Weichselian till stratigraphy and ice movements in , central south

JARLE HOLE & OLE FREDRIK BERGERSEN

Hole, J. & Bergersen, O. F.: Weichselian til! stratigraphy and ice movements in Ottadalen, central south Norway. Norsk Geologisk Tidsskrift, Vol. 61, pp. 25-33. Oslo 1981. ISSN 0029-196X.

Til! stratigraphy and ice movement have been studied in the Skjåk area in the valley Ottadalen. The stratigraphyof one locality, Gubbhågå, is described in detail. Both at Gubbhågå and at other localities in Skjåk waterlaid sediments have been found which are presumed to be the sediments from the end of an interstadial of Mid/Early Weichselian age, most probably Interstadial. On the basis of fabric analyses the overlying tills are correlated with different phases of ice movements and a four-phase model of these movements during the last lee Age has been established. These results are consistent with others from other parts of Ottadalen and Gudbrandsdalen. Investigations show that the effects of the ice erosion, even in large glacial troughs dose to the main water divide, have been surprisingly little during the last lee Age.

Jarle Hole, Jordregisteringsinstituttet, Postboks 115, N-1430 Ås, Norway. Ole Fredrik Bergersen, Geologisk institutt, Avd. B, Universitetet i Bergen, A/Legt. 41, N-5014 Bergen, Norway.

Localities where tbere are waterlaid sub-till sedi­ of basal gneisses of Precambrian age. About 8 ments from a Weichselian Interstadial in the km south of the valley, there is a 13 km wide Gudbrandsdal area have been recently described zone stretching east-west with Eocambrian and (e.g. Bergersen & Garnes 1971) (Fig. 1). Such Cambro-silurian sediments. Further south, large sediments are normally covered by different tills areas with the so-called Jotun rocks, mainly dark which are correlated with reconstructed ice basic crystalline rocks, occur. movement phases from the last ice age ((Garnes & Bergersen 1977, Garnes 1979) cf. Fig. 6). Several till sections where the stratigraphy is Stratigraphy at Gubbhågå thought to show the development of the last glaciation have also been found in Ottadalen (Lie There are a number of localities in Skjåk, mainly 1974, Skjerven 1978, Hole 1979). dose to the valley bottom or along the lower The 1ocalities which have been studied in the parts of the valley sides, where sorted sediments Skjåk area Iie in the upper part of the valley occur below tills that were deposited by different Ottadalen (Fig. 2) and one of these, at ice movements. In most places the general Gubbhågå, situated at about 400 m a.s.l. near the stratigraphy is as sketched in Fig. 2, though this valley bottom, is described here. is usually incomplete because the bedrock is not Ottadalen, a typical glaciated valley with large exposed. The lowermost unit observed consists overdeepened troughs, is the largest tributary of sorted sediment and is overlain by till which valley to Gudbrandsdalen, and extends from appears to have incorporated older (glacio-)flu­ westwards to the water divide towards vial sediments, as suggested by the abundance of Sunnmøre in west Norway. Most of the popu­ rounded stones. One or more additional tills lated areas of Skjåk are concentrated in a large occur at the top (Fig. 3). through with very steep U-shaped valley sides. The locality Gubbhågå Iies at a former gully The bottom of the valley Iies only ca. 400 m section, oriented north-south, where one of the a.s.l., though the rivers have a ca. 480 km long sides has been levelled off due to land reclama­ path to the sea in Oslofjorden area. Surrounding tion (Fig. 7). In this section three different tills mountain plateaus are found up to 2000 m a.s.l. have been distinguished on the basis of textural and from the plateau to the south there are three and structural analyses. Beneath the tills there large tributaries which descend to the valley. are more than 7 m of sorted sand (Fig. 4). The The bedrock in the area consists predominantly sand was homogenous with several cm thick silt 26 J. Hole & O. F. Bergersen NORSK GEOLOGISK TIDSSKRIFT l (1981) '---

..+

--.,"----.... Va l dr es

HARDANGER­ VIDDA Waterlaid sediments presumed older than Late-Weichsel Fig. l. Localities with sediments and fossils presumed older than /v\ammoth finds Late-Weichsel in central south Norway. After Garnes & Other fossils older than Late-Weichsel Bergersen (1977).

