Ole Humlum & Michael Houmark-Nielsen: High Deglaciation Rates in Denmark During the Late Weichselian - Implications for the Palaeoenvironment. Geografisk Tidsskrift, Danish Journal of Geography 94:xx-xx. Copenhagen, Dec. 1994.

Available geologic evidence suggests that the mean deglaciation rate in Denmark 18,000-17,000 calendar years BP was at least about 100 m/year, probably requiring a total vertical ice ablation of 30-35 m/year. This ablation value is large when compared to the amount of ice ablation that could be expected on physical grounds. The reasons for this apparent discrepancy are discussed and factors such as glacier bed strength characteristics, presence of marginal water bodies and occurrence of strong catabatic winds are suggested as environmental phenomena that should be taken into consideration when formulating dynamic deglaciation models and reconstructing the Late Weichselian palaeoenvironment in Denmark; climate alone does not explain the observed patterns and rates of deglaciation.

Keywords: Late Weichselian, deglaciation, ablation, Denmark

Ole Humlum, Associate professor, Institute of Geography, University
of Copenhagen, Øster Voldgade 10, DK 1350 Copenhagen
K, Denmark.

Michael Houmark-Nielsen, Associate professor, Institute of
Geology, University of Copenhagen, Øster Voldgade 10, DK
1350 Copenhagen K, Denmark.

The causes for the growth and collapse of the large Pleistocene ice sheets are topics of intense scientific debate, and the amount of new data and ideas has been large during the last twenty years. Significant progress has also been made in the modelling of land ice masses as well as in the understanding of the role they play in the climate system, and vice versa (see, e.g. Manabe and Broccoli 1985). However, especially concerning the deglaciation period, the amount of new stratigraphic information such as datings has been substantial, and deglaciation isochrone maps have been published both for the Weichselian North European Ice Sheet and the Wisconsin Laurentide North American Ice Sheet (see, e.g. Andrews 1973, Berglund 1979, Lagerlund et al. 1983, Boulton et al. 1985, Dyke and Prest 1987, Lagerlund and Houmark-Nielsen 1993). In what follows, published conventional 14_C ages have been corrected according to the results obtained by Bard et al. (1990, 1993) from studies on corals. This is important for ages higher than 9kaßP (¹⁴C years), which must be corrected toward higher ages. The correction factor is about Ikaat "old" 14_C ages of about 10 ka BP, increasing to 3.5 ka for "old" 14_C ages of about 20 ka BP. Figure 2 illustrates the 14_C-year to calendar-year conversion for the Late Weichselian in Denmark. In what follows, ages will be given as calendar years.

The stream of new information on Late Weichselian deglaciation has resulted in several important analyses and discussions on the palaeoenvironment during the general deglaciation period from 22 ka BP to 8 ka BP (e.g. Andrews and Barry 1978, Berthelsen 1979, Berglund 1979, Lagerlund et al. 1983,Boultonetal. 1985, Dyke and Prest 1987, Houmark-Nielsen 1987, Lidmar-Bergstrom et al. 1991, Lagerlund and Houmark-Nielsen 1993). One common denominator, however, emerges from the steady flow of new data: both in North America and in Northern Europe the associated frontal ice-recessional rates were surprisingly high during this period; at least when compared to rates observed at modem glaciers. Several years ago Andrews (1973) drew attention to this problem for the Wisconsin Laurentide Ice Sheet, and it is the purpose of the present paper to look into the analogue situation for the SW-part of Scandinavia during the Weichselian deglaciation. In the present paper we concentrate on Late Weichselian deglaciation rates and the different mechanisms that might have caused the rather rapid collapse of the Danish sector of the Weichselian ice sheet. The analysis will be attempted by applying the improved possibility of obtaining numeric 3D-reconstructions on the surface form of former ice sheets, given their former margins and the physical character of the substratum. Also results obtained from insolation calculations as well as empirical knowledge on ablation rates from modern glacier surfaces obtained from proper mass balance measurements will be used.

Late Weichselian Deglaciation of Denmark

From a palaeo-glaciological point of view, the Danish sector of the Weichselian North European Ice Sheet was dynamically complex. This was the case because a southerly flow of ice from Norway here coalesced with ice masses nourished by vast accumulation areas northeast and east of Denmark, over the central part of Scandinavia and the Baltic. Furthermore, during most of the Weichselian the Danish area was probably close to the contemporary sea level. In the Late Weichselian the ice sheet came

in direct contact with the ocean within the area from time to time (Lagerlund and Houmark-Nielsen 1993). The climate must have undergone large variations, being coldarid at times when the extension of the ice sheet prevented the ocean from having access to the Danish area, and cold-maritime at times when the ocean had access to the area. Naturally, the overall mass balance characteristics of the Danish ice sheet sector must have changed according to this.

