Integrated decision-making model for groundwater potential evaluation in mining areas using the cusp catastrophe model and principal component analysis

Abstract

Study RegionPanxi mining area (15061 km2, located in Sichuan, China). Study FocusThis study aims to delineate groundwater potential zones in mining areas using a new method based on the cusp catastrophe model (CCM) and principal component analysis (PCA). First, 13 indicators were selected from natural and anthropogenic dimensions, and a comprehensive analysis of the indicators was performed using PCA. Second, the results of the PCA were considered as control variables, and the CCM was used for groundwater potential evaluation modeling. Finally, the receiver operating characteristic (ROC) curve was used to validate the new method and compare it with catastrophe fuzzy membership functions (CFMFs). New Hydrological Insights for the RegionThe results suggest that the area under the ROC curve of the success and prediction rate accounted for approximately 0.85 and 0.76, respectively, in the new method, which were higher than those in the CFMFs. The largest area (39 %) with groundwater potential in the study area has a “moderate” groundwater potential status, followed by an area (28 %) with a “good” status, an area (20 %) with a “poor” status, and areas (12 % and 1%) with “very good” and “very poor” statuses, respectively. The groundwater potential in the study area was unevenly distributed and changed drastically. Topography, drainage density, and land use/land cover had the highest contribution in the modeling process.