Abstract: Based on the hourly total cloud amount (TCA) product provided by FY-2G satellite and ground observation data of 26 observation stations in Jiangxi province from January of 2016 to December of 2018, the spatio-temporal distribution characteristics of correlation between two types of data were analyzed.The result shows that the TCAs based on FY-2G satellite and the observation data have good consistency in value and change trend with a population correlation coefficient of above 0.60.However, the spatial correlation is unevenly distributed.Specifically, the correlation coefficient in the vicinity of large water bodies (Poyang Lake and Zhelin Lake), is relatively small and less than 0.45, and that in southern Jiangxi province is less than 0.50.In addition, the correction coefficient of two TCA data in time is unevenly distributed and is the largest at 14 o'clock.Besides, based on the FY-2G satellite TCA product, the regression model of TCA with surface observation is established by using the least square method with segment modelling.As a result, the TCA and its change trend simulated by the model are closer to those of the ground-based observations.