Base on the association rules in data mining technology, the relationship among the different elements of surface meteorological observation is analyzed, so to provide reference for internal consistency checks for different elements of meteorological observation data. The distribution ranges with strong association rules are obtained by analyzing surface meteorological automatic observation data of 2013 from the Beijing Nanjiao experimental station. Analysis of the correlation between the relative humidity and wind speed variation is made and the change thresholds are obtained. According to the correlation between the relative humidity and visibility, visibility changes are more obviously affected by precipitation. An analysis is made of the different levels of rainfall intensity corresponding to the interval values of visibility for the specific precipitation process to reveals the changing relationship between visibility and rainfall intensity. The results show that using the correlation to identify meteorological observation data can provide reference for meteorological observation data quality control.