Landscape of the core-mantle boundary
Seismological models of the mantle are routinely developed using a range of techniquaes applied to different data types. While models of S- and P-wave velocity show a large degree of consistency (see e.g. the SubMachine website), controversy remains regarding models of lower mantle density and core-mantle boundary (CMB) topography, which are vital for determining the nature of mantle structures.
I reviewed existing models of CMB topography and lower mantle density, focusing on seismological models (Koelemeijer (2020)). These models were used to develop average models (similar to SMEAN or PMEAN) as well as vote maps (see Shephard et al. (2017)). In addition, combined vote maps were computed (combination of a positive and negative mean vote map). While these aid in finding model consistencies, the average models and vote maps do not necessarily fit seismological data and are not intended as replacement of the original models.
DENSITY: Most density models consistently image two areas of dense anomalies beneath South Africa and the North Pacific, though their exact location and relationship to seismic velocity differs. CMB topography strongly influences the retrieved density structure (model KDR2017-pos vs KDR2017-neg), which helps to resolve differences between recent studies based on Stoneley modes and tidal data, particularly for degree 2 structure.
CMB TOPOGRAPHY: Average models and vote maps do not agree for CMB topography, indicating that particular models dominate results. A disparity (evident as low overall vote) also exists between models based on body-wave and normal-mode data, which needs to be resolved in future. Most existing models feature elevated topography in the South Pacific and Central Africa below the LLSVPs, thus likely ruling out strongly thermochemical structures (heavy piles).
| Download the average models and vote maps here: | |
|---|---|
| I provide a ZIP file with figures and input files for all parameters (seismic velocities, density, CMB topography) or alternatively ZIP files related to a particular parameter can be downloaded. In each case, a readme file is provided that explains the model selection and file convention. | |
| landscape_all.zip: Figures and input files for all parameters together, with separate files for different model selections and spherical harmonic degrees. | |
| landscape_velocities.zip: Files relating to S-wave and P-wave velocities, up to degree 6 or 20. | |
| landscape_density.zip: Files relating to different density model selections, up to degree 2 or 6. | |
| landscape_topography.zip: Files relating to different CMB topography model selections, up to degree 2 or 6. | |
| If you are using these average models or vote maps, please reference: | |
| Koelemeijer, P. (2020). Towards consistent seismological models of the core-mantle boundary landscape. In press in AGU monograph: "Mantle Convection and Surface Expressions", edited by Marquardt, Ballmer, Cottaar & Konter, doi:10.1002/9781119528609.ch9. |
Average models and vote maps of density (top) and CMB topography (bottom) compiled by Koelemeijer (2020) based on existing models from the literature. For density, separate versions are computed for the two model classes of Koelemeijer et al. 2017, while for CMB topography versions are given for models based on body wave or normal mode data.