skip to primary navigation skip to content

SPRI Review 2002: Polar Physical Science

Polar Physical Science

Thickness and extent of subglacial till beneath an Antarctic palaeo-ice stream

Marine geophysical records have enabled us to reconstruct for the first time the three-dimensional thickness of the soft sedimentary bed beneath a large fast-moving ice stream. Such subglacial material, whose deformation accounts for the rapid flow of many ice streams, is usually hidden under several kilometres of ice in the case of modern ice sheets. We have used geophysical and geological data on the morphology of the sea floor, and the acoustic, sedimentological and geotechnical properties of the sediments directly beneath it, to infer the presence of an ice stream the last time ice expanded to inundate the 370 km-long Marguerite Trough offshore of the Antarctic Peninsula about 18,000 years ago. Fast flow of this palaeo-ice stream is implied by elongate, streamlined lineations in sea-bed sediments, orientated parallel to the long-axis of the trough. Directly beneath the sea floor, a layer of low shear-strength unsorted sediment, averaging 4.6 m and up to 19 m in thickness, is observed using reflectors on shallow-acoustic records. This layer, with its streamlined upper surface and underlain by overconsolidated glacial sediment, is interpreted as the product of subglacial deformation. This work was undertaken jointly with Dr Carol Pudsey (BAS) under the NERC Antarctic Funding Initiative.

Julian Dowdeswell, Colm Ó Cofaigh

Ice shelf disintegration and climate warming

Since 1980, ice shelves of the Antarctic Peninsula have retreated by 550 km2 each year in tandem with a regional climate warming, and it is widely regarded that these events are related. For several decades this topic has been the subject of much scientific debate, and has fuelled speculation as to the sustainability of the remaining ice shelves and the source of the regional climate change.

A decade of satellite radar altimeter measurements has revealed a widespread surface lowering of the Larsen Ice Shelf. The lowering may reflect a 7% thinning of basal ice as a result of enhanced ocean melting during a decade in which large sections of the ice shelf disintegrated. The thinning preceded the 2002 collapse of the Larsen-B section by at least 9 years, and has progressively weakened the remaining Larsen-C section. If the thinning were to continue unabated for a further ~ 40 years, sections of the Larsen-C ice shelf would approach the thickness of the Larsen-B at the time of its collapse. Basal ice-shelf melting may provide the simple link between regional climate warming and the successive disintegration of Larsen Ice Shelf sections.

Andrew Shepherd

West Antarctic ice thinning and sea level rise

Whilst the Antarctic continent as a whole has had only a modest impact upon global sea level during the past century, the Amundsen Sea sector of western Antarctica has experienced a widespread thinning of grounded ice. A combination of satellite radar-altimetry and interferometry has shown that that ice thinning was concentrated along the fast-flowing trunks of the Pine Island, Thwaites and Smith glaciers - the largest ice streams in western Antarctica. Since 1992, the glaciers have thinned by more than 10 m and have retreated inland by up to 10 km, losing over 150 km3 of ice to the ocean.

The pattern of thinning affects over 10,000 km2 of inland ice, and is associated with ice dynamics. The presumed longevity of the thinning raises the important question as to its likely evolution, and it is possible that such a retreat could signal an accelerating and irreversible process. With enough ice to raise global sea levels by more than a metre, understanding the future rate of ice discharge from this sector of Antarctica is of pressing concern.

Andrew Shepherd

The physiography of Antarctic subglacial lakes

In addition to the well-known Lake Vostok, more than 70 lakes have been identified beneath the Antarctic Ice Sheet using mirror-like reflectors observed on airborne radar records. The total volume of water stored in lakes beneath the Antarctic Ice Sheet is between 4,000 and 12,000 km3. Almost 60% of lakes are found within 200 km of an ice divide and only about 15% are located >500 km distant. Bedrock topography of the ice-sheet interior is characterised by large subglacial basins separated by mountain ranges. Many lakes occur in areas of relatively low bed relief, in and on the margins of subglacial basins. First, there are lakes located where subglacial topography is relatively subdued, often near the centre of subglacial basins. Secondly, some lakes occur in significant topographic depressions, closer to subglacial basin margins. Lakes are also found perched on subglacial mountainsides. Sixteen lakes are located close to the transition to enhanced ice-sheet flow. Warm-based fast-flowing ice streams provide a possible route by which subglacial lakes may establish a hydrological connection with the ice-sheet margin. At a continental scale, the locations of Antarctic subglacial lakes match the modelled distribution of pressure melting at the ice-sheet bed. This work was undertaken jointly with M.J. Siegert, Bristol University.

