Geostatistical Model Construction And 3D Metallogenic Prognosis Of Shambesai Gold Deposit

Resource optimized estimation and qualitative and quantitative prediction of mineralized target areas are very important for geological exploration institutions and mining companies.Resource estimation of defined ore bodies and prediction and location of deep and sideward concealed ore bodies are critical throughout the whole life cycle of exploration.It is of great significance to study the related theories,methods,and key parameters of resource assessment and three-dimensional prediction.In this paper,the Shambesai Carlin-type gold deposit research area in Kyrgyzstan is taken as the main study object.Surpac,Leapfrog,Supervisor,and other software are used as the main research tools.Based on the collecting,sorting and analysis of geological multi-source data,and the geostatistics and three-dimensional metallogenic prediction theory as the main research methods,the research including three-dimensional geological modeling,geostatistical estimation parameter optimization,resource estimation and evaluation,and three-dimensional metallogenic quantitative prediction of Shambesai gold deposit have been completed.The main works completed in the paper are as below:(1)Based on the analysis and collation of the original geological data of the mining area,the multi-source geological database is established.Three-dimensional modeling of various surface models and solid models,such as topography,strata,faults,geophysical anomaly areas,geochemical anomaly areas,and mineralized bodies,has been completed.Meanwhiles,a serial of geology modeling related data including alteration type and extent,rock type,mineral type,structure and texture,oxidation extent,weathering zone,and so on are analyzed and summarized.(2)Based on the analysis of the main factors affecting the stability of variogram of typical metal deposit data experiment,including the selection of sampling length,sample composited method and length selection,outlier samples selection and capping,data skew distribution proportion effect,etc.,it's confirmed that various factor optimization methods can ensure a more stable experimental variogram curve,and will greatly simplify the variogram fitting process.(3)By using Supervisor geostatistics software,the main ore bodies of the deposit are fitted with borehole directional variation function to extract nugget value,and three directional variogram is fitted to extract search ellipsoid parameters and Kriging estimation parameters.At the same time,the variogram fitting curve equations under the exponential model are calculated,and the corresponding transformation matrix is created.(4)Kriging efficiency(KE),Slope of regression(SR),Covariance(CO),and other parameters are used to optimize the parameters of the Kriging estimation process.After parameter optimization,the block size,maximum samples,maximum samples of per hole,and discretization coefficient are all optimized.The relative errors between multiple Kriging block estimation results and original sample results are reduced,and the accuracy of estimation results is highly improved.(5)The ore bodies'resources are estimated by the Kriging method,and the evaluation results are evaluated by swats plots and grade-tonnage cumulative distribution plots.At the same time,the IDW method and IK method are used to carry out the secondary estimation of the ore bodies,and the results of the estimation are cross-validated and the applicable condition of different estimation methods is summarized.(6)Through the correlation analysis of the distance and density field of each ore controlling factor,the strata constraint,fault constraint,geophysical anomaly constraint,three geochemical element anomaly areas of mercury,antimony and arsenic,and the density distribution of pyrite indicator minerals are determined as the main ore controlling factors;the data of ore-forming correlation degree is obtained through the composite calculation based on the membership degree of grey system theory;the data of ore-forming advantage degree is obtained by combining the ore-forming correlation degree with the Fuzzy-Hierarchy evaluation method while the fuzzy complementary matrix and the fuzzy consistent matrix are constructed;the multiple linear regression equation of the mineralization factors and the distribution of the mineralization grade is established,which is applied to the prediction of the mineralization block to get the predicted grades;finally,three main target areas are defined in the study area of the Shambesai gold deposit by combining the data of the ore-forming advantage degrees and the predicted grades of the ore blocks.The development and research results of 3D metallogenic prediction of Shambosai gold deposit provide a favorable basis for 3D metallogenic prediction of concealed ore bodies in similar mines.