For over a decade, radar surveys of the Greenland ice sheet have revealed giant swirling plume-like structures deep inside the ice. The structures were puzzling — organized, large-scale, and persistent. No existing model of ice flow explained them.
Researchers applied the mathematical framework used to describe mantle convection — the slow churning of Earth's interior that drives continental drift — to ice sheet physics. The result: the plumes are thermal convection cells. Vertical temperature differences within the ice drive a slow circulation, warmer ice rising, cooler ice sinking, creating organized patterns within a medium no one expected to convect.
Ice is at least a million times softer than Earth's mantle. That difference in viscosity seems like it should prevent convection — the material is too weak to sustain organized circulation. But the physics works in the opposite direction. The softness doesn't prevent convection. It permits it at lower temperature gradients. The Rayleigh number — the ratio of buoyancy-driven forces to viscous resistance — crosses the critical threshold because the viscosity is low, not high. Mantle rock convects despite being stiff. Ice convects because it is soft.
We think of ice as a solid that deforms slowly under pressure — glacier flow, creep, basal sliding. These are mechanical responses to stress. Thermal convection is different. It is a self-organizing circulation driven by heat, the same process that moves Earth's plates and stirs the sun's interior. Finding it in ice places a familiar material in an unfamiliar category.
The through-claim: the material's identity as a solid made its convective behavior invisible. Convection was not looked for in ice because ice is not classified as a convective medium. But classification is not physics. The equations don't check the label.