## Outer core structures for explaining complexity in PKP traveltime data: a Polar Caps- and Tangent Cylinder model

##### Summary

It has been known for a long time that, when calculating differential times, we find that polar PKPdf waves arrive sooner than predicted compared to equatorial PKPdf waves, while polar PKPbc waves arrive at similar times as predicted times. Inner core anisotropy is the general accepted explanation for this observation (Morelli (1986)), but alternative models might be possible. Here we test alternative models, proposed by Romanowicz (2000) for explaining these data: 'polar caps model' and a 'tangent cylinder model', present in the liquid outer core. The ultimate data for this research are PKPab-PKPdf and PKPbc-PKPdf differential times. The goal of this thesis is to determine whether the observations in the PKPbc-PKPdf and PKPab-PKPdf real data can be explained by waves passing through a faster velocity region inside the outer core in the form of a tangent cylinder or polar caps, without the need for inner core anisotropy. This has been done by writing Fortran programs for determining the depths where the waves enter and leave the distinct regions and use TauP for the arrival times and time spend in these regions. For the real PKPbc-PKPdf data an "L-shape" feature is observed when the residual times are plotted as a function of the angle of the ray paths with the Earth's rotation axis and a "curve" feature for the PKPab-PKPdf dataset. The misfit and variances between the real data and cap or cylinder differential times are determined for every model. It becomes clear that for the PKPbc dataset the L-shape is indeed observed for the tangent cylinder model and a curve structure is visible for the cylinder times of the PKPab dataset with almost the same pattern as the real PKPab differential times, indicating a well fit of the tangent cylinder model with the real data. The velocities in the tangent cylinder are 1%-2% higher than in the surrounding outer core. This is in accordance with the findings of Romanowicz (2003). The polar caps models show high misfits with the real data and is therefore an unrealistic model. The polar cap theory does not need to be pursued further. Even though the misfits of the tangent cylinder models with the real data are relatively low, inner core anisotropy has a lower misfit and is therefore a better fit. Still, the tangent cylinder model cannot be excluded from considerations, since the misfit with the data is low and the velocity increase within the tangent cylinder are possible. The sustainability of heterogeneities within the liquid outer core, however, is still in question.
Another question is the presence of the hemispheres within the core. In this study, no hemisphere are included in the outer more models, still, differences in cap and cylinder times between the eastern and western hemisphere are observed. Therefore the paths are such that it seems like there are hemispheres in the core, and we should be careful with the interpretation of the inner core having hemispherical differences.