Simulation of the regional groundwater-flow system of the Menominee Indian Reservation, Wisconsin

A regional, two-dimensional, steady-state groundwater-flow model was developed to simulate the groundwater-flow system and groundwater/surface-water interactions within the Menominee Indian Reservation. The model was developed by the U.S. Geological Survey (USGS), in cooperation with the Menominee Indian Tribe of Wisconsin, to contribute to the fundamental understanding of the region’s hydrogeology. The objectives of the regional model were to improve understanding of the groundwater-flow system, including groundwater/surface-water interactions, and to develop a tool suitable for evaluating the effects of potential regional water-management programs. The computer code GFLOW was used because of the ease with which the model can simulate groundwater/surface-water interactions, provide a framework for simulating regional groundwater-flow systems, and be refined in a stepwise fashion to incorporate new data and simulate groundwater-flow patterns at multiple scales. Simulations made with the regional model reproduce groundwater levels and stream base flows representative of recent conditions (1970–2013) and illustrate groundwater-flow patterns with maps of (1) the simulated water table and groundwater-flow directions, (2) probabilistic areas contributing recharge to high-capacity pumped wells, and (3) estimation of the extent of infiltrated wastewater from treatment lagoons. The groundwater-flow model described in this report simulates the major hydrogeologic features of the modeled area, including surficial unconsolidated aquifers, groundwater/surface-water interactions, and groundwater withdrawals from existing high-capacity production wells. Areas contributing recharge to pumped high-capacity wells on the Menominee Indian Reservation were delineated by tracking simulated water particles from the water table to wells in combination with Monte Carlo techniques, and maps of the probability of capture for each well nest were produced. Groundwater-agebased areas contributing recharge to wells were simulated by using the calibrated set of parameters and porosity values adjusted to account for bias in simulated saturated thickness. Simulations were performed for current (2013) pumping rates. The simulations show a range in sensitivity of the simulated areas contributing recharge to wells given the parameters evaluated through the Monte Carlo analysis. The areas contributing recharge to supply wells for the villages of Zoar and Neopit are long and narrow, with a sharp gradation from high to low probability of capture. The areas contributing recharge to supply wells for Middle Village and the village of Keshena exhibit a sharp gradation from high to low probability over a relatively small area between the well and a local groundwater mound. The highest probability areas contributing recharge to the supply wells for the Villages of Onekewat and Redwing are in the immediate vicinity of the wells. These wells also have an extensive area with low probability for capturing water that is likely due to a locally low hydraulic gradient and the large degree of uncertainty associated with the lakebed resistance parameters that control interaction between groundwater and local lakes. Additional field investigations and associated local model refinements would facilitate further reductions in uncertainty associated with simulated areas contributing recharge to the wells. The likely extent of the Neopit wastewater plume was simulated by using the groundwater-flow model and Monte Carlo techniques to evaluate the sensitivity of predictive simulations to a range of model parameter values. Wastewater infiltrated from the currently operating lagoons flows predominantly south toward Tourtillotte Creek. Some of the infiltrated wastewater is simulated as having a low probability of flowing beneath Tourtillotte Creek to the nearby West Branch Wolf River. Results for the probable extent of the wastewater plume are considered to be qualitative because the method only considers advective flow and does not account for processes affecting contaminant transport in porous media. Therefore, results for the probable extent of the wastewater plume are sensitive to the number of particles used to represent flow from the lagoon and the resolution of a synthetic grid used for the analysis. Nonetheless, it is expected that the qualitative results may be of use for identifying potential downgradient areas of concern that can then be evaluated using the quantitative “area contributing recharge to wells” method or traditional contaminant-transport simulations.