Influence of observed mantle anisotropy on isotropic tomographic models

There is a potential for isotropic tomographic models to be biased by ignored anisotropy; anisotropy could be mapped as artificial isotropic velocity perturbations. We investigate the distribution and strength of this potential bias for three tomographic S velocity models based on regional S and Rayleigh waveforms. We use observed SKS delay times and fast‐axis orientations to compute equivalent S velocity perturbations for each wave path used for the tomographic models. These synthetic perturbations are then combined into a three‐dimensional isotropic model which reflects the potential bias in the actual tomographic model. We quantify anisotropic bias and find that it indeed exists in the isotropic tomographic models investigated here. This bias gets weaker with increasing depth of the anisotropic material, and compared to the isotropic velocity anomalies typically interpreted by tomographers, the bias from anisotropy is small.