Spatio-temporal variation in karst groundwater cycling and quality: New insights from isotope and BI-LISA
摘要
Karst groundwater plays a crucial role in sustaining water resources. However, the seasonal variations in hydrological cycling and quality remain insufficiently understood. This study integrated stable isotopes (δ2H and δ18O), entropy-weighted water quality indices (EWQI), and Bivariate Local Indicators of Spatial Association (BI-LISA) to investigate the seasonal groundwater dynamics and water quality evolution in the Baotu Spring karst groundwater system, a representative karst region in northern China. The seasonal sampling of 147 groundwater and river water samples revealed limited intra-annual isotopic variability. Groundwater recharge was primarily driven by wet season precipitation, and spatially heterogeneous mechanisms were identified. The area was divided into three groundwater subsystems (southern groundwater (SGW), central groundwater (CGW), and northern groundwater (NGW)) based on isotopic (δ18O and d-excess) and hydrogeological characteristics. SGW demonstrated the rapid atmospheric recharge with minimal evaporation, producing northward flow and isotopically depleted signatures. CGW reflected the mixed recharge sources (precipitation, SGW, and river water) with moderate evaporation during the southwest-to-northeast transport. NGW was mainly recharged by Yellow River infiltration, exhibiting riverine isotopic characteristics. Over 80% of the samples maintained the “Excellent” quality year-round, while localized Cl−, SO42−, and NO3− contamination existed. The dry-season water quality was influenced by anthropogenic inputs, whereas wet-season changes were governed by internal hydrogeochemical processes. BI-LISA revealed seasonal EWQI-isotope coupling. During the dry season, rapid SGW recharge transported surface pollutants, whereas river-fed CGW maintained water quality. During the wet season, δ18O enrichment and quality degradation in CGW’s eastern Yufu River zone were attributed to runoff-induced pollution, whereas SGW exhibited dilution-driven quality improvement. This study demonstrated that BI-LISA effectively addressed interpolation limitations and mapped pollution-hydrology-spatial interactions in karst systems to support sustainable water resource protection.