Transdimensional Imaging of the Cratonic Lithosphere – TransCraton

This project focuses on investigating the seismic structure of cratons, the oldest continents on Earth, whose understanding is essential to decipher early processes in Earth history, such as the formation and evolution of early continents. Competing hypotheses based on geochemical and petrologic constraints propose different possible models for craton formation; these include under-plating by hot mantle plumes and accretion by shallow subduction zones in continental or arc settings.

Recent observations of seismic discontinuities and anisotropy have contributed to the debate. Unfortunately, different data types sampling the earth at different scales (surface waves, SKS data, and scattered body waves) have been inverted and interpreted separately. This has resulted in different models that are often incompatible, and there is no consensus yet on the nature of layering in cratons. In short, images obtained from converted and reflected body waves see sharp discontinuities, and are interpreted in terms of thermo-chemical stratification, whereas seismic models constructed from long period seismograms see the upper mantle as anisotropic, and are usually interpreted in terms of mantle flow, and past or present deformation processes. A problem is that the small-scale heterogeneities that cannot be resolved by long period seismic waves may also contribute to the observed anisotropy. This is because a stack of horizontal isotropic layers is equivalent, at large scales, to a homogeneous anisotropic medium. Knowledge of the location and amplitude of upper-mantle discontinuities is therefore necessary for the interpretation of long-wavelength anisotropic images in terms of a strain field, and hence to better locate deformation in the mantle.

The proposed work consists on developing and applying an entirely new approach to geophysical data interpretation, where different data types are jointly embraced into a single Bayesian procedure. Our recent work has shown that trans-dimensional sampling methods (where the number of inverted parameters is variable) have some considerable benefits for multi-scale seismological problems involving multiple data types, uncertain data errors and uncertain model parameterizations. Rather than being forced to make decisions on parameterization, level of data fit and weights between data types in advance, these choices can be relaxed and instead constrained by the data themselves.

This novel approach will be applied to a large quantity of seismic data measured beneath several cratons. We propose to produce a number of probabilistic 1D and 2D vertical seismic velocity profiles by combining a set of observables with different sensitivities to seismic anisotropy that are usually interpreted separately: (1) Ps and Sp receiver functions; (2) dispersion measurements of surface waves and their azimuthal variations; (3) SKS waveforms. At each station (1D case), or beneath each line of stations (2D case), we shall carry out a trans-dimensional Monte-Carlo inversion in which the number of layers and the number of parameters in each layer (1 for isotropic layers and 3 for anisotropic layers) are unknown. The produced models will provide new constraints on both upper mantle discontinuities and anisotropy in these regions, thus allowing us to compare the structure between and within cratons, and hence allowing us to better understand the processes involved in the creation of the continental lithosphere.