How Polar Ice Strain Deforms and Fractures Ice Shelves

Polar ice strain refers to the deformation that glaciers, ice sheets, and sea ice undergo when forces act on them, whether from their own weight, ocean currents, tides, or warming temperatures. It is one of the most important variables in predicting how quickly ice will break apart, flow into the sea, and raise ocean levels. The concept sounds straightforward, but the way ice actually responds to stress turns out to be surprisingly complex, varying with temperature, crystal structure, and whether the ice sits on rock, floats on water, or is being pulled apart by its own motion.

How Ice Deforms Under Stress

Ice is a crystalline solid, but under sustained pressure it behaves more like an extremely thick fluid. Rather than snapping immediately, glacial ice slowly creeps, deforming over days, months, and centuries. The relationship between the stress applied to ice and the rate at which it deforms is described by flow laws that glaciologists have been refining for over seventy years. Getting these laws right matters enormously, because ice sheet models use them to predict how fast ice will move and how much it will contribute to rising seas.

One persistent challenge is that flow laws calibrated from cold laboratory experiments tend to underestimate how fast ice deforms at warmer temperatures, especially above about −10 °C. Near the melting point, water begins to form along the boundaries between individual ice crystals, lubricating them and allowing faster creep. Some models try to handle this by switching between different calibration values at an arbitrary temperature threshold, but this creates artificial jumps in predicted strain rates that don’t appear in real data.

1PubMed Central. Flow laws for ice constrained by 70 years of laboratory experiments

Ice isn’t uniform inside, either. As snow compresses into glacier ice over thousands of years, individual ice crystals rotate and align in preferred directions, forming what glaciologists call a crystal orientation fabric. This fabric makes the ice respond very differently to stress depending on which direction the stress comes from. In an Antarctic ice stream, radar measurements have shown that spatial variability in crystal fabric translates directly into variability in how stiff or soft the ice is, and that widely used ice-flow models that treat ice as the same in every direction simply cannot capture this effect.

2Journal of Geophysical Research: Earth Surface. Radar Characterization of Ice Crystal Orientation Fabric and Anisotropic Viscosity Within an Antarctic Ice Stream

A recent field experiment on a small Swiss glacier used fiber-optic cables to measure strain rates at sub-meter scales down a borehole, a level of detail no conventional instrument could achieve. The results supported a nearly linear relationship between stress and strain for ice near its melting point, consistent with lab experiments on temperate ice. That finding challenges the standard assumption that glacial ice always deforms in a strongly nonlinear way, and it suggests that models may need different parameters for ice that is close to melting versus ice deep in the cold interior of an ice sheet.

3EGUsphere. Fibre-optic strain sensing on an alpine glacier

When Strain Becomes Fracture

Slow creep is one thing. But when the rate of deformation exceeds what ice can absorb by flowing, cracks form. Crevasses, the deep fissures that make glacier travel dangerous, are the visible result of tensile strain exceeding the ice’s strength. Surface crevasses are common where ice flows over a bump in the bedrock or where an ice stream accelerates and stretches. On the Filchner-Ronne Ice Shelf in Antarctica, researchers tested fracture mechanics models against satellite imagery and found they could accurately predict where crevasses would appear, as long as they assumed an initial flaw size in the range of roughly 5 to 50 centimeters.

4Journal of Geophysical Research: Solid Earth. Experimental and theoretical fracture mechanics applied to Antarctic ice fracture and surface crevassing

Crevasses don’t just open at the surface. Under the right combination of ice movement and the weight of overlying layers, cracks can initiate at depth, forming so-called basal crevasses that open upward from the bottom of floating ice shelves. Fracture mechanics analysis shows that ice movement generates a tensile stress that competes with the compressive weight of the ice above. If a pre-existing flaw at depth is long enough, the tensile stress wins and the crack grows.

5Journal of Geophysical Research: Solid Earth. Subsurface crevasse formation in glaciers and ice sheets

Surface meltwater makes everything worse. When water fills a crevasse, it pushes outward on the walls with hydraulic pressure, driving the crack deeper in a process called hydrofracture. Under certain conditions, a water-filled crevasse can slice through the full thickness of a glacier, triggering calving or altering the plumbing system beneath the ice. This process is not hypothetical: hydrofracture occurs at tens of thousands of sites on the Greenland ice sheet each summer, and its frequency is increasing as the atmosphere warms.

6Extreme Mechanics Letters. A poro-damage phase field model for hydrofracturing of glacier crevasses

Ice Shelf Buttressing and What Happens When It Weakens

Floating ice shelves act as brakes on the ice streams that feed them. By pressing against coastlines, islands, and each other, they hold back the flow of grounded ice behind them. When an ice shelf thins, fractures, or collapses, the restraint disappears, and the ice streams behind it speed up. This acceleration contributes directly to sea-level rise. Modeling studies have shown that the speedup is greatest for ice streams that are wide, have stickier beds, or experience higher driving stresses.

