Runoff from open feedlots has the potential to cause degradation of surface and groundwater if handled improperly. Due to this pollution potential, the United State Environmental Protection Agency (US EPA) regulates runoff control systems on concentrated animal feeding operation (CAFO) sized feedlots. For the first time, the 2003 effluent limitation guidelines allowed consideration of alternative manure treatment systems for National Pollutant Discharge Elimination System (NPDES) permitted CAFO operations. Concentrated animal feeding operations that utilize alternative manure treatment systems under an NPDES permit are required to demonstrate, through modeling, that their alternative runoff control system had an equal or lesser nutrient mass release than a conventional manure management system would. This permitting requirement renewed interest in the modeling of traditional containment systems and generated interest in modeling "alternative technology" systems. One possible "alternative technology" systems being considered are vegetative treatment systems (VTS). A VTS is defined as a runoff control system that uses a series of treatment components, at least one of which uses vegetation, to treat open lot runoff. In particular, much of the VTS research thus far has focused on vegetative treatment areas (VTA's). A VTA is an area planted to permanent vegetation that reduces pollutant transport via sedimentation, filtration, and infiltration of the feedlot runoff. This modeling requirement led to the development of the Iowa State University (ISU) - Effluent Limitations Guidelines (ELG) model and the ISU - VTA model, which predict the performance of traditional containment systems and of vegetative treatment area systems, respectively.

This thesis reviews the accuracy of the ISU-ELG model by comparing the modeled runoff control performance of a traditional containment system to that predicted by the Soil-Plant-Air-Water (SPAW) model. Specifically, the criterion used to determine if a particular day is a "dewatering day," i.e., suitable for land application of basin effluent, is investigated to determine its effect on basin performance, with the objective of verifying that the ISU-ELG model is providing a reasonable prediction of the runoff control provided by a containment basin in Iowa.

The ISU-ELG model is based on a model developed by Koelliker et al. (1975) to predict the performance of a holding basin at controlling feedlot runoff and uses a set of general criteria to determine if land application is acceptable, while the SPAW model uses a soil moisture criterion to determine if conditions are acceptable for land application. The results show that the ISU-ELG model over-predicts performance of traditional containment systems in comparison to the SPAW model at all five Iowa locations investigated. For wetter areas in Iowa, the number of drying days has a large affect on basin performance, whereas for the drier northwest region of Iowa this affect is limited. Possible methods of improving the ISU-ELG model predictions include adding a soil moisture accounting function to model moisture levels in the land application area or calibrating the number of drying days required before land application can commence.

In addition to modeling traditional containment systems, this thesis also examines possible methods of modeling VTA's, as previous research has shown that the ISU-VTA model greatly over-predicts VTA performance. In this study, two different approaches, both using the SPAW model, were investigated to determine their ability to predict hydraulic performance of the vegetative treatment areas (VTA's). Three of the four locations used in this study had a high water table; this water table elevation limited the space available in the soil profile to infiltrate and store water. For these locations, the performance of the VTA was limited by the storage available in the soil profile and SPAW simulations provided a realistic prediction of the monitored results. Modeling results verified that for these locations VTA performance was limited by the space available in the soil profile. Modeling statistics were calculated to determine the model's ability to predict VTA performance. For the four locations investigated, Nash-Sutcliffe efficiencies ranged from 0.45 to 0.99 while the percent bias of the model ranged from -3% to 100%. The results show that the SPAW pond module can be used to determine if VTA performance will be limited by presence of a high water table. Additionally, these methods provided insight into possible modifications to improve the performance of the ISU-VTA model.