This study was conducted in the upstream watershed of Du River, central China, and ASTER GEDM was processed in the system for Automated Geoscientific Analyses (SAGA GIS) package in the R-platform. On the basis of obtaining watershed soil topographic index (STI) and its threshold setting, using Python module programming, three different scenarios (STI≥9, STI≥10, and STI≥11) were selected to delineate and understand the land use types and their areas and spatial distributions within hydrologically sensitive area (HSAs) in the watershed. The results showed that STI technique improved the identification of HSAs, which could reasonably predict the spatial differentiation of runoff in the watershed. Forest land was the major land use in the HSAs under all three scenarios, and about 50% of water was identified as HSAs. Under the three scenarios, the agricultural landscape and development landscape in the HSAs should be prohibited from carrying out intensive land development, and should arrangement priority vegetation buffer zones to protect water quality.