Mapping of Regional-scale Multi-element Geochemical Anomalies Using Hierarchical Clustering Algorithms

Mapping of multi-element geochemical anomalies is the basic goal of stream sediment sampling in worldwide, and especially at 1:100,000 scale in Iran. In the central part of the Lut-Block in eastern Iran, 855 stream sediment samples from an area of 3000 km(2) have been collected. The existence of sub-volcanic rock units along with argillic and sericite alterations provides potential for poly-metallic mineralization in the study district. Hierarchical clustering analysis of the stream sediment geochemical data in R-mode shows that it is possible to group the 44 analyzed elements into four clusters. The first cluster, with Ag, Au, Ba, Pb, Sr, Te, Tl and Zn elements, and the third cluster, with Cd, Co, Cr, Cs, Cu, Fe, Ge, Ni, Th, Ti, U and V elements, comprise the strategic metals in the study district. Four hierarchical clustering algorithms-OS-AHC-av, OS-AHC-wa, BIRCH and BHC-have been used to determine multi-element geochemical anomalies in the data. The results show four and three areas with mineralization potential for metals of the first and third clusters, respectively. Because the four algorithms resulted in anomalous areas with almost the same shapes and locations, the results indicate the ability of these clustering algorithms to help in mapping of multi-element geochemical anomalies. However, the comparison of these results with those of principal components analysis indicates the relative superiority of the BIRCH clustering algorithm over the others. Therefore, the areas occupying 186-365 km(2) are the first priority for exploration in the next stage and the areas occupying 872-1189 km(2) are the second priority.