On the limitations of finite-frequency XKS-splitting tomography

Over the last two decades, there has been significant progress in the development of tomographic schemes to infer the anisotropic properties of the upper mantle from teleseismic XKS phases. The methods are based on linking the anisotropic material properties and their changes with surface observables and waveform proxies, such as splitting parameters and intensities, through finite-frequency sensitivity kernels. These approaches are supported by increasingly dense seismic networks that allow for a more precise characterization of short-scale waveform variations due to lateral variations of anisotropy.

Here we focus on the general capability of the tomographic schemes to uniquely resolve the anisotropic structure of the upper mantle from surface observations. For this purpose, we perform full-waveform calculations for relatively simple, canonical models of upper-mantle anisotropy. Our approach involves checkerboard-style tests similar to those typically used to assess the resolving power of tomographic schemes. The models are characterized by four zones of different anisotropic properties. Specifically, we assume orthorhombic symmetry with arbitrarily chosen strength of the anisotropy and orientation of the horizontal a-axis. XKS waveforms, generated from plane-wave initial conditions, traverse through anisotropic models and are recorded at the surface by a dense station profile. In addition to waveforms, we also consider the effects of different anisotropic models on splitting parameters and splitting intensities.

The results show that it is, generally, not possible to uniquely resolve the eight anisotropic parameters (a-axis orientation and strength of anisotropy in four zones) of a given model, even if complete waveforms (under noise-free conditions) are considered. This is related to the fact that waveforms for significantly different anisotropic models, often, are indistinguishable. We conclude that finite-frequency XKS-splitting tomography, alone, is not suited to resolve the anisotropic structures of the upper mantle and that combinations with alternative methods, based on e.g. receiver-function splitting or surface waves, are required.