A Fuzzy Cellular Automata Modeling Approach – Accessing Urban Growth Dynamics in Linguistic Terms

Abstract. This paper presents a methodological framework for urban modeling which accesses the multi-level urban growth dynamics and expresses them in linguistic terms. In this approach a set of parallel fuzzy systems is used, each one of which focuses on different aspects of the urban growth dynamics, diffe-rent drivers or restriction of development and concludes over the suitability for urbanization for each area. As a result the systems ’ structure and connection merge the input variables into a single variable providing an information flow familiar to the human conceptualization of the phenomenon. At the same time, the structure does not pose severe data requirements while the utilization of pa-rallel connection between fuzzy systems allows the user to add or remove va-riables without altering the ways in which other variables affect the knowledge base. Following, a fuzzy system that incorporates cellular automata techniques simulates the horizontal and vertical urban growth. The proposed model is ap-plied and tested in the Mesogeia area in the Attica basin (Athens) and fits reality in average by 76 % (LeeShalle index) while the average cell error is 19%. Ne-vertheless, the benefits obtained in the herein presented approach lie in the in-formation management and representation