Continuous watershed-scale hydrologic modeling of conservation practices for peak flow reduction

<p>Iowa first started seeing largescale changes to its landscape with the arrival of Europeans and ensuing conversion of forest and prairie to row crops and pasture and would see its landscape altered again in the early 1900s through the conversion of wetlands to row crops. Watersheds in Iowa, and the Midwest at large, have been drastically altered hydrologically—through land use change, tile drainage, digging of drainage ditches, and channelizing of meandering streams. Though drainage practices maximize arable land, they also induce higher flood peaks. Along with these practices, climate change also has the potential to increase flood peaks. Conservation practices —typically employed to reduce erosion and agricultural pollution—have been proposed to be used to reduce flood peaks, but little analysis has been done on their ability to do so at the watershed-scale.</p> <p>To quantify the impact implementing conservation practices at the watershed-scale has on flood peaks, a novel hydrologic model is run to simulate five conservation scenarios under both historic and increased precipitation continuously for seventeen years. The Generic Hydrologic Overland-Subsurface Toolkit (GHOST) is used to model the Boone River, an agriculturally dominated watershed in Iowa. The Boone River model is calibrated against the United States Geological Survey gaging station near the outlet of the watershed and achieves notable success. For the seventeen year study period from 2002 to 2018, calibration achieved a Nash Sutcliffe efficiency of 0.79, percent bias of -6.0 percent, and R2 of 0.80.</p> <p>To simulate the change from the baseline to a conservation practice, changes were made to the parameters of the baseline, calibrated model to reflect the effects of conservation practices. Scenarios run were the return of row crop acres to native vegetation, improved soil health via cover crops and no-till farming, distributed wetlands, conversion of river-adjacent row crop elements to native vegetation, and conversion of stream order one river-adjacent row crop elements to native vegetation. Results for the seventeen year study period show the average peak flow reductions simulated for the conservation scenarios are 82, 39, 9, 13, and 9 percent respectively for annual maximums and 75, 29, 10, 11, and 7.5 percent respectively for the peaks over the 2-year flood threshold. Of the five scenarios modeled, only native vegetation and cover crops were able to offset the increased precipitation anticipated from climate change.</p>