Abstract:On the basis of IDW, combining with temperature change with height and relationship between temperature field and pressure field at each observation station, the interpolation algorithm of temperature data from ground meteorological stations is improved. The experimental data are analyzed from the aspects of the relationship between search radius and distance weight coefficient and the number of stations in the fixed search radius and distance weighting coefficient to find the best algorithm parameters and minimum error. The improved interpolation algorithm is applied to the quality control of temperature observation data. The results show that the improved algorithm can evidently improve the interpolation accuracy resulted from the near surface temperature inversion.In the eastern China where stations are densely distributed, the sensitivity of temperature test can be up to 1 to 5 ℃.