Local multivariate outliers as geochemical anomaly halos indicators, a case study: Hamich area, Southern Khorasan, Iran

Abstract

Anomaly recognition has always been a prominent subject in preliminary geochemical explorations. Among the regional geochemical data processing, there are a range of statistical and data mining techniques as well as different mapping methods, which serve as presentations of the outputs. The outlier's values are of interest in the investigations where data are gathered under controlled conditions. These values in exploration geochemistry indicate the mineralization occurrences, and therefore, their identification is vital. Both the robust parametric (based on Mahalanobis distance) and non-parametric (based on depth functions) techniques have been developed for a multivariate outlier identification in geochemistry data. In this research work, we applied the local multivariate outlier identification approach to delineate the geochemical anomaly halos in the Hamich region, which is located in the SE of Birjand, South Khorasn province, East of Iran. For this purpose, 396 litho-geochemical samples that had been analyzed for 44 elements were used. The obtained results show a good agreement with the geological and mineral indices of Pb, Zn, and Cu in the southern part of the area. Such studies can be used by a project director to optimize the core drilling places in detailed exploration steps.