7Geophysical Research Letters. Assessment of the importance of ice‐shelf buttressing to ice‐sheet flow

Pine Island Glacier, one of the fastest-changing outlets of the West Antarctic Ice Sheet, illustrates this process in real time. Between 2017 and 2023, satellite measurements of surface deformation, velocity, and strain rates revealed localized acceleration and structural weakening of the glacier’s southern ice shelf. The peripheral shelf appears to be entering a progressively unstable configuration, with weakening margins that can no longer restrain the ice behind them effectively.

8Journal of Glaciology. Deteriorating structural integrity of Pine Island Glacier’s Southern Ice Shelf (2017–23) identified with satellite-derived surface deformation, ice velocity and strain rates

Estimates of future sea level rise have converged in recent years between different modeling approaches, but the biggest remaining uncertainty comes from ice sheet dynamics, particularly from how ice streams and their shelves will respond to warming oceans. Progress has been made in understanding ice stream flow, the migration of grounding lines where ice lifts off the bedrock, and the coupling of ice sheet models with high-resolution climate models. But ice dynamics remains the component with the largest potential to push sea level projections higher.

9Reviews of Geophysics. Semiempirical and process‐based global sea level projections

Tides, Waves, and Wind as Strain Drivers

The forces straining polar ice aren’t limited to the ice’s own weight and motion. Ocean tides flex floating ice shelves up and down with every tidal cycle, and this cyclic bending generates stress that can open and propagate rifts. On the Brunt Ice Shelf in Antarctica, a study found that the tide-driven stress during spring tides exceeded the stress from the ice shelf’s own spreading by more than a factor of two. Five separate rift-propagation events of between 50 and 500 meters occurred there in early 2019, each triggered when high winds coincided with spring tides. The combined effect pushed stress above a threshold of roughly 240 kilopascals, within the range known to open surface crevasses.

10PubMed Central. Ocean tides trigger ice shelf rift growth and calving

Ocean swells play a related but distinct role. When sea ice that normally buffers ice shelves from open-ocean waves disappears seasonally, long-period swells can reach the outer margins and cause repeated flexure. This wave-driven bending is especially effective in regions where bottom crevasses have already thinned the ice from below. The collapses of the Larsen A, Larsen B, and Wilkins ice shelves all followed periods when the protective sea ice buffer had diminished, exposing outer margins to wave action that weakened them to the point of calving.

11Nature. Antarctic ice shelf disintegration triggered by sea ice loss and ocean swell

A pan-Antarctic assessment of ice shelf responses to ocean waves reinforced these findings, showing that wave-induced flexure played a clear role in large calving events on the Wilkins shelf and may have contributed to calving on the Larsen C and Conger shelves as well.

12Journal of Geophysical Research: Oceans. Pan‐Antarctic Assessment of Ice Shelf Flexural Responses to Ocean Waves

Tidal flexure also affects glaciers that terminate in fjords. At Jakobshavns Glacier in Greenland, tidal oscillations within the fjord cause the glacier’s floating tongue to bend. An analysis combining elastic and viscoplastic deformation estimated that only about 40% of the glacier’s total thickness behaved as a coherent structural unit 2.6 kilometers from the calving front. Closer to the calving front, that figure dropped to roughly 21%, reflecting how fractures progressively weaken the ice as it nears the point of breaking off.

13Journal of Geophysical Research: Solid Earth. Tidal flexure of Jakobshavns Glacier, west Greenland

What Happens at the Base

Much of the strain in a glacier doesn’t happen where you can see it. Near the bottom, where ice grinds against bedrock or sediment, conditions are radically different from the cold, brittle surface. The basal ice layer is often warmer, contains trapped sediment and debris, and deforms much more readily. In Taylor Glacier, Antarctica, detailed monitoring of a 4.5-meter-thick basal ice sequence showed that about 60% of the glacier’s total surface velocity could be attributed to deformation within that thin basal layer alone. An additional 15% of motion came from sliding or extremely localized shear at the very base.

14Journal of Geophysical Research: Earth Surface. Deformation, Strength and Tectonic Evolution of Basal Ice in Taylor Glacier, Antarctica

The nature of what lies beneath the ice also matters. Where the bed is made of deformable sediment rather than hard rock, the sediment itself can shear, allowing the ice to advance faster. Numerical modeling of the former Lake Michigan Lobe of the Laurentide Ice Sheet suggested that a soft-bedded lobe could reach a steady configuration in about 20,000 years, compared with 50,000 to 60,000 years for a hard-bedded lobe. A soft bed also made the ice roughly twice as responsive to shifts in snowfall or melting over thousand-year timescales.