Julian Dowdeswell

Convective chimneys in the Greenland Sea

Until recently it was believed that winter convection in the Greenland Sea, which helps drive the global thermohaline circulation, occurred in the form of plumes of descending dense water, generated by the salt rejected by newly formed sea ice. The effect of global warming was to reduce ice formation, and thus to weaken convection. However, during field work in 2001 it was discovered that the preferred mechanism may be an extraordinary feature known as a chimney, a rotating vortex of uniform water extending from the surface to 2500 m depth. The chimney, discovered in a cruise by RV Lance funded by the EU under the CONVECTION project, was only 20 km in diameter. Later cruises, in summer 2001 and winter 2002, showed that the chimney was remarkably long-lived and stationary at 75ºN 0ºE, properties which at present are unexplained. Each summer the chimney becomes capped by fresher surface water, only to open up again the next winter. Further work is planned for May 2003.

Peter Wadhams

First use of an AUV under Arctic sea ice

In winter 2002, the Danish Maridan Martin autonomous underwater vehicle (AUV) was deployed in the Arctic seas, and obtained the first under-ice sonar profiles from an AUV under Arctic sea ice. The vehicle was deployed from the Norwegian RV Lance in the East Greenland pack ice, and made a number of runs under the ice. It was equipped with an upward looking side-scan sonar and water-conductivity and temperature sensor. The AUV was able to keep to a well-defined overlapping grid of tracks through the use of an acoustic tracking and control system. The deployment provided valuable operational experience for future use of the more sophisticated UK Autosub AUV under sea ice in the Weddell Sea region of Antarctic and the Greenland Sea.

Peter Wadhams

Ice-surface albedo and melt rates

Surface conditions, such as snow cover and albedo (surface reflectance) form one of the key controls which determine glacier and ice sheet melt rates, as they influence the amount of energy, especially sunlight, absorbed by the surface. A sophisticated mathematical model of the energy inputs to glacier and ice sheet surfaces has been developed, and a key focus of research over the last year has been improving the representation of snow cover within the model. Previous models have only allowed snow cover to vary slowly over glacier surfaces but, in reality, snow depth can vary considerably over quite short distances - metres to tens of metres. Including this variability should improve estimates of melt rates from glaciers and ice sheets. Research efforts have focused on the possibility that this variability may be fractal, and field work has shown that this is the case. This is an important result, as a wide variety of fractal techniques are available to simulate variable surfaces from relatively simple field measurements. Current research is focused on developing fractal snow cover models, and investigating their impacts on glacier and ice sheet melting. This will be supplemented by a field visit to the Arctic to acquire airborne remotely-sensed data of small scale surface topography and albedo over an Arctic glacier, following a successful application to the NERC Airborne Remote Sensing Facility.

Neil Arnold

Optical properties of Arctic lichens

Arctic and sub-arctic tundra vegetation, north of the treeline, consists largely of grasses and sedges, mosses, lichens and dwarf shrubs. In many areas this vegetation is out of equilibrium, subjected to pressures from global climate change, air and ground pollution and other human disturbances, and from grazing by reindeer. In particular, it is probable that in large areas of the Eurasian north lichen-dominated tundra is being displaced by grasses and shrubs. This could impact the global climate system through its effect on the carbon and water cycles and on the Earth's albedo. Studies of the changes that are taking place in the distribution of high-latitude vegetation depend on the analysis of satellite imagery. Accurate interpretation requires detailed knowledge of the optical reflectance properties of the vegetation cover. Until now, such information has been conspicuously lacking for most lichen genera. However, new experiments based in Arctic Sweden have begun to provide reflectance measurements with an unprecedentedly high level of detail: the spectra are measured from blue light, at a wavelength of around 400 nanometres, well into the infrared region at around 2400 nanometres, in 2000 steps. The Swedish test site provides extremely clean, undisturbed conditions. In future, studies will be extended to polluted areas of the Russian arctic to investigate the effect of physiological stress on the optical properties. These results will bring us much closer to the goal of being able to assess the health of an arctic ecosystem by analysis of satellite images.

