Subducting-slab transition-zone interaction: Stagnation, penetration and mode switches

Roberto Agrusta,Saskia Goes,Jeroen Van Hunen

doi:10.1016/j.epsl.2017.02.005

Roberto Agrusta, Saskia Goes + Show 1 more

Open Access

https://doi.org/10.1016/j.epsl.2017.02.005

Copy DOI

Abstract

Seismic tomography shows that subducting slabs can either sink straight into the lower mantle, or lie down in the mantle transition zone. Moreover, some slabs seem to have changed mode from stagnation to penetration or vice versa. We investigate the dynamic controls on these modes and particularly the transition between them using 2D self-consistent thermo-mechanical subduction models. Our models confirm that the ability of the trench to move is key for slab flattening in the transition zone. Over a wide range of plausible Clapeyron slopes and viscosity jumps at the base of the transition zone, hot young slabs (25 Myr in our models) are most likely to penetrate, while cold old slabs (150 Myr) drive more trench motion and tend to stagnate. Several mechanisms are able to induce penetrating slabs to stagnate: ageing of the subducting plate, decreasing upper plate forcing, and increasing Clapeyron slope (e.g. due to the arrival of a more hydrated slab). Getting stagnating slabs to penetrate is more difficult. It can be accomplished by an instantaneous change in the forcing of the upper plate from free to motionless, or a sudden decrease in the Clapeyron slope. A rapid change in plate age at the trench from old to young cannot easily induce penetration. On Earth, ageing of the subducting plate (with accompanying upper plate rifting) may be the most common mechanism for causing slab stagnation, while strong changes in upper plate forcing appear required for triggering slab penetration.

Highlights

Subducted slabs in today’s mantle display a wide variation in morphology, where some slabs seem to stagnate in the transition zone, e.g., beneath the Izu-Bonin region, South-Kurile and Japan, whereas others seem to penetrate into the lower mantle, for example beneath Peru, the Marianas and Central America
It has been shown that slab stagnation is often accompanied by trench retreat and that high trench migration rates promote slab flattening in the transition zone (Christensen, 1996; Cížková et al, 2002; Griffiths et al, 1995), and low dip angles hamper penetration (Tagawa et al, 2007; Torii and Yoshioka, 2007)
In this study, using a series of 2D self-consistent thermomechanical subduction models, we investigate which parameters are most efficient in allowing slabs to penetrate into or stagnate above the lower mantle and if there are plausible mechanisms to change the slab-transition zone interaction from one mode to the another

Summary

Introduction

Subducted slabs in today’s mantle display a wide variation in morphology, where some slabs seem to stagnate in the transition zone, e.g., beneath the Izu-Bonin region, South-Kurile and Japan, whereas others seem to penetrate into the lower mantle, for example beneath Peru, the Marianas and Central America (van der Hilst et al, 1997; Fukao and Obayashi, 2013). Comparisons of plate reconstructions with tomography indicate that most slabs that subducted in the last ∼200 Myr are in the lower mantle (e.g., Ricard et al, 1993; van der Meer et al, 2010) This suggests that to avoid longterm accumulation of slabs in the transition zone, slabs that previously stagnated, as many slabs do today, must have changed from stagnation to penetration. Most models investigated these mechanisms in isolation, and prescribed part of the system, e.g. trench position, trench motion, slab dip In this way, it has been shown that slab stagnation is often accompanied by trench retreat (van der Hilst and Seno, 1993) and that high trench migration rates promote slab flattening in the transition zone (Christensen, 1996; Cížková et al, 2002; Griffiths et al, 1995), and low dip angles hamper penetration (Tagawa et al, 2007; Torii and Yoshioka, 2007).

Objectives

Methods

Discussion

Conclusion