Ground-based GPS Data Processing And Application In Weather Analysis

Water vapor is a major component of the atmosphere and is also a greenhouse gas. Although it is a minor component of the atmosphere, water vapor has considerable impact on local atmospheric conditions due to its uneven spatial distribution and rapidly shifting concentrations. The water vapor is a component that changes most in the atmosphere, whose changing scale is finer than the wind and temperature. One of the basic problems of meteorology and weather forecasting is obtaining accurate measurements of the distribution and variation of water vapor in the atmosphere.First appearing in the1990s, global positioning system (GPS) techniques are new monitoring tools that provide mechanisms to estimate water vapor concentrations in the atmosphere. Due to high precision, high capacity, high spatial and temporal resolution, and ability for all-weather operation, and in obtaining water vapor data from the atmosphere at near real-time speeds, it has received much attention. National, sector, and local have invested substantial manpower and material resources in GPS techniques, and as a result, have accumulated a great quantity of original GPS data. However, the ground-based GPS water vapor monitoring network in China is still in its rudimentary stages and its detection techniques involve cross-integration of surveying and meteorological knowledge; several problems are yet to be further solved and studied.Using GPS delay to determine precipitable water vapor (PWV) in the atmosphere entails several problems:too many processing links and complicated techniques, which restrain the effective use of the ground-based GPS data; the inversion process of water vapor data by GPS; it lacks a unified standard of modeling and application of weighted average temperature; in the processing of ground-based GPS data, GPS-PWV cannot be determined if either the original GPS data or the ground meteorological data is missing; GPS-PWV in weather analysis is restricted to research on changes of PWV itself; and researches on the deep-seated reasons for water vapor change is rare. In response to these questions, this paper carries out a deep analysis and study, and arrives at the following conclusions.(1) The analysis determines that the two key problems in meteorological operation department’s current techniques of water vapor inversion by ground-based GPS data are the usage of software in processing GPS data and integrated treatment in the operation;(2) It focuses on the entire process of water vapor inversion by ground-based GPS data, qualitatively and quantitatively analyzes the main error sources, presents a solution to this error source, and calculates the range of error source reduction required to meet the accuracy demands of meteorological applications;(3) The article uses mathematical methods and intelligent algorithms to propose two complementary schemes for addressing the lack of measurement of precipitable water vapor. Through principle component analysis, the selection problems of fitting factors are resolved, and through a sensitivity test, the structural composition of the neural network model that best fits precipitable water volume is obtained;(4) As the fitting formula for the weighted average temperature is not unified, each region has its own problem of Tm calculation model. The article puts forward a Tm calculation model that can be applied in China, and this model achieves the requirement of having sufficient accuracy for use in meteorological applications;(5) It conducts a systematic analysis of the changing characteristics of GPS-PWV in two strong rainfalls in Jiangsu and Chongqing, assiduously researches the dynamic and thermodynamic processes in line with NCEP reanalysis data, conventional sounding data, ground meteorological data, and WRF numerical simulation results, and reveals the deep-seated reasons for the water vapor change that is the internal reason for PWV change, as well as deepens the significance of GPS-PWV as forecast index;(6) It uses GPS-PWV data for the first time to analyze the spatial distribution and cycle of GPS at Chengdu plain in autumn. The research establishes a negative correlation between PWV and the altitude, and reveals that PWV changes features in one quarter of a season, i.e., that mid-autumn is a turning point where PWV changes from more to less, where rainfall of Chengdu in autumn occurs mainly at night, and there is sufficient correspondence between the accumulation or release of the total amount of water vapor in atmosphere and the actual precipitation on ground, and where the precipitation peak usually appears earlier than the maximum of precipitation intensity.