基于成分数据分析和多重分形滤波技术识别青海红垭豁地区地球化学异常  

Identification of geochemical anomalies in the Hongyahuo area,Qinghai Province based on the component data analysis and multifractal filtering technique

作　　者：刘永胜 罗先熔 高文 刘攀峰 梁鸣 杨青松 石佳磊 孙瑶 LIU Yong-sheng;LUO Xian-rong;GAO Wen;LIU Pan-feng;LIANG Ming;YANG Qing-song;SHI Jia-lei;SUN Yao(Institute for Prediction of Hidden Deposits,School of Geoscience,Guilin University of Technology,Guilin Guangxi 541004,China;Gansu Non-ferrous Engineering Exploration&Research Institute,Tianshui Gansu 741024,China)

机构地区：[1]桂林理工大学地球科学学院,隐伏矿床预测研究所,广西桂林541004 [2]甘肃省有色金属地质勘查局天水矿产勘查院,甘肃天水741024

出　　处：《矿物岩石地球化学通报》2023年第3期602-614,674,共14页Bulletin of Mineralogy, Petrology and Geochemistry

基　　金：国家自然科学基金资助项目(42203067);国家重点研发计划项目(2017YFC0602604HT02);广西科技计划青年基金项目(2022GXNSFBA035548);广西高校中青年教师科研基础能力提升项目(2022KY0261)。

摘　　要：红垭豁地区位于青藏高原东北缘,秦祁昆晚加里东造山系中祁连构造带内,区内岩浆岩分布广泛,成矿条件优越。为有效提取矿致地球化学异常,本文采用主成分分析(PCA)和多重分形模型(S-A)方法对研究区沟系641件土壤样品异常进行了提取及ilr(对数比)变换,并结合密度曲线及箱线图,对比三类数据(原始、对数、ilr变换)元素的多元统计含量特征,运用S-A提取研究区地球化学异常及背景。结果显示,ilr变换后数据消除了数据闭合效应,元素含量更接近正态分布。主成分分析得到2类不同的元素组合,PC1为Cu、Fe、Mn、Ni、Pb、Zn元素组合,即中高温热液成矿元素组合形式,代表与侵入体和断裂有关的Cu-Pb-Zn多金属矿化,反映了铜多金属矿成矿过程的叠加;PC2为Au、As、Sb元素组合,代表与金锑矿化有关的低温热液活动。PC1分解图显示,PC1低背景、高异常主要分布在安山岩地层中,且周围附近有中酸性侵入岩及构造破碎带,为铜多金属矿提供物质来源,经异常查证,发现铜矿体1条。研究认为,红垭豁地区具有良好的找矿潜力,S-A所圈定的综合异常可为下一步的地质找矿工作提供依据。The Hongyahuo area is located in the northeast margin of the Tibetan plateau and is tectonically located in the central Qilian tectonic zone of the QiCaledonian orogenic belt.There are widely distributed magmatic rocks in the area,with excellent metallogenic conditions.In order to effectively extract geochemical anomalies related to the mineralization,in this paper,by using the principal component analysis(PCA)and multifractal model(S-A)methods,we have made the extraction and the ilr(log-ratio)transformation of the anomalies outlined based on geochemical data of 641 soil samples from the ditch system within the study area.In combination with density distribution curve chart and box plot,with comparison of the multivariate statistics content characteristics of three types of data(original,logarithmic and ILR transformed data)for various elements,we have extracted geochemical anomalies and backgrounds of various elements in the area by using the S-A method.The results show that the ILR transformation of raw data has resulted in the elimination of the original data closure effect and the near normal distribution of contents of various elements.Two different types of element association have been obtained by using the principal component analysis.The PC1 is a Cu-Fe-Mn-Ni-Pb-Zn element association which is a form of the medium-to-high temperature hydrothermal ore-forming element association,representing the Cu-Pb-Zn polymetallic mineralization related to the intrusive body and fault,and reflecting the superposition of copper polymetallic ore-forming process.The PC2 is an Au-As-Sb element association which represents the low temperature hydrothermal activity related to the gold-antimony mineralization.The PC1 decomposition diagram shows that the PC1 ones with low backgrounds and high anomalies are mostly distributed corresponding to the andesite outcropped parts with intermediate acid intrusive rocks and structural fracture zones nearby for providing material source to the copper polymetallic mineralization.Based on th

关 键 词：成分数据 ilr数据变换 主成分分析 S-A多重分形 红垭豁地区 青海 

分 类 号：P595[天文地球—地球化学]