Development And Application Of A Decision Support System Supporting Multidiscipline Model Integration

As watershed science researches involve multiple field knowledge likeecology, economics, and sociology, a sustainable watershed management willrequire much more knowledge than traditional. Obviously, it is hard to make arational decision based on the traditional and simple management strategies andtools. Decision Support System (DSS) is developed as a management tool thathas strong capability in certain aspects of watershed management such asevaluating present watershed situation, designing management plans, andmodeling development and climatic scenarios. However, a traditional DSS isusually aiming at one specific objective by making use of one or two models,so that it usually does not have sufficient capability to handle complexwatershed problems that are multidisciplinary in nature.In this paper, a novel DSS architecture was proposed, which offers theability of supporting multidiscipline models. In this architecture, models, withdifferent spatial and temporal scales, can be coupled with each other, to supportmultidisciplinary watershed management decision making. The design andimplementation was elaborated and followed that, two cases studies in theHeihe River basin were carried out to demonstrate its ability in supportingwatershed management. Especially, this paper focuses on following aspects,1. Developing a new smart DSS architecture that supports multidisciplinemodel coupling. By introducing techniques used in modeling software to DSS,a new watershed management DSS was developed.2. A new graphic-driven decision support flow was proposed, which isable to support scenario evaluation. This flow exactly matched our specificscenario definition, including steps like problem definition, scenario definition,simulation, and evaluation.3. An open interface to incorporate any models to the DSS. Byintroducing indicator computation script to link the model to the decisionindicator, the model management module can hold any models and enable interaction between models as well.4. Introducing a new GUI based model integration scheme on basis ofNode-Link. In this scheme, models or data sources were treated as Node, andall model connections and data exchanges between models were implementedby Link. One of the advantages of this scheme is that it can couple models thatare with different spatial discretization and temporal scales.5. A data management module was developed so as to supply data todrive model runs. For often used models, the module can import/export datafrom/to their customized formats.6. Providing GIS support to bring GIS mapping and spatial analysisfunctions to watershed management.Finally, two cases were carried out in Heihe River basin. In the first case,SWAT model and Minimum data method were used to analysis the influence ofeco-compensation price on grazing control and water conservation. In anothercase, a series of Heihe River basin land use scenarios was built to analysishydrological phenomena of their scenarios by coupling SWAT and CLUE in thesystem. The scientific significance and technical reliability of the system wereverified by these two cases.