HR: 13:55h
AN: H12G-02 [PDF]
TI: Addressing Uncertainty in Predictive Groundwater Modeling in the Context of a Major Regulatory
Permitting Process
AU: * Carlson, C P
EM: christopher.carlson@dnr.state.wi.us
AF: Wisconsin Department of Natural Resources, Bureau of Waste Management
P.O. Box 7921, Madison, WI 53719
AU: Feinstein, D T
EM: dtfeinst@usgs.gov
AF: U.S. Geological Survey, Water Resources Division
8505 Research Way, Middleton, WI 53562
AU: Hunt, R J
EM: rjhunt@usgs.gov
AF: U.S. Geological Survey, Water Resources Division
8505 Research Way, Middleton, WI 53562
AU: Gotkowitz, M B
EM: mbgotkow@facstaff.wisc.edu
AF: Wisconsin Geological and Natural History Survey, 3817 Mineral Point Road, Madison, WI 53705
AU: Kenoyer, G J
EM: Galen.Kenoyer@rmtinc.com
AF: RMT, Inc., P.O. Box 8923, Madison, WI 53708
AB:
Regulatory assessments of the environmental impacts from proposed natural-resource development often require detailed
information regarding potential project effects on water resources. Understanding these impacts is important for both
disclosure to the public and determination of regulatory compliance in permitting. Many such projects involve pumping
substantial quantities of groundwater to access the desired resource; in some cases, nearby surface water resources may be
adversely affected by this groundwater withdrawal. In addition, some projects involve the management of waste materials that
have the potential to contaminate area groundwater and surface waters. Because the regulatory assessment needs to address
potential future conditions, predictive groundwater modeling has been used to help provide the necessary information. A case
study of a proposed mine in northern Wisconsin is the focus of this discussion. In this project, both the applicant and the
state regulatory agency review team recognized that substantial uncertainty would remain in model output even after
conducting detailed analyses of sizeable project data sets to develop reasonable model inputs and completion of a rigorous
calibration and an extensive sensitivity analysis. As a consequence, the modeling work and associated impact assessments
submitted by the applicant addressed uncertainty by incorporating two scenarios: an expected case and a worst case. The
agency review team, however, concluded that the level of uncertainty was so great in this situation, and the results of such
importance to the impact assessment and permitting processes, that a more extensive approach to addressing uncertainty was
necessary to ensure a technically defensible product. In response, the state's review team completed a number of detailed
analyses of the project area information and the applicant's 3D MODFLOW regional flow model. These included: development of a
2D analytic-element model (GFLOW) to evaluate the submitted MODFLOW model boundary conditions and recharge rates; completion
of Monte Carlo analyses to assess lake outlet flow and lakebed hydraulic conductivity; calculation of GFLOW-UCODE confidence
intervals for prediction of mine pumping rates and associated reductions in stream baseflow; and the development of detailed
analyses of project-site historical pump tests using both the regional and inset models. Numerous sensitivity analyses were
performed after completing updates to the submitted model, allowing the review team to identify and evaluate the most
sensitive parameters. This work on model uncertainty provided the review team's "reasonable" range for model inputs, which in
turn provided two end-member scenarios that established high and low estimates of the reasonable range of project effects on
the groundwater system for use in the regulatory process. The concept of reasonable ranges of model inputs and outputs was
also subsequently used in the review team's work on the solute transport models submitted by the applicant to assess
groundwater chemistry impacts from the proposed mine and waste facilities. This approach provides a means for communicating
uncertainty to both the public and regulatory decision-makers. Moreover, it underscores the uncertainty inherent in
simplified numerical representations of complex, real-world systems.
DE: 1800 HYDROLOGY
DE: 1803 Anthropogenic effects
DE: 1829 Groundwater hydrology
DE: 1831 Groundwater quality
DE: 1832 Groundwater transport
SC: Hydrology [H]
MN: 2003 Fall Meeting