Regional estimation of base flow and groundwater recharge in the Upper Mississippi river basin
Academic Article
Overview
Research
Identity
Additional Document Info
Other
View All
Overview
abstract
Groundwater recharge and discharge (base flow) estimates from two methods were compared in the Upper Mississippi River basin (USGS hydrologic cataloging unit 07). The Upper Mississippi basin drains 491,700 km2 in Illinois, Iowa, Missouri, Minnesota, and Wisconsin and outlets in the Mississippi River north of Cairo, Illinois. The first method uses the water balance components from the soil and water assessment tool model (SWAT). The model was used to simulate the daily water balance of approximately 16 soil/land use hydrologic response units (HRU) within each of the 131 USGS 8-digit watersheds. The water balance of each HRU is simulated with four storages: snow, soil (up to ten layers), shallow aquifer, and deep aquifer. Groundwater recharge is defined as water that percolates past the bottom soil layer into the shallow aquifer. Recharge is lagged to become base flow and can also be lost to ET. The second method consists of two procedures to estimate base flow and recharge from daily stream flow: (1) a digital recursive filter to separate base flow from daily flow and (2) a modified hydrograph recession curve displacement technique to estimate groundwater recharge. These procedures were applied to 283 USGS stations ranging in area from 50 to 1200 km2. A smoothed surface was obtained using a thin plate spline technique and estimates were averaged for each 8-digit basin. Simulated flow was calibrated against average annual flow for each 8-digit. Without further calibration, simulated monthly stream flow was compared against measured flow at Alton, Illinois (445,000 km2) from 1961-1980. To validate the model, measured and simulated monthly stream flow at Alton from 19811985 were compared with an R2 of 0.65. No attempt was made to calibrate base flow and recharge independent of total stream flow. Base flow and recharge from both methods were shown to be in general agreement. The filter and recession methods have the potential to provide realistic estimates of base flow and recharge for input into regional groundwater models and as a check for surface hydrologic models.
