Mapping vegetation in a late Quaternary landform of the Amazonian wetlands using object-based image analysis and decision tree classification

摘要: Fan-shaped morphologies related to late Quaternary residual megafan depositional systems are common features over wide areas in northern Amazonia. These were formed by ancient distributary drainage that great contrast tributary networks typify the modern Amazon basin. The surfaces of Amazonian megafans constitute vegetacional mosaic wetlands with different campinarana types. A fine-scale-resolution investigation is required provide detailed classification maps for various and surrounding forest types associated megafans. This approach remains be presented, despite its relevance analysing relationship between stages plant succession sedimentary dynamics evolution In this work, we develop a methodology classifying vegetation fan-shaped palaeoform from wetland. included object-based image analysis OBIA data-mining DM techniques combining Advanced Spaceborne Thermal Emission Reflection Radiometer ASTER images, land-cover fractions derived linear spectral mixing model, synthetic aperture radar SAR digital elevation model DEM acquired during Shuttle Radar Topography Mission SRTM. DEM, fraction, band 3 most useful parameters defining classes. normalized difference index NDVI, 1, Land Observing Satellite ALOS/Phased Array type L-band Synthetic Aperture PALSAR transmitting receiving horizontal polarization HH vertical HV all effective distinguishing wetland classes Mauritia. Tests statistical significance indicated overall accuracies kappa coefficients κ 88% 0.86 final map, respectively. allocation disagreement coefficient 5% quantity value 7% further attested results. Hence, addition water, exposed soil, deforestation areas, successful differentiating large number open forest, wood, shrub, grass campinaranas, terra firme, varzea, igapo, alluvial, as well Mauritia inner outer studied megafan.