An outline of the main features of the Late Weichselian deglaciation of Denmark is shown in Fig.l, compiled from various sources (Mangerud and Berglund 1978, Berglund 1979, Lagerlund et al. 1983, Lagerlund 1987: Houmark-Nielsen and Lagerlund 1987, Ringberg 1988,

Houmark-Nielsen 1987, Lidmar-Bergström et al. 1991. Smed 1992, Lagerlund and Houmark-Nielsen 1993). The conversion from conventional 14_C years to calendar years is shown in Figure 2. Around 22 ka BP global sea level was near its eustatic minimum, and off the North European seaboard, the shoreline was probably 110-120 m below present sea level on the outer shelf (Clarck and Kingle 1979, Kjemperud 1986). At this time the North European Ice Sheet achieved its overall maximum size, having its south-western margin standing along the Main Stationary Line (MSL) in central Jutland, Denmark (Fig. 1). From 22 ka BP the ice sheet melted back, although interrupted by several important readvances. and shortly before 17 ka BP the last active ice lost contact with the present southeasternDenmark.

easternDenmark.The geomorphology of Denmark indicatesthat stagnated ice masses were detached from the ice sheet during the deglaciation at many localities, while the ice sheet melted back by way of frontal deglaciation at other localities.

During the early part of the Late Weichselian, the ice sheet melted back towards north and east (Fig.l). This pattern was, however, interrupted shortly after 18.5 ka BP. A major readvance from the east and southeast now again brought active ice close to the Main Stationary Line in southern and central Jutland at 18 ka BP (EJ, Fig.l). This overall maximum position was even overridden a few places in southern Jutland.

At the time of this major Baltic readvance (18 ka BP), the Norwegian ice no longer had contact with the present Danish land area, probably due to calving induced by the rising global sea level, and the ocean now covered the northern part of present Denmark, and was in contact with the ice sheet margin in the southern part of the present Kattegat Sea area. This ocean extension is known as the Younger Yoldia Sea. It transgressed northern Jutland slightly before 18.5 ka BP, and had its maximum extension around 16.5 ka BP. The present Swedish west coast was transgressed around 17 ka BP (Fig. 1). Without any doubt, this significant geographical change was associated with environmental changes also, probably in the direction of a warmer and certainly moister summer climate. This assumed environmental change may be the climatic background for the following activity of the Baltic ice sheet sector due to increased precipitation. Lagerlund (1987) even suggested the evolution of a number of marginal ice domes in order to explain the Late Weichselian Baltic ice advances over Denmark and southern Sweden. At the same time the rising sea level by way of calving prevented the Norwegian ice from reaching the Danish area, even though this part of the ice sheet may also have experienced increased precipitation. In what follows, an overall examination of the deglaciation rates and mechanisms within the Danish sector in the Late Weichselian will be performed.

Late Weichselian Deglaciation Rates

Consider the palaeoglaciological consequences of Figure
1 for a moment. Mean deglaciation rates vary from 60 to

200 ma'^m/year), lowest in the northern area, which was affected by the Younger Toldia transgression, and highest in the southeastern part of Denmark, where terrain was generally above sea level. Obviously, the dates and ice sheet margins given in Fig. 1 are only approximate, but the corresponding mean deglaciation rates should neverthelessbe within the correct order of magnitude. On the other hand, these deglaciation rates can be minimum values only, as the geomorphology indicates the presence of interveiningreadvance lines. This is why, real deglaciation rates may well have been in excess of 200 ma"¹ in southern Denmark. As a conservative estimate, a mean frontal retreat of 100 ma"¹ is adopted as representative for the period 18-17 ka BP in the southeastern part of the Danish area.

Modern Deglaciation Rates

Deglaciation rates measured at modern glaciers within the present century are usually of an order of magnitude smaller than the Danish Late Weichselian conservative estimate of 100 ma"¹. Especially from 1930 to 1960 AD was a time of what is usually seen as a period of substantial back wasting of modern glaciers, but the associated frontal retreat rates were usually within a range from 5 to 15 ma"¹ (see, e.g. ÖAV-Mitteilungen, various dates). These modern deglaciation rates are far below the estimated Danish Late Weichselian rates. The same applies to modern rates obtained from the ice cap Myrdalsjökull in Iceland, where maximum mid-century deglaciation rates were about 10 to 25 ma"¹ (Krüger and Humlum 1981).

At localities where a glacier terminates in a large water body deglaciation rates of about 100 ma"¹ are normal. Recent investigations by one of us (Humlum, in prep.) at maritime outlet-glaciers along the present ice sheet margin in central West Greenland have demonstrated 20thcentury rates of both advances and retreats of 50-200 ma"¹.

