The geology of the Torfinnsbu window, central ,

TREVOR F. EMMETT

Emmett, T. F: The geology of the Torfinnsbuwindow, central Jotunheimen, Norway. Norsk Geologisk Tidsskrift, Vol. 60, pp. 255-261. 1980. ISSN 0029-196X. A sequence of metasediments (sparagmites, schists, amphibolites) are exposed in a recently discovered window (3 1/2 x l km) through the Jotun stem gneisses around Torfinnsbu, on the north shore of the lake . The metasediments display three phases of deformation which correlate weU with the deformation phases in the metasediments at Beito and Bygdin. The Jotun gneisses were thrusted over the sediments during the first and most intense phase of deformation. Metamorphic grade increases upward in the metasediments to epidote - amphibolite facies at the base of the Jotun gneiss. The occurrence of such a window indicates that the Jotun nappe in this region is much thinner than predictions based on gravity anomaly data would suggest.

T. F. Emmett, Department of Geology, The University, Keele, Staffordshire, ST5 5BG, England.

On the geological map of Norway (Holtedahl & Small exposures of metasediments are seen in Dons 1%0) the lake Bygdin (Fig. 1,B) is shown the lake shore 200 m east and 150 m west of to Iie almost entirely within the crystalline rocks Torrmnsbu hostel, but these are partly sub­ of gabbroic composition that make up the Jotun merged when the level of the lake is high. The nappe. The extreme eastem tip of the lake is metasediments on the lake shore east of the hut crossed by the thrust bounding this Nappe (the have within them a thin downfaulted wedge of basal thrust of Hossack 1968), which separates Jotun gneiss (Fig. 2,B). The boundary faults the crystalline rocks to the west and north from have very little displacement so that here the the underlying conglomerates and feldspathic base of the Jotun rocks cannot be far above lake sandstones (sparagmites) of the Group level. The base of the Jotun rocks is at 1250 m to the east and south. Torfinnsbu tourist hostel a.s.l. in Langedalsåi, and at 1370 m a.s.l. in the stands on the northem shore of the lake, 10 km Nybui section. The window is therefore of a west of this thrust. At Torfinnsbuand westwards dome-like form, and this is reflected in the from it along the lakeside for 31/2 km a tract of disposition of lineations within the metasedi­ metasedimentary rocks has been discovered ments. emerging from beneath the Jotun rocks and The arched form of the basal thrust is picked constituting a tectonic window within the Jo tun out by a feature on the south flank of Turfins­ nappe. This paper describes the principal fea­ tindane and the southwest slopes of Kalvåhols­ tures of the Torrmnsbu window. tind. This feature also suggests that the basal thrust is very near to the lake level east of Torrmnsbu. Form of the window

The window (Fig. l ,A) is generally elongated in Structure and petrography of the an east-west direction with a north-trending Jotun rocks minor axis approximately 0.6 km long. Metasediments are exposed in the numerous The crystalline rocks above the metasediments gullies which cut through the wedges of drift at Torfinnsbu are meta-igneous rocks of the mantling the south slopes of Turrmstindane. By Bergen-Jotun Stem (Goldschmidt 1916). The Jo­ farthe most complete section occurs in Lange­ tun gneiss which forms the frame of the window dalsåi, and it is here that the structure of the is here termed the Svartdalen gneiss (Emmett sediments and their relationship to the Jotun 1980) and it is separated from more highly rocks are most clear. metamorphosed rocks to the north by a major

18 -GeologiskTidsskr. 4/80 256 T. F. Emmett NORSK GEOLOGISK TIDSSKRIFT 4 (1980)

THE TORFINNSBU} WINDOW SVARTOOEN DGNEISS . JOTUN STEM PYROXENITE

� RETROGAf:SSfD �GNEISS ...... _ BASAL THRUST

lZZJ vARIOUS AUGEN GNEISSES

I!II!II BIOTITE·SCHIST ARKOsf-AMPHIBOLITE-SCHIST D SEQUENCE

E3QUARTZ-SCHIST•AiiKOst' SEQUENCE fiiiiD llUARTZ!TE (I;] 0SPARAGHITE AHPHIBOUTE

�AMPHIBOLE-SCHIST IESJINTERLAYERfD PELITES b::!:J;JANO SPilRAGMITES 1t<. DIP AND STRIKE Of BYGDIN S... .2b..-LAG �' b AXIS LINEATION i:======';iF=='<""'��------r---�-�...::::- FAULTS B

6' 33'E LONG.

