Review of currently available multispectral and hyperspectral imaging spectroscopy satellite sensors for critical minerals exploration

<div>Earth observation is a fast and cost-effective method for greenfields exploration of critical minerals at a continental to regional scale. A broad range of optical satellite sensors are now available to mineral explorers for collecting Earth observation information (EOI) at various spatial and spectral resolutions, with different capabilities for direct identification of mineral groups and/or species as well as selected chemical elements. The spectral resolution of many of the latest imaging spectroscopy satellite systems (e.g., PRISMA - https://www.asi.it/en/earth-science/prisma/; EnMap - https://www.enmap.org/; EMIT - https://earth.jpl.nasa.gov/emit/) allow the mapping of the relative mineral abundance and, in selected cases, even the chemical composition of hydrothermal alteration minerals and pegmatite indicator minerals, such as white mica, chlorite and tourmaline. More specialised hyperspectral satellite systems, such as DESIS (https://www.dlr.de/eoc/en/desktopdefault.aspx/tabid-13614/) feature a very high spectral resolution (235 bands at 2.55 nm sampling and 3.5 nm full width half maximum) across parts of the Visible to Near-Infrared (VNIR) wavelength range, opening up the possibility for direct mapping of rare earth elements, such as neodymium. The pixel size of the imaging spectroscopy satellite systems is commonly 30 m, which can be sufficient to map hydrothermal footprints of ore deposits or surface expressions of typical rare element host rocks, such as pegmatites and carbonatites. However, airborne hyperspectral surveys still provide a higher spatial resolution, which can be essential in a given mineral exploration campaign. Selected multispectral satellite systems, such as ASTER (https://terra.nasa.gov/data/aster-data) and WorldView3 (https://resources.maxar.com/data-sheets/worldview-3) do have bands at important wavelength ranges in the shortwave infrared, but not with high enough spectral resolution&nbsp;to clearly identify many indicator minerals for critical minerals deposits. Most publicly available satellite imagery comprises multispectral systems that are focussed on the VNIR, such as Landsat and Sentinel, but which allow the direct identification of only very few mineral groups (mainly iron oxides) and not hydroxylated vector minerals (e.g., white mica, chlorite, tourmaline). This work aims to provide a summary of currently available optical satellite sensors and high-level comparison of their applications for critical minerals exploration. In addition to the spatial and spectral resolution, the impact of, for example, signal-to-noise ratio, striping and band width on accurate mineral and element mapping is discussed. For this, case studies are presented that demonstrate the potential use of the respective sensors for different stages of an exploration campaign and also the opportunities for integration with other geoscience data across scales. This abstract was presented to the 13th IEEE GRSS Workshop on Hyperspectral Image and Signal Processing (WHISPERS) November 2023 (https://www.ieee-whispers.com/)