Layered Structure Of The North American Upper Mantle Inferred By The Combination Of Long And Short Period Seismic Constraints

Abstract:

The nature of the lithosphere-asthenosphere boundary (LAB) and mid-lithospheric discontinuities (MLDs) across the continents remains a subject of debate, and different approaches have been used to image the structure of the lithosphere at continental scale, in particular in North America (NA).

Surface wave tomography provides smooth models and cannot uniquely resolve sharp interfaces, such as the Moho, the LAB or the possible presence of MLDs. In order to image such discontinuities, methods based on the analysis of the scattered wavefield have been developed, such as receiver functions (RF), which are able to image mantle interfaces under single stations.

Since surface wave data and receiver functions provide complementary constraints on shallow earth structure, it is natural to combine them to obtain more robust models of the crust and upper-mantle (e.g. Julia et al., 2000; Lawrence and Wiens, 2004). Here we present a set of 1D shear-wave models, depicting the first 300km of the upper mantle, under a number of stations in North America, using Bayesian inversions.

There are several original aspects to our approach: 1) we consider both Love and Rayleigh wave dispersion data, allowing us to constrain radial anisotropy; 2) instead of a standard RF methodology, scattered body waveforms (i.e. the coda of P phases) are directly inverted through a cross-convolution misfit function(Bodin et al., 2014), thus avoiding the instability issues arising from deconvolution, and, importantly, allowing us to account for crustal multiples without ambiguity; 3) we use a Bayesian trans-dimensional MCMC approach in which the number of isotropic and of anisotropic layers are treated as unknowns, allowing us to obtain models compatible with the data, with the least number of parameters. In contrast to standard RF analysis, our approach allows us to constrain not only the position of discontinuities, but also the variations of shear wave velocity between them, thus providing better constraints on the characteristics of intra-lithospheric low velocity layers.