Multi-Scale Modeling of Groundwater to Protect Drinking Water Supplies

Recent environmental policy in the United States requires unprecedented insight into the subsurface environment. Many new laws have been passed to clean up existing contamination and protect water resources that require information about the physical and chemical properties of subsurface materials. The general principle behind each of these laws is that it is not possible to protect the environment from damage if we do not understand the physics, chemistry, and toxicology of the processes we are attempting to regulate. Most federal and state water pollution control legislation is based on our ability to determine impacts by estimating the effects of an environmental release. Increasingly, models are used to identify the range of probably impacts. Consequently, models of subsurface processes have become tools of public policy.

While policy objectives may always be quite clear (e.g., protect groundwater from pollution) it is less obvious how to assure the desired outcome. Groundwater flow and contaminant transport modeling can be used to gain insight into the environmental factors controlling the range of options available. However, there are several problems with using this particular class of models to guide pulic policy:

• Areas of interest often overlap. Consequently, the models need to "fit together" and ideally, the information should be consistent from one area to the next.

• Models are valid at a particular scale. There is always a risk that processes that are understood at one scale, are misrepresented at the scale of interest.

• Some models are in their formative stages of development. A great deal of research is currently being done to develop better theories of contaminant transport in the subsurface.

This dissertation is organized by the scale of the analysis. Modeling was used to assess the potential impacts of global climate change in a regional bedrock aquifer. In order to assess these impacts, new techniques were developed to build and interpret very large scale, supra-regional models. New model-building tools and practical techniques for calibration are presented. A set of basin-scale, regional models are used to demonstrate how the hydrogeological characteristics of a basin and surface water-groundwater interaction can affect hydrologic response to large-scale, long-term drought. New techniques were developed to nest a local groundwater flow model in a regional domain and used to evaluate how the conceptual model of flow through a seperating clay layer can change our interpretation of the vulnerability of the water supply to contamination. Not all releases present the same threat to drinking water. Numerical experiments were conducted using a saturated-unsaturated flow and transport model to create a drinking water contamination index. The index is developed that is based on the toxicity, mobility, and persistence of contaminants in the subsurface.

This thesis illustrates how the results of modeling subsurface processes can be used to inform public policy. In particular:

• I have developed a tool that allows the construction of locally refined analytic element models from regional AEM model components. This approach improves the use of AEM models by providing a reliable method for archiving the modeling analysis.

• I have investigated the potential for climate change to alter the availability of groundwater in regional aquifers in the relatively humid Midwest. This work used the power of the analytic element method to construct supra-regional flow models that could be constrained by base flow analysis of streams in the area of interest.

• I have used the AEM code GFLOW to evaluate a conceptual model of discrete leakage in a multiple-aquifer setting. Traditional techniques of modeling flow between adjoining aquifers assume that the properties of the system are relatively uniform. This work illustrates how discrete leakage could affect the risks posed contamination at the surface.

• I have used conceptual models of contaminant transport to develop a new groundwater contamination risk index. This index can be used to determine the risk posed by a release of any contaminant based on properties of the contaminant, the hydrologic setting and the release.

These developments push modeling of subsurface processes out of the purely technical arena so that the results can be used directly to inform public policy.