DI24A-01 INVITED
New seismological attempts to study the top of the Earthfs core
The seismological structure at the top of the Earthfs core has been masked by the D", the base of the mantle, that is adjacent above the core. As increasing the high quality digital seismic data, the studies of the region have been revisited. First is the analysis of SmKS phases. Previously, the travel times of SKS, SKKS, and S3KS have been examined by using a regional array or an old global network of which distribution was sparse. Now I show that a new data set consisting of 1211 SmKS (m > 1) waveforms has been obtained from the recent permanent and temporary networks that exist between 1990 and 2003. The new data has been analyzed to investigate the radial seismic velocity structure around the core-mantle boundary (CMB). A stacked waveform at each distance bin coincides with reflectivity synthetic one for PREM very well, whereas those for other global models (iasp91, ak135, and SP6) yield disagreements. Furthermore, a waveform modeling for the D" structure results in a 30 km thick layer with a 10 percent S-wave velocity reduction at the mantle bottom as the best model while the SmKS modeling is insensitive to the lowermost mantle structures with thickness of several hundred kilometers. The possibility of a low P-wave velocity layer in the outermost core is remained because that the waveform fitness for the part of S4KS is improved by further introducing a 140 km thick layer with a 0.8 percent P-wave velocity reduction at the core top. However, a linear velocity gradient is assumed in the modeling of the outermost core. More complicated structure, such as the change of the velocity gradient, would be suffered from the trade-off between the velocity and the core radius. As discussed above, an independent approach is required to investigate to the core radius and topography of the CMB. Thus I have started another project. The combination of P4KP and PcP is suitable for canceling the hypocenter uncertainty and the regional variations in the mantle and the crust. To date, I have obtained 94 P4KP|PcP times from the International Monitoring System (IMS) arrays, the J-array and IRIS stations. The times of P4KP and PcP are carefully picked by hand. The picking points are similar to each other. The ray theoretical travel times of PcP and P4KP-AB are calculated with PREM as a reference. The resultant residuals obtained are scattered from +0 to +5 s. After correcting the travel times due to the ellipticity at the CMB for which the hydrostatic equilibrium are considered, the corrected P4KP|PcP are distributed around 2|3 s. Correction with a global P wave tomography yields a small change as large as 0.2 s. Therefore the P4KP|PcP residuals by 2 to 3 s should be explained by excess core radius by 2 to 3 km comparing to those of PREM if the velocity structure obtained by SmKS phases is adopted. Furthermore, scatterring of P4KP|PcP times are investigated by three CMB topography models (Morelli and Dziewonski, 1987: MD, Dorrnbos and Hilton, 1989: DH, and Sze and van der Hilst, 2003: SH). The correction using the DH model makes scattering of the P4KP|PcP residuals very small. This suggests that the P4KP|PcP data is useful to image the CMB topography if we have an enough data. Furthermore, a simultaneous inversion with SmKS would be important to elucidate the both side structures of the CMB.
DI24A-02
Thermal coupling between the mantle, outer core and inner core: an experimental model
Thermal coupling between the mantle and the outer core has been proposed based upon the correlation between the patterns of stationary geomagnetic field and the seismic heterogeneity of the lower mantle (Bloxham and Gubbins, 1987). We have studied how such thermal heterogeneity can affect the outer core flow, using laboratory experiments in a rapidly rotating hemispherical shells (Sumita and Olson, 1999, 2002). Here I review the results obtained from these experiments, and offer their implications to the Earth's core. We use a hemispherical shell with an outer diameter of 30 cm and spin it at 207 rpm to achieve an Ekman number of 4.7 × 10-6. By circulating a cooling water through the inner sphere, we impose a radial temperature gradient, and achieve a Rayleigh number of up to 44 times the critical value. For most Rayleigh numbers (Ra/Rac > 8), thermal convection consists of meandering plumes that originate from inner and outer boundaries and are advected westward by the mean zonal flow (Sumita and Olson, 2000). When we impose a thermal anomaly at the outer boundary using a strip heater we find that the warm fluid generated by the heater flows eastward. When Q\ast = \mbox{(Applied total heat flow)/(Total heat flow at ICB)} > 0.7, we find that a stationary front forms at the east of the heater which separates the warm eastward flow and cold westward flow. The stationary front take the form of a spiral and extends from the outer boundary towards the inner boundary, along which a jet flows towards the inner boundary. Simple estimate shows that the condition Q\ast > 1 can be satisified in the Earth's core. Since centrifugal force is used to simulate the radially dependent gravity, a heater in the experiment corresponds to a cold anomaly at the CMB. Seismic tomography suggest that such an anomaly exists beneath east Asia. Our experiments suggest that there is a cold eastward flow in the Pacific and a warm westward flow elsewhere which is consistent with the core flow model obtained from geomagnetic secular variation (Bloxham and Jackson, 1991). Our experiments also suggest that inner core growth rate is fast at the western hemisphere, which coincides with the region of large P-wave anisotropy of the inner core (Tanaka and Hamaguchi, 1997). According to the model of Yoshida et al. (1996), elastic strain energy of deformed crystals, which is the cause for preferred orientation, scales as \propto \mbox{(growth rate)}2, and thus explains the region of large anisotropy in the western hemisphere. Sumita and Olson, 1999, Science, 286, 1547-1549. ibid, 2002, J.Geophys. Res., 107, 10.1029/2001JB000548.
