Water Vapor Retrieved From Ground-based GNSS And Its Applications In Extreme Weather Studies

The Global Navigation Satellite Systems(GNSS),as a disruptive navigation technology,has been successfully applied in wide range of important fields.This includes,for example,surveying and mapping,meteorology,transportation,environment and agriculture etc.As a relatively new method for retrieving water vapor,GNSS meteorology is regarded as an effective novel tool with a huge potential.The technique is capable of overcoming many shortcomings(e.g.high cost,low temporal resolution,large instrumental biases and drifts,weather condition sensitivity etc.)of traditional water vapor acquisition means.One of its most distinctive advantages is that it can operate under all weather conditions and obtain real-time global atmospheric water vapor in a high accuracy.However,the applicability of GNSS meteorology in multi-scale meteorological disaster events'monitoring and forecasting is still an area of very limited research.The main aims of this thesis are to study the theory and new applications of monitoring extreme weather using ground-based GNSS water vapor retrieval technique.Various atmospheric correction models are first studied,followed by an investigation of the optimization of the key parameters in GNSS water vapor determination.The performance of the GNSS derived water vapor products are evaluated using different GNSS data processing techniques and observation models.Firstly,a high-precision GNSS water vapor monitoring system was established,and then the variations of multiple meteorological parameters such as water vapor,temperature,air pressure,wind speed and precipitation under extreme weather conditions were studied,in particular,for the effective retrieval of short-term and imminent early warning signals of extreme weather events during the process of water vapor changes.The main contributions of this research are summarized as follows:(1)The performance(i.e.accuracy distribution)of five mainstream empirical models for determining atmospheric temperature and pressure in China and the characteristics of annual,semi-annual and diurnal variations of these models were studied,which can provide a theoretical reference for the selection and optimization of the corresponding models.(2)To overcome the limitation of the low temporal and spatial resolution of meteorological data,a spatio-temporal IAGA-Kriging model by considering a spatio-temporal correlation of temperature and pressure was proposed to improve the resolution.The model can solve the problem of spatio-temporal discontinuity of meteorological data,and better performance in comparison with traditional spatio-temporal interpolation models was achieved.(3)The performance of several popularly used T_m regression models based on surface meteorological parameters in China was evaluated.By taking into account of nonlinear characteristics of the T_m models were optimized by combining Artificial Neural Network(ANN)or Support Vector Regression(SVR)model.The results showed that the new T_m model developed can effectively reduce the BIAS of linear T_mmodel.(4)A BNC+Bernese GNSS data processing system for monitoring water vapor was developed.The accuracy of the water vapor retrieval of the system with various observation and processing modes was also evaluated.The results showed that the PWV retrieved using a double difference method(DD-PWV)was dependent on the accuracy of satellite ephemeris,and there was little difference between real-time and post DD-PWV,even in typhoon weather conditions.The accuracy of real-time precise point positioning(PPP)PWV was lower than that of post PPP-PWV when the PPP method was used.(5)The super typhoon Mangkhut occurred in the Hong Kong region in 2018was selected as a study case,and the variations of temperature,pressure,wind speed,precipitation and PWV before and after typhoon landing were studied.A new method to monitor typhoons by using water vapor products with a high temporal resolution was proposed,and a theoretical geometric model of typhoon movement was established.The moving speeds of five different grades of tropical cyclones were studied by using the proposed method,and similar results were achieved by comparing independent information from meteorological authorities.It is believed that this research can provide a valuable theoretical basis and application reference for the further development of GNSS meteorology,in particular in the new applications of the GNSS-based technology for tropical cyclones.There are 51 figures,33 tables,and 198 references in this paper.