Njg 1 vol93

B-Economique B-Blad 2013 Number 1, Volume 93

Return: NGF c/o NGU NO–7491 Trondheim NORWAY

Norwegian Journal of Geology

Per Inge Myhre, Fernando Corfu, Steffen G. Bergh & Kåre Kullerud

www.geologi.no/njg

P.P. Norge/Norvège

No. 1, Vol. 93

NORWEGIAN JOURNAL OF GEOLOGY

Porto betalt ved innleveringen

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex, North Norway ................................. 1

Raymond S. Eilertsen, Nils R.B. Olsen, Nils Rüther & Peggy Zinke Channel-bed changes in distributaries of the lake Øyeren delta, southern Norway, revealed by interferometric sidescan sonar ............ 25

Abdus Samad Azad, Henning Dypvik, Fridtjof Riis & Elin Kalleson Late post-impact sedimentation in the Ritland impact structure, western Norway .................................................................... 37

Rannveig Øvrevik Skoglund & Stein-Erik Lauritzen Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway .............................. 59

ISBN 978-82-92-39481-6

ISSN 0029-196X ISBN 978-82-92-39481-6

9 788292 394816

2013 Number 1 Volume 93

NORWEGIAN JOURNAL OF GEOLOGY Norsk Geologisk Tidsskrift

Norwegian Journal of Geology Norsk Geologisk Tidsskrift The main journal for Norwegian geological research, distributed to all members of the Norwegian Geological Society. First published in 1905, this geological journal is issued quarterly by the Norwegian Geological Society (Norsk Geologisk Forening). The journal publishes research articles, review articles, papers, notes and discussions relevant to the varied and complex geology of Norway, the Arctic regions, and adjacent offshore areas.

Scientific papers from all geological, geophysical and geo­chemical disciplines are considered for publication.

The contributors­, referees and readership are widely international.

Instructions to authors Instructions to authors are found on www.geologi.no/njg

Indexed in

Trond Slagstad, Trondheim

Biological Abstracts (-1985); British Geological Literature; Chemical Abstracts; Current Contents/ Physical, Chemical and Earth Sciences; Energy Research Abstracts; Geological Abstracts Economic Geology; Geographical AstractsI/ Physical Geography; GeoREF (Online Database for Bibliography and Index of Geology); Life Sciences Collection; Norske Tidsskriftartikler; Metal Abstracts; Mineralogical Abstracts; Petroleum Abstracts; Science Citation Index; World Aluminum Abstracts

Editorial Board

Subscription price

Silje Berg and Christine Fichler, Statoil, Terje Bjerkgård, NGU, Henriette Linge, UiB, Sverre Planke, VBPR, Anders Solheim, NGI.

Subscription rates 2013: EUR 280 In Scandinavia: NOK 1600

Editor

Manuscript address Manuscripts shall be addressed to: Editor of Norwegian Journal of Geology N-7491 Trondheim, Norway Tel: +47 73 90 40 00, Fax: +47 73 92 16 20 E-mail: trond.slagstad@ngu.no

P

RI

NT

R

241

731

Graphic production by Skipnes AS

T E D M AT

E

Front cover The cave stream in the main passage in Grønligrotta at base flow.

Photo by Stein-Erik Lauritzen.

NORWEGIAN JOURNAL OF GEOLOGY

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

1

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex, North Norway Per Inge Myhre, Fernando Corfu, Steffen G. Bergh & Kåre Kullerud Myhre, P.I., Corfu, F., Bergh S.G. & Kullerud, K.: U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex, North Norway. Norwegian Journal of Geology, Vol 93, pp. 1–24. Trondheim 2013, ISSN 029-196X. An Archaean geotransect is well preserved within the West Troms Basement Complex, a basement outlier west of the Caledonides in North Norway, whose structural architecture is the finite result of both Archaean and Palaeoproterozoic (Svecofennian) tectonic events. Geochronology by U–Pb TIMS and SIMS of nine samples of magmatic and migmatitic rocks collected in a transect perpendicular to the NW–SE strike of the complex record three main stages of Archaean magmatism and one superimposed, Neoarchaean, metamorphic high-grade event. The oldest Mesoarchaean ages of 2.92–2.80 Ga are found within a paired tonalitic complex and greenstone belt in the northeast. These rocks are separated by a high-grade, ductile shear zone from younger Neoarchaean rocks with only local, inherited, Mesoarchaean zircons. The period from 2.75–2.70 Ga was particularly active in the rest of the complex, with intrusion of diorite-granodiorite plutons on Kvaløya and Senja, followed by local migmatisation and another pulse of diorite-granite magmatism between 2.70 and 2.67 Ga. This period was concluded by the intrusion of mafic dykes at 2.671 Ga, documented in one locality on Kvaløya. A latest Neoarchaean age for stromatic migmatite or metamorphism is indicated in two localities by secondary zircon in neosome, with restitic zircon derived from the 2.70–2.67 Ga event. The Archaean rocks were variably reworked, metamorphosed and intruded by felsic and mafic plutons during the Svecofennian (1.8–1.7 Ga) orogeny, and locally also formed the substrate to Palaeoproterozoic supracrustal rocks.

Possible correlatives are found in the Lofoten and Vesterålen area to the south and in the Karelian and Norrbotten provinces in the Fennoscandian Shield. Given the tectonostratigraphic position of the West Troms Basement Complex, correlations with Archaean provinces elsewhere in the North Atlantic region are also possible.

Per Inge Myhre*, Steffen Bergh, Kåre Kullerud, Department of Geology, University of Tromsø, Dramsveien 201, 9037 Tromsø, Norway. Fernando Corfu, Department of Geosciences, University of Oslo, P.O. Box 1047 Blindern, 0316 Oslo. *Present address, Per Inge Myhre, Statoil ASA, Mølnholtet 42, Harstad, Norway.

E-mail corresponding author (Per Inge Myhre): peringem@gmail.com

Introduction The West Troms Basement Complex occupies the coastal islands of Troms, North Norway (Fig. 1). The region is characterised by Archaean gneisses with varied protoliths, Archaean and Palaeoproterozoic greenstone belts and Svecofennian bimodal intrusions with ages around 1.8 Ga. The West Troms Basement Complex is bounded against the Caledonian nappe stack in the east by extensional and local thrust faults, and represents a basement outlier in a similar tectonic position as Lofoten and Vesterålen. An understanding of the main geological units in the West Troms Basement Complex, including the occurrence of Archaean crust, had already been established (Zwaan 1995; Zwaan et al., 1998; Corfu et al., 2003; Kullerud et al., 2006a; Bergh et al., 2007), see synthesis in Bergh et al. (2010), and recent advances have highlighted many interesting questions that remain to be resolved. For example, an important question concerns the extent of Archaean crust in the region, and whether the many supracrustal belts represent major tectonic boundaries between older crustal segments, or episodes of basin

formation within the Archaean continent (Myhre et al., 2011). Another issue involves possible correlations with other Archaean–Palaeoproterozoic terranes; whereas the West Troms Basement Complex and the Precambrian rocks in Vesterålen and Lofoten to the south may simply be considered as an extension of the Fennoscandian craton (Zwaan, 1995), the tectonostratigraphic position within the Caledonian framework remains to some extent uncertain. For example, the lack of Caledonian structures and metamorphism stands in sharp contrast to basement outliers farther south in Norway, where Caledonian high-grade metamorphism of the Precambrian rocks is widespread (Andersen et al., 1991) or, in the case of Lofoten, at least evident (Steltenpohl et al., 2011).

In order to contribute to these issues and to the questions raised (Bergh et al., 2010), we present field observations and U–Pb geochronology data from seven localities of Meso–Neoarchaean metamorphic rocks along a NE–SW geotransect. The main goal of this paper has been to apply U–Pb geochronology to identify episodes of Archaean magmatism, metamorphism and tectonism, and to better

2

P.I. Myhre et al.

constrain the significance of these events with respect to crustal generation, evolution and assembly history.

In particular, the identification of major crustal shear zones and potential terrane boundaries can be resolved by dating. A supplementary goal involves the possibility of correlation with other Neoarchaean provinces in Fennoscandia (Bergh et al., in press).

Existing data and regional setting Ringvassøya The main lithologies on Ringvassøya include tonalitic orthogneisses (the Dåfjord gneiss), the Ringvassøya greenstone belt with low- to medium-grade volcanic and sedimentary rocks (two metavolcanic rocks yielding ages of 2835 ± 14 and 2849 ± 4 Ma, Motuza et al., 2001) and a banded gneiss unit called the Kvalsund gneiss (Zwaan et al., 1998). The Dåfjord gneiss with 2849 ± 3 Ma tonalite (Zwaan & Tucker, 1996) is considered to be related to tonalitic rocks on the islands to the north of Ringvassøya, and a sample from Vannøya has a U–Pb age of 2885 ± 20 Ma (Bergh et al., 2007). A small alkaline stock intruded tonalitic rocks of the Dåfjord gneiss at 2695 ± 15 Ma (Zozulya et al., 2009). Presumed Neoarchaean structures on Ringvassøya include a widespread gneissosity and metamorphic banding of the tonalites and comprehensively migmatised, ductile shear zones that are parallel to the main gneissic foliation. The Dåfjord and Kvalsund gneisses, as well as the Ringvassøya greenstone belt, are cut by an extensive mafic dyke swarm with a crystallisation age of 2403 ± 3 Ma, and with metamorphic titanite at 1768 ± 4 Ma (Kullerud et al., 2006a). A major synformal folding of the Ringvassøya greenstone belt and a network of ductile shear zones (Fig. 2) that truncate all other structures, including the 2403 Ma mafic dyke swarm, are thought to be of Svecofennian age (Bergh et al., 2010).

Kvaløya The geology of Kvaløya is dominated by Neoarchaean gneisses and intrusions, the 1792 ± 5 Ma Ersfjord granite (Corfu et al., 2003), and narrow, NW–SE-trending, Palaeoproterozoic (2.00–1.95 Ga, Myhre et al., 2011) metasupracrustal belts (Fig. 1). The Archaean rocks are subdivided into the Kvalsund gneiss, Gråtind migmatite, Bakkejord pluton and Kattfjord gneiss (Zwaan et al., 1998; Bergh et al., 2010). A Neoarchaean age for these units is based on dating of a granite near Torsnes at 2689 ± 6 Ma (Corfu et al., 2003), the age of the Bakkejord pluton and mafic dykes documented here (also in Kullerud et al., 2006b) and regional correlation. The supracrustal belts are made up of both metasedimentary and metavolcanic rocks present as amphibolites, with local ultramafic bodies (Armitage & Bergh, 2005; Myhre et al., 2011).

NORWEGIAN JOURNAL OF GEOLOGY

The metasupracrustal belts separate different Archaean gneisses (Fig. 1) and played an important role in Svecofennian deformation, recording a threephase evolution of thrusting, folding and strike-slip deformation (Armitage & Bergh, 2005; Bergh et al., 2010). The boundary between the Bakkejord pluton and the Mjelde–­ Skorelvvatn belt is strongly sheared as a result of NE-directed Svecofennian thrusting and superimposed transpression (Armitage & Bergh, 2005). In the final stage of this event, a major antiform developed after the initial thrusting, with the Gråtind Migmatite in its core, and with the Bakkejord pluton constituting the southwestern limb and the Kvalsund gneiss the northeastern limb. Mafic dykes are prominent within the Neoarchaean rocks (Zwaan et al., 1998; Armitage & Bergh, 2005; Bergh et al., 2010), displaying various degrees of reworking (Armitage & Bergh, 2005; Bergh et al., 2010).

Senja The northeastern part of Senja is occupied by the c. 30 km-wide, NNW–SSE-trending Senja Shear Belt (Zwaan 1995; Bergh et al., 2010), bounded by the Torsnes belt in the northeast and the Svanfjellet belt in the southwest.

The lithologies within the Senja Shear Belt are diverse, with a predominance of banded felsic and mafic gneisses with abundant migmatisation in the northeastern part and more homogeneous, locally mylonitic granitic intrusions with abundant tectonic lenses of greenstone along strike in the southwestern part (Fig. 1; Zwaan, 1995; Zwaan et al., 1998; Zwaan & Fareth, 2005). Considerable uncertainty is associated with the age and origin of the rocks within the Senja Shear Belt, and presently the only available radiometric date is a U–Pb zircon age of 1778 ± 3 Ma (Zwaan et al., 1998) for neosome in the ‘Motind migmatite’. By lithological correlation, and since the gneisses also contain older components than neosome, both Neoarchaean and Palaeoproterozoic origins can be assumed for this part of Senja (Fig. 1; Bergh et al., 2010) but the relative proportions of either component are not known. Minor remnants of greenstone belts (marked on Fig. 1) within the banded gneisses display local evidence of volcanic and sedimentary origins (Zwaan et al., 1998), and may be either Archaean or Palaeoproterozoic. Some of the greenstone units contain graphite deposits of economic significance (Henderson & Kendrick, 2003). To the southeast of the banded gneisses in the Senja Shear Belt, a large, more massive, granitic unit is mapped as Palaeoproterozoic, the age inferred by correlation with similar granites on Senja and Kvaløya (Fig. 1; Bergh et al., 2010).

The southwestern margin of the Senja Shear Belt is defined by the Svanfjellet belt (Fig. 1; Zwaan et al., 1998; Bergh et al., 2010). The rocks on either side of the Svanfjellet belt are dominated by orthogneisses with various compositions ranging from diorite to granite (Zwaan et al., 1998, 2003).

The Svanfjellet belt itself is made up of sheared versions of

NORWEGIAN JOURNAL OF GEOLOGY

Figure 1. Geological map of the West Troms Basement Complex, modified from Bergh et al.

(2010). Locations of detailed maps, samples and cross section (Fig. 10) are marked.

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

West West Troms Troms Basement Basement Complex Complex

d. d.

Vesterålen Vesterålen

3

Vannøya Vannøya

Lofoten Lofoten

Ringvassøya Ringvassøya greenstone greenstone beltbelt c. c. e. e.

Caledonian Caledonian nappes nappes Neoproterozoic Neoproterozoic andand Phanerozoic Phanerozoic rocks rocks Proterozoic Proterozoic rocks rocks (1.7(1.7 - 0.9 - 0.9 Ga)Ga) Trans-Scandinavian Trans-Scandinavian Intrusive Intrusive BeltBelt (1.85 (1.85 Ga Ga - 1.65 - 1.65 Ga)Ga) Palaeoproterozoic Palaeoproterozoic rocks rocks (1.95 (1.95 - 1.75 - 1.75 Ga)Ga) Palaeoproterozoic Palaeoproterozoic rocks rocks (2.5(2.5 - 1.95 - 1.95 Ga)Ga) Archaean Archaean rocks rocks (3.5(3.5 - 2.5 - 2.5 Ga)Ga)

c01-107 c01-107

Ringvassøya Ringvassøya c01-109 & 110 c01-109 & 110 Dåfjord Dåfjord gneiss gneiss

Kattfjord Kattfjord gneiss gneiss

pim07-67 & 69 pim07-67 & 69 Kvaløya Kvaløya

Gråtind Gråtind migmatite migmatite

Sb Sb b. b.

Hamn Hamn gabbro gabbro

c04-31 c04-31 & 35 & 35

Granitoid Granitoid intrusives intrusives (1.8-1.7 (1.8-1.7 Ga) Ga)

Bakkejord Bakkejord pluton pluton

Gabbroid Gabbroid intrusives intrusives (1.8-1.7 (1.8-1.7 Ga, Ga, 2.22.2 GaGa onon Vannøya) Vannøya) Diorite Diorite - granodiorite - granodiorite (Neoarchaean (Neoarchaean / Palaeoproterozoic) / Palaeoproterozoic)

Senja Senja

Supracrustal Supracrustal belts belts (Meso-Neoarchaean (Meso-Neoarchaean / Palaeoproterozoic) / Palaeoproterozoic)

Fig.Fig. 4 4

c04-46 & 47 c04-46 & 47

b MSb MS

s ne rs s To senlet rb To elt b

pa04-2 pa04-2

Southwest Southwest Senja Senja gneiss gneiss complex complex

Tromsø Tromsø

.

Ab .

Ab

lt lt be be ar ar heshe s njanja SeSe

Kvalsund Kvalsund gneiss gneiss

Fig.Fig. 3 3

Ca Caled ledon onian ian na nappe pp s es

10km 10km

Fig.Fig. 2 2

Svanfjellet Svanfjellet beltbelt

a. a.

Mafic Mafic dykes dykes Felsic Felsic and and mafic mafic gneisses, gneisses, partly partly migmatitic migmatitic (Neoarchaean (Neoarchaean with with Palaeoproterozoic Palaeoproterozoic components) components) Tonalitic Tonalitic gneiss gneiss (Meso-Neoarchaean) (Meso-Neoarchaean) Thrust Thrust /normal /normal fault fault (Caledonian) (Caledonian) Steep Steep ductile ductile shear shear zone zone Trace Trace of of gneiss gneiss foliation foliation Ductile Ductile thrust thrust Abbreviations: Abbreviations: AbAb = Astridal = Astridal belt belt (Svecofennian) (Svecofennian) MSb MSb = Mjelde-Skorelvvatn = Mjelde-Skorelvvatn belt, belt, SbSb = Steinskartind = Steinskartind belt belt

Myhre Myhre etet al.,al., Figure Figure 11

these surrounding rocks and minor supracrustals (Zwaan et al., 2003; Zwaan & Fareth, 2005), and the rocks along the southwestern margin are thought to have been thrust northeastwards on top of presumed younger lithologies based on the presence of a stretching lineation and asymmetric folds (Bergh et al., 2010). The structures in the orthogneisses tend to wrap around the Hamn gabbro (Fig. 1) dated to 1.80 Ga (Zwaan et al., 1998; Kullerud et al., 2006b), which probably acted as a tectonic lens during the deformation. Thus, at least some of the deformation can be constrained to be Svecofennian (Bergh et al., 2010).

To the southwest of the Svanfjellet belt there is a 150– 200 km2 unit of felsic and mafic gneisses with abundant migmatisation (informally labelled ‘Southwest Senja gneiss complex’ in Fig. 1), considered to consist of Neoarchaean rocks with Palaeoproterozoic components. The margins of this unit are intruded by voluminous gabbros, (quartz)diorite and granites with Palaeoproterozoic ages (Zwaan et al., 1998).

Field observations Skarsfjord area, Ringvassøya The study area on Ringvassøya (Fig. 1), shown in Fig. 2A, includes the Ringvassøya greenstone belt and the Dåfjord gneiss in the northern and central parts of the area, respectively. The contact between the two units is a steeply dipping thrust fault (Fig. 2A) that is laterally sheared and offset by younger ductile shear zones (Bergh et al., 2010).

In the southwestern part of Ringvassøya, the dominantly tonalitic Dåfjord gneiss is in contact with the Kvalsund gneiss (Fig. 2A), which is a much more heterogeneous unit of banded gneisses with extensive migmatisation.

This NW–SE-striking contact is exposed along the inner parts of Skarsfjorden and southwards to Kvalsund (Fig. 2). Along the shore section in inner Skarsfjorden, the contact is characterised by a transition from weakly foliated tonalite with gentle NE dips to steeply NE-dipping mylonitic tonalite, eventually grading into banded gneisses along the contact shown in Fig. 2A, from Zwaan et al.

4

P.I. Myhre et al.

NORWEGIAN JOURNAL OF GEOLOGY

a.

A Photo location and direction of view Palaeoproterozoic granite Meta­dolerite dykes, 2.4 Ga c01­107

Kvalsund gneiss

Skarsfjorden

Ringvassøya greenstone belt c.

C

d.

D

7765000

Bb

Dåfjord gneiss Trace of gneissosity Thrust fault (ductile shear zones) Semi­ductile fault / shear zone Trace of synform / antiform Fault with shear sense

7760000

Shear zone c01­109 & 110 Kv al su

nd

645000

e.

E

1 650000

b.

B

d.

D

5 kilometres

655000

c.

C

Figure 2. (A) Geological map of the Skarsfjord area (based on Bergh et al., 2010): WGS 84 UTM 33N coordinates in metres are used throughout. Location of analysed samples is shown by sample number; location and direction of photos are shown with blue arrows. (B) and (C) rotated gabbroid clasts within sheared gneiss with rims of neosome and accumulation of neosome in dilatant sites, indicating partial melting during shearing. (D) Stromatic migmatite characteristic of the Kvalsund gneiss.

Note 2–3 dm-thick mafic dyke cutting the NW–SE-trending gneiss fabric. (E) dmthick neosome sheets­ alternating with palaeo­some layers in Kvalsund gneiss. The analysed sample c01– 110 is from a sheet of this type and sample c01–109 represents a mafic enclave.

e.

E

0.5 m

Myhre et al. Figure 2

(1998). The banded gneisses are made up of alternating quartzo-feldspathic and mafic bands, clasts or rafts (Fig. 2B–E) that range in composition from amphibolite to leucocratic metagabbro. The mafic bodies within the shear zone occur either as rotated clasts enveloped within the shear fabric or as strongly deformed bands that make up the banding of the gneisses (Fig. 2B, C). Rims of neosome are common around the clasts and particularly in dilatant sites, indicating that there was some partial melting

*Migmatite terminology following Sawyer (2008).

during deformation (Fig. 2B, C). Besides these dynamic migmatitic* structures, stromatic migmatite dominates within the sheared parts (Fig. 2D, E), and also in the remainder of the Kvalsund gneiss on Ringvassøya and Kvaløya, with common metre-scale neosome sheets (Fig. 2E). Locally, the shear fabric and stromatic migmatite are cut by mafic dykes (Fig. 2D).

NORWEGIAN JOURNAL OF GEOLOGY

a.A

617000

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

619000

Photo location and direction of view

Brensholmen 7723000

Amphibolite +/­ garnet Psammite Basal conglomerate Felsic dykes

c.

C

pim07­69 b.

B 7722000

5

Torsnes belt

Mafic dykes Kattfjord gneiss, granitoid gneiss

d.

D

Kattfjord gneiss, migmatitic part

pim07­68 pim07­67

Trace / orientation of foliation Thrust Orientation of fold axis

e.

E

KV2 (Corfu et al 2003)

Torsnes 1 kilometre

b.

B

c.

C

d.

D

e.

E

20 cm

Myhre et al., Figure 3

Figure 3. (A) Geological map of the Torsnes area (based on Bergh et al., 2010). (B) Migmatite with dm-thick neosome layers that were sampled for geochronology­(pim07–69). Vertical­trending (on photo) white lines are epidote-clinozoisite-altered cracks unrelat­ed to the migmatitic structure.

(C) Layered and folded migmatitic gneiss of the Kattfjord gneiss from Brensholmen. Note intrafolial asymmetric fold in amphibolite component of the gneiss. The foliation is shown on the map as black stippled lines in the Kattfjord gneiss. (D) Contact between mafic dyke (left) and migmatised Kattfjord gneiss (right). (E) Cliff face, c. 150 m high, with alternating sheets of mafic and felsic dykes in the Kattfjord gneiss. Note how the dykes are offset and disrupted by Mesozoic brittle faults.

Torsnes area, Kvaløya The study area between Brensholmen and Torsnes on Kvaløya is shown in Fig. 3A. The Kattfjord gneiss here is made up of layered migmatitic gneiss grading into granitoid gneiss to the southwest, and is overlain by folded and sheared conglomerate, psammite and amphibolite of the Torsnes belt with a maximum deposition age of 1970

± 14 Ma, given by detrital zircon dating (Myhre et al., 2011). The migmatitic part of the Kattfjord gneiss displays a variety of mafic enclaves and migmatitic structures illustrated in Fig. 3B–D, reflecting local variations in the degree of partial melting and multiple generations of magmatism. An example of intense partial melting is shown in Fig. 3B, with a layered or locally chaotic

6

P.I. Myhre et al.

migmatitic structure. Individual layers of neosome, up to 50 cm thick, occur within this unit and were sampled for geochronology (sample pim07–69). The melanosome here is made up of incoherent amphibolite patches and layers. Fig. 3C shows an outcrop of layered and folded migmatitic gneiss with a more coherent planar structure and overall N–S-trending gneissic foliation. This type of fabric is illustrated on the map (Fig. 3A) and is interpreted as Neoarchaean (Bergh et al., 2010). When approaching the Torsnes belt, this Neoarchaean fabric aligns with the belt and becomes overprinted by a NW–SE-trending foliation of Svecofennian origin. The migmatitic part of the Kattfjord gneiss grades into homogeneous and variably sheared granitoid gneisses approximately along the contact shown in Fig. 3A. The transition from migmatitic to homogeneous granitoid gneisses is rather gradual and likely an original primary igneous contact, with superimposed Svecofennian shear deformation. A granitoid body with a similar appearance is found at the southwestern margin of the Torsnes belt, where it has been dated to 2689 ± 6 Ma by Corfu et al. (2003). A suite of mafic dykes shown on the map and in Fig. 3D, E has a distinctly different appearance compared to the mafic banding within the gneisses and tends to cut the migmatitic fabric (Fig. 3D). Primary boundaries between these mafic dykes and the country rock are somewhat irregular or undulating on the small scale (Fig. 3D), locally with some melting of the country rock, indicating that the dykes were emplaced into hot crust. However, for the most part the mafic dykes display sheared margins, boudinage-like structures and ductile folding and are cut by semiductile faults. Due to this, the original trend of the dykes is rather disturbed.

They are commonly intimately associated with felsic dykes (Fig. 3E), as alternating mafic and felsic dykes or sheets.

Along the contact with the Torsnes belt, the mafic dykes, as well as the country rock, are unconformably overlain by a basal conglomerate, showing that the Kattfjord gneiss with mafic dykes predates these supracrustals. This depositional contact and the overlying strata were modified during the Svecofennian orogeny by NE-directed, subhorizontal shortening, producing the macroscale, synformal attitude of the belt, followed by subvertical folding and sinistral shearing, as described in Bergh et al. (2010).

