Craton boundary detection from full-waveform tomography model reveals links to critical metal deposits
Year: 2025
Type: article
Abstract: • Craton boundaries act as conduits for metal-bearing hydrothermal fluids. • Unsupervised machine learning enables the detection of mineral-rich lithospheric zones. • Seismic data clustering reveals key geological boundaries and craton margins. • Decision-tree framework distinguished cratons using seismic and other geophysical data. • Eighty percent of Cu, Pb, and Zn in target deposits lie within 90 km of craton boundaries. Craton margins play a crucial role in mineral exploration as they host faults, fractures, and shear zones that facilitate hydrothermal fluid movement, transporting and depositing dissolved metals into valuable mineral deposits. We use the high-resolution full-waveform seismic inversion model REVEAL to extract horizontal shear wave velocity ( V SH ), vertical shear wave velocity ( V SV ), and isotropic P-wave velocity ( V P ) across depth slices from 150 to 200 km, a range that captures most cratonic lithosphere based on tectonic age and lithospheric thickness analyses. Machine learning, applied through clustered maps, demonstrates that V SH effectively delineates craton boundaries, aligning with target mineral deposits, including iron oxide copper–gold (IOCG) and sediment-hosted lead, zinc, and copper deposits. These boundaries are characterized by high horizontal shear velocities (4.58–4.68 km/s), and trace the edges of cratons, accreted passive margins, orogens, and thick volcanic arcs. Using published thermal and lithospheric thickness models, we distinguish cratons from other thick lithospheric features and identify their edges and associated deposits. Our results show that ∼85 % of the total metal content (Cu + Pb + Zn) in target deposits lies within ∼120 km of high-velocity cluster boundaries identified as craton edges. Near-craton deposits reveal ∼80 % of the total metal content within ∼90 km of craton boundaries. The weighted cumulative distribution function shows a steeper gradient in metal content closer to craton boundaries, indicating higher concentrations near these tectonic features. Focusing on just 16 % of Earth’s continental areas can reveal over 80 % of known target deposits, highlighting the significance of craton boundaries quantitatively mapped in this study.
Source: Geoscience Frontiers
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,000
Funder: Australian Research Council
OpenAlex ID: W4414804179