Avaliação da aplicabilidade de diferentes geotecnologias e ferramentas associadas em análises urbanísticas de assentamentos informais e adjacências na cidade de Belo Horizonte e no Distrito Federal

Abstract

Cities have increasingly been occupied by informal settlements, whose appearance and growth cause urban and social problems that are difficult to solve. In certain cases, such occupations occur due to the inaction of the public power regarding to their requalification or removal. In an attempt to solve the issue, geoprocessing resources, some of which are free of charge, have been used in urban planning. Thus, this research had as main objective to evaluate how feasible is the use of such resources regarding landscape metrics, thematic quality, and generation of urbanistic information. The evaluation used data from the city of Belo Horizonte in Minas Gerais and the Distrito Federal. Two types of areas were analyzed: one larger, covering the totality of both territories, extracting their metrics; and the other analysis considered the landscape metrics and thematic accuracy of maps of the Aglomerado da Serra, in Belo Horizonte and the Habitational Sector Sol Nascente (SHSN), in the Distrito Federal, both originally informal. We used GIS (Geographic Information Systems) programs, data made available by the MapBiomas Project, by the Belo Horizonte City Government and by the Government of Distrito Federal, the latter two in the form of LIDAR (Light Detection and Ranging-Laser Aerial Scanner) data and orthophotos. Maps were also produced for urbanistic analysis. It was found that the landscape metrics illustrated the major changes in land use and occupation in both Belo Horizonte and the Federal District between 1985 and 2020. The results of the landscape metrics of the Aglomerado da Serra suffered relevant influence from the inaccurate classification provided by LIDAR, in which asphalt was mistakenly represented as vegetation. This inaccuracy also influenced the low global thematic accuracy of its map, which was 39%. For the SHSN landscape metrics, it was verified the absence of the class "Asphalt" in the information extracted from the LIDAR data, which was represented as "Soil", so that it was obtained the global thematic accuracy of 79% for its map. Even with limited accuracy, it was observed that the maps generated have potential for use in preliminary urbanistic analysis.