Changes in the Amundsen Sea sector

The shutdown of Ice Stream C is clear evidence that non-linear behaviour is possible in particular glacier basins, but the discovery by Wingham et al. (1998) of surface elevation change in the Amundsen Sea sector of West Antarctica revitalized the debate about a much more serious type of non-linear behaviour that may govern the fate of the West Antarctic Ice Sheet as a whole.

Even when the maps of Antarctica were still blank, Weertman (1974) presented a theoretical analysis of the junction between an ice shelf and an ice sheet grounded below sea level. This suggested that the grounding line could never be stable, but should always be in the process of migrating, either seaward towards the continental shelf edge or inland. Inland migration, which caused acceleration and thinning of the grounded ice, could lead eventually to complete collapse of the ice sheet, as increasingly large areas thinned sufficiently to float and calve off as icebergs. As much of the West Antarctic Ice Sheet is grounded on rock substantially below sea level, it was argued that this ice sheet was most vulnerable to collapse (Thomas et al., 1979). This vulnerability can be seen in the map of the hydrostatic overburden (Fig. 42.9), which indicates the thickness of ice required to keep the ice sheet in contact with the bedrock for each portion of the ice sheet resting on rock below sea level. In particular, the portion of the West Antarctic Ice Sheet feeding the Amundsen Sea is identified as being particularly vulnerable to collapse because it has a particularly low divide resting on bedrock several thousand metres below sea level (Hughes, 1981) and has only narrow ice shelves to buttress the ice sheet.

Figure 42.9 The hydrostatic overburden (free-board) for areas of the Antarctic Ice Sheet grounded on rock that is currently below sea level. Unshaded areas of the ice sheet indicate areas grounded above sea level. In simple terms, the freeboard in any particular area indicates the thickness change required before the ice sheet begins to float. (See www.blackwellpublishing.com/ knight for colour version.)

Figure 42.9 The hydrostatic overburden (free-board) for areas of the Antarctic Ice Sheet grounded on rock that is currently below sea level. Unshaded areas of the ice sheet indicate areas grounded above sea level. In simple terms, the freeboard in any particular area indicates the thickness change required before the ice sheet begins to float. (See www.blackwellpublishing.com/ knight for colour version.)

Table 42.1 Reported basin-scale changes in the Amundsen Sea sector of the Antarctic Ice Sheet

Reference

Data source

Observation

Period of

Conclusion

observation

(Wingham ERS-1, 2 Mean surface lowering of 0.12 ± 0.01myr-1, for the 1992-1996

et al., 1998) altimetry non-coastal parts of the basins from Pine Island

Glacier to Smith Glacier

(Shepherd ERS-1, 2 Confirmed the mean thinning over the Pine Island 1992-1999

et al., 2001) altimetry Glacier basin as 0.11 ± 0.01 myr-1

Apparent basin-scale imbalance possibly resulting from variability in snowfall or possibly from dynamic change Correlation of surface lowering with fast flow indicates dynamic change in all three glaciers

(Wingham ERS-1, 2 Mean surface lowering of 0.12 ± 0.01myr-1, for the 1992-1996

et al., 1998) altimetry non-coastal parts of the basins from Pine Island

Glacier to Smith Glacier

(Shepherd ERS-1, 2 Confirmed the mean thinning over the Pine Island 1992-1999

et al., 2001) altimetry Glacier basin as 0.11 ± 0.01 myr-1

(Shepherd ERS-1, 2 Mean surface lowering of basins from Pine Island 1991-2001

et al., 2002) altimetry Glacier to Smith Glacier, at rate of 0.09 ± 0.02 myr-1.

But this signal dominated by lowering > 1myr-1 of fast flowing ice streams and ca. 50cmyr-1 on their tributaries, and insignificant change on interior of basins

Apparent basin-scale imbalance possibly resulting from variability in snowfall or possibly from dynamic change Correlation of surface lowering with fast flow indicates dynamic change in all three glaciers

Although the evidence for potentially rapid ice-sheet retreat, which exists in the record of global sea level, is undeniable, the general instability of marine ice sheets is not universally accepted. For example, it has been shown that Weertman's (1974) original analysis did not include elements such as ice streams that may act to allow the grounding line to achieve a neutral equilibrium, in which collapse was not inevitable (Hindmarsh & Le Meur, 2001). Although the surface lowering of Pine Island and Thwaites basins was taken by some as evidence for emergent collapse, Wingham et al. (1998) were rightly cautious in their interpretation. They noted that the elevation change was 'not unusual in comparison with the expected snowfall variability' over the period and that a

'snowfall variation is certainly implicated in the volume reduction'.

