Slab Stagnation: How, When, and Where?

Abstract:

Many slabs appear to stagnate in the transition zone although some slabs appear to stagnate at a depth of 1000 km and others appear to descend into the lower mantle relatively unaltered or, perhaps buckling as they descend. Because tomographic images provide a modern day snapshot of a time-dependent process, it is unclear whether the diversity of subducted slab geometries are a manifestation of the same process captured at different times in the lifetime of the subducting system or whether different subduction zones are subject to different conditions that control their evolution. At one time, stagnation of the slab at the base of the transition zone was thought to be due to the transformation of ringwoodite to bridgmanite plus ferropericlase, although subsequent experimental work showed that this transformation does not produce sufficient buoyancy to stall slab descent. In addition to phase transformations in the olivine system, rheology (specifically a ``viscosity hill’’ in the upper part of the lower mantle), trench migration, depth-dependent thermodynamic parameters, and composition have all been investigated as potential slab stagnation mechanisms. The transformation of pyroxene to majoritic garnet occurs by extremely slow diffusion and thus pyroxene unlikely to transform at equilibrium pressures. We have shown that the presence of metastable pyroxene in the cold cores of subducted slabs is sufficient to cause flat-slab subduction. Given the diversity of slab structures, it is quite likely that a combination of mechanisms control slab dynamics. We will investigate slabs stagnation using numerical experiments in 2D and 3D with dislocation/diffusion creep rheology, phase transformations, and plate reconstructions to control the evolution of the plate system.