Gareth Rees

Measuring the volumes of large ice masses

The ability to measure changes in the volumes of large ice masses (glaciers, ice caps and ice sheets) is a high priority in monitoring potential global climate change. Traditionally this task is undertaken by fieldwork, which is slow and labour-intensive, so it is particularly desirable to develop techniques based on the use of satellite data. One recently developed technique, capable of remarkably high precision, is called Interferometric Synthetic Aperture Radar (InSAR). However, it is technically challenging to implement and not yet suitable as a tool for routine monitoring. Consequently, it is likely that for some years to come it will be necessary to integrate height information about large ice masses from a number of sources, including InSAR, field-based GPS measurements, stereophotography, radar altimetry, radio echo sounding, laser profiling and photoclinometry. All of these techniques have different characteristics, so we are investigating optimum techniques for assimilating data from different sources into a consistent digital elevation model (DEM). By recompiling such DEMs over time, it will be possible to assess changes in the volume of the ice mass and hence to link these changes to climate models.

Gareth Rees

3-D patterns of stress and velocity in glaciers

An integrated field-work and modelling strategy is being used to study how basal conditions affect the 3-D distribution of stress and velocity within a temperate Alpine glacier (Haut Glacier d'Arolla, Switzerland). In the field, we monitored spatial and temporal variations in basal water pressure, sediment thickness, texture and strength, surface motion, internal deformation, sliding and subglacial sediment deformation using down-borehole instruments and terrestrial surveying techniques. We found distinct patterns of surface, internal and basal motion that varied between spring, summer, and autumn/winter, which reflect patterns of basal water pressure and sediment characteristics. These, in turn, are influenced by the proximity to subglacial drainage axes. For example, during the spring, the glacier surface speeds up from ~2cm d-1 to >10cm d-1 over short periods of a few days. The zone of maximum surface velocity shifts from the centre of the glacier towards the major drainage axes where water pressure fluctuations are greatest and sediments are relatively thin and coarse grained. The relative importance of basal motion to surface motion increases during these 'spring events,' particularly towards the drainage axes and less so away from them. The field data have been used to drive and test a 3-D glacier flow model, which can reproduce the spring, summer, autumn/winter, and annual patterns and magnitudes of movement very accurately. This work is being undertaken in collaboration with Bryn Hubbard (Aberystwyth), Pete Nienow (Glasgow), Doug Mair (Aberdeen), Urs Fischer (ETH, Zurich) and Alun Hubbard (Edinburgh).

Ian Willis

Hydrology and dynamics of polythermal glaciers

The hydrology and dynamics of Midre Lovénbreen, Svalbard, are being studied using a combination of field-based and photogrammetric methods. Digital Elevation Models (DEMs) of the glacier surface have been constructed for 1977 and 1995. A DEM of the glacier bed has been produced from radar data. These DEMs have been used to map subglacial hydraulic potential for various assumptions about steady-state subglacial water pressure, and these have been used to construct theoretically the overall structure of the subglacial drainage network and therefore the position of major hydrological pathways under the glacier. Results suggest that the position of the terminus stream is sensitively dependent on subglacial water pressures. Observations over the last decade of the main terminus stream suggest that switches in its position have occurred as a result of variations in the rate of rain and melt water delivery to the glacier bed and therefore subglacial water pressures. Measurements of surface velocity variations during the summer suggest that spatial variations in glacier dynamics are also controlled by the presence of basal water. Temporal fluctuations in glacier dynamics occur in response to weather related surface inputs. Velocity variations may be locally forced by subglacial water pressure fluctuations or non-locally forced through changes in the longitudinal stress gradients. This work is being undertaken in collaboration with our former PhD student David Rippin (Bristol) and with Andy Hodson (Sheffield).

Ian Willis, Neil Arnold