Short period body waves unveil deep Earth’s small-scale structure: from a subducted slab to the core

Abstract

The current thermochemical state of the Earth’s interior is key to understand the evolution of the Earth. Seismic imaging can provide critical constraints on its thermochemical state. Although great progress has been made on large-scale seismic imaging, imaging small-scale structures is more difficult due to more complex wave propagation effects for short period waves. In this dissertation, I investigate several types of short period signals traveling in the deep Earth and use their properties to constrain the structure in Earth’s inner core, outer core and the subducted oceanic crust of the Pacific plate beneath NE Japan.

Strong small-scale (a few km) volumetric heterogeneities in the inner core have been detected using short period (∼1 s) inner core scattering (ICS), termed PKiKP coda, in previous studies. However, most of these data sampled the inner core’s “eastern hemisphere”. I observed PKiKP and its coda waves sampling the “western hemisphere” and confirmed small-scale heterogeneities in the inner core’s “western hemisphere”. However, the ICS from “western hemisphere” is weaker than that from the “eastern hemisphere”. This hemispherical difference, or a regional variation, is also seen other properties of the inner core, such as anisotropy and average compressional wave velocity, Vp.

Previous studies of body waves have imaged a low Vp layer at the top of outer core indicating a stratified layer. To better resolve this low Vp layer, I developed a novel approach using array data to iteratively isolate individual SmKS (m = 3, 4 and 5) signals and better measured SmKS-SKKS differential travel times. I successfully used this method to measure SmKS travel time delays. I also applied this iterative scheme on SPECFEM3D globe synthetic SmKS seismograms to investigate the effects of 3D mantle structure. The results indicate that 3D mantle structure effects are not negligible in our data and could bias the estimation of Vp in the uppermost outer core. After 3D mantle structure corrections, there are still substantial time delays for supporting a low Vp at the top of Earth’s outer core, relative to the PREM.

Subducted oceanic crust plays an important role in material recycling (e.g. water) and the genesis of seismicity. Oceanic crust is usually thinner than 10 km and difficult to image. I observed seismic multi-pathing effects in the subduction zone below NE Japan. These multi-pathing effects are controlled by a seismic attenuation contrast between the back-arc mantle wedge and subducted crust. I used an oceanic crust model based on mineral-physics and conducted SPECFEM2D numerical simulations to investigate these multi-pathing effects. My results support a 9% Vp increase in the oceanic crust at a depth of 130-150 km due to the (lawsonite, talc)-eclogite transition.

Here I demonstrated three examples of using short period body waves to image deep Earth’s small-scale structure. A large amount of short period data accumulated in the last decades is still awaiting seismological analysis. Deciphering these signals will tremendously improve our understanding of small-scale structure in the deep Earth.