Improvement in mineral prospectivity maps results using a hybrid ANP-TOPSIS approach

Accuracy incretion and search space reduction are the main purposes in production of mineral prospectivity maps. Choice of the suitable method for information integration in the geographical information system which has the most consistent with the nature of the used data, is the first step in achieving to this goal. Due to the fact that exploratory data are not independent and most of them has feedback among criteria and sub-criteria and according to the one-way dependence condition of the evaluated data in the Analytic Hierarchy Process (AHP) approach, an Analytic Network Process (ANP) approach is assigned. Exploratory data related to the Bavanat Region in Fars province are used, in case of providing the mineral prospectivity model for besshi type massive sulfide deposits. We tried to optimize two factors of the search area and accuracy to highest accuracy of the results and the least amount of search area. AHP as well as ANP approach, is knowledge-based approaches. The results of the research in AHP and ANP approaches in studied area indicate that, the ANP results is more reliable and has less suggested search area than AHP approach. Based on the results, suggested areas of ANP are 17% less than AHP (1138.4 Km2 in AHP has decreased to 952.6 Km2 in ANP), while, the accuracy of the results has increased more than 6%. The Receiver operating characteristic (ROC) curve index (area under the curve) shows that AHP approach ROC is 0.7609, which was improved and optimized by the ANP approach to 0.8275.