GRSG 34th Conference 2023

Title: Bare-earth satellite imagery and the search for hidden lithium-rich brines: An example from the Lithium Triangle in South America

Author: Livi Rhind

Abstract:

The salars of the Lithium Triangle in South America contain approximately 55% of the world’s lithium resources. The source, transportation and concentration of the lithium-brines involves a complex mineral system that is dependent upon climate, weathering, basin closure, surface drainage, hydrothermal and groundwater systems, lithium-bearing rock distribution and geological structural control. A suite of satellite platforms, geological datasets, advanced data analytics and automated workflows has enabled the delineation and targeting of potential lithium-rich salars and paleosalars buried beneath recent sediments and volcanics. A ‘Source-to-Sink’ model for the mobilisation and deposition of the lithium-rich brines is presented. The lithium is sourced primarily in volcanic ignimbrites and these can be spatially mapped.

The drainage derived from the elevation data was analysed to highlight the closed (endorheic) basin required for lithium-rich brine concentration, with ‘flat-areas’ and slope models being derived to outline potential paleosalar regions that may exist under recent sediments away from the known salars. Analysis of the drainage inlet and outlet points with reference to the underlying ranked lithium-source rocks enables flow-path inlets to be ranked by their lithium potential. The remote sensing component includes a multi-sensor approach, with several satellite platforms utilised in the geological and drainage analysis of the region. Sentinel-2, Copernicus DEM and ASTER bare earth mosaics with derived outputs including vegetation maps and mineral ratios were used in this study.

The ASTER bare earth mosaic over South America was produced from over 36,000 individual scenes, with a proprietary artificial intelligence model applied to ensure no scenes are discounted due to cloud cover. This maximised the number of pixels analysed to increase the identification of ‘bare’ pixels. For the bare earth output, weights were applied to achieve a desired effect, enhancing pixel values on scenes where the vegetation and snow are at a minimum leading to the desired ‘bare earth’ output. The production of a continuous 1:500,000 scale geological map over the Lithium Triangle involved the interpretation of a Sentinel-2 bare earth mosaic and Copernicus DEM topographic model.

The geological interpretation was assisted by published geological survey cartographic products, of variable scales and vintages. The ASTER bare earth mosaic was also used to outline areas of hydrothermal alteration that are related to subsurface hydrothermal fluid flow and that may be associated with the mobilisation of lithium from ignimbrites and other lithium-rich units such felsic basement into the groundwater system. A detailed structural map highlights areas of fluid localisation and possible transportation along faults. Combining the data sets in an advanced data algorithm has enabled a heat map to be derived which delineates areas of lithium-rich brine potential. This has been produced for the entire Lithium Triangle and shows lithium-brine potential for new areas of exploration interest both within known salars and within recent sediment covered paleosalars.

The salars of the Lithium Triangle in South America contain approximately 55% of the world’s lithium resources. The source, transportation and concentration of the lithium-brines involves a complex mineral system that is dependent upon climate, weathering, basin closure, surface drainage, hydrothermal and groundwater systems, lithium-bearing rock distribution and geological structural control. A suite of satellite platforms, geological datasets, advanced data analytics and automated workflows has enabled the delineation and targeting of potential lithium-rich salars and paleosalars buried beneath recent sediments and volcanics. A ‘Source-to-Sink’ model for the mobilisation and deposition of the lithium-rich brines is presented. The lithium is sourced primarily in volcanic ignimbrites and these can be spatially mapped.

The drainage derived from the elevation data was analysed to highlight the closed (endorheic) basin required for lithium-rich brine concentration, with ‘flat-areas’ and slope models being derived to outline potential paleosalar regions that may exist under recent sediments away from the known salars. Analysis of the drainage inlet and outlet points with reference to the underlying ranked lithium-source rocks enables flow-path inlets to be ranked by their lithium potential. The remote sensing component includes a multi-sensor approach, with several satellite platforms utilised in the geological and drainage analysis of the region. Sentinel-2, Copernicus DEM and ASTER bare earth mosaics with derived outputs including vegetation maps and mineral ratios were used in this study. The ASTER bare earth mosaic over South America was produced from over 36,000 individual scenes, with a proprietary artificial intelligence model applied to ensure no scenes are discounted due to cloud cover.

This maximised the number of pixels analysed to increase the identification of ‘bare’ pixels. For the bare earth output, weights were applied to achieve a desired effect, enhancing pixel values on scenes where the vegetation and snow are at a minimum leading to the desired ‘bare earth’ output. The production of a continuous 1:500,000 scale geological map over the Lithium Triangle involved the interpretation of a Sentinel-2 bare earth mosaic and Copernicus DEM topographic model. The geological interpretation was assisted by published geological survey cartographic products, of variable scales and vintages. The ASTER bare earth mosaic was also used to outline areas of hydrothermal alteration that are related to subsurface hydrothermal fluid flow and that may be associated with the mobilisation of lithium from ignimbrites and other lithium-rich units such felsic basement into the groundwater system. A detailed structural map highlights areas of fluid localisation and possible transportation along faults.

Combining the data sets in an advanced data algorithm has enabled a heat map to be derived which delineates areas of lithium-rich brine potential. This has been produced for the entire Lithium Triangle and shows lithium-brine potential for new areas of exploration interest both within known salars and within recent sediment covered paleosalars.