Geografisk Tidsskrift, Bind 96 (1996)

Moraine Systems in the Faroe Islands: Glaciological and Climatological Implications

Ole Humlum, Hanne H. Christiansen, Harald Svensson & Lis E. Mortensen

Side 21

Abstract

Geomorphological investigations on the Faroe Islands have shown the existence of small, but distinct moraine systems, indicating the former occurrence of glaciers. At least two glacial events may be distinguished: an older characterised by valley glaciers 2-4 km long, and a younger characterised by small cirque glaciers. Dating (14C) indicates that the youngest moraines are older than 5,200 yr. BP. During the two glacial events the largest glaciers were located in cirques facing NE, while the smallest were located in cirques facing SW. During the two glacial events equilibrium line altitude was at 200-350 m a.s.l. and 300-450 m a.s.L, respectively. Glaciers of intermediate size were found in cirques facing NW and SE. This pattern is interpreted as being caused by strong snow drift from SW and W. The temperature and precipitation values during the glacial events are discussed with reference to measured meteorological conditions at modern glaciers.

Keywords

Faroe Islands, North Atlantic, Late Weichselian glaciation, North
Atlantic Drift, cirque, moraine systems, palaeoclimate.

Ole Humlum, Hanne H. Christiansen, Harald Svensson: Institute of
Geography, University of Copenhagen, Øster Voldgade 10, 1350
Copenhagen K., Denmark.

Lis E. Mortensen: Museum of Natural History (Føroya Nåtturugripasavn).
Torshavn, Faroe Islands.

Danish Journal of Geography 96: 21-31, 1996.

The Faroe Islands are situated in a sensitive region of the Atlantic Ocean for registering climatic changes. In warm periods, with generally strong, or northward-displaced, circulation features in the atmosphere and ocean, these islands lie continually in the main arm of the North Atlantic Drift (the Gulf Stream). In colder periods, when the North Atlantic Drift weakens or its main arms take a more southerly position, a tongue of polar water from the East Iceland branch of the East Greenland Current approaches the Faroe Islands from the north. When this happens, polar ice may partly surround the islands, as happened several times during the Little Ice Age (last time 1888 AD; see Lamb 1977). By this, the Faroe Island region is so placed as to register large amplitude shifts of the water current boundary in the North Atlantic, which generally are thought to fluctuate with the overall climate in the northern hemisphere (Ruddiman & Mclntyre 1981; Bard et al. 1987).

During the Quaternary the Faroe Islands were several times covered by a local ice cap (Jørgensen & Rasmussen 1986). No foreign erratics have been found and the glacial striae radiate out in all directions from the larger islands. Geomorphic imprints of the Quaternary glaciations are largely in the form of erosional features, such as striae, roches moutonnées and U-shaped valleys, while depositional landforms are sparse (Jørgensen & Rasmussen 1986).

The present study intends, by means of a geomorphological analysis, to extract palaeoclimatic information from this part of the North Atlantic region. Air photo interpretation (1:10.000) of the Faroes with special respect to geomorphological indications of Late Weichselian or Holocene glaciation in high-lying cirques was initiated already in 1987 by two of the authors. Several promising localities were found and during 1994 and 95 field work were carried out in selected areas. Moraine systems were found at several localities, some of which are described in the present paper. The ages of the moraines are discussed, as is the character of the climate that produced the associated glaciers.

Study Area: Topography and Climate

The Faroe Islands (Fig. 1) have a total area of about 1,400

Side 22

DIVL775

Figure 1: Map showing the Faroe Islands and the study areas. 1: The study area at Slættaratindur, N Eysturoy. 2: The study area at Oyndarfjørour, NE Eysturoy. 3: The study area SE of Viöareiöi, N Vidoy. The small inset map shows the general position of the Faroe Islands in the North Atlantic.

km2 and are situated between 61°20'N - 62°24'N and 6°ls'W and 7()4l'W. The distance from the southernmost to the northernmost point is about 113 km and the corresponding east-west distance is about 75 km.

The bedrock is Tertiary plateau basalts, with a total thickness of at least 5,200 m (Rasmussen & Noe-Nygaard 1970, Berthelsen et al. 1984). The landscape is a plateau basalt landscape with typical glacial imprints such as cirque carved lava plateaus and U-shaped valleys and fjords. The upper land surface rises gradually from about 600 m a.s.l. in the southwestern part of the islands to more than 800 m a.s.l. in the northeastern part. The highest mountain, Slættaratindur, reaches 882 m a.s.l. (Fig. 1).

