Application of classical statistics and multifractals to delineate Au mineralization-related geochemical anomalies from stream sediment data: a case study in Xinghai-Zeku, Qinghai, China

Abstract

The Xinghai-Zeku area of western China is one of the most important Au polymetallic metallogenic belts in Qinghai Province, China. To guide the mineral exploration in this area, log and isometric log-ratio (ilr) transformations of stream sediment data are evaluated using principal component analysis (PCA) combined with geological data to study the relationship among different elements. In addition, Mean + 2 standard deviations (Mean + 2STD), Median + 2 median absolute deviations (Median + 2MAD), and concentration-area (C-A) models are applied to identify pathfinder threshold values and geochemical patterns are decomposed using a spectrum-area (S-A) model. The results show that: (1) PCA for the ilr-transformed data can accurately describe three different geochemical assemblages, Au-As-Sb, that represent Au-As-Sb mineralization in fracture zones; (2) anomalies of Au + As + Sb are more suitable for targeting Au-As deposits than those of the single element Au; (3) the C-A model is useful for indentifying geochemical anomalies associated with mineralization because results obtained by the C–A fractal model and the geological characteristics are well correlated; (4) background and anomaly maps for Au + As + Sb from the S-A model in conjunction with the Mean + 2STD successfully identify weaker anomalies by reducing anomalous areas; and (5) using Au + As + Sb anomalies identified by C-A in conjunction with the S-A model and Mean + 2STD method are effective in exploration for Au deposits in the area.