DI24A-03
Temporal Change of the Earth's Inner Core Property: Observations and Interpretations
Recent seismic observations from earthquake waveform doublets provided compelling evidence that the seismic waves associated with the Earth's inner core experience temporal change. In this presentation, we present seismic observations associated with the inner core phases from waveform doublets occurring in the South Sandwich Islands and the Tonga-Kermadec subduction zone, and discuss possible interpretations to the seismic observations. The PKiKP phases recorded for a doublet (1993/2003) occurring in the South Sandwich Islands and reflected off the inner core boundary beneath mid-Africa arrived about 30 to 70 milliseconds earlier for the 2003 event than for the 1993 event, while no temporal change of travel time is observed for the PKiKP phases sampling the inner core boundary beneath South America, the Atlantic Ocean, and the Indian Ocean. The PKIKP and PKiKP phases for waveform doublets occurring in the Tonga-Kermadec subduction zone, with the re- occurrence time greater than eight years and sampling the inner core beneath the Philippine Sea, China, central Asia, the Arctic Ocean, the Indian Ocean, Antarctica, and Alaska, exhibited no temporal change in travel time and waveform. The observed temporal changes of travel time for the PKiKP phases sampling the mid-Africa indicate that the Earth's inner core surface locally enlarged by 1 – 1.75 km between the occurrence of the 1993/2003 doublet. We discuss in detail possible interpretations to the observed/non-observed temporal changes of these core phases, including inner core differential rotation with the presence of irregular topography at the inner core boundary, and rapid localized growth of the inner core due to convection in the top of the inner core or solidification of the outer core. http://geophysics.geo.sunysb.edu/wen/
DI24A-04 INVITED
A whole Earth dynamo model
The observations of modern magnetic, archaeomagnetic and paleomagnetic field of the Earth suggest a close correlation of the geomagnetic field with the structure of the Earth's lower mantle, pointing to the processes of a whole Earth dynamo. We report our progress on constructing a three-dimensional, whole Earth dynamo model based on an EBE (element-by-element) finite element method taking the full advantage of modern massively parallel computers. The dynamo model, driven by thermal or compositional convection, has been carefully compared with the well-known benchmark dynamo using spectral methods, showing a satisfactory agreement between two fundamentally different models. The whole Earth dynamo model consists of the four different zones: an electrically conducting solid inner core, a convective fluid outer core, an electrically conducting solid mantle and an exterior to the Earth. It is demonstrated that the numerical dynamo model can be effectively parallelized with nearly linear scalability on massively parallel computers. Since the model is based on a three-dimensional element-by-element finite element method, the physical properties of the Earth's mantle can be readily taken into account in the model. When the electrically conducting lower mantle has a heterogeneous electric conductivity, we reveal that the dynamo has to be time-dependent and the amplitude of the generated magnetic field is determined by the relative phase between the core magnetic field and the electrically heterogeneous lower mantle.