Late Weichselian Surface Ablation

The infrerred high Late Weichselian deglaciation rates in southern Denmark (terrestial environment) in this way point toward a discrepancy hard to explain for a nonmaritime environment. We will, therefore, now consider the amount of vertical ice ablation corresponding to the horizontal deglaciation rates.

To estimate the amount of vertical net ice wastage associated
with frontal deglaciation rates such as the above
conservative 100 ma-1, it is necessary to make some estimates
concerning the surface form of the Late Weichselian
ice sheet in South-western Scandinavia.

Reconstructions of the Weichselian and Wisconsin ice sheets have been attempted on several occasions (see, e.g. Andrews 1982, Boulton et al. 1985). Notably most theoretically based reconstructions tend to generate symmetrical ice covers with large ice thicknesses achieved short distances behind the ice sheet margin, as do empirical reconstructions based on the two modern ice sheets in Greenland and in Antarctica (see, e.g. Paterson 1972, Sugden 1977, Humlum et al. 1978, Denton and Hughes 1981, Andrews 1982, Boulton et al. 1985, Hughes 1985). In contrast to this Peltier's (1981) reconstruction of the Laurentide Ice Sheet based on uplift data suggests generally thinner ice and especially thin ice cover over the prairies in North America. Also observations made by different workers on ice marginal features (see, e.g. Mathews 1974, Clayton et al. 1985, Beget 1987) indicate this to have been the case at certain other localities in North America. Reeh et al. (1983), Boulton et al. (1985) and Fisher et al. (l985) then introduced deformable beds into the ice flow modelling exercise. The introduction of deformable glacier beds into glacier modelling has the general effect of producing areas of thinner ice and removes much of the geometric ice sheet symmetry known from earlier reconstructions. In the present case, a modified version of the model developed by Fisher et al. (198 5) was used to reconstruct the Late Weichselian North European Ice Sheet around 17.5 ka BP. The software represents a refined version of the model developed by Reeh (1982), which for the Greenland Ice Sheet was able to reproduce successfully all the present major ice divides, ice streams and centres. Also flow line trajectories were accurate and the model's surface elevations were at most in error by 10% for the Greenland case. The graphic result of the Danish Late Weichselian exercise is shown in Fig. 3. The model uses a simple ideal plastic ice rheology that is rather insensitive to unknown parameters such as, e.g. mass balance parameters and takes as input only the margins of the former ice sheet, the present day topography, and an assumed yield shear stress for the glacier ice.

Only the south-western part of the Late Weichselian Ice Sheet is shown in Fig. 3. Glacier surface gradients are visibly larger along the Swedish west coast than over the Danish area, which is the result of an assumed difference as to basal yield shear stress, for which a range of 0.5 to 1.5 bars was adopted as a "normal" range (1 bar = 10⁵Nm"²) in the case of Sweden, where the glacier bed consisted of solid bedrock. Lower values of 0.1 to 0.5 bars were adopted for Denmark and the areas south of Sweden, due to a cover of assumed deformable sediments found there.

According to the model output (Fig. 3), ice surface altitudesof500 m a.s.l. were generally reached 15-20 km behind the glacier margin in the Danish area around 17.5 ka BP, and about 200 km from the margin altitudes of 1,000 m a.s.l. were reached. In contrast to this, calculated ice surface altitudes of 1,000 m a.s.l. were found only 50 km east of the glacier margin along the Swedish west coast, due to the higher basal shear stresses assumed for this area. Correcting for isostasy, and taking the present Danish 17.5 ka BP glacier bed as an almost horizontal surface, an ice surface altitude of 1,000 m approximately corresponds to a simultaneous ice thickness of about 1,500 m over the south-eastern part of Denmark at 17.5

must be turned towards the conditions at the 17.5 ka BP
ice sheet margin, which certainly was part of the ablation
area throughout the period considered.

Glacier surface profiles have been measured on the modern Icelandic ice cap Myrdalsjökull (Humlum et al. 1978). At the site of the measurement this ice cap rests on a glacier bed of loose (deformable?) sediments, and from this point of view probably represents a good analogy to the Late Weichselian ice sheet margin in the Danish area. These observations show that in order to generate a frontal recession rate of 100 m, 25-30 m of ice must be lost vertically from the glacier surface. The existence of a sloping glacier surface above the Late Weichselian terminus must have generated ice flow from the accumulation area towards the ablation area, and the above ice loss value therefore only represents a minimum ablation requirement during the Late Weichselian in Denmark. Furthermore, each winter a seasonal snow cover must have existed on the former ice sheet surface below the equilibrium line. This snow layer would first have to be eliminated by melting and evaporation each spring before any ablation could take place from the solid glacier surface below. As snow is a high-albedo feature, this would tend to narrow the length of the potential solid ice ablation season. Also occurrences of summer snowfalls would have this effect.

