Modeling the transport and evolution of sediment diapirs in subduction systems

Abstract:

Sediment subduction is the dominant process by which material from the Earth’s surface is returned to the mantle, and thus plays an important role in trace element and volatile recycling, and the long-term evolution of continental crust and mantle heterogeneities. However, the fate of subducted sediments remains uncertain, as sediments may be subducted beyond the subduction system or returned to the crust either as melts or as relaminated material. Estimates of trace element and isotopic budgets produce a wide range of recycling rates in arc magmas from <25% to ~100%. The high estimated recycling efficiencies likely require that sediment melts are produced at temperatures well in excess of expected slab-top temperatures, and suggest that sediment diapirs are critical to generating sediment signatures in arc magmas.

To explore the evolution of sediment diapirs, we develop a coupled thermodynamic-geodynamic model of sediment diapirs. Specifically, we calculate the equilibrium melt and residual mineral assemblage and buoyancy of diapirs as they advect through and thermally equilibrate with the mantle wedge. Using this model, we evaluate the conditions under which diapir formation and ascent is favored, and where melting of these diapirs occurs. We find that diapirs exhibit two primary behaviors: 1) decoupling from the slab and rising into the hotter core of the mantle wedge where high degree melting occurs; or 2) remaining coupled to the near-slab top region and either not melting or undergoing only lower temperature melting. The first mode is favored in hotter subduction zones, for more felsic diapir compositions, and for larger diapirs. Additionally, efficient melt extraction decreases diapir buoyancy, and can result in negatively buoyant diapir residues that sink back into the mantle. However, we find that for a range of intermediate to felsic compositions, diapirs remain buoyant even following initial melt extraction, suggesting that relamination of diapirs can occur even after melting. Finally, consistent with experiments, we predict that melting of sediment diapirs produces granitic melts that are increasingly potassic as melting pressure increases. Colder subduction zones and smaller radius diapirs suppress initial diapir-slab decoupling and will favor the production of these high-K₂O melts.