Spatial optimization of watershed best management practice scenarios based on boundary-adaptive configuration units

Abstract

Spatial optimization of watershed best management practice (BMP) scenarios based on watershed modeling is an effective decision support tool for watershed management. During such optimization, existing types of BMP configuration units for configuring BMPs (or BMP configuration units, e.g. subbasins, hydrologic response units, farms) remain fixed boundaries once they have been created through spatial discretization prior to BMP scenario optimization. This sort of “boundary-fixed” method does not allow for adjustments to the spatial characteristics of BMP configuration units. Hence, it runs the risk of missing superior BMP scenarios that could have been obtained by adjusting unit boundaries and may produce less effective spatial optimization. In this article, we propose a new approach to the spatial optimization of BMP scenarios based on boundary-adaptive configuration units. The proposed optimization approach adopts slope positions (basic landform units along hillslopes inherently related to physical hillslope processes) as BMP configuration units and dynamically adjusts their boundaries by using quantitative information about their spatial gradation (i.e. fuzzy slope positions) during the optimization. A case study conducted in the Youwuzhen watershed in Fujian, China, showed that the proposed optimization approach can significantly enlarge the search space and obtain optimal BMP scenarios with better cost-effectiveness and higher optimization efficiency than with boundary-fixed units. The proposed optimization approach provides a new alternative framework for spatial optimization of BMP scenarios, in which other watershed models, intelligent optimization algorithms, and BMP configuration units available for boundary adjustment can be applied to BMP scenario optimization in a boundary-adaptive manner. This study also exemplifies the potential for transforming qualitative, vague, and empirical geographical knowledge about slope position units related to physical hillslope processes and BMPs into quantitative, explicit, and automated geospatial algorithms for effectively resolving environmental management problems in a more geographically meaningful way.