Abstract
The dynamic topography of the core-mantle boundary (CMB) provides
important constraints on dynamic processes in the mantle and core.
However, inferences on CMB topography are complicated by the uneven
coverage of data with sensitivity to different length scales and strong
heterogeneity in the lower mantle. Particularly, a trade-off exists with
density variations, which ultimately drive mantle flow and are vital for
determining the origin of mantle structures. Here, I review existing
models of CMB topography and lower mantle density, focusing on
seismological constraints. I develop average models and vote maps with
the aim to find model consistencies and discuss what these may teach us
about lower mantle structure and dynamics. While most density models
image two areas of dense anomalies beneath Africa and the Pacific, their
exact location and relationship to seismic velocity structure differs
between studies. CMB topography strongly influences the retrieved
density structure, which helps to resolve differences between recent
studies based on Stoneley modes and tidal measurements. Current CMB
topography models vary both in pattern and amplitude and a discrepancy
exists between models based on body-wave and normal-mode data. As
existing models feature elevated topography below the Large-Low-Velocity
Provinces (LLVPs), very dense compositional anomalies may currently be
ruled out as possibility. To achieve a similar consistency as observed
in lower mantle models of S-wave and P-wave velocity, future studies
should combine multiple data sets to break existing trade-offs between
CMB topography and density. Important considerations in these studies
should be the choice of theoretical approximation and parameterisation.
Efforts to develop models of CMB topography consistent with body-wave,
normal-mode and geodetic data should be intensified, which will aid in
narrowing down possible explanations for the LLVPs and provide
additional insights into mantle dynamics.