layers, and showed few clear structures, till unit. The overlying till (C), some 3 m thick although the layers had obviously been dis­ and also without clear structures, was very sand­ turbed, probably due to the overriding ice. rich but poor in coarser fragments. Roundness Neither macrofossils nor pollen have been analyses showed that the stones were more found in the section. There was a 0.5 m gradual angular than those in the till below, while fabric transition up to the overlying 7 m thick, fine­ analyses indicated a direction of ice movements grained til!(D ). The til! contained few pebbles or to the northwest (cf. Fig. · 4) . A diffuse layer blocks and was homogenous and compact with­ marked the transition between tills D andC. out visible structures. Fabric analyses showed a The uppermost till (B) followed without a maximum direction east- west, with a slight dip distinct border to tillC. This till, some2m thick, to the west, indicating that the till had been was loose, gravelly and lacked clay. Fabric deposited by glacier streams eastwards, i. e. measurements showed an east-northeast direc­ along the valley. In this til! we found two very tion, with a slight dip to the west-soutwest. On distinct, dark clay layers, respectively 20 cm (a) the top of the section was a0.5-l m thick layer of and lO cm (b) thick (cf. Fig. 4). The layers, silt and sand. These sediments, which are wide­ dipping slightly to the south, contained42% elay spread in the lower part of the valley sides, are and were dried up. They could be traced through correlated to the deglaciation period when a the whole section, and in another section 250 m series of glacier dammed lateral !akes existed away a similar layer was observed in the same (Hole 1979). NORSK GEOLOGISK TIDSSKRifT l (1981) W eichselian til/ in Ottadalen 27

[0Mammothf ind ed 0 ����'\�i�n•t�\�:��c����um r;;;, Till with older sediments l� incorpora1ed

Fig. 2. Localities with mammoth-finds, waterlaid sediments presumed older than Late-Weichsel, and till with older sediments incorporated in the Ottadalen area. For location see Fig. l.

however, block layers can be useful in till Interpretation of the stratigraphy at stratigraphic studies. Gubbhågå The origin of the two clay layers in till D is Few conclusive results have been drawn from uncertain, though it seems probable that they are the investigations. Mineralogical/petrographical syngenetic and therefore suggest the existence of studies also failed to allow distinctions between a wet base glacier during the till formation. the various tills. The characteristic Jotun rocks, As found in other studies of till stratigraphy which elsewhere in the Gudbrandsdalen region (e.g. Garnes 1979, Garnes & Bergersen 1977), have proved to be useful indicators of transport fabric analyses have been the most important direction, are few in all samples. method in distinguishing tills. In the Gubbhågå Roundness analyses showed that the low­ section five analyses were carried out. The ermost till, D, contained more rounded material orientation and dip of the longest axes of stones than the other tills. were measured and plotted on a Schmidt net. In the transition zone from one till layer to The fabric analyses indicate that the tills cor­ another, for example between the tills D and C, respond to three different ice movement phases; there often occur boulder layers. Evidently namely, along the valley (oldest), across the block layers may have different genesis, some­ valley towards the northwest (medium), and times also occurring interlayered in tills (e.g. across the valley to the northeast (youngest) Dreimanis 1976). U sed with combined analyses, (Fig. 5). These directions are in full agreement

450

SOUTH 400

Fig. 3. The Weichselian till Til\ with older sediments incorporated (0) 350 stratigraphy in Skjåk, schemati­ cally drawn. The letters A-D correspond to the layers at Gubbhågå, Fig. 4. 28 J. Hole & O. F. Bergersen NORSK GEOLOGISK TIDSSKRIFT l (1981)

1.20 ma sl

ROUNDNESS GRAIN SIZE

B + :!!!!!!!!!!!W!Wiii!!!t�::;·;·�6 0-76 c ki======�:!:!:::::::::! E:t======�i:H::::::::::::::::::::::::::::::J::;·;J 2 -77 + E1======�======�::i:m::m::m::m:mm::::I>;=:=>;·=:=J 17 -76 E=i===�======�==�m ::i:i:li:li!!ii::::�:::::::::t-:::::-;<=1 61-76

o r.=:::�::: �:::: ::: g======�:::m:::::::::�-;:_;_j ,a-76 : E = �======:s=====1::::mm::::::m::::::::::::t:::·;·::::;·;·!19 -76 l: =�======�======l:::::::ii:::::::::l::H:::::i::![:::-;-:::>:·:J6 2-76

