Toward a Spatially-Centered Approach to Integrate Heterogeneous and Multi-scales Urban Component Models

This article addresses a model coupling based approach (i.e reusing and combining spatial models) for modeling and simulating complex systems. Our research is conducted by a land use program of Métouia city (Tunisia) for which administration would study (by simulations) different planning scenarios to identify strategies of industrial development. These simulations should take into account demographic, socio-economic and environmental factors. Many urban models are available but they do not integrate these three aspects. This limitation could be solved by a model coupling based approach. In this paper, from an analysis of models and approaches presented in the literature, we identify key points, needs and the basis of an approach to couple models. Then, we introduce an original approach, based on agent paradigm, in which space is the coupling factor to interconnect heterogeneous models (mathematical models, stochastic models, individual based models, and so on). The pertinence of this coupling approach is also raised by the correlation to observe the impact of models on each other.