Modeling And Application Of Real-time High-accuracy Troposphere Key Parameters

The troposphere is an important part of the Earth’s space environment and is the closest atmosphere to human life.It contains about 75%of the mass of the atmosphere,almost all of the water vapor and aerosols,and it plays an extremely important role in the global atmospheric radiation,energy balance,and water cycle.Meanwhile,the signal receive from Global Navigation Satellite System(GNSS)is afifected by its refraction as it passes through the troposphere,resulting in a large delay that affects the accuracy of positioning.But the content of atmospheric water vapor can also be retrieved from the tropospheric delay.GNSS water vapor detection has also become a very important technical means for detecting water vapor.Temperature,pressure,water vapor pressure,atmospheric weighted average temperature and other tropospheric parameters can not only be used for GNSS,Very Long Baseline Interferometry(VLBI)and other space technology to provide correction of tropospheric delay,but also serve the weather forecast ’climate change and other fields.In addition,the tropospheric delay is needed to estimate in the GNSS absolute positioning and long-distance relative positioning.Because the measured meteorological parameters are difficult to obtain,the initial value of the tropospheric delay is calculated by using the standard atmospheric parameters and the tropospheric delay model based on meteorological parameters,which accuracy is poor and often requires a long convergence process.In order to solve the problem of imprecise of tropospheric delay based on standard atmospheric parameters,since 1998,some scholars have established relevant empirical model of non-meteorological parameters to improve the accuracy of tropospheric delay,and some models can reach the equivalent accuracy with tropospheric model based on meteorological parameters.With the construction of the Continuously Operating Reference Station(CORS)in each country,the use of CORS can directly estimate the Zenith Total Delay(ZTD)for high-precision regional troposphere.Relative to other technologies,the regional troposphere model established through the CORS has a higher accuracy.There are three commonly used tropospheric delay models:tropospheric delay model based on meteorological parameters,empirical model of tropospheric parameters and tropospheric delay model based on GNSS measured data.There are many studies on the above three types of tropospheric models,and many progress has been made,but the recently published GPT2w(Boehm et al.,2014)and TropGird2(Schuler,2014)have not yet considered comprehensive aspects of various cycle items.Secondly,the accuracy of the existing tropospheric model based on meteorological parameters does not show an advantage compared with the empirical model.In addition,it is worthy to further study the ZTD and regional tropospheric delay modeling using GNSS observation data in real time.Based on the systematic study of different types of tropospheric models,this paper aims to propose a better method to improve the accuracy of existing models so as to better serve the fields of satellite positioning and meteorological monitoring.Thus,this paper mainly carried out the following work:(1)In this paper,the annual and semiannual cycles and daily cycle characteristics of the troposphere parameters are considered,and the amplitude and the initial phase of the daily period are also estimated as a periodic function.The new model called ITG is established which contains temperature,temperature lapse rate,pressure,zenith wet delay and other tropospheric key parameters.The accuracy of the model is verified by various external data,and the results show that:In contrast to the GPT2w and TropGrid2 models,the global performance of the ITG model is optimal,with ZTD’s global average accuracy of 3.73 cm.In order to adapt to the high temporal resolution characteristics of modeling data sources,according to the idea of time-segment modeling,this paper has further established a new tropospheric empirical model expression which called WGTEM.Under the premise that the time resolution of modeling data source is high enough,it can reflect the multi-peak and valley characteristics of the change of the daily cycle.After modeling with the same data source as ITG,the accuracy of WGTEM is higher than that of ITG,GPT2w,TropGrid2,and it can reflect the double peak and valley characteristics of pressure.(2)The accuracy of tropospheric delay model based on meteorological parameters is analyzed,which pointed out that the accuracy of the model is mainly affected by the error of Zenith Wet Delay(ZWD).In order to solve the shortcomings in applying in real time of the tropospheric delay model AN which has the highest accuracy of ZWD calculation,the relationshipλ’ between ZWD,the water vapor pressure and atmospheric weighted average temperature is directly obtained,and the empirical model of WGTEM proposed by(1)is used to obtain the IAN model which can be used for real-time computation.The results show that the tropospheric delay accuracy is superior to the existing Hopfield,Saastamoinen,Ifadis and empirical models.(3)A systematic study was conducted on the use of GNSS observation data and IGS RTS real-time ephemeris products to provide users with ZTD.The accuracy of real-time estimation of ZTD in ephemeris products published by IGS RTS in several analysis centers are analyzed,which found that the ZTD of the ephemeris real-time estimates using different analysis centers can be compared with the same accuracy level as post estimates,and the product with high stability should be the chose in actual application because of the gross error.When using CNES multi-system real-time ephemeris calculation,the accuracy of ZTD is also at the same level,but the increase of observation can increase the stability of solution.Real-time PPP ambiguity fixed has a certain improvement to accelerate the convergence of ZTD and improve the accuracy of ZTD.An adaptive region tropospheric model is proposed,which can automatically adapt to the distribution of CORS stations in different areas,and obtain the theoretical optimal solution.(4)Based on the theoretical research of(1),(2)and(3),a real-time tropospheric enhancement system is established,which can provide users with various tropospheric key parameter model services.The correctness of the system is verified by GNSS positioning,for GNSS positioning users,it is possible to promote accuracy of 22.8%and 44.5%for GPS real-time PPP users in plane and elevation after 20 minutes of initialization time.A reasonable stochastic model is given for the tropospheric empirical model.Using ITG can speed up about 15%of real-time PPP convergence and accelerate the fixed speed of about 10%of long baseline ambiguity.(5)The atmospheric weighted average temperature empirical model GTm-Ⅲ,which is the most accurate in the world currently,is established to ensure the conversion accuracy of ZWD to Precipitable Water Vapor(P WV).On the basis of platform proposed in(4),the real-time water vapor monitoring system based on PPP technology is realized for the first time in China,and the regional water vapor changes are monitored in real time at 1s sampling rate and put into practical application.Based on the two rainfall events in 2013 and 2014 in Wuhan,the feasibility of real-time atmospheric precipitation monitoring system for storm monitoring is discussed.