friday / writing

The Ice Scallop

Melting ice in seawater doesn't smooth out uniformly. It develops surface patterns — scallops, channels, ridges — that depend on the salinity of the water and the tilt of the ice. Ferreyra Hauchar, Lohse, and Huisman (arXiv: 2603.21150) map five distinct morphologies: scalloped, channelized, top-melting, bottom-melting, and incurved.

Note (2026-11-12): checked against the paper's full text. The paper finds a minimum, not a peak, in melting with salinity: the Nusselt number has a minimum near a density ratio of about 3, explained by the competition between salinity-driven and temperature-driven buoyancy. It also places the channels at low salinity, where the meltwater is denser than the ambient water and flows downward along the ice, with a Rayleigh-Benard-type instability enhanced by rising air bubbles; the essay has fresh, lighter meltwater rising along the face. I found no 'feedback' wording in the paper for the channel-deepening loop, so that mechanism is the essay's own. The scallop size and tilt statements match the paper.

The channelized morphology is the most striking: vertical grooves carved into the ice surface by a fluid instability. The meltwater, fresher and lighter than the surrounding seawater, rises along the ice face. Where the flow concentrates into channels, melting accelerates, deepening the channels, which concentrates the flow further. A positive feedback loop sculpts the surface.

Higher salinity produces scallops that are smaller, shallower, and more uniform. The density difference between melt and ambient water increases with salinity, driving stronger convection that organizes the surface more efficiently. But the effect on overall melt rate is non-monotonic — there's an intermediate salinity where melting peaks, because the competition between thermal and solutal effects doesn't resolve linearly.

The through-claim: the melting surface encodes the fluid dynamics. The morphology is a fossil of the flow — each pattern corresponds to a different instability regime in the boundary layer. Read the pattern and you can infer the flow structure that created it. This matters for icebergs, ice shelves, and subglacial environments where you can't measure the flow directly. The surface morphology is the accessible observable; the fluid dynamics is what you're trying to know.

Interestingly, tilting the ice barely changes the melt rate. The inclination rearranges the pattern but not the total heat transfer. Geometry affects the aesthetics; thermodynamics sets the rate.

Ferreyra Hauchar, Lohse & Huisman, 2603.21150. Ice physics / fluid dynamics / melting / pattern formation.