Reconhecimento e mapeamento das feições fluviais da “Ilha Grande do Tapará” (Santarém-PA) a partir do processamento de imagens RapidEye e SAR-SIPAM

The objective of this paper was to recognize and map the main fluvial environments along the Amazon River from remote sensing data. Optical images of the REIS sensor were used aboard the RapidEye satellite, with the goal of mapping the region in detail, and SAR-SIPAM synthetic aperture radar images, which allow us to visualize the region of study with reduced atmospheric interference. In order to complement these data, the integration of the radar and optical data was made providing the SPC-SAR, product of this synergism. The mapping of the environments was based on the non-supervised classification of the REIS, SAR and SPC-SAR images, which allowed the identification of lake classes with swamp, canal, fluvial delta, river terrace, river beach and water. The generated classes were and their features were associated to the environments of the study area. The general precisions of the classifications were 93.33% and 0.90 for the REIS sensor; 93.33% and 0.89 for SAR; and 90.66% and 0.86 for the SPC-SAR product. These results indicate that both data provide us with important and accurate information about the study surface and that the integrated product can complement these data allowing a more accurate mapping of the fluvial environments in the Amazon region.