Source: Dissertation Abstracts International, Volume: 79-05(E), Section: B.

Adviser: Jeffrey J. Park.

Observation of seismic interfaces with velocity contrasts in the Earth's interior reveals significant Earth structure that offers clues to its structure and history. We interrogate the nature of seismic interfaces within Earth's crust and mantle using high-resolution multi-taper receiver functions. We detect sharp seismic-velocity increase or decrease in the vertical direction by estimating P-to-S converted phases beneath the seismic station. In this dissertation, we investigate crustal structure and anisotropy related to deformation processes and subsequent metamorphism. The distribution of seismic anisotropy within the crust, rather than its averaged anisotropy, is important to infer dynamic models. We infer anisotropic velocity-models to demonstrate whether to interpret Moho-generated Ps waves in terms of shear-wave splitting in the whole crust, or a doublet of P-to-S conversions from the top and bottom of a thin anisotropic crustal layer. We also infer anisotropic crustal structures to distinguish different dynamic models for the evolution of the thick Tibetan crust, including shortening and thickening, underthrusting, and crustal channel flow.

In addition, we use receiver functions to search for evidence of low-velocity anomalies above and below the hydrous mantle transition zone, which may indicate expelled water that causes partial melting. This dissertation presents evidence of depressed 660-km discontinuities that are clearly imaged beneath different locales such as the Japan subduction zone and the European Alpine mountain chains. The depression of 660-km discontinuity may imply downwelling cold materials into the upper lower-mantle due to gravitational instability. Mass transfer across the transition-zone boundaries may transport hydrated minerals from the transition zone into the water-poor upper or lower mantle. Water release in the mantle surrounding the transition zone could cause dehydration melting and produce seismic low-velocity anomalies if some conditions are met. Seismic observations of low-velocity layers surrounding the transition zone could provide clues of water circulation at mid-mantle depths.

In Chapter 2, we perform receiver-function analysis on teleseismic data recorded by 3 permanent seismic stations located on three different types of crust, i.e. normal continental crust, thickened continental crust, and thin oceanic crust. Anisotropic velocity-models reveal distributed anisotropy at various levels within the crust, rather than averaged anisotropy through the whole crust. In Chapter 3, we apply multi-taper receiver-function technique to teleseismic data recorded by 35 temporary seismic stations in northeast Tibetan Plateau. We detect strong (sub)horizontalaxis anisotropy in distinct layers in the middle and lower crust, implying vertical metamorphic fluid domains that formed by shearing processes within the boundary layers of channel flow.

In Chapter 4 and 5, we use high-resolution migrated receiver functions to detect seismic-velocity drops surrounding the transition zone beneath northeast Asia and the European Alps. We implement the common conversion point stacking method to construct continuous receiver-function sections. In both cases, we detect localized low-velocity layers below the depressed 660-km discontinuity and above the 410-km discontinuity, suggesting dehydration melting induced by water flow out of the transition zone. In addition, we note a depth gap between the depressed 660-km discontinuity and the seismic-velocity reductions below it. We propose a "Trojan Horse" model to explain this phenomenon. Metallic iron could exist in the gap layer, which absorbs water, preserves the hydration of rock just below the 660-km discontinuity and prevents partial melting of the lower-mantle rocks. Metallic iron could become unstable at greater depth, releasing water and causing partial melting that produces seismic velocity drops.