机构地区:[1]自然资源实物地质资料中心,河北廊坊065201
出 处:《光谱学与光谱分析》2025年第3期742-752,共11页Spectroscopy and Spectral Analysis
基 金:国家自然科学基金项目(42202330);地质调查项目(DD20230138)资助。
摘 要:近年来,高光谱技术因其可以绿色、快速、无损探测物质成分和提高勘查效率而备受关注,前人应用光谱对稀土元素及其化合物,稀土矿物的定性识别,以及对稀土元素的定量反演等方面开展了一系列研究,但对碳酸岩型稀土矿床稀土含量定量反演的研究相对缺乏。应用HyLogger型岩心光谱扫描仪,获取大陆槽碳酸岩型稀土矿床典型矿石和岩石样品的V-NIR-SWIR波段的高光谱反射率数据,通过镜下鉴定明确矿石、岩石的属性,并通过全岩微量实验测试获取稀土元素含量,进而总结归纳不同类型矿石、岩石的高光谱反射率数据特征,通过回归分析建立稀土元素定量反演模型。研究表明,矿石样品在511、523、580、676、740、798、865、890、1093和1252 nm的典型吸收,其相对吸收深度Dep511≥0.002、Dep523≥0.002、Dep580≥0.002、Dep676≥0.002、Dep740≥0.004、Dep798≥0.006、Dep865≥0.004、Dep890≥0.002、Dep1093≥0.003、Dep1252≥0.002等光谱参数,可作为矿化标志的重要信息。矿石样品的高光谱典型吸收位置及相对吸收深度,和15种稀土元素以及总稀土(含Y)含量具有高度的相关性,并以此建立多元线性回归反演模型,回归系数R 2位于0.759~0.987之间,具有较好的预测效果。研究认为,V-NIR-SWIR波段的高光谱技术在快捷、无损、廉价探测碳酸岩型矿床矿石样品稀土含量方面具有广阔的应用前景,能够为该类矿床深边部矿产勘查、矿体圈定和资源预测提供重要支撑。Hyperspectral technology has great appeal due to its environmental-friendly,rapid,and non-destructive determination for material composition and improvement of exploration efficiency in recent years.The qualitative identification of rare earth elements(REE)and their compounds,rare earth minerals,and the quantitative inversion for REE have been systematically studied through hyperspectral technology.However,the hyperspectral quantitative inversion study of rare earth concentrations for carbonatite-related REE deposits is relatively limited.This paper will collect the spectral reflectance data of representative well rocks and ores in a carbonatite-related REE deposit named Dalucao using HyLogger through hyperspectral(V-NIR-SWIR)technology.Meanwhile,the mineral assemblages of rocks and ores were observed through an orthogonal polarizing microscope,and the contents of rare earth elements were measured through the whole rock trace experiment.Then,the characteristics of hyperspectral reflectance for different types of rocks and ores were summarized systematically,and the quantitative inversion for REE was set up through regression analysis in the Dalucao REE deposit.The following conclusions can be obtained.The typical absorptions(in 511,523,580,676,740,798,865,890,1093,and 1252 nm),as well as the parameters of relative absorption depths(Dep511≥0.002,Dep523≥0.002,Dep580≥0.002,Dep676≥0.002,Dep740≥0.004,Dep798≥0.006,Dep865≥0.004,Dep890≥0.002,Dep1093≥0.003,Dep1252≥0.002)etc.for ores are the significant information of mineralization for exploration.These relative absorption depths have strong correlations with the concentrations of REE.The multiple linear regression inversion models have been established between these relative absorption depths and the contents of 15 kinds of REE and total REE,including Y,respectively,along with high accuracy with R 2(0.759~0.987),which demonstrates that they have good predictions for the contents of REE.The results suggest that it has a wide prospect for rapid,non-dest
分 类 号:P627[天文地球—地质矿产勘探]
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...