Anisotropy in the D'' Layer

Understanding the elasticity of materials is important for understanding the processes occurring in the Earth’s mantle. From observations on the lower mantle, we find that there is a significant amount of seismic anisotropy, which is likely related to the primary mineral constituents of the lower most mantle: the D” region. This region is a unique region about 150-300 km thick at the base of the mantle. We find that there is lateral heterogeneity and anisotropy defined by the elasticity of the minerals in this region. The main challenges in our knowledge of this area lie in the limited ability to observe this region in detail, as well as measuring elastic constants at elevated pressure-temperature conditions in the laboratory. However, the purpose of this research is to connect seismic observations with mineral elastic constants with our current tools to paint a global average of what we currently know about this region. Our work includes compiling observed seismic anisotropy from various research and relating it to measured elastic constants from experimental research done in laboratories. As a result, we hope to create a global model of the D” region where we can predict P-wave seismic velocities and understand the D” region with respect to how it affects mantle dynamics in the Earth.

The presence of seismic anisotropy has been found in many regions of the Earth’s interior and known since the 19th century. In particular, the D” is a unique region about 150 - 300 km thick and at the base of the mantle. It is characterized by unusually low velocity gradients and large seismic variability, in which seismic velocities vary considerably depending on the region where it is observed. Much of the lower mantle appears to be relatively isotropic, except in the lowermost 200-300 km, where seismic anisotropic evidence has been observed since the late 1980’s. Shear wave splitting measurements have contributed to the creation of a picture where strong anisotropy is associated with high shear velocities at the edges of the large low shear velocity provinces (LLSVPs) in the central Pacific and under Africa (Romanowicz 2017). The correlation is consistent with the presence of the highly anisotropic MgSiO3 post-perovskite crystals, which are potentially aligned during the deformation of slabs on the core-mantle boundary (CMB) and the upwelling from LLSVPs. Understanding the deep mantle is a puzzle that continues to be solved in current geophysical research and observations. Applications of seismology and mineral physics have been utilized for the study of the Earth’s interior in hopes of determining the properties and composition of the inaccessible deep Earth. Tools used for determining such properties include natural abilities, such as earthquakes, and controlled abilities, such as explosions and vibrators, to obtain sources of elastic waves and their attenuation. This review focuses on the issue of anisotropy and the importance of its role in theoretical and methodological developments.

Global tomographic studies make it possible to study the radial and azimuthal anisotropy throughout the mantle. Combining observational evidence, such as local studies of shear body waves sensitive to the deep mantle, and tomographic studies help contribute to the studies of seismic anisotropy constrained in the D” region. Studying the differences in propagation speed between surface waves that are polarized differently, free oscillations of the Earth excited by large earthquakes, and normal mode splitting can be measured and inverted to obtain anisotropic models of the mantle and core structure (Long 2010), which is our goal of the study. By studying anisotropy and understanding the observations, we may be able to validate tectonic models and mantle dynamics in Earth. Anisotropy can be probed using body waves or surface waves; body waves propagating through the Earth’s volume, or surface waves that propagate along the surface and involve elliptical, retrograde particle motion.

Along with global tomographic studies, mineral physics in the D’’ layer has proved relevant for understanding the anisotropy occurring in the lowermost mantle. Various laboratory work and research measuring the elastic constants of mineral constituents in the lower mantle have been performed in various research groups. Murakami et al. (2004) are at the forefront of mineral physics experiments on the D’’ region, as they were the first to discover a post-perovskite phase transition in MgSiO3 at lower mantle pressure and temperature conditions. This new phase transition has sparked interest in correlating it with anisotropy observed in seismology and elastic tensors measured from high pressure mineral experiments.

Thesis Motivation
Here, I focus on observing seismic anisotropy compiled from various studies to complete a global model of the lowermost mantle. In order to examine the deepest parts of the Earth in detail, we must use a combination seismic waves and mineral physics observations to create a global model of the D” layer. By creating a global model of the D” layer, we would like to make a prediction on the attenuation of P-waves as well as comparing the measured values of elastic and stiffness tensors of lower mantle minerals with observed seismic S-wave velocities.

Seismic Anisotropy in the Lowermost Mantle
Seismic anisotropy gives rise to the directional dependence of seismic wave speeds and is an intrinsic property of elastic materials (Long 2010). Vinnik et al. (1989) were the first to observe shear wave splitting attributed to anisotropy in the D” region. (Vinnik 1989). Following the first observation, many students starting documenting splitting times in shear waves, such as ScS and Sdiff. The culmination of sampled global data resulted in the general observation that the vertically polarized wave (SV) are generally delayed with respect to the horizontally polarized wave (SH) in the circum-Pacific, and SH was delayed with respect to SV on paths sampling large low shear velocity provinces (LLSVPs) in central Pacific and under Africa. Generally, analyzing splitting and scattering of shear waves is the most popular wave method of studying anisotropy. In the last two decades, reports of splitting in ScS and Sdiff have been attributed to anisotropy in the D” because of its abilities to reach the core’s boundaries. The difference in arrival times of the waves can be measured in accordance with anisotropy observed at the CMB and in the D” region. Other methods of observing seismic anisotropy in the D” region include using normal modes from Earth’s free oscillations. Different normal modes are sensitive to anisotropy in different depth ranges and can give us details of anisotropy in different regions of Earth. There are two types of normal modes: spheroidal modes related to mantle Rayleigh waves, and torosional modes related to mantle Love waves.

A challenge of distinguishing different possible causes of anisotropy through seismology include accounting for the interference of seismic waveforms by the strong upper mantle anistoropy sampled by body waves. It is important to correct for the upper mantle when focusing on the lower mantle to consider shear wave splitting in the D”. Additionally, it is difficult to distinguish if anisotropy is due to the layering, crystal alignment, whether the anisotropy is transverse isotropy with a vertical (VTI) or tilted (TTI) axis of symmetry, or if there is more complex symmetries involved. Another conflict is if there is laterally discontinuity in a laterally heterogeneous, but isotropic structure. More generally, apparent splitting in S waves sampling the D” region may also be due to causes other than anisotropy, such as propagation in a complex heterogeneous medium near the core-mantle boundary (CMB), which would require more sophisticated modeling than infinite frequency ray theory (Romanowicz 2017).

Ultra-low velocity zones (ULVZs) were first noted with SpdKS phase diffracting along the CMB to determine the P-wave velocity at the base of the mantle beneath the central Pacific in research from Garnero and Helmberger, 1995, 1996 (Bower 2011). There have been studies suggesting that they have shear wave velocities that are up to 30% lower than the surrounding material and are mapped to be hundreds of kilometers in diameter and tens of kilometers thick. The composition and origin is still unknown, but have been thought to be related to the location of hotspots and near the edges of LLSVPs. When compiling multiple research reports from different regions of the world to create a global understanding of anisotropy, we observe that VSH > VSV within the ring of fast velocities surrounding the Pacific and African LLSVPs, but vanishing or changing signs (VSV > VSH) inside the LLSVPs.(Panning 2006) suggests that V > VSV is generated in the predominant horizontal flow of a mechanical boundary layer, with a change in signature related to transition to upwelling at the superplumes. This is associated with the large-scale low-velocity superplumes under the central Pacific and Africa. The dynamics occurring in ULVZs and LLSVPs affect the seismic velocities observed in the D” region and within these two regions themselves. It represents the current thermal and chemical state of the Earth’s mantle dynamics, which infers the density structure along with understanding the Earth’s mantle convection.