Using multivariate statistical analyses to evaluate groundwater contamination in the northwestern part of Saudi Arabia

Chemical characteristics of groundwater in the Midyan Basin (northwestern Saudi Arabia) were investigated and evaluated. A total of 72 water samples were collected from existing shallow wells and analyzed for different elements. Two multivariate statistical methods, hierarchical cluster analysis (HCA) and principal components analysis (PCA), were applied to a subgroup of the data set in terms of their usefulness for groundwater classification, and to identify the processes controlling groundwater geochemistry. The subgroup consisted of 46 water samples out of 72 samples and 24 variables included major elements (Ca2+, Na+, Mg2+, K+, Cl-, HCO3 (-), NO3 (-), SO4 (2-)), minor and trace element (SiO2, Al, As, B, Ba, Cd, Cr, F, Fe, Mo, P, Pb, Sb, Sn, Ti, and V). For water samples, four geochemically distinct clusters (i.e., C1, C2, C3 and C4) have been observed by hierarchical cluster analysis. Cr, F and Pb are the dominant ions in cluster C2. Al, As, Cd, Mo, Sb and Ti are the dominant ions in cluster C3, while B, Ca, Cl, HCO3, K, Mg, Na, SO4 and V are identified as dominant ions in the cluster C4. In the PCA, a total of five components are extracted form the data set, which explained 73.37 % of the total data variability. Among them the first component reveals strong associations between As, B, Cd, Cr, F, Mo, Pb, Sb and Ti. The second component reveals the associations between Ca, Cl, HCO3, Mg, Na, SO4 and V.