Grunnfarnes area, southwestern Senja The shore section near Grunnfarnes (Fig. 4A) represents a complex mixture of mafic and felsic gneisses with extensive migmatisation. The gneisses belong to a 100–150 km2 suite of similar rocks on Senja (informally labelled Southwest Senja gneiss complex in Fig. 1). The shore outcrops here are made up of a pre-migmatitic unit (consisting of diorite to granodiorite gneiss and amphibolite), stromatic migmatite, and various mafic and felsic dykes (Fig. 4A).

Fig. 4B shows a raft of pre-migmatitic amphibolite with in situ melting structures, and Fig. 4C shows a dioritegranodiorite raft with folded, leucocratic, neosome layers of variable thickness. The axial plane of folds dips 60–80° to the W or WSW and approximates the overall

NORWEGIAN JOURNAL OF GEOLOGY

trend of the stromatic migmatite. For the most part, the migmatisation is rather more pervasive than in the mafic and granodioritic rafts seen in Fig. 4B, C, and the rocks occur as folded stromatic migmatite and locally chaotically structured migmatite. Individual neosome layers vary in thickness from 1 mm up to at least 50 cm; sample c04–47 represents such layers. The pre-migmatitic rocks and the stromatic migmatite are cut by mafic dykes that truncate the folded, stromatic-migmatite structure (Fig. 4C, D).

These dykes have themselves been metamorphosed, and are nearly completely transformed to amphibolites, locally with garnet. Finally, the pre-migmatitic rocks, stromatic migmatite and mafic dykes were cut by a second generation of aplites and associated sheets of leucogranite (Fig. 4D, E). The leucogranite sheets are commonly rich in assimilated material from the country rock.

Sample descriptions and U–Pb results Samples for U–Pb dating were crushed in a jaw crusher and hammer mill, and heavy minerals separated using Wilfley table, magnetic separators and heavy-liquid flotation at the University of Oslo. Zircons were studied and picked using a binocular microscope. Zircons analysed by SIMS were mounted on double-sided tape and cast in epoxy mounts and gold coated at the Natural History Museum in Stockholm, where initial backscattered electron (BSE) and cathodoluminescence (CL) imaging was also carried out. Post-analytical imaging was done at the University of Oslo. Zircons from seven samples were analysed using the Cameca IMS 1270 ion microprobe at NORDSIM, Stockholm, following analytical parameters and procedures described by Whitehouse et al.

(1999) and Whitehouse & Kamber (2005). The size of the ion microprobe analytical spots was approximately 20 μm in the long dimension (see images in Figs. 5, 6, 7, 8). ID– TIMS analyses were carried out at the University of Oslo following the procedures described in Corfu (2004). All zircons dated by ID–TIMS were abraded, either with air abrasion or chemical abrasion (CA–ID–TIMS, labelled ‘ca’ in Table 1 and plotted with red ellipses in the concordia diagram). The CA approach involved baking of an initial selection of zircons at 900°C in quartz vials for three days.

The annealed zircons were then transferred to Teflon bombs with HF at 194°C overnight, followed by final selection of zircon and cleaning.

Tonalite in a sheared part of the Dåfjord gneiss, Skarsfjord area, Ringvassøya (c01–107) This sample is a medium- to fine-grained, foliated orthogneiss, part of the rather extensive and homogeneous tonalitic Dåfjord gneiss (Fig. 2A). The sample was collected near the shear zone indicated in Fig. 2A. Based on thinsection observation, it may be classified as a tonalite with subordinate amounts of fine-grained biotite. The foliation is expressed as a shape-preferred orientation of quartz and plagioclase and as alternating fine- and coarse-grained

NORWEGIAN JOURNAL OF GEOLOGY

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

a.A

Grunnfarnes area Photo location and direction of view Secondary event aplites / leucogranite sheets Mafic (amphibolite) dykes e.

E

Stromatic migmatite w/ trace of layering Pre­migmatitic amphibolite rafts

7689000

No

Pre­migmatitic diorite ­ granodiorite rafts

xpo

te

Figure 4. (A) Geological map of the Grunnfarnes area. (B) Metabasite raft with in situ partial-melting structures. (C) Pre-migmatitic (see description in text) diorite-granodiorite with folded neosome layers, cut by mafic dyke. (D) Stromatic migmatite cut by mafic dyke, subsequently invaded by secondary aplite and leucogranite sheets. Assimilated fragments can be observed in the leucogranite sheet. (E) Cliff section showing migmatite gneisses and mafic dykes cut by secondary leucogranite sheets.

7

sed

Trace of main foliation Orientation of foliation / fold axis

65

d.

D

quarry

c04­47 c04­46

B b.

10­30

7688750

60

250 metres

575750

576000

80 Road

c.

C

c.

C

Bb.

c. 1 m d.

D

Secondary leucogranite sheet Assimilated clasts

Stromatic migmatite

e.

E

Mafic dyke

c. 250 m

0.5 m

Myhre et al., Figure 4

bands. Quartz and plagioclase grain boundaries show irregular and lobate shapes indicative of high-temperature deformation. In the heavy-mineral separate, zircon occurs as three main types: (a) elongate prisms with length/width ratio of 3 to 4 (Fig. 5E, e.g., grains 6 to 10), (b) shorter prisms, some with core/rim overgrowths (Fig 5e, e.g., grains 14, 12 or 16), and (c) tips of short prisms. Internal zoning is present in both types of prisms (e.g., Fig. 5e, grains 9, 10, 11) and also within some of the overgrowths (e.g., grain 14r), but boundaries between different generations within grains are not well defined in most cases. Five fractions of prismatic zircon and two fractions of tips were analysed by TIMS and the data show variable degrees of discordance, not defining one single discordia line (Fig. 5A, B). Rather, they constitute an array of discordant points bounded by two discordia lines, one with an upper intercept around 3.0–2.9 Ga (lower intercept 1.8–1.0 Ga) and the other with

an upper intercept of c. 2.78 Ga (lower intercept 0.40 Ga), indicating the presence of two or more age populations.

In order to resolve this complexity, zircons of all three morphotypes (elongate prisms, short prisms with cores, and tips/rims) were analysed by SIMS. These data plot on an array of overlapping concordant ages from c. 2.93 to 2.79 Ga, with the core and short prisms giving the oldest ages. Since many of the grains do not have well-defined internal boundaries between young and old parts, many of the SIMS analyses could be sampling mixtures. In that case, each end of the concordant array would estimate the oldest and youngest components, respectively. The oldest component can be calculated using three concordant SIMS analyses on short prisms/cores (Fig. 5B, analyses 8, 14c, 17), giving a weighted average 207/206Pb age of 2922 ± 7 Ma. Two TIMS fractions (Fig. 5B, analyses 157/17 and 82/4) that control the upper intercept of the older of the

1

Wt.2 (ug)

U (ppm)

Th/U3 Pbc4 calc. (pg)

Pb5 Pb

204

206

Pb6 U

235

207

2σ (abs)

Pb6 U

238

206

2σ (abs)

1

251/55 pr ca (1)

272

290

151

28

0.27

0.30

0.31

0.46

0.21

0.27

0.36

1.6

1.0

3.9

5.3

2.5

2.3

4.3

5391

7862

1179

1395

946

5952

3618

13.169

11.642

13.031

14.693

11.977

11.984

13.836

1

82/2

719

473 0.21

1.24 2.1

5.7 8699

19412 8.485

11.267

7

0.034

0.032

0.037

0.032

0.054

0.039

0.049

0.054

0.044

0.3975

0.4135

0.4986

0.4465

0.4827

0.5215

0.4541

0.4642

0.4986

1

253/62 ca tip/shell frgm (1)

234

166

223

>14

>69

194

257

288

38

0.48

0.38

0.62

0.79

0.66

0.55

0.69

0.29

0.82

0.62

0.70

0.66

2.0

1.3

1.9

1.7

8.2

16.3

3.5

2.7

2.2

6.5

9.2

6.2

3700

3159

3774

7161

21367

3849

12330

53709

8521

43706

26524

13018

1

121/56 eu sp incl (1)

370

1

30

117/60 (1)

121/55 eq sp incl (1)

211

11

14

102/4 tips subhedral (15)

117/59 cl (1)

135

4

520

161

236

486

18

102/2 eu spr (1)

102/3 eu (4)

0.78

0.62

0.40

0.65

0.53

0.51

0.52

6.3

1.6

2.5

1.2

3.9

5.2

2.5

11875

9380

59603

9284

21572

14680

23371

11.749

12.784

12.590

11.800

11.931

11.923

12.568

13.101

9.083

12.565

10.963

13.165

13.233

13.499

13.757

12.627

13.324

13.279

12.767

0.091

0.029

0.252

0.028

0.022

0.060

0.024

0.034

0.029

0.035

0.027

0.029

0.068

0.047

0.032

0.031

0.035

0.031

0.035

0.0012

0.0011

0.0012

0.0010

0.0010

0.0024

0.0016

0.0011

0.0011

0.0012

0.0011

0.0013

0.0022

0.4648

0.5004

0.4920

0.4595

0.0036

0.0010

0.0098

0.0010

0.46987 0.00088

0.4702

0.49107 0.00092

0.5134

0.3836

0.5005

0.4565

0.5143

0.5146

0.5291

0.5281

0.5011

0.5195

0.5160

0.5082

0.0016

0.0011

0.0013

0.0011

0.0017

0.0012

0.0018

0.0021

0.0016

1.00

0.96

1.00

0.97

1.00

0.98

0.99

0.93

0.86

0.93

0.96

0.96

0.98

0.95

0.96

0.95

0.96

0.97

0.97

0.98

0.98

0.94

0.93

0.94

0.93

0.83

0.96

0.93

rho

Pb6 Pb 2σ (abs)

Pb6 U

235

207

2158

2231

0.18332 0.00014 2584.6

0.18528 0.00011 2663.9

0.18560 0.00016 2649.5

0.18627 0.00012 2588.7

0.18417 0.000030 2599.0

0.18391 0.00019 2598.3

0.185619 0.000037 2647.8

0.18509 0.00018 2687.0

0.17173 0.00028 2346.4

0.18208 0.00019 2647.6

0.17416 0.00013 2520.0

0.18563 0.00012 2691.5

0.18650 0.00021 2696.4

0.18504 0.00021 2715.2

0.18893 0.00012 2733.1

0.18276 0.00014 2652.3

0.18602 0.00014 2702.9

0.18663 0.00011 2699.7

0.18221 0.00012 2662.6

0.15481 0.00012

0.19761 0.00012

0.19156 0.00018 2691.8

0.18911 0.00019 2576.1

0.19578 0.00029 2681.9

0.20434 0.00020 2795.6

0.19129 0.00044 2602.6

0.18726 0.00024 2603.2

0.20126 0.00025 2738.6

206

207

Pb6 U

2460.8

2615.7

2579.2

2437.2

2483.0

2484.4

2575.3

2671.0

2093.1

2616.0

2424.2

2675.1

2676.3

2737.7

2733.4

2618.6

2697.0

2682.4

2648.9

2284

2545

2607.7

2379.7

2539.1

2705.6

2413.5

2457.9

2607.8

238

206

Ages (Ma)

2683.2

2700.7

2703.6

2709.5

2690.7

2688.4

2703.7

2699.0

2574.6

2671.9

2598.0

2703.9

2711.5

2698.6

2732.9

2678.0

2707.3

2712.7

2673.1

2399.7

2806.5

2755.6

2734.5

2791.3

2861.2

2753.3

2718.2

2836.4

1.2

1.0

1.5

1.0

0.3

1.7

0.3

1.6

2.7

1.7

1.2

1.0

1.9

1.9

1.1

1.2

1.2

1.0

1.1

1.4

1.0

1.6

1.6

2.4

1.6

3.8

2.1

2.0

10

3.8

5.6

12

9.3

9.1

5.8

1.3

22

2.5

8.0

1.3

1.6

-1.8

0.0

2.7

0.5

1.4

1.1

12

24

6.5

15

11

6.7

15

12

9.8

Pb6 2σ Disc.

Pb (abs) (%)

206

207

P.I. Myhre et al.

Tonalite, Bakkejord pluton, Kvaløya (c04-35) UTM: (33N 6253xx 77156xx) 7

1

1

253-61 tip/shell frgm (1)

30

203/54 pr ca (1)

251/62 tip ca (1)

10

79

5

157/13 pink tips (7)

15

157/15a pink frg (1)

157/14 pink tip (1)

203/53 frg ca (1)

16

117/52 clr (1)

164

172

44

53

82/6 red frgm (1)

44

57

117/51 drk red frgm (1)

82/5 pink frgm (1)

Neosome in Kvalsund gneiss, Lille Kårvika on Ringvassøya (c01-110), UTM: (33N 6496xx 77560xx) 7

9

82/1

Mafic enclave in Kvalsund gneiss, Ringvassøya (c01-109), UTM: (33N 6496xx 77560xx)

1

1

157/16 pink tips (4)

251/54 pr ca (1)

8

157/17 pink tip (1)

83

238

2

1

82/3 pink tip (1)

61

8

157/18 clr prs tips (1)

82/4 clear tip (1)

Tonalite in sheared part of Dåfjord gneiss, Skarsfjorden area, Ringvassøya (c01-107), UTM: (33N 6476xx 77644xx) 7

mineral description

Sample

Table 1. U–Pb TIMS results.

8 NORWEGIAN JOURNAL OF GEOLOGY

183

238

112

5

34

105/56 tit med br (23)

197/s.17 tit br a (1)

197/s.21 tit cl na (9)

227

0.49

0.06

0.11

0.13

0.20

1.1

45.8

15.7

207.5

199.5

5861

343

2171

2843

3870

11.88

4.939

10.817

10.842

10.808

0.14

0.025

0.030

0.027

0.031

109

82

96

13

26

112/56 oblate frgm pink (3)

121/58 fr sbr (1)

203/5 brok pr ca (1)

103

1.23

1.21

1.03

1.27

1.20

1.34

1.08

29.4

2.0

2.0

8.3

34.5

4.0

20.5

12.642 0.029 12.199 0.027

30655 12.772 0.030

16848 12.767 0.029

40208 12.697 0.030

5494

12162 12.617 0.028

9387

10149 12.494 0.028

1

184/8 tip (1)

275

478

132

0.45 0.39

0.33

1.2 5.7

0.6 2255

5733

20

184s39 tit abr split no 2 4.10 462.7

4.10 463.7

132 197

197

12.39

23 33855.410774 0.36

184s81 U-rich mineral (20)

10 413

2666

0.15

11.101 0.032

6.704 7.890

7.875

5944

722

8611

0.064

27

1

157/58 clr eu incl (14)

157/59 (4) no U

184

179 0.53

0.66 1.5

14.4

5.2

18009

10605

15865

0.4571 0.0026

0.3726 0.0010

0.3724 0.0014

0.3412 0.0011

0.4586 0.0015

0.4933 0.0060

0.4421 0.0011

12.269

10.979

0.034

0.044

0.4909

0.4569

0.0012

0.0016

0.4390 0.0031 0.05025 0.00034

0.4401 0.0034 0.04972 0.00032

1.00

0.96

0.96

0.99

0.85

0.96

0.40

0.59

0.21

0.98

1.00

0.95

0.99

0.97

0.92

0.96

0.97

0.96

0.97

0.96

0.96

0.96

0.96

0.55

0.95

0.95

0.96

0.18492 0.00017 2594.6

2464.9

1776.6

2420.7

2428.2

2413.9

2697.5

1846.3

2578.6

2576.2

2582.8

2478

2672.7

2073 2219

2217

0.17992 0.00073

0.18125 0.00014 2625.2

0.17428 0.00022 2521.4

0.063363 0.000062 316.05

0.06420 0.00026 312.81

0.054376 0.000047 285.69

0.15357 0.00093

0.15336 0.00101

0.14250 0.0013

2574.8

2425.8

370

370

297

2042

2041

1893

2652.1

2664.3

2599.2

720.4

748.1

386.7

2386

2384

2258

0.17957 0.00012 2552.7 2433.4 2648.9

0.18215 0.00019 2634.4 2584.9 2672.6

0.18217 0.00019

0.18211 0.00017 2531.6 2360.0 2672.2

0.18103 0.00012 2557.2 2426.9 2662.3

0.16208 0.011

0.17727 0.00019 2469.9 2283.0 2627.4

0.17555 0.00021 2451.1 2262.7 2611.3

0.17904 0.00012 2546.6 2426.2 2644.0

0.18161 0.00011 2663.0 2656.9 2667.6

0.18164 0.00012 2662.6 2655.7 2667.9

0.18151 0.00012 2657.4 2645.2 2666.7

0.18001 0.00012 2619.8 2577.1 2653.0

0.18095 0.00012 2651.5 2638.2 2661.6

0.18081 0.00012 2653.4 2644.2 2660.4

0.18045 0.00011 2642.3 2623.1 2657.0

0.11288 0.00048 1808.9

0.17215 0.00015 2507.5

0.17190 0.00014 2509.7

0.17258 0.00015 2506.7

1.5

10

9.8

4.0

14

11

16

16

9.9

0.5

0.6

1.0

3.5

1.1

0.7

1.6

4.3

7.3

6.9

7.8

6.7

1.3

2.1

2.1

8.5

1.9

4.1

8.0

58

60

27

10.2 17

11.2 17

16.1 19

1.1

1.7

1.7

1.5

1.1

>100

1.8

2.0

1.1

1.0

1.1

1.1

1.1

1.1

1.1

1.0

7.6

1.4

1.3

1.5

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

Continued

16

157/57 clr eu (1)

0.4570 0.0010

0.5101 0.0010

0.5098 0.0010

0.5073 0.0010

0.3397 0.0011 0.04531 0.00013

8

0.052

0.064

Granodiorite on the eastern flank of Svanfjellet belt, Senja (pa04-2), UTM: (33N 595433 7695773) 8

184s63 U-rich mineral (15)

30595.929670 0.21

1

16 37694.866507 0.39

184s59 U-rich mineral (5)

0.5057 0.0010 0.4915 0.0010

0.4250 0.0018

15969 11.355 0.038

7636

15423

3412

0.5022 0.0010 0.5071 0.0010

10.387 0.046

10883 11.410 0.067

47

0.0011

0.0010

0.31730 0.00083

0.4557

0.4574

0.0053 0.0011

0.4205 0.0012

Felsic dyke in Kattfjord gneiss, Torsnes, Kvaløya (pim07-68), UTM: (33N 617360 7721381)

180

2.9 0.9

4.93 272.6

19

20

80

184/s42 tit abr brown (17 )

0.61

0.8

2.5

0.49

703

0.82 0.69

520

184s38 tit abr split no 1 (~50) 180

1

1

184/6 pr some clr incl (2)

184/11 pr (1)

1

184/3 pr l/w 5 no u (1)

456

302

1.2

1

1

184/5 pr l/w 3 (1)

0.8

5

184/2 pr l/w 4-5 (3)

184/1 clr prs no u (5)

0.4658 0.4542

10.178 0.031

20016 11.281 0.028

Granite in Kattfjord gneiss, Torsnes, Kvaløya (pim07-67), UTM: (33N 617239 7721220) 8

3

1

184/10 prs (4)

184/9 pr (1)

Neosome in Kattfjord gneiss, Torsnes, Kvaløya (pim07-69), UTM: (33N 617357, 7722498) 8

128

15

48

121/57 frgm sbr (9)

112/58 frgm pink (3)

88

113

84

69

112/57 oblate frgm pink (12)

112/59 frgm pink (14)

Mafic dyke intruding the Bakkejord pluton, Kvaløya (c04-31), UTM: (33N 633 3xx 77236xx) 7

22

258

1

105

203/2 tip cl eu (1) ca

105/55 tit drk br (7)

NORWEGIAN JOURNAL OF GEOLOGY

9

Wt.2 (ug)

U (ppm)

Th/U3 Pbc4 calc. (pg)

Pb5 Pb

204

206

12

277

1

658

441

0.30

0.47

0.09

0.23

0.45

1.10

0.46

0.24

0.66

0.17

0.28

1.5

1.4

1.9

1.5

3.6

0.3

0.6

11.5

3.2

1.3

0.7

26045

13474

5001

7082

8307

5100

5987

1522

45121

37885

44927

1

1

253/55 ca pr (1)

253/56 ca pr (1)

631

12

1001

708

684

139

218

1077

0.12

0.19

-0.42

0.12

0.13

0.14

0.13

0.17

2.4

1.3

1.2

1.0

1.1

7.0

5.4

0.7

12232

252

23833

18617

37401

10581

5812

23604

12.395

9.776

7.94

9.129

10.637

9.480

10.364

6.315

9.789

11.044

7.967

10.472

11.403

8.365

13.506

12.203

11.169

12.359

0.023

0.12

0.035

0.026

0.035

0.056

0.044

0.064

0.026

0.019

0.025

0.030

0.031

0.061

0.028

0.028

0.030

0.029

2σ (abs)

Pb6 U

0.0010

0.0013

0.0022

0.0010

0.0010

0.0011

0.0010

2σ (abs)

0.0027

0.0013

0.0021

0.0020

0.0027

0.0014

0.41442 0.00087

0.3690

0.3944

0.44819 0.00095

0.4090

0.4383

0.2937

0.4220

0.44758 0.00095

0.34943 0.00073

0.42530 0.00092

0.4615

0.3532

0.5018

0.4849

0.4481

0.4811

0.4778

238

206

0.96

0.69

0.98

0.96

0.96

0.96

0.98

0.99

0.95

0.94

0.94

0.96

0.94

0.89

0.96

0.97

0.97

0.96

rho

Pb6 Pb 2σ (abs)

Pb6 U

235

207

718.0

2225 0.17109 0.00012 2413.9

0.15614 0.00171

0.16789 0.00014 2351.1

0.17214 0.00012 2492.0

0.16812 0.00017 2385.6

0.17149 0.00026 2467.8

0.15594 0.00020 2020.5

0.16822 0.00017 2415.1

0.17896 0.00013 2526.9

0.16537 0.00014 2227.3

0.17859 0.00015 2477.5

0.17921 0.00014 2556.7

0.17178 0.00023 2271.4

0.19519 0.00040 2715.7

0.09090 0.00093

0.18250 0.00012 2620.1

0.18075 0.00012 2537.3

0.18632 0.00012 2632.1

0.18815 0.00012 2634.8

206

207

Pb6 U

2235.1

2025

2143.1

2387.2

2210.3

2343.1

1660.0

2269.7

2384.5

1931.9

2284.5

2446.0

1949.7

2621.7

508.2

2548.7

2387.0

2532.0

2517.6

238

206

2726.1

2568.3

2414

2536.7

2578.5

2539.0

2572.2

2412.1

2540.1

2643.3

2511.3

2639.8

2645.6

2575.1

2786.4

1445

2675.8

2659.8

2710.0

2

1

1.2

18.5

1.4

1.2

1.7

2.6

2.2

1.7

1.2

1.4

1.4

1.3

2.2

3.4

19.4

1.0

1.1

1.0

1.0

15

19

18

8.9

15

11

35

13

12

27

16

9.1

28

7.2

67

5.7

12

7.9

9.2

Pb6 2σ Disc.

Pb (abs) (%)

206

207

P.I. Myhre et al.

pr = prism, eu = euhedral, sbh = subhedral, clr = clear, incl. = inclusion, a =air abrasion, ca = chemical abrasion, (1)= no. of grains, frgt = fragment, tit: titanite, l/w = length/width ratio.

Weight is known to within 10%. Weight of chemically abraded grain is before partial dissolution. U content is calculated using this weight.

3 Th/U is modelled based on measured 208/206Pb-ratio and 207/206Pb-age.

4 Pbc = total common Pb in sample (initial + blank).

5 Corrected for fractionation.

6 Corrected for fractionation, blank, spike and initial Pb. 7 Approximate coordinates of sample location.

8 Sample location from GPS with a few metres uncertainty.

1

1

251/56 pr ca (1)

251/57 pr (1)

15

15

157/55 short pr (1)

157/54 tip (1)

157/56 clr pr (23)

2

1

157/53 clr tips (20)

Neosome, Grunnfarnes, southwestern Senja (c04-47), UTM: (33N 5758xx 76887xx)7

continued...

253/60 ca pr (1)

386

1

1

253/58 ca pr (1)

253/59 ca tip (1)

1045

1

253/57 ca pr (1)

77

1

157/60 resorped pr clr l/ w=2.5-3 (1)

108

1

899

157/61 resorped pr (1)

105/s.57 tit cl fr

188

2

25

113/52 eu-sb spr (4)

116/12 (1)

328

106/50 eu - sbh tips (4)

207

4

5

106/49 eu - sbh pr (5)

Pb6 U

235

207

Granodiorite, Grunnfarnes, southwestern Senja (c04-46), UTM: (33N 5759xx 76887xx)7

Sample mineral description1

Table 1. Continued

10 NORWEGIAN JOURNAL OF GEOLOGY

NORWEGIAN JOURNAL OF GEOLOGY

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

11

Tonalite in the sheared part of the Dåfjord gneiss, Skarsfjord area, Ringvassøya (c01-107)

Aa. Pb/ 238U

b.

B

3000

Tonalite crystallization likely between: 2.85 - 2.77 Ga

2800

2400

2600

15a

Discordant data: SIMS (12r, 6a, 15a, 7a, 7b, 16): u. i. = 2801 ± 40 Ma

6a

8

TIMS: (157/17, 82/4, 157/16, 157/18, 251/54, max.): u. i. = 2933 ± 90 Ma l. i = 1376 ± 380 Ma MSWD = 121

251/54

Pb/ 235U

12

MSWD = 0,083 0.083

2600

4r

8r

10c

0.54

2700

1c 5c

2.2

2400

Ga

Age component # 2: ≥ 2560 Ma (SIMS)

u.i. = 2709 ± 1 Ma l.i. = 740 ± 20 Ma

2700

MSWD = 0.015 0,015

5r

10r

8r

207

e.

