Analysis Of Multi-scale Geochemical Data And Mineral Prospectivity Mapping

China’s mineral resource endowment is uneven,and key mineral resources that are critical to the country’s economic survival face long-term supply constraints and high external dependence.For example,China’s local supply of copper mining resources has been scarce for a long time,and the country is 75% reliant on imports.According to the data analyzed by Antaike in 2017,the accumulated consumption of refined copper in China has been nearly 11 million tons,accounting for about 50% of the total global consumption,and it is expected that China will still be the country with the largest demand for copper in 10 to 15 years,and the consumption will account for about 40%of the global consumption.Therefore,rapid exploration of mineral resources in the coming period and increasing the national strategic resource reserves are still important research tasks.Exploration geochemical methods have played a pivotal role in the past mineral resources exploration,while accumulating a huge amount of geochemical data.With the reduction of surface ore and the increasing difficulty of finding ore year by year,it is one of the effective means to realize the increase of mineral resources reserves in China by paying attention to the introduction of advanced data analysis methods,making full use of the existing data and striving for new breakthroughs in ore finding based on the original exploration work.The Tibetan Gangdis is a key copper-polymetallic ore metallogenic belt in China,with three mega copper mines at Jiama,Qulong,and Xiongcun,as well as five notable copper mines,including Zhunuo.The resource reserves of copper in just two deposits,Qulong and Jiama,exceed 20 million tons.Furthermore,the Gangdis’ eastern section is rich in Mo,W,Pb,Zn,Ag,Au,and other metal minerals,making it a promising ore-finding area.As a result,this work uses the Qulong-Jiama copper-polymetallic ore region in the Tibetan Gangdis metallogenic belt as a multi-scale study area,identifying elemental anomalies using catchment basin analysis and Multifractal S-A model,and extracting elemental anomalies.Knowledge-driven and data-driven compositional data analysis approaches combine features,and machine learning and deep learning methods estimate mineral resource potential.A set of multi-scale geochemical mineral prediction and evaluation tools has also been created as a scientific foundation for mineral resource development in the Tibetan Gondian mineralization zone.(1)Catchment basin geochemical analysis was used to extract regional geochemical anomalies in the Lhasa area of Tibet(regional scale),and knowledge-driven compositional data analysis methods were used to extract elemental assemblages of regional carbonates,granites,paleocollision zones,volcanic rocks of the Linzizong formation,and polymetallic mineralization,and a regional integrated exploration model was established by combining regional geological mineralization laws.(2)Twelve mineralization prediction factors,including stratigraphic,tectonic,magmatic rock,geophysical,geochemical,remote sensing,and other multi-source information,were quantitatively extracted under the guidance of the regional integrated exploration model to form a regional quantitative mineralization search model.Mineral search information quantity method,right of evidence method,hybrid Gaussian model,maximum entropy model,and convolutional neural network were used to identify six prospective areas.(3)In the Jiama-Qulong mineralized area(mineralized area scale),the geochemical background and anomalies of mineralized elements and indicator elements were extracted using the Multifractal S-A model,and the characteristics of elemental assemblages were thoroughly analyzed using the catchment basin analysis method,based on which the directions for further exploration of the mineralized area were pointed out.(4)The data-driven compositional data analysis method inferred the stratigraphy,magmatic rocks,tectonics,alteration,mineralization,and mineralization zoning characteristics of the Jiama copper-polymetallic mining area,and the knowledge-driven compositional data analysis method inferred the degree of denudation of the mining area.On the basis,a comprehensive exploration model of the Jiama mining area was constructed,and it was pointed out that the Mogulang area has the potential to become the third mineralization center of the Jiama mining area,and new porphyry-silica type ore bodies may be found in the Xiangbei Mountain-Hongshantou area.In conclusion,this work provides a set of methods for multi-scale geochemical data analysis and mineral prospectivity mapping,and obtains good application results in Lhasa,Tibet,through multi-scale exploration,which can give strong support for quick mineral resources exploration.