Mapeamento aero-geofísico de pegmatitos mineralizados na Província Pegmatítica de Alto Ligonha, no norte de Moçambique

Abstract

Major parts of Mozambique were recently flown to acquire high resolution airborne magnetic and radiometric data. Since then it has been the intent of the National Geology Directorate of Mozambique (DNG) to interpret these data generating value-added maps that are easier to use by the exploration and mining industries than mere airborne geophysical grid data and maps. DNG and Maputo-based Eduardo-Mondlane University (UEM) entered a joint venture with the South African Council for Geoscience (CGS) to conduct a case study over the meso-Proterozoic Alto Ligonha pegmatite province in northern Mozambique to support the regional mining sector. Integration of the airborne geophysical data was achieved using the K-means algorithm for explorative unsupervised crisp clustering of the airborne geophysical data. Discriminant analysis was then used to obtain the degree of class membership of each sample previously classified by K-means. Prior to clustering, Th- and K-surface concentrations as well as apparent magnetic susceptibility and first vertical derivative data of the total magnetic intensity anomalies were determined to project a 100 m-spaced grid comprising 850,000 multivariate sample vectors on the study area. Compilation of the classification results provided the value-added map illustrating that the probability to meet pegmatite bodies is highest where amphibole gneiss occurs. The automated integration of airborne geophysical data using the K-means algorithm is a fast, objective and efficient method to compile value-added maps. The experience made in the Alto Ligonha pegmatite province encourages the adoption of this new methodology over other parts of the Mozambique Fold Belt, becoming integral part of geological mapping ongoing in the country.