Inner core sensitive modes exhibit a spread of singlet frequencies far exceeding that which can be explained by plausible isotropic heterogeneity. This large deviation from sphericity contributes to non-uniqueness of the non-linear inversion for individual mode splitting functions. To alleviate this non-uniqueness, we have implimented a direct and simultaneous inversion of spectra from 25 modes for inner core structure.
We have recently submitted the results of our inversions for inner core anisotropy combining normal modes with differential PKP travel times, which provide sensitivity to the center of the core [Durek and Romanowicz, 1998]. While several authors have demonstrated that tranverse isotropy can explain the character of the splitting, to obtain consistency between mode spectra, travel times, and zonal splitting coefficients requires additional structural complexity. Our final model, based on an axisymmetric expansion of the general elastic tensor (Figure 24.1), shows that P-waves traveling parallel to the rotation axis encounter an elongated region of high anisotropy in the center of the core. This is consistent with earlier results of this study [Romanowicz et al., 1996] based on the linear inversion of splitting functions for inner core anisotropy. We also demonstrate that the success of the direct inversion requires a very accurate spherical model. In addition, the optimal inner core model is obtained using a mantle model presented in this study based on inversion of 10 mantle sensitive normal modes.
Durek, J.J., and B. Romanowicz, Inner core anisotropy Inferred from normal mode spectra (abstract), EOS Trans. AGU, 78, 1997.
Durek, J.J., and B. Romanowicz, Inner core anisotropy inferred by direct inversion of normal mode spectra, submitted to Geophys. J. Int., 1998.
Romanowicz, B., X-D. Li, and J.J. Durek,Anisotropy in the inner core; could it be dueto low-order convection?,Science, 274, 963-966, 1996.