Introduction

Fast ice flow plays a key role in the mass balance and dynamics of glaciers and ice sheets. In West Antarctica, for instance, fast ice streaming accounts for ca. 90% of total ice discharge into the ocean (Bentley, 1987; Vaughan & Spouge, 2002). Fast ice streams are also the most dynamic components of the West Antarctic ice sheet and have been shown to change their velocity by up to several orders of magnitude over time-scales as short as several minutes (Joughin et al., 2002; Bindschadler et al., 2003). On millennial time-scales, variability in flow of the Hudson Bay ice stream has been implicated as the driver of rapid climate changes observed in Greenland ice cores and deep-sea sediments (MacAyeal et al., 1995). Increasing evidence indicates that ice streams also played an important role as drainage pathways in palaeo-ice sheets (Stokes & Clark, 1999; Jansson et al., 2003; Sejrup et al., 2003). Bundles of megascale lineations found on continental shelves ringing Antarctica and other continents suggest that many palaeo-ice streams developed as ice sheets extended toward the shelf break during the Last Glacial Maximum (Shipp et al., 1999; Canals et al., 2000; Wellner et al., 2001; Anderson et al., 2002; Dowdeswell et al., 2004a). Some palaeo-ice streams existed also within terrestrial, rather than marine, portions of Pleistocene ice sheets (e.g. Stokes & Clark, 2003).

Measured glacier velocities range from zero to >10kmyr-1 in Jakobshavn Isbrae, Greenland, considered the fastest flowing glacier on Earth. Peak velocities during glacier surges lasting months to years have been recorded to reach ca. 53 m day-1 (equivalent to ca. 19kmyr-1) (Kamb et al., 1985, fig. 2C). Recent measurements of stick-slip motion on the ice plain of Whillans Ice Stream, West Antarctica, revealed 10-30 min long bursts of velocity in excess of 1mh-1, equivalent to ca. 10kmyr-1. There is no single widely agreed velocity threshold above which ice flow can be always classified as 'fast'. Rather, the concept of fast ice flow is a relative one and must be considered in a regional context. It is somewhat easier to define what velocities do not constitute fast ice flow. Ice moving with velocity of a few dozens of metres per year or less is clearly slow moving. Similarly, ice-flow velocities of a few hundred metres per year or more are fast. The difficulties may arise with velocities between ca. 50 and ca. 100myr-1. Figure

70.1 illustrates how one can distinguish between slow and fast moving ice using a regional context. The ice velocity data of Joughin et al. (1999) revealed that ice flow in West Antarctica is organized into ice streams, their tributaries and inland flow.

The distinction between slow and fast ice flow is often explicitly or implicitly associated with the distinction between motion due to internal ice deformation (slow) and motion caused by basal sliding or till deformation (fast). The inference that slow ice motion is caused by internal deformation whereas fast motion is due to rapid basal motion (a term used here after Blankenship et al. (2001) to encompass both basal sliding and till deformation) is frequently justified. For instance, observations made in West Antarctica show that fast ice streams, where the base is at melting point, move through rapid basal motion, whereas slow-moving interstream ridges are frozen to their beds and move through internal deformation (Bentley et al., 1998; Kamb, 2001). It is still uncertain what is the relative role of rapid basal motion and internal deformation in the motion of ice-stream tributaries (Hulbe et al., 2003). Recent force-balance estimates indicate that tributaries may flow by either of these mechanisms and that the predominant mechanism of motion may change along flow (Price et al., 2002; Joughin et al., 2004). However, the relation between fast motion and rapid basal motion is non-unique. For instance, it has been conjectured that Jakobshavn Isbrae attains its high velocity at least partly through deformation of a thick, temperate basal ice layer (e.g. Luthi et al., 2002).

Despite its importance, the phenomena of fast ice flow and ice streaming are still only partly understood and their representation in quantitative models of ice masses is rudimentary. Further observational studies and advances in modelling of fast ice flow are necessary if glaciology is to make a full contribution to understanding the role of the terrestrial cryosphere in past and future changes of global sea level and climate. This chapter is intended to be mainly a review of developments in the understanding of ice streaming that have taken place since the late 1980s. It builds on reviews provided by Clarke (1987a) and Bentley (1987). The considerable scientific interest in ice streaming is illustrated by the fact that a search for the keyword 'ice stream*' in the ISI Web of Science database at the time of this writing returns >500 publications. Other forms of fast ice flow, such as glacier surging, have been covered by recent reviews (e.g. Harrison & Post, 2003).

Figure 70.1 Satellite image (AVHRR) of the Siple Coast region in West Antarctica (A) and a velocity map (B) for those parts of this region where RADARMAP interferometry data are available (Joughin et al., 2002). The satellite map shows names of the major ice streams and of other important glacio-logical features (source: http://igloo.gsfc.nasa.gov/wais/arti-cles/images/ismap2.jpg). Velocity scale is logarithmic. Velocity data provided by Dr I. Joughin (JPL-Caltech).

Figure 70.1 Satellite image (AVHRR) of the Siple Coast region in West Antarctica (A) and a velocity map (B) for those parts of this region where RADARMAP interferometry data are available (Joughin et al., 2002). The satellite map shows names of the major ice streams and of other important glacio-logical features (source: http://igloo.gsfc.nasa.gov/wais/arti-cles/images/ismap2.jpg). Velocity scale is logarithmic. Velocity data provided by Dr I. Joughin (JPL-Caltech).

Was this article helpful?

0 0

Post a comment