Seismic anisotropy: tracing plate dynamics in the mantle

Elastic anisotropy is present where the speed of a seismic wave depends on its direction. In Earth's mantle, elastic anisotropy is induced by minerals that are preferentially oriented in a directional flow or deformation. Earthquakes generate two seismic wave types: compressional (P) and shear (S) waves, whose coupling in anisotropic rocks leads to scattering, birefringence, and waves with hybrid polarizations. This varied behavior is helping geophysicists explore rock textures within Earth's mantle and crust, map present-day upper-mantle convection, and study the formation of lithospheric plates and the accretion of continents in Earth history.     

Water-induced fabric transitions in olivine

The interpretation of seismic anisotropy in Earth's upper mantle has traditionally been based on the fabrics (lattice-preferred orientation) of relatively water-poor olivine. Here we show that when a large amount of water is added to olivine, the relation between flow geometry and seismic anisotropy undergoes marked changes. Some of the puzzling observations of seismic anisotropy in the upper mantle, including the anomalous anisotropy in the central Pacific and the complicated anisotropy in subduction zones, can be attributed to the enrichment of water in these regions.     

The subduction zone flow field from seismic anisotropy: a global view

Although the morphologies of subducting slabs have been relatively well characterized, the character of the mantle flow field that accompanies subduction remains poorly understood. To analyze this pattern of flow, we compiled observations of seismic anisotropy, as manifested by shear wave splitting. Data from 13 subduction zones reveal systematic variations in both mantle-wedge and subslab anisotropy with the magnitude of trench migration velocity |V(t)|. These variations can be explained by flow along the strike of the trench induced by trench motion. This flow dominates beneath the slab, where its magnitude scales with |V(t)|. In the mantle wedge, this flow interacts with classical corner flow produced by the convergence velocity V(c); their relative influence is governed by the relative magnitude of |V(t)| and V(c).     

Slip systems in MgSiO3 post-perovskite: implications for D' anisotropy

Understanding deformation of mineral phases in the lowermost mantle is important for interpreting seismic anisotropy in Earth's interior. Recently, there has been considerable controversy regarding deformation-induced slip in MgSiO(3) post-perovskite. Here, we observe that (001) lattice planes are oriented at high angles to the compression direction immediately after transformation and before deformation. Upon compression from 148 gigapascals (GPa) to 185 GPa, this preferred orientation more than doubles in strength, implying slip on (001) lattice planes. This contrasts with a previous experiment that recorded preferred orientation likely generated during the phase transformation rather than deformation. If we use our results to model deformation and anisotropy development in the D' region of the lower mantle, shear-wave splitting (characterized by fast horizontally polarized shear waves) is consistent with seismic observations.     

High shear strain of olivine aggregates: rheological and seismic consequences

High-pressure and high-temperature torsion experiments on olivine aggregates in dislocation creep show about 15 to 20% strain weakening before steady-state behavior, characterized by subgrain-rotation recrystallization and a strong lattice preferred orientation. Such weakening may provide a way to focus flow in the upper mantle without a change in deformation mechanism. Flow laws derived from low strain data may not be appropriate for use in modeling high strain regions. In such areas, seismic wave propagation will be anisotropic with an axis of approximate rotational symmetry about the shear direction. In contrast to current thinking, the anisotropy will not indicate the orientation of the shear plane in highly strained, recrystallized olivine-rich rocks.     

Seismic determination of elastic anisotropy and mantle flow

When deformed, many rocks develop anisotropic elastic properties. On many seismic records, a long-period (100 to 250 seconds), "quasi-Love" wave with elliptical polarization arrives slightly after the Love wave but before the Rayleigh wave. Mantle anisotropy is sufficient to explain these observations qualitatively as long as the "fast" axis of symmetry is approximately horizontal. Quasi-Love observations for several propagation paths near Pacific Ocean subduction zones are consistent with either flow variations in the mantle within or beneath subducting plates or variations in the direction of fossil spreading in older parts of the Pacific plate.     

Spin crossover in ferropericlase at high pressure: a seismologically transparent transition?

Seismic discontinuities in Earth typically arise from structural, chemical, or temperature variations with increasing depth. The pressure-induced iron spin state transition in the lower mantle may influence seismic wave velocities by changing the elasticity of iron-bearing minerals, but no seismological evidence of an anomaly exists. Inelastic x-ray scattering measurements on (Mg(0.83)Fe(0.17))O-ferropericlase at pressures across the spin transition show effects limited to the only shear moduli of the elastic tensor. This explains the absence of deviation in the aggregate seismic velocities and, thus, the lack of a one-dimensional seismic signature of the spin crossover. The spin state transition does, however, influence shear anisotropy of ferropericlase and should contribute to the seismic shear wave anisotropy of the lower mantle.     

