<p>First-year land-fast sea ice growth in both the Arctic and the Antarctic is characterised by the formation of an initial ice cover, followed by the direct freezing of seawater at the ice-water interface. Such growth usually results, through geometric selection, in congelation ice. This is, in general, the typical crystal structure observed in first-year ice growth in the Arctic. However, in certain regions of the Antarctic, platelet crystals are observed to contribute significantly to the ice growth, beyond a depth of 1 m. This thesis will investigate a number of ideas as to why the platelet crystals only appear in the ice after a significant amount of congelation growth has occurred. One of the key premises will be that platelet ice forms when smaller frazil crystals, beneath the ice, rise up and attach to the interface. They are then incorporated into the ice cover and become the platelets seen in ice cores. The Shields criterion is used to find the strength of turbulence, associated with tidal flow, required to keep a frazil crystal from adhering to the interface. It is shown that the sub-ice flow is sufficient to keep most crystals in motion. However, this turbulence may weaken or dissipate completely as the tide turns. The velocity associated with brine rejection is suggested as an alternative to keep the crystals in suspension during these periods of low shear turbulence. Brine rejection occurs as the sea ice grows, rejecting salt into the seawater below. By comparing this velocity with a model for the frazil rise velocity it is shown that brine rejection has sufficient strength to keep crystals in suspension. This effect weakens as the ice gets thicker, allowing larger frazil crystals to rise to the interface. The early work in this thesis shows that a flow can keep a single crystal from adhering to the interface. This can be regarded as the competence of a flow to keep a crystal in suspension. However, of equal importance is the capacity of a flow to keep a mass of crystals in suspension. It is shown that, given a sufficiently large mass of crystals beneath the ice, the same flow that can hold a single crystal in suspension will not be able to keep all the crystals in motion. The deposition of crystals is predicted to occur in a gradual manner if there is a steady build-up of crystals beneath the ice. The largest crystals, close to the interface, will settle against the ice as the flow is unable to support the entire mass of crystals Also considered is whether frazil crystals may be similar to cohesive sediments. If this is the case, a sudden influx of crystals from outside of the system may lead to the formation of a layer of unattached crystals beside the ice-water interface. This can cause a critical collapse of the turbulent field, resulting in the settling of a large quantity of frazil crystals. Though the emphasis of much of this thesis is on the effect of the flow on the crystals, it is also found that a mass of crystals can have a stabilising effect on the flow. The change in the density profile induced by an increase in the frazil concentration towards the ice-water interface (and hence a decrease in the density of the ice-water mixture) damps the turbulence produced by shear. The mass and size of crystals in suspension play major roles in the strength of stabilisation. Measurements of turbulence and the suspension of frazil crystals beneath sea ice are difficult to make. This thesis aims to present and analyse a number of models which may explain the platelet puzzle - the delayed appearance of the platelet crystals in ice cores. These are compared with the observations which are available, and conclusions made on the validity of the theories presented.</p>
Textural characteristics of sea ice and the major mechanisms of ice growth in the Weddell Sea
