To assess and predict the Nansi Lake soil pollution risk, we evaluate the soil environmental quality in the Nansi Lake region using machine learning techniques, combined with the SHapley Additive exPlanations (SHAP) model for interpretability. The primary objective was to predict the level of soil pollution caused by heavy metals, incorporating the traditional Pollution Load Index (PLI) and Potential Ecological Risk Index (PERI) methods. Through the integration of statistical characteristics, PLI, and PERI evaluations, a new assessment method was created, categorizing soil pollution into “Class0—no risk”, “Class1—low risk”, and “Class2—high risk”. Various machine learning models, including Support Vector Machine (SVM), Decision Tree Classifier (DT), Random Forest (RF), and XGBoost, were employed to predict the soil quality based on these indices. XGBoost demonstrated the highest accuracy, achieving a prediction accuracy of 93%. SHAP analysis was further applied to explain the machine learning model and determined that the accumulation of key pollutants such as cadmium (Cd) and mercury (Hg) may significantly produce soil pollution risk, and targeted management needs to be developed for these pollution features.

This study analyzed the sources of total metal elements using the positive matrix factorization (PMF) model and conducted human health risk assessment for adults and children using the health risk assessment model recommended by the United States Environmental Protection Agency (USEPA). According to the health risk assessment, As is the main contributor to the non-carcinogenic risk of groundwater in Tengzhou, with drinking water as the main exposure route. Regarding carcinogenic risks (CR), the values of As and Cr for adults and children were higher than 1 × 10⁻⁴, with drinking water as the main exposure route. Therefore, As is the largest contributor to the CR of groundwater for adults and children and drinking water is the main exposure route in the study area. The primary exposure pathways are oral intake and dermal contact, with oral intake presenting a significant risk. The carcinogenic risks according to principal component analysis (PCA) and PMF analysis showed that the main sources of heavy metals in shallow groundwater in Tengzhou City are agricultural, industrial, natural, and industrial deposition sources, with contribution rates of 21.7%, 27.2%, 31.0%, and 20.1%, respectively. In particular, natural sources are the largest contributor to the accumulation of heavy metals.

Groundwater in the Nansi Lake Basin, a key reservoir of the South-to-North Water Diversion Project, supports domestic, agricultural, and ecological needs but faces pressure from overexploitation and pollution. This study clarifies the hydrochemical characteristics, controlling processes, environmental quality, and source contributions of shallow groundwater in the basin. Hydrochemical data from 67 wells were interpreted using Piper and Schukalev diagrams, Gibbs and ion-ratio plots, the entropy-weight water quality index (EWQI), and an absolute principal component scores–multiple linear regression (APCS-MLR) model. Groundwater shows high mineralization and hardness, with 35.82% and 55.22% of samples exceeding standard limits for total dissolved solids and total hardness, respectively. The dominant facies are HCO₃-Ca, HCO₃-Ca·Mg, and HCO₃·Cl-Na·Ca, indicating dissolution and ion exchange involving carbonate and silicate rocks. Gibbs and ion-ratio analyses demonstrate that rock–water interaction is the main control, with secondary influence from evaporation. EWQI results indicate generally good groundwater quality (68.66% Class I, 20.90% Class II). APCS-MLR identifies natural, agricultural, ion-exchange, and anthropogenic sources, contributing 53.34%, 22.71%, 4.79% and 19.14%, respectively. These findings show that protection should focus on pollution control in northern agricultural and mining zones while conserving high-quality groundwater elsewhere in the basin.