Recognition of geochemical anomalies based on geographically weighted regression: A case study across the boundary areas of China and Mongolia

The identification of stream sediment geochemical anomalies related to mineralization from background is critically needed for mineral exploration using data processing method in the diverse lithological background and complex regolith terrains. In this research, a method based on geographically weighted regression (GWR) was presented for the identification of stream sediment geochemical anomalies. The concentrations of rock-forming oxides, lithophile elements, organic carbon and total carbon were taken as proxies for parent lithology and regolith type to adjust for variations in background of trace element geochemical patterns. Robust principal component analysis (RPCA) was conducted, and then the principal components were taken as spatially independent variables. The metallogenic elements were taken as dependent variables in GWR model, and the geochemical residuals were used to indicate local anomalies. The 1:1,000,000 stream sediment geochemical data across the boundary areas of China and Mongolia were analyzed, and the result of GWR was compared with that of a traditional method. It is found that the efficiency of GWR was highly improved compared with that of the traditional method, indicating that the proposed method can model and eliminate the background differences of elements due to lithological settings and landscapes. Anomalies identified by GWR had stronger spatial association with the known deposits, and thus can be used as guides to new exploration targets.