Abstract: In order to improve the quality of meteorological observational data and to provide reliable data sources for researching on atmospheric activity regular pattern and a variety of meteorological applications, the paper presents a quality control method for atmospheric data, which using Drools rule engine and RETE algorithm. The method can check the message with traditional alphanumeric codes exactly and realtime. It also has some flexible ability on selfdefining and maintainability, by which to adapt the developing and changing requirements about atmospheric data. This method may inherit the traditional quality control methods, such as climatic extreme range check, internal consistency check, time consistency check, and spatial consistency check, etc. It can enhance the processing efficiency of realtime data, and can also promote the accuracy of result produced by meteorology applications. The practices prove that the applications with this method are effective and efficient.

Key words: atmospheric observational data, quality control, rule engine