Fig. /. The location and geology of the Torfinnsbu window.

fault, the - fault (Battey & McRitchie 1973), which passes 13 km north­ A F2j north-west of Torfinnsbu. The Svartdalen gneiss f, F3,/_.( F :7 is the only Jotun stem rock seen in the frame of '. 1 the metasedimentary window. --;:�1';t;.-t

Pyriclasite gneisses. - The typical Jotun rock of this area is a medium to coarse - grained NNW pyriclasite gneiss (Berthelsen 1960) with weak or moderate foliation. The less well-foliated rep­ Ø1 SOUTH LANGEOALSÅI resentatives, which are analogous to the 1§12 500m MjØlkedØla purple gabbro of Battey & McRitchie (1973), preserve relict igneous (ophi­ / B F3 l tic) texture. The mineral paragenesis is cpx + /l' clinoamphibole - feldspar plag ± opx. ,'' . \ E + K + l w The cpx is augitic and rimmed with green, ,, \Il strongly pleochroic clinoamphibole (pargasite) \/l l\ dose to the basal thrust (Fig. 3). Opaque grains �- / often have a rim of pale yellow - brown to dark j}_;;· F3 brown pleochroic biotite. t-.::::.-13 EAST TORFINNSBU Ultrabasic rocks. - Two ultrabasic bodies oc­ approx 100m cur in the vicinity of Torfinnsbu, one in the south �4 slope of Turfmstindane, the other 3 km north of Torfinnsbu, in TurfinsdØla. They are believed to Geological cross-sections: represent a single body disrupted by lag-faulting. Fig. 2. Langedalsåi. l = amphibolite, = schists etc, B. Shore sec· The rock is a clinopyroxenite, composed essen­ A. 2 tion 200 m east of Torfinnsbu (sketch on! y). 3 = augen gneiss, tially of cpx + opx + clinoamphibole ± olivine. 4 = pegmatite. NORSK GEOLOGISK TIDSSKRIFT 4 (1980) Geology of Torfinnsbu window 257

.. .. r. ,,..,"- ,. . :'},_: 4 .�:�. t-��-

Fig. 3. Photomicrograph of Svartdalen gneiss showing amphibolisation due to F1 !brust­ ing. F = plagioclase, P= pyrox· ene, A= amphibole.

The olivine occurs as sparse, subrounded crys­ grain size and induced trails of cataclasis tals. The clinoamphibole is a green, strongly throughout the rock and altered feldspars to pleochroic pargasite. Plagioclase is a common saussunttc aggregates. Away from the basal trace mineral, occurring as finely granular rims thrust, the Svartdalen gneiss shows little evi­ to oxide grains (compare Battey & Davison dence of Caledonian deformation. This is similar 1977).In conformity with this finding,McRitchie to the effects of thrusting in the Vang window (1%5) reports that, around the west end of area (Heim et al. 1977). Bygdin, ultrabasics associated with the Mjølkedøla purple gabbro have plag + olivine assemblages which are stable. Base of the Svartdalen gneiss Metamorphism. - The co-existence of olivine and plagioclase in ultrabasic members of the Approximately 1.3 km north of the mouth of Svartdalen gneiss group implies that the condi­ Langedalsåi the overthrust nature of the Svart­ tions of metamorphism did not exceed those dalen gneiss is displayed. Fresh Svartdalen assigned to the plagioclase-lherzolite facies of gneiss passes downwards into a highly O'Hara (1%7) or the low-pressure granulite sub­ epidotised and strongly amphibolised zone in facies of Green & Ringwood (1967). The Svart­ which lenses of fresher gneiss are sometimes dalen gneiss must therefore be distinguished found. Relict foliation is sometimes seen. The from 'the undifferentiated jotunites' ofMcRitchie amphibolised zone is approximately 4 m thick (1%5) north of the Tyin-Gjende fault which and rests on a 1-2 m thick zone of intensely show extensive evidence of a prograde meta­ sheared basic rock which passes downward into morphism which has carried them into the an acidic augen gneiss. Similar occurrences of spinel-lherzolite facies (O'Hara 1967, Battey & augen gneiss are recorded in Espedalen (M. R. McRitchie 1975, Emmett 1980). This is equival­ Garton, pers. comm. 1979). A biotitic phyllonite ent to the intermediate pressure granulite sub­ separates the augen gneiss from underlying facies of Green & Ringwood (1967). In summary, quartz-biotiteschists. A pink feldspar pegmatite the Svartdalen gneiss can be distinguished from was found in the biotite-phyllonite zone which 'the undifferentiated jotunites' by their less cross-cuts and intrudes along the foliation in thorough degree of recrystallisation, and their part. This is interpreted as representing intrusion lower grade of metamorphism. of pegmatitic liquids during the formation of the The mechanical effects of the thrust that bring phyllonite. A similar relationship has been found the Svartdalen gneiss above the metasediments at Eidsbugarden (Battey & McRitchie 1973) and are restricted to a narrow band just above the at Bygdin (Hossack 1968). The total thickness of basal thrust. The deformation has reduced the the rocks described in this section, from the 258 T. F. Emmett NORSK GEOLOGISK TIDSSKRIFT 4 (1980)

