Identification of cellular automata: theoretical remarks

Land use evolution during forty years in a large set of European cities is analysed by means of a cellular automaton. In one hand (the operational level), the use of this modelling tool allows: a: to study the transition rules in land use and the proximity effects on these rules; b: to compare the different case -studies, otherwise very difficult to be confronted; c: to define scenarios of evolution, on the bases of the past trends. On the other hand (methodological level), availability of a large data-base (significant time series for a set of comparable cases) allows: a: to manage, in a scientific way, the problem of calibration and validation of a cellular automaton (a crucial problem - we have to blame - usually neglected in territorial applications); b: to verify, empirically, potentialities and limits of cellular automata, compared to other models for the analysis of spatial dynamics.