Data Analysis Of The Upper-air Detection And The System Design For Quality Control

The upper-air observation is the basic service of the meteorological business and also the main source of the weather forecast, climate analysis, scientific research and international exchange of meteorological information.The thesis focuses on the quality control system design and development of the data generated from the standardized software in China’s upper-air meteorological stations. The key points are the software requirements, the design and algorithms of system structure as well as the realization of each module of the algorithm and thoughts, and software applications, and identification of false data.This paper detailedly elaborates on the problems of L—band upper-air observation such as detection of outlier, mistake of moment data, unusual rising rate, and the jumping in date of detecting wind sounding system. The paper offers some solutions to these issues such as checking the reading data, careful proofreading and being familiar with manual and specification.The thesis uses the Delphi language program to realize the quality control of upper-air materials, based on the observational data from Xi’an meteorological stations and combined with the stations’ pre-trial work practice in order to ensure the timeliness and data quality of meteorological observation and eliminate the influence of error data on the whole record. The paper introduced the balloon program and the data processing process in detail by utilizing the imaging platform Windows and abiding by the current high altitude meteorological observation specifications. By employing the software, balloon tests are carried out to obtain the massive precious raw data many times in the Xi’an Jing He meteorological observatory. Meanwhile, several kinds of abnormal phenomena and its judgement and processing are analyzed and discussed in this paper. The programming principle of real-time quality control software and its application are introduced and several effective approaches to identify the false data are found out, which has paved the way for the better automatic quality control.The above-mentioned three aspects turn out to be a successful simulation and practice and it ensures the reusability and expansibility of the system. The standardized software, in this thesis, meets the requirements of WMO(world meteorological organization) and the visual design of data output can also meet the needs of different users.