Answers are grounded in this paper and cite their sources.
It is demonstrated that rejection of salt at a freezing ice-water interface takes place at least in part by cyclic convective processes occurring within the small interstitial spaces located in the first few centimeters of ice above the interface. This invalidates the assumption of the diffusive boundary layer utilized by previous authors and shows that near-interface temperatures are subject to cyclic variations. This result should be generally true for phase changes in aqueous solutions with floating ice, and it must be considered as a modification to current theories of solidification. It is shown that large vertical tubular structures attended by smaller tributary tubes exist within the growing sea ice sheet; this arrangement almost certainly indicates horizontal migration of brine toward preferred drainage areas in a manner analogous to the catchment area of a river and its tributaries.
No takes yet. Share an insight, caveat, or question.
Lake et al. (1970) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: