This study aimed to analyze the variations in groundwater quality, vulnerability and potential health risk from 2006 to 2016 in the Limin District, Harbin, China. Groundwater geochemical characteristics were described using statistical analysis and Piper diagrams. A modified DRASTIC model that combined factors of intrinsic aquifer vulnerability and land use was applied to assess groundwater vulnerability. The weights of parameters were adjusted by using the analytic hierarchy process (AHP) to optimize the model. The non-carcinogenic health risk was estimated by the Unites States Environmental Protection Agency (USEPA) model. Results suggested that concentrations of NH₄-N, Fe and Mn in groundwater exceeded the limits both in 2006 and in 2016. The concentration of Fe in the groundwater showed more significant variation between 2006 and 2016 than the other parameters. Very high vulnerability zones increased from 6.3% in 2006 to 16.9% in 2016, and distributed on agricultural land, indicating that agriculture was still a major source of pollution. Mn and NO₃-N contributed the most to human health risks in 2006 and 2016, respectively. This study highlights the influence of groundwater quality variation in decadal exploitation on human health.

Soil element background concentration is foundational data for environmental quality assessment, contamination diagnosis, and sustainable land management. However, existing investigation-based methods are time-consuming and inefficient. The machine learning (ML) method has demonstrated excellent performance in predicting soil heavy metal concentration. In this study, based on the nine environmental variables of soil formation from 210 soil monitoring points, including elevation, pH, organic matter, soil type, parent material, plant cover, land use type, topography, and soil texture, decision tree (DT), random forest (RF), extreme gradient boosting (XGB), and support vector machine (SVM) models were used to predict the eleven soil element background concentrations. Among them, SVM and RF models could be used for an effective prediction of the background concentration of all soil heavy metals. Compared with the XGBoost and DT, the SVM for all heavy metals except for cadmium (Cd) and manganese (Mn) performs best. Although the key factors affecting background concentrations vary among different soil elements, organic matter, soil type, and altitude, they play a crucial role in the accurate prediction of soil element background concentration. This study provides simple and efficient ML models for predicting soil element background concentration at the large regional scale. The results of this study can be utilized to distinguish natural geochemical processes from human-induced pollution.

The knowledge of PAH sources in soil plays an important role in making the decision of pollution emission restriction, as well as toxic information. The present study developed a modified receptor model to apportion the contribution of various sources to PAHs and associated toxic. PMF model was selected as the basic source apportionment model to identify sources and apportion their contributions. Toxic assessment models, including carcinogenicity assessment model and mutagenicity assessment model, were then used with PMF to obtain toxic-oriented source contribution. To attain this objective, 16 priority PAHs in the soil around Miyun Reservoir were selected as a case study. The results showed that the dominant source of soil PAHs was wood/biomass combustion (31.02%), followed by vehicular source (27.17%), petroleum source (21.27%) and coal combustion (20.53%). While the highest contributor to carcinogenicity and mutagenicity was a vehicular source, with a proportion of 36.61% and 41.46%, respectively. This study highlights the combined utilization of the source apportionment model and toxic assessment model. The approach was valuable for providing comprehensive source-oriented risk information, and expanding the application of both PMF and toxic assessment.