APPLICATION OF GLACIAL GEOLOGIC KNOWLEDGE TO THE HYDROSTRATIGRAPHY OF A REGIONAL FLOW MODEL IN CENTRAL WISCONSIN

Regional groundwater flow models and their conceptual models are underpinned by an understanding of regional geology. We applied this idea to a groundwater flow model in central Wisconsin, an area with numerous irrigation wells and seepage lakes. The overall project goal was to estimate the effect of irrigation pumping on the lake levels and ecology. A large regional model was constructed to serve as a parent model for child inset models around three lakes of interest. The glacial geologic history provided the framework for constructing the large regional model hydrostratigraphy.

The model domain has different glacial geologic settings and landforms including two glacial lakes, an outwash plain, tunnel channels, end moraines, ground moraines, and hummocky stagnant ice topography and sediments. Each of these settings results in different hydrogeologic properties. For example, the outwash plain is mostly sand and gravel sediments and has a high hydraulic conductivity and is relatively homogeneous. One of the glacial lakes has three distinct sediment packages: a coarse-grained aquifer over a fined-grained aquitard over another coarse-grained aquifer. These individual units are each homogenous. The other glacial lake is mostly fine-grained sediment. In contrast to the mostly homogenous units, the hummocky stagnant ice left a heterogeneous mix of sediments, both coarse- and fine-grained over a wide range of scales. The other regions were calibrated using pilot points over the homogeneous region. For the heterogeneous unit, we applied geostatistics to incorporate the sediment and hydraulic conductivity variation with both depths and extent.

The different depositional settings provided clues to how to treat the different units in the model. Some units were mostly homogeneous while others were heterogeneous. It also provided guidance for setting the boundaries between the different units. The incorporation of geologic knowledge maintained the conceptual model and assisted with model calibration beyond what would have been achieved by geostatistics alone.