The present climate of the Faroe Islands is strongly mantime timein character, reflecting proximity to the sea and the moderating influence of the North Atlantic Drift. Summers are cool and winters mild. In general the present climate can be characterised as windy, changeable and humid. In Törshavn the present mean annual air temperature is 6.5°C (1961-1990), with year-to-year variations of 0.5-I.5°C (Fig. 2). August is the warmest month with 10.5°C and the coldest is January with 3.2°C. The yearly mean precipitation is around 850 mm w.e. (water equivalent) in the west (Mykines) and increases to about 2,750 mm w.e. (Klaksvik) in the mountainous northern and eastern part of the islands. Mean cloud cover is about 80% in all seasons. The dominant wind direction is from the N, W and SE, while especially winds from E and NE are less frequent. During the winter, storms with very strong winds from the N are frequent (Hansen 1990).


DIVL772

Figure 2: Running 5-year mean air temperature (°C) in Torshavn 1867-1990. Black dots indicate measured values for the individual years. Source: Danish Meteorological Institute (DMI), Copenhagen.

Since 1867, when meteorological observations were initiated in Törshavn, the mean air temperature has undergone significant variations (Fig. 2). During the 1920s and early 1930s the mean air temperature rose about 1.8"C, but afterwards a cooling trend has dominated until recent years. These variations probably reflect variations in temperature and intensity of the North Atlantic Drift. In the absence of the North Atlantic Drift winter temperatures would be 5-6°C than at present (Søgaard in press). The geographical position within a climatic sensitive region is emphasised by the I.B°C mean air temperature rise in Törshavn during the 1920-33 period, when the Earth mean surface air temperature rose just 0.1 °C (Lamb 1989).

Side 23

The knowledge of the regional climate of the Faroe highlands is poor. On mountain plateaus above 500 m a.s.l. the widespread occurrence of periglacial features, such as small sorted circles, stone stripes, deflation surfaces and -terraces, solifluction sheets and -lobes (Christiansen 1988), suggests a mountain climate characterised by extreme humidity and strong winds rather than extreme cold. Adopting a normal vertical lapse rate, the mean winter air temperatures in the high mountains probably lie between -O.3°C - O°C, with yearly mean temperatures of I°C to 4°C (Hansen 1990). Permafrost is absent, though annual ground freezing on high ground occasionally may be expected to reach depths of 0.5 m or more. Thus, the high Faroe mountains can be classified as a maritime periglacial region.

Previous Work

The first to describe striae and roches moutonnées from the Faroe Islands as glacial phenomena was the British author Chalmers (1856), who stayed on the islands for a short period in 1855. He visited a locality at the settlement Eiöi in the north of Eysturoy (Fig. 1), already described by Allan (1814), and concluded that the erosional phenomena were caused by ice flow from the north and that the Faroe Islands had been covered by an arctic ice sheet (Chalmers 1856). The work by Heiland (1879) and Geikie (1880), however, showed, that the Faroes only have had a local glaciation. Geikie (1880) also described a possible interglacial glacialdeposit, near the town Klaksvik, a locality which was later studied by Rasmussen (1972).

During the Weichselian a local ice cap accumulated in the Faroes, covering the terrain up to at least 700 m a.5.1., and extending far beyond the present coastline out on the shelf (Jørgensen & Rasmussen 1986). Almost nothing is known about the Late Weichselian deglaciation on the Faroe Islands. No recessional or readvance moraine systems have been described, and only some dead-ice features and ancient medial moraine accumulations have been mentioned from a few major valleys on Streymoy (Sugden & John 1976 (p. 241), Jørgensen & Rasmussen 1986). It is thus not known when the islands became ice-free. The oldest relevant 14C date is obtained from a lake bottom core at Hoydalar and gives an age of 9,660±150 yr. BP (Johansen 1975, 1985), indicating that glacier ice may have been present in the large valleys up to the beginning of the Holocene.

The question of short episodes of Holocene glaciation in the Faroes has not been addressed in the literature, but the possibility should, however, not entirely be ruled out. This issue will be further discussed later in the present paper. The work by Johansen (1985) on the vegetational history indicates that the early Holocene (Preboreal 10,000-9,000 yr. BP) climate on the Faroe Islands was arctic-subarctic with arrival of Betula nana. In Boreal times (9,000-8,000 yr. BP) the climate changed towards more oceanic conditions, with disappearance of Betula nana. Following the introduction of a rather warm and wet climate species such as Juniperus and Salix expanded, covering the lowlands, together with tall-herb vegetation and grass heaths. The Atlantic period (8,000-5,000 yr. BP) was wet, with evidence of strong leaching of soils. Peat began to accumulate, especially on high grounds. From the onset of the Subboreal (5,000-2,500 yr. BP), climate became cooler and wetter leading to widespread peat formation and decreasing amounts of Juniperus and Salix. Subatlantic time (2,500-0 BP) brought the arrival of man in two landnam phases, 600-700 yr. AD (monks from Ireland) and 800-900 yr. AD (the Vikings), and was punctuated by the Little Ice Age (1,400-1,900 yr. AD).

Field Observations

The Faroe cirques and main glacial valleys are well developed(Fig.
3), but at first glance very few depositional
glacial forms appear. No raised coastal deposits were


DIVL808

Figure 3: Glacial trough valley (Vidvik) with adjoining cirques in southeastern Vidoy, area 3 in Figure 1. The highest mountain (Sneis) reaches 634 m a.s.l. Seen towards SSE. May J995.

Side 24

found. Except for the steep headwalls and exposed mountainplateaus, the terrain is usually mantled by Holocene blanket peat deposits, 0.5-1.5 m thick (Fig. 4). Interbedded within the individual peat layers several 0.5-2 mm thick grey fine-grained layers were observed in several localities. They represent Holocene ash layers, probably derived from Icelandic eruptions (Thorarinsson 1981), as has been described by Waagstein & Johansen (1968) and Mangerud et al. (1968) from deposits cored in lake Saksunarvatn, northern Streymoy. Holocene ash layers of probable Icelandic origin have also been described from Iceland (Björck et al. 1992), Ireland (Pilcher & Hall 1992), Scotland (Blackford et al. 1992) and Germany (Merkt et al. 1993).

The widespread blanket peat at even high altitudes is important in that the peat not only insulates the underlying regolith from freeze-thaw cycles, thus preventing the development of periglacial forms, but also obscures many relict features that developed during the Late Weichselian period.

Detailed geomorphological mapping demonstrated the presence of distinct moraine systems in several cirques and trough valleys. Most of these moraine ridges are smallscale features, less than 1-5 m high, but in certain valleys they are prominent, 10-15 m high, and may even be observed in air photos (Fig. 5).

Distinct moraine systems were found in the upper part of valleys in NE Eysturoy, SE and N of Oyndarfjørour and around the highest mountain Slættaratindur (882 m a.s.l), and in N Viöoy (Fig. 1). Moraines are also found elsewhere but, in general, the moraines appear to be especially frequent and well developed in NE Faroe Islands. Three of the localities will shortly be described below.

The moraines at Viöoy are especially found along the northern and southern valley side of a large NE-exposed cirque, 5 km SE of the village Viöareiöi (Fig. 6). The moraine ridges are 1-7 m high and consist of coarse taluslikematerial with very large (1-2 m) boulders scattered on the surface. Some of the protruding boulders are polished by winds from S and SE. The wind direction may be controlled by the local topography. The northernmost moraine ridge is covered by talus deposits at its upper end (Fig. 6), thereby probably indicating an age higher than Late Holocene. The upper end of the moraine ridges indicate a former equilibrium line altitude (ELA) of about


DIVL811

Figure 4: Holocene peat covering Late Weichselian or Holocene sediments (below the compass) in the upper part of a cirque NE of Slættaratindur, Eysturoy. A sample (AAR-2103) obtained from the lowermost peat layer was dated to 2,985±55 yr. BP. The profile is about 0.6 m high. August 1994.


DIVL814

Figure 5: Vertical air photo showing a large terminal moraine in Trølldalur, NofOyndarfjørdur, area 2 in Figure 1. The ridge, which seen shortly above the cirque mouth, is 10-15 m high and indicates the former existence of a small cirque glacier with ELA at about 270 m a.s.l. The valley mouth measures about 750 m across. North is towards the bottom of the photograph. May 1970.

Side 25

DIVL820

Figure 6: Large NE-exposed cirques with left lateral moraine ridge, SE of Viöareidi, N Vidoy, area 3 in Figure 1. The moraine ridge starts at ca. 290 m a.s.l. (at the two lakes) and indicates a contemporary ELA at this altitude. Accumulation of peat inside the moraine ridge started 2,680±70 yr. BP (AAR-2105). Seen towards SE. August 1994.


DIVL823

Figure 7: Remnants of left lateral moraines deposited by a valley glacier flowing towards north (right) in Dalurin, northern Eysturoy, area J in Figure 1. The uppermost occurrence of this moraine system indicates a former ELA at about 270 m a.s.l. The visible part of the valley side is about 150m high. Seen towards W. May 1995.

a 1.5 km long valley glacier with terminus at 60 m a.5.1., with smaller cirque glaciers to the N and S. Accumulation of peat inside the moraine ridge started 2,680±70 yr. BP (AAR-2105).