Mantle seismic structure beneath the MELT region of the east pacific rise from P and S wave tomography

Relative travel time delays of teleseismic P and S waves, recorded during the Mantle Electromagnetic and Tomography (MELT) Experiment, have been inverted tomographically for upper-mantle structure beneath the southern East Pacific Rise. A broad zone of low seismic velocities extends beneath the rise to depths of about 200 kilometers and is centered to the west of the spreading center. The magnitudes of the P and S wave anomalies require the presence of retained mantle melt; the melt fraction near the rise exceeds the fraction 300 kilometers off axis by as little as 1%. Seismic anisotropy, induced by mantle flow, is evident in the P wave delays at near-vertical incidence and is consistent with a half-width of mantle upwelling of about 100 km.     

The mantle flow field beneath western North America

Although motions at the surface of tectonic plates are well determined, the accompanying horizontal mantle flow is not. We have combined observations of surface deformation and upper mantle seismic anisotropy to estimate this flow field for western North America. We find that the mantle velocity is 5.5 +/- 1.5 centimeters per year due east in a hot spot reference frame, nearly opposite to the direction of North American plate motion (west-southwest). The flow is only weakly coupled to the motion of the surface plate, producing a small drag force. This flow field is probably due to heterogeneity in mantle density associated with the former Farallon oceanic plate beneath North America.     

Post-perovskite phase transition in MgSiO3

In situ x-ray diffraction measurements of MgSiO3 were performed at high pressure and temperature similar to the conditions at Earth's core-mantle boundary. Results demonstrate that MgSiO3 perovskite transforms to a new high-pressure form with stacked SiO6-octahedral sheet structure above 125 gigapascals and 2500 kelvin (2700-kilometer depth near the base of the mantle) with an increase in density of 1.0 to 1.2%. The origin of the D" seismic discontinuity may be attributed to this post-perovskite phase transition. The new phase may have large elastic anisotropy and develop preferred orientation with platy crystal shape in the shear flow that can cause strong seismic anisotropy below the D" discontinuity.     

Plastic deformation of MgGeO3 post-perovskite at lower mantle pressures

Polycrystalline MgGeO3 post-perovskite was plastically deformed in the diamond anvil cell between 104 and 130 gigapascals confining pressure and ambient temperature. In contrast with phenomenological considerations suggesting (010) as a slip plane, lattice planes near (100) became aligned perpendicular to the compression direction, suggesting that slip on (100) or (110) dominated plastic deformation. With the assumption that silicate post-perovskite behaves similarly at lower mantle conditions, a numerical model of seismic anisotropy in the D' region implies a maximum contribution of post-perovskite to shear wave splitting of 3.7% with an oblique polarization.     

Anisotropic origin of transform faults

Transform faults appear in the process of stretching during freezing of the surface films on liquid wax. These films are composed of a warp yarn of wax fibers with optical anisotropy. This fabric is absent in materials that fail to produce transform faults. The mechanical anisotropy of these wax films (with high tensile strength and low shear strength in the direction of spreading) is responsible for the initiation of the transform faults. It is suggested that the anisotropy of the ocean upper mantle recorded seismically may likewise be responsible for the creation of the ridge-ridge transform faults in the oceans.     

A complex pattern of mantle flow in the Lau backarc

Shear-wave splitting analysis of local events recorded on land and on the ocean floor in the Tonga arc and Lau backarc indicate a complex pattern of azimuthal anisotropy that cannot be explained by mantle flow coupled to the downgoing plate. These observations suggest that the direction of mantle flow rotates from convergence-parallel in the Fiji plateau to north-south beneath the Lau basin and arc-parallel beneath the Tonga arc. These results correlate with helium isotopes that map mantle flow of the Samoan plume into the Lau basin through an opening tear in the Pacific plate.     

Upper mantle discontinuity topography from thermal and chemical heterogeneity

Using high-resolution stacks of precursors to the seismic phase SS, we investigated seismic discontinuities associated with mineralogical phase changes approximately 410 and 660 kilometers (km) deep within Earth beneath South America and the surrounding oceans. Detailed maps of phase boundary topography revealed deep 410- and 660-km discontinuities in the down-dip direction of subduction, inconsistent with purely isochemical olivine phase transformation in response to lowered temperatures. Mechanisms invoking chemical heterogeneity within the mantle transition zone were explored to explain this feature. In some regions, multiple reflections from the discontinuities were detected, consistent with partial melt near 410-km depth and/or additional phase changes near 660-km depth. Thus, the origin of upper mantle heterogeneity has both chemical and thermal contributions and is associated with deeply rooted tectonic processes.     