E

117/52 251/62 tip

4a

Ga 2550

2 2.

Ga

8 1.

a G

12

f.F

c01-107 10.

6.

7c.

8.

TIMS (157/15, concordant) 207/206 Pb = 2733 ± 1 Ma

Sketch of tip and shell overgrowths.

Analysed by (CA)-TIMS (#157/13, #253/62, #251/62)

Picking-scope photo of conchoidally fractured z. analysed by TIMS (#157/15: 2733 ± 1 Ma, Pb-Pb-age).

12r.

5c.

1. (cl) 11.

14c.

14

c01-110

9.

7r.

Age component #1: SIMS (10 c, 1, 5c)

13

12

14

[117/51] [82/6] 157/13 tip 5c [203/53]

207/206 Pb = 2737 ± 8 Ma, 1.17 MSWD = 1,17

Zircon, amphibolite lens in neosome

10

1

253/62 tip [82/5]

4b

2600

0.50

0.42

253/61

[157/15]

203/53, 253/61):

Ga 1.8

75 0,

10c 157/14 tip

Likely arbitrary discordia line: TIMS (117/51,

4r2

0.52 203/54

d.

D

5r.

8r.

4a.

12c.

15.

cl 10r.

16.

e

17.

14r.

bs

4b.

cl

bs e

Age component # 1: 157/15, 117/51, 82/5: 2203 ± 97 & 2731 ± 8 Ma

157/14

5G a

Inset

0,7

Cc.

2500

0.38

18

16

14

Neosome in Kvalsund gneiss, Ringvassøya (c01-110)

5r

0.46

82/4 157/16

157/18

16

12

Inheritance/xenocrysts: SIMS (17, 8, 14c): Pb/Pb = 2922 ± 7 Ma MSWD = 2.3

82/3 0.45

207

251/55 7c

l. i. = 517 ± 140 Ma MSWD= 7.7

12r

0.50

17

157/17

2200

0.54

8

16

7r

0.45

0.35

12c

9, 1

2800

0.55

7c

3000

14c r

4 1, 1 0, 1

TIMS discordia: (251/54, 251/55, min.) u. i. = 2768 ± 3 Ma l. i. = 415 ± 40 Ma

2600

206

0.55

Inset

10c.

Figure 5. (A–F) U–Pb concordia diagrams and zircon images from the study area in Ringvassøya. Error ellipses and quoted errors are 2σ. Red MyhreTIMS et al.,analyses Figure 5of chemi­cally-abraded (CA) zircon (this convention used throughout). Scale bars are 100 µm and zircon­images ellipses indicate are back-­scattered electron through­out the paper, unless marked otherwise (i.e., CL or binocular microscope).

discordia lines reflect this component. Another component could be represented by a slightly younger core (2892 Ma, Fig. 5B, analysis 12c), or it could be the result of mixing or Pb loss from a 2.92 Ga core. The age of the youngest zircon group is represented by elongate, prismatic zircons and overgrowths, and a minimum age for this group is provided by the youngest TIMS upper intercept (Fig. 5B, analyses 251/54 and 251/55) of 2768 Ma and the somewhat discordant SIMS analyses (analyses

7r, 16) with 207/206Pb ages between 2.75 and 2.80 Ga. Other SIMS analyses range from around 2.85 Ga to 2.79 Ga. The remainder of the SIMS rims and elongate prisms give results that are discordant with a general trend towards 2.80 Ga. From the data, it appears that the sample contains at least two generations of zircon formed at c. 2.92 and 2.85–2.77 Ga with a superimposed component, presumably Pb loss,

U (ppm)

Th/U meas.

Pb Pb

204

206

f206%2

Pb Pb

235

207

σ (%)

Pb Pb

238

206

σ (%)

rho

Pb Pb

206

207

σ (%)

582.9

147.2

6a

8

0.60

0.91

0.47

0.54

79897

11579

50605

46510

242656

137239

14152

883

25069

49439

2379

5354

47888

83897

0.02

0.16

0.04

0.04

0.01

0.01

0.13

2.12

0.07

0.04

0.79

0.35

0.04

0.02

16.82

7.95

14.08

12.88

15.76

15.31

15.70

6.50

16.16

15.57

10.01

17.12

14.46

16.67

0.91

0.95

0.99

1.11

1.06

1.17

1.28

1.12

1.13

1.08

1.26

1.03

1.20

1.08

0.5784

0.3239

0.5232

0.4786

0.5637

0.5574

0.5609

0.2675

0.5628

0.5567

0.3852

0.5807

0.5352

0.5688

0.84

0.91

0.84

1.05

1.00

1.12

0.84

0.84

1.00

1.01

0.85

0.85

1.08

1.03

93.8

8

0.45

1.01

0.85

0.27

78498

24013

27676

24430

54047

32695

29033

11575

0.02

0.08

0.07

0.08

0.03

0.06

0.06

0.16

13.91

12.27

12.87

11.63

13.67

13.50

12.86

11.72

1.00

1.08

0.96

1.00

1.12

1.04

1.04

0.97

0.5341

0.5042

0.5094

0.4932

0.5240

0.5139

0.5157

0.4529

0.94

0.92

0.82

0.84

1.02

0.93

0.91

0.88

643.8

411.3

1

396.1

4

474.1

419.1

3a

1137.4

8b

5

3b

271.2

8a

0.44

0.46

0.34

0.37

0.40

0.45

0.12

50160

72778

149022

154789

68834

52273

7424

0.04

0.03

0.01

0.01

0.03

0.04

0.25

12.19

12.43

12.56

11.74

11.73

4.99

13.69

0.90

0.94

0.95

0.88

1.05

0.98

1.00

0.4884

0.4987

0.4998

0.4731

0.4764

0.2384

0.5256

0.86

0.89

0.92

0.83

0.94

0.92

0.91

3

0.95

0.95

0.97

0.94

0.90

0.94

0.92

0.85

0.86

0.84

0.91

0.89

0.88

0.90

0.95

0.92

0.96

0.86

0.94

0.95

0.96

0.66

0.75

0.89

0.94

0.68

0.83

0.90

0.95

0.1810

0.1808

0.1823

0.1800

0.1785

0.1518

0.1889

0.1765

0.1833

0.1710

0.1892

0.1905

0.1809

0.1876

0.1889

0.2110

0.1779

0.1951

0.1952

0.2027

0.1992

0.2030

0.1762

0.2082

0.2029

0.1884

0.2138

0.1960

0.2126

0.28

0.29

0.23

0.30

0.46

0.34

0.40

0.56

0.50

0.54

0.47

0.47

0.50

0.41

0.32

0.36

0.26

0.51

0.37

0.33

0.31

0.96

0.74

0.52

0.38

0.93

0.57

0.51

0.34

2619

2638

2647

2584

2583

1818

2728

2625

2670

2575

2727

2715

2670

2582

2744

2925

2225

2755

2671

2862

2835

2859

2046

2886

2851

2435

2942

2780

2916

9

9

9

9

8

10

8

9

10

9

9

11

10

10

9

9

9

9

11

10

11

12

10

11

10

12

10

11

10

σ (abs)

Pb U

2564

2608

2613

2497

2512

1378

2723

2632

2654

2584

2716

2673

2681

2408

2759

2942

1809

2713

2521

2882

2856

2871

1528

2878

2853

2101

2952

2763

2903

238

206

21

18

19

20

17

20

11

20

20

18

18

23

20

20

18

20

14

19

22

23

26

20

11

23

23

15

20

24

24

σ (abs)

Pb U

2662

2660

2674

2653

2639

2367

2732

2620

2683

2568

2735

2747

2661

2721

2732

2913

2634

2786

2786

2848

2820

2851

2617

2892

2849

2728

2935

2793

2925

206

207

5

5

4

5

8

6

7

9

8

9

8

8

8

7

5

6

4

8

6

5

5

16

12

8

6

15

9

8

6

σ (abs)

4

2

3

7

6

46

0

-1

1

-1

1

3

-1

14

-1

-1

36

3

11

-1

-2

-1

47

1

0

27

-1

1

1

Disc. (%)

P.I. Myhre et al.

Neosome in Kattfjord gneiss, Torsnes, (pim07-69), n3100, UTM: (33N 617357 7722498) 7

119.9

151.6

4a

4b

250.0

1

0.47

0.50

131.3

189.8

5r

5c

0.52

0.56

366.8

191.8

10c

10r

Neosome in Kvalsund gneiss, Ringvassøya (c01-110), n3099 UTM: (33N 6496xx 77560xx)

159.4

7b

0.60

207.4

211.0

9

7a

0.47

298.9

10

0.16

0.23

382.1

165.7

0.53

11

78.8

12c

0.21

0.39

12r

143.6

182.7

14r

78.4

15a

0.44

112.2

16

14c

0.45

0.40

187.1

17

Pb U

235

207

Ages (Ma)

Tonalite in sheared part of Dåfjord gneiss, Skarsfjorden area, Ringvassøya (c01-107), n3104, UTM: (33N 6476xx 77644xx) 3

spot #

1

Sample

Table 2. U–Pb SIMS results.

12 NORWEGIAN JOURNAL OF GEOLOGY

0.92

0.45

182.0

1c

1.03

0.49

1.20

0.81

140888

87980

65719

53623

22240

3805

105046

307042

65237

129616

0.01

0.02

0.03

0.03

0.08

0.49

0.02

0.01

0.03

0.01

13.24

12.89

12.49

13.10

9.90

12.46

12.27

13.36

13.29

12.60

1.03

1.01

0.97

0.93

2.99

0.95

0.97

1.02

1.13

1.12

0.5215

0.5101

0.4991

0.5131

0.4023

0.4966

0.4899

0.5275

0.5207

0.4991

0.97

0.92

0.86

0.83

2.92

0.88

0.93

0.99

1.05

1.05

0.94

0.92

0.89

0.89

0.98

0.92

0.95

0.97

0.93

0.94

0.1842

174.2

174.4

4

2

0.32

0.39

0.44

0.51

0.44

135573

106554

43432

95411

5881

132288

{0.00}

0.01

0.02

0.04

0.02

0.32

0.01

{0.00}

12.94

13.37

13.07

13.51

10.74

6.12

12.96

13.39

1.16

1.06

1.03

1.13

1.07

1.13

1.25

1.02

0.5140

0.5274

0.5135

0.5272

0.4475

0.3099

0.5112

0.5264

0.89

0.98

0.96

0.95

1.00

1.04

1.16

0.92

63.7

1414.9

132.4

255.1

9c

9r

7

4c

1.21

0.75

0.02

0.49

0.02

0.70

0.02

0.63

0.44

0.11

1.95

0.14

29738

117849

5609

45764

13999

67325

70389

125451

315655

50278

82314

153938

0.06

0.02

0.33

0.04

0.13

0.03

0.03

0.01

{0.01}

0.04

0.02

0.01

13.33

13.35

6.90

14.34

3.25

11.92

8.45

13.34

15.31

7.12

14.00

10.87

1.14

1.12

1.11

0.99

0.88

0.92

1.22

0.87

0.95

0.96

0.96

0.96

0.5193

0.5159

0.3229

0.5251

0.1727

0.4700

0.3710

0.5090

0.5578

0.3231

0.5258

0.4557

1.09

0.79

1.06

0.80

0.83

0.83

1.20

0.83

0.89

0.90

0.83

0.92

0.95

0.70

0.96

0.81

0.95

0.90

0.98

0.95

0.93

0.94

0.86

0.96 0.1932

0.1862

0.1877

0.1550

0.1980

0.1363

0.1839

0.1652

0.1901

0.1991

0.1597

0.35

0.34

0.80

0.33

0.58

0.28

0.40

0.22

0.27

0.35

0.33

0.49

0.27

0.75

0.39

0.38

0.62

0.41

0.43

0.47

0.44

0.40

0.43

0.42

0.65

0.37

0.30

0.24

0.42

0.39

2697

2703

2705

2099

2772

1468

2598

2280

2704

2835

2126

2750

2512

2675

2706

2685

2716

2501

1993

2676

2708

2672

2642

2687

2425

2639

2625

2705

2700

2650

10

11

11

10

9

7

9

11

8

9

9

9

9

11

10

10

11

10

10

12

10

10

9

9

28

9

9

10

11

11

2706

2696

2682

1804

2721

1027

2484

2034

2653

2858

1805

2724

2421

2674

2731

2672

2730

2384

1740

2662

2726

2657

2610

2670

2179

2599

2570

2731

2702

2610

21

24

17

17

18

8

17

21

18

20

14

18

19

20

22

21

21

20

16

25

21

20

19

18

54

19

20

22

23

23

2691

2709

2722

2402

2810

2181

2688

2509

2743

2819

2453

2769

2588

2676

2688

2694

2706

2597

2268

2688

2694

2683

2667

2700

2638

2671

2668

2686

2699

2681

6

6

13

6

9

5

7

4

4

6

6

8

5

12

7

6

10

7

7

8

7

7

7

7

11

6

5

4

7

6

1

2

28

4

57

9

22

4

-2

30

2

8

0

-2

1

-1

10

26

1

-1

1

3

1

20

3

4

-2

0

3

-1

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

Continued

283.5

1835.2

10c

1123.4

14r

10r

174.7

344.6

15c

14c

227.9

425.1

16c

16r

385.8

4r

0.1731

0.1826

0.77 3

0.1839

0.1846

0.1858

0.1740

0.1433

0.1838

0.1845

0.93

0.93

0.84

0.93

0.92

0.93

0.90

Granodiorite, Grunnfarnes, southwestern Senja (c04-46), n3105, UTM: (33N 5759xx 76887xx)

70.2

188.1

10

5

206.9

13r

0.30

0.53

114.4

869.0

14r

0.70

163.9

13c

14c

4

0.1833

0.1815

0.1852

0.1784

0.1819

0.1817

0.1837

0.1851

0.1831

Granodiorite on the eastern flank of Svanfjellet belt, Senja (pa04-2), n3103, UTM: (33N 595433 7695773)

165.5

174.3

2c

2r

258.8

141.4

3r

3c

0.46

357.1

7r

0.95

1.43

173.7

550.8

8c

0.89

7c

256.6

206.3

10c

9

Granite in Kattfjord gneiss, Torsnes, Kvaløya (pim07-67), n3101, UTM: (33N 617 239 7721220) 4

NORWEGIAN JOURNAL OF GEOLOGY

13

Internal Nordsim sample designation is on the form n31xx. r - rim, c - core.

Common Pb as % of 206Pb, determined using 204Pb counts. Numbers in parentheses indicate that 204Pb counts were within 3 standard deviations of the average background count and in these cases, the correction has not been applied.

3 Approximate coordinates of sample location.

4 Sample location from GPS with a few metres uncertainty.

2

1

14 4 2641 18 2341 9 0.26 625.0 6c

0.28

146520

0.01

10.79

0.97

0.4378

0.93

0.96

0.1788

2505

2

7 5

4 2688

2578 17

24 2636

2433 8

11 2666

2513

0.22

0.31

0.1839

0.1721 0.93

0.98 1.09

0.83 0.4585

0.5052 1.12

0.89 10.88

12.81 0.01

0.01 224701

143876 1.09

0.09

577.1

362.1

4c

0.30 0.95 0.95 0.4677 0.99 11.16 0.02 110050 0.11 434.8 4r

3

6

5 5 2587 20 2473 9

13

0.1731

2537

5

12 2367

2546 18

16 2101

2425 9

10 2238

2491

0.70

0.32

0.1518

0.1688

0.78

0.94 0.88

0.87 0.3852

0.4566 0.94

1.12 8.06

10.63 0.01

0.07 25168 257.8

394.3

7

6r

0.15

305964

4.00 0.97 14.98 0.2470 15.50 5.30 0.04 50159 1.20 1418.1 10

0.10

45 66 2408 194 1423 142

11

0.1556

1868

6

7 2667

2549 20

19 2422

2297 10

10 2557 0.40

0.37

0.1815 0.92 0.93

1.02 0.4280

0.4559 1.01

1.09 9.98

11.41 0.09

0.51 3703

19868 577.4

671.1

12c

0.24

0.94

0.1691

2433

NORWEGIAN JOURNAL OF GEOLOGY

12r

Neosome, Grunnfarnes, southwestern Senja (c04-47), n3102, UTM: (33N 5758xx 76887xx) 3

0.09

σ (abs) Pb U 206

207

σ (abs) Pb U 238

206

σ (abs) Pb U 235

207

σ (%) Pb Pb 206

207

rho σ (%) Pb Pb 238

206

σ (%) Pb Pb 235

207

f206%2 Pb Pb 204

206

Th/U meas.

Sample U spot #1 (ppm)

Table 2. Continued

12

P.I. Myhre et al.

Disc. (%)

14

responsible for the more discordant data points. Due to these complexities, assignment of ages to inheritance or magmatic or metamorphic zircon growth is necessarily somewhat ambiguous and subject to interpretation. One interpretation is that the c. 2.92 Ga component represents igneous zircon in the tonalitic protolith, with one or more episodes of metamorphic overgrowth between c. 2.85– 2.77 Ga. The intermediate ages would represent sampling mixtures or several discrete metamorphic events. An alternative interpretation is that the old ages from prisms and cores represent inherited grains, with the younger prisms and tips being products of tonalite crystallisation.

In this case the youngest of the TIMS discordia line and the two youngest near-concordant SIMS analyses would place a minimum age constraint for the age of the tonalite at c. 2.77 Ga. It is also possible that this 2.77 Ga component is related to superimposed metamorphic growth or slight Pb loss from 2.85–2.80 Ga grains, in which case the igneous age would be somewhat older. The morphology and internal zoning of, e.g., grains 9, 10, 11, 16 (Fig. 5e) would tend to support an igneous origin for these 2.85 Ga and younger grains, so we consider the 2.85–2.77 Ga interval to likely encompass the crystallisation time of the tonalite, acknowledging the uncertainties involved for this interpretation.

Neosome in Kvalsund gneiss, Ringvassøya (c01–110) The sample is from a neosome sheet within stromatic migmatite (sample locality in Fig. 2A, neosome sheet in Fig. 2E). In thin section, the sample is granitic with minor biotite and hornblende occurring in clusters. The zircons in this sample mainly belong to three morphotypes: (a) large (100–300 µm), fragmented pink grains (Fig. 5F, picking scope photo), (b) smaller (50–100 µm) prismatic crystals commonly made up of cores with prismatic terminations (Fig. 5F, grains 5, 8), and (c) fragmented tips and shells with concave shapes where they have been attached to a core (Fig. 5F, sketch). The BSE images (Fig. 5F) show a thinly spaced magmatic zoning in some of the cores of prismatic zircon truncated by a tip overgrowth where zoning is not particularly well developed (Fig. 5F, e.g., grains 5, 8). The larger, fragmented zircons display some weak zoning reminiscent of sector zoning (Fig. 5F, e.g., grain 4), or alternating bright and dark lamellae in the case of one large fragment (Fig. 5F, grain 1). Some of the fragmented zircons are also overgrown by Cl-bright rims (Fig. 5F, grain 10). For the TIMS analyses, large fragmented zircons and some tips were chosen, trying to avoid any mixed domains. One of the fragments yielded a concordant age of 2733 ± 1 Ma, with the other analyses of fragments plotting on various discordia lines trending towards this age or a slightly younger age of c. 2710– 2700 Ma (Fig. 5C, D; Table 1). The tips plot on similar trajectories, although somewhat more discordantly.

Three SIMS analyses of two fragments and the core of a prismatic zircon (Table 2) agree with the TIMS age of the concordant fragment, whereas four concordant rim analyses spread out along a c. 130 Myr segment of the

NORWEGIAN JOURNAL OF GEOLOGY

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

concordia line between 2690 and 2560 Ma. From these data, it seems that the cores and the large fragmented zircons have roughly the same age of 2.73 Ga. The position on the concordia diagram of the SIMS data on rims and tips cannot be explained only by Pb loss from 2.73 Ga grains, and they must therefore have grown later, possibly in several distinct events. The tips that were analysed

by TIMS do not record these younger ages very clearly, plotting in an intermediate position between the core and rim ages, but far below a hypothetical mixing line between 2.73 and 2.56 Ga (not shown in Fig. 5).

in the neosome. An alternative interpretation would be that the majority of the zircons are not xenocrystic to the neosome, but crystallised from the melt at 2.73 Ga and went through an overprinting metamorphic event before and/or at 2.56 Ga, the spread in rim ages being due to either analytical mixes or several distinct growth events. A possible third event could be represented by the 2.71–2.70 Ga TIMS discordia lines (Fig. 5C) and 207/206Pb ages of the least discordant TIMS tips, but discordance from a 2.73 Ga component is perhaps a more appropriate explanation for the 2.71–2.70 Ga ages. In summary, the two former interpretations seem equally valid, with neosome formation either at 2.73 Ga or at ≥2.56 Ga. A possible third zircon-forming event at 2.71–2.70 Ga cannot be ruled out but is not a preferred interpretation.

The data record a multistage evolution of this rock, and assignment of the different ages to specific events is a matter of interpretation. One possibility is that the fragmented zircon (Fig. 5F, pink grain) and the cores of prisms are xenocrystic and derived from a precursor rock to the neosome, perhaps akin to the palaeosome enclaves within the migmatite seen in Fig. 2B, C. In this case, the rims could be crystallisation products of the neosome itself at 2.56 Ga and/or earlier events. A mafic enclave (c01–109) from the same locality was collected but very little zircon was recovered (two analyses plotted as crosses in Fig. 5C), and it does not seem that this type of rock could be the source of the large numbers and sizes of the zircons found

0,54

a.

A Tonalite, Bakkejord pluton, Kvaløya (c04-35)

Tonalite, Bakkejord pluton, Kvaløya (c04–35) A sample of tonalite cut by several mafic dykes collected in the southern part of the Bakkejord pluton (Fig. 1) contains about equal amounts of brown biotite and hornblende, the latter with substantial alteration in the form of symplectite. Some alteration in the form of saussuritisation of plagioclase and epidotisation is also evident. In thin section, titanite can be observed in two

0,53

b.

B 2720

Crystallization age (zircon): Combined age from two discordia lines: 2711 ± 13 Ma

2700 Inset 0,50

0,51

2500

15

2680

Line 1: 2699 ± 4 Ma, MSWD = 2.9 Line 2: 2724 ± 7 Ma, MSWD = 4.7

2640 121/55

0,46

117/60

±

0,49

2560

102/2

27

0,42

62

2300

49

M

a

2600

8

10

12

M

2G a

74

0.1

±

4±

Brown titanite

6

0.514

2440

0G a

Pale titanite

7

1900

2480

0,47

0.6

Secondary growth: 1717 ± 74 Ma Igneous component: 2762 ± 49 Ma MSWD = 30

a

Titanite:

17 1

0,34

2520

2100

0,38

102/3 102/4 203/2 121/56 117/59

14

Cc. Mafic dyke intruding the Bakkejord pluton, Kvaløya (c04-31)

2670 2660 2650 2640

0.506

112/57

2620

85 0.

121/56

121/57

2630

0.498

112/59

121/58 203/5

±

06 0.

a G

Crystallization age: Upper icpt, 4 points: 2671 ± 1 Ma, MSWD = 1.3

112/58 0.490 12.4

12.8

Figure 6. (A–C) U–Pb concordia diagrams for the tonalitic Bakkejord pluton, and mafic dykes cutting this pluton. Error ellipses and quoted errors are 2σ.

Myhre et al., Figure 6

16

P.I. Myhre et al.

NORWEGIAN JOURNAL OF GEOLOGY

textural settings; either as subhedral grains, or in close association with allanite-cored epidote and feldspar in a kind of reaction texture. Two types of titanite are also present in heavy-mineral separates, as brown and pale grains. The brown titanite has a U content of 180–250 ppm in contrast to the pale ones with c. 20 ppm U (Table 1). The two types also yield different ages; brown titanite plots 7–8% discordantly (Fig. 6B) with uniform 207/206Pb ages of around 2580–2570 Ma, whereas pale titanite gives 207/206 Pb ages of c. 1850 Ma (6% discordant, Fig. 6A). A discordia line fitted through these analyses give intercept ages of 2762 ± 49 and 1717 ± 74 Ma that approximate the two titanite age components (any translation due to Pb loss during the Caledonian orogeny results in an increase of the upper intercept and a decrease of the lower intercept ages). The majority of zircons in this sample are subhedral, elongate grains that appear to be of magmatic origin, and seven fractions of prisms and one tip fraction were analysed by TIMS. A minor group of zircons with apparent cores and fragmented metamict grains were not analysed. The analysed grains are from 4 to 10% discordant and define a discordia trajectory with an upper intercept between 2700 and 2725 Ma, however in detail with some scatter (Fig. 6B). This agrees with the oldest component represented by brown titanite, and is interpreted as the crystallisation age of the tonalite.

Due to the data scatter, we prefer to estimate the age by combining the maximum and minimum-solution discordia lines shown in Fig. 6B, which gives an age of 2711 ± 13 Ma for the crystallisation of the tonalite.

Neosome in Kattfjord gneiss, Torsnes, Kvaløya (pim07–69) This sample represents a c. 50 cm-thick neosome vein, which is part of the chaotically structured and banded migmatite illustrated in Fig. 3B. Petrographically, the neosome sample is a granite or granodiorite, with c. 10% biotite. Some epidotisation is evident in this sample. In heavy-mineral separates, zircon occurs as subhedral grains with (Fig. 7A, grains 5, 8) or without (Fig. 7A, grain 3) cores of contrasting morphology, and small broken off tips. The cores have a pronounced zoning pattern seen in CL and BSE, whereas the single-domain zircon grains and tips display a somewhat different, weak zoning, reflecting the presence of two distinct generations of zircon in this sample. Few core-bearing zircons were suitable for dating due to cracks or metamict domains, but one SIMS analysis of a core yields an age of 2732 ± 13 Ma (Fig. 7B, analysis 8c), with a massively discordant rim (outside of the bounds of the plot in Fig. 7B; see Table 2 for the data). Analyses of tips and subhedral grains give distinctly younger, discordant ages. An upper intercept age of 2683 ± 17 Ma can be calculated based on SIMS analyses of six tips and subhedral prisms, agreeing with the calculated, three-point, TIMS upper-intercept age of 2679 ± 200 Ma, but with poor precision due to the scatter of the data (MSWD = 63). We interpret 2683 ± 17 Ma as the time of crystallisation of the neosome, incorporating xenocrystic material similar in age to the nearby Bakkejord pluton (2732 ± 13 Ma).

