Effects of global core-mantle boundary topography on outer-core convection and topographic torques

This study uses direct numerical simulations to demonstrate that global core-mantle boundary topography drives a new instability and significantly enhances outer-core convection and topographic torques, providing a mechanism consistent with observed decadal variations in the length of day.

The Gist

Imagine the Earth as a giant, spinning marble. Inside, there is a liquid iron core that swirls and churns, generating our planet's magnetic field. Surrounding this liquid core is the solid mantle, which acts like a thick, rocky shell.

Usually, scientists imagine the boundary between this liquid core and the rocky shell (called the Core-Mantle Boundary, or CMB) to be a perfectly smooth, round sphere. But this paper argues that the boundary is actually bumpy and uneven, much like the surface of a potato rather than a billiard ball. These bumps are caused by giant structures deep within the rocky mantle, some of which are thousands of kilometers wide.

The researchers used powerful supercomputers to simulate what happens when this "liquid core" swirls against a "bumpy shell." Here is what they found, explained simply:

1. The "Smooth Slide" vs. The "Bumpy Track"

In a perfectly smooth sphere, the liquid inside wants to flow in neat, circular rings around the Earth's spin axis. It's like a skater spinning on a perfectly smooth ice rink; they can glide effortlessly in a circle.

However, when the boundary is bumpy, it's like putting a series of speed bumps or hills on that ice rink. The liquid flow gets forced to change direction to go over or around these bumps. The researchers found that these bumps actually help the liquid move faster and transport heat more efficiently. It's as if the bumps act like a catalyst, giving the liquid a "push" that it wouldn't get on a smooth surface. In their simulations, these bumps increased the speed of the flow and the amount of heat moving from the center to the edge by up to 100%.

2. The "New Instability" (The Subcritical Surprise)

There is a rule in physics that says liquid convection (like boiling water) only starts when the heat gets hot enough to overcome the fluid's resistance. The researchers discovered something surprising: the bumps on the boundary can break this rule.

Even when the core isn't hot enough to start moving on its own, the bumps can create a new kind of instability that gets the liquid moving anyway. Think of it like a ball sitting in a deep valley; normally, it needs a big push to get out. But if the valley has a weird, bumpy shape, a tiny nudge might be enough to get the ball rolling. This means the Earth's core might be churning and generating its magnetic field even when it's "cooler" than we previously thought.

3. The "Torque" (The Wobbly Top)

The Earth spins like a top. Sometimes, the length of our day changes by tiny fractions of a second (milliseconds) over periods of 6 to 60 years. Scientists have long suspected that the interaction between the spinning liquid core and the solid mantle is responsible for these tiny wobbles.

The researchers calculated the "torque" (the twisting force) that the liquid core exerts on the bumpy boundary. They found that the bumps create a significant twisting force.

• The Analogy: Imagine pushing a spinning merry-go-round. If you push it on a smooth edge, it's hard to change its speed. But if you push against a bumpy, uneven edge, you can grab onto the bumps and twist the whole thing much more effectively.
• The Result: Their calculations show that the twisting force generated by these bumps is strong enough to explain the observed changes in the length of our day.

4. The "Locking" Effect

One of the most interesting findings was about how the liquid flow interacts with specific shapes of bumps.

• The Analogy: Imagine a dancer trying to move to music. If the music (the flow) and the dance floor pattern (the bumps) match perfectly, the dancer might get "locked" into a specific rhythm.
• The Result: When the bumps had a specific size and shape that matched the natural rhythm of the liquid flow, the flow would "lock" onto the bumps. While this made the flow very organized, it actually reduced the twisting force (torque) because the liquid wasn't fighting against the bumps anymore; it was just riding along with them. This suggests that the shape of the bumps matters just as much as their size.

Summary

This paper uses computer models to show that the "bumpy" boundary between Earth's liquid core and solid mantle is not just a passive wall. It is an active participant that:

1. Speeds up the liquid flow and heat transfer.
2. Starts the flow even when it's too cool to move on its own.
3. Twists the Earth's rotation, explaining why our days get slightly longer or shorter over decades.

The study confirms that to understand how Earth's magnetic field works and why our days change length, we cannot treat the core as a smooth, perfect sphere; we must account for the rough, bumpy reality of the boundary.

Technical Summary

Technical Summary: Effects of Global Core-Mantle Boundary Topography on Outer-Core Convection and Topographic Torques

Problem Statement
The boundary between the Earth's liquid outer core and the solid mantle (the Core-Mantle Boundary, or CMB) is not perfectly spherical or smooth. Seismic and geodynamic studies suggest the presence of topographic heterogeneities, potentially linked to Large Low Shear Velocity Provinces (LLSVPs) in the deep mantle. While it is hypothesized that CMB topography couples the core and mantle—potentially driving observed decadal and subdecadal fluctuations in the length of day ($\Delta$LOD) and influencing the geomagnetic field—the specific dynamical mechanisms remain poorly understood. Previous investigations were limited to linear regimes, simplified two-dimensional models, or isolated topographic features, failing to capture the effects of globally distributed, spectrally broad topography on fully turbulent convection.

Methodology
The authors employed direct numerical simulations (DNS) of rotating shell convection to investigate these dynamical effects. The study utilized the spectral element code Nek5000 to solve the non-dimensional Boussinesq equations for a fluid in a spherical shell with aspect ratio $\chi = 0.35$.

