An Integrated Geophysical Approach to Investigating Thermal and Chemical Variations in Earth's Mantle

The Earth’s mantle is heterogeneous as a result of melting, differentiation, plate subduction, and whole-mantle convection throughout geologic time. Our current picture of the mantle has been informed largely by mapping variations in seismic wavespeed. However, it is challenging to infer the thermochemical nature of the mantle from seismic images because they are often poorly resolved, and velocity variations cannot be uniquely related to either temperature or composition. In this thesis, I take a multi-disciplinary approach that combines constraints from geodynamics, mineral physics, and seismology, in order to investigate how ther- mal and compositional Earth models are compatible with seismic observations. I focus primarily on thermal upwellings (i.e., mantle plumes), and assess how these features can be seismically imaged. In chapter 2, I model plume development in a compressible mantle using physics-based simulations of flow in the mantle, and calculate the travel time delays of P waves and S waves propagating through the narrow plume tails in the lower mantle. In chapter 3, I investigate whether or not mantle plume tails can be seismically imaged using common seismic tomog- raphy approaches. I analyze how imaging artifacts can affect our interpretations of the deep mantle below hotspots and find optimal imaging configurations that will maximize resolution of plume tails. In chapter 4, I analyze images of the mantle beneath the Samoa hotspot in global tomography model S40RTS (Ritsema et al., 2011). Specifically, I explore the range of temperatures and compositions that can explain the observed seismic velocity variations and determine if obser- vations are consistent with a lower mantle plume origin of Samoan volcanism. In chapter 5 I investigate the origin of large low velocity provinces (LLVPs) above the core-mantle-boundary beneath Africa and the Pacific, which are thought to be anomalously hot and compositionally distinct mantle domains. I test the hypoth- esis that the anomalies represent an accumulation of recycled oceanic crust above the core by comparing LLVPs resolved in S40RTS to dynamic mantle mixing simulations of Brandenburg et al. (2008). Chapter 6 focuses on how lateral vari- ations of temperature and composition affect the seismic structure of the mantle transition zone (MTZ). I use P-to-S receiver functions to image the strengths and depths of mineralogical phase changes in the MTZ beneath the United States, and relate these observations to the physical conditions of the transition zone using constraints from experimental and theoretical mineral physics.PHDEarth and Environmental SciencesUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttps://deepblue.lib.umich.edu/bitstream/2027.42/145807/1/romaguir_1.pd