Mafic dyke intruding the Bakkejord pluton, Kvaløya (c04–31)

The sample was taken from the granitic gneisses on the eastern flank of the Torsnes belt (Fig. 3A), and is classified as a biotite granite. The zircons occur as euhedral, elongate, single-domain grains (e.g., Fig. 7C, grain 9), and shorter, stubby grains which commonly contain cores (Fig. 7C, grains 7, 8). The SIMS data show that the cores (pink ellipses) and rims and one prism (grey ellipses) are quite close in age. The core-bearing zircons are concordant with ages between 2720 and 2680 Ma, whereas three analysed rims and one of the prisms plot slightly discordantly with 207/206Pb ages of around 2670 Ma. The TIMS data on elongate prisms and some tips give similar results to these younger SIMS dates. Due to the similar geology of the granite and the similar zircon ages, the data from sample KV02 published in Corfu et al. (2003) are also plotted in Fig. 7D. The TIMS data from the current study and the rims and single-domain prism SIMS data are equivalent to the age reported by Corfu et al. (2003) of 2689 ± 6 Ma (based on four analyses, labelled 1–4 in Fig. 7), and we prefer to assign this age as the crystallisation age for pim07–67 as well. In detail, it is possible that this age could be somewhat lower and align more with the upper intercept of one of the discordia lines in Fig. 7D, at 2684 ± 10 Ma. Regarding the zircon cores (pink SIMS ellipses), we consider them xenocrystic to the granite, analogous with inheritance in the neosome presented above (pim07–69), and with similar or slightly younger ages relative to the tonalites near the Bakkejord pluton.

This sample comes from the centre of a coarse-grained part of a c. 20 m-wide, amphibolitic mafic dyke, located in the Bakkejord pluton a few hundred metres west of the contact with the Mjelde–Skorelvvatn belt (Fig. 1).

Rich in felsic minerals, it represents a rather evolved part of the dyke. The green amphibole has inclusions of quartz and feldspar with shape outlines following the amphibole cleavage. Medium-grained, brown biotite is common as a secondary phase around the amphiboles.

Zircon occurs as 100–300 µm large prisms or fragmented prisms. They have a somewhat brownish coloration due to metamictisation. The zircon morphology suggests they are primary magmatic grains, and the evolved character of the sampled dyke is consistent with zircon crystallising as a primary phase. The TIMS data (Fig. 6C) give a crystallisation age of 2671 ± 1 Ma, calculated using a four-point discordia line with a MSWD of 1.3.

Since two of the analyses on this line plot fairly close to the concordia line, the exclusion of three analyses from the age calculation can be justified and attributed to ancient Pb loss or possibly some new growth, e.g., during Svecofennian thermal events.

Granite in Kattfjord gneiss, Torsnes, Kvaløya (pim07–67)

NORWEGIAN JOURNAL OF GEOLOGY

Figure 7. (A–E) U–Pb concordia diagrams and zircon images for samples from the Torsnes area on Kvaløya. Error ellipses and quoted errors are 2σ.

A a.

b.

B

0.55

pim07-69

8r. (cl)

5.

Inheritance/xenocryst: SIMS (8c): Pb/Pb = 2732 ± 13 Ma

Neosome in Kattfjord gneiss, Torsnes, Kvaløya (pim07-69)

8c. (cl)

2600

0.50

2500

3b

SIMS (1, 3a, b, 4 & 5): u.i = 2683 ± 17 Ma

184/10

0.45

l.i. = 400 - 1000 Ma MSWD = 1.7

4.

TIMS (184/9, 8, 10): u.i = 2679 ± 200 Ma l.i. = not defined

184/8 184/9 0.40

3r.

2c.

12

1c.

3c.

5 z. (TIMS; 1, 2, 4, 184/6 & 184/11) Upper intercept = 2689 ± 3 Ma 2640 MSWD = 2.3 2,3

0,51

0,49

10c.

9.

1c.

2720 2c.

1.

Pink: cores Grey: rims or prism

4.

3.

TIMS analyses 1. -4. are from KV02, reported in Corfu et al.

(2003).

2520

pim07-67

8c.

10c.

2.

184/6

2560

7c.

2680

2600

7r.

2r.

8.

14

Granite in Kattfjord gneiss, Torsnes, Kvaløya (pim07-67 & KV02)

0,53

7c.

MSWD = 63

10

d.

D

pim07-67

3a

Neosome crystallization:

4

3a.

8c

2700

1

5

1. (cl)

3b.

c.

C

17

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

0,47

184/11

0,45

To c.

700 Ma

184/5

184/2

3 z. (TIMS; 1, 3 & 184/2): Upper intercept = 2684 ± 10 Ma MSWD = 3.0 3,0 207

To 500-100 Ma 11,6

12,6

Pb/ 238U

Ee.

206

0.050

320

0.046

Felsic dyke in Kattfjord gneiss, Torsnes, Kvaløya (pim07-68)

U-rich (3-4 %) mineral l.i. = 280 ± 30 u.i = 2300 1600 MSWD = 16

310

0.048

Pb/ 235U

13,6

300

290

0.33

0.37

0.41

0.45

Myhre et al., Figure 7

Mafic and felsic dykes, Kattfjord gneiss, Torsnes, Kvaløya Several samples of the dykes within the Kattfjord gneiss, which are shown in Fig. 3A, D, E, were sampled for U– Pb geochronology. Thin-section investigation of four samples of mafic dykes by electron microprobe at the University of Alberta revealed no euhedral prismatic zircon or baddeleyite of igneous origin in these samples.

Instead, small, <15 µm round zircons were found, primarily at triple junctions or within recrystallised quartz or hornblende, suggesting a metamorphic origin of the zircons. Separation of these zircons was attempted using a combination of Wilfley table and magnetic separation and

an experimental approach involving high-T annealing and dissolution of quartz and feldspar, but very little zircon was recovered and the results are inconclusive and pending further investigation.

A sample of folded felsic dyke (pim07–68, Fig. 3A) did not yield any zircon; instead a brown, U-rich (3–4% U) heavy mineral mistaken for titanite gives discordant, Phanerozoic ages with one grain plotting close to the lower intercept of the three-point discordia at c. 300 Ma (Fig. 7E). These data are also inconclusive but may reflect deposition of U-rich minerals during Phanerozoic faulting. A similar situation was reported by Corfu et al. (2003).

18

P.I. Myhre et al.

NORWEGIAN JOURNAL OF GEOLOGY

Granodiorite on the eastern flank of Svanfjellet belt, Senja (pa04–2) One sample (pa04–2, see Fig. 1) of granodiorite has been dated in order to place some constraints on the proportions of Archaean vs. Palaeoproterozoic crust on Senja. This sample is a medium-grained, amphibolebearing granodiorite with a moderately SW-dipping foliation. Zircons are mostly elongate, subhedral prisms with somewhat rounded crystal faces and flat terminations (rather than developed tips). Such zircons have welldeveloped oscillatory zoning clearly of magmatic origin (e.g., Fig. 8A, grains 2, 4). Two TIMS analyses, four SIMS analyses of elongate prisms and two SIMS analyses of a round grain indicate a uniform origin of the zircon population in this sample (Fig. 8B), and the concordia

age of the SIMS analyses of 2692 ± 6 Ma is interpreted as the age of crystallisation of the granodiorite. One partly metamict grain with core was analysed (Fig. 8A, B, analysis 13c) by SIMS to evaluate the age of potential inheritance, but the result is discordant and could represent either a magmatic or an inherited component.

Granodiorite, Grunnfarnes, southwestern Senja (c04–46) This sample represents pre-migmatitic diorite-granodiorite rafts within stromatic migmatite at Grunnfarnes in southwestern Senja (Fig. 4A, C). Migmatisation of these rafts is manifested as mm- to dm-thick layers of neosome that are parallel to and largely define the gneissic foliation.

The sample displays a minimum amount of neosome. It is a

Granodiorite on the eastern flank of Svanfjellet belt, Senja (pa04-2)

Bb.

a.

A

pa04-2

2.

0.55

14c, 14r, 10, 5, 4, 2

Grandiorite crystallization: Concordia age (6 SIMS analyses): 2692 ± 6 Ma MSWD = 0.77

2800

2600 157/58 13r.

2400

0.45

13c.

157/57

13r 2200 14c. (cl)

4. (cl)

0.35 13 1.

14r. (cl)

±

18 0.

Ga

Discordia line through 7 SIMS-analyses: upper icpt.: 2689 ± 14 Ma MSWD = 0.91

13c 6

8

10

Granodiorite, Grunnfarnes, southwestern Senja (c04-46) 4c.

c.

C

c04-46

D d.

9r.

9c.

Granodiorite crystallization: SIMS (4c & 7, conc. age):

2400

55 % disc.)

14r.

15c.

0.43

2200

14r.

10c.

±

80 0.

14c.

100 µm

16c.

16r.

93 0.

±

2 0.

02 0.

4r. a G

9r.

9

4r.

0.48

206

3.

f.F 7.

Pb/ 238U

c04-47

Xenocrysts/cores: TIMS (157/16 & 157/60): Upper icpt.: 2827 ± 6 Ma, MSWD = 0 SIMS (15c & 9c, Pb/Pb): 2816 ± 10 Ma, MSWD = 0.65 0,65

16r.

Neosome, Grunnfarnes, southwestern Senja (c04-47) e.

E

13

11

Neosome crystallization: TIMS (7 points): Upper icpt.: 2612 ± 28 Ma MSWD = 184

2600

0.40

253/55.

12c.

10.

Myhre et al., Figure 8

12c.

251/56.

6r.

157/56.

12r.

6c. 157/53.

157/55.

7.

10.

4r.

253/56.

2100

12r.

4c.

3.

2400

2200

6r.

15

2500

2300

4c.

9c.

253/58.

a G

157/60.

253/59.

0.35

15c.

157/61.

116/12.

106/50.

10c.

106/49.

253/57.

113/52.

253/60.

MSWD = 344

10r. (not shown,

2800 16c.

2600

113/52): upper icpt.: = 2709 ± 30 Ma

0.51

14

4c.

7.

14c.

= 2707 ± 11 Ma, MSWD = 1.17 TIMS: (253/57 -58 - 59 - 60 - 116/12 -

7.

4r.

12

61 0.

±

2 0.

a G

65 0.

±

2 0.

251/57.

a G

Xenocrysts/cores: SIMS (6c, 12c & 5c): Upper icpt.: 2698 ± 13 Ma MSWD = 1,9 1.9

0.32

6c.

157/54.

8

10

12

207

Pb/ 235U

Figure 8. (A–D) U–Pb concordia diagrams and zircon images for samples from Senja.

Error ellipses and quoted errors are 2σ.

NORWEGIAN JOURNAL OF GEOLOGY

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

medium-grained granodiorite, with biotite more abundant than hornblende. Some coarse-grained hornblende laths have a symplectitic texture. Feldspar commonly occurs as elongate laths resembling primary igneous grains, but with some modification to the grain boundaries and considerable grain-size variation due to deformation.

Quartz grains are completely recrystallised and occur as amoeboid grains with smooth, lobate grain boundaries, also with a considerable grain-size variation. The feldspar and quartz microstructure is consistent with hightemperature deformation. Undulatory extinction is visible within some quartz grains, which may represent a younger deformation event at low temperature. Accessory zircon, titanite and allanite-cored epidote are present, the epidote occurring as a low-grade secondary phase. In heavymineral separates, titanite grains are almost colourless, and one analysis with a U content of 12 ppm (Table 1) gives a highly discordant Mesoproterozoic age, presumably related to Proterozoic and/or Caledonian alteration, but because of the small amount of radiogenic Pb the age is not very meaningful. Zircons are mainly subhedral with somewhat resorbed external morphologies, and their shapes range from almost round via short prisms (e.g., Fig. 8C, grain 15) to elongate prisms with length/width ratios of 5 or more (e.g., Fig. 8C, grain 7). Internally, slender prisms and cores display fairly sharp growth patterns, with more faint zoning in the rim and tip domains. Cores with a different texture are common, especially among the short prisms and they can be identified by a different shade in BSE and locally by cracks along the core-rim interface and cracked outer shells (due to expansion of the core). Slender prisms analysed by TIMS define a six-point discordia line with an upper intercept of 2709 ± 30 Ma (Fig. 8D). At this upperintercept point, two SIMS analyses of elongated prism and core give a concordia age of 2707 ± 11 Ma, with two more (discordant) analyses of cores containing an older component. Slightly older ages of 2816 ± 10 Ma are given

by SIMS core analyses 15c and 9c. This age component is also recorded by two TIMS analyses of resorbed prisms (Fig. 8D, analyses 157/60–61). Rims analysed by SIMS give discordant results that plot somewhat above both of these two discordia lines, and considering Th/U ratios that are consistently 1–2 orders of magnitude lower than the cores and prisms (see Table 2), it appears that these rims represent a distinctly younger growth phase of zircon. Unfortunately, the discordant nature of the rim analyses prevents any sound estimate of their age; they could be Neoarchaean with superimposed Pb loss, or mixed Palaeoproterozoic and Neoarchaean analyses. The external morphology and internal zoning point to a magmatic origin of the two oldest zircon populations at c. 2.80 and 2.70 Ga (Fig. 8D), with rims recording a poorly defined younger event. In the field, the analysed granodiorite (Fig. 4C) is closely associated with neosome and the migmatisation probably had some effect on the granodiorite. Therefore, it is reasonable to assume that pre-migmatitic zircons developed thin rims during partial melting. Essentially two possibilities then exist to explain formation of the two older populations in the sample; either the oldest population of cores may

19

represent inheritance and the slender prisms formed during crystallisation of the granodiorite, or alternatively, the slender prisms could be related to an earlier phase of partial melting but this presumably would require a higher amount of melt than observed to develop magmatic zoning in the slender prisms.

Neosome, Grunnfarnes, southwestern Senja (c04–47) This sample represents dm-thick neosome layers from stromatic, folded migmatite at Grunnfarnes. The composition is granitic with dark-brown biotite and accessory zircon, apatite and epidote. Many zircons are stubby, subhedral to rounded grains usually with quite murky coloration and common cracks, and with common cores, some of which have overgrowths with welldeveloped zoning and crystal faces (e.g., Fig. 8E, grain 4.). A second group of zircons consists of single-domain elongate prisms (e.g., Fig. 8E, grain 10). In BSE, both types of zircon display magmatic-type zoning, with rims showing a rather faint zoning. Tips and elongate prisms were analysed by TIMS and the data for six analyses of prisms and two of tips define a discordia line with an upper intercept of 2612 ± 28 Ma, with four SIMS analyses on rims plotting more or less on the same trajectory (Fig. 8F). Cores, analysed only by SIMS, give a distinctly older age than the slender prisms and overgrowths, defining an upper intercept age (Fig. 8F, three analyses) of 2698 ± 13 Ma. Our preferred interpretation is that the slender prisms and overgrowths are related to crystallisation of the neosome at c. 2.6 Ga and that the cores represent zircons from the palaeosome, essentially derived from rocks equivalent to the premigmatitic granodiorite discussed above.

Discussion The field observations and U–Pb ages presented here document the presence of Archaean rocks in the geotransect through the full length of the West Troms Basement Complex from southwestern Senja to Vannøya in the northeast, and provide new and important information that serves as a basis for discussing Archaean crust-forming events and tectonomagmatic evolution of the West Troms Basement Complex (Figs. 1, 9). The data suggest three main episodes of Archaean crust formation, and one superimposed, late Neoarchaean, high-grade metamorphic event, whose significance with respect to crustal growth and assembly history is discussed below.

Archaean rocks on southwestern Ringvassøya In inner parts of Skarsfjorden, a NW–SE-trending shear zone separates the largely tonalitic Dåfjord gneiss from the banded migmatitic gneisses of the Kvalsund gneiss.

The ages demonstrate that the Dåfjord gneiss is older than both generations of zircon in the Kvalsund gneiss neosome, so it is evident that the shear zone separates two pieces of Archaean crust that differ in both age and

20

P.I. Myhre et al.

NORWEGIAN JOURNAL OF GEOLOGY

2600 Tonalite, Vannøya (Bergh et al., 2007)

2700

2800

2900 Ma Legend Age of inheritance Igneous age

Tonalitic gneiss, southeast Ringvassøya (Zwaan & Tucker, 1996) Mikkelvika alkaline stock (Zozulya et al., 2009)

Migmatization age

Metavolcanic (two samples), lower unit Ringvassøya greenstone belt (Motuza et al., 2001) Quartz-keratophyre, Skogsfjordvann on Ringvassøya (Kullerud et al., 2006a)

Uncertain assignment of age to the specific events.

Tonalite in sheared part of the Dåfjord gneiss, Skarsfjord area, Ringvassøya (c01-107) Neosome in Kvalsund gneiss, Ringvassøya (c01-110)

1.

2.

Tonalite, Bakkejord pluton, Kvaløya (c04-35) Mafic dyke intruding the Bakkejord pluton, Kvaløya (c04-31) Neosome in Kattfjord gneiss, Torsnes, Kvaløya (pim07-69)

3.

4.

Episodes of crust formation Superimposed high-grade event

Granite in Kattfjord gneiss, Torsnes, Kvaløya (pim07-67) + KV02 from Corfu et al. (2003) Granodiorite on the eastern flank of Svanfjellet belt, Senja (pa04-2) Granodiorite, Grunnfarnes, southwestern Senja (c04-46) Neosome, Grunnfarnes, southwestern Senja (c04-47)

Myhre et al., Figure 9.

Figure 9. Archaean age relationships along the geotransect in the West Troms Basement Complex. Sample locations are shown in Fig. 1 and on the cross section in Fig. 10. The proposed subdivision into three Meso- and Neoarchaean episodes of crust formation is represented by pink, orange and yellow squares in the age diagram, and a subsequent, poorly constrained, high-grade event is marked by blue squares.

lithology. In detail, the tonalitic Dåfjord gneiss records two zircon populations at 2.92 and 2.85–2.80 Ga, with some uncertainty regarding assignment of each population to crystallisation, inheritance or metamorphism. The youngest population is, within error, comparable in age to other dated samples of the Dåfjord gneiss and the Ringvassøya greenstone belt (see Fig. 9), but somewhat younger than the tonalite on Vannøya. The oldest population represents a Mesoarchaean event previously only documented in the West Troms Basement Complex by detrital zircon grains in the Vanna group (Bergh et al., 2007). The complexity in the data from the neosome in the Kvalsund gneiss reflects an up to 300 Myr history of this rock, starting with formation of the oldest zircon population at 2730 Ma and two populations of prisms and rims at c. 2.70 and between 2.70–2.57 Ga. The deformation along the shear zone which separates the two dated units took place at rather high-grade conditions as evidenced by partial melting of mafic enclaves in dilatant sites within the Kvalsund gneiss (Fig. 2A, B), and lobate grain boundaries in the sheared tonalite. As outlined in Bergh et al. (2010), both Neoarchaean and Palaeoproterozoic, large-scale tectonics are responsible for the present configuration of the West Troms Basement Complex, and so either of these events could be responsible for juxtaposing the Dåfjord and Kvalsund gneiss units. This question depends on how the Neoarchaean ages from the neosome (c01–110, Fig. 5C, D) are interpreted. If the dated neosome is correlated with the dynamic melting structures within the shear zone, then movement on the shear zone was Neoarchaean (>2.56 Ga). However, this correlation is necessarily somewhat ambiguous since the sampled neosome (Fig. 2E) was not taken in the close vicinity of

the enclaves displaying dynamic melting (Fig. 2B, C). A Neoarchaean timing of the deformation is supported by the fact that the shear fabric is cut by assumed 2.40 Ga mafic dykes (Kullerud et al., 2006a) that are unaffected by shearing (Fig. 2D). This sequence of events fits in a very simple way with the tectonic model envisaged in Bergh et al. (2010), involving a tonalitic terrane that is juxtaposed against younger migmatitic terranes in the southwest.

Neoarchaean rocks on Kvaløya: the Bakkejord pluton and Kattfjord gneiss The new data document that the Bakkejord pluton and Kattfjord gneiss on Kvaløya are both Neoarchaean in age, in agreement with the results from Corfu et al. (2003). The Bakkejord pluton grades into locally migmatitic banded felsic and mafic gneisses of the Kattfjord gneiss. Since the dated neosome and granite in the Kattfjord gneiss from Torsnes contain inherited zircons of a similar age as the Bakkejord pluton, this would indicate that rocks equivalent in age to the Bakkejord pluton make up at least some of the protoliths of the Kattfjord gneiss. The contact between the two units could represent an active continental margin/mobile belt, where a pre-existing 2.75–2.70 Ga continental mass experienced deformation and renewed granitic magmatism and migmatisation in the subsequent c. 10–30 Myr interval. The renewed tectonothermal activity is represented by migmatitic fabrics (Fig. 3) and the dated neosome and granitic orthogneiss with ages of c.

2.70–2.67 Ga. In detail, depending on the interpretation of the data, it is possible that the neosome represents an even younger (c. 10 Myr) event, but this remains unresolved.

Additional constraints on the age of the protolith

NORWEGIAN JOURNAL OF GEOLOGY

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

basement within the Kattfjord gneiss are provided by the Meso- and Neoarchaean populations of detrital zircons in psammite of the Torsnes belt (Myhre et al., 2011), with significant populations at 2.90–2.80 Ga and 2.75–2.65 Ga that correspond to the ages presented in this paper. The Neoarchaean tectonothermal event in the studied part of Kvaløya was terminated by mafic dyke emplacement in the Bakkejord pluton at 2671 ± 1 Ma. These mafic dykes probably correlate with the ones in the Torsnes area (Fig. 3A, D, E), but we do not rule out the possibility that younger mafic dykes could also be present here. Primary margins of the mafic dykes in the Torsnes area show some evidence of emplacement in hot crust, and this interpretation would be compatible with a thermal event around this time since the migmatites and granitoids are only slightly older than the 2671 Ma mafic dykes.

The granitic orthogneiss dated herein from the eastern flank of the Torsnes belt is equivalent in age to granite from the western flank dated by Corfu et al. (2003), and it thus appears that the Neoarchaean rocks are coherent on either side of this supracrustal belt.

Neoarchaean rocks with Mesoarchaean components on Senja Neoarchaean rocks have been assumed to be present within banded and migmatitic gneisses on Senja (Zwaan 1995; Zwaan et al., 1998, 2003; Zwaan & Fareth, 2005), but the age and extent of such rocks have remained uncertain (Bergh et al., 2010) and limited to preliminary results from samples c04–47 and c04–46 reported by Kullerud et al. (2006b). The new data presented here document the presence of 2692 ± 6 Ma granodiorite immediately to the east of the Svanfjellet belt (Fig. 1), within dioritic to granodioritic gneisses. This rather homogeneous dioritegranodiorite unit extends to the western flank of the Svanfjellet belt, eventually grading into more complex banded gneisses farther west (Fig. 1). The shore section at Grunnfarnes can be considered representative of this 150– 200 km2 region, and the U–Pb data, although somewhat imprecise due to scatter and discordance, indicate a threestage Archaean evolution. The oldest event is recorded by c. 2.83 Ga zircon crystals and cores in a granodiorite, which also contains a second generation represented by c. 2.70 Ga zircon overgrowths and prismatic zircon, representing either crystallisation or metamorphism of the rock. This age is also recorded in cores of zircon in the neosome, suggesting that they represent a restite equivalent in age to the diorite-granodiorite rafts, and this is, in fact, implied by the field observations where these rafts are in a partial melting state (Fig. 4B). The crystallisation age of the neosome is likely represented by zircon overgrowths and prismatic zircon with an age of c. 2.6 Ga, distinctly younger than any zircons recorded in the granodiorite. Field observations here show that the pervasive migmatisation event was followed by folding,

21

mafic dyke intrusion and a second event with intrusion of granitoid dykes and sheets (Fig. 4D, E).

An Archaean geotransect within the West Troms Basement Complex Fig. 9 is a timeline illustrating the existing and new Meso- and Neoarchaean U–Pb data from the West Troms Basement Complex. The 2.92–2.80 Ga tonalite and greenstone province in the northeast represents the earliest cratonisation of the West Troms Basement Complex, and the 2.90–2.80 Ga tonalitic rocks here have been acknowledged for some time, both from Ringvassøya (Zwaan & Tucker, 1996; Zwaan et al., 1998; Motuza et al., 2001; Kullerud et al., 2006b) and from Vannøya (Bergh et al., 2007). An even older event at c. 2.92 Ga is documented in this work by one zircon population in tonalite (c01– 107) from Skarsfjorden, thus expanding the time frame for the West Troms Basement Complex slightly. Neoarchaean rocks were subsequently docked against this Mesoarchaean province, probably in the Neoarchaean since the shear fabric along the docking shear zone in Skarsfjorden is cut by mafic dykes that may be correlated with the 2.40 Ga Ringvassøya dyke swarm. A Neoarchaean time of docking may also be supported by the potential correlation of Neoarchaean stromatic migmatite in the Kvalsund gneiss with dynamic melting structures within the shear zone. The most likely process of these cratonisation events, including highgrade ductile shearing and migmatisation, is by crustal contraction and accretion caused by plate convergence, probably also involving juxtaposition of different terranes (Bergh et al., 2010). The main Neoarchaean crust formation to the southwest of this shear zone took place between 2.75 and 2.67 Ga in two distinct events, but this region also contains Mesoarchaean rocks as recorded by a group of 2.83 Ga zircons in a migmatitic granodiorite from southwest Senja (Figs. 9, 10). We have found no major ‘breaks’ in the ages of the Archaean rocks associated with the structural architecture of this part of the geotransect. For example, both flanks of the high-strain supracrustal belt in Torsnes are underlain by age-equivalent granitoids. Similarly, we have documented 2.69–2.70 Ga granitoid gneisses on either side of the Svanfjellet belt on Senja, suggesting that large parts of the Neoarchaean rocks here record the same cratonisation history. However, ages of felsic crust are not uniquely diagnostic, and it is possible that there are some distinct differences in, e.g., geochemical affinities associated with the structural architecture seen in the cross section in Fig. 10. Also, large areas of presumed Neoarchaean crust on Kvaløya and Senja are still undated and documentation of these areas, e.g., within the Senja Shear Belt, would be necessary to fully resolve this question.