In the Dalurin valley, N Eysturoy, remnants of lateral moraines were found along the western valley side (Fig. 7). These moraine ridges are 2-6 m high and indicate the former existence of a 2 km long valley glacier flowing towards N. Several systems (5-6) of ridges are deposited below each other during minor advances during a general retreat following deposition of the outermost moraines. The moraine ridges have later been truncated by Holocene streams flowing down the valley side and are also modified by solifluction. The upper end of this moraine system points towards a former ELA of about 270 m a.s.l. Farther down the Dalurin valley, at the village Gjögv, remnants of an older and even more subdued moraine system are found, deposited by a 3.5 km long valley glacier flowing towards N and NE. During the period where this older system was deposited, again judging from the uppermost occurrence of the moraine ridges, the ELA was about 180 m a.s.l.


DIVL817

Figure 8: Terminal moraine ridges (white arrows) terminating at 385 m a.s.l. in N-exposed cirque, Svinabotnur, northern Eysturoy, area 1 in Figure 1. This well-defined moraine ridge can be followed up to about 440 m a.s.l. (position of former ELA). Accumulation of peat inside the moraine ridge started 5,150±90 yr. BP (AAR-2104). Seen towards SW. May 1995.

In the uppermost part of the Svinabotnur, N Eysturoy, a terminal moraine system consisting of two ridges was found (Fig. 8). The system indicates the former existence of a small cirque glacier, about 350 m long and flowing towards N. Compared to the above mentioned moraine systems these moraines have a more fresh appearance with steep slopes and many angular blocks protruding from the surface and may indicate a limited readvance or reglaciation. The moraine ridges are 1-5 m high and consist of talus-like material derived from the prominent headwall above (Fig. 8). The moraines can be followed up to 440 m a.5.1., which probably indicate the local ELA then. Accumulation of peat inside the moraine ridges started 5,150±90 yr. BP (AAR-2104). Farther down the valley, remnants of other moraine systems were found, but with a much more subdued appearance.

Side 26

Discussion

The moraines ridges observed indicate the extent of glaciers during at least two glacial stages. The uppermost, such as the moraine system in Svmabotnur has a much more fresh-looking appearance than e.g. the moraines at Gjögv, which are more subdued due to solifluction and fluvial activity. In general, the older moraine systems were deposited during at least two glacial stages characterised by 1-4 km long valley glaciers, while the younger moraines correspond to a (third ?) stage with smaller cirque glaciers, only 0.3-0.5 km long (Fig. 9b). It is not yet completely clear if the small cirque glaciers are contemporary with the younger of the two valley glacier stages, or represent an individual third glacial stage. This remains to be investigated by future investigations. However, all moraine systems observed indicate that the former glaciers were especially large in cirques and valleys facing NE, and smaller in cirques facing SW. In cirques and valleys facing NW or SE the glaciers were of intermediate size. Judging from the uppermost part of the individual moraine ridges, the mean ELA for the two stages can be estimated to around 200-350 m a.s.l. and 300-450 m a.5.1., respectively (Fig. 9b).


DIVL853

Figure 9a: Topographic map showing the Slættaratindur (882 m a.5.1.) massive, northern part of Eysturoy, area 1 in Figure I. Contour equividistance 50 m. Water is shown in grey. North is towards the top. Distance between border tick marks is WOO m.

In most cases the younger of these two moraine systems are surprisingly fresh looking, with many protruding angular blocks (Fig. 8). The frequent component of talus-like debris points towards the importance of rockfalls from the free rock faces above the glaciers. The older sets of moraines (Fig. 7) have been much more modified by subsequent solifluction and fluvial dissection. Protruding angular blocks are also less frequent. The general appearance and topographical position of the youngest moraine ridges (Fig. 9b) signal a readvance or renewed glaciation period, while the older moraine ridges may have been deposited during a previous, general deglaciation period.

To obtain information on the age of the observed morainesystems, samples from the lowermost 2-3 cm of the basal peat layer, inside certain moraine ridges, were collectedfor 14C dating (Table 1). Ages are all within a range of about 2,500-5,200 yr. BP, and all the observed moraine


DIVL856

Figure 9b: Computer generated shaded relief map of figure 9a (light source in NW), showing moraine systems (black lines) and outline of reconstructed glaciers (white-grey and white). ELA in metres above present sea level. Values for ELA are positioned at the associated glacier outline. Subglacial longitudinal ridges (D) shown in dark grey. Question marks indicate cirques without visible moraine systems.

