Multi-scale Spatial Distribution Characteristics And Mineral Prospectivity Mapping Of Hydrothermal Tungsten Polymetallic Deposits In The Nanling Metallogenic Belt

Many researches have shown that the structuration of crust and lithosphere is a self-organized critical?SOC?phenomenon.Since the Mesozoic,large-scale extension and thinning of the lithosphere occurred in the eastern part of South China,forming a large number of crustal-remelting granite.The processes that the tungsten polymetallic deposits occurred explosively in the Nanling metallogenic belt and the ore-forming elements highly enriched in the ore-forming fluid,eventually forming large amounts of tungsten-bearing materials accumulated in narrow fractures are typical singularity processes.As a result of singularity process,the mineral deposits should satisfy fractal distribution and hold the properties of anisotropy and inhomogeneity.In this paper,the Xihuashan and Piaotang tungsten polymetallic deposits are selected as the research objects to explore the evolution of structure-fluid,spatial pattern analysis methods are applied to quantify the anisotropy and inhomogeneity of the distribution of tungsten polymetallic deposits on the scales from metallogenic belt to the deposit.On this basis,the spatial distribution pattern of tungsten polymetallic deposits is introduced in mineral prospectivity mapping to reduce the impact of data imbalance,so as to improve the prediction accuracy and reduce the exploration risk which can provide guidance for prospecting in this region.According to the purposes and contents of this research,the following work are conducted:?1?Collecting the geological maps of various scales,geophysical data,geochemical data systematically,and summarizing the previous researchese on the sources of ore-forming materials and the evolution of ore-forming fluid of hydrothermal tungsten polymetallic deposits in the NMB;?2?Taking Xihuashan and Piaotang hydrothermal tungsten polymetallic deposits as research objects,field geological survey,such as documentation of borehole core,gallery geologic sketch,measurement of vein thickness and vein spacing and collection of geological samples,are conducted to study the spatial distribution of ore-bodies,ore-controlling structures,ore structures,paragenetic association of minerals and wall-rock alteration;?3?Carrying out laboratory works,including the grinding of polished thin section and thermoscope-pieces of typical samples,identification of polished thin section to study the ore fabric,the mineralization periods and the assemblages of altered minerals;the temperature of the inclusion was measured,and the composition of the inclusion is analyzed by means of laser Raman to study the variation of parameters during the evolution of ore-forming fluids;and the measurements of vein thickness and vein space are analyzed to discuss the growth mechanism of ore-bearing quartz vein;?4?Quantitative analysis of the spatial distribution pattern of tungsten polymetallic deposits from the metallogenic belt scale to the deposit scale to quantify the anisotropy and inhomogeneous of mineralization;?5?According to the metallogenic model,the predictive model of tungsten polymetallic deposits is established,the spatial distribution pattern of tungsten polymetallic deposits on the metallogenic belt is introduced in the mineral prospectivity mapping and the prediction and evaluation of mineral resources is conducted.Based on the above research,the following conclusions are preliminarily drawn:?1?The homogenization temperature of fluid inclusions in Xihuashan tungsten polymetallic deposit concentrically ranges from 184.45 to 223.80?,the sanity concentrically ranges from 4.96 to 10.11 wt%,the density ranges from 0.68 to 1.01 g/cm³,the ore-forming pressure is concentrated from 49.18 to 60.35 MPa,and the ore-forming depth is nearly 5.89 km.The main components of gas phase in the fluid inclusions are CO₂?CH₄.This results indicate that the ore-forming fluid are reductive fluid which hold the properties of medium temperature,medium-low salinity and medium-low density.And the ore-forming fluid can be divided into three stages.During its evolution,the ore-forming fluid experienced processes of mixing with low-temperature and low-salinity fluid and natural cooling.?2?The width and spacing of ore-bearing quartz-vein in the hydrothermal tungsten polymetallic deposit satisfy fractal distribution.Meanwhile,the fractal dimension can be used to quantify the intensity of mineralization.When the fractal dimension is greater than 1,the mineralization is weak;when the fractal dimension is smaller than 1,the smaller the fractal dimension,the stronger the mineralization.?3?The formation of hydrothermal vein-type tungsten deposits is the result of interactions of many factors,such as structure and fluid.On the one hand,the formation of multiple hydraulic faults provides important channels for fluid migration and convergence.Atmospheric precipitation mixes with fluid enriched in W?Sn and other elements,resulting in a rapid decrease of temperature and pressure,increase of pH,Eh,and oxygen fugacity,loss of CO₂.As a consequence,tungsten-bearing complexes decomposed,releasing large amounts of WO₄^2-.On the other hand,acid-alkali separation occurs in the solution,and the volatile components,such as F and Cl,interacted with wall-rocks,forming minerals such as fluorite.And the alkali metal elements,such as K,Na,and Al,interacted with wall-rocks,forming kfeldsparization,albitization and greisenization.During this process,the oxygen fugacity,pH and Eh of the system decreased,and elements,such as Fe and Mn,in the wall-rocks entered into the solution in the form of low-valent,such as Fe²⁺,Mn²⁺.Under the condition of low oxygen fugacity,pH and Eh,these low-valent metal ions bound with WO₄^2-in the solution and formed the wolframite aggregates in the multiple hydraulic faults.?4?The spatial pattern analysis can quantify the anisotropy and inhomogeneity of the distribution tungsten polymetallic deposits on different scales.Results of Fry analysis show that the tungsten polymetallic mineralization hold the property of aisotropy and the distribution of tungsten deposits on different scales are controlled tectonics since the Yanshanian period.Specifically,the distribution of tungsten deposits on metallogenic belt and ore concetration scales are dominated by NE-NEE striking faults,followed by NW-,and NS-striking faults;while the distribution of tungsten vein on deposit scale is dominated by NEE-striking faults.And the results of average nearest neighbor distance analysis,K function and fractal analysis indicate that the tungsten polymetallic mineralization hold the property of inhomogeneity and the distribution of tungsten deposits satisfy fractal cluster distribution from the metallogenic belt scale to the deposit scale.?5?Introducing the fractal cluster distribution of tungsten deposits on the metallogenic belt scale to the mineral prospectivity mapping using ensemble learning algorithms can be used to scrren non-deposit locations.And the SMOTE sampling technique can be applied to generate balanced datasets to reduce the impact of data imbalance,which can improve the prediction accuracy and reduce the exploration risk.Meanwhile,in the case of data imbalance,when evaluate the predictive models,the improved P-A plot is more reliable than ROC curve and P-A plot,which may give an overly optimistic evaluation.?7?According to the mineral prospectivity mapping,the geological background,metallogenic probability and economic geography factors,seven A-type prospective areas,such as Dayu-Chongyi-Shangyou,Shixing-Quannan,Qitianling-qianlishan,etc.are mapped and five B-type prospective areas including Le'an-Yihuang,Chaling-Lianhua,Jiangyong-Guanxian,Fengkai-Huaiji and Lianzhou-Ruyuan are delineated.?8?The radial-density models of tungsten polymetallic deposits and metal contents are established.According to the radial-density models,it is conservatively estimated that the number of undiscovered tungsten deposits in this region is 70,and the amount of undiscovered metal amount?WO₃?is about 1.3798 Mt,indicating that this region still has good prospecting potential for tungsten polymetallic deposits.