Age constraints for Palaeoproterozoic and younger events Most parts of the Archaean geotransect in the West Troms Basement Complex have been overprinted by the Svecofennian (1.8–1.7 Ga) tectonometamorphic events and invaded by intrusions as outlined in Bergh

22

P.I. Myhre et al.

c04-46 c04-47

NORWEGIAN JOURNAL OF GEOLOGY

pa04-2

a.

b.

Senja shear belt: uncertain extent of Archaean rocks

Southwest Senja gneiss complex Svanfjellet belt

pim07-67 pim07-69 kv02

c04-31 c04-35

c01-107 c01-110

Bakkejord pluton

B et al. '07

M et al. 01 K et al. 06

e.

Dåfjord gneiss

Gråtind migmatite

Kattfjord gneiss

Z et al '09

Z & T '96 c. d.

Kvalsund gneiss

0

10

20 kilometres

Myhre et al., Figure 10.

Figure 10. Cross section of the geotransect in the West Troms Basement Complex (from Bergh et al., 2010). The units with Archaean rocks discussed in this paper are labelled on the cross section. Names of other units are omitted for simplicity, but see Fig. 1 and Bergh et al. (2010) for details. The trace of the cross section is marked as a black stippled line in Fig. 1, and the legend (colour scheme) of the cross section is the same as in Fig. 1.

et al. (2010). In spite of the widespread Svecofennian intrusive and metamorphic activity, the U–Pb systematics of the Archaean minerals presented here provide little direct information about the timing of these events. The only exception is the pale titanite in the tonalite of the Bakkejord pluton on Kvaløya that plots very close to the lower intercept of a Neoarchaean–Palaeoproterozoic mixing line at 1717 ± 74 Ma (Fig. 6). One could perhaps expect a similar, Neoarchaean–Palaeoproterozoic, twostage evolution of zircon manifested as, e.g., mixing lines or Palaeoproterozoic zircon rims, but this is not the case in any of the samples. In fact, the zircon data for the Bakkejord tonalite (Fig. 6) plot close to a Neoarchaean upper-intercept age with a very long projection towards a lower-intercept age between 0.7 and 0 Ga, which is likely related to Pb loss. Mixing between two Archaean age components, on the other hand, is evident in some of the samples, e.g., c01–110, where the data reveal the presence of a Neoarchaean zircon population present as cores and large fragments, reflecting either a xenocrystic population or a population related to the neosome, with rims that provide a younger Neoarchaean age interpreted either as the time of migmatisation or of a secondary metamorphic overprinting. In summary, there is little evidence from the TIMS data that, for example, zircon growth as a response to Svecofennian reworking is responsible for the discordance.

This can be confirmed by SIMS analyses, where zircon rims were targeted specifically to resolve this question.

Consequently, we attribute the discordance mainly to post-Archaean Pb loss, in some cases superimposed on mixing between Archaean age components. The Pb-loss lines commonly give various Neoproterozoic lower intercepts, which were also noted by Corfu et al. (2003).

These lower intercept ages are imprecise due to very long projections, but many overlap with Neoproterozoic Ar–Ar ages presented by Dallmeyer (1992), interpreted by him to result from Sveconorwegian tectonothermal activity.

Again, in the data presented here (or in other U–Pb data from the West Troms Basement Complex) there are unambiguous indications of zircon growth at that time.

Regional correlatives Archaean rocks in Vesterålen, southwest of Senja (Fig. 1), have magmatic protolith ages of between 2.85 and 2.70 Ga, and they record a high-grade event at c. 2640 Ma (Griffin et al., 1978; Corfu, 2007). Along with the seemingly equivalent, present structural position on the basement high west of the Scandian nappes, there seem to be just as good reasons to correlate between the Neoarchaean rocks in Vesterålen and the West Troms Basement Complex, as internally within the West Troms Basement Complex. The closest area of Archaean rocks in the Fennoscandian Shield is the Norrbotten province in northern Sweden and Finland (Hölttä et al., 2008), with Meso- to Neoarchaean TTG rocks of 2.83 and 2.67 Ga (Öhlander et al., 1987). Archaean rocks are also thought to underlie Palaeoproterozoic rocks to the southwest of the Norrbotten province, defining the NW– SE-trending Archaean–Proterozoic boundary in the Fennoscandian Shield (Öhlander et al., 1987; Öhlander & Skiöld, 1994). Thus, in broad terms, the Meso- to Neoarchaean tectonomagmatic evolution is similar, and the West Troms Basement Complex could constitute the northwestern extension of the Norrbotten province.

Even though the West Troms Basement Complex and the nearby Lofoten–Vesterålen province may be considered as an autochthonous part of the Fennoscandian Shield, the tectonostratigraphic position within the Caledonian framework still remains enigmatic. The current position on the edge of the Fennoscandian Shield raises the question as to what role it played in past plate-tectonic scenarios, such as the likely communion of Baltica and Laurentia at the end of the Archaean (Bleeker & Ernst, 2006; Mertanen & Korhonen, 2011) and the situation prior to Palaeoproterozoic orogeny (e.g., Bridgwater et al., 1990; Connelly et al., 2000; Myhre et al., 2011).

Whether or not the West Troms Basement Complex is allochthonous, it does occupy a position on the very edge of the Fennoscandian Shield and therefore it makes sense to point out other areas in the North Atlantic realm where Archaean rocks are prominent. Palaeogeographical maps discussed in the literature look rather familiar in many cases, and illustrate these possible correlative cratons both in the Palaeoproterozoic (e.g., Bridgwater et al., 1990;

NORWEGIAN JOURNAL OF GEOLOGY

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex

23

Connelly et al., 2000) and at the end of the Archaean The rest of the geotransect (including the Kvalsund and (Hölttä et al., 2008).

Kattfjord gneisses and gneiss units on Senja) records a uniform two (three)-stage Neoarchaean cratonisation.

In spite of its complex post-Archaean evolution, the Mafic magmatism West Troms Basement Complex contains a remarkably Whereas mafic magmatism previously documented in well-preserved record of Meso and Neoarchaean crustal the West Troms Basement Complex at 2.40 Ga (Kullerud evolution. The rocks were formed and modified at globally et al., 2006a), 2.221 Ga (Bergh et al., 2007) and 1.99 Ga important times of Archaean cratonisation, and there are (Myhre et al., 2011) can be correlated with either Baltica many potential correlatives in the North Atlantic realm or Laurentia, to our knowledge the 2.671 Ga mafic dyke and Fennoscandian Shield. For example, a correlation swarm reported here does not have any direct equivalents with the nearby Archaean rocks in Vesterålen seems to be in those cratons. Since this dyke swarm is only slightly supported by the data. younger than some of the associated granitoid intrusions, it is perhaps more appropriate to consider it as a result of the same orogenic event rather than a distinct continental Acknowledgments. The constructive and thorough reviews by Bernard dyke swarm. Although somewhat younger, in terms of Bingen and an anonymous reviewer are greatly acknowledged. The first author wishes to thank David Root and Pritam Nasipuri for tectonic position the mafic dykes invite a comparison discussions and help with the fieldwork. Gunborg Bye Fjeld is thanked with the 2.695 Ga Mikkelvik alkaline stock in Ringvassøya for preparing most of the mineral separates. Paul E.B. Armitage is thanked for conducting fieldwork and producing several University of (Zozulya et al., 2009).

Conclusions New U–Pb data from the West Troms Basement Complex in North Norway document the extent of an Archaean geotransect perpendicular to the structural grain of the region. The U–Pb data, together with published data, outline a three-stage magmatic evolution of the West Troms Basement Complex, with a subsequent, local, Neoarchaean high-grade metamorphic event. The oldest rocks are found in the northeastern part of the West Troms Basement Complex, with various 2.92–2.80 Ga granitoids and a greenstone belt. This region is bounded against younger banded migmatitic gneisses of the Kvalsund gneiss by a high-grade shear zone. The Kvalsund gneiss and the remainder of the West Troms Basement Complex record important events of crustal magmatism and reworking at 2.75–2.70 Ga and 2.70–2.67 Ga, and concluded with the emplacement of a mafic dyke swarm on Kvaløya at 2.671 Ga. The 2.75–2.70 Ga event saw intrusion of the major Bakkejord pluton and formation of the protoliths to some of the banded gneisses, followed by granitoid intrusion and migmatisation at 2.70–2.67 Ga. On Senja, granodiorite on the eastern flank of the Svanfjellet belt crystallised at 2.69 Ga, and a similar age is also recorded in two samples from southwest Senja where magmatism and/or metamorphism occurred during distinct events around 2.80, 2.70 and 2.60 Ga. Two samples of neosome from Ringvassøya and Senja record evidence of a latest Neoarchaean high-grade event, but the details regarding the timing and nature of this event remain to some degree uncertain.

The architecture of the geotransect is the combined result of Archaean accretionary processes and subsequent, dominantly Svecofennian, deformation. The high-grade shear zone separating the dominantly Mesoarchaean Dåfjord gneiss and the dominantly Neoarchaean Kvalsund gneiss is likely an example of the Neoarchaean tectonism.

Tromsø internal reports on the geology of northwest Ringvassøya and the Svanfjellet area on Senja, and for providing sample pa04–2. We thank Martin Whitehouse, Lev Ilyinsky and Kerstin Lindén at the Nordsim laboratory for technical assistance. SIMS analyses were supported by NFR ‘Småforskmidler’ granted to Kåre Kullerud. The Nordic geological ion microprobe facility (Nordsim) is operated and funded under an agreement between the funding agencies of Denmark, Norway, Sweden and Finland, and the Swedish Museum of Natural History. This is Nordsim publication no. 326.

References Andersen, T.B., Jamtveit, B., Dewey, J.F. & Swensson, E. 1991: Subduction and eduction of continental crust: major mechanisms during continent-continent collision and orogenic extensional collapse, a model based on the south Norwegian Caledonides. Terra Nova 3, 303–310.

Armitage, P.E.B. & Bergh, S.G. 2005: Structural development of the Mjelde–Skorelvvatn Zone on Kvaløya, Troms: A metasupracrustal shear belt in the Precambrian West Troms Basement Complex, North Norway. Norwegian Journal of Geology 85, 117–132.

Bergh, S.G., Kullerud, K., Myhre, P.I., Corfu, F., Armitage, P.E.B., Zwaan, K.B. & Ravna, E.J.K. in press: The Archaean elements of the basement outliers west of the Scandinavian Caledonides in North Norway: architecture, evolution and possible correlation with Fennoscandia. Archean earth and early life, Springer.

Bergh, S.G., Kullerud, K., Corfu, F., Armitage, P.E.B., Auvray, B., Johansen, H.W., Pettersen, T. & Knudsen, S. 2007: Low-grade sedimentary rocks on Vanna, North Norway: A new occurrence of a Palaeoproterozoic (2.4–2.2 Ga) cover succession in northern Fennoscandia. Norwegian Journal of Geology 87, 301–318.

Bergh, S.G., Kullerud, K., Armitage, P.E.B., Zwaan, K.B., Corfu, F., Ravna, E.J.K. & Myhre, P.I. 2010: Neoarchaean to Svecofennian tectono-magmatic evolution of the West Troms Basement Complex, North Norway. Norwegian Journal of Geology 90, 21–48.

Bleeker, W. & Ernst, R. 2006: Short-lived mantle generated magmatic events and their dyke swarms: The key unlocking Earth’s paleogeographic record back to 2.6 Ga. In Hanski, E., Mertanen, S., Rämö, T. & Vuollo, J. (eds.): Dyke Swarms – Time Markers of Crustal Evolution. Balkema Publishers, Rotterdam, pp. 3–26.

Bridgwater, D., Austrheim, H., Hansen, B., Mengel, F., Pedersen, S. & Winter, J. 1990: The Proterozoic Nagssugtoqidian mobile belt of southeast Greenland: A link between the eastern Canadian and Baltic shields. Geoscience Canada 17, 305–310.

24

P.I. Myhre et al.

Connelly, J.N., van Gool, J.A.M. & Mengel, F.C. 2000: Temporal evolution of a deeply eroded orogen: the Nagssugtoqidian Orogen, West Greenland. Canadian Journal of Earth Sciences 37, 1121–1142.

Corfu, F. 2004: U–Pb age, setting and tectonic significance of the anorthosite-mangerite-charnockite-granite suite, Lofoten– Vesterålen, Norway. Journal of Petrology 45, 1799–1819.

Corfu, F. 2007: Multistage metamorphic evolution and nature of the amphibolite–granulite facies transition in Lofoten–Vesterålen, Norway, revealed by U–Pb in accessory minerals. Chemical Geology 241, 108–128.

Corfu, F., Armitage, P.E.B., Kullerud, K. & Bergh, S.G. 2003: Preliminary U–Pb geochronology in the West Troms Basement Complex, North Norway:Archaean and Palaeoproterozoic events and younger overprints. Geological Survey of Norway Bulletin 441, 61–72.

Dallmeyer, R.D. 1992: 40Ar/39Ar mineral ages within the Western Gneiss Terrane, Troms, Norway: evidence for polyphase Proterozoic tectonothermal activity (Svecokarilian and Sveconorwegian).

Precambrian Research 57, 195–206.

Griffin, W.L., Taylor, P.N., Hakkinen, J.W., Heier, K.S., Iden, I.K., Krogh, E.J., Malm, O., Olsen, K.I., Ormaasen, D.E. & Tveten, E. 1978: Archaean and Proterozoic crustal evolution in Lofoten–Vesterålen, N Norway. Journal of the Geological Society 135, 629–647.

Henderson, I. & Kendrick, M. 2003: Structural controls on graphite mineralisation, Senja, Troms. NGU Report 2003.011, 111 pp.

Hölttä, P., Balagansky, V., Garde, A.A., Mertanen, S., Peltonen, P., Slabunov, A., Sorjonen-Ward, P. & Whitehouse, M. 2008: Archean of Greenland and Fennoscandia. Episodes 31, 13–19.

Kullerud, K., Skjerlie, K.P., Corfu, F. & de La Rosa, J.D. 2006a: The 2.40 Ga Ringvassøy mafic dykes, West Troms Basement Complex, Norway: The concluding act of early Palaeoproterozoic continental breakup. Precambrian Research 150, 183–200.

Kullerud, K., Corfu, F., Bergh, S.G., Davidsen, B. & Ravna, E.K. 2006b: U–Pb constraints on the Archean and Early Proterozoic evolution of the West Troms Basement Complex, North Norway. Bulletin of the Geological Society of Finland, Special Issue 1, p. 79.

Mertanen, S. & Korhonen, F. 2011: Paleomagnetic constraints on an Archean–Paleoproterozoic Superior–Karelia connection: New evidence from Archean Karelia. Precambrian Research 186, 193– 204.

Motuza, G., Motuza, V., Beliatsky, B. & Savva, E. 2001: The Ringvassøya greenstone belt (Tromsø, north Norway): implications for a Mesoarchaean subduction zone. EUROPROBE time-slice symposium ‘Archaean and Proterozoic Plate Tectonics: Geological and Geophysical Records’, 43–44.

Myhre, P.I., Corfu, F. & Bergh, S. 2011: Palaeoproterozoic (2.0–1.95Ga) pre-orogenic supracrustal sequences in the West Troms Basement Complex, North Norway. Precambrian Research 186, 89–100.

Sawyer, E.W. 2008: Atlas of migmatites. The Canadian Mineralogist, Special Publication 9, 389 pp.

Steltenpohl, M.G., Kassos, G., Andresen, A., Rehnström, E.F. & Hames, W.E. 2011: Eclogitization and exhumation of Caledonian continental basement in Lofoten, North Norway. Geosphere 7, 202– 218.

Whitehouse, M.J. & Kamber, B.S. 2005: Assigning dates to thin gneissic veins in high-grade metamorphic terranes: a cautionary tale from Akilia, southwest Greenland. Journal of Petrology 46, 291–318.

Whitehouse, M.J., Kamber, B.S. & Moorbath, S. 1999: Age significance of U–Th–Pb zircon data from early Archaean rocks of west Greenland – a reassessment based on combined ion-microprobe and imaging studies. Chemical Geology 160, 201–224.

Zozulya, D., Kullerud, K., Ravna, E.K., Corfu, F. & Savchenko, Y. 2009: Geology, age and geochemical constraints on the origin of the Late Archean Mikkelvik alkaline massif, West Troms Basement Complex in Northern Norway. Norwegian Journal of Geology 89, 327–340.

Zwaan, K.B. 1995: Geology of the West Troms Basement Complex, northern Norway, with emphasis on the Senja Shear Belt: a preliminary account. Geological Survey of Norway Bulletin 427, 33–36.

NORWEGIAN JOURNAL OF GEOLOGY

Zwaan, K.B. & Fareth, E. 2005: Bedrock geology map, Mefjordbotn 1433 IV, scale 1: 50,000, preliminary edition, Norges geologiske undersøkelse.

Zwaan, K.B. & Tucker, R. 1996: Absolute and relative age relationships in the Precambrian West Troms Basement Complex, northern Norway (Abstract). 22nd Nordic Geological Winter Meeting, Åbo, Finland, p. 237.

Zwaan, K.B., Fareth, E. & Grogan, P.W. 1998: Bedrock geology map, Tromsø, scale 1:250,000, Norges geologiske undersøkelse.

Zwaan, K.B., Fareth, E. & Johannesen, G.A. 2003: Bedrock geology map, Gryllefjord 1333 I, scale 1: 50,000, preliminary edition, Norges geologiske undersøkelse. Öhlander, B. & Skiöld, T. 1994: Diversity of 1.8 Ga potassic granitoids along the edge of the Archaean craton in northern Scandinavia: a result of melt formation at various depths and from various sources. Lithos 33, 265–283. Öhlander, B., Skiöld, T., Hamilton, P.J. & Claesson, L.-Å. 1987: The western border of the Archaean province of the Baltic shield: evidence from northern Sweden. Contributions to Mineralogy and Petrology 95, 437–450.

NORWEGIAN JOURNAL OF GEOLOGY Channel-bed changes in distributaries of the lake Øyeren delta, revealed by interferometric sidescan sonar

25

Channel-bed changes in distributaries of the lake Øyeren delta, southern Norway, revealed by interferometric sidescan sonar Raymond S. Eilertsen, Nils R.B. Olsen, Nils Rüther & Peggy Zinke Eilertsen, R.S., Olsen, N.R.B., Rüther, N. & Zinke, P.: Channel-bed changes in distributaries of the lake Øyeren delta, southern Norway, revealed by interferometric sidescan sonar. Norwegian Journal of Geology, Vol 93, pp. 25–35. Trondheim 2013, ISSN 029-196X. Multiple high-resolution bathymetric data from interferometric multibeam surveys over a three-year period were used to investigate channel-bed morphology and changes in distributary channels of the Øyeren delta plain, the largest freshwater delta in northern Europe. Most of the erosion over the three-year period occurred at the outside of channel bends and upstream ends of islands, while deposition mostly occurred at the inner bends. Overall, there appears to be a net aggradation within the channels over the three-year period. The dominant bedforms registered are dunes.

The dune morphology is complex with large variations in wavelength, height, lee-side angles, scour depth and crest-line curvature. Their size shows a correlation with flow depth. Another prominent feature registered is that of scours occurring at different settings and at different scales. The largest scour recorded was 23 m deep (below water level) in 2004, and 19.5 m in October 2007. It has migrated more than 20 m downstream and up to 7.5 m of sediments have been removed vertically over the three-year period. Over the same period, 10 m of vertical deposition has taken place at the upstream end of the scour. Between June and October 2007, relatively little erosion has taken place at this locality, apart from a bank collapse at the scour margin. The smaller and more localised scours occur at the lee sides of some dunes, commonly extending well into the stoss side of the downstream dune and with occasional prominent spurs on the flanks.

Raymond S. Eilertsen, Geological Survey of Norway, Polarmiljøsenteret, 9296 Tromsø, Norway. Nils R.B. Olsen, Nils Rüther, Peggy Zinke, Department of Hydraulic and Environmental Engineering, Norwegian University of Science and Technology in Trondheim, Norway.

E-mail corresponding author (Raymond S Eilertsen): raymond.eilertsen@ngu.no

Introduction Many people rely on rivers for water supply, food, power, transport, recreation, waste disposal, and their deposits as a source of raw materials. As a consequence, it is important to understand rivers in order to deal with problems such as floods, water supply, design and construction of infrastructures near or at river banks, river-bank erosion, sedimentation in navigated waterways, restoration of freshwater habitats, and remediation of polluted surface water and groundwater (Bridge, 2003). Rivers as erosive and depositing agents are potentially hazardous to foundations along their courses, such as bridges, riverbank protection structures, cables and other installations. Also, erosion along the riverbank may destabilise slopes along the river causing landslides (Suárez, 1996; Bogen et al., 2002; Eilertsen & Hansen, 2008; Lévy et al., 2012). Man-made structures on the river floor (e.g., cables, bridge pillars) may also affect and be affected by the river flow. In addition, Earth scientists study modern rivers in order to understand how water flows, transports, erodes and deposits sediments, and sedimentologists use such knowledge to interpret the origin of ancient river deposits.

Traditionally, studies of modern rivers have been conducted using aerial photographs and single-beam echo sounders, and by sediment sampling, diving and dredging.

Other geophysical tools such as Doppler equipment have helped to understand the fluid dynamics in rivers (e.g., Lane et al., 1998). A general limitation of all these methods is the lack of resolution and detail of subsurface morphology. New geophysical equipment designed to obtain high-resolution swath bathymetry in shallow waters may provide information on morphology and processes not previously obtained by traditional methods such as single-beam echo sounders and sidescan sonars. Such new information may not only be valuable to Earth scientists (e.g., Parsons et al., 2005; Eilertsen & Hansen, 2008), but also to land-use planners and decision makers. Over the last decade or more, the use of such high-resolution tools has become standard methodology in sea-floor mapping (e.g., Fornari et al., 1988; McAdoo et al., 2000; Laberg et al., 2007; Ottesen et al., 2008; L’Heureux et al., 2009). However, their application in lacustrine and river environments, using portable systems, has been introduced only during the last few years (Bacon et al., 2002; Bini et al., 2007; Eilertsen & Hansen, 2008; Fanetti et al., 2008; Ledoux et al., 2010; Sastre et al., 2010, Nittrouer et al., 2011).

26

R.S. Eilertsen et al.

Dunes are common bedforms in rivers (Bridge, 2003), and they are abundant within the distributary channels of the Øyeren delta in southern Norway (Fig. 1). Their presence influences the water flow and exerts a strong control on the entrainment, transport and deposition of sediment (Parsons et al., 2005). Thus, understanding the nature and origin of dunes may help predict flow resistance, sediment transport and deposition within many rivers (Best, 2005). For instance, cross

NORWEGIAN JOURNAL OF GEOLOGY

stratification formed through migration of dunes is the most common sedimentary structure of many ancient alluvial successions. These deposits may be of varying and complex geometries (Harms et al., 1982; Allen, 1984; Rubin, 1987), and thus may cause heterogeneous and anisotropic permeability fields in, e.g., aquifers and hydrocarbon reservoirs (Weber, 1980, 1986; Van de Graff & Ealey, 1989; Best, 2005). The study of riverdune dynamics has mainly focused on small-scale

Figure 1. Overview of the lake Øyeren delta plain with the October 2007 dataset shown.

The aerial photo is from www.norgedigitalt­.no.

NORWEGIAN JOURNAL OF GEOLOGY Channel-bed changes in distributaries of the lake Øyeren delta, revealed by interferometric sidescan sonar

laboratory experiments and with small field sites and 2D survey tools (e.g., single-beam echo sounders), although in recent years with new technology larger rivers have received increasing interest (e.g., Carling et al., 2000; Best, 2005; Parsons et al., 2005). Furthermore, advances in the understanding of fluid flow and sediment dynamics over dunes have mainly come from studies involving simplified boundary conditions, e.g., 2D dunes under steady flow, which have not included the more complex flow and morphology typical of natural rivers (Best, 2005). As natural bedforms are invariably three-dimensional in alluvial, estuarine and marine environments (Allen, 1984; Baas, 1994, 1999), this oversimplification has imposed inherent limitations on the interpretation and understanding of dune form and flow dynamics (Parsons et al., 2005). However, highresolution swath bathymetry provides the possibility to overcome these limitations and improve our understanding of fluvial processes and dune morphology (e.g., Parsons et al., 2005; Eilertsen & Hansen, 2008).

The aim of this paper is to: (1) describe and interpret channel bedforms in distributaries on a lacustrine delta plain at Øyeren, South Norway; (2) describe changes to these bedforms that have occurred over 24 hrs, 4 months, and a three-year period; (3) contribute to the overall understanding of river-bed processes; and (4) show the ability of interferometric, high-resolution sonar data to be used for ‘full-scale’ studies of natural rivers.