Side 27

DIVL859

Figure 10: Altitude-age relationship of basal peat layer datings (14C) from the Faroe Islands. Datings from Johansen 1985 (triangles), Persson 1971 (square) and the present study (circles). AAR-2103 is from the innermost part of the cirque NE of Slættaratindur (Fig. 4 and 9), AAR-2104 is from Svinabotnur (Fig. 8 and 9), AAR-2105 is from a cirque SE of Vidareidi, N Viöoy (Fig. 6 and 1). The sample AAR-2106 is from a mountain plateau with evidence of a former ice cap on southern Bordoy (Fig. 1). AAA-2435 is from a locality north of Fjallavatn, Vagar. All other samples (Persson 1971, Johansen 1985) are from bog localities in trough valleys on Vagar and Eysturoy, respectively, and were collected without special reference to moraine systems.


DIVL865

Table 1: 14C dates (AMS) from the present study, reported in conventional radiocarbon years BP (before present - 1950). Samples were taken from the lowermost 2-3 cm of the peat layer.

In Figure 10 all known dates from the Faroe Islands on the initiation of peat accumulation are shown in an altitude-age diagram. Johansen (1985) suggested from five 14C datings (K-1197, K-1201, K-2468, K-2721, K-2939) that peat formation started in low-lying areas around 7,100-8,700 yr. BP and extended upwards later in the Holocene. This hypothesis is further substantiated by the inclusion of the date St-2078 (Persson, 1977) and the five new dates from the present study (AAR-2103, AAR-2104, AAR-2105, AAR-2106 and AAR-2435). The date AAR-2104 (490 m a.5.1.), however, does not fit entirely within the general pattern. Peat accumulation may apparently have been initiatedbefore 5,000 yr. BP even at rather high altitudes. The North Atlantic Holocene peat accumulation history has been investigated in Britain, where blanket peat formation was initiated in western uplands as early as 7,000 yr. BP (Pears 1975, Birks 1977), and peat accumulation on the Faroe Islands may have followed this general evolution, being situated only 300 km NNW of Scotland. The date AAR-2104 was obtained inside the youngest moraine systemobserved (Fig.B) and the corresponding glacial stage is thus more than about 5,200 yr. BP old. From this, a Mid Holocene (Atlantic) age can probably be ruled out for the last glaciation on the Faroe Islands.

The Early Holocene (Boreal) period before 7,000 yr. BP was characterised by warmer conditions than at present in Britain. In Late Boreal and Early Atlantic times mean annual air temperatures in England and Wales were 1.3-I.6°C higher than at present (Lamb et al. 1966), suggesting a Mid- or Early Holocene glaciation in the Faroe Islands to be unlikely.

From the above, the observed moraines derived from the youngest cirque-glacier episode (Fig. 9b) are tentatively ascribed a pre-Holocene age. The majority of reconstructed ELA are at about 300-450 m a.5.1., which is close to what would be expected by extrapolation of Younger Dryas ELA reconstructed in England, Scotland and Iceland (250-450m ; Sissons 1979 and 1980, Cornish 1981, Gray 1982, Ballantyne 1989, Ingölfsson 1985, 1987, 1988 and 1991). A proper dating of the moraines observed on the Faroe Islands has, however, not yet been possible. The geometry of the youngest moraine systems observed on the

Side 28

Faroe Islands signals a readvance or renewed glaciation. The asymmetrical spatial glaciation pattern with especially large glaciers facing N, NE and E (Fig. 9b) suggests significant winter snow drift by winds from SW and W at that time.

The moraine ridges associated the previous glacial stage with valley glaciers (Fig. 7 and 9b) were perhaps probably deposited during a period of general retreat. The spatial pattern of glacier evolution was probably caused by snow drift from SW and W.

The altitude of the upper end of the observed moraines enables an estimate of the contemporary ELA, and from this, a reconstruction of certain aspects of the stadial palaeoclimate can be performed. The reconstructed ELA for the period with valley glaciers is about 200-350 m a.5.1., while it is about 300-450 m a.s.l. during the last glaciation episode with small cirque glaciers (Fig. 9b).

From a geomorphological point of view, glacier equilibrium lines are very important because they represent the lowest boundary of climatic glaciation. The climate that prevails at glacier equilibrium lines is considered to be just sufficient to maintain the existence of glaciers (Ohmura et al. 1992). Analysing the climatic control on ELA, it is usually considered appropriate to use at least two variables, precipitation and temperature, which represent the effects of accumulation and ablation, respectively. The climate at the ELA is known accurately on about 70 modern glaciers world-wide, where annual mass-balance measurements and meteorological observations have been carried out for a sufficiently long period. Figure 11 shows this as a plot of the mean yearly precipitation versus summer mean air temperature observed at these glacier equilibrium lines. From this empirical plot it clearly stands out, that as summer mean temperatures get higher, accumulation (precipitation) must increase to compensate the increased ablation.