upper limit of the retrograded Svartdalen gneiss S\".-\RTD.o\LEN GNEISS bt lr1sf 1000111 thic:kl to the base of the biotitic phyllonite is about 20 m. .\\tPHIBOLISED G�EISS l•ilh l�11ws of rr�sh aøei.,!!il

AlG['i G!IOEISS 1 BIOTITIC PHYLLONITE (•ilh s�·nttctøit "amatildj The Torfinnsbu metasediments The rocks occurring within the Torfinnsbu window are regarded as metasediments rather than highly sheared Jotun rocks for the following reasons: Quartz is ubiquitous in the metasedi­ ments and rare in the Svartdalen gneiss. In addition, hands of pure white quartzite occur in the Nybui section. Similarly, muscovite is omnipresent in the sediments and almost totally absent from the Svartdalen gneiss. Nodules of pink calcite and stringers of brown carbonate are common in certain horizons within the metasediments (Fig. 4). Carbonates are rare in the Jotun rocks. The metasediments are very schistose. Even dose to mylonitic shear zones the Svartdalen gneiss does not take on a prominent schistosity. The metamorphic grade of the sediments in­ creases upward towards the most intensely de­ formed zone at the base of the Svartdalen gneiss, ACICl'LAit AMPHIIOLE SCHIST whilst the latter shows increasing retrogression downward towards the same zone. This argues

GAP against the Torfinnsbu rocks being intensely deformed representatives of the Svartdalen gneiss.

Petrography of the metasediments

. lASE !IOOT SEEN Description. - The rocks below the basal thrust consist of a variety of sparagmites, quartz-mica schists, and amphibolites (Fig. 4). Quartz - mica schist is the commonest rock (IIIIHS llllotllt ... toiiMI) type and, although the exact mineral proportions � s,araa•itt are variable, the paragenesis is biotite + epidote

� ...,lli ltolilt + quartz + plagioclase ± muscovite ± K­ feldspar. The biotite is pleochroic from very pale ��"j vario1s whisU yellow-brown to deep olive green. It usually has a very strong preferred orientation parallel to the

1. BASAL THitUST COM PLEX ( F1 ) main foliation in the rock. The plagioclase is commonly about An10• The centres of feldspar 2. GIIHNSCHIST-HIPHIIOLITE TltANSITIONAL FACIES grains are oftensaussuritised, suggesting that the

3. GltEENSCHIST FACIES (?) present albitic plagioclase has been produced by metamorphism of plagioclase with initially more B ltasic sttnrH rock of BaSill ThrltSI calcic cores. T thnst

Fig. 4. Approximate lithotectonic stratigraphy of the Tor­ finnsbumetasediments. The sequence is probably inverted (cf. Hossack 1968, Heim et al. 1977). NORSK GEOLOGISK TIDSSKRIFT 4 (1980) Geology of Torfinnsbu window 259

