Spectral linear mixture modelling approaches for land cover mapping of tropical savanna areas in Brazil

摘要: It is estimated that approximately 60% of the natural vegetative cover Brazilian savanna, locally known as Cerrado and second largest biome in South America, have already been converted. Despite this rapid conversion pace, there only limited attempts to operationally monitor major farming frontier with remote sensing data. In study, we evaluated performance spectral linear mixture models (SLMM) for mapping physiognomies. Two SLMMs were considered: a general model, comprising vegetation, soil shade components, specific restricted 'true' We also considered potential effects atmospheric contamination, influence endmember sources on fraction images derived from models, respectively. The apparently resistant atmosphere respect land discrimination, primarily enhanced forested domains non-vegetated targets (water bodies bare soils). By contrast, regardless source, significantly distinguished Such contrasting complementary behavior suggests synergism between monitoring complex environment such biome.