Démarche géo-prospective et modélisation causale probabiliste

This paper proposes the development of causal probabilistic models for spatial strategic foresight, using decision graphs (decisional Bayesian Networks). The proposed models integrate uncertainty, spatial interaction among a limited number of key sub-spaces and the evolution of the spatial system over time. This modeling approach is applied to scenario building for future development of the French Riviera metropolitan area. The limits of the proposed methodology are finally discussed, as well as its potential of integration in strategic planning and its perspectives of future development.