GRSG 35th Conference 2024 Abstract

Title: Multiscale hyperspectral imaging of ree-bearing minerals at Mountain Pass, California

Author: Todd Hoefen

Organisation: USGS

Imaging spectroscopy, or hyperspectral imaging, has primarily been used to mapsurface materials over relatively small areas from airborne sensors for more than 20years. As part of the U.S. Geological Survey’s Hyperspectral Imaging of MineralResources from New and Old Origins: Minerals for the Nation’s Economy andUtilization of Legacy Mine Lands project, we collected multiscale imagingspectrometer data including laboratory, field, and airborne data from the MountainPass carbonatite rare earth element (REE) mine.

This data set offers a uniqueopportunity to map surface mineralogy at scales ranging from microns to metersper pixel. In this study, we utilized several imaging spectrometers, including theAirborne Visible and Infrared Imaging Spectrometer (AVIRIS), SpecTIR AisaFENIX 1k,and HySpex VNIR-1800 and SWIR-384. The resulting maps, created using spectralanalysis of multiscale hyperspectral data, revealed mineral assemblages typical ofcarbonatite deposits. Key REE-bearing minerals identified, such as bastnaesite,hydroxylbastnaesite, and monazite, are distinguishable by their characteristic 4f-4ftransitions.

Spectroscopic analysis of REE minerals has enriched our understandingof their 4f-4f transition features across the ultraviolet to shortwave infrared regionsof the electromagnetic spectrum. Subtle variations in wavelength absorption bandcenters and spectral shape allow for precise identification of REE-bearing mineralsbased solely on these transitions. By applying specific user-defined criteria throughthe MICA module of the U.S. Geological Survey (USGS) PRISM software, wedeveloped a method for mineral identification at Mountain Pass based on trivalentlanthanide absorptions. Mapping results, instrument specifications, data collectionparameters, and the software used for data processing and analysis will bepresented.