Hadean tectonics: Insights from machine learning

The tectonic affiliations and magma compositions that formed Earth's earliest crusts remain hotly debated. Previous efforts toward this goal have relied heavily on determining the provenance of Hadean zircons using low-dimensional discriminant diagrams developed from Phanerozoic samples, which are inadequate for capturing systematic differences without considering secular changes in zircon composition. Here, we high-dimensional machine learning (ML) approaches chemistry data (spanning 19 elements over 4.0 b.y.) to characterize crystallized some typical settings (e.g., arcs, plume-related hotspots, rifts) either igneous (I-type) or sedimentary (S-type) magmas. proposed ML method, a nonuniformitarian perspective, identifies granitoid types given (from Archean Phanerozoic) at higher prediction accuracy >89% compared ~66%–82% traditional U/Yb vs. Y rare earth (REE) + P). ML-based discriminators depend chemistry, notably, significant U, Th, heavy REE settings, P Hf I- S-type Application trained models Jack Hills, Australia, suggests these were mainly continental arc–forming magmas (90%) with 45% belonging melts. This result provides clear evidence sediment recycling associated subduction activity Hadean.