Mantle Convection: The Slow Engine Beneath Plate Tectonics

Solid rock that flows like syrup

Earth's mantle is solid rock, yet over geological time it behaves like an extremely viscous fluid -- roughly 10^21 Pa*s, about a trillion trillion times more viscous than water. Under sustained stress over millions of years, crystal defects migrate through the rock's lattice (solid-state creep) and the whole mass flows, slowly, without ever melting. Heat from the core-mantle boundary (around 3,500-4,000 kelvin) and from radioactive decay within the mantle itself makes the rock at the bottom hotter and therefore less dense than the rock above it. Less-dense material rises, cooler and denser material sinks, and the two together set up a slow, continuous overturning circulation -- mantle convection -- that takes tens to hundreds of millions of years to complete one cycle.

Rayleigh-Benard convection and the Rayleigh number

The canonical laboratory version of this is Rayleigh-Benard convection: a fluid layer heated from below and cooled from above. Whether convection happens at all -- as opposed to heat simply conducting through a static layer -- is decided by a single dimensionless number, the Rayleigh number Ra:

Ra = (g * alpha * deltaT * h^3) / (kappa * nu)

g        gravitational acceleration
alpha    thermal expansion coefficient
deltaT   temperature difference across the layer
h        layer thickness
kappa    thermal diffusivity
nu       kinematic viscosity

Ra compares the buoyancy force driving the overturn against the two things resisting it: viscosity (nu) damping the motion and thermal diffusion (kappa) smoothing away the temperature difference before it can drive much flow. Below a critical value (about 1,708 for idealised rigid boundaries) any perturbation decays and heat moves by conduction alone. Above it, convection cells spontaneously form and grow, and for Earth's mantle the estimated Ra is somewhere around 10^6 to 10^8 -- vastly supercritical, meaning the mantle convects vigorously rather than marginally.

The Boussinesq approximation

Modelling a compressible fluid whose density genuinely varies with depth and temperature is expensive. The standard simplification, the Boussinesq approximation, treats density as constant everywhere except in the buoyancy term of the momentum equation, where the small temperature-driven density variation is exactly what drives the flow. This captures the essential physics of thermal convection -- buoyancy is the engine, everything else is treated as incompressible -- while making the governing equations (continuity, momentum with a buoyancy source, and heat advection-diffusion) tractable to solve on a grid with a finite-difference or finite-volume method.

continuity:  div(u) = 0
momentum:    rho0 * (du/dt) = -grad(p) + nu*laplacian(u) + rho0*g*alpha*(T - T0)
heat:        dT/dt + u . grad(T) = kappa * laplacian(T)

From convection cells to plate tectonics

Mantle convection is not just a curiosity -- it is the engine behind plate tectonics. Upwelling limbs of the convection pattern correspond to mid-ocean ridges, where hot material rises and spreads, creating new oceanic crust; downwelling limbs correspond to subduction zones, where old, cooled, denser oceanic plates sink back into the mantle. The plates themselves are essentially the rigid, cold upper thermal boundary layer of the convecting system -- they are dragged along by, and in turn help drive, the same circulation. This is why simulations of mantle convection use viscosity and heat-flux sliders: real mantle viscosity varies by orders of magnitude with temperature and pressure, and that variation controls whether convection organises into a few large cells (a pattern closer to Earth's actual handful of major tectonic plates) or many small ones.

What the grid solver is actually doing

A 2-D finite-difference implementation discretises temperature and velocity on a grid, solves the momentum equation for the flow field given the current buoyancy distribution (often via a stream-function/vorticity formulation to automatically satisfy the incompressibility constraint), then advects and diffuses the temperature field forward one timestep using that velocity, and repeats. The stable timestep is limited by both the CFL condition on advection and a diffusive stability bound on the heat equation, which is why increasing the Rayleigh number in the demo -- more vigorous convection, faster flow -- typically forces smaller timesteps to stay numerically stable.

Frequently asked questions

Try it live

Everything above runs in your browser — open Mantle Convection and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.