Deep Earth Structure – The Earth’s Cores

Abstract

The Earth’s core is made of a fluid outer core, with radius 3480 km, and a solid inner core, with radius ∼1220 km, which results from the solidification of the liquid core during the cooling of the Earth. Traveltimes and amplitudes of waves propagating through the core, and normal modes having energy deep inside the Earth, are the most powerful tools for investigating core structure. Important information is also provided by Earth dynamics, high-pressure physics, magnetic field analyses, and geochemical constraints. From the seismological point of view, the liquid core is homogeneous and nonstratified, except perhaps close to its boundaries. The smooth increase in P-velocity and density with depth reflects the increase in pressure and temperature. Its density, lower than the density of pure iron, indicates the presence of light elements (∼10%) in the liquid core iron alloy. The liquid core has a very high quality factor (almost no wave attenuation). Its surface, the core–mantle boundary, is a first-order discontinuity. Its topography is small, less than ∼4 km, and present at long wavelengths only, but it cannot be mapped safely because of the perturbing effect of the strong heterogeneities in the D″ layer at the base of the mantle. The lowermost 100 km at the base of the liquid core exhibit a low P-velocity gradient, likely due to light elements released during inner core freezing. The inner core boundary (ICB) appears as a first-order discontinuity. It corresponds to a small P-velocity and density increase with respect to the liquid core, which indicates a depletion in light elements compared to the liquid core. The rigidity of the inner core is well established from normal mode analyses. Body-wave analyses show that the P- and S-velocities and the quality factor may exhibit a strong gradient in the uppermost 100 km of the inner core, the lower values at ICB being ascribed to the possible presence of a mush of iron alloy. No topography could be detected at theICB. The main property of the inner core is its anisotropy, evidenced from both P-wave propagation and normal modes. P-wave velocity is 1–3% higher in the direction parallel to the Earth’s rotation axis than in the direction parallel to the equatorial plane. Depth variations are observed in anisotropy: the top 100–150 km are nearly isotropic, whereas the lowermost 300–500 km exhibit different characteristics of anisotropy, which have still to be specified. An anisotropy in attenuation is also present. The preferred explanation of velocity anisotropy invokes oriented anisotropic iron crystals but other mechanisms, such as oriented fluid inclusions or scatterers, are also possible. A hemispherical pattern is observed in the inner core anisotropy, the Western Hemisphere being more anisotropic than the Eastern Hemisphere, denoting a variation in the degree of crystal orientation, or a thicker isotropic layer beneath the Eastern Hemisphere. A correlated hemispherical variation is also observed in the mean velocity and attenuation at the very top of the inner core, suggesting that the thermal or chemical conditions vary at ICB. A possible faster rotation of the inner core with respect to the mantle is predicted by some geodynamo models, but forbidden by gravimetric coupling. A search for this differential rotation has been attempted by using the analysis of several long series of seismic data sampling specific paths, or, at worldwide scale, by using either body waves or normal modes. The results vary strongly from one study to another, the highest values (0.3–1.1 °yr−1) being obtained for the highly heterogeneous path from South Sandwich Island to Alaska. Worldwide methods come to lower rotation rates (≤0.2 °yr−1), and generally allow for a null rotation rate. Future research will focus on the depth variations of both P- and S-velocities and anisotropy in the inner core, and on the properties of the core interfaces (core–mantle boundary and ICB), which may provide important information to better understand the Earth’s differentiation process. The hemispherical pattern and a possible differential rotation are other intriguing features to elucidate. Progress in our knowledge of the core will however depend on the quality of the Earth sampling by seismic data, which relies heavily on the deployment of observatories in poorly sampled regions, in particular, oceans and high-latitudecountries.