By way of both modelling and field observations and by introducing the concepts of a seasonal winter snow cover and active ice flow, it is suggested that a total ablation of at least 30 to 35 m of ice per year was required to generate a frontal deglaciation rate of about 100 ma"¹ during the Late Weichselian in Denmark, provided the existence of a terrestial environment. The required total surface ablation would probably increase to about 50 ma"¹, if still higher frontal deglaciation rates such as 200 ma"¹ were accepted, as discussed above.

Main Controls on Ice Ablation

The loss of mass from any point on a glacier surface by way of melting and evaporation depends upon the specific surface energy balance, which is composed of several components. At least four sources of energy are important for ablation at most modern glaciers in a terrestial environment: (1) short-wave (solar) radiation, (2) long-wave (terrestrial) radiation, (3) turbulent energy fluxes, and (4) refreezing of melt water.

The relative importance of these processes varies from glacier to glacier, but the short-wave radiation (1) balance is found to represent the most important factor at most modern glaciers (see, e.g. Föhn 1973 and Braithwaite 1981). In the ablation area of the modern Greenland Ice Sheet radiation, is found to supply about two-thirds of the energy used for ablation and turbulent fluxes supply most of the remaining one-third (Braithwaite and Olesen 1984, 1990). The importance of refreezing melt water (4) still needs clarification, but can be of considerable importance for the glacier mass balance, especially around the equilibrium line (Bøggild 1991). The latent heat flux is small on average but this is the result of substantial fluctuations between negative and positive daily values, i.e. evaporation and condensation, respectively, which nearly cancel out over longer periods (Braithwaite and Olesen 1990).

Calculated Late Weichselian Ablation

As an experiment, and in order to test the realism of Late Weichselian total ablation rates of 30 to 35 ma"¹, the midsummer short-wave net radiation balance was calculated for each of the segments of the 3D model of the Late Weichselian Ice Sheet shown in Fig. 3. Necessary inputs in the model were aspect and slope of all individual terrain elements (N=1288), the overall surface albedo for each terrain segment, the time (17.5 ka BP), the date (July 1) and degree of cloud cover (0-100%). Since about 20 ka BP, the mean annual solar radiation at the top of the atmosphere for mid- and high latitudes of both hemispheres has been above the Quaternary average, reaching a maximum of about 9 ka BP. The present extra atmospheric solar radiation (1367 Wm'V¹ or 236.8 Jm'V¹) is about the average for the entire Quaternary, and is comparable to that of the last glacial maximum at around 22 ka BP (Ohmura 1987), although the climate is milder now. The variation of the solar radiation has an apparent period of about 22 ka due to the precession of the equinoxes. For purpose of the present calculation experiment, the extra atmospheric radiation at 17.5 ka BP was assumed to be 1374Wm"².

Albedo values for different terrain surface types were obtained from standard tables in literature, from satellite imagery analysis (Kamper et al. 1991) and from personal (Humlum, unpublished) field work in Greenland. The albedo was taken as being 0.8 for the snow covered areas above the estimated ELA (assumed at about 1,000 m a.s.l), and 0.35 for the ablation area below the equilibrium line.

For the date chosen the calculation was run iteratively using time steps of one hour, and from this, mean daily net short-wave radiation values were calculated for each terrainsegment. For each time step in the calculation the

software automatically checks for a topographic-induced shadow. The calculations were run for different values of cloud cover (0%, 25%, 50%, 75% and 100%). Figures 3 and 2 show graphic and numerical outputs of the calculation for clear-sky conditions (cloud cover 0%) on the mean short-wave net radiation, at 17.5 ka BP, July 1. These input values were chosen to obtain information on the maximum short-wave net radiation during the ablation season in the Late Weichselian. Figure 3 shows the overall areal distribution of net radiation values, while Fig.4 shows the corresponding altitudinal distribution.

From both Fig. 3 and 4, thus representing a sunny midsummer day in Denmark during the Late Weichselian, average values for the mean daily short-wave net radiation are seen to have been within a probable range from 240 Wm"² to 250 Wm"² in the lowermost part of the ablation zone. Overcast conditions (cloud cover 100%) reduce the short-wave balance to about 150 Wm"². Adopting the high value of 250 Wm"² to represent a maximum mean daily summer value, and extending this to the months of June, July and August, representing the Late Weichselian ablation season, we calculate a total maximum short-wave energy input of about 1950 MWm"² or 338 MJm-2 for the whole ablation period. If all of this energy is used for melting ice, about 4.5 m of ice would then have been lost each summer from the glacier surface near the ice sheet margin. As stated above, this clearly represents a maximum value for ablation caused by shortwave energy input, and a more realistic value may have been about 3 m of ice, when taking periods of overcast conditions, etc., into consideration. On the other hand, also other mechanisms are active ablating ice, although usually being subordinate to the short-wave energy input (see discussion above), and the above calculation should only be taken as a rough indication of an order of magnitude.

Summing up, the calculation experiment indicates the Late Weichselian mean yearly total ice ablation in Denmark at 17.5 ka BP probably to be within a range from 4to 6 m ice per year, including all mentioned ablation mechanisms. Almost certainly, it must have been less than 10 ma"¹. Introducing the effect of active ice flow towards the glacier terminus and the presence of a seasonal snow cover probably reduce the yearly net ablation to a realistic value below 5 m of ice.

A yearly net loss of about 5 m of ice (total ablation less than 10 ma"¹) from the glacier surface is far below the above estimate of 25-30 ma"¹ (total ablation 30-35 ma"¹) made on geologic grounds, and suggests that the Late Weichselian geologic estimate should be tested against other independent criteria. This can be achieved by way of Fig. 5. The three graphs in the figure represent empirical relations concerning snow/ice ablation used by the US Army Corps of Engineers, incorporating the effect of both air temperature and wind velocity. According to these relations, about 10 m of ice would today melt per year in Denmark (B°C, using a mean wind velocity of about 5 ms"¹). This estimate on modern potential ice ablation is thus far below the Danish Late Weichselian estimate. Also results obtained by Braithwaite and Olesen (1984) indicate the presence of an substantial and interesting discrepancy between the estimated Late Weichselian ablation rates and the energy input available then.

Dicussion

From the above analysis we have apparantly identified a major unsolved problem concerning the Late Weichselian deglaciation of Denmark and adjoining areas. On one hand the knowledge of Late Weichselian ice-marginal lines and associated datings are well founded. On the other hand data on the then available energy input suggest that estimates of Late Weichselian deglaciation rates derived from geologic knowledge are much too high.

Both the geologic knowledge and the energy calculations must be considered sound and internally robust, and we do not consider it likely that major errors will be found within either approach. Assuming this to represent a correct assessment, the reason for the apparent deglaciation paradox should probably be sought within our overall conception of the Danish Late Weichselian environment.

Several deglaciation models are presently at hand ( see, e.g. discussions in Berthelsen 1979, Berglund 1979, Lagerlund et al. 1983, Lagerlund 1987, Houmark-Nielsen 1987, Lidmar-Bergström et al. 1991), but most models appear to operate with the implied condition that deglaciation generally took place within a terrestrial environment, with no further specification of environmental characteristics. However, three environmental factors may deserve consideration in order to gain insight into the reasons for the above apparent deglaciation paradox.

(1) Surface form of the ice sheet. The influence of different shear stress ranges at the glacier base was shortly outlined earlier in this paper. Assuming lower than normal shear stresses due to a deformable substratum will produce a flatter ice sheet surface profile, requiring less net ablation in order to satisfy the requirement of frontal deglaciation rates of about 100 ma"¹. This point was taken into consideration during calculation of the 3D-model (Fig.3), assuming yield stresses from 0.1 to 0.5 bar for the

ice sheet covering Denmark and areas south of Sweden. We do not find a further reduction to values below 0.1 bar realistic. Several field observations from Denmark show that considerable friction must have existed between the Weichselian ice sheet and the substratum. Signs of this are, e.g. push moraines, thrust features, glacial flutes, drumlins and glacial striae on the surface of large rock fragments and solid bedrock (see, e.g. Krüger 1970, Berthelsen 1978, Humlum 1979, Houmark-Nielsen 1987). Experimentally reducing the slope of the marginal ice sheet surface profile to about half of that shown in Fig. 3, still results in a total ablation requirement of impressive 15-20 ma"¹ at the glacier margin.

(2) Marginal water bodies. Isostatic depression of the crust below the ice sheet must have been considerable. Both on theoretical and empirical grounds, it has been suggested that the isostatic depressed region would extend 50 to 100 km beyond the margin of a huge ice sheet such as the Weichselian in North Europe, due to flexure of the crust (see, e.g. Brotchie and Silvester 1969, Birchfield and Grumbine 1985, Hyde and Peltier 1985). Near the glacier margin the depression probably amounted from about 50 to 100 m, thus producing a regional topographic low along the ice sheet margin. If so, melt water must have accumulated in several places within this depression, producing numerous large and small 'water bodies (Fig.6 and 7), tending to migrate with the ice sheet margin as this undergoes advances and retreats. At certain times, the ocean itself may even have transgressed into this isostatic induced topographic low. As the outflow of material in the asthenosphere takes time, the isostatic adjustment must lag the change in load upon the crust. During periods of back wasting, when the ice sheet margin retreated across still depressed regions, the tendency towards generating marginal lakes must have been especially pronounced. The numerous Late Weichselian kame-hills and former

ice-dammed lakes found within the Danish area (Fig. 8
and 9) also clearly point towards the frequent occurrence
of temporary water bodies at the ice sheet margin.