Setting The Øyeren delta is the largest freshwater delta in northern Europe (56 km2; Berge et al., 2002). It is situated at the northern end of the regulated lake Øyeren, with a surface area of 87.4 km2 and a mean water level of 101.37 m above present sea level (Fig. 1). The Øyeren delta has been divided into four morphological units by Bogen & Bønsnes (2002): (1) the delta plain is located between 103 and 101 m a.s.l., and is composed of vegetated islands and five intermittent distributary channels that are from

Figure 2. Lake level for lake Øyeren and discharge for the river Glomma at Rånåsfoss (18 km upstream from the study area) between 2000 and 2008. Arrows point to time of data collection.

27

45 to 600 m wide; (2) the delta platform, positioned between 101 and 96 m a.s.l., and extending about 9 km downstream from the delta plain; and (3) the foreset slope that grades into; (4) the delta bottomset at around contour 36 m. This study focuses on the delta plain.

Three rivers feed the delta: Leira, which joins the river Nitelva, and Glomma (Fig. 1). The Glomma river is the largest, having a mean annual discharge of 682 m3 s-1 measured at Rånåsfoss, 18 km upstream of the study area. It delivers a mean suspended load and bed load of 500,000 tons and 75,000–150,000 tons yr-1, respectively, to the lake delta (Bogen et al., 2002). The smaller rivers Leira and Nitelva deliver suspended loads of 90,000 and 18,000 tons yr-1, respectively. The total catchment area draining into Lake Øyeren is c. 40,000 km2 (Pedersen, 1981), and has a mean annual precipitation and temperature of 820 mm and 4.1°C (Norwegian Meteorological Institute, http://met.no/index.shtml).

Prior to the onset of river regulation in 1862, natural water levels in the lake varied by up to 14 m between spring flood and lowstand during the winter, with a mean fluctuation of 8 m. At present, the water level rarely fluctuates more than 4 m between seasons (Fig. 2; Bogen et al., 2002). Sediments were transported and deposited at the delta front before regulation. However, as a consequence of regulation, sediments are at present being deposited on the delta platform (Bogen et al., 2002). The relationship between the river discharge and lake level is important for sediment erosion and transport within the channels, as the waterline gradient and thus stream velocity controls erosion and transport. For instance, with a high lake level and relatively low discharge the waterline gradient and subsequent stream velocity decreases, causing a reduced transport of sediment onto the delta platform. However, with low lake levels, the water gradient can be sufficient to cause sediment erosion and transport even under relatively low discharges (Bogen et al., 2002). Bogen et al. (2002) measured the bed-load transport along a transect from Fetsund to Årnestangen and noted large variations downstream with the highest

28

R.S. Eilertsen et al.

NORWEGIAN JOURNAL OF GEOLOGY

transport rate close to Fautøya. In contrast, bed-load transport in the lower parts (south) of Storråka was almost negligible, suggesting that most of the bed load was deposited somewhere between these two areas, i.e., around the island Flatsand (Fig. 1; Bogen et al., 2002).

The deltaic sediments are up to 60 m thick in the study area, and are underlain or flanked mainly by glaciofluvial sands and gravels, and glaciomarine sands, silts and clays (Longva, 1991). The beds in the distributary channels consist mainly of medium sand (D50 ~0.4 mm), apart from the Nitelva and lower Kusandråka channels with a mean grain size around that of fine sand (D50 ~0.1mm) (Bogen et al., 2002).

Methods During 12 days in June 2004 (14th–25th of June), 3 days in June 2007 (4th–6th of June), and 5 days in October 2007 (27th–31st), high-resolution bathymetric data were collected using a 250 kHz GeoSwath interferometric sidescan sonar from GeoAcoustics (Fig. 3). During the 2004 and October 2007 cruises, all channels of the delta plain were measured, whereas only parts of the Storråka, Nitelva and Gjushaugsandråka channels were measured during the June 2007 cruise (Fig. 1). In addition, parts of the Kusandråka channel were measured twice in October 2007 with only one day in between. In total, four datasets were collected at different times from the distributary channels.

The swath bathymetry system scans the bed topography in a line perpendicular to the track of the survey vessel, typically with a width several times that of the water depth. The bathymetry is determined by measuring the phase differences between multiple receive staves within a transducer from a returning acoustic wave (Hiller & Lewis, 2004). Vessel speed was around 4 knots, and depending on the width of the swath, 16 to 30 pulses per second (pps) were used, giving a reading for every 6 to 12 cm of the river bottom. Six sound-velocity profiles (SVP) were measured using a Valeport 650 SVP. The water level during the survey was measured digitally using a Valeport 740 instrument submerged into the Nitelva channel that was calibrated with water-level measurements at a fixed station measured by Glomma og Lågen Brukseierforening (ww.glb.no). The water level in Øyeren varied in total between 101.34 and 101.41 m a.s.l. during the surveys (Fig. 2). All datasets have been adjusted to the same water level (101.35 m a.s.l.), which is roughly 1 m below the bank-full level. Differential GPS was used for positioning, giving an accuracy in the horizontal direction of ±1 m. A gyroscope was also used for navigation. Multiple overlapping runs over the same area with fixed bottom structures like bedrock or ‘obstacles’ revealed a consistency in the depth measurements and an accuracy at cm to dm scale. Two types of data were recorded, bathymetric and backscatter data. The former gives high-resolution

Figure 3. (A) The boat with the sonar mounted in the bow. (B) Part of the Rossholmråka channel; the channel is 50 m across. (C) Eastern side of Fautøya. Note the considerable erosion and bank collapse causing tree trunks to fall into the channel. See Fig. 1 for location.

depth information, whilst the latter provides information about the river-floor reflectivity, which depends on the bed character (e.g., grain size, bedrock, roughness).

Processing was conducted using the GeoSwath software, and included sound-velocity correction and calibration to reduce signal-to-noise ratios. The processed dataset was then gridded at a 0.3 m grid scale and imported into

NORWEGIAN JOURNAL OF GEOLOGY Channel-bed changes in distributaries of the lake Øyeren delta, revealed by interferometric sidescan sonar

29

ERMapper 7.1 for analyses of the datasets. Comparisons between datasets were also carried out using the ERMapper 7.1 software. Measurements of individual bedforms included height, length, stoss- and lee-side gradients using maximum elevation (crest height), minimum elevation (troughs), and average lengths of individual dunes as reference (Allen, 1984; Collinson & Thompson, 1989). Bedform three-dimensionality was measured using the non-dimensional span method as defined by Venditti et al. (2005).

Channel-bed morphology The most prominent bed forms are described below and their changes over time are discussed in the next section.

We have distinguished between the more local, smallscale scours occurring at the lee side of dunes (hereafter termed lee-side scours) and the more prominent, largescale scours. The latter are described in a separate section.

Dunes and local scours The dune morphology within the distributary channels is rather complex and dominated by sinuous transverse dunes with well-defined crest lines and more complex 3D forms with almost linguoid shapes (Allen, 1968). The dunes have individual wavelengths between 3 and 53 m (average 16.7 m) and heights between 0.14 and 2.24 m (average 0.64 m; Fig. 4). Typically, the dune size (length and height) increases with depth, although exceptions do occur (Fig. 4C, D). The dune-form index (or aspect ratio, H/L) is between 0.01 and 0.07 (average 0.03). The dunes are asymmetric with lee-side angles between 3 and 25° (average 9.4°), although individual dunes with angles as low as 1.1 and as high as 34.3° also exist (Fig. 5A). The high values typically occur where there are prominent local scours on the lee side of the dunes (see below). The shape of the stoss side ranges from planar to more commonly concave up, with angles between 1.2 and 6.3° (average 2.9°). Superimposed bed forms are rare; however, smaller bed forms such as ripples or small dunes may be present but are not visible due to the resolution of the dataset.

The dune crest lines can be followed for up to 200 m, and their planform curvature (sinuosity, NDS – nondimensional span) varies between 1.04 and 1.63 (average 1.18 for parts of the October 2007 dataset; Fig. 5B).

Commonly, the dunes exhibit crestbrink parting with gentle, rounded dips at the crest followed by a sharper break of slope into the lee side. Also, the heights of individual dunes may vary strongly along individual crest lines, in some cases more than 1 m. Discontinuous crest lines also occur, along with bifurcating and merging of individual crest lines. In channel bends, crest lines can often be traced from the deep thalweg to the shallower inner bend, covering almost the entire channel width.

Figure 4. (A) Plots of dune height as a function of length. (B) Plots of form index (steepness) of dunes as a function of length. The stippled line in (A) and (B) indicates the ’equilibrium’ dune function of Ashley (1990) as H=0.16L0.84. (C) Plots of dune height as a function of flow depth. (D) Plots of dune wavelengths as a function of flow depth. (E) Histograms showing dune lengths.

30

R.S. Eilertsen et al.

NORWEGIAN JOURNAL OF GEOLOGY

level), occurring at a channel confinement laterally to a mid-channel bar (see below). The river bends are characterised by a deep thalweg scour close to the incised bank, with the deepest part close to the downstream end of the bend. Typically, the well-defined thalweg scour begins at the start of the river bend. The river bend scours are between 160 and 600 m long, and from 5 to 14 m deep. Point-bar slopes characterise the inner bend, locally with small chute channels on the surface.

Dunes are common in the scours, with oblique crest lines reaching from the deeper part of the scour onto the point-bar slope.

The confluence scours are between 110 and 200 m long and from 8.5 to 12 m deep, and have both steep upstream slopes and sidewalls with avalanche faces. Downstream slopes are more gently inclined. These scours tend to parallel the direction of the dominant channel (cf., Ashmore & Parker, 1983; Best, 1987) and can be separated into symmetrical and asymmetrical types based on their

Figure 5. (A) Histograms of lee-side slope angles for the October 2007 dataset (N=293). (B) Histograms of dune crest-line sinuosity for the October 2007 dataset (non-dimensional span (NDS) sensu Venditti, 2003. N=164).

In places, local scours are present on the lee side of dunes (Fig. 6). They are up to 20 m long, 10 m wide, and more than 1 m below mean bed level, and in some cases extend almost to the crest line of the downstream dune. Spurs along the scour margins are quite common.

These are between 15 and 60 cm high, mostly with wellrounded crests, and often extend over the whole length of the scour (Fig. 6). Dunes are generally relatively high at the scour location and lee-side angles are also steeper (22 to 34°). Crest-line curvature also appears to increase at localities with scours, and the deeper scours are commonly associated with downstreamconcave crest lines (i.e., saddles; Allen, 1984; Parsons et al., 2005). Saddles are in places succeeded downstream by downstream-convex crest lines (i.e., lobes), especially where the dunes have a strong three-dimensional shape, although this is not always the case. Downstream, the scours and the associated spurs commonly appear to be aligned in ‘bands’ at the lee sides of successive dunes. This is most evident at channel bends (Fig. 6B).

‘Large-scale’ scours Eilertsen & Hansen (2008) reported a series of deep scours on the delta plain based on the 2004 dataset, of which all are also found in the 2007 dataset. A short synthesis of these findings is given in the following.

The scours are generally most pronounced at river bends and channel confluences. However, the deepest scour registered in 2004 was 23 m deep (below water

Figure 6. (A) Lee-side scours in the Storråka channel. (B) Lee-side scours with spurs in the Sniksand channel (note the alignment of the scours and spurs). See Fig. 1 for location.

NORWEGIAN JOURNAL OF GEOLOGY Channel-bed changes in distributaries of the lake Øyeren delta, revealed by interferometric sidescan sonar

31

planform morphology, where lower confluence angles favour the former type and large confluence angles favour the latter (Mosley, 1976; Ashmore & Parker, 1983; Eilertsen & Hansen, 2008).

The largest scour recorded (50 m wide, 100 m long and 23 m deep in 2004) is positioned laterally to the Flatsand mid-channel bar (Fig. 1). The scour has steep walls and a relatively even floor, and is the deepest scour registered on the delta plain. The depth of the channel upstream of the scour increases from c. 5 to 15 m towards the scour, with well-defined sinuous dunes covering the channel floor. A small slide scar is present on the western flank of the scour in the 2007 dataset (see below). The scour was probably generated as a result of channel confinement due to the build up of the mid-channel bar followed by an increase in velocity and vertical erosion (Eilertsen & Hansen, 2008).

Temporal changes of the channel bed The four datasets allow us to observe changes in the river bed on several time scales. However, the time span between the datasets, apart from the two-day survey in the Kusandråka channel, is mostly too long to recognise and record the individual dunes and their migration, although other features can be recognised.

Three-year period The bathymetric changes during the period from 2004 to October 2007 are shown in Fig. 7. The changes within the –2 to 2 m range are generally mostly related to the migration and changes of the dune field within the channels. More significant changes have occurred at bends and the ‘large-scale’ scour at Flatsand (Figs. 7, 8).

The tip and the eastern part of Fautøya show most signs of river-bank erosion between 2004 and 2007. Here, the main channel Storråka flows directly towards the island, causing bank undercutting and collapse (Figs. 3C, 7). This is also where the highest stream velocities occur (Zinke et al., 2010). In 1980, the channel was more than 23 m deep at this location (Pedersen, 1981), whereas today it is ~15 m. Bogen et al. (2002) reported a 3 m-long and 15 m-wide bank collapse in this area following a large flood in 1995. The river bank has retreated more than 10 m along the eastern part of Fautøya between 2004 and 2007.

The bathymetry and difference in bathymetry in parts of the Storråka channel between June 2004 and June 2007, and June 2007 and October 2007, are shown in Fig. 8.

The deep scour probably appeared during a major flood in 1995 (Eilertsen & Hansen, 2008). It has since migrated more than 20 m downstream and locally up to 8 m of sediments have been removed vertically over 3 years (Fig. 8D). Over the same period, 7 m of vertical deposition has taken place at the upstream end of the scour, thus ‘keeping pace’ with the scour migration. Between June and October 2007, the data show that relatively little

Figure 7. (A) Changes in bathymetry between 2004 and October 2007. Note compressed scale. The largest recorded erosion (8 m) and deposition (10 m) occur laterally of the mid-channel bar, Flatsand.

(B) Depth along transect A–A’ shown in (A) for the years 2004 and 2007. Arrows point to areas with a significant aggradation. The aerial photo is from www.norgedigitalt.no.

erosion has taken place, apart from a 30 m-wide, 22 m-long and 3 m-deep bank collapse at the scour margin (Fig. 8E). In contrast, evidence of dune migration is clearly visible in the upstream channel, and more than 3 m of sediments were deposited at the upstream end of the scour.

Overall, there appears to have been a net aggradation of sediments within the Storråka channel, suggesting that deposition exceeded erosion within this particular channel between 2004 and 2007. This is especially evident at the scours and in the shallow areas downstream of these features (Fig. 7B).

Four-month period The bathymetry and difference recorded between the June and October 2007 datasets in parts of the Gjushaugsandråka channel are shown in Fig. 9. A small

32

R.S. Eilertsen et al.

NORWEGIAN JOURNAL OF GEOLOGY

Figure 8. Bathymetry at the Storr책ka channel west of Flatsand showing the largest scour recorded. (A) June 2004 dataset. (B) June 2007 dataset. (C) October 2007 dataset. (D) Changes in bathymetry between June 2004 and June 2007. Note up to 8 m of erosion and 7 m of deposition. (E) Changes in bathymetry between June 2007 and October 2007. Note the bank collapse within the scour hole. In total, the scour has migrated more than 20 m downstream over the three-year period. Flow is from top to bottom. See Fig. 7 for location .

Figure 9. Part of the Gjushaugsandr책ka channel showing the bathymetry measured on (A) June 7th, and (B) October 30th, 2007, respectively.

Four, distinct, lee-side scours are shown (arrows and encircled). (C) Changes recorded between the datasets. White arrows point to areas of accumulation of sediment as the scours migrated downstream. Flow is from top to bottom. See text for further explanation and Fig. 7 for location.

NORWEGIAN JOURNAL OF GEOLOGY Channel-bed changes in distributaries of the lake Øyeren delta, revealed by interferometric sidescan sonar

33

Figure 10. Bathymetry in the Kusandråka measured on (A) Oct. 30th and (B) Oct. 31st. C) Changes recorded over the 24-hour period. Note that, overall, the bed elevation remains the same, though with most deposition at the dune crest line, causing migration of the dune. Flow is from top to bottom. See Fig. 7 for location.

‘obstacle’ provides a marker positioned at the exact same position in both datasets. Three distinct lee-side scours (white arrows in Fig. 9) can be traced in both datasets, showing that they have migrated only about 6–7 m over 4.5 months. A more prominent scour (encircled in Fig. 9) can also be traced showing roughly the same migration rate. While the scours migrated downstream, infilling occurred at their original position with up to 2 m of sedimentation over the 4.5 month period. The dune immediately downstream of the scours appears to have grown during the same period, forming a lobe-shaped crest line (Fig. 9B).

24-hour period The bathymetry and difference over a 24-hour period recorded in the datasets from parts of the Kusandråka channel are shown in Fig. 10. The dune crest lines are easily correlated between the datasets, and although the changes recorded are relatively small compared to the resolution of the system, they do form coherent geological structures and are thus inferred to be real.

The largest changes recorded are at the dune crests and show up to 1 m of dune migration and up to 30 cm of deposition (Fig. 10C). Most of the erosion appears to be concentrated evenly on the stoss side of the dunes.

The crest-line curvature (sinuosity; NDS) measured over the same reach of the Gjushaugsandråka channel in June 2004, June 2007 and October 2007 shows relatively small changes, with mean curvature of 1.13, 1.16 and 1.18, respectively. NDS measurements for the area shown in Fig. 9A, B give values of 1.20 and 1.22, respectively.

Discussion The abundance of low-angle dunes (<10°) in the delta distributaries compares well with results reported from other rivers like the Jamuna River in Bangladesh (Roden, 1998; Best et al., 2007), the Rhine and Waal rivers in the Netherlands (ten Brinke et al., 1999), the Fraser River in Canada (Kostaschuk & Best, 2005), and the Mississippi River in the USA (Harbor, 1998). Although more than 50% of the lee-side angles are lower than 10°, such lowangle dunes may possess intermittent separation in the dune lee (Best & Kostaschuk, 2002), whereas dunes with larger lee-side angles may generate a permanent flow separation. In the delta distributaries, the highest dune angles are normally found in combination with the more prominent scours, although the lee-side angle may be half or less laterally of the same individual dune crest. This lateral variation would likely cause an important topographic steering of the flow, affecting the distribution of shear stresses and sediment transport over the dunes (Allen, 1984; Maddux et al., 2003). The crest-line parting observed at some dune crests may also significantly influence shear and turbulence on the lee side (Best & Kostaschuk, 2002).

The prominent lee-side scours and associated spurs would suggest that a narrow zone of high turbulence is operating at specific locations, probably related to some kind of topographic steering. The convergence of flow in the lee of saddle-shaped dunes may explain some of the scours, but not all are related to saddles. Also, Venditti (2003) argued that the convergence of flow in the lee of saddle-shaped crest lines would reduce the turbulence intensities but increase the average flow velocity.

Another important aspect is how the local scour relates to the dune migration subsequent to its formation, and whether or not the scour will enhance the 3D shape, or if it migrates at the same pace as the lateral dune crest, and/or whether it is infilled. The example shown from

34

R.S. Eilertsen et al.

Gjushaugsandråka (Fig. 9) may suggest that there is little or no effect on the 3D planform pattern of the dunes over time with an NDS increase from 1.20 to 1.22, respectively.

Significant net sedimentation took place at the deep scour in the Storråka channel between June and October 2007 (Fig. 8E). This period was characterised by high lake levels and a drop in discharge (Fig. 2). This suggests that most of the scouring takes place during periods of lake-level lowstand when the river’s gradient is at its largest causing high flow velocities and subsequent erosion, especially during high discharges (Fig. 2). This is in agreement with the conclusions reached by Bogen et al. (2002). However, the flow velocities were still high enough to transport and deposit substantial amounts of sediment (mainly medium sand) during the lakelevel highstand, causing an overall net deposition at this locality. Also, it appears as if the erosion during the flood causes the scour to maintain itself, although migrating more than 10 m yr-1. Such ‘self-maintenance’ has been reported from the study of an artificial scour in the Ohio River (Moore, 1970) and from the Mackenzie Delta, N.W.T. Canada (Fassnacht & Conly, 2000). Evidence of sliding at the scour margin may indicate that bank oversteepening and collapse are important mechanisms in scour formation and maintenance.

The difference in mean crest-line curvature between the datasets is small, ranging only from 1.13 to 1.18. This is slightly lower than the value of 1.2, which Venditti et al.

(2005) defined as the boundary between 2D and 3D bed forms. However, the lowest value occurs early in a period of a more stable lake level following lake-level lowstand, and the higher value occurs in late autumn after a period of several months with a relatively stable lake level. This may suggest that the dunes were adjusting from an initial 2D to a more 3D appearance as the water flow was relatively steady for a longer period of time (Fig. 2).

This is in accordance with other studies showing that if a flow persists for a sufficiently long period, 2D bed forms will evolve to 3D bed forms (Baas et al., 1993; Baas, 1994, 1999; Venditti et al., 2005).

Summary and conclusions High-resolution bathymetry data from distributary channels on a lake-delta plain at lake Øyeren, southern Norway, were collected during four surveys within a period of three years by using interferometric sonar.

The data show that the dune morphology is complex with large variations in wavelength, height, lee-side angles, scour depth and crest-line curvature. Prominent features are localised scours that occur at the lee sides of some dunes, often extending close to the crest line of the downstream dune and with spurs on the flanks. Their presence seems to be related to an increased curvature and 3D appearance of the dunes.

NORWEGIAN JOURNAL OF GEOLOGY

Changes in planform morphology of the crest line with time suggest that the dunes will develop towards a more 3D appearance with persistent flow level, in accordance with flume studies.

Erosion of the bed is at its highest during lake-level lowstand when the gradient between the inflowing river and the lake itself is at its largest, causing high flow velocities and subsequent erosion. Erosion, deposition and migration of dunes are also prominent during lakelevel highstand, but to a lesser extent than during floods, in accordance with previous studies. The present study highlights the potential of the interferometric sidescan sonar for use in fluvial studies.

Acknowledgments. We are indebted to Gunnar Andersen and Oddbjørn Totland for assistance during fieldwork, as well as the Miljøvernavdelingen at Fylkesmannen i Akershus. The work was financed by the Geological Survey of Norway as a part of the GEOSproject (Geology of the Oslo Region). Comments made by two anonymous reviewers and the editor of NJG helped improve the manuscript.

References Allen, J.R.L. 1968: Current Ripples: Their relation to patterns of water and sediment motion. Elsevier, New York, 433 pp.

Allen, J.R.L. 1984: Sedimentary Structures: Their character and physical basis. Vol. 1, Developments in Sedimentology 30, Elsevier, New York, 593 pp.

Ashley, G.M. 1990: Classification of large-scale subaqueous bedforms: a new look at an old problem. Journal of Sedimentary Petrology 60, 160–172.

Ashmore, P. & Parker, G. 1983: Confluence scour in coarse braided streams. Water Resources Research 19, 392–402.

Baas, J.H. 1994: A flume study on the development and equilibrium morphology of current ripples in very fine sand. Sedimentology 41, 185–209.

Baas, J.H. 1999: An empirical model for the development and equilibrium morphology of current ripples in very fine sand.

Sedimentology 46, 123–138.

Baas, J.H., Oost, A.P., Sztano, O.K., de Boer, O.L. & Postma, G. 1993: Time as an independent variable for current ripples developing towards linguoid equilibrium morphology. Terra Nova 5, 29–35.

Bacon, C.R., Gardner, J.V., Mayer, L.A., Buktenica, M.W., Dartnell, P., Ramsey, D.W. & Robinson, J.E. 2002: Morphology, volcanism, and mass wasting in Crater Lake, Oregon. Geological Society of America Bulletin 114, 675–692.

Berge, D., Martinsen, T., Bogen, J., Bønsnes, T.E., Elster, M., Rørslett, B., Sloreid, S-E., Halvorsen, G., Brabrand, Å., Dale, S. & Andersen, R. 2002: Environmental investigations in Lake Øyeren, 1994–2000, Main Report. Akershus Fylkeskommune, Norway, ISBN 82–91036– 46–2 (in Norwegian).

Best, J. 1987: Flow dynamics at river confluences: implications for sediment transport and bed morphology. In Ethridge, F.G. & Flores, R.M. (eds.): Recent and ancient nonmarine depositional environments, Society of Economic Paleontologists and Mineralogists Special Publication 31, pp. 27–35.

Best, J. 2005: The fluid dynamics of river dunes: A review and some future research directions. Journal of Geophysical Research 110, doi: 10.1029/2004JF000218.

Best, J.L. & Kostaschuk, R.A. 2002: An experimental study of turbulent flow over a low-angle dune. Journal of Geophysical Research 107

NORWEGIAN JOURNAL OF GEOLOGY Channel-bed changes in distributaries of the lake Øyeren delta, revealed by interferometric sidescan sonar

(C9), 3135, doi: 10.1029/2000JC000294.

Best, J.L., Ashworth, P.J., Sarker, M.H. & Roden, J.E. 2007: The Brahmaputra–Jamuna River, Bangladesh. In Gupta A (ed.): Large Rivers: Geomorphology and Management, John Wiley & Sons, Ltd., pp. 395–433.

Bini, A., Corbari, D., Falletti, P., Fassina, M., Perotti, C.R. & Piccin, A. 2007: Morphology and geological setting of Iseo Lake (Lombardy) through multibeam bathymetry and high-resolution seismic profiles. Swiss Journal of Geosciences 100, 23–40.

Bogen, J. & Bønsnes, T.E. 2002: The impact of reservoir regulation on the processes of erosion and sedimentation of the delta in Lake Øyeren, Norway. International Association of Hydrological Sciences Publication 276, 103–112.

