Comparative Analysis between Thematic Maps of Land Use and Land Cover on the East Coast of the State of Rio de Janeiro

Abstract

The availability of large volumes of data observed today means that we must redouble our attention to the differences between them, always making a critical analysis of their limits and potential. In this sense, this study aims to carry out a comparison between land use and land cover mappings carried out from different methodologies and orbital images. These thematic mappings were produced by different groups, MapBiomas and DAGEOP, based on pixel-by-pixel classification by machine learning and Geobia classification, respectively. The main results point to a correspondence of 79.70% between the mappings, and a good part of the divergences can be explained by different aspects, such as the characteristics of the images, chosen subtitles and adopted methodologies.