Fig. 4. Sketch of the Gubbhågå + section showing the stratigraphy and the localities of the samples. E The diagrams illustrate the 50 roundness analyses. grain-size % % distribution, and fabric analyses 40 in the section. Roundness cias­ CLAY SILT SAND GRAVEL ses are angular (k), abraded ang­ 20 ="' = les (kr), rounded (r), and well L E i ======rounded (gr). Definitions, see o = = =3::m:::::::::::::t·::=::·;·::::;·;·::=::·l k kr gr <0.002 0002-0.6 0.6-2 2-16mm Bergersen (1973). NORSK GEOLOGISK TIDSSKRIFT l (1981) Weichselian till in Ottadalen 29

Orumlin field or distinct ftuted surface Oldest � Glacial striae Youngest o

Fig. 5. Striae and drumlinoid forms in Ottadalen and in the upper Gudbrandsdalen area. Only a limited number of the total striae measured are presented. Some of the striae in the lower part of Ottadalen are taken from Lie (1974), Garnes (1978), and Skjerven (1978). For location, see Fig. l.

with the conclusions of Garnes & Bergersen tion. In recent studies in central south Norway, (1977) as to the ice movements during the last ice the last ice age has been separated into four age over the central part of south Norway. phases as first shown by Garnes (1975) and later Before further conclusions are discussed, the confirmed by others (Garnes & Bergersen 1977, results of the investigation of striae, fluted Garnes 1979, Lie 1974, Skjerven 1978, Hole surface, and drumlins are presented. 1979, Olsen 1979), cf. also Fig. 6.

The present investigations Striae and drumlinoid forms in the The analyses of striae in the Skjåk area and its Ottadalen - Lesja area surroundings are presented on Fig. 5. Those The ice movements in the Skjåk-Lesja area have striae interpreted as the oldest are found on the been little studied up to now. Rekstad (1898) was valley bottom, and show a direction towards the the first to present results of striae measure­ east, down the valley slope. Younger regional ments from this area. He postulated a divide and topographically independent ice movements between eastern and western ice streams located are found over the whole Skjåk area. From the at the eastern border of Skjåk. Rekstad pointed Torsvatn area and eastwards the north-western out that erratics had been transported to the directions seem to have been followed by move­ north and the north-west in Skjåk. For· the area ments towards the north and later on to the south of the upper part of Ottadalen, Ørjasæter north-east. Since these movements show no (1955) found that the ice moved towards the indication of having been clearly influenced by west, parallel to the valley. Based on striae the more than 1500 m high relief, the ice must analyses in KjØlenfjella,the mountain area north have had a considerable thickness. In the area of Ottadalen, Tollan (1963) assumed that in the west ofTorsvatn, the ice moved to the west until beginning of the last glaciation the ice move­ the mountains were deglaciated. In the highest ments went towards the east (in Finndalen). mountains, in the vicinity of the main water Later, the movements turned towards the north divide between east and west Norway, striae and the north-west, probably corresponding to a from local glaciers have been found. These period dose to maximum for the last glaciation. glaciers, probably of Preboreal age, originated At the end of the last glaciation, the ice move­ after the inland ice had melted away, and,co­ ments were found to have a northeasterly direc- vered only restricted areas. 30 J. Hole & O. F. Bergersen NORSK GEOLOGISK TIDSSKRIFT l (1981)

