| Water vapor is one of the most important parts in the lower atmosphere,it accounts for less percentage but plays the key role in various climates and a series of weather phenomenon.The variation of water vapor is affected by many factors,like temperature,topography and seasons with characteristics of changing fast with time and heavily in horizontal and vertical directions,which makes it difficult to monitor with high precision.The processing of water vapor changing is accompanied with the absorbing and releasing of enormous energy,which has an important signicance on the vertical stability of atmospheric structure as well as the formation and evolution of rainstorm and severe convection weathers.The sharply variation of water vapor phase in a short time will lead to the disorder of atmosphere spatial structure and the occurance of damaging weathers,like rainstorm,typhoon,flood,etc.Recently,the research on water vapor attracts much attentions with the coming of disastrous weathers with high frequency and the global warming.However,how to obtain the water vapordistribution with high spatial-temporal resolution is still a key problem currently,which limits the development of numerical weather prediction,especially for the precision of rainfall quantitiative forecast.Global Navigation Satellite System(GNSS)can be used to remote sensing water vapor with advantages of continuous operation,all-weather,high precision,high spatial-temporal resolution,low construction costs,etc,which make it used for water vapor monitoring with a wide range and high density,when compared to the traditional water vapor probing means.GNSS can not only use for obtaing the two-dimensional spatial distribution of precipitable water vapor(PWV),but also for reconstructing the three-dimensional profiles of atmospheric water vapor vertically using the tomography thchnique,which increasingly becomes one of the most potential tools for acquiring the atmospheric water vapor.The data derived from GNSS is of important practical application and research value.The real-time GNSS observations can be used for the monitoring of atmospheric water vapor content and its spatial distribution as well as the warning 'for the short-term disastrous weathers like strong convective condition.The accumulated GNSS observations with a long-time span can be used for the analyzing and warning of the large-scale long-term climate change.This thesis focus on the research of some important issues related to water vapor remote sensing using GNSS technique.For the calculating of PWV,a global conversion factor model is established without using meteorological data based on the conversion formula from zenitih wet delay(ZWD)to PWV.For the reconstruction of three-dimensional water vapor,the issues for the division of tomography region,the selection of tomography height and the determination of weights between different equations of tomography modeling are discussed and investigated.Some studies are carried out about how to use the satellite rays penetrating from the side face of research area,and developed a comprehensive theory and method of three-dimensional water vapor tomography.For the application of water vapor,a short-term precipitation forecasting method is proposed for disastrous weathers based on the high-precision ZTD/PWV products from GNSS observations.The main work and contributions of this hesis are as follows:1.Building a global conversion factor model(G? model)without using meteorological parameters by analyzing the impact factors of the conversion factor from ZWD to PWV based on the GGOS Atmosphere products.The validated result shows that the model's inner accuracy is at 10-3 level with the source data,while the accuracy of calculated PWV using the established modell is about 3mm when compared with the PWV estimated using the meteorological data in HKSC and ANJI stations.In addition,a method to unify the height to the same elevation system is presented for the parameters from different data analysis centers.2.For the reconstruction of three-dimensional water vapor,a non-uniform symmetrical division method of horizontal voxels is proposed based on the distribution of GNSS signals penetrating through the research area,and the tomographic top boundary is determined according to the variation of water vapor profile with height.Numerical result manifests that the number of voxels is increased by about 9%,while the accuracy of reconstructed water vapor field is improved about 18%..3.In order to improve the number of satellite rays used for water vapor tomography,a method to sophisticated troposphere tomography modeling is proposed by virtualized the slant water vapor contents using wet mapping function.Experimental result reveals that the number of voxels crossed by rays is increased by about 3%,while the quality of reconstructed result is improved by 11%.4.A method to determine the weights of the same equation in the tomography modeling is first presented,and the weights for different equations is determined by introducing the idea of co-integration test in the macro econometric analysis.Tomographic result are compared with that from the radiosonde data and ECMWF products,and the result shows that a reasonable result can be obtained under different weather conditions.5.Extensive studies has been carried out for the using of satellite rays penetrated from the side face of tomography area,which cannot be used in most existing studies,and the related theory and method for improving the utilization rate of satellire rays for water vapor tomography is increasing developed.The main contribution includes three aspects:? a troposphere tomography method is proposed with an assisted area to improve the utilization rate of signals.The reconstructed water vapor information of the assisted area is considered as the initial value of iterative computation,which futher reflects the valuable of signals crossing from the side face of research area.? By introducing the water vapor unit index,a water vapor tomography method is proposed.The signals crossing from the side face can be used to calculate the initial water vapor density using the water vapor unit index in a certain voxel,and those estimated values are considered as priori constraint to impose into tomography modeling.? Maximally using GPS observation for water vapor tomography method is proposed to improve the satellite rays used.This method allows all satellite rays with a certain satellite elevation angle used to build observation equation and effectively improved the utilization rate of GNSS rays,which is the most mature and perfect theory currently.6.For the application of water vapor products,? a short-term rainfall forecasting method is proposed using PWV time series by analyzing the relationship between PWV and corresponding precipitation.A satisfied result is obtained when applied such method in a typhoon event of "Can Hong" 1509 passed through the Zhejiang province in July,2015.?a relationship between ZTD,PWV,ZHD and precipitation is analyzed and the result shows that the most of ZTD changes is caused by the variation of ZWD or water vapor content,therefore,a method to forecast short-term rainfull is proposed using ZTD instead of PWV.In addition,the relationship between ZTD and precipitation duiring the process of cold front is also investigated.? Finally,the change of three-dimensional water vapor is tracked and analyzed duiring a rainstorm event using the troposphere tomography technique. |
PDF Full Text Request