An Alberta model traces decadal day-length swings to the inner core's gravitational pull

Length-of-day fluctuations of several milliseconds occur on timescales of 10-70 years, as shown between 1964 and 2019 in the paper's Fig. 1a.

Proposed mechanisms for the core-mantle angular momentum exchange include electromagnetic and topographic coupling at the core-mantle boundary and a gravitational torque by the inner core, and the precise nature of the torque has remained unknown.

The seismically reconstructed differential rotation of the inner core and core flows derived from magnetic field changes suggest that multidecadal length-of-day changes are driven primarily by the gravitational torque and resisted by electromagnetic and topographic torques, consistent with results from Earth-like dynamo models.

The reconstructed torque histories support a lowermost mantle featuring near-neutrally buoyant thermochemical piles, a post-perovskite phase with a low viscosity, and a highly conducting iron-enriched layer a few kilometres thick at its base.

Geophysicists measure changes in the length of day by observing Earth's position relative to distant cosmic objects.

The results also suggest a low-viscosity inner core deforming in only a few years, and yield an upper limit on the stratification at the top of the fluid core.

On a time average, the topography of the inner core boundary should coincide, except for a small eastward offset, with the gravitational potential imposed by mantle density anomalies.

Time-dependent azimuthal core flows exert electromagnetic stresses at the inner core boundary that rotate the inner core back and forth out of longitudinal alignment, causing a fluctuating gravitational torque on the mantle.

Flows near the top of the fluid core can be reconstructed from the observed secular variation of the geomagnetic field, and predictions of length-of-day change based on the angular momentum carried by these flows match the observed changes well, confirming their core origin.

Mathieu Dumberry, a geophysicist at the University of Alberta in Edmonton, Canada, and his colleague Huifeng Zhang, also at the University of Alberta, developed the model, reported in Nature.

A key input was research published in 2023 that used seismic waves from earthquakes to measure the inner core's rate of rotation relative to the rest of the planet since the 1960s; it found the inner core spun slightly faster relative to the rest of the planet until around 2010, when it began rotating more slowly.

Dumberry explained that the inner core 'wants to be aligned' with the mantle because its dense regions are attracted to the mantle's dense regions, but over time it gets pulled out of alignment by the generally westward flows of the outer core before being slowly pulled back eastward into position by gravity.

Earth's core is a Mars-sized ball at the centre of the planet made up mostly of iron; the inner core is solid and can rotate independently within a molten, flowing outer core.

The reconstructed torque histories are tied to the accuracy of the inner core rotation and core flow models.

Core flows exert only a weak viscous stress on the core-mantle boundary, too small to explain the required torque.

If a thin layer of iron-enriched material several kilometres thick, with a conductance of about 10^8 S, is present at the base of the mantle, as suggested by studies of Earth's nutations, electromagnetic coupling can produce a torque of the required magnitude.

The match between predicted electromagnetic torque and the observed decadal length-of-day record is generally poor, unless core flows are specifically designed to match the required torque.

Topographic torque predictions based on geostrophic pressure reconstructed from core flows and seismically derived maps of core-mantle boundary topography produce length-of-day changes typically 100 times too large, are highly sensitive to small changes in topography and core flows, and the validity of the approach has been questioned.

No length-of-day prediction based on the topographic torque arising from the deflection of core flows over core-mantle boundary bumps has yet been presented.

The length-of-day series shown spans 55 years, which is shorter than the 70-year upper end of the 10-70 year band of fluctuations being explained.

About nine years of the 1964-2019 length-of-day window fall after the roughly 2010 reversal in the inner core's rotation relative to the rest of the planet.