Mantle flow around subduction zones: evolution through time

Abstract

When cold and dense oceanic lithosphere sinks into the mantle at subduction zones, it pushes weaker asthenospheric mantle away, creating specific flow patterns. Traditionally mantle flow is divided into two components: trench-perpendicular poloidal flow operating in a vertical plane and 3D toroidal flow around the slab edges. In the past years, we have learned that both poloidal and toroidal mantle flow around slabs effectively connects nearby subduction zones, deforming their slabs and upper plates, and modifying their patterns of volcanism and uplift/subsidence. In turn, the two-way dynamic interaction between the subduction zones also affects the flow pattern, and thus impacts the volcanism, surface uplift and lithospheric deformation (Király et al., 2021).At present, our best constraints on mantle flow patterns around subduction zones originate from seismic anisotropy observations, which can be interpreted based on 3D geodynamic models. In the mantle, seismic anisotropy originates from crystallographic preferred orientation (CPO), which derives from the anisotropic nature of olivine crystals. Due to olivine’s orthorhombic symmetry and the different strengths of its three slip systems, olivine crystals are anisotropic in their elastic and viscous properties. Hence, when many olivine crystals are aligned within mantle rock (i.e., CPO is developed in an area of the mantle), the mantle will deform anisotropically, both for seismic wave transmission and viscous flow. Since CPO occurs as a response to deformation, seismic anisotropy directions are often read as the recent mantle flow direction in an area. However, there are a few complications to this simple one-to-one interpretation. First, because the CPO depends on the deformation history of the mantle, it might not reflect the current flow orientation if the deformation direction has changed through time (Ribe, 1989). Second, CPO formation depends on stress and on water content (Korenaga and Karato, 2008), which in some cases allows texture to form with a fast axis perpendicular to the deformation direction. Third, the interpretation of seismic anomalies is often difficult because geodynamic models do not incorporate enough complexity to model all the intricacies of the flow. This problem can occur due to complex anisotropic signals from crustal layers, from more complicated geodynamic settings (e.g., multiple slabs), or from a modified flow pattern that arises due to the viscous anisotropy associated with the texture itself.In this presentation, I will use the Mediterranean area to highlight how including multiple slabs and accounting for viscous anisotropy can eventually help us to interpret the seismic observations from this geodynamically complex region.