Groundwater Potential Map Data of Cheongju Area Using Probabilistic Methods

This study aims to assess and map the groundwater potential zones in Cheongju area, South Korea, using a probabilistic approach based on the frequency ratio (FR) model. To construct a reliable groundwater potential map data, multiple thematic layers were compiled and analyzed using a GIS-based spatial analysis framework. These included channel network distance, digital elevation model (DEM), depth to water, distance from fault, hydrogeology, normalized difference vegetation index (NDVI), slope, soil drainage, soil land use, soil material, soil type, soil topography, total catchment area, topographic wetness index (TWI), and valley depth. The FR model was applied to evaluate the spatial relationship between existing well locations and the influencing factors. Based on this, the study area was categorized into five classes of groundwater potential, from very low to very high. The performance of the model was validated using the receiver operating characteristic-area under the curve (ROCAUC), with results indicating a predictive accuracy showing 76%. The final groundwater potential map data serves as a valuable tool for sustainable groundwater development, optimal site selection for drilling, environmental protection, and integrated water resource management. Additionally, it offers practical implications for mitigating groundwater depletion, improving disaster preparedness against drought, and supporting decision-making in local and national water planning policies.