Aim and background

The surface elevation of the Antarctic ice sheet has fluctuated considerably in the past. During Northern Hemisphere glacial periods (such as during the last glacial maximum 20 000 years ago), peripheral areas of the Antarctic ice sheet experienced a considerable expansion and thickening as a result of lowered sea level. In contrast more central areas of the continent may even have been lowered as they experienced a reduced snow accumulation. However, the net result was an increase in Antarctic ice volume during the last glacial maximum (LGM). Although this general pattern is well established, the details of the Antarctic glaciation history remain uncertain in many areas.

Most of the Antarctic ice sheet today ”melts” through calving of icebergs, particularly from ice shelves surrounding much of the continent. A fraction of the loss of ice occurs through sublimation (evaporation directly from ice to water vapour) in blue-ice areas. However, there are very few direct methods that can be used to assess the relative importance of different melting processes during earlier glacial periods, and relatively little is known about the role of sublimation during such times. Therefore one of the key issues in current research on blue ice areas is their evolution through glacial-interglacial times (Bintanja, 1999).

In this project we have studied an area in Scharffenbergbotnen, Heimefrontfjella, Dronning Maud Land, east Antarctica. Here we used the distribution of glacially transported rock material (till), often pushed into moraine ridges on the glacier surface, to reconstruct earlier ice surface elevations. We also show how they can be used in a discussion of former melting processes and the persistence of blue ice areas.

Scharffenbergbotnen is situated in the central part of the Heimefrontfjella nunatak range, which dams the ice sheet on the high polar plateau in the southeast and separates it from the lower coastal portions of the ice sheet. Scharffenbergbotnen itself is an 8 km long valley facing towards the northwest, away from the polar plateau . The valley has steep sides, which rise 150-500 m above the ice surface, and it ends in a l 000 m high back wall. Large areas in and around Scharffenbergbotnen have blue-ice conditions, a result of local wind effects in the high relief terrain, causing erosion of snow and enhanced sublimation in föhn-winds (warm, dry winds) in lee-side positions (Jonsson, 1990; Bintanja, 1999).

Previous studies of the Quaternary glaciation history of this part of Dronning Maud Land indicate that the ice sheet previously reached much higher altitudes than it does today. Patzelt (1988) noted evidence for two former limits of glaciation in Heimefrontfjella. One lower limit was at 100-150 m above the present ice surface, where unweathered erratics (glacially transported boulders of foreign rock type composition) were common, and a second zone was up to 430 m above the present ice surface, where more weathered erratics were found on highly weathered bedrock. Later, Lintinen and Nenonen (1997) confirmed these results by studying till around Scharffenbergbotnen and elsewhere in Heimefrontfjella, and they concluded that the ice sheet during LGM reached up to 100- 200 m above the present ice surface.

The prevalence of blue-ice fields in Antarctica has hitherto primarily been reconstructed using the frequency and age of meteorites, which have been found in great numbers in some blue-ice areas in Antarctica (Cassidy et al., 1992). The reasoning for using meteorite stranding surfaces as tools for reconstructing blue-ice areas is that in areas where meteorites are accumulated, stable glacial conditions with a net loss of ice must have prevailed for as long as the age of the oldest meteorites, otherwise the meteorites would have been transported out of the glacial system. Commonly, terrestrial ages for meteorites reach up to a few 100000 years, although a few meteorites have ages close to and exceeding 1Ma (Bland et al., 2000). These ages give an estimate of the possible prevalence of blue-ice fields. Such reconstructions have been focused on the main meteorite stranding sites such as Alan Hills in Victoria Land and Yamoto Hills in Dronning Maud Land. However, to date no meteorites have been found in Heimefrontfjella and no reconstruction of earlier blue-ice conditions has been attempted in the region.

Methods

To assess the former extension of blue-ice conditions and of ice surface elevations in Scharffenbergbotnen, Heimefrontfjella, we have primarily used the distribution of supraglacial moraines (fields of till on the ice sheet surface) and their continuations up neighbouring slopes and hills. These were first identified prior to the expedition by mapping in black-and-white aerial photographs at a scale of 1:50000. During field work in February 2004 we spent about one week in and around Scharffenbergbotnen. During this time we

• made additional mapping of the areal and vertical distribution of supraglacial debris and tills on adjacent slopes. This was done by visiting as many moraine areas and nunataks as possible. We also climbed hills to measure the vertical limit of till.
• measured debris cover thickness of the supraglacial moraines. We made ̴ 20 test pits by digging into the moraines.
• took notes on supraglacial debris characteristics such as bedrock composition of stoned and boulders, roundness of stones, grain size distribution and presence of striated stones.
• searched for ice flow indicators in bedrock (glacial striae, chatter marks etc.) that could give information on earlier ice flow directions, occurring when the ice surface was higher than it is today.
• took samples of bedrock and boulders for cosmogenic isotope dating. By this technique it is possible to study the time a rock surface has been exposed to cosmic radiation and hence how much time has passed since ice sheet retreat from the surface. These samples have not been analysed yet.

