Spaceborne mineral maps for critical metals exploration in the Gascoyne and Pilbara
New geoscience-tuned hyperspectral satellite sensors now enable the characterization of minerals on the Earth’s surface at an unprecedented level of detail, representing a step change from multispectral sensors. Optical Earth observation imagery, such as that produced by the Italian Space Agency’s (ASI) PRecursore IperSpettrale della Missione Applicativa (PRISMA), and the German Aerospace Center’s (DLR) Environmental Mapping and Analysis Program (EnMAP), complement the already successfully applied, “geoscience-tuned” multispectral sensors (e.g. ASTER and WorldView3) in geological remote sensing.
In the frame of a collaborative research project between the CSIRO and the Geological Survey of Western Australia (GSWA), hyperspectral satellite imagery was acquired from the Western Australian Gascoyne and Pilbara provinces and processed into a suite of mineral abundance, mineral composition and crystallinity mosaics, as well as green and dry vegetation layers. In order to evaluate the geological and mineralogical information provided by the satellite imagery, the mineral maps were compared with 1) GSWA’s 500K and 100K geological and regolith maps, 2) publicly available geophysical layers (i.e. radiometrics and magnetics) and 3) previously published mineral maps, processed from spaceborne multispectral (i.e. ASTER) and airborne hyperspectral (i.e. HyMap) VNIR-SWIR imagery. This work was complemented by a ground validation campaign in the Gascoyne, during which VNIR-SWIR and TIR field spectrometer data were acquired along transects and samples collected for further analysis.
The publicly available report (Laukamp et al., 2025) describes variations in the PRISMA-derived mineral map mosaics evident in Paleoproterozoic bedrocks of the tectonostratigraphic zones of the Gascoyne Province. For example, a strong ESE-trending structural grain in the Limejuice Zone is highlighted by the white mica abundance and composition maps (Fig. 1), especially towards the Ti Tree Shear Zone and southern border with the Mutherbukin Zone. The combined chlorite-biotite-epidote abundance index maps out mafic intrusive units as well as the Narimbunna Dolerites, but also points to previously unmapped occurrences of mafic intrusives in granites of the Limejuice Zone. Within the Gifford Creek Carbonatite Complex, the mineral maps suggest mineral patterns that may help to revise the current geological mapping of this important REE province. Furthermore, kaolin crystallinity index and ferric oxide abundance index maps show patterns that overlap with already mapped regolith landforms, but also highlight locations of, for example, previously unmapped mesas. Further ground validation is recommended in the Gifford Creek Carbonatite complex and other areas to assess the opportunity for revising existing geological and regolith landform mapping.
1a) PRISMA-derived white mica abundance index (WMai), calculated using a polynomial fit method and visualised using a linear stretch (ENVI colour scheme: “rainbow”, Min/Max: 0.02/0.18). d) PRISMA-derived white mica composition index (WMci), calculated using a polynomial fit method and visualised using a linear stretch (ENVI colour scheme: “rainbow”, Min/Max: 2185 nm/2215 nm). Thick magenta rectangle – Gascoyne ROI. Thick black rectangle – larger case study areas to encompass neighbouring hyperspectral satellite imagery.
In the Pilbara region, the PRISMA-derived mineral map mosaics successfully discriminate major geological features of the Pilbara Craton, including Archean granitoids and greenstone belts, at the regional-scale to tenement-scale. Furthermore, various types of hydrothermal alteration patterns were mapped, including hydrothermal quartz veins, white mica compositional variations (Fig. 2), and advanced argillic alteration diagnostic of highly-acidic epithermal environments. Furthermore, PRISMA and EnMAP satellite imagery both identified known Li-bearing pegmatite occurrences and new pegmatite targets. Mafic and ultramafic host rocks of the pegmatites each show significantly different mineralogical signatures in the satellite imagery. Further ground validation is planned, for example, 1) across newly identified pegmatites and ultramafic sills in the Tambourah area and 2) across hydrothermal quartz veins and potential metasomatic silica alteration within granites of the Split Rock Supersuite.
Fig. 2: Panorama area: a) 100K bedrock geology; b) Al-sheetsilicate abundance index (ASai ; stretch: colour scheme: “spectrum-full bright”, Min/Max: 0.04/0.16); c) White mica composition index (WMci; rainbow colour stretch: warm colours = Al-poor white mica, cool colours = Al-rich white mica, dark blue = background). HyMap coverage (Van Ruitenbeek et al., 2012) shown by white hashed line in c); d) Pyrophyllite ± kaolinite abundance index (rainbow colour stretch: warm colours = high values, cool colours = low values); e) Example PRISMA reflectance spectra showing white mica, kaolinite and pyrophyllite; f) Pyrophyllite SWIR reflectance spectrum of field sample in purple (location: purple arrow in d) and spectral reference library spectrum of pyrophyllite (black).
The comparison between the PRISMA-derived mineral map mosaics with EnMap imagery at selected sites showed that the signal-to-noise ratio of PRISMA is sufficient across key wavelength ranges required for characterising Al-sheetsilicates as well as chlorites (i.e. between 2100 and 2300 nm). However, EnMap hyperspectral imagery has a significantly higher SNR across the VNIR-SWIR, allowing, in addition to the characterisation of Al-sheetsilicates and chlorite, also the mapping of amphiboles/talc as well as carbonates. Furthermore, EnMap imagery can be used to generate superior iron oxide abundance maps. This evaluation of hyperspectral mineral maps from the regional to tenement scale demonstrates that the hyperspectral resolution of the new optical sensors is a step change when compared to the historically used multispectral imagery, such as provided by ASTER and WorldView3. The hyperspectral satellite sensors enable the characterisation of regolith and mapping of outcropping/subcropping bedrock lithologies as well as hydrothermal and supergene mineral alteration patterns potentially associated with mineral deposits approaching the detail that is known from airborne hyperspectral imagery. In addition to PRISMA, EnMap and EMIT, many hyperspectral satellites have been launched since 2019 (e.g. DESIS, DLR; Tanager, Planet Labs), which can be used for cost-effective and sustainable mineral exploration, but also for monitoring the sustainable use of terrestrial ecosystems, areas exposed to desertification and land degradation in support of many of the UN’s 17 Sustainable Development Goals (SDGs).
References
Laukamp, C., Miles, J., Lampinen, H., Williams, M., Clarke, M., Lau, I., Caccetta, M. (2025): Spaceborne mineral maps for critical metals exploration in the Gascoyne and Pilbara.- CSIRO report EP2025-1592.
Van Ruitenbeek, F.J.A., Cudahy, T.J., Van Der Meer, F.D., Hale, M. (2012): Characterization of the hydrothermal systems associated with Archean VMS-mineralization at Panorama, Western Australia, using hyperspectral, geochemical and geothermometric data. Ore Geology Reviews 45, 33–46.