Enhancing health risk assessment for soil heavy metal (loid)s using a copula-based monte carlo simulation method
Xiaohui Wang, Feng Xu, Yanjin Gui, Yafeng Liu
Abstract
In probabilistic health risk assessment (HRA) of multiple soil heavy metal(loid)s (HMs), Monte Carlo Simulation (MCS) typically ignores inter-element correlations during concentration simulation, potentially compromising risk estimation accuracy. In this study, a total of 109 surface soil samples collected in 2023 from a mining area in Chizhou, China, were analyzed for HM concentrations, which followed the order Zn > Cr > Cu > Pb > As > Ni > Hg > Cd. An enhanced HRA model by integrating Copula functions with MCS was employed to account for HM interdependencies. Among five tested Copula functions, the t Copula performed the best, effectively capturing the tail dependence and nonlinear associations between HMs. Risk assessment results showed children's 95th percentile hazard index (HI) increased by 9.2 % (from 3.170 to 3.461) and adult females' 95th percentile total carcinogenic risk (TCR) elevated by 11.8 % (from 1.19E-04-1.33E-04) when considering correlations, both aligning more closely with actual risk levels. Sensitivity analysis revealed that HM correlations redistribute risk contributions among elements, reducing the dominance of primary contaminants. This Copula-MCS integrated framework significantly enhances HRA reliability in polymetallic contamination scenarios, providing critical methodological support for soil risk management.