世界核地质科学

高光谱遥感技术以其高光谱分辨率在矿产勘查领域展现出了重要的潜力，但在高光谱遥感蚀变矿物提取技术领域，由于地层原生的岩性矿物与热液蚀变矿物很难区分，容易对蚀变矿物信息提取形成干扰，且很难通过遥感技术手段解决。针对蚀变提取中非铀成矿地层带来的干扰问题，以青海海德乌拉地区为研究区，使用资源一号02D高光谱数据作为数据源对高光谱蚀变矿物提取方法进行了改进。首先对研究区的含铀地层进行分析，明确蚀变矿物提取的范围，采用掩膜方法去除非有利成矿地层干扰，在此基础上使用资源一号02D高光谱数据，采用经典的高光谱沙漏方法完成了目标地层的蚀变矿物提取。从提取结果来看，改进后的方法在研究区所提取的蚀变分布更有规律性，且与铀矿化分布关系密切，通过野外查证也证实了改进方法的有效性。由此可见，结合地质矿产先验知识，针对目标地层的精细化高光谱遥感蚀变矿物提取，可大幅降低无关地层信号干扰，改善蚀变矿物提取可靠性；蚀变信息空间分布与海德乌拉地区铀矿化发育范围高度耦合，也体现了国产资源一号02D高光谱数据在铀矿勘查工作中良好的应用前景。

Hyperspectral remote sensing technology has demonstrated significant potential in the field of mineral exploration due to its high spectral resolution. However， in the procedures of hyperspectral remote sensing-based alteration mineral extraction， it is often difficult to distinguish the minerals of hydrothermal alteration genesis form primary diagemesis of stratigraphic rocks ， which tend to interfere with the extraction of alteration information and hard to resolve solely through remote sensing means. To address the interference caused by non-uranium-mineralized strata in alteration extraction， this study focuses on the improvement of hyperspectral alteration mineral extraction method using ZY-1 02D hyperspectral data of Haidewula area in Qinghai Province. First， the uranium-bearing strata in the study area were analyzed to define the scope of alteration mineral extraction， and a masking method was employed to remove the interference from non-favorable mineralization strata. On this basis， the ZY-1 02D hyperspectral data were used， and the classical hyperspectral hourglass workflow was applied to complete alteration mineral extraction for the target strata. The extraction results show that the improved method yields more geologically reasonable alteration distributions in the study area， which are closely related to the distribution of uranium mineralization. Field verification also confirmed the effectiveness of the improved method. Therefore， incorporating prior geological and mineralogical knowledge in the refined hyperspectral remote sensing alteration mineral extraction of targeting specific strata can substantially reduce signal interference from irrelevant strata and improve the reliability of alteration mineral extraction. The highly spatial distribution consistance of remote sensing extracted alteration information with the extent of uranium mineralization development in the Haidewula area further demonstrated the application potential of the domestically produced ZY-1 02D hyperspectral data in uranium exploration.