S. Asari, H. Shimizu and H. Utada
Earthquake Research Institute, University of Tokyo, Tokyo, Japan.
Topographic coupling is one of the possible mechanisms for dynamical coupling between the core and the mantle. The topographic torque acting on the mantle from the core can be calculated by integrating fluid pressure on the aspherical CMB (e.g. Roberts 1988). In the past studies, models of CMB topography and core surface flow were employed to derive the torque under the assumption of tangential geostrophy at the top of the fluid core. It was found that axial component is larger by an order of magnitude than the required value for decadal fluctuation of LOD (Jault and Le Mou\"el 1989). Equatorial component was estimated to be smaller by several factors of magnitude than the torque for decadal polar motion if it is to be explained by core-mantle interaction (Hide et al. 1996). Mapping of core surface flow by applying diffusionless induction equation at the CMB is subject to unavoidable non-uniqueness. Various flows can reproduce observed geomagnetic secular variation with acceptable level of misfit. It is highly anticipated that the net torque associated with different flows can take wide range of values for certain CMB topography. Our aim of this study is to examine the possible strength of topographic torque, and to seek the possibility to use the torque as additional information to obtain core surface flow. In order to discuss above issues, we focused on the magnitude of the topographic torque at a particular epoch. We calculated horizontal flow at core surface under frozen-flux approximation by using the geomagnetic field model DGRF80. Tangential geostrophy was assumed to reduce the non-uniqueness problem and to obtain the pressure from calculated flow. Two kinds of damping (kinetic energy and smoothness of the flow) were imposed when the flow map with truncation degree 8 was estimated. Hyper parameters for damping were widely changed to examine their effect on the flow and subsequent pressure distribution. In this study, recent CMB topography model by Boschi and Dziewonski (2000) which includes the core ellipticity was adopted. The amplitude of this topography model is comparable with the bulge of the core flattening, which is quite large compared with other topography models. Three components of topographic torque were calculated following the equations given by Greff-Lefftz and Legros (1995). The calculated torque changes continuously as the damping parameters for the flow are varied. Axial component tends to be larger by an order of magnitude than what is necessary for LOD fluctuation. However, it turned out that its sign can be changed depending on the damping parameters. It is possible that the axial torque takes appropriate value for LOD observation. Equatorial components are typically larger by an order of magnitude than axial component due to eminent pressure deviation around 50 to 70 degrees in latitude that acts on the flattening of the core. Variability of the torque is also larger by an order of magnitude for equatorial components than that for axial component. This seems to be caused by the large uncertainty of flow around the region beneath southern Africa where the magnitude of radial component of the magnetic field is small. The RMS misfit to the secular variation model is kept below its error, when such variation of the torque is created. It seems that observations of Earth's rotation can be used as additional data to determine core surface flow under the hypothesis that angular momentum is transferred between the core and the mantle by topographic coupling.