In the current study, time-series MODIS images were selected to calculate the monthly temperature vegetation dryness index (TVDI) from 2000 to 2015 of the Songnen Plain. There was a strong correlation between the TVDI value and soil relative moisture, which signified that TVDI could be used as a substitute indicator for soil relative moisture. The spatial dataset for the surface soil moisture during the crop growing season (from April to October) was rebuilt with a 1 km cell size. The temporal and spatial variation characteristics and dynamic trends of soil moisture were analyzed by means of ensemble empirical mode decomposition (EEMD) and the impact of soil moisture on crop yield was discussed. The results showed that soil moisture was gradually increasing from the southwest to northeast in the Songnen Plain in terms of spatial distribution. Regarding temporal changes, soil moisture had increased over the past 16 years, although the frequency of extreme water shortage events had increased. There was a significant correlation between yearly average soil moisture and crop yield, which showed that crop yield was primarily affected by the yearly average soil moisture. Crop yield was affected by soil moisture of different seasons, especially that of summer.