Abstract: With the acceleration of global energy transition, pumped storage power stations (PSPS) have become a critical technology for stabilizing new power systems due to their regulation capabilities and economic advantages. However, traditional site selection methods relying on manual surveys suffer from inefficiency and limited coverage, hindering large-scale development. The study proposes an automated site selection methodology based on Digital Elevation Models (DEMs), by integrating terrain feature analysis and hydrological analysis constraints modeling, to construct a collaborative extraction framework for natural and semi-natural basins. The method innovatively combines key technologies such as depression identification, catchment area search, main stream analysis, and dam site optimization, combined with dynamic threshold-based screening of core parameters (e.g., effective storage capacity, water head, and length-to-height ratio), significantly improving site selection efficiency and engineering adaptability. Case studies demonstrate that this framework accurately identifies candidate reservoirs, reduces survey costs, and resolves feasibility assessment challenges in artificial dam construction. The results provide a scientific basis for large-scale PSPS development, aligning with the goals of the "Medium- and Long-Term Development Plan for Pumped Storage (2021–2035)" while supporting flexible regulation capacity in new power systems and advancing carbon neutrality targets.