A linear spectral mixture model to estimate forest and savanna biomass at transition areas in Amazonia

The objective of this paper is to analyze the potential application of fraction images from a spectral mixture model to characterize the different vegetation types at transition zones between forest and savanna formations. This study presents also the summarized methodology of biomass data collection during a field survey, the spatial distribution of “vegetation”, “soil” and “shade” proportions for each class identified at the TM/Landsat image, as well as the relation among biomass and the variables obtained from satellite data, derived from this spectral mixture model. This approach has been very useful to monitor the changes due to land occupation that have occurred in these transition zones of forest/savanna in Brazilian Amazonia