Abstract: Taking ground meteorological data of Huangshi National Meteorological Station as the samples, we compare and analyze the data of main and standby automatic meteorological stations in the same software. By combining the absolute differences and standard deviation values, we aim to improve the functions of automatic debugging and automatic error correcting of the state- level meteorological stations. Besides, data of temperature and air pressure of ground meteorological data are used for the automatic identification test with the data processing methods proposed by this study. Results show that the expected effects of automatic identification of wrong data source can be achieved.