Search for a command to run...
Maps showing clay-hosted rare earth element (clay-REE) mineral potential in Australia, generated by (1) a knowledge-driven weighted sum model (KDWS) and (2) a machine learning random forest (MLRF) model. <br>The mineral potential maps are a synthesis of four component layers: 1. Source of metals, fluids and ligands (KDWS & RFML - subsets of the Layered Geology of Australia 1:1M and Alkaline Rocks Atlas of Australia; RFML additionally includes subset of Surface Geology of Australia 1:2.5M). 2. Energy sources and fluid flow drivers (KDWS & RFML - Revised Weathering Intensity Model of Australia - values 4.5-5.5; RFML additionally included the full national weathering intensity grid). 3. Fluid flow pathways and architecture (KDWS & RFML - Depth of regolith [regolith thickness] grid of Australia - vales 5-60 m inclusive). 4. Ore depositional gradients and traps (KDWS & RFML - Slope-relief classification of Australia - class 11 (lowest slope and relief landform pattern).<br>The input maps and dataset representing mineral deposits and occurrences used to train and validate (KDWS) and train, test and validate (RFML) the models are provided. For clarity, the small number of documented 'transported' deposits (see accompanying GA Record) are not included in this analysis.<br>These datasets are accompanied by an assessment criteria table which contains information on the mappable criteria applied. Model weightings applied to each mineral system component (KDWS) and random forest statistics (MLRF) are reported in the accompanying GA Record.<br>The development of the maps is documented in GA Record 2026/12 (eCat 150834).