Modelagem espacial da dinâmica da mancha urbana de Viçosa-MG

• Autor(a) principal: Caixeta, Paulo Ricardo Corrêa
• Data de Publicação: 2014
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/3837
• topic: Planejamento urbano - Viçosa (MG); Solo - Uso; Planejamento; Geoprocessamento; Modelagem espacial; Urban planning - Viçosa (MG); Land - Use; Planning; Geoprocessing; Spatial modeling; CNPQ::CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS

Urban Planning, as the key to solve the negative impacts of population growth, obtained a considerable gain in guidelines and legal tools for the regulation of urban policies on the approval of the Cities' Statute (Law No. 10.257/2001). However, the implementation of urban planning requires geoinformatic infrastructure that assists the planning and management of cities, providing analysis and diagnostic mechanisms for decision making. Thus, this study evaluated the ability of dynamic models of land use based on the structure of cellular automata for simulating the landscape in current and future events. The model is implemented in the DINAMICA EGO free software. In this model, the input data constitute cartographic variables that influence the phenomenon of urban expansion. Besides these variables, "landscape maps" were also used, which refer to land use in the beginning and in the end of the interval studied. In order to obtain historical information regarding land use, it was used images taken from the ThematicMapper (TM) sensor, on board Landsat 5 satellite, which were classified using techniques of spectral mixture and vegetation index for discriminating the dominant classes. By combining these variables through the sum of the weights of evidence, it was obtained the areas which are most likely to transitions. The models were set at intervals of 13, 9 and 4 years, and their assessment was made by comparing reference data using neighborhood analysis, as known as similarity indices. In the projections for the years of 2015 and 2019, it was observed a trend toward the consolidation of semi-urbanized areas, and the expansion in radial direction from the town center, bounded by regions of high declivity, or with fragments of Mata Atlantica, which hinders urbanization, following the trend shown in the previous periods. Through quantitative assessment of the similarity of the results of simulation models with reference data, values were above 0.45 for analysis of exponential decay, and 0.84 for analysis of steady decay at a 11x11 window, which were considered satisfactory. Therefore, the modeling based on cellular automata (CA) spatialized the changes which occurred over the intervals analyzed, similarly to the ixstandards that actually occurred in the same interval, and indicated the ability to extrapolate the model's period in order to find out future projections of possible transitions.