Application of the partial least square regression method in determining the natural background of soil heavy metals: A case study in the Songhua River basin, China

Machine learning-supported determination for site-specific natural background values of soil heavy metals

Heavy metal natural background values play a crucial role in distinguishing anthropogenic sources from natural sources to assess human impacts in polluted areas, thereby accurately formulating environmental policies. However, due to limitations imposed by human activities, research methods, and regional constraints, the determination of heavy metal background values, particularly on site or profile scale, is often challenging, highlighting the urgent need for new methodologies. To establish a comprehensive dataset containing heavy metal concentrations and soil properties, the study systematically collected and screened 82 soil profiles from areas minimally affected by human activities, resulting in a total of 2185 data sets. Using soil depth, pH, organic matter, weathering indices (SAF, BA), Fe2O3, MgO, Na2O, CaO, and K2O as model input variables, the predictive performance for site-specific background levels of Cd, Cr, Cu, Ni, Pb, and Zn was compared across four advanced machine learning models (RF (random forest), XGBoost (extreme gradient boosting), ANN (artificial neural network), SVR (support vector regression)). The results indicated that the optimal model for predicting background values of Cd, Cr, and Ni was XGBoost (MAE = 0.14 - 0.17; MSE = 0.04 - 0.06; R² = 0.82 - 0.87), while RF was used for Cu, Pb, and Zn (MAE = 0.01 - 0.18; MSE = 0.02 - 0.06; R² = 0.89 - 0.95). Importance assessments using RF and SHAP revealed that pH is a key controlling factor for Cd and Ni, Fe2O3 significantly impacts Cr, Cu, and Zn background levels, and K2O is the main controlling factor for Pb. The machine learning models developed can effectively predict the background levels of these six heavy metals based on major elemental and soil physicochemical properties, particularly achieving accurate predictions for Cu and Zn using just two input variables. This machine learning prediction framework is based on major elemental compositions and the physical/chemical properties of soil, enables precise and cost-effective point-to-point environmental assessments, thereby offering significant potential for practical applications.

Mapping of heavy metal pollution density and source distribution of campus soil using geographical information system

In this study, the pollution intensity, spatial distribution, and index-based risk distribution in campuses, which are a small prototype of cities, were mapped and the sources of heavy metals in the soil were investigated. Soil samples were taken from 9 different points from the Aksaray University Central campus, which was determined as the study area. It has been determined that the pH value in the collected soil samples varies between 8.7 and 11.0. This situation created an effect on reducing the accumulation and mobility of heavy metals in the soil. When the study area was evaluated based on the geo-accumulation index, Pb heavy metal was much denser in the places indicated as circulation areas and where students were actively present. Based on the pollution load index, it was concluded that 75% of the study area was moderately/highly polluted, and the rest consisted of unpolluted soils. Pearson correlation analysis and APCS-MLR analyses conducted to determine the source distribution showed that the contributions of natural sources, mixed sources of industrial and traffic activities, agricultural activity-based sources, and other sources were 57.49%, 21.44%, 12.67%, and 8.40%, respectively. Pb is mainly related to the mixed sources of industrial and traffic activities. Therefore, to clear up its long-term impact on the accumulation of heavy metals in the soil, it is important to conduct continuous heavy metal monitoring in the soil throughout the campus.

Source-specific probabilistic health risk assessment of heavy metals in surface water of the Yangtze River Basin

The detrimental effects of heavy metal accumulation on both ecosystems and public health have raised widespread concern. Source-specific risk assessment is crucial for developing effective strategies to prevent and control heavy metal contamination in surface water. This study aims to investigate the contamination characteristics of heavy metals in the Yangtze River Basin, identifying the pollution sources, assessing the risk levels, and further evaluating the health risks to humans. The results indicated that the average concentrations of heavy metals were ranked as follows: zinc (Zn) > arsenic (As) > copper (Cu) > chromium (Cr) > cadmium (Cd) > nickel (Ni) > lead (Pb), with average concentrations of 38.02 μg/L, 4.34 μg/L, 2.53 μg/L, 2.10 μg/L, 1.17 μg/L, 0.84 μg/L, and 0.32 μg/L, respectively, all below the WHO 2017 standards for safe drinking water. The distribution trend indicates higher concentrations in the upper and lower reaches and lower concentrations in the mid-reaches of the river. By integrating the Absolute Principal Component Scores-Multiple Linear Regression (APCS-MLR) receptor model and Positive Matrix Factorization (PMF) model, the main sources of heavy metals were identified as industrial activities (APCS-MLR: 41.3 %; PMF: 42.1 %), agricultural activities (APCS-MLR: 30.1 %; PMF: 27.4 %), and unknown mix sources (APCS-MLR: 29.1 %; PMF: 30.4 %). The calculation of the hazard index (HI) for both children and adults was <1, indicating no non-carcinogenic or carcinogenic risks. Based on the source-specific risk assessment, agricultural activities contribute the most to non-carcinogenic risks, while industrial activities pose the greatest contribution to carcinogenic risks. This study offers a reference for monitoring heavy metals and controlling health risks to residents, and provides crucial evidence for the utilization and protection of surface water in the Yangtze River Basin.