Classificação espectral de fitofisionomias em área de floresta tropical utilizando dados do sensor Aster

Abstract

This article aims to evaluate the data capacity created by a sensor named Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER)/Terra, in the phytophysiognomies description of Amanã Sustainable Development Reserve (RDSA). The ASTER data analyzed include the spectral intervals of visible (0.52-0.69 μm), near-infrared (0.78-0.86 μm) and shortwave infrared (1.60 to 2:43 μm), wherein these intervals bands were applied the spectral classification techniques adapted to the data from this sensor as Spectral Angle Mapper (SAM), Mixture Tuned Matched Filtering (MTMF) plus NDVI. By SAM technique was possible to distinguish six predominant phytophysiognomies in the RDSA. By MTMF technique that involves a more robust classification algorithm, equivalent information was obtained. It was also possible to associate and detect spectral patterns of vegetation, showing the close relationship with the NDVI index.