Optimizing health risk assessment for soil trace metals under low-precision sampling conditions: A case study of...

Cost limitations often lead to the adoption of lower precision grids for soil sampling in large-scale areas, potentially causing deviations in the observed trace metal (TM) concentrations from their true values. Therefore, in this study, an enhanced Health Risk Assessment (HRA) model was developed by combining Monte Carlo simulation (MCS) and Empirical Bayesian kriging (EBK), aiming to improve the accuracy of health risk assessment under low-precision sampling conditions. The results showed that the increased sampling scale led to an overestimation of the non-carcinogenic risk for children, resulting in potential risks (the maximum Hazard index value was 1.08 and 1.64 at the 500 and 1000 m sampling scales, respectively). EBK model was suitable for predicting soil TM concentrations at large sampling scale, and the predicted concentrations were closer to the actual value. Furthermore, we found that the improved HRA model by combining EBK and MCS effectively reduced the possibility of over- or under-estimation of risk levels due to the increasing sampling size, and enhanced the accuracy and robustness of risk assessment. This study provides an important methodology support for health risk assessment of soil TMs under data limitation.