Bogen, J., Bønsnes, T.E. & Elster, M.S. 2002: Environmental studies in lake Øyeren. Erosion, sedimentation and delta development (in Norwegian). Norwegian Water Resources and Energy Directorate Report 3/2002, 103 pp.

Bridge, J.S. 2003: Rivers and Floodplains: Forms, Processes and Sedimentary Record. Blackwell, Malden, 504 pp.

Carling, P.A, Gölz, E., Orr, H.G. & Radecki-Pawlik, A. 2000: The morphodynamics of fluvial sand dunes in the River Rhine, near Mainz, Germany. I. Sedimentology and morphology. Sedimentology 47, 227–252.

Collinson, J.D. & Thompson, D.B. 1989: Sedimentary Structures. 2 ed., Unwin Hyman Ltd, London, UK, 207 pp.

Eilertsen, R.S. & Hansen, L. 2008: Morphology of river bed scours on a delta plain revealed by interferometric sonar. Geomorphology 94, 58–68.

Fanetti, D., Anselmetti, F.S., Chapron, E., Sturm, M. & Vezzoli, L. 2008: Megaturbidite deposits in the Holocene basin fill of Lake Como (Southern Alps, Italy). Palaeogeography, Palaeoclimatology, Palaeoecology 259, 323–340.

Fassnacht, S.R. & Conly, F.M. 2000: Persistence of a scour hole on the East Channel of the Mackenzie Delta, N.W.T. Canadian Journal of Civil Engineering 27, 798–804.

Fornari, D.J.M.R., Perfit, J.F., Allan, R., Batiza, R., Haymon, A., Barone, W.B.F., Ryan, T., Smith, T., Simkia, T. & Luckman, M.A. 1988: Geochemical and structural studies of the Lamont seamounts: Seamounts as indicators of mantle processes. Earth and Planetary Science Letters 89, 63–83.

Harbor, D.J. 1998: Dynamics of bedforms in the lower Mississippi River. Journal of Sedimentary Research 68, 750–762.

Harms, J.C., Southard, J.B. & Walker, R.G. 1982: Structures and sequences in clastic rocks, Lecture Notes Short Course 9. Society of Economic Paleontologists and Mineralogists, Tulsa, Oklahoma, 249 pp.

Hiller, T.M. & Lewis, K. 2004: Getting the most out of high resolution wide swath sonar data. Proceedings of the 14th International Symposium of the Hydrograph Society, Plymouth, UK, Paper 8, 8 pp.

Kostaschuk, R.A & Best, J. 2005: Response of sand dunes to variations in tidal flow: Fraser Estuary, Canada. Journal of Geophysical Research 110, F04S04, doi:10.1029/2004JF000176.

Laberg, J.S., Eilertsen, R.S., Salomonsen, G.R. & Vorren, T.O. 2007: Submarine push moraine formation during the early Fennoscandian Ice Sheet deglaciation. Quaternary Research 67, 453–462.

Lane, S.N., Biron, P.M., Bradbrook, K.F., Butler, J.B., Chandler, J.H., Crowell, M.D., McLelland, S.J., Richards, K.S. & Roy, A.G. 1998: Three-dimensional measurement of river channel flow processes using acoustic Doppler velocimetry. Earth Surface Processes and Landforms 21, 1247–1267.

Ledoux, G., Lajeunesse, P., Chapron, E. & St-Onge, G. 2010: Multibeam bathymetry investigations of mass movements in Lake Le Bourget (NW Alps, France) using a portable platform. In Mosher, D.C., Shipp, R.C., Moscardelli, L., Chaytor, J.D., Baxter, C.D.P., Lee, H.J. & Urgeles, R. (eds.): Submarine mass movements and their consequences, 4th International Symposium. Springer, Dordrecht, pp.

423–434.

35

Lévy, S., Jaboyedoff, M., Locat, J. & Demers, D. 2012: Erosion and channel change as factors of landslides and valley formation in Champlain Sea Clays: The Chacoura River, Quebec, Canada.

Geomorphology 145–146, 12–18.

L’Heureux, J.-S., Hansen, L. & Longva, O. 2009: Development of the submarine channel in front of the Nidelva River, Trondheimsfjorden, Norway. Marine Geology 260, 30–44.

Longva, O. 1991: Fet, Quaternary geology map 1914 I, scale 1:50,000, Geological Survey of Norway.

Maddux, T.B., McLean, S.R. & Nelson, J.M. 2003: Turbulent flow over three-dimensional dunes: 2. Fluid and bed stresses. Journal of Geophysical Research (F1), 6009, doi: 10.1029/2003JF000017.

McAdoo, B.G., Pratson, L.F. & Orange, D.L. 2000: Submarine landslide geomorphology, US continental slope. Marine Geology 169, 103–136.

Moore, B.R. 1970: Scour and fill processes in a deep river hole, Ohio River, Louisville, Kentucky. Journal of Sedimentary Research 40, 449–456.

Mosley, M.P. 1976: An experimental study of channel confluences.

Journal of Geology 84, 535–562.

Nittrouer, J.A., Mohrig, D., Allison, M.A. & Peyret, A.-P.B. 2011: The lowermost Mississippi River: a mixed bedrock-alluvial channel.

Sedimentology 58, 1914–1934.

Ottesen, D., Dowdeswell, J.A., Benn, D.I., Kristensen, L., Christiansen, H.H., Christensen, O., Hansen, L., Lebesbye, E., Forwick, M. & Vorren, T.O. 2008: Submarine landforms characteristic of glacier surges in two Spitzbergen fjords. Quaternary Science Reviews 27, 1583–1599.

Parsons, D.R., Best, J.L., Orfeo, O., Hardy, R.J., Kostaschuk, R. & Lane, S.N. 2005: Morphology and flow fields of three-dimensional dunes, Rio Paranà, Argentina: Results from simultaneous multibeam echo sounding and acoustic Doppler current profiling. Journal of Geophysical Research 110, doi: 10.1029/2004JF000231.

Pedersen, L. 1981: Glommas delta i Øyeren. En fluvialgeomorfologisk studie med en oversikt over de siste 200-års utvikling. MSc thesis, University of Oslo, 107 pp.

Roden, J.E. 1998: The sedimentology and dynamics of mega-dunes, Jamuna River, Bangladesh. PhD thesis, University of Leeds, 310 pp.

Rubin, D.M. 1987: Concepts in Sedimentology and Paleontology, 1, Cross-bedding, bedforms and paleocurrents. Society of Economic Paleontologists and Mineralogists, Tulsa, Oklahoma, 187 pp.

Sastre, V., Loizeau, J.-L., Greinert, J., Naudts, L., Arpagaus, P., Anselmetti, F. & Wildi, W. 2010: Morphology and recent history of the Rhone River Delta in Lake Geneva (Switzerland). Swiss Journal of Geosciences 103, 33–42.

Suárez, J. 1996: Erosion induced landslides in tropical environments.

In Senneset, K. (ed.): Landslides, Proceedings 7th International Symposium on Landslides, Trondheim, 17–21 June, 2, pp. 1115–1119. ten Brinke, W.B.M., Wilbers, A.W.E. & Wesseling, C. 1999: Dune growth, decay and migration rates during a large-magnitude flood at a sand and mixed sand-gravel bed in the Dutch Rhine river system. In Smith, N.D. & Rogers, J. (eds.): Fluvial Sedimentology VI, Special Publication International Association of Sedimentologists 28, pp. 15–32.

Van de Graff, W.J.E. & Ealey, P.J. 1989: Geological modelling for simulation studies. American Association of Petroleum Geologists Bulletin 73, 1436–1444.

Venditti, J.G. 2003: Initiation and development of sand dunes in river channels. PhD thesis, University of British Columbia, 291 pp.

Venditti, J.G., Church, M. & Bennett, S.J. 2005: On the transition between 2D and 3D dunes. Sedimentology 52, 1343–1359.

Weber, K.J. 1980: Influence on fluid flow of common sedimentary structures in sand bodies. Society of Petroleum Engineers Paper 9247.

Weber, K.J. 1986: How heterogeneity affects oil recovery. In Lake, L.W. & Carrol Jr., H.B. (eds.): Reservoir Characterization, Elsevier, New York, pp. 487–541.

Zinke, P., Olsen, N.R.B., Bogen, J. & Rüther, N. 2010: 3D modeling of the flow distribution in the delta of Lake Øyeren, Norway.

Hydrology Research 41, 92–103.

36

NORWEGIAN JOURNAL OF GEOLOGY

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

37

Late post-impact sedimentation in the Ritland impact structure, western Norway Abdus Samad Azad, Henning Dypvik, Fridtjof Riis & Elin Kalleson Azad, A.S., Dypvik, H., Riis, F. & Kalleson, E.: Late post-impact sedimentation in the Ritland impact structure, western Norway. Norwegian Journal of Geology, Vol 93, pp. 37–59. Trondheim 2013, ISSN 029-196X. The crater-infilling successions of the 2.7 km-diameter, Ritland impact structure are classified as: (A) late syn-impact, (B) early post-impact and (C) late post-impact sediments. The late post-impact sediments represent later stages of crater sedimentation during stable crater conditions. The transition between the early post-impact and late post-impact crater sedimentation is marked by a ~6 m-thick succession exposed in the southcentral part of the crater. The lowermost part of this succession consists of fine- to medium-grained sandstone, deposited from turbidity flows during the retreating or abandonment stage of earlier prograding submarine fans. Fine sandstones intercalated with silty shales in the middle part represent alternating episodes of turbiditic and suspension deposition. Upward transition of this facies into thick, dark-grey to black shales suggests establishment of anoxic to hypoxic bottom-water conditions where sediments were deposited during an extensive period of suspension deposition.

Local coarse clastics exposed in the easternmost crater wall represent small-scale scree deposits, suggesting that submarine slides and reworking of the sediments were active processes for a long time after the impact.

Abdus Samad Azad, Henning Dypvik, Department of Geosciences, University of Oslo, PO Box 1047 Blindern, 0316 Oslo, Norway. Fridtjof Riis, Norwegian­Petroleum Directorate, PO BOX 600, 4003 Stavanger, Norway. Elin Kalleson, Fyrstikkalleen skole, PO BOX 6660 Etterstad, 0609, Oslo, Norway.

E-mail corresponding author (Abdus Samad Azad): m.a.s.azad@geo.uio.no

Introduction Crater-infilling sediments of the Cambrian Ritland impact structure of southwestern Norway have been grouped into three broad units: (A) late syn-impact, (B) early post-impact and (C) late post-impact (Azad et al., 2012). The late syn-impact and early post-impact sediments were discussed in Azad et al. (2012). The present paper deals with the later stages of crater sedimentation, representing a shift from the initial dramatic episodes of crater sedimentation (during the late syn- to early postimpact stages) to relatively stable sedimentation during the late post-impact stage. ‘Late syn-impact’ sedimentation (Dypvik & Kalleson, 2010) within impact craters starts by collapse of the transient cavity and slumping back into the crater by different gravity and massflow processes. In an idealised simple crater model, these could include rock avalanches, debris avalanches or debris flows, depending on the physical properties of the target rock, dry or wet target surface (e.g., subaerial, subaqueous/shallow marine, deep marine), crater size, degree of slope failure, and temporal fluidisation of the sediments (Melosh, 1989). The ‘early post-impact’ crater sedimentation is more erosion dominated and occurred under water-saturated conditions in the case of marine craters (Dypvik & Kalleson, 2010). Sediments are derived from the crater walls and rim and deposited by

gravity-controlled processes, e.g., debris flows, turbidity flows (for more details see Dypvik & Kalleson, 2010 and Azad et al., 2012). The late post-impact sedimentation is suspension dominated, consists mostly of fine-grained sediments and commonly represents part of a widespread regional succession. The sedimentation of the late post-impact crater-filling deposits has major importance as it preserves the record of the transition between impact-related (late syn-impact and early post-impact) crater sedimentation and late post-impact geological history.

The Cambro–Ordovician successions in the Norwegian Caledonides were largely affected by thrust nappes directed to the southeast and east (Rey et al., 1997).

Within the Caledonian Orogen, only a few thin successions are preserved autochthonously between the Precambrian peneplain and the Caledonian nappes (Bergström­& Gee, 1985). The crater-filling sediments of the Ritland impact structure were deeply buried below the lowermost Caledonian nappe, and were scarcely affected by later thrusting. These well-preserved sedimentary successions were later exposed by tectonic activity and glacial erosion, providing opportunities to study the autochthonous Cambrian successions. The main object­ive of this study is to interpret the depositional mechanisms, sedimentary environments and possible sources of the late post-impact crater infills of the

38

A.S. Azad et al

Ritland impact structure. The study focuses in particular on the distribution, texture, mineralogical compo­sition, structure and geometry of the late post-impact craterfilling sediments.

Geological setting The Ritland structure is a 2.7 km diameter and 350 m deep, recently confirmed, simple impact structure, probably formed during the Early to Middle Cambrian period (Riis et al., 2011). At the present day, the structure shows remnants of an originally circular depression located in the mountainous terrain of Hjelmeland municipality, Rogaland, southwestern Norway (Fig. 1A). During the Cambrian, the granitic/gneissic sub-Cambrian peneplain was covered by a shallow but extensive epicontinental sea in which a thin unit of unconsolidated marine clays (10– 20 m thick) accumulated, forming the target surface of the Ritland bolide impact. The water depth of the existing epicontinental sea was probably ≤100 m, assumed to be close to the size of the bolide (~115 m) (Azad et al., 2012; Shuvalov et al., 2012). Today, the sub-Cambrian peneplain is well exposed in the southeastern part of the crater, representing the remnants of a wide, flat, slightly undulating surface where the crater forms a topographic depression (Riis et al., 2011). The crater depression was initially filled by impact-derived coarser clastic material, e.g., breccias, conglomerates (Fig. 1B), mainly related to rock avalanches, while debris-flow and concentrated density-flow deposits dominated over rock avalanches during the early post-impact stage (Azad et al., 2012).

Turbidity-current deposition dominated afterwards, when the crater was completely submerged by seawater.

Different gravity-controlled sedimentary processes were also active, especially along the crater margins in the early post-impact stage (Azad et al., 2012). During the late post-impact stage, dark-grey to black marine clays were deposited (Fig. 1B). The Cambro–Ordovician sandstone covering the marine shales (Fig. 1B) represents a shallowing of the existing epicontinental sea (Knaust, 2004). At that time, the stable crater depression was apparently completely filled with sediments (Riis et al, 2011). The crater infills were subsequently deeply buried by the Caledonian thrust nappes (Fig. 1B) during the development of the Caledonian Orogen in Mid Silurian to Early Devonian time (Gee et al., 2008). Succeeding tectonic episodes during Late Palaeozoic to Cenozoic time, and finally several glacial episodes in the Quaternary period, helped to erode and expose the present-day impact structure.

Materials and methods Detailed field investigations were carried out to map the late post-impact sediments exposed in the Svodene–Ritlandsfjellet and Dormålsknuten areas (Fig. 2). Sedimentary logs from both these sections

NORWEGIAN JOURNAL OF GEOLOGY

were prepared to gain a better understanding of the lithofacies distribution. The lithofacies distribution map (Fig. 2) is based on a high-resolution, aerial-photo map with a contour interval of 5 m. Two geological profiles, from the crater centre to Svodene–Ritlandsfjellet (southwestern crater wall) (Fig. 3) and Dormålsknuten (northeastern crater wall) (Fig. 4), have been constructed to define the stratigraphical relationships of the late post-impact sediments to the late syn-impact and early post-impact sediments. Stratigraphical correlation has been made based on three logged sections to clarify our understanding of the Sandstone Unit–2 (Figs. 5, 6), which represents a transition in crater sedimentation (discussed in this paper) from early post-impact to the late post-impact stage. Another log has been prepared to plot the late post-impact clastics exposed in the Dormålsknuten area (Figs. 7, 8). Gamma measurements were made in the field using the portable RadiogramTM 4000 device. Natural gamma-ray activity was measured in counts per second (cps) for almost every individual bed within the succession to record significant variations in the radioactivity that would indicate any changes in the lithological continuity especially for the shale beds.

The counter was placed in direct contact with the fresh bedding surfaces and kept there for at least 10 seconds to obtain optimal values for the natural gamma activity.

Comparisons of the detailed sedimentological (Fig. 9) and mineralogical characteristics (Figs. 10, 11) between Sandstone Unit–1 (early post-impact) and Sandstone Unit–2 were made in order to establish the transition in depositional processes and sedimentary environment from the early post-impact to the late post-impact stage.

The observations are presented in Table 1. A total of 37 thin sections have been examined under optical and scanning-electron microscopes. Modal compositions (400 grain counts) were determined for 29 of these (Table 2). Semiquantitative determination of mineralogical composition for 17 bulk samples is based on peak height (Morris et al., 2008) of data from XRD (X-ray diffraction) analysis, using a Philips X’Pert MPD, at the University of Oslo. The results are presented in Table 3. Quartz, K-feldspar, plagioclase, calcite, dolomite, mica, chlorite and pyrite were quantified with respective peak values of (d = 4.26 Å), (d = 3.24 Å, 3.25 Å), (d = 3.18 Å, 3.19 Å), (d = 3.03 Å), (d = 2.89 Å), (d = 10 Å), (d = 7 Å), (d = 2.71 Å). The Wentworth (1922) grain-size scale was used for grain-size classification. Folk’s (1974) triangular diagram has been used for the classification of sedimentary rocks, e.g., conglomerate, conglomeratic sandstone, sandstone, etc. Based on field observations and mineralogical analyses, depositional models for the late-impact crater sedimentation are presented in Figs. 12 and 13.

Field observations The late post-impact sediments of the Ritland impact structure have primarily been studied in two different depositional areas representing two different

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

39

Figure 1. (A) Location map of the Ritland impact structure. (B) Geological map of the Ritland impact structure based on topographic maps with rivers, lakes (blue), roads (black) and 20 m-interval contour lines (brown), showing the general outline of the crater-filling sedimentary succession and major lithological units inside and outside the crater (slightly modified from Riis et al., 2011); the bold red line indicates the area shown in Fig. 2.

stratigraphical levels. The sediments exposed in the south-central part (Svodene–Ritlandsfjellet area) (Fig. 2) represent the transition in crater sedimentation from the early post-impact to the late post-impact stage, showing an overall fining-upward relation and indicating a gradual change from high-energy to low-energy sedimentation. The sediments exposed higher up along the northeastern crater wall (Dormålsknuten area) (Fig. 2) are interbedded with dark-grey to black shales. The

Dormålsknuten section is located at higher elevations and stratigraphically younger compared to the Svodene– Ritlandsfjellet section, consisting of coarse clastics intercalated with dark-grey to black shales.

Svodene–Ritlandsfjellet area The late syn-impact and early post-impact sedimentary succession in the Svodene–Ritlandsfjellet area was grouped

40

A.S. Azad et al

into three general stratigraphical units: (i) Lower Breccia Unit, (ii) Sandstone Unit and (iii) Upper Breccia Unit, described in detail by Azad et al. (2012). The sedimentary succession described in the present paper lies stratigraphically above the early post-impact Sandstone Unit, displaying a gradual upward transition into dark-grey to black shales in the central part of the crater (Figs. 2, 3). In this study, the early post-impact Sandstone Unit is renamed Sandstone Unit–1 and the overlying transitional sandstone succession is named Sandstone Unit–2 for clarity (Figs. 2, 3).

Sandstone Unit–2 has been logged at three different locations extending from south to north over a distance of about 100 m (Fig. 2). The beds dip with an angle of 4–7° towards the crater centre and represent a more basincentral location compared to Sandstone Unit–1 (Figs.

2, 3). Presently, the uppermost part of this succession is exposed 450 m above sea level, approximately 270 m below the reference level of the peneplain and stratigraphically located in the field on top of Sandstone Unit–1 and the Upper Breccia Unit (Fig. 3). However, the direct contact between Sandstone Unit–1 and Sandstone Unit–2 has not been found exposed in the field.

The thickness of Sandstone Unit–2 is ~6 m, consisting dominantly of sandstones, silty shales and shales. The succession shows a gradational transition from fine- to medium-grained sandstone at the base to alternating layers of fine sand and silty shale in the middle, and darkgrey to black shales towards the top (Fig. 5). Within the succession it was possible to make just a few (four) palaeocurrent-direction measurements from the faint ripple marks. The ripple marks strike within the range 55–100°, indicating a northeastward flow direction (Fig. 2). Sandstone Unit–2 can be further subdivided into three facies types: (i) fine to medium sandstone, (ii) fine sandstone and iii) dark-grey to black shale (Fig. 5).

Fine to medium sandstone These sandstones are light grey in colour, ~3 m thick, moderately sorted and planar bedded (Figs. 5, 6A). The thickness of the individual beds varies from 10 to 30 cm. The beds are commonly capped by thin silt laminae (10–20 mm thick) and show irregular or undulating erosive boundaries to the underlying beds (Fig. 6A).

In a few places, ripple marks have been observed on the bedding-­plane surfaces. The rippled beds are a few centi­ metres in thickness, consisting of silty sands with a typical ­ripple wavelength of 4 cm and height of 1 cm (Fig. 6B). In the lower part of the succession, a few beds display a ­concave-­up relationship to the underlying beds, probably representing small scale (0.5 m) channels. The individual beds show internal fining-upward features and the overall development of this succession is homogeneous (Fig. 5). Gamma activity measurements in this succession show higher values (350–415 cps) compared to the overlying fine-sandstone and silty-shale facies (255–310 cps) (Fig. 5). A subrounded granitic clast, 40 cm dia­meter, has been observed embedded within the parallel-­bedded sandstones, with associated conformable

NORWEGIAN JOURNAL OF GEOLOGY

deformation of both overlying and underlying strata (Fig. 6C). Small-scale convolute laminations are seen in places within these sandstones. An increase in clay content in the sandstones has been recorded in the more distal part (northern part) of the succession. Sedimentary pyrites (dia­genetic, 0.5 cm) and mud clasts (subrounded, 2 mm) have also been found. Cruziana ichnofacies, possibly planolites or thalassinoides (N.M. Hanken, pers. comm., 2012), have been observed within a few sandstone beds (Fig. 6D). A similar ichnofacies within these sediments has also been reported by Bruton et al. (1989).

The traces are mostly horizontal, undulating, unbranched and convex­in nature. Commonly, the length of the trace fossil varies from 10 to 25 cm and the width is 5–10 mm, and the track filled with material similar to that of the host rock. The trace fossils are found in silty sandstone, mostly on the bedding-plane surfaces (Fig. 6D).

Fine sandstone The middle part of Sandstone Unit–2 is ~1.5 m thick and consists of alternating layers of dark-grey, parallelbedded, fine sandstones and thinly laminated silty shales (Fig. 5). The thickness of individual fine-grained sandstone beds varies from 2 to 4 cm (Fig. 5). Some mediumto coarse-grained sandstone beds were also found, especially in the upper part of the succession (Fig. 5). The silty shale beds vary in thickness from 0.5 to 2 cm and show an upward increase in thickness (Fig. 5). The boundaries between individual beds are irregular, undulating and locally rippled. Individual beds show a fining upwards and the overall development is irregular/alternating (Fig. 5). Traces of Cruziana ichnofacies have also been found within the silty shale layers.

Dark-grey to black shale The dark-grey to black shales overlie the fine-grained sandstone facies (Fig. 5). In the south-central part of the crater, the exposed thickness of this unit within the logged profile is only a metre (Fig. 5). In the central part of the crater, the dark-grey to black shale attains a maximum thickness of about 180 m (Riis et al., 2011) (Figs.

2, 3, 4). This shale has been referred to as bituminous shale by Knaust (2004) and contains early Middle Cambrian faunas of trilobites, brachiopods, hyoliths, sponge spicules and various problematica (Henningsmoen, 1952; Bruton et al., 1989). In the upper part of this shale unit, a bed of limestone concretions has been found containing an abundant inarticulate brachiopod fauna (Bruton & Harper, 2000). Based on the fossil content, Bruton & Harper (2000) suggested a correlation of this shale unit with the upper Middle Cambrian Limestone of southern Sweden. These shales are rich in organic matter, flaky, thinly laminated and commonly found with silty partings towards the overlying beds. They display a gradational transition from the silty shale at the base to darkgrey to black shale towards the top. The gamma readings in the lower part of the shale unit in the logged profile are low (215–270 cps) (Fig. 5).

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

41

Dormålsknuten conglomerate beds interfingering in Cambrian Shale

Figure 2. Lithofacies distribution map of the Ritland impact structure (modified from Azad et al., 2012). The Sandstone Unit–2 exposed in the RITF (Ritlandsfjellet) and SVD (Svodene) areas and the conglomerate beds in the DOR (Dormålsknuten) area are shown in two index figures.

Other locations on the map are STH (Stemhaugen), BBH (Bjødnabuhaugen) and BDB (Bjødnabu). Rose diagram showing the palaeocurrent flow directions measured in Sandstone Unit–2. Two cross sections along X–Y and X–Z are shown in Figs. 3 and 4, respectively. A stratigraphical correlation (red dotted line, in index photo) of Sandstone Unit–2 is shown in Fig. 5.

42

A.S. Azad et al

NORWEGIAN JOURNAL OF GEOLOGY

Figure 3. Geological cross section (X–Y in Fig. 2) illustrating the stratigraphical relationship of Sandstone Unit–2 to other sedimentary units of the late syn-impact, early post-impact and late post-impact stages (Modified from Azad et al., 2012). The profile is drawn from the crater centre towards the Svodene–Ritlandsfjellet area. The base of the crater and the subsurface distribution of the different sedimentary units are drawn based on field observations and the numerical crater model of Shuvalov et al. (2012). The red arrow indicates the locations of the exposed sections of Sandstone Unit–2 within the profile.

