Interpretation of surface geochemical data and integration with geological maps and landsat-tm images for mineral exploration from a portion of the Precambrian of uruguay

Abstract

Geological map information, geochemical data and digital Landsat-TM imagery were processed and integrated for a portion of the Precambriam basement of Uruguay based on statistical and image processing techniques and geographic information systems. Stream sediments and soil samples were collected at 2,172 sites and analysed for 22 elements by DCP spectrometry. The elements Fe, Mn, P, Ba, Co, Cr, Cu, Ni, Pb, V, Y and Zn showed adequate variance and analytical errors lower than 10%. A number of samples had the elements Ag, As, B, Be, Cd, Mo, Nb, Sb and W with contents lower than the analytical detection limits and analytical errors greater than 15%. For these reasons, stastistical processing was carried out for the samples from the first group of elements, whereas, for the second group, only samples with contents near or above the detection limits were used. Some problems were found in relation to the sampling methodology and statistical outliers. These problems were corrected: (i) through the use of median values for groups of elements; (ii) eliminating the outliers. The final data set showed a more homogeneous signature. Several regional anomalies and nine anomalous sectors were defined and related to lithological factors or to known mineralization. Classification techniques using geochemical data and features extracted from Landsat-TM imagery discriminated regional-, lithological- and mineralization- related anomalies and new areas with potential for containing mineral deposits were detected. The use of CIS techniques allowed to reduce the size of the potential areas and to establish hierarchical levels related to the mineral potential.