Phase 1 Phase 2

Phase 3 Phase 4

� Direction of ice flow Main water divide

� lee culmination zone "", Recent glacier

Fig. 6. A reconstruction of the ice movements in 4 phases during the last glaciation in the Ottadalen - upper Gudbrandsdalen area. P hase l was a glaciation phase. The glaciation probably started in the highest westernparts along the main water divide gi ving ice streams lining the main valleys. Phase 2 was an early regionalp hase with a large ice culmination zone around the water divide. At this time most of the Skjåk area was situated below the ice di vide. Nearly topographic independent ice streams flowed towards the east and southeast in the area around the lower part of Ottadalen. Phase 3 was an inland ice phase. The main ice di vide bad migratedto the area southeast of Skjåk and ice streams were now flowing towards the north west in Ottadalen and to the north in Gudbrandsdalen. The culmination zone had probably a secondary branch stretching over Vest- to the Jostedalsbreen area between Luster and . Phase 4 was a late regional phase. The ice movements around Ottadalen and Gudbrandsdalen were directed from culmination centres in the Vest-Jotunheimen and areas. The ice streams flowedto the north-northeast over Ottadalen and middle and upper Gudbrandsdalen. The movements shown by arrows are not synchronous.

In the area between the lake Aursjøen and the 1290 m a.s.l. The ridges follow broad gaps valley Lordalen many localities with drumlins through the mountains and some are more than 3 and fluted surfaces have been observed (Fig. 5), km long. They are 50-200m wide but only 2-5 m the majority of which show a flow direction high (Fig. 8). The direction of these ridges cor­ towards the northeast. The drumlins in the upper responds well with the youngest striae found in part of Lordalen were first described by Tollan this area. The phenomena are thought to have (1963), who correlated the forms with the young­ been formed at a late stage during the deglacia­ est ice movements to the northeast. As seen tion period. Clearly the last inland ice flows bad from Fig. 5, the flow direction has a tendency to a diffluence in the upper part of Lordalen at this deflecttowards Lordalen. time showing influence of young flows from the In the mountains between Lordalen and Lesja west. The northem branch, being more and more (upper part of Gudbransdalen), some drumlins deflectedto the west, followed the valleys across and fluted forms with direction approximately the main water shed to west Norway. north-south have been found in an elevation of In the Råkåvatn area there are several NORSK GEOLOGISK TIDSSKRIFT l (1981) W eichselian till in Ottadalen 31

Fig. 7. The section at Gubbhågå, viewed towards the west. The Jettered layers A-E correspond to the layers on Fig. 4. Most of the sub-till sediments are covered by slumped material. The dark arrowed layers in lill D are clay layers.

drumlins and fluted surfaces with direction 315- sited in an ice-free period before the last ice age. 335" (Skjerven 1978). The phenomena are Furthermore, both the sediment texture and situated 13 60-1390 m a.s. l. in a broad mountain their position are very similar to what wetind at valley descending to the southeast. The other localities with sub-till sediments in the drumlins, having their highest and broadest end Gudbrandsdalen area, cf. Fig. l. The latter most in the southeast, indicate deposition by ice probably were all deposited during the which moved towards the northwest, up the Gudbrandsdalen Interstadial (Bergersen & valley slope. The forms therefore do not cor­ Garnes1981). The age of this ice-free period, respond with the youngest ice movement, which which seems to have lasted several thousand presumably was towards the northeast. The au­ years, is still unknown, though recent14C datings thors have observed several localities in the of mammoth fragments, which are thought to Jotunheimen area where older depositonal derive from this period, gave an infinite age of forms, even older end moraines, are preserved more than ca.40 ,00 0 years B.P. (Heintz, Garnes from younger ice movements. Distinct drumlins & Nydal1979). from other parts of south Norway have also Sub-till sediments are only found in the lower survived the youngest ice movements (e.g. Rye parts of the northern valley side in Skjåk. Since & Pollestad 1972, Garnes19 76) . there is neither erosion border between the sedi­ ments and the lower till, nor traces of fossil soils or organic matter atthe top of the sediments, we presume that the sediments were deposited at Discussion the beginning of the last ice age. The sediments Sub-til/ sediments were probably deposited in a glacial lake dam­ med by ice masses entering Ottadalen from The sub-till sediments in Skjåk give little infor­ Bøverdalen. It is also possible, however, that the mation about their genesis and age. However, sediments accumulated in a lateral lake which their stratigraphic position below three till units existed along the valley glacier in Skjåk. The corresponding with glacial phases of the last ice most interesting aspect of the sediments is that age indicates that the sediments have been depo- they show clear evidence of a very modest ice 32 J. Hole & O. F. Bergersen NORSK GEOLOGISK TIDSSKRIFT l (1981)

