Dynamic monitoring of plant cover and soil erosion often uses remote sensing data, especially for estimating the plant cover rate (vegetation coverage) by vegetation index. However, the latter is influenced by atmospheric effects and methods for correcting them are still imperfect and disputed. This research supposed and practiced an indirect, fast, and operational method to conduct atmospheric correction of images for getting comparable vegetation index values in dif- ferent times. It tries to find a variable free from atmospheric effects, e.g., the mean vegetation coverage value of the whole study area, as a basis to reduce atmospheric correction parameters by establishing mathematical models and conducting simulation calculations. Using these parameters, the images can be atmospherically corrected. And then, the vegetation index and corresponding vegetation coverage values for all pixels, the vegetation coverage maps and cover- age grade maps for different years were calculated, i.e., the plant cover monitoring was realized. Using the vegetation coverage grade maps and the ground slope grade map from a DEM to generate soil erosion grade maps for different years, the soil erosion monitoring was also realized. The results show that in the study area the vegetation coverage was the lowest in 1976, much better in 1989, but a bit worse again in 2001. Towards the soil erosion, it had been mitigated continuously from 1976 to 1989 and then to 2001. It is interesting that a little decrease of vegetation coverage from 1989 to 2001 did not lead to increase of soil erosion. The reason is that the decrease of vegetation coverage was chiefly caused by urbanization and thus mainly occurred in very gentle terrains, where soil erosion was naturally slight. The results clearly indicate the details of plant cover and soil erosion change in 25 years and also offer a scientific founda- tion for plant and soil conservation.