Surface elevation change

The glacial-geological record and the historical record show that, generally, the marginal zone of the ice sheet has thinned and retreated over the past 100yr (Weidick, 1968). Whether this mass loss was compensated, partly or fully, by thickening in the interior is unknown. Although several expeditions have crossed the ice sheet since the late nineteenth century, the earliest data of sufficient precision to permit calculation of surface-elevation change appear to be those of the British North Greenland Expedition (BNGE) that crossed the ice sheet in 1953-1954. Comparing these data with modern surface elevations measured by satellite radar altimetry and airborne laser altimetry shows that between 1954 and 1995, ice thickness along the BNGE traverse did not change much on the northeast slope, whereas the northwest slope thinned at a rate of up to 30cmyr-1 (Paterson & Reeh, 2001), see Fig. 44.1. Repeated height measurements in 1959, 1968 and 1992 along another profile across the ice sheet in central Greenland showed thickening on the west slope between 1959 and 1968, but subsequent thinning between 1968 and 1992 (Moller, 1996), probably reflecting decadal-scale fluctuations in accumulation rate.

Since 1978, satellite-borne radar altimeters have allowed precise monitoring of the surface elevation of interior regions of the ice sheet south of 72°N. In contrast to the previous traverse data, these measurements provided extensive area coverage. Other satellite-borne radar altimeters extended the area coverage of surface-elevation measurements to almost the entire ice sheet, showing the interior regions to be close to balance (within 1cmyr-1) for the past few decades. However, local areas showed quite rapid thickening or thinning (Thomas et al., 2001) (see Plate 44.1b), probably resulting from short-term variations of snow accumulation (McConnell et al., 2000b). Airborne laser-altimeter measurements from the period 1994-1999 confirmed this result for the interior ice-sheet region, but showed substantial thinning along the periphery of the ice sheet (Krabill et al., 2000). Thinning rates of more than 1myr-1 were measured in many outlet-glacier basins in southeast Greenland, in some cases at elevations up to 1500m (see Plate 44.1c), possibly indicating dynamic thinning caused by increased glacier velocities resulting from increased meltwater supply to the glacier bottom.

In order to interpret volume changes derived from repeated accurate measurements of surface-elevation change in terms of mass changes—the relevant quantity influencing global sea

S 20

Figure 44.1 Measured changes in surface elevation between 1994 and 1995 along the traverse route of the British North Greenland Expeditions. For location, see map in Plate 44.1. (From Paterson & Reeh, 2001.)

Longitude (oW)

Figure 44.1 Measured changes in surface elevation between 1994 and 1995 along the traverse route of the British North Greenland Expeditions. For location, see map in Plate 44.1. (From Paterson & Reeh, 2001.)

level—temporal changes of specific mass balance and near-surface temperature, and their influence on surface density and hence ice-sheet volume change, must be understood and accounted for. Changing snow fall over the central accumulation region and changing ice-melt rates in the marginal ablation area will result in immediate surface-elevation (volume) changes that can be translated directly into mass changes by using the density of firn or ice. In the accumulation zone, temporal variations of surface energy balance and accumulation rate will, in addition, cause temporal changes of the rate of densification of the near-surface layers, giving rise to a surface-elevation (volume) change, not accompanied by a corresponding change of mass. Model studies of firn densification (Arthern & Wingham, 1998; Cuffey, 2001) analysed this effect in the dry-snow region of ice sheets where surface melting/refreezing is insignificant. Arthern & Wingham (1998) showed that, in central Greenland, the maximum expected rate of change of surface-elevation due to time-dependent density changes is 3cmyr-1, 1.5cmyr-1 and 0.5cmyr-1 for, respectively, a 10% step-change in surface density, a 10% step-change in accumulation rate, and a 1K step-change in surface temperature. These surface-elevation changes are significant, but small. Much larger surface-elevation changes without a corresponding change of mass may occur in the percolation and wet-snow zones of the ice sheet (for definition of zones in an ice sheet, see Paterson (1994, chapter 2)). Here, the density-depth profile is to a large extent controlled by melting-refreezing processes, by which low-density surface snow is transformed into high-density ice lenses deeper in the snow pack (Paterson, 1994, chapter 2). A temperature change—or rather a change of the energy balance at the surface—will change the amount of surface melting and the subsequent refreezing of ice lenses or formation of superimposed ice. The resulting density change will cause an immediate change of surface elevation, the magnitude of which in the wet snow zone may reach values of 0.1-0.2 myr-1. As first pointed out by Braithwaite et al. (1994), this has important consequences for how observed changes of icesheet surface elevation should be interpreted: a change of surface elevation does not necessarily mean that the mass of the ice sheet has changed. Part of the elevation change, or all of it, may be due to a change of surface-layer density, and therefore may merely reflect a change of ice-sheet volume without a corresponding change of ice-sheet mass. This is illustrated in Plate 44.2, which, for a surface-temperature increase of 1 K over the Greenland Ice Sheet, shows the calculated increase of ice-lens formation (Plate 44.2a), and the part of the surface-elevation change not contributing to the mass change (Plate 44.2b). The accumulation rate is assumed to be unaffected by the temperature change. Moreover, the small effect on snow density of the temperature change considered by Arthern & Wingham (1998) is neglected. The increase of ice-lens formation is derived by means of the melting-refreezing model presented by Reeh (1989). For the total Greenland Ice Sheet, the increased volume loss resulting from a 1 K change of surface temperature amounts to 128km3yr-1, whereas the increased mass loss by runoff amounts to only 96km3yr-1 of ice equivalent, showing that only 75% of the volume change for a 1K warming actually represents a change of mass.

Was this article helpful?

0 0

Post a comment