Evolutionary algorithms for multiobjective evaluation of watershed management decisions

Misgana K. Muleta, Southern Illinois University Carbondale
John W. Nicklow, Southern Illinois University Carbondale

Article comments

©2002 IWA Publishing. The definitive peer-reviewed and edited version of this article is published in Journal of Hydroinformatics, 4(2): 83-97. The original is available at http://www.iwapublishing.com.

NOTE: At the time of publication, the author Misgana K. Muleta was not yet affiliated with Cal Poly.

Abstract

The comprehensive and systematic management of watersheds is essential for reducing the adverse environmental impacts arising from anthropogenically caused erosion and subsequent sedimentation. This paper describes a computational methodology that is designed to serve as a watershed decision support system and is capable of controlling environmental impacts of non-point source pollution resulting from erosion. In the decision process, the methodology also accounts for other inseparable objectives such as economics and social dynamics of the watershed. This decision support tool was developed by integrating a comprehensive hydrologic model known as SWAT and state-of-the-art multiobjective optimization technique within the framework of a discrete-time optimal-control model. Strength Pareto Evolutionary Algorithm (SPEA), a multiobjective optimizer based on evolutionary algorithms, has been used to generate Pareto optimal sets. For demonstration purposes, the tool was applied to the Big Creek watershed located in Southern Illinois. Results indicate that the methodology is highly effective and has the potential to improve comprehensive watershed management.