Potencial das imagens Landsat – OLI e RapidEye para identificar áreas de degradação florestal em Querência e Canarana – MT comparadas com imagens LiDAR

Favorite this paper

The degradation process happens when a reduction in forest’s quality occurs. In this context, the state of Mato Grosso - MT, Brazil, is known to have the largest forest degraded areas that increased during the last years. A recent technique to study the forest degradation is the Light Detection and Ranging – LiDAR, which allows the assessment of forests in a 3D form. The area studied in this work is located at northern part of Mato Grosso state, comprising 1006 ha. We used two approaches to identify the degradation areas in the OLI and RapdEye images. These results obtained by these approaches were compared with a LiDAR image result which have a better spatial resolution. The techniques used to estimate degradation areas were: the Linear Spectral Mixture Model (LSMM) and the Maximum Likelihood Classification. The RapidEye image was better identify forest degradation in isolated and small areas; and on the other hand, the OLI image was better to depict the sum of degraded areas. Overall, the LSMM showed a more accurate classification than the Maximum Likelihood Classification. Forest degradation is better identified with LiDAR image, but optical images are a possibility when there isn’t the option to use the cloud points 3D.