Spatial distribution, origins, and driving mechanisms of groundwater nitrate contamination in a rural-agricultural-mining catchment: Insight from PMF, machine learning and multi-isotopic techniques

Study region A rural-agricultural-mining catchment of Chengde city, the critical water conservation area of Beijing-Tianjin-Hebei region, northern China. Study focus To comprehensively elucidate the sources and fate of nitrate in the shallow aquifer of the study region, this study established an integrated analytical framework combining hydrochemical and soil geochemical multivariate statistics, geostatistics, redundancy analysis, Random Forest-XGBoost-SHapley Additive exPlanations, PMF-MixSIAR source apportionment modeling, and multi-isotope tracers to identify spatial patterns, influencing factors, sources, and driving mechanisms of groundwater nitrate migration and transformation processes in the aquifer. New hydrological insights for the region The most influential drivers of nitrate include Cl⁻, total nitrogen, total dissolved solids (TDS), electrical conductivity (EC), groundwater depth, residential and industrial-mining land use, soil biogenic elements. The average contributions to groundwater nitrate were: manure/sewage (33.94%), soil nitrogen (24.74%), atmospheric deposition (21.91%), chemical fertilizer (19.18%), geological and mining sources (1.06%), and surface water–groundwater exchange (1.03%). Nitrogen biogeochemical cycling in the catchment is jointly governed by groundwater system and pollutant source–sink dynamics. Groundwater discharge zones—the hyporheic zone—serve as denitrification hotspots, where organic matter drives nitrogen biogeochemical cycling. In riparian wetlands of the V-Ti-Fe-P mining area, nitrate contamination and high-intensity organic carbon loading, microbially mediated processes including nitrate-reducing Fe(II) oxidation, sulfur-oxidizing autotrophic denitrification, and heterotrophic denitrification were observed, though their overall effect on nitrate attenuation was limited.