Plume‐Slab Interactions Can Shut Off Subduction

Abstract

Mantle plumes are typically considered secondary features of mantle convection, yet their surface effects over Earth's evolution may have been significant. We use 2‐D convection models to show that mantle plumes can in fact cause the termination of a subduction zone. This extreme case of plume‐slab interaction is found when the slab is readily weakened, for example, by damage‐type rheology, and the subducting slab is young. We posit that this mechanism may be relevant, particularly for the early Earth, and a subdued version of these plume‐slab interactions may remain relevant for modern subduction zones. Such core‐mantle boundary–surface interactions may be behind some of the complexity of tomographically imaged mantle structures, for example, in South America. More generally, plume “talk back” to subduction zones may make plate tectonics more episodic.

Plume‐Driven Subduction Termination in 3‐D Mantle Convection Models

“…As may be expected, and explored more fully in 2‐D (Heilman & Becker, 2022), not every plume‐slab interaction ends in subduction termination. We find at least five instances where a plume interacts with a subducting slab without causing a complete termination, that is, a roughly 60% chance of plumes shutting down subduction if they get close to slabs, for our chosen parameter values.…”

“…This configuration of damage remains frozen in the lithosphere and is advected along the surface until a new subduction zone is initiated from the damaged arc (cf. Foley & Bercovici, 2014; Fuchs & Becker, 2019; Heilman & Becker, 2022).…”

“…The reference model was analyzed for a total of 3 model overturns, beginning from an initially dynamically steady-state model run. During the qualifying model run time of 3 overturns, we observe 7 instances of plume-slab termination of the kind we explored in 2-D (Heilman & Becker, 2022), that is, an average of 2.3 terminations every overturn. Termination of subduction was determined from visual, temperature thresholding analysis when no part of the cold, downwelling structure below a temperature threshold typical of continuous subduction was connected to the surface anymore.…”

“…Combining plasticity and such a damage rheology can approximate the behavior of micro‐physical weakening processes like those inferred from grain‐size dependent rheologies (Fuchs & Becker, 2021), which is one of the suggested mechanisms for strain localization in the lithosphere (e.g., Auth et al., 2003; Bercovici & Ricard, 2016; Landuyt & Bercovici, 2009; Landuyt et al., 2008). The strain‐weakening factor ϕ is set to reduce the yield stress from the reference σ y at γ = 0 linearly to ϕσ y at γ ≥ 5; here, a 75% drop from 140 to 35 MPa (Table 1), with parameters based on Heilman and Becker (2022). This damage evolution formulation allows plastically weakened regions to persist and be advected in cold lithosphere while damage in the hotter mantle is healed more readily (cf.…”

The Role of the Overriding Plate and Mantle Viscosity Structure on Deep Slab Morphology

“…However, in nature, the strength of the subducting slab is expected to be controlled by a range of factors, including evolving grain size, and other damage memory (e.g. Montési, 2013;Bercovici & Ricard, 2016;Dannberg et al, 2017;Heilman & Becker, 2022). The implications of these contributions on the effective viscosity and yield stress are only approximately represented by our relatively simple rheological setup.…”