Earth's Inner Core as a Conglomerate of Anisotropic Domains

The days when the Earth’s inner core (IC) was viewed as a homogenous solid body with a cylindrical anisotropy having a fast axis nearly parallel to the Earth’s rotation axis are now behind us. A number of concepts for the IC structure and dynamics have been proposed to explain different types of seismological observations, but due to a lack of an experimentally controlled environment in the seismology of the deep Earth, it is difficult to scrutinize competitive hypotheses. In Earth sciences, we often come closer to the truth through negative instances rather than the verification of existing hypotheses. A conglomerate of anisotropic domains in the IC combined with the inhomogeneous structure of the mantle is a likely concept or a working hypothesis that, inter alia, reconciles observed complexities in differential PKP travel times while preserving a net IC anisotropy that is required to explain the anomalous splitting of Earth's free oscillations. The conglomerate model is compatible with the observations of laterally- and radially-varying anisotropy and negates the concept of a strong uniform cylindrical anisotropy in the inner core. Varying and small PKP travel-time residuals from polar paths suggest weak average anisotropy, and means that the fast axis of anisotropy cannot be preserved over the entire volume of the IC. Columnar convection and convective heat flux in the outer core result in heat transfer variations, which is one of the mechanisms that influence IC growth and crystal alignment, and this has been suggested through the observation of variations in crystal alignment, texture, and modeling of randomly oriented anisotropic patches. With the current configuration of receivers and earthquakes worldwide, it is difficult to achieve a satisfactory sampling of the inner core, except for the paths nearly parallel to the equatorial plane. An exception is a dataset associated with the ray-paths from the South Atlantic earthquakes recorded at the northernmost stations, whose corresponding PKP differential travel times account for a large percentage of the most anomalous travel time data and a notion that the quasi-western hemisphere of the IC is more anisotropic than the quasi-eastern one (it is important to distinguish between the IC hemisphericity in absolute velocity that is well documented and the hemisphericity in anisotropy). PcP-P differential travel-time residuals from the South Atlantic region earthquakes are similar in range to that of the varying PKP(BC-DF) residuals (four seconds), yet entirely insensitive to Earth’s core structure. This observation strongly suggests that mantle structure can affect PKP travel time residuals from the South Atlantic earthquakes more than previously acknowledged. This emphasizes an inevitable trade-off role of the mantle in conjunction with the interpretation of IC structure. For example, if elastic anisotropy in the IC is weak or cancels out over different domains sampled by body waves, then some very anomalous travel times for those ray paths are a result of inhomogeneous or anisotropic structure outside the IC, such is probably the case, at least partly, for the earthquakes from the South Atlantic region. The role of inhomogeneities in the mantle must be recognized in any future seismological attempts of imaging structure of the IC.