Periodic Oscillation Analysis Of GNSS Water Vapor Time Series And Autogeneration Of Meteorological Files

As an important part of the earth’s atmosphere, water vapor plays a key role in global climate change and many synoptic processes. The traditional measuring means have many disadvantages such as low spatial or temporal resolution, finite precision, and small scope of application, etc. Ground-Based GPS meteorology is a very promising technique which can retrieve highly resolved and accurate water vapor fields in space and time. It has advantages of near-real-time processing, no requirement for instrument calibration, all-weather, high precision, low cost for each station and so on. Based on the relevant knowledge of the troposphere, in order to study several important issues in Ground-Based GPS meteorology, this thesis analyzes the theory and method of hybrid neural atmosphere delay models. Explore the rule of the water vapor changes and physical factors. Combine with numerical weather model data and sounding data to generate GPS meteorology file, and use a variety of methods to validate the precision of generated meteorological files. Provide the basis for further improving the GPS water vapor inversion precision.The research contents of this paper are as follows:1 、 The theory and analysis of neural atmosphere delay models: Hopfield neural atmosphere delay model、Saastamoinen neural atmosphere delay model、UNB3 neural atmosphere delay model 、 UNB3 m neural atmosphere delay model 、 EGNOS neural atmosphere delay model. Analyze the adaptability of different models in different regions(China, America, Europe), latitudes and altitudes. The analysis shows that UNB3 m neural atmosphere delay model and EGNOS neural atmosphere delay model have the higher precision.2、Contrast and analysis of the mapping function: Introduce the commonly used mapping function: NMF、GMF、VMF1. And analyze the calculating results based on the different mapping functions by using the GAMIT software. Lay the foundation for subsequent analysis.3、Research the spatial and temporal characteristic of the PWV: GPS water vapor time series of Taiwan were chosen as data sources. By analyzing the corresponding relation between tropospheric water vapor and geographical environment, the conclusion that the water vapor depending upon the latitude, topography and climatic conditions was found. Then, the long and the short GPS water vapor signals of Taiwan were decomposed by Empirical Mode Decomposition(EMD). The result shows there were annual, semiannual, diurnal, and semidiurnal periodic oscillation graph in each GPS station. Combined with geographical climate factors to analyze the physical reasons of these oscillation periods, conclusions were as following. Annual temperature change and summer monsoon impact the annual cycle of water vapor. semiannual oscillation is mainly due to the interaction of the warm and cold currents.. Diurnal cycle of the water vapor is caused by the change of temperature and the direction of the wind. Amplitude of semidiurnal cycle which is mainly caused by atmospheric vertical movement is small.4、Based on ECMWF and sounding data to generate GPS meteorological file: By using ECMWF and sounding data acquiring station location of meteorological parameters, to produce the meteorological files which can calculate PWV. Program to generate meteorological data, and calculate the GPS meteorological file worldwide and inverse point of water vapor. The results show that the global m document based on ECMWF data has good accuracy worldwide. The precision of the m file on a global scale based on the sounding data shows an obvious polarization. The global m file based on the integrated data was the best.