STORM EVENT FLOW AND SEDIMENT SIMULATIONS IN AGRICULTURAL WATERSHEDS USING DWSM

Abstract

DWSM, the dynamic watershed simulation model, was expanded with a subsurface and a reservoir flow routingschemes. The hydrology and sediment components of the model were applied to three agricultural watersheds in Illinois, BigDitch (100 km2), Court Creek (250 km2), and Upper Sangamon River (2,400 km2), to simulate spatially and temporallyvarying surface and subsurface storm water runoff, propagation of flood waves, upland soil and streambed erosion, andsediment transport; to evaluate these simulation capabilities through calibration and validation; and to conduct variouswatershed investigative analyses. The new schemes were selected from the literature. DWSM was able to simulate the majorhydrologic, soil erosion, and sediment transport processes, and generate reasonable water and sediment discharges in theBig Ditch and Court Creek watersheds, considering complexities of the physical processes simulated and sizes of the drainageareas evaluated. Comparisons of predicted and observed sediment discharges during recession portions of the hydrographswere much better in Court Creek watershed than in Big Ditch because of depth-integrated observation samples in the former,which is necessary during recession and low flow periods when pronounced concentration gradients are expected. Somediscrepancies in model predictions were found, which may be due to limitations of the model, especially its single-event natureand lack of backwater simulation, limitations and uncertainties of input data, and temporally constant values of inputparameters. Addition of the subsurface flow (tile drain and base flow combined) routing scheme improved predictions of therecession and base flow portions of the subwatershed (100 to 290 km2) hydrographs in the Upper Sangamon River watershed.Significant improvements were noticed in larger subwatersheds. Scaling effect investigations on the Big Ditch watershedshowed different overland Mannings roughness coefficients, effective lateral saturated hydraulic conductivities, and flowdetachment coefficients for a coarser and a finer representations (subdivisions) of the watershed. These input parametersrequired recalibration when watershed subdivision sizes were altered. After recalibration, simulated water and sedimentdischarges were approximately the same for both representations. DWSM provided a robust tool in ranking overland planesand channel segments in the Court Creek watershed based on comprehensive criteria for flooding and sediment productionpotentials. The rankings were useful to stakeholders in prioritizing critical parts of the watershed and planning restorationand education programs. The model also provided a robust tool for evaluating detention basins in controlling downstreamwater and sediment discharges, although evaluations on sediment discharges were limited to large basins.