| The troposphere where most of the water vapor in atmosphere are collected is an important part of the near-Earth space environment.Various disastrous weather such as rainstorm and hail caused by tropospheric water vapor can cause serious damage to economic construction and people’s life and property safety of state.Therefore,the importance of monitoring extreme precipitation events is increasing gradually.Not only millimeter-level static positioning and centimeter-level kinematic positioning can be achieved by Precise Point Positioning(PPP)technology which based on Global Navigation Satellite System(GNSS)technology,but also high-precision estimation of zenith tropospheric delay(ZTD)that can be used to calculate Precipitable Water Vapor(PWV)which can be used for monitoring extreme precipitation events.However,the accuracy of PPP and ZTD estimation are both affected by many factors,such as satellite precise correction products and observation models,as well as atmospheric model parameters.In addition,with the extensive usage of GNSS,there are many GNSS monitoring stations used in public fields such as agriculture and deformation monitoring.Poorer observation conditions are appeared in these stations when comparing with the International GNSS Service(IGS)stations that have been professionally selected sites.However,due to the large number and wide distribution,the applicability of these GNSS monitoring stations to PPP and water vapor retrieval is still worth being studied.Based on the above content,the impact of different precise correction products and mathematical models on the positioning performance of PPP and the accuracy of ZTD estimation are firstly studied in this thesis,to obtain a refined PPP parameter estimation strategy.On this basis,real-time PWV results are retrieved using GNSS observation data during heavy rainfall,and the correlation between PWV and the data of measured rainfall is analyzed.Finally,based on the simulation of various shelter environments where GNSS monitoring stations are located,the impact of these shelter environments on the positioning accuracy of PPP and retrieval accuracy of PWV are studied,supporting for more GNSS monitoring stations to reinforce the network of PPP and ground-based water vapor observation.The content and main results of this thesis are as follows:(1)The effects of aforementioned factors on the accuracy of PPP parameter estimation are studied based on the Multi-GNSS Experiment(MGEX)observation data in the Asia-Pacific region.The results show that the three-dimensional accuracy of PPP using the precise products from the GFZ Research Center for Geosciences can reach up to 0.67 cm,which is significantly better than the respective results of 0.89 cm and 0.72 cm which are from using products of the Shanghai Astronomical Observatory(SHAO)and Wuhan University(WUH).The performance based on PPP from ionosphere-free combination model is better than the result of the dualfrequency uncombined model,and the performance of ZTD retrieval from ionosphere-free combination model is equivalent to that of the dual-frequency uncombined model.The accuracy of ZTD retrieval based on the above two observation models can be both slightly improved by using gradient estimation.Finally,based on all of the above results,a refined PPP parameter estimation strategy is obtained.(2)On the basis of the verification of the PPP estimation strategy for the data from Chinese Crustal Movement Observation Network(CMONOC),using the weighted average temperature retrieved from HGPT model,the retrieval accuracy of PWV,and the availability of PWV monitoring the “7.20”rainfall events in Zhengzhou and its surrounding areas,as well as the correlation between PWV and rainfall are studied from the observation data of CMONOC.The results shows that the absolute error and standard deviation(STD)of the real-time PWV results are 10 mm and 3mm,respectively,compared with the Fifth Generation Reanalysis PWV product(ERA5)released by the European Centre for Medium-Range Weather Forecasts(ECMWF),which can meet the requirements for rainfall monitoring of PWV.Comparing with the PWV and measured rainfall data from three GNSS stations in Zhengzhou and its surrounding areas with different rainfall intensities,it can be concluded that the sequence feature of PWV rising rapidly to more than 50 mm and then dropping to about 30 mm can not only monitor long-lasting heavy rainfall events,but also can detect relatively short-lasting and lower-intensity ordinary rainfall events accurately.However,the above sequence feature may also result in false rainfall reports.Further analysis shows that the correlation between PWV and rainfall is stronger than that of tropospheric gradient,but the correlation coefficient is only about 0.48,indicating weak correlation.(3)Based on the simulation of three shelter environments of GNSS monitoring stations,including head-space shelter,one-side shelter and surrounded shelter environment,the performance of PPP and PWV retrieval under each shelter environment are studied with the refined PPP estimation strategy mentioned above.Comparing the PPP performance under different shelter environments,it can be seen that the accuracy of PPP under head-space and one-side shelter environment is reduced by 49% and 65%,respectively,compared with the result in open-space environment,while still providing PPP result of millimeters-level accuracy in horizon and sub-centimeters accuracy in vertical.In comparison,the impact of surrounded shelter environment on positioning accuracy in horizon is less,but the vertical accuracy of positioning in this shelter environment is only up to 3.8cm,which is significantly reduced compared with the results in other environments.Comparing the retrieved results of PWV under various observation environments,it can be showed that the PWV in head-space and one-side shelter environments can still maintain good consistency with the result from open-space environment,with Root Mean Square Error(RMSE)of 1.0mm and 1.5mm,respectively.Therefore,GNSS monitoring stations in these two environments can still be used for groundbased water vapor retrieval.The PWV in surrounded environment cannot reflect the trend of PWV change in open-space environment,and the RMSE in this condition is only 5.4mm.Thus,the data from GNSS monitoring stations in this environment is not suitable for water vapor retrieval. |
PDF Full Text Request