Union score and fuzzy logic mineral prospectivity mapping using discretized and continuous spatial evidence values

Two common problems affect integration of exploration criteria for mineral prospectivity mapping (MPM) in geographic information system (GIS): (a) stochastic error associated with sufficiency in number of known mineral occurrences (KMOs) used to estimate evidential weights and (b) systemic error associated with subjectivity of expert judgment applied to process, analyze, and assign weights to evidential data. In this paper we used logistic sigmoid (or S-shaped) function to transform continuous-value evidential data into logistic space without using KMOs as in data-driven MPM and without discretization of evidential data into classes by using arbitrary intervals based on expert judgment as in knowledge-driven MPM. We generated a prospectivity model using discretized evidential data as well. Then, we compared the prospectivity models generated using continuous- and discretized-value evidential data and demonstrated that the former is better model for selecting target areas for further exploration.