| abstract |
 |
 |
|
An entirely new dynamic picture of Earth's deep interior is emerging that is primarily the result of significant advances in high pressure experimentation and theoretical mineral physics, but which also brings together decades of progress in seismic imaging, mass spectrometry, and geomagnetic research. Zeroeth-order questions such as the depth extent of subduction of oceanic lithosphere and the location of large chemically distinct deep reservoirs have now been solved. Rapid progress is now being made in understanding the formation of the Earth, the rate of heat and mass flow from the core, the composition and origin of chemically distinct reservoirs, the dynamics of convection in the core which generate Earth's magnetic field, and thermal evolution of Earth's interior. While the primary sources of uncertainty still remain with physical and chemical properties of materials at high pressures and temperatures, we are now able to identify the most critical missing links, and have spawned dramatic new integrated research efforts spanning the globe to address these remaining problems.
The emerging picture is astonishing, and contains more surprises than could possibly have been imagined only a decade ago. First, temperature variations are being revealed in the deepest mantle in unprecedented (and unexpected) detail via the discovery of a post-perovskite phase transition and its association with remarkable seismic structures in the D" region. Perovskite->Post-perovskite isn't like your grandfather's phase transitions, instead it is embedded in an active thermal boundary layer above the core (where the mantle cools the core, giving rise to core convection and a geodynamo), and the high pressure phase can appear and disappear in surprising ways, forming lens-like structures floating above the CMB. The post-perovskite evidence points to high core-mantle heat flow, and therefore to high core temperatures in the past. Second, theoretical and experimental work on the physics of molten rocks at pressures of the deep mantle now provide ample evidence for a cross-over in liquid-solid density at some pressure in the mid-lower mantle (~80 GPa), such that melts in the deepest mantle will settle downwards instead of rising upward. Hints that melt might be stable at the base of the mantle were first discovered over a decade ago by seismologists probing the core-mantle boundary, and the finding of ultralow-velocity zones that seemed to bear the hallmarks of a partially molten mush. The inevitable conclusion of these findings is that the deepest mantle must have been more extensively molten in the Earth's hotter past, once forming a vast gravitationally stable basal magma ocean that formed very early in Earth's history. This synthesis is motivating an entire re-appraisal of the evolution of Earth's deep interior, and involves gradual cooling and fractional crystallization of this giant basal magma chamber, with solid cumulates forming chemically distinct regions in the deep mantle (formerly known as "super-plumes") and ultralow-velocity zones containing the residual liquid. This molten body reacted with the metallic core to produce a buoyant slag which influences the geodynamo and the observed geomagnetic secular variation. |
