Multivariate Statistical Analysis of Soil Geochemical Data from Bogosu Area, Southwestern Ghana.

ABSTRACT

Geochemical mapping and multivariate factor analysis were performed on multi-element soil geochemical data from Bogosu area in southwestern part of Ghana. Multi-element soil geochemical data have been used in soil geochemistry for the characterization of underlying lithology as their surface expressions are often based on the elements’ spatial distribution (Grunsky and Smee, 2003). The objectives of the study were to determine the spatial distribution of the elements, determine elemental associations which could help in the identification of pathfinder element(s) and to identify possible geochemical factors that could explain the spatial distribution patterns associated with the elements in the study area. The study initially focused on eight elements which included Fe, Mn, Zn, As, Pb and Au. These elements had been previously analyzed in one thousand and twenty-one (1021) samples which were taken from the study area. The samples were analysed using Atomic Absorption Spectroscopy. Contour maps were used to show the spatial distribution of all eight elements. The contour maps were plotted using threshold values determined as (Median+2MAD) of datasets of each of the elements to delineate anomalous and background zones.

Histogram and Box plots were initially used to verify the suitability of the distribution of the datasets of all eight elements for factor analysis. Factor analysis was however performed on seven of the elements excluding Pb since an initial test of factor analysis showed a very low communality for Pb. The threshold values obtained were 5.0 ppb, 31.0 ppm, 64.0 ppm, 23.0 ppm, 7.9 % wt., 500.0 ppm, 17.0 ppm and 29.0 ppm for Au, As, Ag, Cu, Fe, Mn, Pb and Zn respectively. Overlying anomalous regions on contour maps of the elements confirmed the results of factor analysis to show element associations. Factor analysis explained 68.42 % of the total variance of the data through three factors. The gold factor was factor 2 and Au was strongly associated with Ag accounting for 15.42% of the total variance. It was also noticed that gold had some associations with as in factor 3, and Fe, Mn, Cu and Zn in iv factor 1. Therefore, Ag alongside as, Cu, Zn, Mn and Fe could be used as pathfinders for gold in the study area. It can also be inferred from the results of the analysis that the occurrence of gold and its associated elements can be linked to primary dispersion from underlying rocks and secondary processes like alteration and leaching. There is however the need for further investigation to compare the chemical composition of the soil and the rock outcrops found in the study area to establish the significance of anomalies.