Quantifying the spatial characteristics of geochemical patterns via GIS-based geographically weighted statistics

• Geographically weighted statistics are applied to explore geochemical patterns.
• The enrichment and depletion of elemental concentrations are revealed.
• Geochemical patterns exhibit spatially non-random and scale independent.

The spatial characteristics of geochemical patterns are significant for mineral exploration and environmental studies. In this contribution, geographically weighted means, coefficients of variation, correlation coefficients, geographically weighted principal components analysis (GWPCA) and local autocorrelation were applied to investigate the spatial characteristics of geochemical patterns in regional stream sediment data in southwestern Fujian Province, China. The results demonstrate that (1) the spatial variations of geographically weighted means and coefficients of variation, based on various window sizes, are scale-dependent, which reveal enrichment/depletion of elemental concentrations, and quantify the heterogeneity of geochemical patterns, respectively; (2) the resulting geographically weighted correlation coefficients suggest heterogeneous and geochemical behavior relationships between Fe2O3 and Cu, Mn, Pb, Zn; (3) the local Moran's I indicates that the observed geochemical patterns are spatially non-random and scale-dependent; and (4) the GWPCA is a powerful tool to investigate the spatial relationships between the known Fe deposits and mineralized elements. These investigations are useful for the next round of mineral exploration for Fe polymetallic deposits in the study area.