New constraints on deep-earth heterogeneity and anisotropy

The research that is presented in this dissertation is devoted to the study of the heterogeneity of the deep mantle and its potential implications for studies of the inner core. A new approach is developed that allows to gather informations on the very narrow regions of D″. Using this method, it is qualitatively shown that the deep mantle beneath the Central Pacific is strongly heterogeneous and anisotropic (Chapters 2). The signal due to the anisotropy of D″ can be contaminated by upper mantle effects. In order to separate those two signals, a new method had to be devised (Chapter 3). Chapter 4 describes an extension of the experiment presented Chapter 2, and a comparison between the results that we obtained and tomographic images of the deep mantle. It is showed that seismic tomography underpredicts the effects of D″ heterogeneity by a factor of 2 to 3. Information contained in tomographic studies are then used to produce improved local models of the deep mantle. Those models strongly suggest that a large part of D″ heterogeneity has a chemical origin (Chapter 5).; Most studies of the structure of the inner core are based on waves that propagate through the heterogeneous D″. An analysis of the potential effect of the complex structure of the deep mantle demonstrates that most of the signal observed on PKP(AB-DF) travel times and usually attributed to inner core anisotropy can also be explained by D″ heterogeneity (Chapter 6). In Chapter 7 are finally analysed the most robust evidence for inner core anisotropy, PKP(BC-DF) differential travel time residuals. Those residuals are shown to vary very rapidly, which implies a complex inner core anisotropy or deep mantle structure. The latter is proposed as the most plausible explanation since D″ heterogeneity is already well documented.