Dynamical modeling of trench retreat driven by the slab interaction with the mantle transition zone

Abstract

We present the 2-D self-consistent dynamical model of interactions of a subducting slab with the 410-km and 660-km phase boundaries to further our understanding of the relation between the slab stagnation/penetration and the trench migration. Our model takes into account freely-movable plate boundaries and the difference between tensional and compressional yield strengths in the lithosphere. For the case in which the tensional strength is weaker than the compressional one, the negative buoyancy of the subducting slab produces extension of the overriding lithosphere and, accordingly, the trench retreats. Interactions with the 410-km and 660-km phasetransition boundaries further promote the trench retreat, and the dip angle of the slab is substantially decreased. This enhances the resistance of the 660-km phase boundary against the slab penetration. Slab weakening caused by the grain-size reduction in the transition zone may result in a horizontally-lying slab and trench retreat.

Subduction-transition zone interaction: A review

“…Dynamic subduction models show that subduction accompanied by trench retreat is the most basic form of subduction (Kincaid and Olson, 1987). Trench motion is controlled by downgoing plate buoyancy and strength (Bellahsen et al, 2005;Capitanio et al, 2007;Tagawa et al, 2007;Ribe, 2010) and by upper-plate deformability and mobility (Garel et al, 2014;Holt et al, 2015). As a result, old strong slabs have a stronger tendency to retreat than young weak slabs (Agrusta et al, 2017).…”

“…King and Ita (1995) also found that slab strength does not exert a major influence on whether slabs penetrate or not, when the trench is free to move. Others further tested the effect of local slab weakening in the transition zone and confirmed that the effect is not dominant, but it may be sufficient to stagnate a slab that is otherwise marginally penetrating (Tagawa et al, 2007;Agrusta et al, 2017). The buckling of weaker slabs can actually increase the slab's Stokes' sinking velocity (which scales with diameter squared) and hence help it penetrate a high-viscosity lower mantle rather than stagnating it.…”

“…On the other hand, if there is no viscosity jump, it is difficult to develop any flat slabs (Tagawa et al, 2007;Torii and Yoshioka, 2007;Yanagisawa et al, 2010; and our own models for Fig. 6).…”

Numerical model for the generation of the ensemble of lithospheric plates and their penetration through the 660-km boundary

“…Most studies only analyze a brief evolution of the con vection in a local area of the mantle with one or two plates. These plates are either modeled by the bound ary conditions or are specified in the initial state of the model together with their sizes, locations, and initial velocities, e.g., (Kneller et al, 2005;Tagawa, 2007;Kaus and Becker, 2008;Bonnardot et al, 2008;Andrews and Billen, 2009). A significant advantage of these models is their suitability for exploring the mechanism of subduction in fine detail.…”

Mid-crustal horizontal shear zone in the forearc region of the mid-Cretaceous SW Japan arc, inferred from strain analysis of rocks within the Ryoke metamorphic belt

“…Based on the observations and the results of dynamic model simulations, Goes et al (2008) indicated that the younger lithosphere drives less trench retreat than old lithosphere; zones subducting young lithosphere seem unable to initiate extension in the forearc region, which hampers trench retreat. However, the slab interactions with the mantle transition zone promote the trench retreat, and the dip angle of the slab is substantially decreased (Tagawa et al, 2007). Although the relationship between the forearc stress regime and the trench migration is not clear yet, the subduction of a spreading ridge or very young oceanic lithosphere would affect the stress configuration in the forearc region.…”