Three-dimensional, first-order models of glacier flow have the capacity to incorporate fully spatially distributed variations in softness. To achieve this, however, modellers need to be provided with quantitative schemes describing the spatial distribution of ice flow at real ice masses. Thus, the key physical controls over ice softness need to be identified and their influence over softness quantified over a realistic range of temperatures and stress regimes. In parallel with this, schemes describing the spatial evolution of those key ice properties, such as ice crystal size, shape and fabric, at ice masses need to be developed. One means of moving towards such schemes is to acquire sets of ice cores from along flow-lines at ice masses and to compare their physical characteristics and softness variations. In such studies, softness can be measured either directly in the laboratory or inferred from comparing measured three-dimensional motion fields with model output. Although the pilot study by Hubbard et al. (2003) summarized above has demonstrated the potential importance of including spatial variations in ice character to the modelling of temperate glaciers, the analysis is limited because it is based solely on the bulk ionic composition of the ice. Future studies need to consider suites of rheologically-important characteristics. Where such schemes may be considered to be too logistically demanding, it may be possible to quantify key properties such as bulk water content by non-intrusive geophysical techniques such as radar (e.g. Murray et al., 2000a).
Was this article helpful?