Why Ice Is Slippery Without Being Simply Wet
The Film Is Not Just Water
Picture a skate blade moving across a frozen pond. The tempting explanation is almost shorter than the glide: pressure melts the ice, a film of water appears, and the blade slides on it.
The trouble is not that water has nothing to do with slippery ice. It is that this explanation treats the most interesting part—the material between the blade and the solid—as if naming it settled how it behaves.
A small laboratory experiment gave me a better question. What if the layer can flow, yet does not behave like ordinary liquid water?
Three ways to get the story wrong
There are at least three distinct ideas hiding inside “a layer of water.”
Pressure melting changes the temperature at which ice and liquid water can coexist. Ordinary ice occupies more volume than the same mass of liquid, so pressure favors the denser phase. But lowering the melting temperature does not magically supply all the energy needed to melt the ice. Significant melting below freezing requires the right local pressures and an energy source. This is a possible ingredient in a contact, not a complete explanation of every glide. [1]
Frictional melting supplies heat through motion. In the familiar self-lubrication picture, sliding generates a film that helps the surfaces slide. That proposal has obligations: the heat must go somewhere, water can be squeezed out, and the surviving film must be thick enough to separate the surfaces it supposedly lubricates. [1]
Surface premelting can exist before pressure or sliding enters the picture. Molecules at an ice surface do not have exactly the same neighborhood as molecules locked inside the crystal. Below the bulk melting temperature, the surface can develop disorder and increased molecular mobility—a quasi-liquid layer. But a mobile molecular skin is not necessarily a supporting puddle, and its presence does not guarantee that a rough object cannot reach the solid beneath. [1]
These mechanisms can interact. Calling them all “the water film” conceals the differences we need to explain.
The independent research team behind a 2021 review makes another useful objection: ice is not merely a reservoir waiting to melt. Under rapid loading it can fracture. A slider can crush or scrape it, creating particles between the two original surfaces. Those fragments may matter as much as any liquid. The review argues for taking abrasion and ice-rich slurries seriously; it does not establish a single replacement mechanism for every skate, tire and snow grain. [1]
The bead that stroked and listened
In 2019, L. Canale and colleagues studied a much more controlled contact than our imagined skate: a glass bead, about 1.5 millimeters in radius, attached to a tuning fork. [2]
The bead oscillated sideways against ice. At the same time, a smaller perpendicular oscillation probed the interface. One motion sheared the contact; the other asked how it resisted a gentle mechanical disturbance. Force and phase responses let the researchers investigate both sliding friction and the layer's mechanical behavior.
The main experiments used ice temperatures between −16°C and 0°C. The surface had been prepared by preliminary approach-and-withdrawal movements, and the bead repeatedly stroked a small region. This was not a blade crossing untouched ice once. That difference matters when deciding how far the result travels. [2, Section II]
As the bead withdrew while continuing to shear the ice, its response followed a pattern consistent with a fluid-like interface. Fitting a hydrodynamic model allowed the researchers to infer a layer roughly 100–500 nanometers thick—about 0.1–0.5 micrometers—with thickness increasing towards melting. This was a mechanical inference, not a photograph of a transparent film. [2, Section III C]
The more revealing result was how that layer resisted motion.
Its inferred response was viscoelastic: partly dissipating energy as a viscous material does, partly storing and returning it elastically. Near melting, its effective dissipative viscosity could approach a hundred times that of ordinary water. Those properties were inferred under the experiment's imposed motions and frequencies; they are not a complete description of every way the material might flow. [2, Section III D]
The mechanical response supported a fluid-like interfacial layer. “Ordinary water” was not an adequate mechanical description.
Why thicker is not the same as slipperier
At first, a more viscous lubricant sounds like the wrong answer. Doesn't thicker, stickier fluid mean more resistance?
There are two different jobs here. A layer must resist being squeezed away if it is to keep solid surfaces apart. But shearing that layer also costs energy. A material can help with the first job while making the second harder. “More viscous” is therefore not a universal recipe for lower friction.
The experiment supplies a useful warning against the simpler slogan. As the ice approached melting, both the inferred film thickness and the measured friction rose. A thicker layer did not, by itself, mean an easier glide. [2, Sections III A–D]
The team also changed the bead's surface with a water-repelling molecular coating. Friction fell—by as much as a factor of ten in their apparatus—without a major change in inferred film thickness. The inferred dissipative viscosity fell too, especially near melting. That comparison makes a thickness-only explanation even less convincing. It does not turn a glass bead into a ski-wax prescription. [2, Section III E]
What changed my picture was not simply the discovery of a film. It was the need to distinguish keeping the solids apart from how much energy is lost while they move.
The plausible mixture we have not yet seen
Canale and colleagues proposed that sliding might create a suspension: liquid mixed with submicron fragments worn from the ice. A crowded mixture could help explain why the layer both flows and has an elastic response. The later review places their result alongside other evidence for crushing, abrasion and ice-rich slurries. [1; 2, Section IV]
But a mechanical signature is not a census of ingredients. The primary paper explicitly leaves full structural characterization of the material under shear to future work. Ice fragments in this particular layer are an explanation to test, not particles the apparatus directly identified and counted. Nor does a review discussing the experiment constitute independent replication.
This is the limit I want to keep rather than edit away. A smooth sphere oscillating over a prepared patch has taught us something concrete about an ice contact. It has not settled every one-pass blade, rough boot or moving tire.
“Water makes ice slippery” used to feel like an answer. Now it feels like the beginning of an investigation: where did the interfacial material come from, what is in it, and how does it carry load while letting motion happen?
The next thing I want to see is that material identified directly as it forms under a fresh, single-pass contact. Not just whether a layer exists—but what the layer becomes.
Sources and reading limits
[1] James H. Lever, Emily Asenath-Smith, Susan Taylor and Austin P. Lines, “Assessing the Mechanisms Thought to Govern Ice and Snow Friction and Their Interplay With Substrate Brittle Behavior,” Frontiers in Mechanical Engineering 7 (2021). Critical review; its main narrative and captions were read. Its cited studies are not all independently inspected here, and its proposed unifying slurry account remains a hypothesis.
[2] L. Canale et al., “Nanorheology of Interfacial Water during Ice Gliding,” Physical Review X 9, 041025 (2019). Complete published main paper, methods, equations, captions and Appendix read. Supplementary material, original datasets and figure images were not inspected. The 2021 review cites this experiment; it is not a second experiment reproducing it.
This is a close reading of those sources, not a claim to survey all subsequent ice-friction research. I performed no sliding, skating, imaging or temperature experiment.