Three-dimensional Troposphere Water Vapor Tomography Combining Ground-based GPS And Space-based COSMIC

It can effectively increase the accuracy of climate monitoring and weather forecast through obtaining accurate three-dimensional distribution of the water vapor in the atmosphere. In addition, it will usefully improves the initial humidity field of numerical weather prediction and increase the accuracy of numerical weather prediction. GPS technology provides a new method for detecting the three-dimensional distribution of the water vapor with the advantages of a24-hour job, unaffected by weather, high detection precision and real-time continuity of operations, etc which has availably made up for the weakness of traditional detection methods.In this paper, we combine ground-based GPS and COSMIC (The Constellation Observing System for Meteorology, Ionosphere and Climate) occultation data for3D troposphere distribution of Hong Kong. For the limitations of the conventional iterative reconstruction algorithm, we propose an improved algebraic reconstruction technique (IART) for3D water vapor tomography. Moreover, in order to overcome the disadvantage of the single reconstruction algorithm, we propose a combined reconstruction algorithm (CRA) for3D troposphere water vapor tomography. The trial calculated by the GPS data of12meteorological stations in Hong Kong to verify the feasibility and superiority of the new methods (IART and CRA). The main contents and achievements of this thesis are as followings:1) We used the Precise Point Positioning to obtain the Zenith path delay(ZPD) and get the Zenith hydrostatic delay(ZHD) by the Saastamoinen model. We can access to Zenith wet delay (ZWD) by removing ZHD from ZTD. Finally, we can acquire the slant water vapor(SWV).2) For the shortcomings of classical algebraic reconstruction algorithm that has a slower rate of convergence and lower reconstruction precision in a large-scale water vapor inversion iteration, this paper proposes an improved algebraic reconstruction iteration algorithm. IART can automatically adjust the various elements of the relaxation parameter vector in the algorithm and improve the accuracy of the inversion, and, it also increase the speed of inversion at the same time.3) To overcome the drawbacks of the single reconstruction algorithm, we present a combined iterative and non-iterative reconstruction approach for the three-dimensional (3D) water vapor inversion. The CRA has higher accuracy than the single construction algorithm in obtaining the vertical distribution of water vapor. What’s more, CRA can effectively solve the problem of the initial value selection.4) We obtain the three-dimensional water vapor distribution of July20,2012to August2,2012, and turn it into the total column integrated water vapor. Then we study the process of the torrential rain in Hong Kong according to the distribution and change of those values.