The Ice That Negotiates

In a paper published September 10 in Physical Review Letters, Bryan Shaddy, P. Alex Greaney and Bhargav Rallabandi model sea ice as a granular medium driven by fluctuating winds. “Granular” here does not mean tiny. The grains happen to be floating slabs. Physics is refreshingly indifferent to whether its particles would flatten your house.

The researchers report that their simulations reproduce three measured properties of ice movement in the Fram Strait, between Greenland and Svalbard: how the floes disperse, the distribution of their velocities, and the power spectrum of their motion—how the fluctuations are distributed across different timescales. Their inputs are measured local environmental and ice properties. A UC Riverside account, carried by ScienceDaily on September 17, describes the model as including ocean drag as well as collisions.

The lovely part is how much the collisions accomplish.

According to the paper’s abstract, encounters between floes rapidly dissipate energy supplied by the wind while shortening the distance a floe travels between collisions. In a crowded field, motion cannot be understood simply by handing each piece its own portion of wind and watching it go. Each piece keeps meeting the consequences of everyone else being there.

That helps explain why the ice spreads more slowly than a simple wind-based picture would suggest. The release describes dense fields in which collisions happen more frequently than the wind changes. The large-scale movement carries the signature of those repeated interruptions.

There is no negotiation in the conscious sense. No floe has formed a committee. But the title’s metaphor has a mechanical foothold: the eventual motion is worked out through encounters, not dictated by the push alone. The neighbors are part of the dynamics, not scenery surrounding the real experiment.

This is also a useful way to think about what a good simplification preserves. A model cannot contain an entire Arctic Ocean at full detail. The question is which relationships it can afford to leave out. Here, keeping the contacts between pieces lets a relatively simple description recover several kinds of observed behavior at once. Simplicity is not the same as isolation.

The result has boundaries. Matching these transport statistics in the Fram Strait does not establish that collisions explain every ice regime, or that winds and ocean processes can be demoted to background noise. The model needs its driving environment. Nor is this a demonstrated forecast of where Arctic ice will end up under future warming.

Rallabandi’s proposed application, as described in the university release, is to help investigate those future changes. Ice coverage and floe size affect the frequency of collisions; altered collisions could change how readily ice spreads toward warmer waters. That is a question the framework may help answer, not an answer already delivered.

For now, the achievement is satisfyingly physical. An enormous moving field becomes more intelligible when its pieces are allowed to interfere with one another. Nothing mystical has been added. The explanation becomes richer because the objects are less alone.

Even a kilometer-wide slab does not get the ocean to itself.