Our research strategy is similar to that used earlier for meteorite fields: If we can show for how long supraglacial moraines has been at the surface of the ice sheet, we will also have a too) to reconstruct the time of stable glaciological conditions in the area.

Results

During field work we found the upper limit of more or less unweathered till and erratics in Scharffenbergbotnen to be at 200-250 m above the present ice surface. This figure is much less outside the valley, noting that the valley floor of the Scharffenbergbotnen is 100-300 m lower than surrounding outside areas. We also note that there are continuous till veneers covering at least the low hills around the research station Svea and the innermost part of the cirque, and onlapping the lower reaches of also higher nunataks, such as Boysenuten. The till veneers appear to be thin, probably less than 1 m on average, as bedrock ourcrops are common both on the slopes and near the crests of the hills. Some of these bedrock outcrops, particularly in cols, have the fresh unweathered appearance of a glacially plucked surface. Other outcrops, however, display signs of extensive surface weathering.

The till veneers of the hills extend down and onto to the surface of the glacier in Scharffenbergbotnen (as a supraglacial debris cover), where they make up supraglacial moraines. These moraines range from a few tens of m2 up to 1.5 km2 . The more detailed surface morphology of most moraines is characterised by arced ridges and furrows, indicating flow structures. Particularly the larger moraines are pitted with sink holes caused by melting of buried ice, commonly reaching 3-4 m in depth. The moraine morphology is almost invariably a result of irregularities in the underlying ice. The thickness of the de bris cover normally varied between 5 cm and 40 cm. Massive ice was encountered below the debris cover. The composition of the de bris cover is fairly consistent, with mixed grain sizes ranging from sand and gravel to boulders. The rocks were normally more or less unweathered, apart from some slight oxidation varnish on some surfaces. The rock types were, as far as could be determined, all of local origin.

Interpretation

We interpret the supraglacial debris as scree deposits from surrounding slopes, which have been glacially transported and accumulated in the inner parts of Scharffenbergbotnen. Because we found glacier ice also in the highest hills in the supraglacial moraine complexes, just below a thin debris cover, we argue that the supraglacial moraines must have been persistent for quite some time. That is because glaciers do not push ice up into such hills. The only reasonable explanation for the glacier ice high up in isolated hills, at least in this case, is that the whole ice surface in the valley was previously higher. Subsequent negative mass balance has later lowered the surrounding ice surface while it has been protected from melting in areas of supraglacial de bris cover. Hence we conclude that the supraglacial moraines have been persistent in the same position at least since the time that the ice surface was >40 m higher.

Furthermore, in some areas the supraglacial debris cover continues seamlessly as a till sheet up the slopes on hills to an altitude 200-250 m above the present ice surface. Hence we agree with Lintinen and Nenonen (1997) that the ice surface was around 200-250 m higher during the last local ice maximum. However it appears that the ice thickening in Scharffenbergbotnen itself was much greater than in surrounding areas. This appears glaciologically sound, because a minor (<50 m) thickening of the outside ice surface would lead to an inflow of ice through a number of new cols. This would result in a greatly increased influx of ice into Scharffenbergbotnen from the southwest and northeast, compared to the single sluggish ice fall of today, and a large increase in ice surface elevation in Scharffenbergbotnen. This is also the time when we infer that the moraine material was brought into its present position in the inner part of Scharffenbergbotnen, by lateral erosion from surrounding slopes.

From these observations and interpretations we argue that the inner part of Scharffenbergbotnen must have been a local ablation area for a long time. The only way to accumulate and keep the supraglacial material in place in Scharffenbergbotnen is to have sustained blue-ice conditions during the last local ice maximum (probably LGM), as well as continuously since. Our preferred reconstruction of ice flow directions and supraglacial moraine distribution in Scharffenbergbotnen at LGM, compared to today’s situation (figure 4 och 5).

In conclusion: On a local scale our results indicate that during the last local ice maximum the ice surface elevation in this Part of Heimefrontfjella was 50-100 m higher on the outside slopes of Scharffenbergbotnen nunataks, and 200-250 m higher in the inner part of the valley. We can also show that supraglacial moraines and their links to surrounding till sheets can be used as a tool to reconstruct not only ice surface elevations but also the former extent of local ablation centres such as blue-ice fields. Such reconstructions have previously only been attempted in meteorite stranding fields. Supraglacial moraines are by far more common than meteorite stranding sites and could probably be exploited more in such reconstructions in other blue-ice areas of Antarctica.