Designing BMPs at a Watershed-Scale Using SWAT and a Genetic Algorithm

An optimal control model (OCM) is developed by coupling a semi-distributed hydrological model, Soil and Water Assessment Tool (SWAT), with a genetic algorithm (GA) to identify the least cost design (sizes, types, and locations) of structural best management practices (BMPs) while meeting treatment goals at a watershed-scale. Treatment goals considered are constraints on daily peak flow values and annual sediment load. SWAT performs the hydrological simulation incorporating combinations of BMPs while the GA searches for the least cost combination. Structural BMPs included in the OCM include detention ponds, infiltration ponds, parallel terraces, filter strips, grade stabilization structures, and grassed waterways. The model is demonstrated on Silver Creek watershed, a sub-watershed of the Lower Kaskaskia basin in Illinois. The OCM is able to find optimal or near optimal solutions for different scenarios of treatment goals.