Observing seismic attenuation in the Earth’s mantle and inner core using normal modes

Abstract

Seismic tomographic models based solely on wave velocities have limited ability to distinguish between a temperature or compositional origin for variations in Earth’s structure. Attenuation or damping, which is the loss of energy as a wave travels through the Earth, is able to make that distinction because it is directly sensitive to temperature, partial melt and water content. In this thesis, we use whole Earth oscillations, or normal modes, to study 1D variations in inner core attenuation and 3D variations in mantle attenuation. Focussing and scattering, which are problematic when measuring seismic attenuation, are automatically included in our normal mode calculations when jointly modelling elastic and anelastic structure using first order perturbation theory. For the inner core, we measure the controversial mode pair 10S2-11S2, and find that our measurements indeed agree with a strongly attenuating inner core, just as seen in previous measurements of other inner core sensitive modes, and that our new observations can be used in future inner core tomographic studies. We attribute the previous disagreement to the strong energy exchange between 10S2 and 11S2, that changes the characteristics of the modes as a result of a 0.5% perturbation to either inner core shear velocity or radius. We also measure radial modes, both in isolation and resonating with other modes, to study bulk attenuation and inner core structure with an extended data set. We find indications of lower global bulk attenuation and lower inner core radial anisotropy than previously suggested. For the mantle, the main focus of this thesis, we measure lateral variations in attenuation for both the upper and lower mantle. In the upper mantle, we find anti-correlation between velocity and attenuation, with strong attenuation found in regions with low seismic velocity. These results suggest a thermal origin for the low-velocity oceanic spreading ridges, agreeing with previous studies. In the lower mantle, we find the strongest attenuation in the ‘ring around the Pacific’ high velocity region, which is thought to be the ‘graveyard’ of subducted slabs, and not in the Large Low Shear Velocity Provinces (LLSVPs) under Africa and the Pacific. These findings might potentially be explained by either a small grain-size in combination with cold temperatures in the slabs resulting in strong attenuation, or the presence of the strongly attenuating post-perovskite mineral in the slab regions. The weakest lower mantle attenuation is found near Hawaii and southern Africa, at the edges of the LLSVPs. This suggests that these two low attenuation regions might be dominated by compositional variations instead of temperature, potentially due to iron enrichment in agreement with larger density found in the same places.