Recent investigations (Humlum, in prep.) along the present ice sheet margin in central West Greenland has shown 20th-century rates of both advances and retreats to be of about 50-200 ma"¹ where the ice sheet margin ends in a marine or lacustrine environment, without being floating, while corresponding rates are 0-10 ma"¹ at places where the margin stands in a terrestrial environment. Where floating of the glacier terminus occurs, rates are usually 100-400 ma"¹. Calving and accelerated melting of ice caused by contact with liquid water is considered the two main reasons for the observed high retreat rates in the marine/lacustrine environment. Furthermore, calving will tend to produce a cuffed margin, a feature which in itself will tend to increase ice ablation during periods of low sun angles (Chinn 1987), thus extending the effective length of the ablation season.

(3) Increased ablation due to strong winds. During glaciological field work in both West and East Greenland, one of the authors have several times observed that the action of strong, dry winds within few hours can result in heavy ablation from a glacier surface. Wind action as this especially appears to be of importance in areas such as the Ammassalik area in south-eastern Greenland, where dry catabatic winds are frequent. One visible result of the wind controlled ablation is the rapid production of a very smooth glacier surface in the ablation area. Under normal summer conditions this part of a glacier is usually characterised by high surface roughness. The importance of wind velocity for ice ablation can also be seen from the empirical relations established by the US Army Corps of Engineers (Fig. 5). Certain sectors of the ablation area of the Late Weichselian ice sheet may have experienced heavy catabatic winds, depending upon the ice sheet topography and the track of wandering meteorological depressions. In sectors such as these, ablation may well have been considerably above average, partly due to increased evaporation, and partly due to much of the winter's snow being swept away from the glacier surface, exposing the solid low-albedo glacier surface below to short-wave radiation. An increased storm activity along the southern ice sheet margin (Oerlemans and Venekar 1981, Manabe and Broccoli 198 5, Oerlemans 1991) during the Late Weichselian may have been the meteorological background for an increased outflow of strong catabatic winds, as is presently the case in SE Greenland.

Conclusion

Simple calculations have shown the existence of a considerable inconsistency between mean Late Weichselian deglaciation rates in Denmark and maximum ablation rates as deducted from a physical point of view. Climatic variations do not alone appear sufficient to explain observed patterns and rates of deglaciation. A combination of flat ice sheet surface profiles due to low basal shear strength, presence of ice marginal water bodies and high (catabatic?) winds may possibly explain the abnormally high Late Weichselian deglaciation rates, and should be considered when formulating dynamic deglaciation models. Further work on this problem is much needed. Also the distribution in space and time of stagnant ice bodies should be investigated further.

The very high deglaciation rates following 17 ka BP in the Baltic Sea area may have been caused by calving in a deep water body (Lagerlund et al. 1983), but supplementary explanations should be sought for both the Danish

areas in general and the Swedish area south of the 12 ka
BP marginal line in central Sweden.

Acknowlwdgements

One of the authors (OH) are indebted to N.Nielsen, B.Hasholt, J.Krüger, H.H.Christiansen (Geographical Institute, Copenhagen) and C.E.Böggild (A.Wegener Institute, Bremerhaven), with whom he has carried out glaciological and geomorphologic field work in Greenland and in Iceland. Especially results obtained from field work at calving and not-calving glaciers in the Disko Bay area have been very useful in the present context. These observations were made in collaboration with N.Nielsen. The software used for the production of the 3D-model and for the calculation of short-wave radiation balance was developed by OH in connection with glaciological-geomorphologic projects at the Mitdluagkat glacier within the Ammassalik area (SE-Greenland) and in the Disko Island region (W-Greenland). The observations on glacier surface profiles were obtained during investigations at Myrdalsjökull (Iceland). The Icelandic part of the project was funded by The Carlsberg Foundation and the work in Greenland by The Danish Natural Research Council; Journ no. 117511, 118270, 1100711. Also the Arctic Station (University of Copenhagen) in Qeqertarsuaq/Godhavn, Greenland, supported the investigations in Disko Bay. The paper benefited from comments made by J.Krüger and H.H.Christiansen on an early version of this paper; the responsibility for any remaining shortcomings is, however, entirely our own.