Using modern meteorological observations from the Faroe Islands, where August is the warmest month with 10.5°C at sea level, and assuming a mean vertical lapse rate of 0.0065° Cm~I, a modern mean summer air temperature of about 6.5-B.O°C can be estimated for the cirque valley bottoms (about 300-450 m a.5.1.), where moraine systems have been observed. As mentioned earlier, there are at present no data available on the mountain climate of the Faroes. However, adopting the above estimate, and using measured precipitation values, the higher part of the present Faroe valleys plots in Figure 11 as the shaded ellipse in the diagram. It correctly appears that the modern Faroe Island cirques plots in the no-glaciation part of the diagram, although some of the highest mountains with maximum precipitation (northern and northeastern part of the islands) must be rather close to incipient glaciation. The higher part of the mountains (500-882 m a.5.1.) today is in a periglacial nival zone with widespread nivation features such as nivation hollows, fluvial snowmelt scars and niveo-eolian deposits, and some small perennial snowpatches. A lasting summer air temperature decrease of rather few degrees, say 3-4°C, would probably initiate small glaciers in several cirques in these mountains, even with a small decrease in precipitation.


DIVL862

Figure 11: Annual total precipitation (winter mass balance plus summer precipitation) and the mean summer air temperature measured at the ELAsfor 70 modern glaciers, using raw data from Ohmura et al. (1992). The solid line indicates a polynomial fit, yielding a coefficient of determination (R-squared) of 0.809 (N=7o). The polynomium takes the form: Y = 713.098 + 300.597*X + 2.35328*X2 + 0.540662*X3. The hatched ellipse indicates the present position of mountains and cirques in the Faroe Islands. The open ellipse 'YD?' indicates the inferred position during Younger Dryas.

In both England and Scotland (300 km SSE of the Faroe Islands), a Younger Dryas glacier readvance is both well documented and widespread. The limits of this readvance (the Loch Lomond Stadial) are usually defined by remarkablefresh glacial features such as moraines, drift limits and periglacial trimlines that mark the upslope transition from glacial drift and ice-scoured bedrock to terrain that bears the imprint of contemporaneous frost action (Gray and Coxdon 1991). These features have enabled geomorphologists to reconstruct equilibrium line

Side 29

altitudes (ELA) during the Younger Dryas, and from this, to reconstruct aspects of the stadial palaeoclimate. In Scotland, average 'regional' reconstructed ELA rise gradually southwards from 495 m a.s.l. for the western Southern Uplands of Scotland to 540 m a.s.l. for the English Lake District and 600 m a.s.l. for northern Wales (Sissons & Southerland 1976, Sissons 1980, Cornish 1981, Gray 1982, Ballantyne 1989). Across the Scottish Highlandsand Inner Hebrides, the ELA rise from around 300 m a.s.l. in the west to 1000 m a.s.l. in eastern Scotland (the Cairngorms). Sissons (1979, 1980) interpreted the spatial pattern of the Loch Lomond readvance to be controlled by southerly snow-bearing winds associated with eastward passage of warm or occluded fronts across northern Britain,with depressions following more southerly tracks than at present.

Analysis of reconstructed ELA in Britain, especially in Scotland, has also permitted an estimate of summer temperatures during the Loch Lomond Stadial. For NW Scotland, at sea level, a mean summer air temperature of 6°C is calculated (Sissons and Southerland 1976, Ballantyne 1989). This is consistent with estimates made on the basis of contemporaneous coleoptera (Coope 1977), and about 6°C below modern values. Adopting a Younger Dryas summer air temperature decrease of 6°C for the Faroe Islands would place many of the higher mountains within the zone of glaciation (Fig. 11, YD-outline), even though precipitation probably then was somewhat below present values due to the lower air temperature, perhaps as much as 700-100 mm w.e. This palaeoclimatic scenario would further have had a high frequency of SW winds during the winter. There is thus reason to expect a glaciation in many Faroe cirque valleys during the Scottish Loch Lomond Stadial, and therefore we suggest a tentative Younger Dryas age for the youngest moraine system observed.

Conclusion

The Faroe Islands is the only land area situated within the North Atlantic Drift, and in our opinion this area represents a key study area in the North Atlantic context within the current debate on global climate change.

Several moraine systems have now been identified in the Faroe Islands, associated with at least two glacial stages: a stage (the older) with 1-4 km long valley glaciers and one characterised by cirque glaciers less than 1 km in length.

