Mineral prospectivity mapping for podiform chromite deposits using continuously-weighted evidence maps in Sabzevar ophiolitic belt

In regional exploration, various evidential exploration layers, i.e., remote sensing, geology, geochemistry and geophysics, are extracted, analyzed and integrated to delimit prospective areas for further exploration phases. In this paper, the evidential maps of host rock (elicited from Sentinel-2B satellite images), fault density and geochemical signature were fuzzified using the continuous (based on logistic function) and knowledge-driven (discretization of evidential data into classes and then assigning the same weight to the values in each class of evidential features based on expert judgment) approaches and then integrated using the fuzzy gamma and geometric average operators to generate mineral prospectivity mapping. The receiver operating characteristics (ROC) and prediction-area (P-A) plots were utilized to appraise the generated prospectivity models. The outcomes of this paper demonstrated that the prospectivity models generated with continuously-weighted spatial evidence values is superior to those of the prospectivity models generated with discretely-weighted exploration evidence data. Consequently, the former prospectivity models could be utilized to selecting target areas for further exploration of deposit-type sought in the study area.