运用光谱测量技术来测量土壤、 矿物、 植物的光谱吸收特征, 并将其用于物质成分的分析中, 是高光谱遥感技术近些年来的一个新的发展方向, 具有速度快、 效率高、 费用低、 损耗少的优点。 稀土是战略元素, 因具有特殊的物理化学性质, 在多个领域都能发挥其独特的作用。 近年来对稀土资源需求量的不断增加使得稀土价值不断提高, 使得如何快速大面积探测稀土资源, 合理对稀土矿进行布设开采成为了当前社会面临解决的重要科学问题。 通过对稀土矿物进行光谱采集与分析从而对稀土元素及其化学特征之间的相关性进行研究分析展开了一系列的工作。 在研究过程中, 实地采集了广西崇左市六汤稀土矿区内的12个稀土矿样, 并在实验室控制环境内使用SVC HR1024I便携式地物波谱仪测得其对应的反射光谱数据。 对所测得的矿样波谱特征进行了连续统去除处理, 并对凸显的特征吸收波段进行相对吸收深度的计算分析, 从而运用其波谱特征建立起波谱与矿样稀土总含量、 各稀土元素含量间的线性关系。 通过实验发现, 稀土元素的五个特征吸收波段在可见光和近红外波长分别为370, 950, 1 400, 1 900和2 200 nm。 这五个吸收带的强度与总稀土含量呈线性相关, R2达到0.69, 同时发现稀土含量值与可见光波段相关性较大, 对可见光波段与样本稀土总含量进行相关分析, 提取出部分可见光波段中与稀土总含量建模相关性最强的10个波段, 分别为340, 350, 360, 370, 390, 400, 420, 480, 550和760 nm。 运用线性回归的方法得到上述波段处的反射率值与样本总稀土含量值的预测模型精度较高, R2大于0.95。 运用可见光波段与15种稀土元素含量值也进行了线性建模, 相关系数均可以达到0.89以上, 表明各单一稀土元素与可见光波段之间的相关性同样较好。 通过研究稀土矿样的光谱特性及化学分析, 通过5个稀土特征吸收波段和可见光波段与矿样的稀土总含量、 15种稀土元素含量进行回归建模, 得到了矿样稀土含量的定量预测模型, 对稀土矿快速定量-半定量评估具有一定参考价值, 同时为进行稀土矿物及其元素高光谱遥感信息提取研究提供了科学启发, 为最终实现对稀土资源高效开采的基础上, 从源头上降低消耗和生产成本, 减少对环境的破坏和污染, 为促进中重稀土资源的战略开发和使用提供了科学有效的理论依据。

Spectral absorption characteristics are used in the analysis of soil, mineral and plant material composition with spectral measurement technology, which is a hyperspectral remote sensing technology developed in recent years. It has many advantages, such as fast speed, high efficiency, low cost, low loss and so on. Rare earth is a strategic element with rich and unique physical and chemical properties, such as magnetic, optical, electrical, etc., which is widely used in aerospace, electronics, petrochemical, metallurgy, machinery, energy, agriculture and other fields. It is indispensable strategic material for the development of high-tech and cutting-edge national defense technology and the transformation of traditional industries in the world today. In recent years, As a result of the increasing demand and the enhancing value of rare earth resources, it has become an important research area to discover how to detect rare earth resources rapidly in large area and implement rare earth mining properly. Through spectral collection and analysis of rare earth minerals, this study carried out a series of researches on the correlation between rare earth elements and their chemical features. During the study, 12 rare earth mineral samples were collected from Liutang rare earth ore area of Chongzuo City, Guangxi, and the corresponding reflectance spectrum data were measured by using SVC HR1024I portable ground object wave spectrometer in laboratory. Continuous dispatch is implemented for the measured spectral features of the samples, and relative absorption analysis is carried out for prominent diagnostic absorption wavelength. Thus the linear relationship between the spectrum and the total content of rare earth elements and the contents of rare earth elements in ore samples was established according to its spectral characteristics. The results reveal that the five characteristic absorption bands of rare earth elements are 370, 950, 1 400, 1 900 and 2 200 nm in visible light and near infrared, respectively. The intensity of the five absorption band is related to the total rare earth content linearly, with R2 reaching 0.69, also discovered that the correlation between rare earth content and the visible light band is larger, and the correlation analysis between the visible light wave band and the total rare earth content of the sample was carried out. The 10 bands which have the strongest correlation with the total content of rare earth are 340, 350, 360, 370, 390, 400, 420, 480, 550 and 760 nm, respectively. The linear regression method is used to get the prediction model of the reflectance value and the total sample content of visible bands with high accuracy with R2 greater than 0.95. Also linear modeling is established by using the visible light wave band and 15 rare earth elements content values, with the correlation coefficient may reach above 0.9, which also shows that each single rare earth element has a strong correlation with the visible light wave region. By studying the spectral characteristics and chemical analysis of rare earth mineral samples, a linear regression analysis was carried out for 5 diagnostic absorption wavelengths and the visible band and the total rare earth element content of the samples, and for 15 kinds of rare earth element contents. The quantitative evaluation model of rare earth content in ore samples is established, which has certain reference value for rapid quantification-semi-quantitative evaluation of rare earth ores, and lays a theoretical foundation for extracting mineral information from hyperspectral remote sensing of rare earth ores and elements. It provides a scientific and effective theoretical basis for the ultimate realization of efficient exploitation of rare earth resources, reducing consumption and production costs at the source, reducing environmental damage and pollution, and promoting the strategic development and utilization of medium-heavy rare earth resources.