One of the great accomplishments in the Earth Sciences in the last few decades has been an increasing appreciation for geologic, thermal, and chemical complexity of Earth's interior, spanning a distance from a few kilometers depth down to the outer boundary of the liquid outer core. This complexity is the result of several interconnected geophysical observations that reveal not only a current snapshot of Earth's state of planetary evolution, but also a 4.6 billion year historical record of global scale geodynamic processes and events that have shaped the surface upon which we live as well as the oceans and atmosphere. Our research focuses on furthering our understanding of Earth's 'deep history' by investigating enigmatic structures known as thermochemical piles, which reside near the base of the mantle, just above the boundary with the core. The lateral extent, size, temperature, composition and geologic significance of these structures is presently a subject of great interest because of their potential relationship to the origin of volcanoes, driving forces for present-day tectonic plates and the fate of long-ago subducted tectonic plates. However, their remoteness from Earth's surface - roughly 3000 kilometers down - makes them an especially elusive target for geophysical investigation.
Because of inherent resolution limits and selective sensitivity to the full spectrum of geomaterial properties, no single geophysical method is sufficient for painting a complete picture of this region of the Earth. We propose to study the thermochemical state of the lower mantle through a comprehensive set of numerical experiments, and comparison to observational data, for mutually compatible electromagnetic and geodyamic models. The union of these methods is based on the strong sensitivity of electromagnetic observations to thermochemical variability as manifested in the mantle's electrical conductivity, and the sensitivity of geodynamics modeling to material parameters such as density and viscosity that control the accumulation of material in the lower mantle and that material's thermal structure. Central to these experiments is the over-arching question: How do we, or can we, constrain our knowledge of the thermo-physio-chemical state of the lowermost mantle through geodynamically-compatible electromagnetic observations? Computational experiments will be conducted with a new, parallel electromagnetic MANTle Induction Simulator (project MANTIS) for distributed memory compute architectures, which shares the same underlying mathematical discretization as the CitcomS code that will be used for the geodynamic simulations. This compatibility between discretization methods between the two codes is important for minimizing numerical artifacts as material properties from the electromagnetic domain are mapped into corresponding properties in the geodynamic domain.
