Date of Award

May 2022

Degree Type


Degree Name

Doctor of Philosophy



First Advisor

Woonsup Choi

Committee Members

Alison Donnelly, Neal O'Reilly, Shangping Xu


Baseflow, Groundwater, High-capacity wells, MODFLOW, Regression model, SWAT




bySusan Borchardt The University of Wisconsin-Milwaukee, 2022 Under the Supervision of Professor Woonsup Choi

In Wisconsin, the number of high-capacity wells has increased substantially, and concerns have been raised about their impact on both groundwater levels and streamflow. At the same time Wisconsin’s climate has been changing, and both the annual precipitation (5%) and temperature (1.5oC) have been trending upward over the last 68 years and both are predicted to increase into the future. This study attempted to demonstrate the simultaneous effects of climate change, physical basin changes and changes to the groundwater withdrawal rate from high-capacity wells by employing both analytic methods and simulation models. Linear regression was used to determine how variables representing climate, land use, soil characteristics, and groundwater withdrawals would affect baseflow variability. While double-mass curve analysis was used to find that twenty out of the thirty-five basins studied exhibited a deviation in the slope when the cumulative value of precipitation was plotted against the cumulative value of baseflow suggesting an anthropogenic variable was affecting baseflows, most likely groundwater withdrawals. Panel data analysis (PDA) was then used to evaluate the simultaneous effects of climate, withdrawal rate, and physical basin variables on baseflow variability across the state. The PDA found that the climate variables (precipitation and temperature) were significant in explaining the temporal variability of baseflow, whereas land use and the drainage conditions were important in explaining the spatial variability of baseflow as expected. But the groundwater withdrawal was not, which was not expected. The Soil & Water Assessment Tool (SWAT) and the United States Geological Survey’s Modular Hydrologic Model (MODFLOW) were used to simulate changes in hydrology in a single basin in the state that has experienced declining baseflows over the last 30 years but steady population numbers and land use percentages over the same period. Using the variables found to affect baseflow (climate, land use, and soil characteristics), SWAT was used to simulate the change in the recharge rate, and then using the withdrawal rate from high-capacity wells, MODFLOW was used to simulate the change in hydraulic head. The SWAT model predicted that future increases in Wisconsin’s annual precipitation of approximately 7.5% will cause increases in both groundwater recharge (16.74%) and streamflow (14.13%). The future increases in temperature of 2.2–3.3oC by the middle of the 21st century, however, are predicted to leave the state with a net reduction in both streamflow (–23.39%) and groundwater recharge (–19.63%). In addition, the MODFLOW model predicted a mean head elevation decrease of over 2 meters due to increased predicted temperatures, this is despite predicted increases in annual precipitation and an additional decrease in groundwater elevation surrounding high-capacity wells due to predicted increases in annual withdrawal rate. Overall, analytically the most important variable in determining the variability of streamflow is the amount of annual precipitation, but the modeling portion of the study showed that the predicted increases in temperature will ultimately lead to decreases in the available fresh water in the study area. This study also highlights that if the escalating use of irrigation for Wisconsin’s agriculture outpaces the increases in annual precipitation, declines in stream baseflow will result. The study also predicted that some of these decreases can be mitigated by abandoning just a select number of high withdrawing wells.

Keywords: Baseflow, Groundwater, High-capacity wells, Double-mass curve analysis, SWAT, MODFLOW, Streamflow, Aquifer, Regression model