Abstract
Long-term leaching from historical tailings sites causes persistent soil contamination, posing direct health risks to exposed populations. However, accurate health risk assessment (HRA) remains challenging due to substantial uncertainties and a lack of source-specific precision. This study integrates two-dimensional Monte Carlo simulation (2D-MCS), absolute principal component score-multiple linear regression (APCS-MLR), and the HRA model, and applies this combined approach to a typical pyrite tailings area in Southwest China, with the aim of systematically quantifying the health risks posed by potentially toxic elements (PTEs) in soil. Compared to conventional one-dimensional MCS, 2D-MCS significantly improved assessment accuracy by synergistically accounting for parameter variability and uncertainty. The refined assessment revealed that non-carcinogenic risks for all populations are negligible, but the overall carcinogenic risk is at a cautionary level, with chromium and arsenic identified as key risk-contributing elements. Source apportionment coupled with 2D-MCS further enabled source-specific risk quantification. Among the three major sources identified, tailings pollution should be prioritized for health risk management and control, followed by a mixed source of atmospheric deposition and traffic emissions, which also requires attention. This study optimizes HRA methodology for soil PTEs by enhancing precision and reducing uncertainty. The findings provide more reliable scientific support for risk-based decision-making, thereby improving resource allocation efficiency and strengthening protection for sensitive populations.