Dynamics of free subduction from 3‐D boundary element modeling

Abstract

[1] In order better to understand the physical mechanisms underlying free subduction, we perform three-dimensional boundary-element numerical simulations of a dense fluid sheet with thickness h and viscosity h 2 sinking in an 'ambient mantle' with viscosity h 1 . The mantle layer is bounded above by a traction-free surface, and is either (1) infinitely deep or (2) underlain by a rigid boundary at a finite depth H + d, similar to the typical geometry used in laboratory experiments. Instantaneous solutions in configuration (1) show that the sheet's dimensionless 'stiffness' S determines whether the slab's sinking speed is controlled by the viscosity of the ambient mantle (S < 1) or the viscosity of the sheet itself (S > 10). Time-dependent solutions with tracers in configuration (2) demonstrate a partial return flow around the leading edge of a retreating slab and return flow around its sides. The extra 'edge drag' exerted by the flow around the sides causes transverse deformation of the slab, and makes the sinking speed of a 3-D slab up to 40% less than that of a 2-D slab. A systematic investigation of the slab's interaction with the bottom boundary as a function of h 2 /h 1 and H/h delineates a rich regime diagram of different subduction modes (trench retreating, slab folding, trench advancing) and reveals a new 'advancing-folding' mode in which slab folding is preceded by advancing trench motion. The solutions demonstrate that mode selection is controlled by the dip of the leading edge of the slab at the time when it first encounters the bottom boundary.

How weak is the subduction zone interface?

“…Previous work has argued that the upper mantle is dominantly non-Newtonian and strain rate softening [Billen, 2008;Cizkova and Bina, 2013], which implies that the effective viscosity ratio between a slab and the ambient upper mantle should be relatively high. Nevertheless, the effective viscosity ratio that we use in our experiments (~160) is well within the previous estimates of approximately 50-500 [e.g., Schellart, 2008;Funiciello et al, 2008;Ribe, 2010;Wu et al, 2008;Loiselet et al, 2009;Li and Ribe, 2012]. For the density difference we consider a natural upper limit of Δρ = 80 kg m 3 , which would correspond to an old lithospheric slab in which all the oceanic crust has been altered to eclogite and with no heat diffusion into the slab.…”

Parameters controlling dynamically self‐consistent plate tectonics and single‐sided subduction in global models of mantle convection

“…A first major assumption is (i) the absence of important 3‐D mantle flow. Models run in 2‐D space effectively represent infinitely wide subduction zones and neglect the important effect of variable trench width [ Stegman et al , ], possible 3‐D influences on slab‐sinking resistance [ Li and Ribe , ] or overriding plate deformation [ Schellart and Moresi , ], and any toroidal motion like trench‐parallel flow beneath the slab [e.g., Buttles and Olson , ; Kincaid and Griffiths , ] or the return flow around slab edges [ Funiciello et al , ] caused by slab retreat. Additionally, there are some important aspects due to the sphericity of a planet that might not be captured by experiments in Cartesian model domains [e.g., O'Farrell and Lowman , ].…”

A review of analogue modelling of geodynamic processes: Approaches, scaling, materials and quantification, with an application to subduction experiments

“…Using different lines of evidence, a variety of works imply that in nature the effective viscosity ratio Á SP /Á SLUM is in the range 50-500 (e.g. Moresi and Gurnis, 1996;Billen et al, 2003;Schellart, 2008;Funiciello et al, 2008;Loiselet et al, 2009;Ribe, 2010;Li and Ribe, 2012).…”