Repository logo
 

Transient Creep in Subduction Zones by Long‐Range Dislocation Interactions in Olivine

Published version
Peer-reviewed

Change log

Abstract

Abstract Large earthquakes transfer stress from the shallow lithosphere to the underlying viscoelastic lower crust and upper mantle, inducing transient creep during the postseismic interval. Recent experiments on olivine have provided a new rheological model for this transient creep based on the accumulation and release of back stresses among dislocations. Here, we test whether natural rocks preserve dislocation‐induced stress heterogeneity consistent with the back‐stress hypothesis by mapping olivine from the palaeosubduction interface of the Oman‐UAE ophiolite with high‐angular resolution electron backscatter diffraction. The olivine preserves heterogeneous residual stresses that vary in magnitude by several hundred megapascals over length scales of a few micrometers. Large stresses are commonly spatially associated with elevated densities of geometrically necessary dislocations within subgrain interiors. These spatial relationships, along with characteristic probability distributions of the stresses, confirm that the stress heterogeneity is generated by the dislocations and records their long‐range elastic interactions. Images of dislocations decorated by oxidation display bands of high and low dislocation density, suggesting that dislocation interactions contributed to the organization of the substructure. These results support the applicability of the back‐stress model of transient creep to deformation in the mantle portion of plate‐boundary shear zones. The model predicts that rapid stress changes, such as those imposed by large earthquakes, can induce order‐of‐magnitude changes in viscosity that depend nonlinearly on the stress change, consistent with inferences of mantle rheology from geodetic observations. Plain Language Summary Large earthquakes are generated in relatively cold and brittle rocks but also change the stresses acting on hotter rocks beneath the fault zone. These stress changes cause flow of the hot rocks in the Earth's upper mantle. Both laboratory experiments and observations of deformation at Earth's surface indicate that the viscosity of mantle rocks typically increases by orders of magnitude over the initial increments of deformation after an earthquake. However, the microscale processes that cause this viscosity evolution remain unclear. Recent laboratory experiments suggest that the viscosity evolution may be caused by elastic interactions among linear defects, dislocations, in the crystal lattice of the mineral olivine. Here, we test the relevance of this model to natural rocks by assessing whether the stresses that generate elastic interactions among dislocations are preserved in rocks from Earth's mantle that have been brought to the surface by the motion of tectonic plates. The new observations reveal significant stresses, imparted by dislocations, within individual grains of olivine. This finding supports the relevance of the laboratory‐derived model, which in turn allows the model to be used to predict the magnitude of viscosity changes affecting similar rocks in Earth's mantle. Key Points Olivine from Oman–UAE ophiolite preserves significant stress heterogeneity imparted by dislocations The stress heterogeneity supports a recent model for transient creep based on elastic interactions among dislocations The dislocation‐interaction model predicts order‐of‐magnitude changes in mantle viscosity due to earthquake loading

Description

Journal Title

Journal of Geophysical Research: Solid Earth

Conference Name

Journal ISSN

2169-9313
2169-9356

Volume Title

127

Publisher

American Geophysical Union (AGU)

Rights and licensing

Except where otherwised noted, this item's license is described as http://creativecommons.org/licenses/by/4.0/
Sponsorship
MRC (MR/V021788/1)
This work was supported by the Natural Environment Research Council, grant NE/M000966/1; the Netherlands Organisation for Scientific Research, User Support Programme Space Research, grant ALWGO.2018.038; and startup funds from Utrecht University.