Fig. 8. Fluted forms,1290 m a.s.l.,in the mountains between Lordalen and Lesja, viewed towards northwest.

erosion during the last ice age. The same has This regional direction now casts doubt on been pointed out for Gudbrandsdalen and its eariier suggestions by Ramsli (1947) and StrØm tributaries (e.g. Garnes & Bergersen 1977). But (1956), and later repeated in many papers, that the widespread remnants of sediments in Skjåk Jotunheimen was a culmination zone during the show that even in one of the !argest trough whole last ice age. From the field data it seems valleys close to the highest mountains of Nor­ more likely that the culmination zone was way, there has been little glacial erosion during situatedto the east of Jotunheimen in this phase, the Weichsel. with a possible secondary culmination zone stretching westwards towards the Jos­ tedalsbreen area, cf. Fig. 6, phase 3. The north­ lee movements and ice divide migration in westerly ice movement gradually turnedtowards Slqåk the north and later towards the northeast prob­ Striae, as well as fabric, show that there has ably due to a growing culmination centre in been ice movements parallel to Ottadalen to­ Vest-Jotunheimen. These youngest movements wards the east, down the valley. The have been demonstrated with the help of striae, stratigraphy in the till sections indicated that this till fabric, drumlins, and fluted surface. The was the oldest ice movement in the area (Fig. 4). same development has been found in many sur­ Further east and southeast there is abundant rounding areas, cf. Fig. 5. An ice divide was evidence that the valley glacier phase was fol­ probably located over Vest-Jotunheimen­ lowed by a regional main glaciation phase with Hestbrepiggane, at 1east in part of this phase ice movements towards the east-south-east. This (Vorren 1973, 1977), (Fig. 6, phase 4). has been reported in all previous recent in­ The northeastwards movements persisted un­ vestigations in the Gudbrandsdalen area, though til the highest mountain areas were deglaciated. this phase has not been found in Skjåk, possibly Mter that time there was a downwasting of the because the ice divide at that time was located in inland ice remnants now limited to the valleys. the vicinity of Skjåk. Garnes (1979) postulated The ice masses along Ottadalen melted with that when an ice divide migrated over an area, extremely small gradients, giving conditions for there was minimal glacial activity beneath the formationof numerous meltwater phenomena. culmination zone, and therefore one or more phases could be difficult to trace in such an area. The oldest regional and topographic indepen­ Summary dent ice movements which can be demonstrated in Skjåk showed a flow direction towards the From our investigations in the neighbourhood of northwest. This flow has been found from striae Skjåk, the following conclusions have been and till frabric. The drumlins at Råkåvatn (Fig. 5) drawn: are presumed to belong to this phase. The direc­ tion is also reported in earlier works (e.g. Skjer­ The survival of widespread waterlaid sediments ven 1978). from a presumed Early/Mid Weichselian Inter- NORSK GEOLOGISK TIDSSKRIFT I (1981) W eichselian til/ in Ottadalen 33

