Imaging the D" Structure using Waveform Modelling

Abstract

The D" region of the Earth, the lowermost few hundred kilometres of the lower mantle, is designated as a thermal boundary layer and chemically distinct layer above the core-mantle boundary (CMB). It plays an important role in mantle convection, core convection and the resulting geodynamics. Characterising the structure of this region is crucial to a better understanding of the mantle's thermo-chemical evolution and the nature of core-mantle interactions. The top of D'' region characterises a velocity discontinuity - the D'' discontinuity, which likely originates from the phase transition from bridgmanite (Bm) to post-perovskite (pPv) at the pressure-temperature conditions of the lowermost mantle. This discontinuity has been widely observed in a few regions, while much of the Earth is still unmapped. This poses a challenge in deciphering the origin of the D'' discontinuity and resolving the debate whether it is a global feature.

This thesis presents a compilation of published, accepted and submitted work. The main objective is to provide new insights on the origin of the D'' region and its relationship with mantle dynamics. There are three chapters of the thesis. Chapter I contributes to probing the global existence of the D'' discontinuity by adding new observations in unexplored areas on the Earth. Chapter II focuses on the development of a new method to fill in the gaps of the existing methods. Chapter III is dedicated to investigating the potential origin of the D'' region by applying this new method to the D'' region beneath Alaska, Northern Pacific and Central America.

To explore the distribution of the D'' discontinuity in the globe, we first add a new observation of D'' discontinuity beneath the Central Atlantic with new data that greatly improved prior poor data coverage. Through forward waveform modelling, we observe a D'' discontinuity with varying depth and sharpness along a high velocity corridor from north to south beneath the Central Atlantic. Our results also provide evidence on the potential origin of the D'' discontinuity, which is attributed to a Bm-pPv phase transition within both pyrolitic mantle and slab debris beneath Central Atlantic.

To address some of limitations of the current methods on imaging the detailed seismic structures in the D'' region, we develop a grid-search scheme to constrain the detailed 1-D shear-wave velocity structure in the D'' region with quantitative assessment of the uncertainty of 1D models. This new method addresses problems of non-uniqueness and subjectivity in the conventional trial-and-error waveform modelling method which evaluates the fit between data and synthetics qualitatively. A good recovery of the input model with a vertical resolution of 10 km from synthetic experiment demonstrate the feasibility and effectivity of applying grid search method to investigate the 1-D shear wave velocity structure in the D'' region.

In order to probe the global existence and origin of the D'' discontinuity, we apply the grid search method to the lowermost mantle beneath Alaska, northern Pacific and Central America. Our detailed seismic imaging studies allow for speculation that D'' layer is caused by both thermal and chemical effects. Beneath Central America, the D'' layer is dominantly caused by thermal effects associated with Farallon slab remnant, but also affected by varying chemical compositions in the lowermost mantle. By contrast, in regions such as northern Pacific and eastern Alaska, subduction-related chemical heterogeneities are likely to play a dominant role to explain the sharp velocity change in adjacent areas. This difference in the D'' origin could be tightly linked to the slab history in different regions. This thesis shed a light on a better understanding of the D'' origin and its associated mantle dynamics.