Against the background of informatization,numerical weather prediction(NWP) has become an important means of meteorological prediction.The quality control of surface temperature observation data serves as the foundation of data assimilation,which is helpful for improving the accuracy of numerical weather prediction.Based on this,in this paper the B spline surface fitting algorithm is applied to the quality control of surface air temperature observation.Considering the spatial correlation between each reference station and target station with the region,a quality control algorithm for surface temperature observation data based on spatial correlation and B spline surface fitting(BSF) is proposed.Nanping,Nanjing,Taiyuan,Lasa,Jinghong and Changchun Stations are selected as the target stations,and the surrounding stations within 300 km are chosen as the reference stations.The time temperatures at 02:00,08:00,14:00 and 20:00 are selected as the observation data.BSF,SRT and IDW are used to control the quality of the target stations' surface temperature data.Combined with the mean absolute error(MAE),root mean square error(RMSE),index of agreement(IOA),nash coefficient(NSC) and rate of error detection,the results of the comparison lead to the conclusion that the BSF algorithm is able to effectively identify suspicious values in surface temperature observation data,and that it has a higher error detection rate and stronger adaptability than the traditional method.