2002 Volume 13 Issue 3 Pages 167-173
In recent years, it became easy to analyze the two dimensional data which covers a wide area using the various data with the progress of the GIS software and superior performance of the personal computer. As for the mineral resource exploration field, too, GIS and the program which customized this become used for the selection of the promising area, the selection of the drilling point using geological feature data, geophysical prospecting data, geochemical prospecting data, and so on. The application of the fuzzy set theory is thought as one of the ways of analyzing prospecting data. This method, at first, select known deposits theme and buffer theme applied the distance from exploration index. Buffer theme mean polygon set like tree rings made by constant distance from each exploration index, i.e. geology, deposit, geochemical, geophysical data and so on. Next, decide the membership function to every exploration index and to newly created all grid points, and outputs favorability mapping for mineral exploration. In this study, the fuzzy buffer program was created which can process a series of such processes of the analysis easily by customizing the macro language of GIS. The favorability mapping which is output by the created fuzzy buffer program was verified. As the object area, Hokuroku area, Akita Prefecture was taken up (31 km of east-west direction, 44km of north-south direction). In this area, a lot of Kuroko deposits were discovered, and gathered many kinds of prospecting data. Here, as the known deposit theme, 1km grid points within 1km from the ore bodies were set. As the buffer theme of exploration criteria, seven themes were used in this research, namely, K2bufs(the edge of submarine dacitic sequence and upper layer), K3bufs(upheaval circular structure in submarine dacitic sequence and upper layer), K6bufs(dacite/rhyolite dyke), K8bufs(andesite dyke), K10bufs(vein type deposits, hanging wall stratum is host rock), K12bufs(alteration index) and gradient of gravity, they showed effective criteria in the result of multiple regression analysis. The fuzzy set which is allocated with 1/exp function as a membership function, and the GAMMA operator was used to estimate fuzzy score. As a result, the favorability mapping got to have harmonized with the known deposit well with the multiple regression analysis, reflecting a deposit genesis having to do with learning after using only an effective prospecting index, not using an unnecessary prospecting index. As one of the evaluation methods of the favorability mapping result by the various analysis techniques, the ROC curve (receiver operating characteristic curve) was created. As a result, in case of the fuzzy set application, it is possible to estimate that 50% of ore bodies are discovered at 10% of prospecting area and that the discovery percentage exceeds 80% at 23.5% of prospecting area. It is expected that application of fuzzy set theory is available to discover ore bodies in Hokuroku area, and to do prospecting more efficiently
