端元
沉积岩
地质学
矿物
矿物学
磷酸盐矿物
高光谱成像
地球化学
磷酸盐
比例(比率)
矿产勘查
采样(信号处理)
粘土矿物
地质图
砂矿开采
空间变异性
方解石
自生的
演习
变质岩
碳酸盐矿物
作者
Houda Lkhaoua,Otmane Raji,Abdellatif Elghali,Mahdi Khodadadzadeh,Radouan El Bamiki,Mustapha El Ghorfi,Mostafa Benzaazoua
标识
DOI:10.1016/j.rineng.2026.113525
摘要
Sedimentary phosphate deposits are characterized by complex mineral assemblages dominated by carbonate-fluorapatite (CFA), carbonates, and siliceous phases, whose spatial variability strongly controls ore quality and processing behavior. In this study, we investigate the potential of hyperspectral infrared imaging in the mid-wavelength infrared (MWIR) range for the identification and quantitative mapping of the main mineral constituents of the sedimentary phosphate sequence. Representative endmember spectra were extracted from selected samples and used to perform mineral mapping through linear spectral unmixing solved by Non-Negative Least Squares (NNLS). The NNLS approach was validated against SEM-based quantitative mineralogy at the sample scale and p-XRF geochemical profiles at the log scale. The resulting continuous mineralogical profiles demonstrate a strong consistency with reference lithostratigraphic logs and geochemical validation data, successfully capturing high-frequency variations that conventional discrete sampling often misses. Overall, the results provide a reliable and efficient framework for rapid, continuous mineral mapping of sedimentary phosphate deposits, with direct implications for exploration, orebody characterization, and geominerallurgical modeling.
科研通智能强力驱动
Strongly Powered by AbleSci AI