At the base of the mantle, strong lateral variations in S velocity structure and anisotropy have been documented, in recent years, at the borders of the two large superplumes in the central Pacific and under Africa. These features undoubtedly hold important clues for understanding the dynamics of the earth's mantle. Current tomographic models of S velocity represents good starting models to fit travel times and some aspects of S and ScS waveforms along selected profiles sampling these regions (e.g. Bréger and Romanowicz, 1998; Breger et al., 2001). Increasing the strength of heterogeneity while keeping the original distribution of velocity highs and lows goes a long way towards explaining the observed trends.

However, these results point to the complex 3D nature of this heterogeneity, appropriate forward modeling tools need to be developed to handle strong 3D heterogeneity in this region, at relatively short periods and including diffracted waves. The coupled mode/spectral element method developed by Capadeville et al.(2001) and adapted to a heterogeneous layer (i.g. D'') between two sphericaly symmetric shells affords both numerical accuracy and the computational efficiency for this purpose.

Our goal is to extend the forward modeling approach of D'' structure ( Bréger and Romanowicz, 1998 ), so far limited to travel times, to also include waveforms, applying the hybrid spectral element method and further explore 3D structure and anisotropy at the base of the mantle.

First, we will collect the waveforms of core sensitive phases (Sdiff,SKS,SKKS,ScS) and measure differential travel times within them. Then, starting from the tomographic model, we will adjust the strength and boundaries of the velocity structure to match differential travel times between Sdiff -SKS,Sdiff-SKKS and ScS-S . The advantage of this approach is that while it uses an infinite frequency approximation and therefore inaccurate to model broadband S travel times, it is very expedient and allows to rapidly get closer to an adequate final model. Finally, we will refine the model by computing numerical synthetics in 3D.

Fig 36.1 shows the preliminary result of the comparison between synthetics from 1D, 3D model and observations. 1D synthetics are calculated from PREM (Dziewonski and Anderson, 1981) by normal mode summation. 3D synthetics are calculated from the model which has 3D structure of SAW24b16 (Megnin and Romanowicz, 2000) at the bottom 370 km of the mantle and sandwiched by PREM for the other part of the Earth. It is calculated by hybrid spectral element and modal summation method. Sdiff and sSdiff phase are shown. 3D model fits better to the observation especially for stations with longer distance (KBS,BOSA,LMN). For some stations, time shift for 3D is too strong (INK,YKW3) and sometimes both amplitude and time shift are poorly explained (CCM).

We thank the National Science Foundation for support of this research.

Bréger, L. and B. Romanowicz, Thermal and Chemical 3D heterogeneity in D", *Science, 282,* 718-720, 1998

Bréger, L., B. Romanowicz and C. Ng, The Pacific plume as seen by S,ScS and SKS, *Geophys. Res. Lett., 28
,* 859-1862, 2001.

Capdeville Y., E. Chaljub, J.P. Vilotte and J.P. Montagner, Coupling the Spectral Element Method with a modal solution for Elastic Wave Propagation in Global Earth Models, **accepted** in *Geophys. J. Int,* 2001

Dziewonski A.M. And D.L. Anderson, Preliminary reference Earth model, *Phys. Earth Planet. Inter., 25,* 297-356, 1981.

Megnin, C. and B. Romanowicz, The shear velocity structure of the mantle from the inversion of of body, surface and higher modes waveforms, *Geophys. J. Int, 143,* 709-728, 2000

Berkeley Seismological Laboratory

215 McCone Hall, UC Berkeley, Berkeley, CA 94 720-4760

Questions or comments? Send e-mail: www@seismo.berkeley.edu

© 2002, The Regents of the University of California.