Scientists studying the Greenland Ice Sheet have found evidence that massive layers of refrozen meltwater deep beneath the surface are warping and softening the ice from within, producing internal structures that bear a striking resemblance to convection plumes in molten rock. The discovery, drawn from ice-penetrating radar surveys and supported by gravity measurements, suggests the ice sheet is far more internally active than standard models assume. As surface melting accelerates under warming temperatures, this hidden churning could change how quickly ice flows toward the ocean and contributes to sea-level rise.
Refrozen Meltwater Builds Thick Hidden Layers
The process starts at the base of the ice sheet, where geothermal heat and pressure melting generate liquid water. That water does not always drain away. Instead, it refreezes onto the underside of the overlying ice, building up enormous formations known as basal ice units. In northern Greenland, these units have been reported at up to roughly 1,100 m thick, making them a significant fraction of the total ice column in some locations. Their sheer scale was not widely appreciated until radar surveys revealed their extent, showing thick, patchy bodies of ice that differ markedly from the layered structure above.
These basal ice units form through a process distinct from the familiar snowfall-and-compaction cycle that builds most of the ice sheet. Because they originate from liquid water that refreezes under pressure, their crystal structure and thermal properties differ from the glacial ice above. That difference matters: refrozen basal ice is warmer and softer than the surrounding material, and its presence alters how the ice sheet deforms under its own weight. The result is a deep interior that behaves less like a rigid frozen slab and more like a slowly churning fluid, with consequences that ripple upward through the entire ice column and outward toward fast-flowing outlet glaciers.
Radar Reveals Disrupted Internal Architecture
Ice-penetrating radar has been the primary tool for mapping these hidden features. When radar pulses travel through an ice sheet, they bounce off internal layers that formed at the same time across wide areas, creating what glaciologists call isochrones. In a well-behaved ice sheet, these layers stack neatly like pages in a book. But across large swaths of Greenland, the layers are anything but neat. A broad synthesis of Greenland radiostratigraphy documents widespread zones where internal reflections are buckled, folded, or entirely scrambled, recording a history of deformation that standard ice-flow models struggle to reproduce.
The disruption is not random. It concentrates in regions where basal conditions favor melting and refreezing, and it follows patterns consistent with active deformation rather than passive burial of old features. Researchers interpret these warped isochrones as evidence that basal processes can strongly disrupt and reshape the ice sheet’s internal layering, sometimes over distances of hundreds of kilometers. Comparisons with satellite gravity measurements have independently supported the presence of dense refrozen units, confirming that the mass anomalies expected from large basal bodies match what is detected from orbit. Together, the two measurement techniques build a case that the base of the Greenland Ice Sheet is far more dynamic than its frozen surface implies.
Plume-Like Structures Challenge Convection Theory
One of the most striking features in the radar record is the appearance of plume-shaped structures rising from the bed into the ice column. These features look remarkably like the thermal plumes that drive convection in Earth’s mantle, which is what prompted comparisons to processes in molten rock. The radar images show narrow, vertically oriented zones where internal layers bend sharply upward, as though columns of softer ice were pushing through the surrounding material. At first glance, this geometry seems to invite an explanation based on buoyant upwelling driven by temperature differences within the ice itself.
But a competing explanation has gained traction. Modeling work published in Nature Communications proposes that basal freeze-on alone can generate plume-like internal structures without invoking true thermal convection. In these simulations, water refreezing at the base adds new ice that is warmer and mechanically weaker than the overlying column. As overlying ice flows toward outlet glaciers, the newly formed basal ice is drawn into elongated, rising bodies that distort the surrounding layers, naturally producing the observed plume shapes. The geometry of the plumes evolves with the rate of freeze-on and the speed of horizontal flow, and convergent flow patterns can concentrate the features into narrow, dramatic columns.
Why Internal Churning Changes the Ice Loss Equation
This distinction between convection and freeze-on carries real scientific weight. If the plumes result from convection, it would mean the ice sheet contains self-sustaining circulation cells, a process with no confirmed precedent in glaciology and potentially large implications for how heat is transported through the ice. If they result from freeze-on, the mechanism is better understood but still capable of producing large-scale softening and deformation. The freeze-on explanation does not make the ice sheet less active; it simply relocates the energy source from internal heat gradients to basal water dynamics. Either way, the internal architecture recorded by radar shows an ice sheet whose deep layers are being reworked on scales that matter for predicting future behavior.
Standard ice-sheet models treat the interior as a relatively passive body that flows slowly under gravity, with most of the interesting dynamics happening at the surface, where snow falls and melts, and at the margins, where glaciers calve into the ocean. The discovery of thick, warm, deformable basal units challenges that framework. When refrozen meltwater softens the ice from below, it reduces the effective viscosity of the column, allowing it to flow faster under the same gravitational stress. This means that regions with extensive basal freeze-on could be losing ice to the coast more quickly than models predict, even before accounting for any increase in surface melting, and that changes in basal hydrology could have outsized effects on ice discharge.
Implications for Sea-Level Projections and Future Research
The practical stakes are tied to sea-level projections. Greenland holds enough ice to raise global sea levels by about seven meters if it all melted, and even small accelerations in ice discharge translate to measurable changes at coastlines worldwide. Current models already struggle to match observed ice velocities in some drainage basins, and unaccounted-for basal softening could explain part of the gap. A detailed radiostratigraphy review of the ice sheet’s age structure shows that disrupted zones coincide with areas of complex flow, suggesting that internal deformation linked to basal processes is not a marginal curiosity but a key control on how ice is routed from the interior to the sea.
To narrow the uncertainties, researchers argue that models must better represent the thermomechanical effects of basal freeze-on and the evolving geometry of plume-like structures. That will require more than remote sensing alone. Direct measurements of basal temperatures, melt rates, and water pathways beneath the thickest parts of the ice sheet remain extremely sparse, yet they are essential for distinguishing between competing mechanisms and for testing model predictions. Future campaigns that combine deep ice coring, borehole instrumentation, and expanded radar and gravity surveys could reveal how widespread these hidden basal units are, how rapidly they grow, and how strongly they influence ice flow. Until then, the emerging picture from Greenland is that what happens at the base of the ice sheet may be just as important for sea-level rise as what happens at its surface.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