Gamet has been found only in one rock, a strained by the Phase Rule, it is possible for up gamet-biotite-schist which occurs in Lan­ to four phases to form an equilibrium assemb­ gedalsåi, a few tens of metres below the basal lage. It seems, therefore, that the Torfinnsbu thrust. The gamet forms euhedral post-kinema­ metasediments betong to the greenschist­ tic crystals and its paragenesis is clinoamphibole amphibolitetransitional facies. + biotite + epidote + quartz + gamet. The gametiferous amphibolite has oligoclase Amphibolites are the second most common as its plagioclase, rather than albite. Hence its rock type. The essential paragenesis is paragenesis of clinoamphibole + biotite + epi­ clinoamphibole ± epidote ± quartz ± plag. An dote + quartz + gamet + oligoclase places it amphibolite from the lake shore east of firmly in the amphibolite facies (Turner 1%8: Torfinnsbu (and therefore not far below the basal 309, assemblage C4). thrust) has the assemblage clinoamphibole + biotite + quartz+ garnet. This rock contains a plagioclase (An25) somewhat more calcic than is usual in the Torfinnsbu sequence. The Structure of the Torfinnsbu amphibole in all the amphibolites is strongly metasediments pleochroic (very pale brown-green to deep olive Fo/ds. - The metasediments have one dominant green) and is probably homblendic. The biotite foliation which, although it is intensely refolded, is typically pleochroic from colourless to dark has a regional northerly dip of approx. 30". The brown. formation of this foliation, which is parallel to Sparagmites (arkoses) occur as layers varying lithological variations (isoclinally folded bed­ in thickness from 50 cm to approx. 5 cm. They ding), seems to be the earliest metamorphic­ are commonly white or off-white in colour. Dis­ structural event in these rocks. The base of the continuous green (or sometimes reddish) layers Jotun crystalline rocks, the basal thrust, is paral­ less than 5 mm thick are common. Occasional lel to this foliation (S0_1). layers take on the aspect of an epidosite, consist­ Two sets of minor coaxial intrafolial folds are ing almost entirely of quartz and epidote, but present, of which the earlier set, F., is only generally the rock is an arkose ( quartz + rarely seen. The se are isoclinal in form with axial feldspar + epidote + muscovite). The feldspar is planes parallel to S0_1, but they do appear to fold usually potassic and is often perthitic S0• This is identical to the situation in the (?mesoperthite). Plagioclase of intermediate Valdres Sparagmite Group at Bygdin where composition also occurs. Petrographically, the Hossack (1%8, fig. 4) illmtrates a fold similar to mineral assemblage indicates that the Torfinnsbu the F1 folds seen in Langedalsåi. Hossack post­ sparagmites have affinities with the Olefjell ulates a much larger scale F1 folding (which is sparagmite type of Loeschke (l%7). actually responsible for S1) on the basis of the Quartzites are of limited and sporadic distribu­ vergence of F1 minor structures. Similar evi­ tion. A particu1arly prominent band 2 m thick dence is not available from the Torfinnsbu area occurs in the Nybui section. They all are quite (Fig. 2). white and apparently pure. The F2 structures are fairly open folds, with Mineral facies. - The non-gametiferous as­ axial planes which dip more steeply than S0_1• semblages appear to represent a facies trans­ They may be divided into major and minor itional from greenschist to amphibolite (Turner representatives and they refold F1 folds, though 1968: 303-307). Similar parageneses have been this is rarely seen. recorded from the Dalradian succession in Scot­ F3 folds are of an open form with upright axial land (Barrow 1912). In the AKF diagram given planes and a wavelength of a few metres. They by Turner (1968: 305, fig. 7-22), the possible fold F2 b-lineations, and also S0_1• The base of parageneses include: the Jotun rocks is also folded by F3• It is this phase of folding which is most obvious in the K-feld + biotite + muscovite field. biotite + gamet + muscovite Lineations. - The b-axes of F., F2, and F3 folds both with quartz, albite, and epidote as addi­ all define a linear element within the rock fabric. tional phases. Miyashiro (1972: 129) pointed out Whenever F1 is seen, F1 and F2 are coaxial. F1 that since AKF diagrams are not strictly con- and F2 lineations all lay in So-1• Mineral defined 260 T. F. Emmett NORSK GEOLOGISK TIDSSKRIFT 4 (1980)

lineations are poorly developed. Early lineations closely with the deformational regime deduced are refolded by later folds, and this accounts for by Hossack (1968) in the Valdres Sparagmite the anomalous disposition of lineations (mainly Group of the Bygdin- region 10 km to the b-axes of F2 folds) in the mouth of Langedalsåi. east and below the crystalline nappe. Taken in The form of the culmination which has been conjunction with the deep embayment into the dissected to expose the window may be dis­ crystalline massif along Sjodalen, the new evi­ cemed in the regional disposition of lineations. dence lends strong support to the view that the The lineations dip approximately 20" northwest southeastem and eastem part of the Jotun massif in upper Langedalsåi and Nybui, northwards is a comparatively thin sheet.

