Geochemical anomaly separation by multifractal modelling

Abstract

Geochemical anomaly separation using the concentration–area method as applied to two different geological settings in Portugal is investigated. Soil geochemistry data sets were used in both settings. Geochemical surveys were conducted for the exploration of gold mineralisations in quartz veins associated with major shear zones (Arouca, NW Portugal). Similar surveys were also undertaken for V and Ti exploration related to the accumulation of oxide minerals in gabbroic rocks belonging to an exotic oceanic terrane (Mombeja) and to a late calc-alkaline igneous complex (Odivelas), both in SE Portugal. Results have been interpreted using the extensive information available for each area, which includes detailed geological mapping, structural information, whole-rock chemical analysis, and mineralogical data. Anomalous thresholds for each area were computed and compared mainly with whole-rock chemical analyses and mineral chemistry data. In the Arouca area, anomalies at the local scale were identified for Au (630 ppb), As (912 ppm), and Sb (630 ppm), and at a regional scale for As (96 ppm). In the Mombeja area, anomalies at a regional scale were identified for V (88 ppm), Cr (232 ppm), and Ni (76 ppm), and the available results for Odivelas area suggest the existence of local scale V anomalies, whose magnitude is generally above 400 ppm. In the Arouca area, the correlation between these threshold values and grades in rocks is clearly shown, especially by discriminating two different mineralising events. The second, and most important mineralising event is responsible for grades above 1 ppm for Au in the mineralised structures, contrasting with grades below 500 ppb from the first event. The anomalies identified in Mombeja reflect mainly a regional distribution of the outcropping rock types, mainly metagabbros and serpentinised peridotites, with special emphasis on the correlation between V and the gabbroic rocks. The thresholds obtained for each element are always lower than their mean content in the rocks. The Odivelas area shows threshold values for V clearly above the mean rock content, being a consequence of the occurrence of anomalous accumulations of oxide minerals within the gabbroic rocks. In order to overcome data scarcity, a scheme to super-sampling the population by using estimated values with ordinary kriging in the computation of the anomalous thresholds was essayed. These results, albeit needing further validation, gave some indicative references to the economic potential of this area for V mineralisations.