Fall M., Baratoux David, Ndiaye P. M., Jessell M., Baratoux Lenka. (2018). Multi-scale distribution of potassium : thorium and uranium in paleoproterozoic granites from eastern Senegal. Journal of African Earth Sciences, 148 (No spécial), p. 30-51. ISSN 1464-343X.
Titre du document
Multi-scale distribution of potassium : thorium and uranium in paleoproterozoic granites from eastern Senegal
Fall M., Baratoux David, Ndiaye P. M., Jessell M., Baratoux Lenka
Source
Journal of African Earth Sciences, 2018,
148 (No spécial), p. 30-51 ISSN 1464-343X
Potassium (K), thorium (Th) and uranium (U) are good tracers of magmatic or alteration processes and the surface concentrations of these elements may be mapped at the regional or local scale by airborne radiometric (gamma) surveys. Geological mapping and mineral exploration are therefore often supported by airborne radiometric surveys, especially in areas of scarce outcrops (e.g., tropical regions). However, a quantitative use of these data poses a problem of scale to the petrologist. The comparison between rock sample analyses and airborne values, corresponding to average concentrations over hundreds of square meters, requires knowledge of the distribution of K-Th-U concentrations at different scales. For this purpose, we undertook a multi-scale analysis of K-Th-U concentrations on four Paleoproterozoic granitoid massifs in eastern Senegal (Balangouma, Badon, Moussala and Saraya). Our method combines geochemical bulk rock analyses, four high-resolution grids (150 m*150 m, 5 m spacing) using a Field Portable Gamma Ray Spectrometer (FPGRS), and an airborne regional survey of eastern Senegal (similar to 30 000 km(2), 250 m line spacing). We find a large scatter of K-Th-U concentrations at the small scale, with the existence of right-skewed frequency distributions approaching a lognormal behavior. This implies that airborne concentrations may systematically differ from the most frequent values measured on the ground at the FPGRS or sample scale. The scale-dependence of K-Th-U distributions is considered to be, in part, a consequence of the central limit theorem applied on radiometric data.
