Lithospheric deformation and Mantle flow in the asthenosphere-lithosphere system of the flat slab subduction beneath the Peruvian Andes with probabilistic finite-frequency SKS splitting intensity tomography

Flat or near-horizontal subduction of oceanic lithosphere is suggested to occur for ~10% of Earth’s subduction zones. While it is therefore not the dominating geometry, it has been suggested to have significant impact on tectonic processes both currently and in the geologic past. As an example, the ongoing subduction of the aseismic Nazca Ridge beneath South America has been associated with the onset of flat subduction and the termination of arc volcanism in Peru.

In this study, we investigate the impact of flat-slab subduction on the mantle flow and deformation in the larger asthenosphere-lithosphere system beneath the northern portion of the South American subduction zone. Strain in the asthenospheric and lithospheric mantle causes an alignment of intrinsically anisotropic mantle minerals, particularly olivine. The resulting bulk anisotropy can be measured as splitting of core-mantle converted phases, parameterized by the delay time and the fast splitting direction. While shear phases are commonly investigated for average splitting parameters, the tomographic inversion of shear wave splitting data for upper mantle anisotropy has been a longstanding challenge for classical analysis techniques. Recent developments involve the calculation of finite-frequency sensitivity kernels for SKS splitting intensity observations, which allow us to take advantage of overlapping sensitivity kernels at adjacent stations to localize anisotropic structure at depth.

Here we apply probabilistic, finite-frequency SKS splitting intensity tomography to all available datasets across the Andes in Peru and Bolivia to improve our understanding of mantle flow and deformation in the lithosphere in the complex flat slab subduction scenario. While the data sets are mostly comprised of dense lines of seismic stations, the broad lateral distribution of the different networks allows us to combine the data set in a 3D tomographic inversion for upper mantle anisotropy.