基于Landsat 8遥感数据的离子吸附型稀土矿定量反演研究

Quantitative Inversion of Ion Adsorption Rare Earth Deposits Based on Remote Sensing Data by Landsat 8

稀土矿作为国家的战略资源,其勘探方法是近几年的研究热点,文章采用研究区的化学勘探数据和Landsat 8遥感影像数据作为研究数据,建立BP神经网络模型反演,分析利用遥感定量反演方法勘探离子吸附型稀土矿的可行性。研究结果表明:研究区的赋矿位置主要集中在全风化层的中下部,配分分析表明研究区的矿石主要属于低钇低铕稀土矿-中钇中铕稀土矿;相关性分析表明样本的稀土矿品位含量与Landsat8影像第1、2、3、5、6、7波段光谱反射率呈显著负相关性,而与第4波段光谱反射率的相关性不显著;BP神经网络模型的检验R2为0.64,可以用来反演研究区的离子吸附型稀土矿品位;植被覆盖程度影响离子吸附型稀土矿遥感定量反演结果的准确性。综上所述,在植被覆盖度低的地区可以采用遥感定量反演方法来勘探离子吸附型稀土矿。

Rare earth as a national strategic resource,the exploration method of which is a hot research topic in recent years.In this paper,the chemical exploration data and remote sensing image data from Landsat 8 of the research area are used as the research data,and BP neural network model inversion is established to analyze the feasibility of using the quantitative inversion method of remote sensing to explore ion adsorption rare earth ore.The results show that the ore hosting location of the study area is mainly concentrated in the middle and lower part of the completely weathered layer.The distribution analysis shows that the ore in the study area mainly belongs to the low yttrium low europium rare earth ore-the middle yttrium middle europium rare earth ore;the correlation analysis shows that the rare earth ore grade content of the sample has a significant negative correlation with the spectral reflectance of bands 1,2,3,5,6 and 7 of Landsat 8 image,but has no significant correlation with the spectral reflectance of band 4;The test R2 of BP neural network model is 0.64,which can be used to invert the grade of ion adsorption rare earth ore in the study area;the accuracy of remote sensing quantitative inversion results of ion adsorption rare earth ore is affected by the degree of vegetation coverage.In conclusion,remote sensing quantitative inversion method can be used to explore ion adsorption rare earth deposits in areas with low vegetation coverage.