Estimation of basal area from Amazon tropical rain forest using airborne P-band SAR data

Shows the capacity of P-band polarimetric images to define the space of attributes of primary forest and secondary succession, through the analysis of the relation among backscatter and basal area data. To ensure that different landscapes of Amazon upland forest are represented, a test-site located in the Tapajo´s region was selected. Statistical regression model was used to verify the relation between basal area values (37 plots) and backscatter data derived from P_HH, P_HV and P_VV images. An analysis of the spatial configuration of backscatter values for primary forest and secondary succession is made, considering the intrinsic variations of these formations in terms of biometric parameters. For this thematic stratification, a contextual Markovian classification technique (Iterated Conditional Modes CM) was used in polarimetric P-band images, to delineate primary forest and regeneration areas, in association with basal area intervals. The advance that was reached with the treatment of P-band data could improve the regional monitoring process of the land cover change, a process whose speed was accelerated as a result of human action in the Amazon during the last two decades.