Large low-shear-velocity provinces

Large low-shear-velocity provinces (LLSVPs), also called large low-velocity provinces (LLVPs) or superplumes, are characteristic structures of parts of the lowermost mantle, the region surrounding the outer core deep inside the Earth. These provinces are characterized by slow shear wave velocities and were discovered by seismic tomography of deep Earth. There are two main provinces: the African LLSVP and the Pacific LLSVP, both extending laterally for thousands of kilometers and possibly up to 1,000 kilometres vertically from the core–mantle boundary. These have been named Tuzo and Jason respectively, after Tuzo Wilson and W. Jason Morgan, two geologists acclaimed in the field of plate tectonics. The Pacific LLSVP is 3,000 km across and underlies four hotspots on Earth's crust that suggest multiple mantle plumes underneath. These zones represent around 8% of the volume of the mantle, or 6% of the entire Earth.

Other names for LLSVPs and their superstructures include superswells, superplumes, thermo-chemical piles, or hidden reservoirs, mostly describing their proposed geodynamical or geochemical effects. For example, the name "thermo-chemical pile" interprets LLSVPs as lower-mantle piles of thermally hot and/or chemically distinct material. LLSVPs are still relatively mysterious, and many questions remain about their nature, origin, and geodynamic effects.

Seismological modeling
Directly above the core–mantle boundary is a 200 km thick layer of the lower mantle. This layer is known as the D″ ("D double-prime" or "D prime prime") or degree two structure. LLSVPs were discovered in full mantle seismic tomographic models of shear velocity as slow features at the D″ layer beneath Africa and the Pacific. The global spherical harmonics of the D″ layer are stable throughout most of the mantle but anomalies appear along the two LLSVPs. By using shear wave velocities, the locations of the LLSVPs can be verified, and a stable pattern for mantle convection emerges. This stable configuration is responsible for the geometry of plate motions at the surface.

The LLSVPs lie around the equator, but mostly on the Southern Hemisphere. Global tomography models inherently result in smooth features; local waveform modeling of body waves, however, has shown that the LLSVPs have sharp boundaries. The sharpness of the boundaries makes it difficult to explain the features by temperature alone; the LLSVPs need to be compositionally distinct to explain the velocity jump. Ultra-low velocity zones at smaller scales have been discovered mainly at the edges of these LLSVPs.

By using the solid Earth tide, the density of these regions has been determined. The bottom two thirds are 0.5% denser than the bulk of the mantle. However, tidal tomography cannot determine how the excess mass is distributed; the higher density may be caused by primordial material or subducted ocean slabs. The African LLSVP may be a potential cause for the South Atlantic Anomaly.

Origins
Several hypotheses have been proposed for the origin and persistence of LLSVPs, depending on whether the provinces represent purely thermal unconformities (i.e. are isochemical in nature, of the same chemical composition as the surrounding mantle) or represent chemical unconformities as well (i.e. are thermochemical in nature, of different chemical composition from the surrounding mantle). If LLSVPs represent purely thermal unconformities, then they may have formed as large mantle plumes of hot, upwelling mantle. However, geodynamical studies predict that isochemical upwelling of a hotter, lower viscosity material should produce long, narrow plumes, unlike the large, wide plumes seen in LLSVPs. It is important to remember, however, that the resolution of geodynamical models and seismic images of Earth's mantle are very different.

The current leading hypothesis for the LLSVPs is the accumulation of subducted oceanic slabs. This corresponds with the locations of known slab graveyards surrounding the Pacific LLSVP. These graveyards are thought to be the reason for the high velocity zone anomalies surrounding the Pacific LLSVP and are thought to have formed by subduction zones that were around long before the dispersion—some 750 million years ago—of the supercontinent Rodinia. Aided by the phase transformation, the temperature would partially melt the slabs to form a dense melt that pools and forms the ultra-low velocity zone structures at the bottom of the core-mantle boundary closer to the LLSVP than the slab graveyards. The rest of the material is then carried upwards via chemical-induced buoyancy and contributes to the high levels of basalt found at the mid-ocean ridge. The resulting motion forms small clusters of small plumes right above the core-mantle boundary that combine to form larger plumes and then contribute to superplumes. The Pacific and African LLSVP, in this scenario, are originally created by a discharge of heat from the core (4000 K) to the much colder mantle (2000 K); the recycled lithosphere is fuel that helps drive the superplume convection. Since it would be difficult for the Earth's core to maintain this high heat by itself, it gives support for the existence of radiogenic nuclides in the core, as well as the indication that if fertile subducted lithosphere stops subducting in locations preferable for superplume consumption, it will mark the demise of that superplume.

Another proposed origin for the LLSVPs is that their formation is related to the giant-impact hypothesis, which states that the Moon formed after the Earth collided with a planet-sized body called Theia. The hypothesis suggests that the LLSVPs may represent fragments of Theia's mantle which sank through to Earth's core-mantle boundary. The higher density of the mantle fragments is due to their enrichment in iron(II) oxide with respect to the rest of Earth's mantle. This higher iron(II) oxide composition would also be consistent with the isotope geochemistry of lunar samples, as well as that of the ocean island basalts overlying the LLSVPs.

Dynamics
Geodynamic mantle convection models have included compositional distinctive material. The material tends to get swept up in ridges or piles. When including realistic past plate motions into the modeling, the material gets swept up in locations that are remarkably similar to the present day location of the LLSVPs. These locations also correspond with known slab graveyard locations.

These types of models, as well as the observation that the D″ structure of the LLSVPs is orthogonal to the path of true polar wander, suggest these mantle structures have been stable over large amounts of time. This geometrical relationship is consistent with the position of Pangaea and the formation of the current geoid pattern due to continental break-up from the superswell below.

However, the heat from the core is not enough to sustain the energy needed to fuel the superplumes located at the LLSVPs. There is a phase transition from perovskite to post-perovskite from the down welling slabs that causes an exothermic reaction. This exothermic reaction helps to heat the LLSVP, but it is not sufficient to account for the total energy needed to sustain it. So it is hypothesized that the material from the slab graveyard can become extremely dense and form large pools of melt concentrate enriched in uranium, thorium, and potassium. These concentrated radiogenic elements are thought to provide the high temperatures needed. So, the appearance and disappearance of slab graveyards predicts the birth and death of an LLSVP, potentially changing the dynamics of all plate tectonics.