Recognition and mapping of the significant geochemical signatures of mineral deposits are the main objective of the exploration geochemical surveys. However, due to complexity of ore-forming processes, multiple significant geochemical signatures may reflect the deposit of the type sought. Therefore, combination of significant geochemical signatures to an enhanced one, which could be used for targeting undiscovered mineralization, is another challenging aspect of exploration geochemistry. This study tackles the foregoing challenges by (a) using the receiver operating characteristics (ROC) curves for discriminating significant and nonsignificant geochemical signatures and (b) proposing an ROC-based weighted aggregation matrix, called RWM, for synthesis of individual geochemical signatures to a single enhanced signature of the deposit-type sought. To demonstrate the effectiveness of the proposed methodology, stream sediment geochemical data from the Varzaghan District, northwestern Iran, were employed to target skarn copper deposits. The proposed methodology and fuzzy logic operators were applied to integrate individual geochemical signatures, the comparison of which revealed that the former was superior to the latter. The proposed RWM procedure not only effectively combines individual geochemical signatures, but also serves as a bivariate data-driven procedure for integration of layers of evidence for mineral prospectivity modeling.
