An integrated framework for knowledge - based modeling and simulation of natural systems

This paper proposes a new approach to simulation modeling of natural systems in the context of water quality modeling in streams affected by point source pollution. The approach has a potential for application to other domains of natural resource modeling. Its conceptual basis is knowledge-based simulation and systems analysis. In the approach presented in this paper, a stream or its section is viewed as a collection of components, i.e. stream segments. The structure of a stream is its segments and their couplings. For each stream segment, a single constituent model describing processes affecting the segment's water quality is defined. Models are coupled in a hierarchical manner. The hierarchical , modular model specification results in a stream model comprised of a finite number of sub-segment models. Fundamental theoretical concepts supporting such a specification are described. A prototype simulation modeling environment to support prediction of water quality in streams has been implemented and tested. The proposed approach differs from other natural systems modeling frameworks in that it offers a) modular model specification facilities, b) high degree of model reusability, and c) support for model selection and coupling.