Integrated Modeling of Surface-Subsurface Processes to Understand River-Floodplain Hydrodynamics in the Upper Wabash River Basin

Majority of the flood inundation maps in the United States are created using 1D steady-state hydraulic models that are unable to simulate the surface water-groundwater interactions within the floodplain that can play a major role in determining the water depth and extent of the flooding. Saturated conditions in subsurface and floodplain can lead to more severe inundation from low intensity but continuous flood events. In addition, these models are typically calibrated for design flow conditions (e.g., 100-year flow) by changing the roughness parameter. As a result, the calibrated roughness parameter value may or may not represent the actual roughness conditions within the floodplain during more extreme events. To understand the effect of subsurface and floodplain storage, this study combines hydraulic, hydrologic and groundwater modeling approaches for creating flood inundation maps. The methodology involves creating a large-scale 2D integrated surface-groundwater model for the Upper Wabash River Basin using ICPR and comparing the effect of subsurface storage and antecedent groundwater conditions on the overall floodplain dynamics by obtaining stage-duration and flow-duration relationships for multiple locations within the floodplain. The results for a 50-year return period storm event show that the integrated model more accurately captures the streamflow and flood depths and the traditional hydraulic model over-predicts the streamflow and flood depths as it does not account for subsurface storage. Moreover, the mesh resolution of the integrated model determines the accuracy in prediction of streamflow and stage at the outlet. In addition, the integration of sub-surface processes improves the accuracy of hydraulic models even without calibration which suggests that an integrated modeling approach reduces the uncertainty in flood prediction.