Zheming Shi

Dongping Lake is a key regulating and storage lake on the Eastern Route of the South‐to‐North Water Diversion Project, and characterizing the hydrochemistry of groundwater around the lake is essential for safeguarding regional water security. We collected 70 groundwater samples in the Dongping Lake area and used a combination of statistical analysis, self‐organizing maps (SOMs), Piper and Gibbs diagrams, ion ratios, and positive matrix factorization (PMF) to identify the hydrochemical characteristics and controlling mechanisms of the groundwater. The groundwater is weakly alkaline and oxidizing, with HCO ₃ ‐Ca as the dominant hydrochemical type. Freshwater prevails, but anthropogenic impacts are already detectable. SOM clustering resolved four groundwater groups: natural background water, transitional water, agriculturally polluted water, and mixed polluted water. Water–rock interactions—chiefly the weathering and dissolution of silicate, carbonate, and evaporite minerals—exert the primary control on hydrochemical composition, with cation exchange also playing an important role. Evaporative concentration and precipitation infiltration have comparatively weak effects. PMF further identified four source factors: natural geological background, urban domestic pollution, carbonate and evaporite dissolution, and agricultural nonpoint source pollution. These findings illuminate the mechanisms driving groundwater chemical evolution under the combined influence of natural processes and human activities and provide a basis for regional groundwater protection, pollution control, and sustainable utilization.

Terrestrial legacy heavy metal (HM) contamination may threaten riverine drinking-water sources, yet its groundwater-mediated transfer remains poorly resolved. To elucidate this process, metal speciation, groundwater hydrogeochemistry, and 222Rn-based discharge quantification were combined with source apportionment in an industrialized drinking-water source area along the middle and lower reaches of the Yangtze River. Solid-phase As and Cd were enriched, reaching 18.5 and 28.0 times their risk screening limits, but showed contrasting mobility: As occurred predominantly in the residual fraction (95.6%), whereas Cd was dominated by non-residual fractions (93.2%) and contributed most strongly to ecological risk. Groundwater contamination was dominated by Fe and Mn, with localized As, Zn, and Ni enrichment. Weakly reducing conditions and As-Fe-Mn co-enrichment suggested secondary As mobilization associated with reductive dissolution of Fe/Mn oxides. River-water composition corresponded more closely to Cu-rich near-river groundwater than to Ni-rich river-distal groundwater. This correspondence, together with 222Rn-derived groundwater discharge of 3.75 mm/d, supported groundwater-mediated transfer of legacy contamination to the Yangtze River. Despite compliance with drinking-water concentration limits, river-water As was associated with elevated potential carcinogenic risk under long-term exposure. These findings highlight the roles of metal speciation, hydrogeochemical mobilization, and groundwater transport in determining the threat of legacy HM contamination to riverine drinking-water sources.

Groundwater contamination by nitrate (NO ₃ ⁻ ) and fluoride (F ⁻ ) is pervasive, yet the coupled mechanisms governing their evolution along flow paths remain poorly understood. Most studies examine them separately, neglecting their potential to serve as complementary tracers of anthropogenic and natural processes. Here, we propose a coupled SOM‐PMF framework to identify the sources and co‐evolution of NO ₃ ⁻ and F ⁻ in the Ganjiang River Basin. Self‐organizing maps (SOM) applied to 67 field samples resolved a five‐stage hydrochemical sequence from the recharge zone (C1) to the confined zone (C5), capturing both systematic trends and anomalous sites. NO ₃ ⁻ peaked at 14.42 mg/L in the transition zone (C3), reflecting intense local anthropogenic inputs, whereas F ⁻ reached a maximum of 6 mg/L in the confined zone (C5), associated with strongly alkaline conditions at sites such as XF043. Positive matrix factorization (PMF) then quantified the factors controlling this evolution. NO ₃ ⁻ was dominated by anthropogenic factor F4 (78.3%), tracking agricultural and domestic inputs and their attenuation under reducing conditions. F ⁻ enrichment was driven primarily by the dissolution of fluoride‐bearing minerals under alkaline conditions (F3, 76.1%) and further amplified by evaporative concentration in the confined zone (F5, 14.5%). Although NO ₃ ⁻ and F ⁻ show no direct statistical correlation, together they form a complementary tracer system that distinguishes the combined influence of anthropogenic and natural processes. The coupled SOM‐PMF approach provides a transferable, data‐driven tool for source apportionment and risk zoning of multi‐contaminant groundwater system.

To address the challenges of identifying mine water inrush sources and the low efficiency of risk control under complex karst hydrogeological conditions in the Beiya Gold Mine, Yunnan, this study proposes an intelligent identification model integrating nonlinear feature extraction and intelligent parameter optimization. Utilizing 42 sets of measured water samples (comprising karst springs, surface water, and solution caves), a coupling identification model was constructed based on 11-dimensional features including hydrochemical indices and hydrogen–oxygen isotopes. The model employs Kernel Principal Component Analysis (KPCA) to extract discriminative low-dimensional features from nonlinear data, while the critical parameters of the Support Vector Machine (SVM) are optimized via an Improved Sparrow Search Algorithm (ISSA) to enhance generalization performance. The results demonstrate that the following: (1) the proposed model achieves an identification accuracy of 91.7% on the independent test set, significantly outperforming benchmark models such as RF and standard SVM; (2) three sets of comparative experiments indicate that the fusion of multi-source features yields superior identification performance compared to single-source inputs; and (3) SHAP (shapley additive explanation) interpretability analysis reveals that HCO₃⁻, Mg²⁺, Ca²⁺, and F⁻ are the core discriminative factors, with their contribution patterns aligning closely with the hydrogeochemical evolution mechanisms of the mining area. This model achieves a synergy between high-precision identification and mechanical interpretability, providing reliable technical support for water disaster prevention in karst mining areas.

While large earthquakes can alter groundwater systems far from the epicenter, their hydromechanical changes remain elusive. We investigate the 2011 Mw 9.0 Tohoku earthquake's impact on a deep well‐aquifer system ∼2,000 km from the epicenter. By employing the groundwater tidal and barometric pressure response methods, we perform tidal response analysis combined with Markov chain Monte Carlo (MCMC) inversion and numerical simulations in the MOOSE. The results indicate that permeability and the vertical leakage coefficient increased by one to two orders of magnitude, whereas the ability of the rock to store water (porosity, specific storage, and bulk compressibility) decreased by 20%–80%. These contrasting changes suggest enhanced hydraulic connectivity accompanied by reduced storage capacity. We interpret them as the result of simultaneous compaction (from volumetric strain) and fracturing (from shear strain) within the deep rock matrix. The study demonstrates that distant megaquakes can cause long‐lasting modifications on deep groundwater systems.