Volcanology


Earth Science programmes

Microscopic water content of glaciers

This ongoing project, run by Dr Sebastien Mueller, is on the 'Effect of spatial variability of physical parameters on the rheology of temperate glaciers and their response to climate change'.  The project aims to constrain the microscopic water contained within the glacial ice and follows on directly from the topic of a PhD thesis on 'The water vein system in polycrystalline ice' by HM Mader, which was published in Mader 1992a, b. The issue here is that polycrystalline ice, such as is found in glaciers, at subzero temperatures contains a liquid-water phase. In essence, glacial ice is a 'partial melt', provided the temperature is above the eutectic temperature of the impurities contained within the ice. The existence of this water phase has some profound implications for the behaviour of the bulk ice:
  1. The ice 'softness' will be affected and hence its flow behaviour
  2. The response of the ice to temperature changes will involve both latent as well as specific heat
  3. It can facilitate the transport of both heat and impurities through the ice.

In considering the microscopic structure of ice, it is important to distinguish between 'single' crystals and 'polycrystals'. A single crystal is one in which the strict periodic arrangement of the molecules continues in all directions throughout the material with no large-angle boundaries. By contrast, a polycrystal is essentially a composite of many, more-or-less randomly-oriented single crystals such that large-angle crystal boundaries arise where the individual single crystals meet.

Natural ice found on Earth in glaciers and ice sheets is polycrystalline. The image to the right shows a thin section of polycrystalline ice from Dome C, Antarctica (ex: Barnes, PhD thesis, 2002) viewed between crossed polarisers that shows the individual crystals very clearly. The crystals are a few millimetres across.

NaCl skeleton
A water phase exists within polycrystalline ice because the ice lattice tends to reject foreign ions; in other words, the solubility of compounds in the ice (single) crystals is generally extremely low. As a result, as water is progressively frozen, most foreign ions are expelled from the growing crystals and remain in the liquid-water phase, which becomes increasingly concentrated. Ultimately, a polycrystal is formed which contains an interconnected network of highly-concentrated water-filled 'veins' and films around the crystals. The image below (ex: Barnes et al. 2003) shows an SEM image of the NaCl skeleton that remains after sublimation of the grain interiors and shows clearly how the NaCl is has become concentrated in the films around the crystals.

The vein system has been observed by numerous workers and indeed it can be readily seen with the naked eye - if you allow an ice cube from your fridge to warm up a bit, the veins look like tiny silvery lines. The geometry of the vein system has been determined (see e.g. Mader 1992a) and is controlled by the 'dihedral angle', which is the angle subtended at a crystal boundary in contact with water.

ice veins The composite figure to the left shows the dihedral angle in figure (a) and the cross-section of a vein at a 'triple-junction' (i.e. the line where three crystals meet) in figure (b). Four veins meet in a 'node' at the points where four crystals meet. The images in figures (c) and (d) are photographs of the vein system in laboratory-grown ice using transmitted white light (ex: Mader 1992a. The veins can be seen because water has a different refractive index from ice. The optics of the system is described in detail in Mader 1992a. In (c), a vein cross-section is clearly visible. The vein cross-section is not strictly uniform. This is primarily because the vein is not exactly perpendicular to the angle-of-view, but may also be due to slightly different values of interfacial energy for the three crystal-crystal boundaries associated with the cross-section. The photograph in (d) shows a node where four veins meet. All the veins in these images are approximately 100mm across. The diagram in figure (e) illustrates the vein network around a crystal (redrawn after Price 2000).

A paper on the implications of the liquid water content for radar measurements has recently been published (West et al 2007) and several other manuscripts are currently in preparation.

Those involved in this work include Dr. Heidy Mader,  Dr David Chandler (NERC PDRA 2004-2007), Dr Bryn Hubbard (Institute of Geography and Earth Sciences, Aberystwyth), Dr David Rippin (NERC PDRA 2004-2006), Dr Jared West (School of Earth and Environment, Leeds), Professor Tavi Murray (Department of Geography, Swansea).

Funded by the Natural Environment Research Council (UK)

Related publications: