Velimir Vesselinov, Kay Birdsell and David Broxton, LANL
The presented work focuses on the regional aquifer located beneath the Los Alamos National Laboratory, New Mexico, USA. There are various alternatives conceptual models proposed in the literature that characterize the flow and transport in this aquifer. We performed a detailed comparison of the alternatives related to various conceptual model elements such as recharge, discharge, hydrodynamics, flow directions, medium anisotropy and heterogeneity. We rank the alternative conceptual models based on their plausibility. The plausibility is estimated based on how accurately they represent the observed quantitative data, qualitative knowledge about the system and the physics of governing processes. In addition, the plausibility is estimated based on the required complexity in the groundwater system for the conceptual model to be feasible. If alternative conceptual models describe the observed quantitative data and qualitative knowledge about the system equally well, the simplest conceptual model (requiring the least complexity and assumptions about the system) is the most probable, i.e. the parsimony principle. Alternative conceptual models are additionally ranked by the associated environmental risk. For example, a less probable model that predicts higher environmental risk of potential contaminant movement may be still ranked high. Then the alternative conceptual models are applied through numerical models to make predictions of potential contaminant transport and the impact of conceptual model uncertainty is evaluated using alternative statistical approaches.
2008 Ground Water Summit