DI24A-05
A New Thermal Equation of State for Iron at Megabar Pressure
Because the core is composed primarily of iron alloyed with light elements, the physics of iron must be well understood to assess the effect of impurities on the properties the real core material could be made of. The structure of iron at inner core conditions is still an open issue, since some theoretical calculations ( e.g. Belonoshko et al. Nature 424, 1032, 2003) or a recent experimental observation on a nickel-iron alloy (Dubrovinsky et al. Science 316, 1880, 2007) indicate that a body centered cubic one could be more appropriate. It is however still reasonable to envisage that iron adopts the hexagonal close packed structure (hcp) at Earth's inner core conditions. In this report, the high-pressure and high-temperature behavior of iron has been investigated to 140 GPa and 3400 K with in situ synchrotron X-ray diffraction at ESRF (Grenoble). We present series of experiments carried out on hot-pressed samples of iron and periclase in a laser-heated diamond-anvil cell, that were combined with ATEM examination of recovered samples. Among all structures proposed so far, hcp-iron only is found to be present at the highest pressure and temperature investigated. In addition, these experiments allow us to better bracket the triple point between the fcc and hcp phases and the liquid, and to address the evolution of the c/a ratio with pressure and temperature and the related elastic anisotropy of iron at core conditions. Finally, we propose a new thermal equation of state for iron at megabar pressures, which put tighter constraints on the light element content of the inner core.
DI24A-06 INVITED
An Experimental Dynamo in a Highly Turbulent Flow.
The dynamo effect is a magnetohydrodynamics instability, allowing the existence of a self-sustaining magnetic field in a flow of an electrically conductive fluid. This effect is considered to be at the origin of the magnetic field of the Earth and the stars. Until now, it had only been evidenced experimentally in cases where internal boundaries guide the flow. We will present the first observation of a dynamo in a homogeneous flow of liquid sodium (VKS2 experiment), where the turbulent fluctuations are of the same order of magnitude as the mean flow. The behaviour of this instability close to the threshold will be described, as well as scaling laws for its saturation. One striking feature of such a dynamo, that makes it very different from the previous constrained dynamos, is that it displays a large variety of dynamical regimes, including chaotic reversals, strongly reminiscent of the observed reversals of the Earth's magnetic field.
DI24A-07
PKiKP Coda Observations Interpreted in Terms of the Earth's Inner Core Heterogeneities
Although resulting from gradual few billion years long crystallization process, Earth's inner core (IC) is rather heterogeneous than a single crystal of iron. IC fabric variations such as changes in crystallographic alignment or supposed increase in crystal size with depth often constitute physical grounds for hypothesizing IC heterogeneities of various scale lengths. To constrain upper IC heterogeneities, we analyze reflections from the Earth's inner core boundary and the coda waves following the reflections (PKiKP) on array records of underground nuclear explosions. In particular, we compare PKiKP codas observed after reflections with close bounce points on the surface of the IC and diverse ray paths in the Earth's crust, mantle and outer core. Such PKiKP coda doublets show similar shape, frequency content, intensity and duration, and feature powerful arrivals originating from IC heterogeneities. The proposed interpretation favors misaligned anisotropic iron crystals up to 10 km in size to cause the observed PKiKP codas rather than classical scattering on inclusions in the outermost IC. Also, the IC fabric's image observed as PKiKP coda appears to be time stationary in 20 years span, failing to exhibit any pronounced effect of IC differential rotation.
DI24A-08
Topography of inner core boundary
Precise determination of the topography of a major internal boundary of the Earth is difficult because of the trade-off with the unknown velocity structure above it. However, the discoveries of the inner core (IC) rotation and high-quality teleseismic waveform doublets make the precise mapping of the topography of the inner core boundary (ICB) possible, as demonstrated in recent studies. Here we examine IC refracted (PKP-DF) and reflected (PKP-CD) waves recorded at the Yellowknife Array and global stations from 13 high-quality doublets, among a large collection of doublets in S. Sandwich Islands that we have assembled. Our results show clear evidence for spatial and temporal variations of IC reflections in travel times and in waveforms. If the time separation (dT) between the two members of the doublet is less than 3 years, the IC arrivals show little temporal change in travel times or waveforms. If dT is greater than about 6 years, some doublets show large variations but some others do not. The ICB regions beneath Atlantic Ocean and Indian Ocean show little temporal change. The regions show large variations are beneath Africa and the Central America, which coincide with large seismic anomalies at the core-mantle boundary (CMB). Inside these two ICB regions, there are fine-scale (km scale) variations. The largest temporal changes of IC reflections are about 0.10 to 0.15 s, corresponding to a topographic variation of up to 3.7 to 5.6 km. The results suggest ICB topography of a few kms on fine to regional scales. Dynamical models include a bumpy ICB rotating with the IC itself or a transient slurry boundary sloshing about in the turbulence at the base of the convecting outer core. The geographical coincidence of the ICB and CMB anomalies may suggest strong thermal coupling of the mantle and the core.