Constraining Mantle Viscosity Structure From a Statistical Analysis of Slab Stagnation Events

Abstract

The mantle viscosity is a key factor controlling the dynamics of the Earth's interior, but even the Earth's 1-D radial viscosity profile remains under debate. It has been suggested that mantle viscosity generally increases by ∼10-100 times from the upper mantle to the lower mantle, in order to explain the Earth's long-wavelength geoid anomaly (Hager & Richards, 1989; Panasyuk & Hager, 2000). Traditionally, the viscosity structure is modeled as a stepwise increase at 660 km depth where the ringwoodite to bridgmanite + ferropericlase phase transition happens (e.g., Hager & Richards, 1989; Liu & Zhong, 2016; Mitrovica & Forte, 2004). However, some geoid inversion models with many-layered viscosity structure reported an increase in viscosity around 1,000-1,200 km depth accompanied either with or without a viscosity increase at 660 km depth (e.g., Forte et al., 1991; Kido et al., 1998; King & Masters, 1992). More recent long-wavelength geoid analysis reported a viscosity increase at depths of 800-1,200 km (Rudolph et al., 2015), while studies of plume structures from seismology argued for increased mantle viscosity below ∼1,000 km (French & Romanowicz, 2015). Experimental studies suggested that the viscosity in the uppermost part of the lower mantle may gradually increase down to ∼1,000 km depth to reach its maximum due to potential pressure effects (Marquardt & Miyagi, 2015). The possible compositional variations of lower-mantle enrichment in

Constraints on Mantle Viscosity From Intermediate‐Wavelength Geoid Anomalies in Mantle Convection Models With Plate Motion History

“…The second is on considering additional mantle and lithospheric viscosity features, including non‐Newtonian viscosity (e.g., Hines & Billen, 2012) and depth‐variations in viscosity for the lower mantle (e.g., Mitrovica & Forte, 2004; Rudolph et al., 2015; Steinberger & Calderwood, 2006), although the viscosity increase at 1,000 km depth proposed by Rudolph et al. (2015) has been debated (e.g., Ghosh et al., 2017; Wang & Li, 2020; Yang & Gurnis, 2016). In particular, the implications of non‐Newtonian rheology for non‐uniform plate margin viscosity and lithospheric net rotation need to be further explored.…”

Formation of Horizontally Deflected Slabs in the Mantle Transition Zone Caused by Spinel‐to‐Post‐Spinel Phase Transition, Its Associated Grainsize Reduction Effects, and Trench Retreat

“…At some subduction zones (e.g., the North America and the Central America), slabs penetrate into the lower mantle and could reach the core‐mantle‐boundary (e.g., Grand et al., 1997; van der Hilst et al., 1997); while at other subduction zones (e.g., in the Honshu, Bonin, and Chile), slabs appear to be deflected and extend horizontally over a long distance in the mantle transition zone above the 670 km depth (French & Romanowicz, 2014, 2015; Fukao & Obayashi, 2013; Ritsema et al., 2011) (note that the horizontally deflected slabs are sometimes referred to as “stagnant” slabs in literature, although the slabs are not stationary in the mantle transition zone). Various factors affect subduction zone dynamics and contribute to the formation of deflected slabs in the mantle transition zone including trench retreat, viscosity jump from the mantle transition zone to the lower mantle, the endothermic phase change of spinel‐to‐post‐spinel at ∼670 km depth, slab age and rheology, and nonequilibrium pyroxene garnet transition (e.g., Agrusta et al., 2017; Christensen, 1996; Garel et al., 2014; Goes et al., 2017; Holt et al., 2015; King et al., 2015; Lee & King, 2011; Stegman et al., 2006; Wang & Li, 2020; Yang et al., 2018; Zhong & Gurnis, 1995). However, the relative importance of each factor is still in debate (Goes et al., 2017).…”

Vastly Different Heights of LLVPs Caused by Different Strengths of Historical Slab Push

“…F. Davies, 2008;Machetel & Weber, 1991; P. J. Tackley et al, 1993;Wang & Li, 2020). These experiments also show that slab avalanches are always accompanied by strong return upwelling flows (G. F. Davies, 1995Davies, , 2008Machetel & Weber, 1991;P.…”

“…Seismic observations have shown that whereas some slabs penetrate into the lowermost mantle, others are flattened and appear to be stagnant at ∼660–1,200 km depth (e.g., Fukao & Obayashi, 2013; Grand et al., 1997). Geodynamic modeling experiments have shown that stagnant slabs will eventually sink to the lower mantle, a process known as slab avalanche if it happens in a punctuated scenario (Agrusta et al., 2017; Billen, 2008; G. F. Davies, 2008; Machetel & Weber, 1991; P. J. Tackley et al., 1993; Wang & Li, 2020). These experiments also show that slab avalanches are always accompanied by strong return upwelling flows (G. F. Davies, 1995, 2008; Machetel & Weber, 1991; P. J. Tackley et al., 1993), which may lead to changes in trench motion, surface topography and basin rifting (Capitanio et al., 2009; Pysklywec et al., 2003; Pysklywec & Mitrovica, 1998; Yang et al., 2016, 2018).…”