Application of geographically weighted regression to identify spatially non-stationary relationships between Fe mineralization and its controlling factors in eastern Tianshan, China

Marine volcanic-sedimentary Fe mineral deposit is one of the most productive mineralization types in eastern Tianshan mineral district, Uyghur Autonomous Region, China. Previous researches have revealed that Fe deposits in this area are mainly hosted by the Lower Carboniferous Yamansu bimodal volcanic rocks and characterized by extensive skarn alteration. Due to variable effects of controlling factors across the study area, Fe mineralization occurs at limited locations with favorable geological environment. In order to understand ore genesis of different Fe deposits and to assist in Fe mineral exploration in this area, it is essential to investigate the spatial relationships between Fe mineralization and its controlling factors across the study area. In this paper, Fe mineralization and its controlling factors are characterized and delineated on the basis of spatial distributions of geochemical signatures obtained by principal component analysis. The spatially non-stationary relationships between these spatial distributions of geochemical signatures are further analyzed by a geographically weighted regression (GWR) method. The comparison of applications of the ordinary regression and GWR shows that GWR is superior in depicting localized spatial relationships between Fe mineralization and its controlling factors. The GWR results demonstrate that influences of controlling factors on Fe mineralization are spatially varied and imply that geological environments favorable to Fe mineralization are diverse at different locations. Based on currently achieved results, suggestions to support future Fe mineral exploration at different subareas are proposed at the end of this paper.