References

Andrews, J.T. (1973): The Wisconsin Laurentide Ice Sheet: Dispersal
Centers, Problems of Rates of Retreat, and climatic
implications. Arctic and Alpine Research, 5(3): 185-199.

Andrews, J.T. (1982): On the reconstruction of Pleistocene ice
sheets: A review. Quaternary Science Reviews, 1: 1-30.

Andrews, J.T. & Barry, R.G. (1978): Glacial inception and disintegration
during the last glaciation. Annual Review of Earth
and Planetary Sciences, 6: 208-228.

Bard, E., Hamelin, 8., Fairbanks, R.G. & Zindler, A. (1990): Calibration of the 14_C time scale over the past 30,000 years using mass spectrometric U-Th ages from Barbados corals. Nature, 345:405-410.

Bard, E., Arnold, M., Fairbanks, R.G. & Hamelin, B. (1993):
210Th-234U and 14C ages obtained by mass spectrometry on
corals. Radiocarbon, 35(1): 191-199.

Beget, J. (1987): Low Profile of the Northwest Laurentide Ice
Sheet. Arctic and Alpine Research, 19(0:81-

Berglund, B.E (1979): The deglaciation of southern Sweder
13,500-10,000 BP. Boreas, 8:89-

Berthelsen, A. (1978): The methodology of kineto-stratigraphy a;
applied to glacial geology. Bulletin geological Society Den
mark, 27: 25-38.

Berthelsen, A. (1979): Contrasting views on the Weichselian glaciation
and deglaciation of Denmark. Boreas, 8: 125-132.

Birchfleld, G.E. & Grumbine, R.W.(\ 985): 'Slow' physics of large continental ice sheets and underlying bedrock and its relation to Pleistocene ice ages. Journal of Geophysical Research, 90: 11294-11302.

Braithwaite, R. J. (1981): On glacier energy balance, ablation, and
air temperature. Journal of Glaciology, 27(97): 381-391.

Braithwaite, R.J. cfe Olesen, O. (1984): Ice ablation in West Greenland in relation to air temperature and global radiation. Zeitschrift fur Gletscherkunde und Glazialgeologie, 20: 155-168.

Braithwaite, R.J. & Olesen, O. (1990): A simple energy-balance
model to calculate ice ablation at the margin of the Greenland
Ice Sheet. Journal of Glaciology, 36(123):222-228.

Boulton, G.S., Smith, G.H., Jones, A.S. & Newsome, J. (1985):
Glacial geology and glaciology of the last mid-latitude ice sheets.
Journal of Geological Society of London, 142: 447-474.

Brotchie, J.F. & Silvester, R. (1969): On crustal flexure. Journal of
Geophysical Research,74(22): 5240-5252.

Bøggild, C.E. (l991): En smeltende snepakkes masse- og energifluxe.
Unpublished M.Sc.thesis, University of Copenhagen.

Chinn, T.J.H. (1987): Accelerated ablation at a glacier ice-cliff
margin, Dry Valleys, Antarctica. Arctic and Alpine Research,
19(1):71-

Clarck, J.A. & Lingie, C.S. (1979): Predicted Relative Sea-Level Changes (18,000 Years B.P. to Present) Caused by Late-Glacial Retreat of the Antarctic Ice Sheet. Quaternary Research, 11(3): 279-298.

Clayton, L., Teller, J.T. & Attig, J.W. (1985): Surging of the
south-western part of the Laurentide Ice Sheet. Boreas, 14:235-241.

Denton, G.H. & Hughes, T.J. (eds.) (1981): The Last Great Ice
Sheets, John Wiley, New York.

Dyke, A.S. & Prest, V.K. (1987): Late Wisconsin and Holocene
history of the Laurentide Ice Sheet. Geographie physique et
Quaternaire, XLI(2): 237-263.

Fisher, D.A., Reeh,N. & Langley, K. (1985): Objective reconstructions of the Late Wisconsin Ice Sheet and the significance of deformable beds. Geographie physique et Quaternaire, 39(3): 229-238.

Föhn, P.M.B. (1973): Short-term snow melt and ablation derived
from heat- and mass-balance measurements. Journal of Glaciology,
12(65): 275-289.

Hyde, W.T. & Peltier, W.R. (1985): Sensitivity experiments with
a model of the ice-age cycle: the response to harmonic forcing.
Journal og Atmospheric Sciences, 42: 2170-2188.

Houmark-Nielsen, M. (1987): Pleistocene stratigraphy and glacial
history of the central part of Denmark. Bulletin geological
Society Denmark, 36(1-2): 1-189.

Houmark-Nielsen, M. & Lagerlund, E. (1987): The Helsingør
diamicton. Bulletin Geological Society Denmark, 36:237-247.

