Three-dimensional prospectivity modeling based on fuzzy logic and geometric average: a case study from Sansong district, Yongyu apatite deposit, DPR Korea

In this paper, we applied two kinds of knowledge-driven methods, one using the fuzzy logic and another using geometric average to evaluate 3D mineral prospectivity for Sansong district, Yongyu apatite deposit, in the Phyongnam Basin, DPR Korea. Based on the ore geology studies of apatite deposits and the available spatial datasets in the study area, we used four independent evidential maps for 3D apatite deposit prospectivity modeling. They include (1) carbonatite, (2) biotite gneiss, (3) granitic gneiss, and (4) P2O5 values from the borehole data. The evidential factors were modeled into 3D space, and 3D P2O5 values from the borehole data were transformed into continuous values of the [0, 1] range using logistic sigmoid. In 3D MPM just as 2D MPM, it is very economic and efficient to simultaneously apply the fuzzy logic and geometric average methods for mineral prospectivity modeling of the study area because two predictive models can use the same fuzzification methodology based on fuzzy membership function. Our strategy is to fuzzify the evidential maps before applying the geometric average as well as in the fuzzy logic. The results for the two predictive models were validated by the prediction efficiency method. The results demonstrated that most of the validation data were distributed in voxels with high prospectivity values. Although the validation results were slightly worse than those in 2D MPM case study, our case studies suggested that both predictive models and their modeling results are useful for evaluating 3D prospectivity of apatite deposits in Sansong district, Yongyu apatite deposit, DPR Korea.