Developing Hydrogeological Site Characterization Strategies based on Human Health Risk

In order to provide better sustainable groundwater quality management and minimize the impact of contamination in humans, improved understanding and quantification of the interaction between hydrogeological models, geological site information and human health are needed. Considering the joint influence of these components in the overall human health risk assessment and the corresponding sources of uncertainty aid decision makers to better allocate resources in data acquisition campaigns. This is important to (1) achieve remediation goals in a cost-effective manner, (2) protect human health and (3) keep water supplies clean in order to keep with quality standards. Such task is challenging since a full characterization of the subsurface is unfeasible due to financial and technological constraints. In addition, human exposure and physiological response to contamination are subject to uncertainty and variability. Normally, sampling strategies are developed with the goal of reducing uncertainty, but less often they are developed in the context of their impacts on the overall system uncertainty. Therefore, quantifying the impact from each of these components (hydrogeological, behavioral and physiological) in final human health risk prediction can provide guidance for decision makers to best allocate resources towards minimal prediction uncertainty. In this presentation, a multi-component human health risk-based framework is presented which allows decision makers to set priorities through an information entropy-based visualization tool. Results highlight the role of characteristic length-scales characterizing flow and transport in determining data needs within an integrated hydrogeological-health framework. Conditions where uncertainty reduction in human health risk predictions may benefit from better understanding of the health component, as opposed to a more detailed hydrogeological characterization, are also discussed. Finally, results illustrate how different dose-response models can impact the probability of human health risk exceeding a regulatory threshold.