Hughes, T.J. (1985): The great Cenozoic Ice Sheet. Palaeoeeo
graphy, Palaeoclimatology, Palaeoecoloey, 50:9-43

Humlum, O. (1979): "Fluted moraine" på Omø - isbevaegelses
retning og aflejringsmåde. Dansk Geologisk Forening Årsskrif
for 1978, 15-22.

Humlum, O. (in prep.): Holocene ice sheet margin fluctuations ir
the Eqip Sermia area, central West Greenland

Humlum, 0., Kruger, J. & Sjørring, S. (1978): Danmark i sidste
istid: En diskussion. Naturens Verden 1978, 9, 272-286.

Kamper, J., Jacobsen, A. & Carstensen, A.R. (1991): Normalisering af satellitregistreret reflektans fra en glacial overflade, rumlig analysis af BRF og albedo samt afsmeltningsmodellering i et geografisk informationssystem. Unpublished thesis at Geographical Institute, University of Copenhagen.

Kjemperud, A. (1986): Late Weichselian and Holocene shoreline
displacement in the Trondheimsfjord area, central Norway.
Boreas, 15:61-

Kruger, J. (1970): Till fabric in relation to ice movement. Geografisk
Tidsskrift, 69(2): 133-170.

Kruger, J. (1979): Structures and textures in till indicating subglacial
deposition. Boreas, 8:323-340.

Kruger, J. & Humlum, O. (1981): The proglacial area of Myrdalsjokull:
general report on the Danish Geomorphologic Expedition
to Iceland 1977. Folia Geographica Danica, Tom.XV(l).

Lagerlund, E. (1987): An alternative Weichselian glaciation model
with special reference to the glacial history of Skåne, south
Sweden. Boreas, 16: 433-459.

Lagerlund, E., Knutsson, G., Åmark, M., Hebrand, M., Jonsson, L.-0., Karlgren, 8., Kristiansson, J., Moller, P., Robison, M.J., Sandgren, P., Terne, T. & Waldemarsson, D. (1983): The deglaciation pattern and dynamics in South Sweden, a preliminary report. Lundqua Report, 24: 7 p.

Lagerlund, E. & Houmark-Nielsen, M. (1993): Timing and patten
of the last deglaciation in the Kattegat region, southwest Scandinavia.
Boreas, 22:337-347.

Lidmar-Bergstrom, K., Elvhage, C.&Ringberg, B. (1991): Landforms
in Skåne, South Sweden. Geografiska Annaler, 73A(2),
61-91.

Manabe, S. & Broccoli, A.J. (1985): The influence of continental
ice sheets on the climate of an ice age. Journal of Geophysical
Research, 90:2167-2190.

Mangerud, J. & Berglund, B.E. (1978): The subdivision of the
Quaternary of Norden: adiscussion. Boreas, 8: 179-181.

Mathews, W.H. (1974): Surface profiles of the Laurentide ice
Sheet in its marginal areas. Journal of Glaciology, 13(67): 37-43.

Oerlemans, J. (1991): The role of ice sheets in the Pleistocene
climate. Norsk Geografisk Tidsskrift, 71:155-161.

Oerlemans, J. & Vernekar, A.D. (1981): A model study of the relation between Northern Hemisphere glaciation and precipitation rates. Contributions to Atmospheric Physics, 54: 352-361.

Ohmura, A. (1987): Heat budget of the climate system between
the last glacial maximum and the present. Bulletin of the Department
of Geography University of Tokyo, 19: 21-28.

ÖAV-Mitteilungen (various dates): Gletscherbericht. Mitteilungen
des Österreichen Alpenverein. Innsbruck, Austria.

Paterson, W.S.B. (1972): Laurentide Ice Sheet: estimated volumes
during late Wisconsin. Reviews of Geophysics and Space
Physics. 10(4): 885-917.

Peltier, W.R. (1981): Ice age geodynamics. Annual Review of
Earth and Planetary Sciences, 9: 199-226.

Reeh, N. (1982): A plasticity theory approach to the steady-state
shape of a three-dimensional ice sheet. Journal of Glaciology,
V01.28, No. 100:431-

Reeh, N., Fisher, D.A. &Langley, K. (1983): Ideal plastic reconstruction
of Late Wisconsin Ice cover; maximum concept. Un-

published poster session at Symposium on Ice and Climate Modelling, Northwestern University, Evanston, Illinois, USA. International Glaciological Society and American Meteorological

Ringberg, B. (1988): Late Weichselian geology of southernmost
Sweden. Boreas, 17: 243-263.

Smed, P. (1992): Personal communication.

Sugden, D.E. (1977): Reconstruction of the morphology, dynamics
and thermal characteristics of the Laurentide Ice Sheet at
its maximum. Arctic and Alpine Research, 9(1): 21-47.