资江流域下游石煤矿周边农田土壤Cd含量空间分异特征及影响因素分析OA

This study aims to explore the spatial variation characteristics of cadmium(Cd)pollution in farmland soils around a typical stone coal mine.The methods of geostatistics,geographical detectors,and machine learning were employed to analyze the spatial variation characteristics and influencing factors of Cd pollution in a typical stone coal mining area and surrounding farmland in the lower reaches of the Zijiang River Basin.The results showed that the average Cd content in the soil of the study area was 1.99 mg/kg,and 73.48%of the sampling points had the Cd content exceeding the risk screening value,which indicated severe pollution.The soil Cd exhibited strong spatial autocorrelation,and its spatial distribution showed obvious anisotropy,with the east-west direction being the main variation direction.The high-concentration areas were highly consistent with the historically dense distribution areas of mine adits.The soil Cd content in the study area was jointly affected by anthropogenic factors and natural factors.Moreover,the interactions between factors significantly enhanced the explanation for spatial variations of soil Cd content.The XGBoost model combined with SHapley Additive exPlanations(SHAP)identified pH value,distance from mine adit,slope,soil organic matter,distance from railway,and distance from waterway as key influencing factors.