Predicting long-term potential human health risks from contaminants in the multimedia environment requires the use of models. However, there is uncertainty associated with these predictions of many parameters which ca...Predicting long-term potential human health risks from contaminants in the multimedia environment requires the use of models. However, there is uncertainty associated with these predictions of many parameters which can be represented by ranges or probability distributions rather than single value. Based on a case study with information from an actual site contaminated with benzene, this study describes the application of MMSOILS model to predict health risk and distributions of those predictions generated using Monte Carlo techniques. A sensitivity analysis was performed to evaluate which of the random variables are most important in producing the predicted distributions of health risks. The sensitivity analysis shows that the predicted distributions can be accurately reproduced using a small subset of the random variables. The practical implication of this analysis is the ability to distinguish between important versus unimportant random variables in terms of their sensitivity to selected endpoints. This directly translates into a reduction in data collection and modeling effort. It was demonstrated that how correlation coefficient could be used to evaluate contributions to overall uncertainty from each parameter. The integrated uncertainty analysis shows that although drinking groundwater risk is similar with inhalation air risk, uncertainties of total risk come dominantly from drinking groundwater route. Most percent of the variance of total risk comes from four random variables.展开更多
基金Project supported by the National Natural Science Foundation of China (No.40572141).
文摘Predicting long-term potential human health risks from contaminants in the multimedia environment requires the use of models. However, there is uncertainty associated with these predictions of many parameters which can be represented by ranges or probability distributions rather than single value. Based on a case study with information from an actual site contaminated with benzene, this study describes the application of MMSOILS model to predict health risk and distributions of those predictions generated using Monte Carlo techniques. A sensitivity analysis was performed to evaluate which of the random variables are most important in producing the predicted distributions of health risks. The sensitivity analysis shows that the predicted distributions can be accurately reproduced using a small subset of the random variables. The practical implication of this analysis is the ability to distinguish between important versus unimportant random variables in terms of their sensitivity to selected endpoints. This directly translates into a reduction in data collection and modeling effort. It was demonstrated that how correlation coefficient could be used to evaluate contributions to overall uncertainty from each parameter. The integrated uncertainty analysis shows that although drinking groundwater risk is similar with inhalation air risk, uncertainties of total risk come dominantly from drinking groundwater route. Most percent of the variance of total risk comes from four random variables.
