Assessing Uncertainty in Groundwater-Driven Human Health Risk Predictions

Abstract:

Uncertainty quantification is an important component in groundwater-driven human health risk assessment. Reliable estimation of adverse health effects on humans due to contaminated groundwater exposure requires knowledge from multiple fields such as hydrogeology and public health. In general, risk predictions are subject to a large degree of uncertainty. Uncertainty in risk analysis stems from the incomplete characterization of the subsurface formation, source zone identification, exposure variables and the human physiological response to contamination. For such reasons, human health risk assessment should be viewed as a multi-component stochastic system. Understanding the impact from each of these components in risk estimation can provide guidance for decision makers to better manage contaminated sites and best allocate resources towards minimal prediction uncertainty. This presentation aims in investigating the impact of aquifer heterogeneity in human health risk within a stochastic framework. Spatial variability of the subsurface properties can lead to the formation of preferential flow paths which control the plume spreading rates and travel time statistics, both which are critical in assessing the risk level and its corresponding uncertainty. The developed integrated hydrogeological-health stochastic model accounts for the key parameters characterizing the subsurface formation and engineering devices (e.g. volume of the sampling device). Through a series of examples, we illustrate how fundamental knowledge on the main physical mechanisms affecting solute pathways are necessary to understand the human health response to varying drivers. Finally we demonstrate how hydrogeological site characterization campaigns should be goal-oriented.