Fuzzy Decision-making Assessment Model For Cu-Zn Deposits By Multi-source Information In Yushigou

This paper describes an application of a method called compromise fuzzy multipleattribute decision which is built to prospect Cu-Zn deposits in Yushigou, Qinghai,China. Multi-source data sets involving geological, geochemical and remote sensingdata layers are related and analyzed by GIS technology and applied in thecompromise fuzzy multiple attribute decision making method for prospecting.Compromise fuzzy multiple attribute decision making method develops from classicalmulti-criteria fuzzy multiple attribute decision making. The basic principle of thismethod is described as follows: starting from the original sample data, virtual fuzzypositive idea index and fuzzy negative ideal index are computed, of which the fuzzypositive ideal is the maximum value of fuzzy index and fuzzy negative ideal is theminimum index value. Then distance between each candidate object and the fuzzypositive ideal and fuzzy negative ideal is calculated by Weighted Euclidean Distancemethod. At last, relative-membership grade of each candidate object belongs to thefuzzy positive ideal is calculated for interpretation of mineralization possibility ofeach candidate area.The data sets are collected involving remote sensing images of AdvancedSpaceborne Thermal Emission and Reflection Radiometer (ASTER),and EnhancedThematic Mapper data, foundation geological map (1:50,000scale), and geochemicaldata (1:200,000scale).Mapping surface alterations using remote sensing sensorsprovides beneficial information for the reconnaissance stages of copper and zincexploration especially at regional scales. Hydrothermal alteration zones associatedwith copper and zinc deposit in study area such as silicification, carbonatization andchloritization mineral assemblage can be discriminated from one another by virtue oftheir spectral absorption features, which are detectable by ASTER and ETM data.Principal Component Analysis (PCA) and Band Ratio methods applied forinterpretation and identification of hydrothermally altered rocks and associatedmineralization. Ore-controlling faults and host rock areas are abstracted from remotesensing data and which the result is the verified by geological map. According to the geology of the study area, peridotite, diabase and basalt is the host rock of the Cu-Zndeposit. Thus, host rock units are mapped by thermal infrared bands. Geochemicalinformation layers containing copper anomaly and zinc anomaly were interpolatedfrom geochemical data derived from chemical exploration (1:200,000scale) in1980s.They were interpolated by an inverse distance weighted method to map theiranomalies, and each of the Cu and Zn dataset were transformed into standardnormalized distributions.6layers are considered as key criteria and To use the compromise fuzzy multipleattribute decision making method. Cell values of raster data sets associated with thesecriteria are extracted and stored in a database as alternatives. The comprehensivemembership is obtained by applying all evidential layers and its weights in thecompromise fuzzy multiple attribute decision making method. A program is written inMatlab platform to execute all procedure of compromise fuzzy multiple attributedecision making method and to compute the final comprehensive membership, whichis mapped to an isogram for interpreting the prospecting area.Compared with the results obtained by fuzzy comprehensive evaluation model,which is a classical comprehensive evaluation method based on Fuzzy Mathematicsand applied well in mineral resources exploration, the result from the compromisefuzzy multiple attribute decision making method has better effect than the classicalone. Metallogenic prospect area is located at the junction of faults and secondaryfracture and indicates that the strong fracture development convergence zone has goodore prospects.