Spatial distributions and the identification of ore-related anomalies of Cu across the boundary area of China and Mongolia

Abstract

The 1:1, 000, 000 geochemical mapping across the boundary area of China and Mongolia has acquired high-quality basic geochemical data, yielding more than 9000 stream sediment samples and over 400 rock samples. Each stream sediment and rock sample site was assigned to the regional tectonic unit, geological background and geomorphic landscape maps, respectively, resulting in spatial data for three categorical variables. The contents and spatial distributions of Cu in stream sediment and rock samples across the boundary area of China and Mongolia were studied. The stream sediment geochemical data set was centered logratio transformed (clr) to avoid the closure effect. Random forest regression (RFR) was applied to predict Cu geochemical background for each sample with tectonic unit, geological background, and geomorphic landscape as characteristic variables, then the residuals were used to map geochemical anomalies. The study shows that Cu contents in stream sediments and rocks are high in the west and low in the east of the study area, and the median Cu varies with different tectonic units, geological backgrounds, and geomorphic landscapes. The background values of Cu are strongly spatial heterogeneous. The distributions of predicted Cu background values by random forest regression algorithm are similar to the distribution characteristics of Cu in stream sediments and rocks. The influence of tectonic units on the spatial variations of Cu geochemical background was greatest, followed by geological background, the influence of geomorphic landscape was smallest. The geochemical anomalies drawn by the residuals produced in random forest regression are in good agreement with the known deposits, indicating that the predicted geochemical background of Cu is reasonable and accurate, and has certain theoretical and practical significance.