CHRISTIAN-ALBRECHTS-UNIVERSITAET ZU (DE)
Density variations inside the mantle not only drive mantle convection but are also important indicators of rock composition variation. Satellite gravity measurements, like GOCE, are directly sensitive to large-scale density variations inside the Earth, but their potential is not yet fully used. Instead, density is typically estimated based on variations of seismic shear wave velocity. The gravity field is only used in a second step to estimate the viscosity structure of the Earth. Thus, in the classic approach, resolution of Earth’s structure and dynamics become entangled and there is no possibility for density variations unrelated to velocity variations.
In this project I will rely on gravity data and seismological constraints to estimate the density distribution inside the mantle, without including any dynamical modelling. To achieve a fair combination of seismology and gravity, a good understanding of their respective uncertainties is required. For the gravity field, this mainly relates to uncertainties due to crustal structure and has already been studied, while seismic tomography models suffer from uncertainties due to different smoothing approaches. To estimate these, an ensemble of recent seismic tomography models will be converted to its equivalent representation as surface wave phase speeds, eliminating vertical smoothing. The gravity field and the surface wave speed maps will be used to find discrete anomalous volumes in the mantle in terms of their location, shape, density and seismic wave speed. Since both the number as well as the properties of the anomalous volumes are unknown, a novel Bayesian inversion method will be developed, that uses the transdimensional Monte-Carlo-Markov-Chain algorithm. With this technique an in-depth study of required model complexity, resolution limits and trade-offs is possible.