The individual moraine ridges are small, usually 1-5 m high, and it was probably the widespread occurrence of Holocene blanket peat that prevented earlier recognition of the moraines. An asymmetrical spatial glaciation pattern with especially large glaciers facing N, NE and E is interpreted to represent the result of significant winter snow drift by winds from SW and W at that time. Dating of the blanket peat initiation indicate the moraine ridges to be older than 5,200 yr. BP. Based on comparisons with other North Atlantic regions (e.g. Scotland), modern Faroe Island climate and meteorological observations from modern equilibrium lines, a tentative Younger Dryas age is suggested for the youngest glacial stage on the Faroe Islands.

Acknowledgements

The initial investigations were carried out in co-operation with Dr. Richard Aahmann, Institute of Physical Geography, University of Lund, Sweden. The initial study in 1994 was funded by Institute of Geography, University of Copenhagen (OH), Professor N. K. Jakobsen's Fond (HHC), and the Royal Physiographical Society in Lund (HS). The field work in 1995 was funded by The Danish Natural Research Council (SNF). All these institutions and individuals are thanked for their support. The manuscript benefited significantly from the comments of two anonymous

References

Allan, T. (1814): An Account of the Mineralogy of the Faroe
Islands. Transactions Royal Society Edinburgh, VII: 229-267.

Ballantyne, C.K. (1989): The Loch Lomond Readvance on the
Isle of Skye, Scotland: glacier reconstruction and palaeoclimatic
implications. Journal of Quaternary Science, 4: 95-108.

Bard, E., Arnold, M., Duprat, J., Moyes, J. and Duplessy, J.-C. (1987): Reconstruction of the last deglaciation: deconvolved records of 180 profiles, micropaleontological variations and accelerator mass spectrometric I4C dating. Climatic Dynamics 1: 101-112.

Berthe Isen, 0., Noe-Nygaard, A. & Rasmussen, J. (1984): The deep drilling project 1980-1981 in the Faeroe Islands. Annales Societatis Scientiarum Færoensis, Supplementum IX, Törshavn.

Birks, H.J.B. (1977): The Flandrian forest history of Scotland: a
preliminary synthesis. Pp.119-135 in F.W.Shotton (ed.): British
Quaternary Studies: Recent Advances. Clarendon, Oxford.

Björck, S., Ingolfsson,O., Haflidason,H., Hallsdottir,M. & Anderson, N.J. (1992): Lake Torfadalsvatn: a high resolution record of the North Atlantic ash zone I and the last glacial-interglacial environmentalehanges in Iceland. Boreas, 21: 15-22.

Blackford, J.J., Edwards, K.J., Dugmore, A.J., Cook, G.T. og Buckland, P.C. (1992): Icelandic volcanic ash and the mid- Holocene Scots pine (Pinus sylvestris) pollen decline in northern Scotland. The Holocene, 2, 3: 260-265.

Chalmers, R. (1856): Tracings of Iceland & the Faroe Islands.
Annal, societ. scient. Færoensis, Supplem. IX, Thorshavn.

Christiansen, H.H. (1988): Færøernes geomorfologi (Geomorphology of the Faroe Islands). In 'Naturgeografisk Hovedfagskursus, Vagar, Færøerne, 1988', pp. 13-29. Institute of Geography, University of Copenhagen.

Coope, G.R. (1977): Fossil coleopteran assemblages as sensitive indicators of climatic changes during the Devensian (last) cold stage. Philosophical Transactions of the Royal Society of London, 6280:313-340.

Cornish, R. (1981): Glaciers of the Loch Lomond Stadial in the
western Southern Uplands of Scotland. Proceedings of the
Geologists' Association, 92: 105-114.

Geikie, J. (1880): On the Geology of the Færoe Islands. Transactions
Royal Society Edingburgh, XXX: 217-269.

Gray, J.M. (1982): The last glaciers (Loch Lomond Advance) in
Snowdonia, N Wales. Geological Journal, 17: 111-133.

Gray, J.M. and Coxdon, P. (1991): The Loch Lomond Stadial glaciation in Britain and Ireland, pp.B9-105 in J. Ehlers, P.L. Gibbard and J. Rose, (eds): 'Glacial Deposits in Great Britain and Ireland'. Rotterdam, Balkema.

Hansen, L. E. (1990): Jordbundsudviklingen i Leynavatnområdet.
M.Sc. thesis, Institute of Geography, University of Copenhagen.

Helland, A. (1879): Über die Vergletscherung der Färöer, sowie
der Shetland- und Orkney-Inseln. Zeitschrift Deutsche
Geologischer Gesellschaft, XXXI: 149-755.

Ingölfsson, O. (1985): Late Weichselian glacial geology of the
lower Borgafjöröur region, western Iceland: a preliminary
report. Arctic. 38: 210-213.

