Professional scientists and engineers are often tasked to predict the environmental
impact of human activities by combining the understanding of complex physical
systems with advanced analytical and numerical tools. At the same time, uncertainty
has always been a critical aspect of most engineering projects. In this paper, we
present a simple probabilistic framework that incorporates model parameter
uncertainty and translates the results to predictive uncertainty, as required for riskbased decision making. The framework is applied to three relatively simple, yet well
understood groundwater contaminant transport problems adapted from real-world
case examples. We use well-known analytical solutions documented in most
contaminant hydrogeology textbooks, coded using freely available tools. The
presented examples provide a useful illustration of the general methodology that
should be applied, irrespective of problem type, when data uncertainty needs to be
accounted for in design and decision-making. We have found the framework to be
well understood by water resource managers and well received by decision-makers.
Keywords: Groundwater modelling, Contaminant transport, Environmental
engineering, Model Uncertainty, Stochastic Modelling, Decision Support, Risk
Based Modelling
[ Download ]
