Processamento e integração de imagens multiespectrais, de radar e aerogeofísicas do bloco Serra Leste (Carajás, PA) e discussão do potencial para mineralizações de ferro

Abstract

This paper applied multispectral (ETM+/Landsat 7), SAR and airborne geophysical data both to aid the geologic characterization and to highlight tracts with potential for iron deposits in the Serra Leste Block, located in the Serra dos Carajás (Pará State, Brazil). The Minimum Noise Fraction (MNF) technique was applied to six bands of the ETM+/Landsat sensor/satellite, producing the enhancement of iron-bearing surfaces, SAR data allowed the observation of the macro-topography and lineaments that were also interpreted on the basis of airborne geophysical and multispectral data. Geophysical and SAR (RADARSAT-1) data were combined by means of the IHS (intensity-hue-saturation) transform and the products yielded were used to improve the limits of the geologic units and to define targets for iron exploration. The Spectral Angle Mapper (SAM) supervised classification method was applied to a set of selected airborne geophysical images based on the rationale of the iron mineralization descriptive model and geophysical signatures observed at Serra Leste, Using documented iron occurrence as training point for the SAM classification, iron favorability maps were produced and compared against several early maps, proving to be equivalent or superior in accuracy. © 2008 Sociedade Brasileira de Geofísica.