Adaptação da metodologia PRODES para a detecção de áreas desmatadas da Mata Atlântica do Estado do Espírito Santo, utilizando imagens RapidEye

The Brazilian Atlantic Forest is one of the most important biodiversity hotspots in the Earth but is currently under pressure and its deforestation is on the increase. Several programs and initiatives to map and monitor this biome are being developed. However often the resolution and the scale are not compatible with the size of the fragments of this Brazilian biome. In this way, the main objective of this study was to propose a methodology for monitoring the vegetation cover of the Espirito Santo State, appropriate to the size of state forest fragments.. The work was developed in five stages: 1- selection and acquisition of the Rapideye images, 2- creation of the reference mask “vegetation cover”, 3- identification of the deforestation polygons through the process of segmentation and automatic classification of the images, 4- verification and manual edition of the deforestation maps and 5- calculation of the updated vegetation cover. The results showed updated vegetation cover areas for each municipality and the identification and quantification of the loss of vegetation for the following vegetation types, for 2007/2008 to 2014/2015: native forest, mangrove and sandbank. The total area lost was, respectively, 2364.69, 61.68 and 13.95 hectares for the native forest, sandbank and mangrove.This study demonstrated the applicability of RapidEye images to detect deforested areas, as well as the use of high-resolution images provided by Google Earth for results verification.