首页|基于GF-5卫星的西藏珠勒—芒拉地区矿物蚀变信息提取及找矿前景分析

基于GF-5卫星的西藏珠勒—芒拉地区矿物蚀变信息提取及找矿前景分析

扫码查看
原文链接
  • 国家科技期刊平台
  • NETL
  • NSTL
  • 万方数据
[研究目的]近年来,遥感在地质调查和矿产勘查领域取得了广泛的应用,基于多光谱遥感数据的蚀变矿物填图为地质找矿工作提供了重要技术支撑,然而基于国产高光谱遥感数据在此领域的研究却为数不多.高分五号(GF-5)较小的波谱间隔提供了相比于多光谱更为丰富的目标地物波谱信息,为矿物的精细识别提供了良好的数据源.本文主要基于GF-5开展西藏革吉南地区的矿物蚀变信息提取,同时结合Landsat-8、ASTER多光谱数据提取结果叠加对比,综合野外调查验证,进一步深化遥感在地质矿产资源调查领域的应用.[研究方法]基于多光谱数据建立了不同类别蚀变矿物的光谱指数模型,在GF-5数据蚀变信息提取方面,摒弃了传统的光谱角匹配等方法,提出了基于决策树分类辅助混合调谐匹配滤波技术进行矿化蚀变信息的提取方法,最后综合区域构造、蚀变信息提取结果等要素,圈定成矿有利区,并开展野外调查验证.[研究结果]基于Landsat-8、ASTER两种多光谱数据对铁染、羟基类(Mg-OH、Al-OH)、碳酸盐类矿物信息进行了增强与提取;基于GF-5数据识别出了方解石、钠云母、普通白云母、多硅白云母、明矾石、高岭石、地开石、绿帘石 8种蚀变矿物.[结论]结合不同数据源的提取与叠加结果,证实了本文提出的矿化蚀变信息提取方法的可行性.根据野外验证情况综合揭示了该地区发育高硫型浅成低温热液蚀变矿物组合,具有斑岩-浅成低温热液矿床的成矿潜力.本文认为高光谱与多光谱数据相结合有助于后续蚀变分带的分析与更精确的成矿预测,从而更好地服务于矿产勘查工程等领域.
Extraction of mineral alteration information and mineralization prospecting analysis based on GF-5 hyperspectral in Zhule-Mangla,Tibet
[Objective]Remote sensing has been widely used in geological survey and mineral exploration in recent years.Alteration mineral mapping using multispectral remote sensing data provides important technical support for geological prospecting.However,only a limited number of surveys were carried out based on the Chinese hyperspectral remote sensing data.The smaller spectral interval of the GaoFen-5(GF-5)hyperspectral remote sensing data provides a richer spectral information of geological targets than multispectral data,which offers an ideal data source for mineral identification.This paper was mainly focusing on the extraction of alteration minerals based on GF-5 data in the south of Geji,Tibet.Moreover,the alteration minerals extracted from GF-5 were overlaid and compared with the results extracted from Landsat-8 and ASTER data.The results were verified by field survey,which could deepen the application of remote sensing in mineral resources investigation.[Methods]A spectral index model for different alteration minerals was proposed based on the multispectral data.During the extraction of alteration information from GF-5 data,traditional methods such as spectral angle mapper were abandoned,while the decision tree algorithm was adopted to support mixture tuned matched filter for the extraction of mineralized alteration information.Finally,regional structure,alteration information and other factors were integrated to delineate mineralization anomalous targets.The field investigation was carried out to verify the technical reliability.[Results]The information of iron-stained,hydroxyl(Mg-OH,Al-OH)and carbonate minerals was enhanced and extracted using Landsat-8 and ASTER.Eight alteration minerals including calcite,paragonite,muscovite,phengite,alunite,kaolinite,dickite and chlorite were identified by GF-5.[Conclusions]Combined with the extraction and superimposition results from different data sources,the reliability of the mineralization alteration information extraction method proposed in this paper was confirmed.The field investigation results showed that the area is characterized by high-sulfur epithermal alteration mineral assemblage,which has the potential for porphyry-epithermal hydrothermal deposit.It is suggested that the combination of hyperspectral and multispectral could help the subsequent analysis of alteration zonation and provide accurate mineralization prediction for prospecting,so as to serve the sector of mineral exploration engineering.

mineralized alteration informationGF-5spectral indexdecision treemixture tuned matched filterporphyry depositsmineral exploration engineeringTibet

白龙洋、代晶晶、王楠、李宝龙、刘治博、李志军、陈伟

展开 >

中国地质科学院矿产资源研究所自然资源部成矿作用与资源评价重点实验室,北京 100037

中国地质科学院自然资源部深地科学与探测技术实验室,北京 100094

成都理工大学地球科学学院,四川成都 610059

矿化蚀变信息 GF-5 光谱指数 决策树 混合调谐匹配滤波 斑岩矿床 矿产勘查工程 西藏

国家重点研发计划中国地调局项目中国地调局项目科技部国家重大研究计划国家自然科学基金中央级公益性科研院所基本科研业务费专项

2022YFC2905001DD20230054DD202300332021QZKK030442172332KK2102

2024

10.12029/gc20220701002

中国地质
中国地质调查局

中国地质

CSTPCD北大核心
影响因子：2.317
ISSN：1000-3657
年,卷(期)：2024.51(3)
  • 1
  • 39