Study On Tropospheric Wet Delay Model And Application Of Water Vapor Inversion

Tropospheric wet delay(ZWD)is one of the factors affecting the positioning accuracy of GNSS navigation.At the same time,ZWD can also directly retrieve the atmospheric zenith precipitation(PWV).Because of its own advantages,the model modification method has become the current method to study deeply the tropospheric delay.However,the traditional ZWD model without meteorological parameters has a large amount of data.And it is not suitable for the vast and complicated areas of our country.Sounding data was used to obtain approximate true ZWD values in China,to analyze the spatiotemporal characteristics of ZWD in China,and to build a ZWD model of zenith tropospheric wet delay in China without meteorological parameters.,which was applied to atmospheric water vapor inversion in Kunming.Followings are the main research contents and conclusions:(1)Method for obtaining ZWD in China based on sounding dataStarting from the theory of atmospheric refraction,zenith tropospheric delay(ZTD)was obtained by sounding site data.Then with help of the tropospheric delay products provided by the GGOS Atmosphere analysis center,the accuracy of the zenith tropospheric dry delay(ZHD)models of several traditional models were compared.And the ZHD data of 4 domestic IGS((International GNSS Service)stations were combined to obtain the conclusion that the ZHD of the Saastamoinen model(S)value has the highest accuracy and can reach to a millimeter level.Using the relationship of ZWD=ZTD-ZHD,the ZWD value of the zenith tropospheric wet delay in China was obtained.The VMF1 grid tropospheric delay product was used to verify the ZWD value obtained in China through ZWD elevation naturalization and grid interpolation.The results were reliable and the accuracy of ZWD reached about 3.2mm.(2)Analysis of spatiotemporal characteristics of ZWD in ChinaAnalyze the relationship between the mean ZWD and the latitude,longitude and altitude of the sounding stations from 2013 to 2017,and compare and analyze the ZWD data of eight sounding stations in China.It is concluded that ZWD has a strong negative correlation with the latitude and altitude of the stations and has nothing to do with the longitude of the site.And the average annual ZWD and amplitude of the station decreases with increasing latitude ? and altitude h.After analyzing the ZWD time variation rules of 86 sounding stations from 2013 to 2017,it was found that the ZWD value is low in winter and high in summer,and the annual change is oscillating.The oscillation range is basically 0?400mm,and the geometry of the ZWD curve is It is closest to the "cosine function"or "quadratic function" curve.(3)Research on ZWD model in ChinaBased on the analysis of the spatiotemporal characteristics of ZWD in China,two types of ZWD models in China,CZWD(cosine function model)and QZWD(quadratic function model),with parameters of station latitude,altitude,and annual accumulated day as parameters were established by 2013-2017 data,and accuracy verification was performed using 2018 data.After comparative analysis of the models,the average deviation of the EGNOS-ZWD model(abbreviated as ZWD(E))from 2013 to 2017 was-51.1mm,indicating that the ZWD(E)model has systematic errors in the Chinese region.Secondly,the average median error of the CZWD model from 2013 to 2017 was ±38.3mm,while the error of the QZWD model was ± 32.7mm,which was better than the ZWD(E)model(± 56.3mm).In 2018,the accuracy of the CZWD model and QZWD model(± 38.7mm and ±32.1mm)was improved by about 27.8%and 40.1%compared to the ZWD(E)model(±53.6mm),respectively.The example results showed that the QZWD model in the form of quadratic function has better accuracy than the CZWD model.From the perspective of Harbin's single site,the main reason is that the QZWD model takes into account the sharp changes in ZWD during summer.(4)QZWD model applied to atmospheric water vapor inversionAfter analyzing the spatiotemporal characteristics of the regional weighted average temperature Tm in China,a weighted average temperature Tm model without meteorological parameters-the NTm model was established and compared with the GPT2w-Tm model,and the accuracy was found to be comparable.In addition,there is little difference between the NTm model and the traditional BTm model on the water vapor inversion.Therefore,the NTm model is considered to be suitable for PWV inversion in China.From 2013 to 2018,the calculated PWV deviation and medium error are close to each other,with an average deviation of 0.2mm and a medium error of ± 4.1mm.Using the 2013 sounding data from Kunming to retrieve atmospheric water vapor,it was found that the PWV 12-month average median error calculated by the QZWD model inversion was ± 3.3mm,which was higher than the PWV accuracy calculated by the EGNOS model inversion of ±5.2mm,about 36%higher.In the absence of meteorological parameters,it is convenient and simple to calculate the zenith atmospheric precipitation in China using the QZWD model and NTm model in China,but it is also practical.