Application of Tabu search algorithm with a coupled AnnAGNPS-CCHE1D model to optimize agricultural land use

A principal contributor to soil erosion and nonpoint source pollution, agricultural activities have a major influence on the environmental quality of a watershed. Impact of agricultural activities on the quality of water resources can be minimized by implementing suitable agriculture land-use types. Currently, land uses are designed (location, type, and operational schedule) based on field study results, and do not involve a science-based approach to ensure their efficiency under particular regional, climatic, geological, and economical conditions. At present, there is a real need for new methodologies that can optimize the selection, design, and operation of agricultural land uses at the watershed scale by taking into account environmental, technical, and economical considerations, based on realistic simulations of watershed response. In this respect, the present study proposes a new approach, which integrates computational modeling of watershed processes, fluvial processes in the drainage network, and modern heuristic optimization techniques to design cost effective land-use plans. The watershed model AnnAGNPS and the channel network model CCHE1D are linked together to simulate the sediment and pollutant transport processes. Based on the computational results, a multi-objective function is set up to minimize soil losses, nutrient yields, and total associated costs, while the production profits from agriculture are maximized. The selected iterative optimization algorithm uses adaptive Tabu Search heuristic to flip (switching from one alternative to another) land-change variables. USDA's Goodwin Creek experimental watershed, located in Northern Mississippi, is used to demonstrate the capabilities of the proposed approach. The results show that the optimized land-use design with BMPs using an integrated approach at the watershed level can provide efficient and cost-effective conservation of the environmental quality by taking into account both productivity and profitability.