Modelagem dinâmica do uso e cobertura do terra na área de influência da usina hidrelétrica Dona Francisca - RS

Abstract

The understanding of changes in land cover and use in influence area of Dona Francisca Hydroelectric Power Plant requires methodological applications that allow not just to map, but also to analyze the spatial patterns generated by these changes and simulate reliable future scenarios that can be applied at local level, for planning and decision analysis. From this perspective, this study aimed to make the modeling of the dynamics of land cover and use in influence area of Dona Francisca Hydroelectric Power Plant and, based on that model, make projections for future scenarios. The methodology consists in thematic mapping static variables represented by economic, social and landscape attributes, mapping land cover and use for the years 1992, 2003 and 2013 from Landsat images 5 and 8, TM and OLI sensors. The developed maps were used as input data of a dynamic simulation model based on cellular automata, which was performed in the Dinamica EGO modeling platform. Annual scenarios simulations until 2023 were generated, designed by probabilistic method of Evidence weights. The results demonstrated the predominance of forest and farming classes, showing a gradual increase of these over the considered period. In the transition process, farming influenced both reduction and forestry expansion. The U.H.E Dona Francisca contributed to the maintenance of forest and farming areas, as these had lower reduction rates after installation. Just the same, forestry and farming reductions were lower after the installation of the plant. The construction and implementation of the dynamic model of land cover and use for the period of 2003-2013 allowed to obtain a simulated map for year 2013, with fuzzy similarity between 0,55 to 0,88. The simulated scenarios obeyed the observed tendency during the considered period for modeling. Therefore, if conditions are maintained, forest and farming classes will be continually expanded until the year 2023. The exposed soil and water classes, in contrast, will present gradual reduction, however, in smaller proportions.