Methodology of modelling: Simulation of urban scenarios

Abstract

Urbanization has led to a series of problems in medium and large cities, resulting from a lack of planning and technologies to support decision making. This work aims to contribute to the methodology of modeling by cellular automata, aiming the study of the processes that condition the expansion of the urban area. Having as a study case of the municipality of Novo Hamburgo, RS, Brazil. For this purpose, the urban areas of 1967, 1977, 1987, 1997, 2009 and 2015 were used, totalling a historical analysis of 48 years, besides the thematic maps: Declivities, Hypsometric, Land Use, Water Resources and Road System. The compiled data was inserted in the environment of the platform Dinamica EGO, which is a free software that uses the system of Cellular Automata. The software employs a set of input maps composed of an initial landscape, a final landscape and a compilation of thematic variables that are combined from the definition of weights of evidence to generate the transition probability maps. The results showed similarity indexes of 0.9 for the first three models, 1977, 1987 and 1997, as well as 0.83 for the simulation of 2015. It was also possible to analyze the predominant factors for each expansion process, besides the simulation of areas of prognosis for short and medium term, that is, 5, 10 and 15 years.