It has an important application to blend multi-source remote sensing to generate high spatial-temporal resolution data. In this study, three spatial-temporal fusion algorithms were analyzed and compared with each other. They are spatial and temporal fusion method using high/low resolution time-series images(STIFM), spatial and temporal data fusion method based on decomposition of mixed pixels(STDFM) and an enhanced spatial and temporal adaptive reflectance fusion method(ESTATFM). The study area was located in Yingke irrigation districts. Temporal change information was detected from sequence MODIS and high-resolution spatial information was provided by ASTER/TM. Results showed that STDFM can make an optimal effect in red reflectance(r=0.91) and ESTATFM in near-infrared reflectance(r=0.71) which was compared with actual observations of ASTER. The blended NDVI from above three method were similar with r higher than 0.84. Results also indicate that ESTATFM is well adapted to the heterogeneous region, such as corn and wheat.