• Geometry: The inner boundary was spherical, while the outer boundary was deformed according to $r_{cmb} = r_o - \epsilon h(\theta, \phi)$, where $\epsilon$ is the dimensionless topographic amplitude and $h$ is the shape function.
• Parameters: Simulations were conducted at an Ekman number $Ek = 10^{-6}$ (and a subset at $10^{-4}$) and a reduced Rayleigh number $\widetilde{Ra} = 40$ for the turbulent regime, with a Prandtl number $Pr = 1$.
• Topographic Models: Five distinct topographic shapes were tested:
  1. Single spherical harmonics: $h_{2,1}$, $h_{8,4}$, $h_{32,16}$, and $h_{32,17}$.
  2. A "tomographic model" ($h_t$) derived from seismic cluster analysis maps of LLSVPs, smoothed to represent continuous intrusions.
• Analysis: The study examined heat transport (Nusselt number, $Nu$), momentum transport (Reynolds number, $Re$), flow morphology (kinetic energy spectra, geostrophic contours), and the resulting axial topographic torques ($\Gamma_z$).

Key Contributions and Results

1. Confirmation of Topographic Instability in Global Geometry:
   The study confirms the existence of a linear instability driven by topography at Rayleigh numbers below the critical onset for a spherical shell ($Ra < Ra_c$), a phenomenon previously predicted by Bell & Soward (1996) in a 2D annulus model. While the instability mechanism (buoyancy driving flow along deformed geostrophic contours) is confirmed, the quantitative scaling of flow speeds differs from the annulus theory due to geometric differences in the background temperature profile and the nature of the geostrophic contours in a sphere.

2. Enhancement of Transport via Geostrophic Contour Deformation:
   The shape of the geostrophic contours (lines of constant axial height) is central to the dynamics. In a perfect sphere, these are circles, and buoyancy cannot perform work on the geostrophic flow. Topography deforms these contours, allowing buoyancy to do work on the time-averaged flow.

   • Transport Increases: For topographies with significant equatorial symmetry (e.g., $h_{8,4}$, $h_{32,16}$), the deformation allows buoyancy to drive the geostrophic flow directly. This results in increases in both Reynolds and Nusselt numbers of up to $\sim 100\%$ relative to the spherical control case.
   • Symmetry Dependence: Topographies with odd $\ell-m$ (e.g., $h_{2,1}$, $h_{32,17}$) exhibit equatorial anti-symmetry, leading to smaller amplitude deformations of the geostrophic contours ($O(\epsilon^2)$) compared to even $\ell-m$ cases ($O(\epsilon)$). Consequently, the enhancement of geostrophic flow is suppressed in anti-symmetric cases, though transport can still increase via resonance with convective scales.

3. Topographic Torque Scaling and Flow Locking:
   The study investigates the axial torque exerted by the core on the CMB topography.

   • Scaling Law: The results confirm and extend the scaling law $\Gamma'_z \sim \epsilon Re^2$, where the torque scales linearly with topographic amplitude and quadratically with flow speed.
   • Flow Locking: For the $h_{32,16}$ case, where the topographic wavelength is close to the critical convective wavelength, the flow exhibits "locking" to the topographic pattern at high amplitudes. This causes the out-of-phase pressure component (which generates torque) to decrease, leading to a saturation of the torque magnitude despite increasing flow speeds.
   • Generalized Theory: The authors propose a procedure to estimate torques for spectrally broad topography (like the tomographic model) by convolving the topographic spectrum (assumed to follow Kaula's rule, $\eta_{\ell,m} \sim \ell^{-2}$) with the fluctuating pressure spectrum. This yields torque estimates consistent with the magnitude required to drive observed $\Delta$LOD variations.

4. Viscous vs. Topographic Torques:
   The ratio of viscous to topographic torque fluctuations ($\gamma$) was found to be less than unity for most parameters, supporting the conclusion that topographic torques, rather than viscous coupling, are the dominant mechanism for $\Delta$LOD variations on decadal timescales.

Significance and Claims
The paper claims four primary findings:

1. It validates the Bell & Soward (1996) instability mechanism within a fully three-dimensional, global spherical geometry.
2. It demonstrates that global topography deforms geostrophic contours, enabling buoyancy to perform work on the geostrophic flow, thereby significantly enhancing heat and momentum transport (up to 100% increases).
3. It shows that large-amplitude topography can override the canonical viscous $Ek^{1/3}$ paradigm for the onset of motion and time-averaged geostrophic flow, although small-scale turbulence retains its viscous fingerprint.
4. It establishes the generality of the dynamical torque scaling $\Gamma'_z \sim \epsilon Re^2$ across diverse global topographic shapes. By developing a generalized scaling theory for spectrally broad topography, the authors estimate that extrapolated torques reach magnitudes of $\sim 10^{18}$ N m, which is consistent with the torque required to drive observed decadal and subdecadal $\Delta$LOD variations.

The authors note that while their results provide a robust dynamical framework, the specific structure of CMB topography remains poorly constrained by seismology. Future work is identified as necessary to incorporate magnetic fields (magnetohydrodynamics), refine the topographic models with smoother profiles, and explore the transition to regimes where topographic scales fully replace viscous scales at Earth's core conditions.