Scenario and Parametric Uncertainty in GESAMAC: A Methodological Study in Nuclear Waste Disposal Risk Assessment

We examine a conceptual framework for accounting for all sources of uncertainty in complex prediction problems, involving six ingredients: past data, future observables, and scenario, structural, parametric, and predictive uncertainty. We apply this framework to nuclear waste disposal using a computer simulation environment--- GTMCHEM---which "deterministically" models the one-dimensional migration of radionuclides through the geosphere up to the biosphere. Focusing on scenario and parametric uncertainty, we show that mean predicted maximum doses to the man due to I \Gamma 129 and uncertainty bands around those predictions are larger when scenario uncertainty is properly assessed and propagated. We also illustrate the value of a new method for global sensitivity analysis of model output called extended FAST . Key words: Bayesian prediction, extended FAST, Level E/G test case, parametric uncertainty, scenario uncertainty, sensitivity analysis. 1 School of Mathematical Sciences, Unive..