This project is a three-year continuation of a theoretical and experimental study of the dynamics of liquid water existing at the boundaries of ice crystal at temperatures below the freezing point of pure water. This process, called premelting, has significant consequences in the redistributing of brine in sea ice, the apparent migrating of contaminants through glacier ice, and the rate of frost heave and weathering in soils and rocks. Premelting is common to virtually all types of materials, but has been most extensively studied in ice. It can be caused by adsorption forces, wetting, size effects, or substrate disorder, and involves many aspects of surface physics and phase transition phenomena. The macroscale consequences of these molecular-scale processes include the relationship between seasonal freezing and the oceanic thermohaline circulation, and the quantitative interpretations of paleoclimate proxy records. On the longer climatic time scales, the overall features of recent glacial-interglacial cycles have been revealed in the Antarctic ice cores, but higher resolution is required to decipher the mechanisms that drive abrupt climate changes. The accuracy of any higher resolution paleoclimatic information derived from ice cores will have to rely on the proper treatment of processes that affect the trace constituents during their long (100 kyr) residence times in the ice sheets. Grain-boundary melting is believed to be the primary process controlling the rate of material transport through the premelted-liquid networks of veins and nodes that line the grain intersections of glacier ice. Moreover, the role of grain-boundary melting can further increase the liquid content and enhance trace-constituent mobility. The project has three distinct, but interrelated components. Direct measurements of premelting of grain boundaries in the presence of different impurities will be made as a function of temperature and grain mismatch, and experiments on instabilities that are predicted to occur at the growing sea ice/ocean interface in the presence of a shear flow will be conducted. Both will be complemented by theoretical calculations to prove experimental issues and to pursue the implications of premelting in numerous contexts.
