Savannas are the main vegetation type in Central Brazil, covering approximately 23% of the national territory. Locally known as Cerrado, Brazilian Savannas are formed by amosaic of different physiognomies such as grassland, shrubland and woodland that have atypical phenological cycle. ln this context, the MODIS data provide daily measurements well suited to monitor the seasonal phenology of vegetation. The present work aims to evaluate the advantages of the temporal signatures to detect Brazilian Savanna vegetation types in the Chapada dos Veadeiros National Park, Brazil. The adopted methodology may be subdivided into the following steps: (a) elaboration of the 3D cube of NDVI from temporal MODIS images, where the z profile corresponding to temporal signature, (b) noise elimination by combining Median Filter and Minimum Noise Fraction techniques, (c) endmember detection, and (d) spectral classification using Spectral Correlation Mapper method. The results demonstrate that the savanna physiognomies present typical temporal signatures. The endmembers correspond to the three major physiognomic domains: (a) Cerrado grassland, herbaceous dominated region; (b) Cerrado, mostly amixture of grasses and shrubs; and (c) Cerrado woodland, densely covered by trees. Comparison with Landsat 7/ETM+ image demonstrates the classification efficiency of the temporal series. The study concluded that the NDVI series is useful in differentiating the amount of vegetation types The methodology efficiency has been proved for regional delimitation of savanna physiognomies even considering the low spatial resolution of the 250m MODIS sensor and the high spectral mixture.
Cerrado; changing detection; multitemporal analysis; digital image processing