Ingölfsson, O. (1987): Investigations of the Late Weichselian
glacial geology of the lower Borgafjöröur region, western
Iceland. Lundqua Thesis 19, Lund.

Ingölfsson, O. (1988): Glacial history of the lower Borgafjöröur
region, western Iceland. Geologiska Foreningens i Stockholms
Förhandlingar. 110: 293-309.

Ingolfsson, O. (1991): A review of the Late Weichselian and early Holocene glacial and environmental history of Iceland. Pp. 13-29 in J.K.Maizels and C.Caseldine (eds.): Environmental Change in Iceland: Past and Present. Kluwer Academic Publishers, London.

Johansen, J. (1975): Pollen diagrams from the Shetland and
Faroe Islands. New Phytol., 75: 369-387.

Johansen, J. (1985): Studies in the vegetational history of the
Faroe and Shetland Islands. Annales Societatis Scientiarum
Færoensis Supplementum XI.

Jørgensen, G. and Rasmussen, J. (1986): Glacial Striae, roches
moutonnées and ice movements in the Faroe Islands.
Geological Survey of Denmark, DGU Series C, N0.7.

Lamb, H.H. (1977): Climate, Present, Past and Future, 2.
Climatic history and the future. Methuen & Co. Ltd., London.

Lamb, H.H. (1989): Weather, Climate & Human Affairs. Routledge,
London & New York.

Lamb, H.H., Lewis, R.P.W. and Woodroffe, A. (1966): Atmospheric circulation and the main climatic variables between 8000 and 0 BC: meteorological evidence. Pp. 174-217 in J.S. Sawyer (ed.): World Climate, 8000-0 BC. Meteorological Society, London.

Mangerud, J., Fumes, H. and Johansen, J. (1986): A 9000-yearold
ash bed on the Faroe Islands. Quaternary Research, 26:
262-265.

Merkt,J., Müller,H., Knabe,W., Müller,?., and Weiser,T. (1993):
The early Holocene Saksunarvatn tephra found in lake
sediments in NW Germany. Boreas, 22(2): 93-100.

Ohmura, A., Kasser, P. and Funk, M. (1992): Climate at the
equilibrium line of glaciers. Journal of Glaciology, 38, 130:
397-411.

Pears, N.V. (1975): Radiocarbon dating of peat macrofossils in
the Cairngorn Mountains, Scotland. Transactions of the
Botanical Society of Edinburgh, 42: 255-260.

Persson, C. (1971): Tephronological investigation of peat deposits
in Scandinavia and on the Faroe Islands. Sveriges
Geologiska Undersökning, Årsbok 65(2), Ser.C, 656, 34 pp.

Pilcher, J.R. og Hall, V.A. (1992): Towards a tephrochronology
for the Holocene of the north of Ireland. The Holocene, 2,3:
255-259.

Rasmussen, J. (1972): Mörena å Boröoyarvfk, sum bendir å eitt millumbil i glersetingini har noröuri. (Glacial deposits in Boröoyarvfk indicating an intervening stage in the glaciation). Frööskaparrit, 20: 54-70.

Rasmussen, J. and Noe-Nygaard, A. (1970): Geology of the
Faroe Islands. Danm. Geol. Unders, l, 25.

Ruddiman, W.F. and Mclntyre, A: (1981): The North Atlantic

during the last deglaciation. Palaeogeography, Palaeoclimatology,
Palaeoecology, 35: 145-214.

Sissons, J.B. (1979): Palaeoclimatic inferences from former
glaciers in Scotland and the Lake District. Nature, 278: 518-521.

Sissons, J.B. (1980): Palaeoclimatic inferences from Loch Lomond Advance glaciers, pp.3l-44 in J. L. Lowe, J. M. Gray and J. E. Robinson (eds.): 'Studies in the Lateglacial of North- West Europe, Pergamon, Oxford.

Sissons, J.B. and Southerland, D.G. (1976): Climatic inferences
from former glaciers in the south-east Grampian Highlands,
Scotland. Journal of Glaciology, 17: 325-346.

Sugden, D.E. and John, B.S. (1976): Glaciers and Landscape.
Edward Arnold, London.

Søgaard, H. (in press): Klima, vind og vejrlig på Færøerne (Climate, wind and weather on the Faroe Islands), pp.24-27 in R. Guttesen (ed): Topografisk Atlas Færøerne, Royal Danish Geographical Society.

Thorarinsson, S. (1981): Greetings from Icelands. Ash-falls and
volcanic aerosols in Scandinavia. Geografiska Annaler, 63A:
109-118.

Waagstein, R. og Johansen, J. (1968): Tre vulkanske askelag fra
Færøerne. Meddelelser fra Dansk Geologisk Forening, 18:
257-264.