UCLA

We use data from seismic networks with unprecedented dense coverage to study the Earth's structure under Mexico.

First, we develop a three-dimensional (3-D) model of shear-wave velocity and anisotropy for the Mexico subduction zone using fundamental mode Rayleigh wave phase velocity dispersion measurements. The 3-D nature of our surface-wave-based results allows for better understanding of the interaction between the subducting slab, mantle lithosphere, and asthenosphere in the top 200 km. Our phase velocity maps reveal lateral variations at all periods consistent with the presence of flat and steep subduction. We also find that the data are consistent with two layers of anisotropy beneath Mexico: a crustal layer and a deeper layer that includes the lithosphere and asthenosphere, with the fast direction interpreted as aligned with the toroidal mantle flow around the slab edges. Our combined azimuthal anisotropy and velocity model enables us to analyze the transition from flat to steep subduction and to determine whether the transition involves a tear resulting in a gap between segments or is a continuous deformation caused by a lithospheric flexure. Our anisotropy results favor a tear, which is also consistent with the geometry of the volcanic belt.

Next, we conduct a shear wave splitting analysis that results in delay times of 1-2 s and the fast direction that coincides with the absolute plate motion for the Mesoamerican Seismic Experiment (MASE) stations as well as stations east of the MASE array. The significant difference of the anisotropy in the upper 200 km, as detected by the surface wave analysis, and the average anisotropy between the CMB and the surface, as resolved by the shear wave splitting, implies that the shear wave splitting results are dominated by a structure deeper than 200 km. Since the time delays are significantly longer for the shear wave splitting results, the deeper structure is either much larger than 200 km, or has stronger anisotropy than the top 200 km, or a combination of both. At the same time, several relatively subtle features in the shear wave splitting results reveal potential influences of the shallow structure and its deeper extensions. This includes a small change in the fast direction around the southern edge of the Trans-Mexican Volcanic Belt (TMVB), which is located above the transition from the flat to steep subduction, as well as a different pattern of fast directions west of the MASE array, the region on top of two smaller subducting slabs.

Finally, we determine phase velocities of higher modes of Rayleigh waves, in order to constrain the depth of the anisotropy revealed by the shear wave splitting. Our analysis shows that the phase velocities for a number of overtones and periods are fastest in the direction predicted by shear wave splitting, suggesting that they are affected by the same deeper structure. Remarkably, the results for different directions are consistent with the presence of azimuthal anisotropy. Inspection of obtained phase velocities together with the sensitivity kernels tentatively indicates that a layer at the 200-400 km depth is a likely candidate for the source of the anisotropy. We find that such a layer can reproduce the observed shear wave splitting delays for reasonable values of anisotropy. The 200-400 km depth likely corresponds to the bottom of the asthenosphere, and it may be affected by the plate motion, explaining why the fast shear wave splitting direction is aligned with the plate motion. This tentative estimate of the anisotropy depth is consistent with findings in Northern Australia.