Modelagem espacial dinâmica para a concepção de cenários prospectivos de uso e cobertura da terra no município de Araraquara – SP

Abstract

The present dissertation sought to analyze in which ways a dynamic spatial modeling can contribute to issues related to territorial planning, discussing the possibilities arising from the prospection of future land use and cover scenarios provided to those responsible for territorial planning. The models are a simplified structuring of reality, presenting in a relatively superficial way some of its important characteristics or relationships, and can be divided into 4 groups, mathematical models, systems models, graphic models and predictive models, where in this work the last mentioned model, through dynamic spatial modeling. The model was created using the MOLUSCE plugin, present in the Qgis software, using the mapping of land use and land cover from the MapBiomas project for the years 2000, 2010 and 2020, adopting the method of Artificial Neural Networks to generate the transition potential between land uses and land cover and an automaton-cellular model in the Monte Carlo approach, prospecting a land use and land cover scenario for the year 2030. The study area was the municipality of Araraquara–SP, chosen due to to the strong pressures imposed by agricultural expansion in its landscape, specifically sugarcane, which jumped from a total representation of the municipality's area from 37.12% in 2000 to 48.58% in 2010. The results of the transition models were satisfactory, where the Kappa value obtained was 0.70185, while the kappa index obtained for the simulation was 0.70034, providing a correspondence rate between the mapping used as reference and the simulated one of 80.32820%. It was possible to observe that the dynamics of land use and land cover in the municipality was harmful to the prospect of future scenarios, and it is important to analyze the trends of change in the study area when using the MOLUSCE. The plugin proved to be capable of analyzing and capturing trends in changes in land use and land cover classes, satisfactorily and coherently projecting these trends for future years, recommending its use in other areas of study.