Statistical analysis of airborne gamma‐ray data for geologic mapping purposes: Crixas‐Itapaci area, Goias, Brazil

Abstract

Q‐mode factor analysis, K‐means clustering, and G‐mode clustering were used on digitized gamma‐ray spectrometer data from an aerial survey of the Crixas‐Itapaci area, Goias, Brazil. The data points including seven variables—eU, eTh, K, total count, U/Th, U/K, and Th/K—were digitized for a 2 km square grid. For the northwest corner of the area the data were gridded at 1 km. The Q‐mode classification method supplied results that do not show a good correspondence with the known geology. The K‐means clustering procedure barely identified the main lithologic features of the area. The G‐mode technique produced results that correlate well with the known geology and identified the greenstone belts present in the area by discriminating their ultramafic and mafic components from adjacent felsic rocks. Statistical analysis of aerial gamma‐ray spectrometer data can be very helpful in mapping geologic units in poorly known areas. It can also be used for mineral exploration purposes if mineralization is known to be associated with lithologies that can be identified by the techniques used in this study.