Shipboard geophysical indications of asymmetry and melt production beneath the east pacific rise near the MELT experiment

Near the Mantle Electromagnetic and Tomography (MELT) Experiment, seamounts form and off-axis lava flows occur in a zone that extends farther to the west of the East Pacific Rise than to the east, indicating a broad, asymmetric region of melt production. More seamounts, slower subsidence, and less dense mantle on the western flank suggest transport of hotter mantle toward the axis from the west. Variations in axial ridge shape, axial magma chamber continuity, off-axis volcanism, and apparent mantle density indicate that upwelling is probably faster and more melt is produced beneath 17 degrees15'S than beneath 15 degrees55'S. Recent volcanism occurs above mantle with the lowest seismic velocities.     

The process of formation of ocean crust

Ocean crust is the outermost layer of earth under the oceans. It is separated from the underlying mantle by a seismic transition zone called the Moho. A widely held view is that the Moho represents a petrologic change from basaltic-type rocks to a mantle composed mostly of olivine and pyroxene. According to this view, crust is formed by a steady segregation of basaltic melt, derived from partial melting of the mantle, into a crustal magma chamber wherein cooling and crystallization bring about steady-state accretion to the continuously spreading plates. There is sufficient disagreement between the predictions of this hypothesis and marine geophysical data to cause one to doubt the validity of this formation process. At least two other processes are more compatible with the geophysical data. In one, the crust is formed from the episodic injection of basaltic dikes from a mantle reservoir and the Moho is a primary petrologic boundary. In the other, the crust is treated as a mechanical boundary layer in which thermal contraction results in cracking; by comparison, in the mantle thermal contraction is accommodated by flow. The upper part of the crust is formed from episodic extrusion and intrusion of basaltic melt. The lower crust is formed by rapid hydrothermal alteration of mantle that may be continuously or episodically injected by viscous flow at temperatures below the melting temperature.     

First-principles determination of elastic anisotropy and wave velocities of MgO at lower mantle conditions

The individual elastic constants of magnesium oxide (MgO) have been determined throughout Earth's lower mantle (LM) pressure-temperature regime with density functional perturbation theory. It is shown that temperature effects on seismic observables (density, velocities, and anisotropy) are monotonically suppressed with increasing pressure. Therefore, at realistic LM conditions, the isotropic wave velocities of MgO remain comparable to seismic velocities, as previously noticed in athermal high-pressure calculations. Also, the predicted strong pressure-induced anisotropy is preserved toward the bottom of the LM, so lattice-preferred orientations in MgO may contribute substantially to the observed seismic anisotropy in the D" layer.     

Inferences on flow at the base of Earth's mantle based on seismic anisotropy

We applied global waveform tomography to model radial anisotropy in the whole mantle. We found that in the last few hundred kilometers near the core-mantle boundary, horizontally polarized S-wave velocities (VSH) are, on average, faster (by approximately 1%) than vertically polarized S-wave velocities (VSV), suggesting a large-scale predominance of horizontal shear. This confirms that the D" region at the base of the mantle is also a mechanical boundary layer for mantle convection. A notable exception to this average signature can be found at the base of the two broad low-velocity regions under the Pacific Ocean and under Africa, often referred to as "superplumes," where the anisotropic pattern indicates the onset of vertical flow.     

The Gutenberg discontinuity: melt at the lithosphere-asthenosphere boundary

The lithosphere-asthenosphere boundary (LAB) beneath ocean basins separates the upper thermal boundary layer of rigid, conductively cooling plates from the underlying ductile, convecting mantle. The origin of a seismic discontinuity associated with this interface, known as the Gutenberg discontinuity (G), remains enigmatic. High-frequency SS precursors sampling below the Pacific plate intermittently detect the G as a sharp, negative velocity contrast at 40- to 75-kilometer depth. These observations lie near the depth of the LAB in regions associated with recent surface volcanism and mantle melt production and are consistent with an intermittent layer of asthenospheric partial melt residing at the lithospheric base. I propose that the G reflectivity is regionally enhanced by dynamical processes that produce melt, including hot mantle upwellings, small-scale convection, and fluid release during subduction.     

Trench-parallel anisotropy produced by foundering of arc lower crust

Many volcanic arcs display fast seismic shear-wave velocities parallel to the strike of the trench. This pattern of anisotropy is inconsistent with simple models of corner flow in the mantle wedge. Although several models, including slab rollback, oblique subduction, and deformation of water-rich olivine, have been proposed to explain trench-parallel anisotropy, none of these mechanisms are consistent with all observations. Instead, small-scale convection driven by the foundering of dense arc lower crust provides an explanation for the trench-parallel anisotropy, even in settings with orthogonal convergence and no slab rollback.