Enhancement and integration of airborne gamma-ray spectrometric and Landsat imagery for regolith mapping ? Cape York Peninsula

Abstract

Airborne gamma-ray spectrometric data (400-m line spacing), gathered for the National Geoscience Mapping Accord North Queensland Project, has proved invaluable for differentiating regolith types based on their potassium, thorium and uranium signatures. The ability of the gamma-ray signal to penetrate the vegetation cover and as much as 40 cm below the surface was a considerable advantage in North Queensland, where traditional mapping techniques are hindered by the thick vegetation cover and where deeply weathered sandy soils frequently mask underlying geology. Two image-processing techniques were used to enhance the gamma-ray data and integrate them with Landsat Thematic Mapper (TM) data. Firstly, a sun-angle illuminated high-pass filtered component (hillshade) of the gamma-ray image was added to a three-band false-colour composite image of the corrected gamma-ray data. The hillshade component highlighted local changes in the gamma-ray signal which related to changes in regolith materials and lithology, and sharpened boundaries associated with geomorphic features. Secondly, the gamma-ray image was integrated (pixel by pixel) with the high-frequency component of the corresponding TM band 5 scene. This allowed the imagery to be interpreted in a geomorphological and structural context. Broad lithological divisions and structural domains were identified in the imagery, but most of the variation within these groups relates to the regolith cover and to geomorphic features in the landscape. Subtle variations in lithology were effectively interpreted only when the responses of the regolith cover are known. The gamma-ray imagery provided an insight into the weathering and geomorphic history of the region. In addition, the imagery can also be interpreted as a 'geomorphic activity map', separating stable landforms with deep weathering from younger landforms with active erosion and stripping of the surficial cover. This allows past and present geomorphic processes to be interpreted from the imagery, which in turn provides more informed extrapolation of field results and prediction of regolith types. The imagery was highly useful for basement and regolith mapping. There is considerable potential for assisting with interpretation of both stream-sediment geochemical data and environmental studies, since the resulting imagery effectively California maps the distribution of sediments in the landscape, and identifies areas of erosion and land degradation.