west of Torfinnsbu and east in the exposures U nfortunately, lithological correlation be­ east of Torfinnsbu. This disposition may be tween the metasediments of the Torfinnsbu explained by assuming the formation of a some­ window and the established successions outside what elongate post-F3 dome which has its high­ the crystalline massif is not easy. The rocks are est point somewhere between Langedalsåi and broadly dissimilar to the Olefjell and Hecklefjell Nybui. Its flanks will dip fairly steeply away formations (Hossack 1976) which are im­ from its crest and it is this dome which has been mediately adjacent to the Jotun nappe at Bygdin, eroded at its crest to give the window. nor do they resemble the dark slates and shales

These data can be summarised in the following of Sjodalen. However, quartzite + arkose + way: three phases of deformation are recognised pelite ± amphibolite sequences are fairly com­ in the Torfinnsbu window and these are identical mon in central Norway (e.g., Dietrichson 1950) to those recorded at Bygdin by Hossack (l%8). and any attempt at lithological correlation must The Svartdalen gneiss was emplaced over the be very tentative at this stage. D. Twist (pers. metasediments during the first (and most in­ comm. 1978) has noted that the rock types and tense) phase of deformation. The areal distribu­ sequences at Torfinnsbu appear to be similar to tion of lineations can be related to the semi-el­ the much thicker successions he has studied lipsoidal form of the window. from the northwest margin of the Jotunheimen massif in Bøverdalen (D. Twist 1979), whilst M. R. Garton (pers. comm. 1979) has indicated that Discussion very similar rocks are to be found in the highest thrust sheet of the V aldres nappe. Despite the work of several geologists (Battey & The Jotun crystalline rocks were undoubtedly McRitchie 1973, Gjelsvik 1946, Dietrichson emplaced during F �> as is shown by the 1950, Hossack 1968, 1978, Heim et al. 1977) our structural conformity of the basal thrust with understanding of the emplacement of the Jotun So-1• This was accompanied by 1ow-grade pro­ nappe is still rather incomplete. Field evidence gressive metamorphism in the sediments and by

from the sub-Jotun units in Valdres (Hossack retrogression in the Jotun rocks. As l. Bryhni 1978) and from the northwest margin (Roberts (pers. comm. 1978) has pointed out, the basal 1977) conflicts with the geophysical (Smithson et thrust is a zone of metamorphic convergence, al. 1974) and some structural (Banham et al. with the metamorphic grade of the metasedi­ 1979) evidence. What is needed is much more ments increasing upwards towards the thrust data both from the Jotun stem rocks themselves (prograde metamorphism) and the grade of the and from the nappes beneath Jotunheimen. Jotun crystalline rocks decreasing downward Especially it must be determined whether sedi­ (essentially retrograde). The effect of retrogres­ ments of the Jotun-Valdres nappe (Heim et al. sion within the Jotun rocks is, however, essen­ 1977) reappear along the northwest margin of tially confined to a narrow zone immediately Jotunheimen, and whether such units can be above the basal thrust. traced right round the ends of the Jotun nappe. The Torfinnsbu window demonstrates that the Acknowledgements. -Dr. M. H. Battey, Dr. J. R. Hossack, and Mr. M. R. Garton read and criticised the initial draft of this sheet of crystalline rocks around the middle part paper. Mr. D. W. Kelsall prepared the figures for publication. of Bygdin, which is the southeastward extension The work was supported by an N. E. R. C. Research Student­ of the Jotun nappe, is thinner than that shown on ship. the geophysical proftle of Smithson et al. (1974).Moreover the structures in the metasedi­ ments within the window may be correlated NORSK GEOLOGISK TIDSSKRIFT 4 (1980) Geology of Torfinnsbu window 261

