A framework for model integration in spatial decision support systems

Predictive simulation systems, embedding models of real-world processes, ar e usually designed as stand-alone computer packages. Users wishing to work on broader problems of extended scape need access to several models. This a ccess can be provided by integrating several packages into a decision suppo rt system offering a coherent map-based problem-oriented interface to the u nderlying models. This paper proposes the FMISDS integration approach that offers a problem-o riented method for designing the integrated system. It distinguishes syntac tic and semantic issues in model integration, enabling a declarative descri ption of the models and their conceptual relationships. The integration eff ort is supported by a reusable toolkit that interprets the declarative desc riptions. FMISDS is demonstrated in a spatial decision support system for w ater catchment management.