The research mainly focused on the key parameters of water vapor detection based on ground-based GNSS,analysis the determination of zenith hydrostatic delay(ZHD)without collocated meteorological data,the spatial-temporal variations of atmospheric weighted mean temperature(T_m)and its modeling methods.A ground-based GNSS water vapor processing software is built,which was used to obtain GNSS water vapor over the Xuzhou region.The main research contents and results are as follows;1)From the aspects of spatial distribution and seasonal changes,the ERA5 hourly pressure data and the GPT3(global pressure and temperature 3)model are used to solve ZHD in China and the accuracies are discussed.The results shows that the overall accuracy of ZHD derived from ERA5 data is 1.3 mm,of which the accuracy is slightly worse in parts of the Qinghai-Tibet and the northwest,and there is no obvious seasonal difference;the overall accuracy of ZHD derived from the GPT3 model is 10.3mm,and its accuracy is mainly affected by latitude and the influence of seasonal changes;the accuracy will decrease when latitude increases,and there is a difference of 1-3mm in different seasons.2)The characteristics of the spatial-temporal changes of T_m in China are analyzed,and it is found that the spatial-temporal changes are caused by seasonal,latitude,altitude and climate factors.According to spatial-temporal characteristics of T_m,based on the there-year observation data of 89 radiosonde stations in China from 2016 to 2018,a geographical zoning model of T_m in China was established.The result shows that the overall accuracy of the model in China is about 3.52K.And the accuracies ware 2.8K,4.0K,4.2K,3.5K in the south,north,northwest and Qinghai-Tibet,respectively,which are 11%,7%,17%and 44%higher than the Bevis model,especially in the Qinghai-Tibet area.3)A GNSS water vapor processing software is built,which can automatically process multiple days in batches.It was used to obtain the GNSS water vapor over the Xuzhou region.The sounding data is as a reference,the accuracies of GNSS-PWV based on ERA5 hourly pressure data and GPT3 model are 1.6mm and 3.2mm,respectively,which can be satisfied the accuracy threshold for long-term climate change research and real-time water vapor detection.Finally,the analysis of the temporal characteristics of PWV before and after a rainfall event in the Xuzhou region indicates that PWV will continue to grow for a long period of time before the rain,and the rainfall can be predicted according to the continuous growth time of PWV.