A knowledge-based model of watershed assessment for sediment

A watershed is a complex ecosystem. Assessment of watershed condition entails consideration of numerous issues and factors. The problem is complex, the issues are not well defined, and data are often lacking. These characteristics suggest that a knowledgebased approximate reasoning approach is especially useful for watershed assessment. This paper describes a knowledge base for watershed assessment for sediment (WAS). The knowledge base is designed for protection of fish habitat and control of excessive sediment, and is evaluated in the Ecosystem Management Decision Support (EMDS) system. The WAS model allows experts from diverse fields to contribute to an integrated assessment of watershed condition. As a decision support tool, the model provides a means to assemble key pieces of information and reasoning that support land use or regulatory decisions, and to communicate among diverse audiences the basis for those decisions. The paper also presents an application of the model to assess the condition of a coastal watershed in northern California.