基于质控方法的地面气象资料预警系统设计

准确、可靠的地面气象资料是气象和地球相关学科研究的重要基础,是提高气候预测水平的重要前提,为使气象观测数据具有更好的代表性、准确性和比较性,地面气象资料预警系统在地面气象资料数据入库之前,需进行一系列的数据质量控制,控制方法除了传统的极值,时间、空间、内部一致性检查之外,还包括数据挖掘中的BP神经网络算法。此外,系统设计过程包括后台数据库结构、前台入库及预警系统设计,该系统采用了多线程操作机制进行气象资料解析入库,提高了数据处理效率及预警信息的准确性,同时通过图形化的界面展示预警信息,这在日常生活、农业生产及社会稳定等方面都有重要的意义。

Design of the Ground Meteorological Data Pre-Warning System Based on the Quality Control Method

Accurate and reliable ground meteorological data is important basis for the related discipline research meteorological and earth as well as important prerequisite for improving climate prediction level. In order to make the meteorological obser-vation data to be more representative, accuracy and comparability, a series of quality control of the ground meteorological da-ta should be conducted before it was stored in the pre-warning system of ground meteorological data system data. Except for traditional control methods such as extreme value, consistency checks of time, space and internal, the BP neural network al-gorithm of data mining was also feasible. The system design process includes the design of backstage database system struc-ture as well as front storage and pre-waring system. The system adopts multithreading mechanism into meteorological data analysis to improve the efficiency of data processing and the accuracy of the pre-warning information. Meanwhile, through the pre-warning information could be displayed through graphical interface, which has important significance in daily life, agri-cultural production and social stability.