金矿床遥感影像模式及证据权重法在找矿预测中的应用

河南熊耳山地区为我国金矿的重要成矿区,为加速金矿资源评价和找矿勘探工作,利用遥感技术对该区金矿成矿规律进行了研究,并进行了金矿找矿靶区优选。首先利用多种方法进行TM数字图像处理,以获取理想的遥感图像,然后进行遥感影像信息解译分析,结合区域地质调查和典型金矿床解剖研究,确立主要控矿因素(如构造、含矿地层、岩体、蚀变等).根据金矿分布规律和从遥感图像上提取的综合影像特征,建立了不同金矿床类型遥感影像模式。

APPLICATION OF REMOTE SENSING IMAGERY MODEL FOR GOLD DEPOSIT AND "EVIDENCE-RIGHT-WEIGHT"METHOD IN GOLD DEPOSIT PREDICTION

Xiongershan region, Henan province is an important gold-mineralized area in China. Research of gold mineralization and selection of prospective targets in the region have been carried out by using remote sensing technology. First of all, TM image data are processed through various methods to get ideal remote sensing images, then, the image information is interpreted and analyzed in combinaticon with regional geological survey and typical gold deposit study to determin main ore control factors, such as structures,ore-bearing strata, rock bodies, and alteration, etc. The remote Sensing imagery models of different types of gold deposit have been set up on tho basis of extraction of comprehensive image signatures and gold deposit distribution pattern.In the light of the models, prospecting targets have been chosen by using "evidence-right-weight" graphics information integration. The method,which presents favourable geological factors of ore control as binary graphics and calculates their "evidence-right-weight", is based on Bayes rule, By use of geographic information system (GIS), mutual independent geological factors relative to the deposit, mineralization evidence, are integrated statisticaly for calculating posterior probability of gold deposit distribution,which can be used to indicate potential gold-mineralized area, The method,combination of remote sensing, statistics and expert’s experiences, is provided with a rigorous theoretical basis and visual, fast, and flexible characterastics,and is convinent for geological application.After practice, it is concluded that according to remote sensing imagery models of different type of gold mineralization and "evidence-right-weight"graphics information integration method to choose prospecting targets is succeesful. Various extents of gold mineralization have been found in 8 first class and 5 second class potential areas predicted by us after field testing . The result is very obvious.