15Journal of Geophysical Research: Solid Earth. Numerical modeling of subglacial sediment deformation: Implications for the behavior of the Lake Michigan Lobe, Laurentide Ice Sheet

Sea Ice Strain Is a Different Beast

Strain in floating sea ice follows a different set of rules from the slow creep of glaciers and ice sheets. Sea ice is thinner, colder, and more brittle, and it responds to wind and ocean currents on timescales of hours rather than decades. When a sea ice floe fractures, it can be sudden and dramatic. GPS buoys deployed on a single Arctic floe recorded three deformation events over nine months, each lasting less than a day. During these events, the area occupied by the buoy array shrank by more than half. The strain rates measured, on the order of 10⁻⁵ per second, were about a thousand times larger than previous estimates for brittle fracturing of cracks roughly 100 meters long.

16Deep Sea Research Part II: Topical Studies in Oceanography. Small-scale deformation of an Arctic sea ice floe detected by GPS and satellite imagery

These events matter for shipping, for the accuracy of sea ice forecasts, and for understanding how the Arctic ice pack fragments as it thins. Thinner ice fractures more easily under the same wind forcing, and as the Arctic continues to lose multi-year ice, the remaining ice is increasingly vulnerable to rapid deformation events. Each fracture episode rearranges the ice pack, creating leads of open water that absorb more sunlight and accelerate further melting.

Measuring Strain from Space and on the Ground

Tracking how ice deforms requires measuring tiny displacements over vast, remote areas. Two main satellite approaches dominate the field. Feature tracking uses pairs of optical or radar images to follow recognizable patterns on the ice surface as they move between acquisitions, producing velocity fields from which strain rates can be derived. A statistical toolkit called GLAFT was developed to evaluate the quality of these velocity products, and it computes along-flow and across-flow strain rates by analyzing how velocities change across the ice surface.

17The Cryosphere. GLAcier Feature Tracking testkit (GLAFT): a statistically and physically based framework for evaluating glacier velocity products derived from optical satellite image feature tracking

Interferometric synthetic aperture radar, or InSAR, takes a different approach. By comparing the phase of radar signals returning from the same spot on two different passes, InSAR can detect displacements with centimeter-level accuracy. This technique has been used to measure strain in landfast sea ice in Terra Nova Bay, Antarctica, separating out the contributions of glacial flow and tidal forcing. The catch is that InSAR needs image pairs acquired close together in time, typically one day apart, to avoid losing the signal as the ice surface changes.

18Remote Sensing of Environment. Glacial and tidal strain of landfast sea ice in Terra Nova Bay, East Antarctica, observed by interferometric SAR techniques

On the ground, new fiber-optic distributed strain sensing is opening up a scale that satellites cannot reach. A fiber-optic cable installed in a borehole can measure strain continuously along its entire length at sub-meter resolution. The Alpine glacier experiment mentioned earlier demonstrated that both Brillouin and Rayleigh techniques resolved highly variable strain rates that conventional tiltmeters or borehole inclinometers would completely miss.

19EGUsphere. Fibre-optic strain sensing on an alpine glacier

Seismology adds yet another window. Glaciers and ice sheets generate their own seismic signals, collectively studied under the banner of cryoseismology. These icequakes range from tiny clicks produced by individual cracks to large events triggered by calving or sudden basal motion. Seismic monitoring has yielded insights into basal sliding, calving dynamics, and precursory signals that may indicate an ice structure is approaching instability.

20Reviews of Geophysics. Cryoseismology

Strain Recorded in Ice Cores

Ice cores contain a record of strain stretching back tens of thousands of years, though reading that record requires careful detective work. As new snow accumulates on top, older layers are buried and compressed, stretching horizontally and thinning vertically. This thinning means that a single annual layer that was originally a few tens of centimeters thick may be squeezed to a fraction of a millimeter near the bottom of a deep core. To reconstruct past climate from the core, scientists need to know how much each layer has been thinned, which means they need a model of how the ice has deformed over time.

At the South Pole, data from the SPC14 ice core were used to reconstruct temperature, snowfall rates, and layer thinning over the last 54,000 years. The analysis revealed millennial-scale variations in the thinning function, meaning the ice didn’t deform at a steady rate. At greater depths, layers were less thinned than a simple one-dimensional flow model would predict, suggesting that bumps and valleys in the bedrock beneath the ice sheet were steering the flow in ways that a vertical-only model couldn’t capture.

21Journal of Geophysical Research: Atmospheres. Reconstruction of Temperature, Accumulation Rate, and Layer Thinning From an Ice Core at South Pole, Using a Statistical Inverse Method

This matters because any error in reconstructing past strain translates directly into an error in the age assigned to deeper ice, and therefore into the timeline of past climate events. It also feeds back into ice sheet modeling: if the strain history at a core site doesn’t match what a flow model predicts, that mismatch tells scientists something about bed conditions, ice rheology, or past changes in flow patterns that the model has wrong. In this way, ice cores serve as both an archive of past climate and a test of how well we understand the physics of ice deformation itself.