Is the top of Earth's liquid core stably stratified?
In plain words
Some seismic and magnetic data suggest the outermost few hundred kilometers of the liquid core form a calm layer that does not churn. Other analyses find no such layer, and its presence changes how the field is generated and how fast the core cools.
Precise statement
Some SmKS seismic phases (waves reflected several times beneath the core-mantle boundary) show reduced P velocity in the top few hundred km of the outer core, and Buffett (Nature 2014, https://doi.org/10.1038/nature13122) interpreted a 60-yr geomagnetic fluctuation as magnetic-Archimedes-Coriolis waves requiring a stratified layer about $140\,\mathrm{km}$ thick; other studies fit the data without a layer. Determine whether a stable layer exists, its thickness and buoyancy frequency $N$, and whether its origin is thermal (subadiabatic heat flow) or chemical. An answer is thickness and $N$, or an upper bound, consistent with seismology, secular variation and core thermal history.
What would settle it
Joint inversion of mantle-corrected SmKS travel times and geomagnetic secular variation for wave modes that exist only in a stratified layer.