Watershed models for nonpoint source pollution management decision support

Nonpoint source pollution and stormwater are generated from diffuse processes within a watershed. The management of nonpoint source pollution is a growing area of interest, as past success with managing point source pollution has provided approaches that may be adapted to the challenging problems of nonpoint source pollution, and has shown that point source pollution control alone may not be sufficient to meet water quality goals in many locations.;Watershed models are likely to play an important role in the future of nonpoint source pollution management because of the assistance they provide in analyzing complex processes in the environment and in potential management systems. There is increasing need for the formulation of models to allow for their application to the optimization of watershed management strategies. Application of watershed models in an optimization context to nonpoint source pollution problems is a developing field, and the approaches described here are intended to contribute to the growing knowledge in that area.;The approaches described here include the development of a lumped-parameter watershed model that is coupled with best management practice models to explore management practice allocation for nutrient reduction at the watershed scale. A distributed-parameter non-linear model is also developed to allow for the spatial representation of management practice implementation for sediment load reduction. Additionally, alternate formulation approaches to distributed stormwater and nonpoint source pollution reduction are developed to allow for the use of classic optimization techniques that have not been widely applied in this contemporary context. The results of the alternate approaches indicate the solution to stormwater peak flow reduction is influenced primarily by distributed flow arrival time, and a linear program analog to a non-linear optimization model efficiently reproduces much of the same solution structure. Linear and dynamic program solutions to the storm sediment management problem indicate natural sediment trapping is an important consideration, and a solution to the sediment management optimization problem can be efficiently found using a dynamic program.