Potencial dos dados Sentinel-2 e Landsat-8 para a classificação do uso e cobertura da terra de um ambiente costeiro
Considering the provision of timely, and accurate data from remote sensing system, satellite images are important source of creating land cover/use information. This study assessed the performance of the Sentinel-2 and Landsat-8 data for classification of a subtropical coastal zone. Two approaches were compared: maximum likelihood (MAXVER) and random forest (RF). Sentinel-2 data resulted in Kappa index 0.97 and 0.94 with MAXVER and RF classifier, respectively, while Landsat-8 Kappa index were 0.92 and 0.90. All methods differed significantly from one another, indicating that the use of Sentinel-2 satellite images had superior results to Landsat-8. The analysis of the variables relevance with RF classifier showed that the new bands of Sentinel-2, like red-edge and near infrared plateau, were decisive for the successful classification of Sentinel-2 data. Additional research is needed to assess the full potential of Sentinel-2 data and to explore potential applications of this data in other environments or land cover change monitoring.