Limitations on the study of tile drainage using a distributed parameter hydrologic model and stable isotopes

Recent large-scale flood events in Iowa demonstrate a need to understand how drainage alteration, specifically through tile drainage and ditches related to intensive agriculture, affect peak flow events and streamflow hydrology. This research tested the hypothesis that there is a relationship between tile drainage and peak flow events, specifically that tile drainage increases the occurrence of peak flow events. The first objective was to implement a physically based, coupled surface water/groundwater model (HydroGeoSphere -- HGS) in the South Fork watershed in north-central Iowa. The second objective was to sample and analyze precipitation, surface water, groundwater, and tile water for stable isotopes from August 2011 to July 2013 in order to implement and model isotopic hydrograph separation during a peak flow event. The intent of that work was to delineate the source of water in the flood. Due to the inability of the HGS model to converge, the model was abandoned and efforts were refocused on using stable isotopes to predict water sources during peak flow events. The drought of 2012 provided an opportunity to investigate how tiles interact with other hydrologic components during a hydrologic extreme. Results suggest that tile drainage water---which is often a mixture of surface intakes water, saturated macropore flow, and groundwater---was fed exclusively by groundwater during the drought. In contrast, tile water is a combination of these sources under normal hydrologic conditions. During a precipitation and flood event in late May 2013 water pooled around surface intakes, stream discharge reached two-year peaks, and the tile water isotopic composition moved significantly away from groundwater towards the composition of precipitation. These results suggest that surface runoff to the tiles would be the dominant component fed to streams during peak flow, and that isotopic hydrograph separation may prove useful when peak flow occurs again. Although HGS was not useful for simulating peak flow and hydrograph separation in this research, it will be a useful tool, along with stable isotopes, to continue investigations of this type in the future.