Glacier mass balance and the global glacier inventory GLIMS

The two methods discussed above for determining ice-sheet mass balance can be, and have been, used to determine the mass balance, not only of ice sheets, but also glaciers. Airborne laser altimetry was combined, for example, with cartographic maps, derived from aerial stereo photogrammetry from the 1950s, to estimate dh/dt over a period as long as 40 yr for some 67 Alaskan glaciers (Arendt et al., 2002). The results suggested a mass wastage much higher than previous estimates, equivalent to twice the loss from the Greenland ice sheet over the same period. A similar approach was taken for the Southern Patagonian Icefield but using a DEM derived from SRTM data instead of laser altimetry and comparing this with historical cartography (Rignot et al., 2003). Again, the results showed thinning rates far greater than expected, based on previous mass-balance estimates. In this study, the accuracy of the SRTM data was found to be around ±2m. Thus, these data could have considerable value for use in elevation change measurements for other low-latitude glaciers, where reasonable historical cartographic data exist. Unfortunately, the northern limit of SRTM was 60°N, missing all of the major glaciated areas of the Arctic such as Svalbard, Greenland, Severnaya Zemlya and so on. For subpolar glaciers, the flux divergence approach has been less useful as there is, in general, no well-defined gate across which the flux can be measured, ice thickness is rarely known and difficult to infer, as is the net mass balance.

Subpolar glaciers are clustered within specific mountain regions. They can range in size from ca. 1 km2 to ca. 1000km2 and can be partially (or even wholly) debris covered, snow covered or bare ice. The problems associated with monitoring the behaviour and mass balance of these types of ice mass are, therefore, often different compared with the ice sheets of Greenland and Antarctica. Thus, although the instruments used may be common, the approach may not be. For example, velocity data can be obtained using either feature tracking or InSAR techniques and these methods have been used to monitor the development of a surge in an Arctic glacier (Luckman et al., 2002). To use these data, however, to determine mass balance requires, as mentioned, the addition of estimates for accumulation, ablation and ice thickness: data that are rarely available for more than a handful of glaciers. Although elevation change estimates have been obtained for some areas, as mentioned above, in many cases no historical cartographic data exist. Consequently, the preferred approach for assessing, indirectly, the mass balance ofthese 'data sparse' regions has been to monitor, remotely, changes in spatial extent and/or parameters such as the equilibrium line altitude (ELA). This is the principal aim of an international programme known as Global Land Ice Monitoring from Space (GLIMS). The objectives of GLIMS are to establish a global inventory of land-ice characteristics, including surface topography, to measure the changes in extent of glaciers and, where possible, their surface velocities (Paul et al., 2002). The project is designed to use data primarily from the ASTER instrument, supplemented with Landsat ETM+ imagery where necessary/appropriate. The key objective is to establish a baseline digital inventory of glacier extent for comparison with measurements at later dates (Paul et al., 2002). It is the first project of its kind, with a global remit, relying on some 23 regional centres (all separately funded), which have a responsibility to manage the data acquisition and processing for specific areas.

The aim is to achieve annual coverage based on four to five acquisitions during a year. To reduce the influence of seasonal effects, sampling should be achieved at around the same time of year. This problem is illustrated in Fig. 73.1, which shows the variability in surface characteristics common to many glaciers and the issues associated with attempting to undertake a consistent estimate of glacier extent. Highly automated, tailored software is required and is being developed by the GLIMS consortium (Paul et al., 2002) with the aim of allowing automatic classification of terrain into glacier and non-glacierized surfaces.

73.5 Concluding remarks

The small sample of applications discussed above are, inevitably, highly selective, but serve to illustrate the immense capabilities of satellite remote sensing for applications related to glaciology, and also some limitations. Great advances have taken place in the past decade, particularly in two key areas. The first is active radar remote sensing, instigated by the launch of ERS-1 in 1991, which carried both a SAR and radar altimeter onboard. The second major advance has been in the use of laser altimetry, both on airborne and, since 2001, satellite platforms. This is a result of advances in laser technology but also, crucially, in platform navigation and pointing accuracy. Both laser and radar sensors still have much more to offer the glaciological community. ICESat was only launched in 2001 and will, despite problems encountered with the laser subsystem, provide a wealth of new data on land-and sea-ice characteristics. CryoSat-2 should in due course also offer many new insights. Both are altimeter missions with a primary focus on ice, and this is a current area of strength in cryospheric remote sensing. In fact, these are the first satellite missions with a specific focus on the cryosphere, highlighting the importance of understanding and monitoring this sensitive component of the Earth system. Beyond these two missions, however, the future is less clear. There are, for example, currently no planned SAR missions that provide the right sort of conditions for optimum use of interferometry over ice and there still remain gaps in the capabilities of current and planned missions for monitoring changes to the cryosphere. Observations of summer sea-ice, for example, are problematic, with large measurement errors in concentration and extent and no agreed 'gold standard' for the most appropriate algorithm. Mass-balance measurements of land ice have improved but the error budget is still too large to determine whether the Antarctic ice sheet (the largest ice mass on the planet by an order of magnitude) is growing or shrinking (Rignot & Thomas, 2002). There is still no satisfactory approach for remotely determining accumulation rates over the vast expanse of the ice sheets although recent research shows promise in helping to address this (Winebrenner et al., 2001). Glacier mass balance from space still provides many challenges, which have no immediate solution and the space agencies are coming under increasing pressure to support ever more specialized satellite missions. Nonetheless, one thing is certain, the next decade will certainly provide many surprising, important and unexpected advances in our ability to observe and understand the cryosphere from space.


I would like to thank Toby Benham for providing the ASTER data used to produce Fig. 73.1 and Ron Kwok for his permission to use some of his material on sea-ice remote sensing and Chris Shuman for providing Fig. 73.3.

0 0

Post a comment