Integration of fuzzy logic and geographic weighted regression modeling for enhanced groundwater potential mapping using remote sensing and GIS

Abstract

Groundwater resource management is critical in regions with limited surface water availability. Traditional mapping methods often face challenges in addressing spatial heterogeneity and data uncertainties, creating a need for more advanced approaches. This study addresses the research gap by introducing an innovative methodology combining fuzzy logic and Geographically Weighted Regression (GWR) to enhance groundwater potential mapping. Fuzzy logic addresses uncertainties through membership functions, while GWR captures spatial variations via localized regression coefficients. Utilizing conditioned data sets of geological factors, land cover, and topographic indices, the proposed method produced high-accuracy groundwater potential maps validated through metrics, such as local R2 values and Moran’s Index. Key findings reveal significant spatial variability in groundwater potential, with southern regions showing enhanced recharge capacity due to favorable geological conditions. The integration of fuzzy logic and GWR demonstrated robust predictive performance and the ability to account for local spatial patterns. These results provide valuable insights for hydrological readers, guiding sustainable groundwater management practices and targeted water resource interventions in diverse environmental settings.