The debate remains unresolved but the results of Wingham et al. (1998) provided the first clear evidence that a significant portion of the West Antarctic Ice Sheet is thinning at a measurable rate. This, together with the observation by Rignot (1998) that a portion of the grounding line of Pine Island Glacier was retreating, spurred many researchers to focus on this area and investigate the changes, and their causes, in greater detail. The results for the basin-scale imbalance and the particular changes in Pine Island Glacier are summarized in Tables 42.1 & 42.2, and the pattern that has emerged is coherent, if not yet fully interpretable.

Table 42.2 Recently reported changes on Pine Island Glacier

Reference

Data source Observation

(Rignot, 1998) ERS-i InSAR

(Rignot, 2002a)

(Bindschadler, 2002)

ERS-i, 2 altimetry

ERS-i, 2 InSAR

Aerial photography, various satellite imagery

Landsat imagery

ERS-1, 2 SAR Feature-

tracking Landsat feature-tracking, ERS 1, 2 InSAR

Retreat of 1.2 ± 0.3kmyr 1 in the central portion of the grounding line, implying thinning of 3.5 ± 0.9myr-1 Mean thinning of 0.75 ± 0.07 m yr-1, on the trunk, with greatest rates of thinning nearest to the grounding line, and with surrounding areas showing only minor changes An increase of 18 ± 2% in speed of the downstream part of the glacier over 8yr, with increase in the rate of acceleration in 1996-2000, over 1992-1996 No discernable migration of the ice front of the floating portion of Pine Island Glacier, although most recent images show cracks further inland than previously noted.

Progressive retreat of adjacent ice-shelf front Further retreat of central portion of the grounding line, implying thinning of 21 m over 8 yr Progressive widening (ca. 5 km) of the floating portion of Pine Island Glacier into adjacent ice shelf

Thinning of up to 134 m of the ice shelf, adjacent to north of Pine Island Glacier New areas of crevassing close to the grounding zone of Pine Island Glacier A progressive increase in velocity of Pine Island Glacier of ca. 12% over 8 yr

Two roughly equal periods (1974-1987, and

1996-2000) of acceleration (totalling 22%) on the floating and grounded glacier. These periods of acceleration separated by a period of steady flow

Period

1992—1996

1992—1999 1992—2000 1947-2000

1947-2000 1992-2000

1973-1997 (possibly since 1953) 1973-2001

1973-2001 1992-2000

1974-2000

Conclusion

First indication of non-steady behaviour on recent time-scales Glacier thinning too large to be due to snowfall variability, indicating probable dynamic origin

Confirmation of sufficient dynamic variability to implicate acceleration as a cause of thinning Floating portion of the glacier is not retreating dramatically, although continued thinning is noted. Suggests that a significant oceanographic change is possible

Confirms thinning of ice shelf adjacent to Pine Island Glacier and likely oceanographic change

Independent confirmation of earlier result (Rignot et al., 2002)

Refinement of timing of glacier acceleration and indication of stepped change

In summary, it appears that much of the floating portions of Pine Island Glacier and adjacent ice shelves have been thinning for many decades. This thinning appears to be broadly consistent with the rate of grounding-line retreat. There have been some periods of substantial acceleration of the lower reaches of the glacier that are unlikely to be due to a change in snowfall rate, or even a shift to general erosion, and so must be due to dynamic change. Its neighbour, Thwaites Glacier, appears not to have accelerated but to have increased its flux by widening (Rignot et al., 2002), and both Pine Island Glacier and Thwaites have reached a quite distinctly negative mass balance. The most recent estimate for the mean thinning across the entire Amundsen Sea sector suggests a contribution of ca. 0.04 mm of sea-level rise. This is, however, dominated by thinning in the coastal parts of the glaciers, and it is possible that the interior is not thinning at all (Shepherd et al., 2002). This, combined with the fact that changes in the velocity of Pine Island Glacier occurred quite recently and over a very short period, should cause us to be extremely cautious in predicting that this negative mass balance will persist long enough to have a significant impact on sea level.

Thinning is concentrated in the downstream portions of the glaciers and is similar across several neighbouring basins, which implies a cause other than an Ice Stream C-type instability. It is tempting to suspect as a root cause a change in the oceanic boundary condition, such as an increase in the supply of warm water to the ice sheet, which would act regionally at the glacier fronts (Payne et al., 2004; Shepherd et al., 2004). Such a hypothesis would fit in with the thinning of ice shelves adjacent to Pine Island Glacier. We have, however, no time-series of oceanographic measurements that could prove this, and it is difficult to rule out the possibility that the change is part of an ongoing, stepwise Holocene retreat of the ice sheet, or even a sign of emergent collapse. The causes of change in the Amundsen Sea sector remain uncertain, but our increasing understanding of the many potential factors that could cause such change should now make us cautious of interpreting this as evidence of imminent ice-sheet collapse.

Was this article helpful?

0 0

Post a comment