Data fusion of SST from China FY and HY2 satellites was made by American LAPS(Local Analysis Prediction System). The system can ensure the spatial and temporal characteristics of different elements to fit each other in atmospheric motion,and it can provide fusion data with standard format and common corrodinate. The fusion data is suitable to join the quantitative diagnosetic analysis of variables. As the LAPS system has no a special input channel for satellite SST data, the numerical adjustment was done for data input of marine station. According to the different sensing principles of FY and HY2 satellites, and the characteristics of random spatiotemporal variation of satellite SST data, the inversion data of satellites were preprocessed by several steps, including outliers discrimination and elimination, the anomaly gradient discrimination and elimination, the numerical treatment of different satellite scanning grids, and data smoothing, etc. After the testing of different methods, it was verified that through preprocessing, the fusion effect of FY and HY2 satellites SST data was more reasonable and correct. Based on LAPS SST fusion and further statistical tests, it was confirmed that the SST fusion fields could keep the detail description of spatiotemporal features of SST distribution, and also resolved the problem of satillite SST with blank patches along the coastal zones. Furthermore, the SST fusion fields were applied to the diagnostic analysis of themodynamics spatiotemporal variance in air-sea synoptic process in cold waves and cold air invasion to Yellow sea and East China Sea. The results showed that the SST fusion could provide quantitative information and description for meso-scale strength variance, north-south varing difference of SST distribution and sensible heat flux in China offshore regions.