Artificial Intelligence Techniques in an environmental Geo DSS

The idea of the integration between GIS and Artificial Intelligence, two complimentary technologies, was proposed some years ago. This integration is promising since GIS is very efficient in heterogeneous data integration and geographical analysis, whereas AI, and expert systems in particular, are suitable for handling imperfect knowledge and non structured problems. Environmental problems are a good test bed. We describe the design an implementation of an Environmental Decision Support System to control the quality of water in the Arno river (Tuscany). The system integrates a knowledge base and an inference engine, a model base, a GIS (geo data and processing facility) and a User Interface. The system provides domain experts with information to evaluate the effectiveness and socio-environmental impact of one or more remedies against an abnormal environmental condition. It may be used in emergency conditions or as part of a long term strategy to enhance the quality of river water