References Holtedahl, O. & Dons J. A. 1960: Geo1ogical map of Norway. Nor. Geo/. Unders. 208. Banharn, P. H. , Gibbs, A. D. & Hopper, F. W. M. 1979: Hossack,' J. R. 1968: Structural history of the Bygdin area·. Geological evidence in favour of a Jotunheimen Caledonian Nor. Geo/. Unders. 247, 78-107. suture. Nature 277, 289--291. Hossack, J. R. 1976: Geology and structure of the Beito Barrow, G. 1912: On the geology of lower Dee-side and the window. Nor. Geo/. Unders. 327,1-33. southem highland border. Proc. Geo/. Ass. 23, 274-90. Hossack, J. R. 1978: The correction of stratigraphic sections Battey, M. H. & Davison, W. 1977: Exsolution of plagioclase for tectonic finite strain in the Bygdin area, Norway. J. from clinopyroxene in a pyroxenite from Jotunheimen, Geo/. Soc. Lond. 135, 229--242. Norway. Mineral. Mag. 38, 435-45. Loeschke, J. 1967: Zur petrographie des Valdres Sparagmites Battey, M. H. & McRitchie, W. D. 1973: A geological traverse zwischen Bitihorn und Langnuen/Valdres. Nor. Geo/. Un­ across the pyroxene-granulites of the Jotunheimen in the ders. 243, 67-98. Norwegian Caledonides. Nor. Geo/. Tidsskr. 53, 237�5. McRitchie, W. D. 1965: Structure and geochemistry of the Battey, M. H. & McRitchie, W. D. 1975: The petrology of the layered metamorphic rocks near Eidsbugarden, Jotun­ pyroxene-granulite facies rocks of Jotunheimen, Norway. heimen, Norway. Unpublished Ph.D. thesis, University of Nor. Geo/. Tidsskr. 55, l-49. Durham. Berthelsen, A. 1%0: Structural studies in the Precambrian of Miyashiro, Y. 1973: Metamorphism and Metamorphic Beits. western Greenland Il: geology of Tovqussap Numi. Medd. George Allen and Unwin Ltd. London. 492 p. om Green/and, 123 (1), 14-24. O'Hara, M. J. 1967: Mineral parageneses in ultrabasic rocks. Dietrichson, B. 1950: Det Kaledoniske knuteområde i 393-402. In Wyllie, P. J. (ed. ) Ultrabasic and Related . Nor. Geo/. Tidsskr. 28, 65-143. Rocks. Wiley, New York. Emmett, T. F. 1980: The metamorphism and tectonic relation­ Roberts, J. L. 1976: Allochthonous origin of the Jotunheim ship between gabbros and granulites around Gjendebu, massif in S. Norway - a reconnaissance study along its NW Jotunheimen, central southem Norway. Unpub. Ph. D. margin. J. Geo/. Soc. Lond. 134, 351-362. the sis, University of Newcastle-upon-Tyne. Srnithson, S. B. , & Ramberg, I. B. & Grønlie, G. 1974: Gravity Gjelsvik, T. 1947: Anorthosittkomplekset i . Nor. Geo/. interpretation of the Norwegian Caledonides. Tectono­ Tidsskr. 26, 1-57. physics, 22, 205-222. Green, D. H. & Ringwood, A. E. 1967: An experimental Turner, F. J. 1968: Metamorphic Petro/ogy. McGraw Hill, investigation of the gabbro to eclogite transition and its New York. 403 p. petrological applications. Geochim. Cosmochim. Acta, 31, Twist, D. 1979: The structure and geochemistry of rocks along 767-833. the NW margin of the Jotunheim massif, southern Norway. Goldschmidt, V. M. 1916: Obersicht der Eruptivgesteine im Unpublished Ph. D. thesis, University of Newcastle-upon­ Kaledonischen Gebirge zwischen Stavanger und Trond­ Tyne. hjem. Skr. Vidensk. Selsk. Christ. 2, 140 p. Heim, M. , Scharerer, U. & Milnes, A. G. 1977: The nappe March 1979, revised August 1980. complex in the Tyin-Bygdin-Vang region, , Nor­ way. Nor. Geo/. Tidsskr. 57, 171-178.