| A systematic uncertainty and variability analysis is necessary to enhance the quality of risk assessment, and to aid assessing how various possible decisions should be weighted. Three types of uncertainty, i.e. random, informal, and lexical uncertainty associated with contaminated sites, exposure assessment, and dose-response assessment are identified and their representations are examined. Probability theory and geostatistical methods are used to represent random uncertainty and site variability. Fuzzy set theory, possibility theory, random sets, and evidence theory are used to represent lexical uncertainty and informal uncertainty.;A fuzzy-stochastic partial differential model was developed to couple random and non-random uncertainties in the human health risk assessment process. When one variable is represented using one uncertainty theory, and another one is more suitably represented using another uncertainty theory, the propagation of hybrid uncertainties is necessary. Three methodologies, i.e., transforming a probabilistic parameter to a possibilistic parameter; a possibilistic parameter to a probabilistic parameter, and hybrid method were used for propagating hybrid uncertainties in this research. The methodology of propagating hybrid uncertainty and variability was illustrated using groundwater flow and contaminant transport modeling in a heterogeneous aquifer as a case study. Effects of dependencies amongst uncertain parameters on the distribution of contaminant concentrations at an exposure site were quantified by using evidence theory and the results were compared with that obtained using assumptions of independence.;A framework was developed to propagate uncertainty and variability from contaminated sites to disease endpoints through contaminant fate and transport models, exposure models, and dose-response models. Physiologically-based pharmacokinetic (PBPK) models were used to describe contaminant absorption, distribution, metabolism, and excretion from human body. PBPK models were constructed based on "mode of action (MoA)" of chemicals. Dose-response assessment based on active doses at target tissues significantly reduces model uncertainty. The total risk estimated using a point estimation method is higher by one to three orders of magnitude than the PBPK model simulation results.;A ranking system of contaminated sites based on comparative risk assessment methodology under uncertainty using fuzzy Preference Ranking Organization METHod for Enrichment Evaluation (PROMETHEE) was also developed. |
