Reducing exploration risk by data-driven uncertainty quantification in mineral prospectivity mapping

The most sustainable mine is the one that should not have been mined because of lack of sufficient resources. Mining exploration and appraisal heavily relies in geological, geophysical and geochemical analysis, but is subject to considerable uncertainty. In that context, mineral prospectivity mapping (MPM) is particularly important for exploring undiscovered mineral deposits. However, MPM is subject to considerable uncertainties due to lack of understanding in metallogenesis and limited spatial data samples. In this work, we provide a framework that addresses how uncertainty in the evidence layers can be quantified and how such uncertainty is propagated to the prediction of mineral potential. More specifically we will use Monte Carlo to jointly quantify uncertainties on all uncertain variables, categorized into geological, geochemical and geophysical. On each sample of the multiple input layers, logistic regression is employed to produce different quantification of the mineral potential in terms of probability. Finally, we will provide a novel way to interpret the established uncertainty in a risk-return analysis to decide high probability prospective areas. And hence avoid further unnecessary exploration drilling in high risk areas. Our methods are illustrated on a real case study of prospecting skarn Fe deposition in southwestern Fujian, China. It quantifies the prediction uncertainty of potential mineral deposits in the study area, with results tested against known Fe deposits. More importantly, the risk-return analysis enables the risk and return to be tradeoff on economic analysis and decision making for mineral explorations in this area.