Effective use and interpretation of lithogeochemical data in regional mineral exploration programs: application of Geographic Information Systems (GIS) technology

Abstract

A Geographic Information System (GIS), in concert with statistical and geostatistical software, has been used to compile, manipulate, analyze and visualize a large lithogeochemical dataset collected from the Archean age Swayze greenstone belt (SGB) located in northern Ontario, Canada. Altered samples from this database have been identified using a variety of univariate and multivariate statistical and visualization techniques. Alteration maps are produced for each method, and compared to known gold mineralization using the weights of evidence (WofE) technique. Major oxide data are interpolated using different sample groups ( all, altered, unaltered, normalized samples), and anomalous concentrations are separated from background using breakpoints on probability plots of the data, as well as statistics generated from WofE analysis. These maps are also evaluated, using WofE analysis, with respect to locations of known gold prospects. The atypical normative mineral method is more sensitive for identifying alteration in felsic rocks (felsic volcanic rocks, granitoids), whereas the volatile method is more sensitive for detecting alteration in mafic volcanic rocks and sedimentary rocks. Altered samples are characterized by higher SiO 2, K 2O and MnO concentrations, and lower MgO and TiO 2 concentrations. Areas within the greenstone belt characterized by high concentrations of CaO, Fe, K 2O and Al 2O 3 were found to the best predictors of the known gold prospects. Anomalously high concentrations of these oxides reflect carbonatization, Fe-carbonate alteration and potassic alteration trends, and indicate that these alteration styles are spatially related to gold mineralization. The WofE method is a useful technique with which to evaluate the geochemical maps produced in this paper. The WofE technique is useful not only for comparing geochemical anomalies to the location of known gold prospects, but also for providing a quantitative measure of association between anomalies and gold prospects. Furthermore, the gold potential maps generated using WofE not only highlight (or predict) known areas of gold mineralization, but also identify high potential areas where no known mineralization occurs. A number of areas within the SGB have been identified as having high gold potential on many of the geochemical maps produced in this paper. Many of these areas coincide with areas of known gold mineralization; however, others are enigmatic and await follow-up exploration.