Multi-scale imaging of 3-D electrical conductivity structure under the contiguous US constrains lateral variations in the upper mantle water content

Abstract

The magnetotelluric (MT) component of the USArray survey is a unique data set that enables imaging the electrical conductivity distribution from the surface down to astenospheric depths. Here, we present a new 3-D electrical conductivity model (MECMUS-2022) derived by inverting data from 1291 USArray MT stations covering ∼80% of the contiguous United States on a quasi-regular 70-km grid. The inversion was performed using a novel multi-scale imaging approach that can consistently incorporate a large range of spatial scales and perform 3-D modeling directly in the spherical frame, completely avoiding the conventional flat-Earth assumption. Furthermore, the use of locally refined meshes allows us to take into account the complex coastline and mimic the natural resolution footprint of broadband MT transfer functions. We find conductivity variations that correlate with known continental structures such as due to the active tectonic processes within the western United States (e.g., Yellowstone hotspot, Basin and Range extension, and subduction of the Juan de Fuca slab) as well as the presence of deep roots beneath cratons. We further interpreted conductivity variations in terms of the upper mantle water content by coupling electrical conductivity with constrains on mantle thermo-chemical structure derived from the analysis of seismic data (in the form of P-to-s and S-to-p receiver functions). Our results suggest the existence of a relatively dry upper mantle beneath the United States, with a weak trend in the upper mantle water content from North to South.