Transient rheology of the Sumatran mantle wedge revealed by a decade of great earthquakes

Abstract

Understanding the rheological properties of the upper mantle is essential to develop a consistent model of mantle dynamics and plate tectonics. However, the spatial distribution and temporal evolution of these properties remain unclear. Here, we infer the rheological properties of the asthenosphere across multiple great megathrust earthquakes between 2004 and 2014 along the Sumatran subduction zone, taking advantage of decade-long continuous GPS and tide-gauge measurements. We observe transient mantle wedge flow following these earthquakes, and infer the temporal evolution of the effective viscosity. We show that the evolution of stress and strain rate following these earthquakes is better matched by a bi-viscous than by a power-law rheology model, and we estimate laterally heterogeneous transient and background viscosities on the order of ~1017 and ~1019 Pa s, respectively. Our results constitute a preliminary rheological model to explain stress evolution within earthquake cycles and the development of seismic hazard in the region.

“Measures of Dissipation in Viscoelastic Media” Extended: Toward Continuous Characterization Across Very Broad Geophysical Time Scales

“…While this model is adopted by the majority of estimates in Figure 1b, one postseismic study used a Burgers model (Pollitz, 2003), resulting in two viscosity estimates: ss and t (where in Figure 1a, the lower and upper limits of the yellow box denote t and ss , respectively). Indeed, at other locations, similarly complicated viscoelastic models were required to fit data (e.g., for PSR see Qiu et al, 2018, (Argus et al, 2014;Bao et al, 2016;Bills et al, 1994;Creveling et al, 2017;Dalton et al, 2008;Lau et al, 2016;Mitrovica & Forte, 2004;Pollitz et al, 2000;Pollitz, 2003). For all estimates that adopted the Maxwell model refers to ss , while the upper and lower limits of the Burgers model postseismic estimate represent ss and t , respectively (Pollitz, 2003).…”

“…While this model is adopted by the majority of estimates in Figure b, one postseismic study used a Burgers model (Pollitz, ), resulting in two viscosity estimates: η ss and η t (where in Figure a, the lower and upper limits of the yellow box denote η t and η ss , respectively). Indeed, at other locations, similarly complicated viscoelastic models were required to fit data (e.g., for PSR see Qiu et al, , or for normal modes and tides see Lau & Faul, ). We emphasize that the choice of viscoelastic model does not provide any information on the underlying mechanistic process (e.g., the adoption of the Zener model in the seismic example does not imply anything about the possibility of diffusion or dislocation creep), which can lead to further complication in interpretation across different estimates.…”

The Role of Viscoelastic Stress Transfer in Long‐Term Earthquake Cascades: Insights After the Central Italy 2016–2017 Seismic Sequence

“…By comparing long‐term geologic slip rates with geodetically derived fault slip rates of 15 continental strike‐slip faults, Meade et al () concluded that a two‐layer Burgers model better fit the observations of deformation over the entire earthquake cycle. More recent papers have directly imaged transient viscosity (most likely due to a biviscous Burgers rheology) in the lower crust of the region struck by the 2016 M w 7.1 Kunamoto earthquake (Moore et al, ) and in the Sumatran mantle wedge (Qiu et al, ). These new results show that different rheological models may be plausible also for the central Apennines and that further investigations, possibly based on the new geodetic data available after the 2016–2017 Central Italy sequence, are necessary to better constrain the rheological parameters of the lithosphere in Central Italy.…”

Structural control and system-level behavior of the seismic cycle at the Nankai Trough

“…3a, 4f ). The afterslip period consists of a short pulse of rapidly decaying creep that lasts a few years, as observed after all suitably instrumented large subduction earthquakes (e.g., Hsu et al 2006;Chlieh 2007;Feng et al 2016;Tsang et al 2016;Bedford et al 2016;Klein et al 2016;Hu et al 2016;Qiu et al 2018;Tang et al 2019;Muto et al 2011).…”

“…Thermally activated viscoelastic flow in the asthenosphere facilitates plate tectonics and the Wilson cycle (e.g., Burke 2011). Asthenospheric flow is modulated by the seismic cycle (Barbot 2018b), creating large-scale deformation during the postseismic period (Savage and Prescott 1978;Hirahara 2002;Hu et al 2004;Wang 2007;Pollitz et al 2008;Wang et al 2012;Sun 2014;Masuti et al 2016;Klein et al 2016;Li et al 2018;Qiu et al 2018;Weiss et al 2019). Viscoelastic flow and fault slip are mechanically coupled during postseismic deformation (Agata et al 2019;Muto et al 2011).…”