Lake Michigan Basin Groundwater Flow Model: Potential Approaches and Challenges in Applying a Basin-Scale Model to Local-Scale Questions

Monday, April 12, 2010
Continental Foyer (Westin Tabor Center, Denver)
Christopher J. Hoard , USGS Michigan Water Science Center, U.S. Geological Survey, Lansing, MI
Howard W. Reeves , USGS Michigan Water Science Center, U.S. Geological Survey, Lansing, MI
Daniel T. Feinstein , U.S. Geological Survey, Milwaukee, WI
Hendrik Haitjema , Haitjema Consulting Inc, Bloomington, IN
Michael Fienen, Ph., D. , Wisconsin Water Science Center, United States Geological Survey, Middleton, WI
Randall Hunt, Ph., D., P., H. , Wisconsin Water Science Center, United States Geological Survey, Middleton, WI
Basin-scale groundwater flow models are powerful tools to evaluate regional groundwater issues and explore groundwater budgets on a regional scale.  Often, however, questions are raised that require local-scale analysis.  Can basin-scale models inform local-scale decisions?  The regional model results provide a broad picture of the resource, but the size of the numerical grid used for regional models is not appropriate to address local questions.  In particular, the water table response to pumping is severely limited in regional models because of potentially dense surface-water boundary conditions, and, surface-water features that may be important on the local scale may not be well characterized in regional models.

            Two approaches developed to analyze groundwater/surface-water interaction at local scales derived from a basin-scale model are discussed.  The first approach uses a hybrid analytical-element/MODFLOW model.  In this approach the upper layer(s) of a regional MODFLOW model are replaced by GFLOW, an analytical element model, which is gridless and provides for fine resolution of surface-water/groundwater interaction.  The second approach uses telescoping grid refinement and local grid refinement techniques to create an inset MODFLOW model within the basin-scale model that has grid spacing at the appropriate scale to simulate groundwater/surface-water interaction.  Both approaches allow the local-scale and basin-scale models to interact, and both will be shown to produce reasonable simulation of groundwater/surface-water interaction.  Direct application of parameters from the regional model, however, may limit the accuracy and reliability of the local-scale model.  Structural errors and unreasonable parameterization for the local-scale problem must be addressed for effective use of the regional model to address local questions.