High-resolution tomography of CMB and lowermost mantle coupled by geodynamics

Despite the fast advances of seismic tomography in the last decades provided us with very clear and reliable images of the Earth’s mantle,seismically and/or geodynamically inferred models of core-mantle boundary topography are still poorly correlated both in pattern and amplitude. A major cause for these discrepancies is the difficulty to separate, in travel-time anomalies, the contribution of CMB topography from that of lowermost-mantle and D" heterogeneities. As an attempt to reconcile the contrasting views of the Earth's CMB, we propose an innovative approach to mapping CMB topography from seismic travel-time inversions: instead of treating mantle velocity and CMB topography as independent parameters, as has been done so far (e.g., Soldati et al., 2003), we plan to account for their coupling by mantle flow, as formulated by e.g. Forte & Peltier (1991). In practice, we shall invert direct P waves, and core-sensitive phases, with coefficients of mantle (and, possibly, core) velocity structure as the only free parameters. CMB undulations will not be treated as free parameters, but accounted for via a modification of the tomographic matrix based on the estimated physical relationship between them and mantle velocities. For the first time, the resulting tomographic maps of CMB topography will be, by construction, physically sound, while explaining the inverted seismic data.