stadial tentatively correlated with Gudbrandsda­ Garnes, K. 1975: Øst-Jotunheimen. Beskrivelse og vurdering len Interstadial, shows that the glacier in the last av de geomorfologiske og kvartærgeologiske fo rhold for 'Landsplan for naturområder/forekomster'. Report, ice age exercised surprisingly modest erosion Geological Institut, dept. B, University of Bergen, 25 pp. even in earlier strong-glaciated areas dose to the Garnes, K. 1976: Stratigrafi og morfogenese av drumliner på highest mountains in Norway. Eigerøya, Rogaland. AmS l, 5-53. Garnes, K. 1978: Zur Stratigraphie der Weichseleiszeit im zentralen Siidnorwegen, 195-220. In H. Nag! (ed.), Beilriige The waterlaid sediments are often capped by zur Quartiir- und Landschaftsforschung. Festschrift zum 60. three· different basal tills. The fabricof these tills Geburtstag von Julius Fink, Hirt. Wien. combined with regional investigation of striae, Garnes, K. 1979: Weichselian lill stratigraphy in central south drumlins, and fluted surface allow a reconstruc­ Norway, 207-222. In Ch. Schliichter (ed.), Moraines and Varves. Origin Genesis. Classification. Balkema, Rot­ tion of a four-phase model for the last ice age. terdam. The results support an earlier general model Garnes, K. & Bergersen, O. F. 1977: Distribution and genesis (Garnes 1975) for the last ice age in central south of tills in central south Norway. Boreas 6, 135-147. Norway. Heintz, N., Garnes, K. & Nydal, R. 1979: Norske og sov­ jetiske mammutfunn i kvartærgeologisk perspektiv, 209-225. In: Fortiden i søkelyset. 14C datering gjennom25 år. Lab. for Acknowledgements. - We wish to thank Ellen Irgens, Jane Radiol. Dat. Trondheim. ElliJllsen, and Jan E. Lien for drafting the figures, and Dr. Hole, J. 1979: Kvartærgeologiske granskingar i Skjåkområdet. Ronald Steel and John Larsson for correcting the English Unpubl. cand. real. thesis, Universitetet i Bergen. language. Grants from the Norwegian Research Council for Lie, E. D. 1974: Kvartærgeologiske undersøkelser i Otta­ Science and the Humanities (NAVF) and the University of området. Unpubl. cand. real, thesis. Universitetet i Bergen. Bergen are gratefully acknowledged. Olsen, L. 1979: Kvartærgeologiske undersøkelser i Ø. May 1980 -Lillehammer området, . Unpubl. cand. real. thesis. Universitetet i Bergen. Postscript.- This research was initiated and guided by O. F. Ramsli, G. 1947: Siste istid i Gudbrandsdalen. Nor. Geogr. Bergersen. The authors have both participated in the field Tidsskr. l l, 253-259. work and together prepared the results for publication. Pre­ Rekstad, J. 1898: Mærker efter istiden i Gudbrandsdalen. liminary results were presented in the cand. real. thesis of Arch. Mat. Naturv. 20, 18 pp. J. Hole (1979). Rye, N. & Pollestad, B. A. 1972: The ice movement and the ice divide in the Hardangervidda area. Nor. Geo!. Unders. 280, 25-30. Skjerven, J. 1978: Kvartærgeologiske undersøkelser i Vågå­ References området. Unpubl. cand. real. thesis, Universitetet i Bergen. Bergersen, O. F. 1973: The roundness analysis of stones. A Strøm, K. 1956: The disappearance of the last ice sheet from neglected aid in till studies. Bull. Geo/. Inst. Univ. Upps. 5, central Norway. J. Glaciology 2, 747-755. 69-79. Tollan, A. 1963: Trekk av isbevegelsen og isavsmeltingen i Bergersen, O. F. & Garnes, K. 1971: Evidence for sub-till nordre Gudbrandsdalen fjelltrakter.Nor. Geo/. Unders. 223, sediments from a Weichselian Interstadial in the 328-345. Gudbrandsdal Valley, central East Norway. Nor. Geogr. Vorren, T. O. 1973: Glacial geology of the area between Tidsskr. 25,99-108. Jostedalsbreen and Jotunheimen, south Norway. Nor. Geo/. Bergersen, O. F. & Garnes,K. 1981: Weichsel in central South Unders. 291, 1-46. Ø Norway. A general view of the deposits from the ljasæther, H. 1955: Løsmateriale og isavsmeltingshistorie i Gudbrandsdalen lnterstadial and from the following ice age. området Nordberg-Sota med sidedalene Tundradalen, Østre del av Billingsdalen (kartbladet Skjåk) og nordøstre delen av Boreas JO, (4). Dreimanis, A. 1976: Tills: their origin and properties, 11-48. In Lundadalen (kartbladet Lom). Unpubl. cand. real. thesis. Universitetet i Oslo. R. F. Legget (ed.), Glacial till. An inter-disiplinary study. R. Soc. Canada, Spee. Publ. 12, Ottawa.

�Norsk GeologiskTidsskr. 1/81