Interpreting spatial variations in anisotropy: Insights into the Main Ethiopian Rift from SKS waveform modelling

Abstract

Seismic anisotropy is a common feature in the upper mantle and measuring shear wave splitting in core phases is a common approach in estimating its characteristics. Large lateral variations in estimated splitting parameters are observed over small spatial distances in many differing tectonic regions, including areas of continental break-up such as the Main Ethiopian Rift (MER). We investigate the ability of shear wave splitting analysis to constrain spatial variations in anisotropy using a one-way wave equation modelling scheme to generate bandlimited waveforms for a suite of models representing regions with rapidly changing anisotropy. We show that shear wave splitting can identify lateral variation in anisotropy on the order of 20-50 km, where a change in fast direction demarcates the transition in anisotropy. In addition, variation in the amount of splitting is complicated close to the transition, and is sensitive to the vertical thickness of anisotropy. We have used these modelling results to interpret shear wave splitting measurements for the MER. The model that best fits the observations has a 100- km wide rift zone with a fast direction of 30° outside and 20° inside the rift. The model has 9 per cent anisotropy close to the western margin, with 7 per cent anisotropy elsewhere. In all regions of the model we constrain the anisotropy to begin at a depth of 90 km. The depth of anisotropy is consistent with geochemical estimates of the depth of melt initiation beneath the region. Also the elevated splitting beneath the western margin supports evidence of low velocities and highly conductive zones from seismic tomography and magneto-tellurics, suggesting melt is more focused along the western margin. This study shows how observations of SKS-wave splitting from dense seismic networks can be used to map sharp lateral changes and constrain the depth of the anisotropy. © 2010 The Authors Journal compilation © 2010 RAS.