A Generic Framework for Multi - Disciplinary Environmental Modelling ∗

We present a generic framework for computer-based environmental modelling which supports the coupling of simulation models from various sciences to perform integrative simulations. The framework is, in principle, applicable to any kind of model which simulates spatially distributed environmental processes with an arbitrary, but discrete, time scale. During an integrative simulation the framework coordinates the coupled models which run in parallel exchanging iteratively data via their interfaces. For proving the correctness of the coordination, formal methods of software development have been applied. The framework provides a developer interface for the implementation of natural science and socio-economic simulation models. For the latter, a framework specialisation has been developed which supports the modelling of agent-based social simulation models. The framework design was driven by the idea of enforcing common rules for integrative simulations, which must be respected by all simulation models, while leaving as much freedom as possible for discipline-specific implementations. Within the GLOWA-Danube project, the framework has been successfully applied to construct the distributed simulation system DANUBIA which integrates up to 15 simulation models from various disciplines, like meteorology, hydrology, plant physiology, glaciology, economy, agriculture, tourism, and environmental psychology. Actually, DANUBIA is already in use as a tool for decision makers to support the sustainable planning of the future of water resources in the Upper Danube basin.