Dormålsknuten area The Cambrian marine shales in the Dormålsknuten area show onlapping relations to the early post-impact breccias (Upper Breccia Unit) (Azad et al., 2012) in the northeastern crater wall (Fig. 7). The estimated thickness of this marine shale is ~180 m in the crater centre, thinning to almost zero along the crater margin (Fig. 4). At the present day, the top of the marine shale is located at higher elevations (720–760 m) in the northeastern crater wall compared to the crater centre (580 m), showing a concave-up geometric profile (Figs.

4, 7). The thickness of comparable marine shales found outside the crater (southeastern part) is about 35 m (Fig. 1B) (Kalleson et al., 2012). Knaust (2004) reported a thickness of this marine-shale unit outside the crater of about 20 m, possibly omitting the measurement of the 10–20 m-thick, pre-impact shales below the ejecta layer (Kalleson et al., 2012). Inside the crater, towards the top, this shale unit grades into silty shale and siltstone with numerous thin sandstone beds, containing abundant trace fossils (Knaust, 2004). This heterolithic rock grades further into a massive sandstone unit which can be mapped to a thickness of about 10 m inside the crater

(Figs. 4, 7). This Cambro–Ordovician sandstone has been interpreted as a ‘shoreface deposit’ by Knaust (2004), probably deposited when the crater was totally filled with sediments. The autochthonous/parautochthonous unit of Cambrian sedimentary rocks is covered by a number of thrust sheets that involved the uppermost part of the crater-infill succession (Figs. 4, 7).

The contact of the Upper Breccia Unit (early postimpact) towards the brecciated basement is exposed extensively at higher elevations along the northeastern crater wall (Fig. 7). Deeper down, adjacent to this crater wall, the basement contact with the sedimentary breccias has not been found. The field observations and the modelled crater margin (Shuvalov et al., 2012) suggest that a thin veneer of early post-impact breccia (Upper Breccia Unit) stuck to the crater wall (Figs. 4, 7). The sedimentary succession in the Dormålsknuten area is characterised by clast-supported conglomerate beds interbedded with dark-grey to black shales (Figs. 7, 8A).

The thickness of the lowermost clast-supported conglomerate bed (bed–1) is about 1 m (Figs. 7, 8B).

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

43

Figure 4. Geological cross section (X–Z in Fig. 2) illustrating the strati­ graphical relationship of the Dormålsknuten clast-supported conglomerate beds (drawn in orange, the thickness and extent of the beds are not to scale) to other sedimentary units of the late syn-impact, early post-impact and late post-impact stages. The profile is drawn from the crater centre towards the Dormålsknuten area. The base of the crater and the subsurface distribution of the different sedimentary units are drawn based on field observations and the numerical crater model of Shuvalov et al. (2012).

It appears as a lens within black shale with a lateral extension of about 10 m (Fig. 8A). The bed thins towards the crater centre and dips at an angle of about 20°. The bed is poorly developed and shows an overall increase in the size of the clasts towards the top (Fig. 8B). The clasts are subangular to subrounded and similar in composition to the granitic/gneissic basement (Azad et al., 2012) (Fig. 8B), and range in size mostly from 5 to 15 cm; some larger clasts of 25–30 cm have also been found (Fig. 8B). The matrix content in the bed is fairly high (40–50%), consisting of a mixture of black clay and medium- to coarse-grained sand. The matrix is light grey to dark grey in colour, poorly sorted and has a similar composition to that of the clasts. In a few places, the matrix occurs as centimetre-thick irregular bands within this conglomerate bed (Fig. 8B). The overlying shale beds are deformed around the clasts of the conglomerate bed (Fig. 8B). Towards the top of the bed the larger clasts show possible imbrication or a tendency towards bedding-parallel alignments (Fig. 8B). The maximum clast size vs. bed thickness ratio for this bed is 0.2.

The clast-supported conglomerate bed in the middle

(bed–2) is about 1.5 m thick, with a lateral extent of 15 m (Figs. 7, 8A), and also occurs as a lens in the darkgrey shale (Fig. 8A). The bed is poorly developed and the clasts display a slight increase in size in the upper part of the bed. Clast size typically ranges from 15 to 20 cm, with a few clasts up to 50 cm. The clasts and matrix have similar compositions to those in the underlying conglomerate bed (bed–1). This conglomerate bed (bed– 2) also thickens towards the crater wall and thins to the crater centre. The maximum clast size vs. bed thickness ratio for this bed shows a higher value (0.4) than in the underlying bed (bed–1).

The uppermost, clast-supported, conglomerate bed (bed–3) is approximately 1 m thick and consists of clasts and matrix of a similar composition to those in bed–1 and bed–2, and occurs as a convex-up lens within shales. Detailed clast-size measurements of this bed were not made due to poor accessibility. Observation from a distance indicates that the average clast size is 15–20 cm; a few outsize clasts exceed 70 cm. The overall clast size increases upwards through the bed. Matrix content is fairly high (30–40%). The matrix shows centimetre-thick

44

A.S. Azad et al

NORWEGIAN JOURNAL OF GEOLOGY

Figure 5. Stratigraphical correlation of Sandstone Unit–2 exposed in the Svodene–Ritlandsfjellet area (see Fig. 2 for geographical locations of the logged sections). The gamma-activity measurements (counts per second, cps) for individual beds are plotted next to the lithofacies column.

The rock sample numbers are indicated to the left.

faint bands in a few places within the conglomerate. No obvious alignment or imbrication of the clasts was found.

Two outsize clasts up to ~1.8 m in diameter have also been observed embedded within the dark-grey shale beds (at 5.5–6 m in the log) (Figs. 7, 8C). The clasts are granitic in composition, heavily fractured and subrounded to rounded in shape (Fig. 8C). The overlying shale layers show brittle deformation and a lamination bending around the clasts (Fig. 8C). Thin veneers of shale surrounding the clasts, marked in Fig. 8C, probably relate to the compaction of the unconsolidated clays during burial. A bed (2–4 cm thick) consisting of pebblesize granitic/gneissic clasts and very coarse-grained sand has been found within the thinly laminated, dark-grey to black shale beds (at 10.6 m in the log) (Figs. 7, 8D).

The sandstone is light grey in colour, very poorly sorted

and contains a black clayey matrix, and has a slightly erosional base to the underlying shale beds (Fig. 8D).

Comparison between Sandstone Unit–1 and Sandstone Unit–2 The sedimentological and mineralogical characteristics of Sandstone Unit–1 have been compared with those of Sandstone Unit–2 (Table 1) to decipher whether these sediments represent any clear transition in crater sedimentation. Sediments of Sandstone Unit–2 are texturally more mature compared to Sandstone Unit–1 (Table 1). The beds in Sandstone Unit–2 are parallel, well developed and commonly separated by

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

45

Figure 6. Sedimentary features of the fine- to medium-grained sandstone facies of Sandstone Unit–2. (A) Parallel beds, commonly capped by thin silt laminae, showing irregular or undulating erosional boundaries to the underlying beds. (B) Faint ripple marks on the bedding planes are locally found (this piece was removed and placed vertically for better visualisation). (C) Granitic clast (40 cm in diameter) found embedded in the parallel-bedded sandstones; note the conformable deformation of the underlying and overlying beds. (D) Horizontal, undulating, unbranched and convex trace fossils (planolites) found within the lower sandstone beds.

10–20 mm silt partings (Figs. 6A, 9B), in contrast to the more massive and relatively poorly developed beds in Sandstone Unit–1 (Fig. 9A). The contacts between the beds in Sandstone Unit–2 are mostly erosional and undulating (Fig. 6A), whereas gradational contacts are commonly found between the beds in Sandstone Unit– 1. Both the individual beds and the overall succession of Sandstone Unit–2 show a gradual fining-upward trend accompanied by an upward decrease in bed thickness (Figs. 5, 9B). Sandstone Unit–1 is conglomeratic in its lower and upper parts and mostly homogeneous in the middle (Fig. 9A). Clasts in the conglomeratic beds in Sandstone Unit–1 are subangular to subrounded, granitic/gneissic in composition and vary in size from a centimetre up to a metre (Azad et al., 2012). Sandstone Unit–2 is mostly free of clasts, although one subrounded granitic clast (40 cm) has been recorded (Fig. 6C).

Sedimentary structures are comparatively poorly developed in both units. Very faint ripple marks occur in the parallel-bedded sandstones of Sandstone Unit–1 (see log, Fig. 9A), and some deformational structures are found in the homogeneous sandstone beds in the

upper part of the unit (Azad et al., 2012). Sedimentary structures, especi­ally ripple marks, are better developed in Sandstone Unit–2 (Fig. 6B). The ripple marks are mostly symmetrical and found in the finer-grained, silty sediments. Sandstone Unit–1 contains several 10–20 cm-thick (c. 2 m total thickness), calcite-cemented, low-angle cross-stratified beds (Fig. 9A) displaying erosional relationships to the underlying beds (Azad et al., 2012). The Sandstone Unit–2 beds are parallel and do not contain any calcite cement. Small-scale convolute laminations have been found within the silty shale layers of Sandstone Unit–2. Crenulated mud clasts (1–3 mm) occur in places within Sandstone Unit–1, and subrounded mud clasts (2 mm) are found in Sandstone Unit–2. Diagenetic pyrite has been found locally in Sandstone Unit–2, but only as an accessory phase in Sandstone Unit–1. Evidence of pressure solution has been found in Sandstone Unit–2, whereas no such compactional features have been observed in Sandstone Unit–1. Palaeocurrent flow measurements noted from, e.g., cross beds and ripple marks in both units show similar northeastward transport directions (Table 1).

Trace fossils have not been observed in Sandstone

46

A.S. Azad et al

Unit–1, whereas Sandstone Unit–2 contains a few such tracks, especially on the bedding-plane surfaces.

Sandstone Unit -2 sediments are generally quartz rich and feldspar poor, contrasting with the feldspar-rich and quartz-poor Sandstone Unit–1 (Table 1). Both the clast content (millimetre-sized clasts) and amount of matrix in Sandstone Unit–2 are lower as compared with the Sandstone Unit–1 (Table 1).

Mineralogical study Sandstone Unit–2 The Sandstone Unit–2 sediments are mostly fine- to medium-grained, locally with coarse-grained sand and granitic/gneissic clasts (up to 4 mm). The grains are mostly subangular to subrounded (a few rounded), mode­ rately to well sorted, tightly packed and show evidence of pressure solution. The grains display an upward-­fining trend in a few thin sections. Mineralogical compositions of three different facies types of Sandstone Unit–2 are described below.

NORWEGIAN JOURNAL OF GEOLOGY

The average quartz content of the fine- to mediumgrained sandstone facies is 42% (Table 2). Feldspar is the second most dominant mineral (28%) while mica (mostly biotite) and rock fragments (granitic and gneissic) make up the other primary constituents (Table 2). The XRD analyses show a higher content of plagioclase compared to K-­feldspar (Table 3). Sericite forms the major authigenic phases (Table 2). Authigenic pyrite is common­ly found, both as a grain-replacing (Fig. 10A) and a pore-filling phase. Titanite, phosphate (e.g., apatite, jarosite) and iron oxides (e.g., hematite) are other minor authigenic phases present as pore fillings and locally filling microfractures in grains. Discrete particles of apatite and titanite have also been found (Fig. 10B). Illite is another important authigenic mineral, forming thin coatings around quartz and feldspar grains (Fig. 10C). Organic matter (rounded, granular masses) and detrital heavy minerals, e.g., zircon, garnet, epidote and rutile, are other accessory phases.

The fine-sandstone facies has a slightly higher quartz content (44%) compared to the underlying facies (Table 2). However, the feldspar content of this facies is surprisingly low (3%) (Table 2). This low feldspar content reflects

Figure 7. The late post-impact Cambrian marine shale (green) showing an onlap relationship (indicated by yellow arrows) to the early post-impact breccias (Upper Breccia Unit) (shown in light brick red) in the Dormålsknuten area. The lower contact of the Upper Breccia Unit to the brecciated basement (light grey) has been found in the upper part of the crater wall. The Cambro–Ordovician sandstones overlying the Cambrian shale are shown in light green and the thrust sheets are in pink. The Dormålsknuten conglomerate beds (orange) interfinger with the Cambrian marine shale. To the right, a log profile of the Dormålsknuten conglomerate beds shows the intercalating conglomerate and marine shale. Rock sample numbers are shown next to the lithological column. The photo in the inset represents an uninterpreted section.

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

47

Figure 8. Dormålsknuten conglomerate beds in Cambrian shale. (A) The clast-supported conglomerate beds intercalated with dark-grey to black shale. (B) Clasts show an overall upward increase in size, with some outsize clasts, and a coarse-grained sandy matrix forming thin bands (shown by pink hatched line); photo from the lowermost conglomerate bed. (C) An outsize fractured clast (~1.8 m, granitic in composition) found embedded within the dark-grey to black shales. Note the deformation (dashed line) of the shale layers surrounding the clast. (D) A thin (2–4 cm) bed (indicated by the pencil head) consisting of small granitic clasts and very coarse-grained sand occurring within the thinly laminated, dark-grey to black shale unit.

intense sericitisation (28%) (Table 2). XRD analyses reveal a further decrease in K-feldspar and a corresponding increase in plagioclase (Table 3). Grains of both quartz and feldspar show an upward increase in roundness compared with the underlying facies. Fine-grained, recrystallised quartz and dark-grey to black illitic clay form the major part of the matrix (Fig. 10D). Authigenic pyrite (framboidal) and iron oxides (e.g., hematite) are the other significant authigenic minerals appearing within these sediments. Among other authigenic minerals, titanite and phosphates have been found in greater amounts compared with the fine- to medium-grained sandstones. Organic carbon and zircon are the other accessory phases.

Thin, mostly parallel and locally wavy laminations were observed in thin sections of the silty-shale facies rocks.

The quartz and feldspar grains in the silty shales are angular to subangular, and show some a preferred orientation along the bedding planes. A few, angular, granitic clasts were also found within the silty shales. Fractures in the silty shale samples are filled with hematite. Finegrained framboidal pyrite and granular organic carbon represent other important accessories.

The quartz/feldspar ratio (Fig. 11A) in Sandstone Unit–2 shows an upward increase. A significant increase in the quartz/feldspar ratio has been noted in Sandstone Unit–2 compared to Sandstone Unit–1 (Fig. 11A). XRD data reveal a decrease in K-feldspar content and a corresponding increase in plagioclase feldspar in Sandstone Unit–2 compared to Sandstone Unit–1 (Fig. 11B). The mica content (mostly biotite) shows an upward decrease.

Among the authigenic minerals, Sandstone Unit–2 displays a significant increase in pyrite compared to Sandstone Unit–1 (Table 1). Sericite is the other authigenic phase commonly found in Sandstone Unit–2 sediments, showing an upward increase in the unit (Fig. 11C). Calcite and amphibole are absent in Sandstone Unit–2 sediments, but show a sporadic enrichment in Sandstone Unit–1. Increased amounts of apatite and titanite have also been noted in the sediments of Sandstone Unit–2.

Dormålsknuten conglomerate beds The clast-supported, Dormålsknuten conglomerate beds have a lower average quartz content (26%) compared to Sandstone Unit–2 (43%) (Table 2). The low feldspar content in the conglomerate beds is attributed to the

48

A.S. Azad et al

NORWEGIAN JOURNAL OF GEOLOGY

higher amounts of sericite (Table 2). Dark-grey to black, illitic clay is the major matrix component, constituting 20–45% of the total sample (Table 2). Organic material and traces of pyrite have been found (Tables 2, 3). The coarse-grained sandstone bed in the shale unit (at 10.6 m in the log) (Figs. 7, 8D) has higher quartz and lower feldspar contents compared with the clast-supported conglomerate beds (Table 2).

Discussion The late post-impact sediments in the Ritland impact structure have been studied sedimentologically and mineralogically. Sandstone Unit–2, exposed in the Svodene–Ritlandsfjellet area, consists of light-grey sandstones, dark-grey silty shales and black shales, representing a gradual upward transition from highenergy, oxygenated, crater-filling conditions to a lowenergy, calm, stable, anoxic to hypoxic depositional regime. Field observations suggest that the clastsupported, Dormålsknuten conglomerate beds were deposited at later stages of crater sedimentation. The occurrence of the conglomerates as lenses within the

dark-grey shale implies that mass-flow processes were intermittently active, interrupting long and quiet periods of clay deposition.

Svodene–Ritlandsfjellet area The lowermost part of Sandstone Unit–2 (fine- to medium-grained sandstone facies) (Fig. 5) indicates a clear transition in crater sedimentation as evidenced by a greater textural maturity and increased quartz content as compared with the underlying Sandstone Unit–1 (Fig. 11A; Table 1) (Azad et al., 2012). Fine- to mediumgrained sandstones (Fig. 6A) can be interpreted as turbidity deposits. The individual planar beds are marked by normal grading, erosional bases to the underlying beds, absence of floating clasts and a thin, silty sand capping; each bed represents a single turbidite flow event (Figs. 5, 6A). These turbiditic events may represent longer-duration, surge-like turbidity flows (Mulder & Alexander, 2001) as compared with the short-duration, surge-like turbidity flows of Sandstone Unit–1 (Azad et al., 2012). Relatively well-developed bedding and sedimentary structures (Fig. 9B) with upward-increasing bed thicknesses (Table 1) are suggestive of sustained

Table 1. Comparison of sedimentological and mineralogical characteristics of Sandstone Unit–1 and Sandstone Unit–2.

Sandstone Unit–1

Sandstone Unit–2

Texture and composition

Light-grey, fine- to medium-grained sand with common outsize clasts (cm size). Angular, subangular to subrounded grains. Moderate to poorly sorted.

Light-grey to dark-grey, fine- to medium-­grained sand with few clasts (mm size). Fining-upward trend. Subangular to subrounded grains.

Moderately sorted to well sorted.

Thickness of the succession Facies types

15–20 m

~6 m

1. Upper conglomeratic sandstone.

2. Middle homogeneous sandstone.

3. Parallel-bedded sandstones.

4. Lower conglomeratic sandstone.

1. Dark grey to black shale.

2. Fine sandstone.

3. Fine to medium sandstone.

Bed thickness

Parallel-bedded sandstones are 10–20 cm Fine to medium sandstones are 10–30 cm thick. Homogeneous and conglomeratic thick. Fine-sandstone beds vary from 2 to 4 sandstones vary from 1 to 1.5 m in thickness. cm in thickness. Silty shale beds vary within mm in thickness.

Bed type

Massive, less parallel, relatively poorly developed.

Parallel, well developed.

Bed contact

Graded mostly, the silty partings in the parallel-­bedded sandstones are 20–80 mm.

Erosive, undulating, locally rippled, silt part­ ings in fine to medium sandstone varies from 10 to 20 mm.

Conglomerate beds

Present in the upper and lower part.

Commonly sandstone and conglomeratic beds locally coexist.

Total absence of conglomeratic beds, gradual fining-upward trend of the succession.

Clasts

Subangular granitic clasts (0.5–1m) commonly found.

Subrounded granitic clast (40 cm) found only in one place.

Sedimentary structures

Very faint ripples, low-angle cross beds, local crenulated mud clasts, convolute laminations in sand due to loading of the overlying conglomeratic beds.

Relatively well-developed ripples on the bedding-plane surface, local convolute lamination in silty shales and rounded mud clasts, absence of cross bedding.

Palaeocurrent directions Trace fossils

40° to 110°, northeastward.

55° to 100°, northeastward.

Not found.

Horizontal, undulating, unbranched and convex.

Mineralogical composition (modal analysis, average)

Quartz: 24%, feldspar: 31%, mica: 9%, chlorite: 5%, calcite: 5%, rock fragments: 10%, matrix: 15%, heavy: 0.6%, sericite: 1%.

Other accessory: amphibole, siderite.

Quartz: 43%, feldspar: 16%, mica: 4%, chlorite: 0.3%, calcite: 0%, rock fragments: 8%, matrix: 6%, heavy: 0%, sericite: 21%. Other accessory: pyrite (0.9%), titanite, apatite, hematite.

Sedimentary features

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

49

Figure 9. Comparison of the sedimentary features between Sandstone Unit–1 and Unit–2. (A) Sedimentary log (modified from Azad et. al., 2012) of Sandstone Unit–1 (see legend in Fig. 5). The lower part of the succession shows a coarsening- and thickening-upward trend (indicated by the arrow), while the middle and upper parts are more homogeneous. The beds of this sandstone unit are massive, poorly developed, locally calcite cemented, cross stratified and ripple laminated, and contain subangular to angular clasts. (B) Sandstone Unit–2 is parallel-bedded, relatively well developed and individual beds show undulating and erosional contacts. The overall succession displays a fining- and thinning-upward trend. The overlying fine sand and silty shale beds are thin and occur as alternating layers of light to dark-grey fine sand and dark-grey silty shale.

flows, contrasting with the impact-generated surges or short-duration, surge-like turbidity flows (Mulder & Alexander, 2001) of Sandstone Unit–1 (Azad et al., 2012). Such short-duration, surge-like turbidity flows are commonly triggered by flow transformation through erosion and acceleration from flow types with higher concentrations of sediment. Longer-duration, surge-like turbidity flows are generated by slope failure, consisting of a waxing flow head and waning body and tail (Mulder & Alexander, 2001). Each individual, fining-upward bed marked with an erosional base (Fig. 6A) can be related to individual events of slope failure, renewed base-slope erosion and further reworking of the slope sediments.

Bouma Tb–d facies (Bouma, 1962) are relatively well developed in longer-duration, surge-like flows compared to the short-duration flows (Mulder & Alexander, 2001). However, the lowermost beds (fine- to mediumsandstone facies) of Sandstone Unit–2 do not reveal any complete develop­ments of Bouma Tb–d sequences. The base of a typical sandstone bed of this facies is rather homogeneous and grades upward into silty sands with centimetre-thick, locally rippled beds (Fig. 6B) and can be compared to the Bouma Tb–c facies. The upper,

fine-grained intervals of Bouma Td–e are evidently absent within these sandstones.

The boulder-size granitic clast (40 cm) found within these sandstones (Fig. 6C) is unlikely to have been transported by turbidity currents and has probably derived from a rock fall from the granitic basement of the crater­wall.

The late post-impact crater wall was much more stable compared with the early post-impact stage, when angular granitic/gneissic clasts were released more frequently (Azad et al., 2012). The increased roundness of the clast (Fig. 6C) suggests that the crater infills were subjected to further reworking. The modest deformation of the sandstone beds around the clast (Fig. 6C) indicates that the clast rested in place while the sands were deposit­ed at a slow, steady rate, eventually burying the clast. The occurrence of this isolated clast is unlike Sandstone Unit–1, where rapidly deposited, large, angular clasts are commonly found derived during the early post-impact stage of sedimentation (Azad et al., 2012). Comparable mineralogical compositions and palaeocurrent directions of Sandstone Unit–2 and Sandstone Unit–1 (Table 1) indicate that these sediments had a similar original source.

Dormåls­ knuten

Facies

41 46 37 39 47 54 40 49 53 51 42 38 43 40 43 45 43 44 39 44 46 42 37

SVDX–01–19

SVDX–01–18

SVDX–01–17

SVDX–01–16

SVDX–01–15

SVDX–01–14

SVDX–02–07

SVDX–02–06

SVDX–01–11

SVDX–01–10

SVDX–02–05

SVDX–02–04

SVDX–02–03

SVDX–02–01

SVDX–01–09

SVDX–01–08

SVDX–01–07

SVDX–01–06

SVDX–01–05

SVDX–01–04

SVDX–01–03

SVDX–01–02

SVDX–01–01

DOR–04–03

DOR–04–02

Clast congl.

23

29

37

35

SVDX–02–10

DOR–04–05

42

SVDX–02–12

Clast congl.

Coarse sand

30

12

11

8

35

25

23

30

28

33

30

35

37

21

26

33

12

8

4

6

3

7

4

2

1

2

0.5

1

2

1

Quartz Feldspar

SVDX–02–14

Sample no.

0

0

0

0

0

0

0

0.5

0.5

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

2

3

0

0

0

1

0

0.5

2

0

0

3

2

4

3

7

9

7.5

3

8

5

5

3

3

45

20

22

6

4.5

1.5

0

5

4

3

1

1

0

2

2

0

8

0

0

1

6

0

0

8

0

5

13.5

17.5

6

0.5

17

7

39

0

0

4

1

2.5

0

0

2

2.5

2

2

1

1

Mica1 Matrix

0

0

0

0

0

0

0

0

0

0

0

0

0

Carbon­ Chlorite ates

6

7

7

11

6

3

8

3.5

5

7

4

3

9

4

5

8

0

0

4

10

0

5

15

12.5

10

17.5

17

20

14.5

Rock fragment

0

0

0

1

4

0

2

0

1

2

0

1

0

1

0

0

0

0

1

2

1

1

1

1

0

0

0

1

2

0

0

0

0

0.5

0

0

0

0

0

0

0

1

0

0

0

4

0

12

3

2

0

0.5

2.5

3

4

1

1

1

Pyrite Hema­tite

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

12

29

24

6

13

16

7

16

9

6

10

9

24

18

13

38

30

41

22

30

32

38

28

25

30.5

34.5

24

24

11

2

1

0

0

0

0

0

0

0

0

0

0.5

0

3

1

0

0.5

2

2

3

1.5

0

1

1.5

0

0

3

2

0.5

AmphiOrganic Sericite bole matter

1

1.9

2.64

4.63

1.06

1.68

2.00

1.47

1.39

1.33

1.43

1.29

1.16

1.90

1.65

1.15

3.50

6.38

13.25

8.17

13.33

7.71

11.75

19.50

37.00

23.00

82.00

35.00

21.00

30.00

Qtz/ Felds.

Hematite, pyrite

Pyrite

Apatite, titanite, zircon, illitic clay

Apatite, titanite, zircon, garnet

Titanite, apatite, zircon, pyrite

Pyrite, organic

Pyrite, organic

Zonation of pyrite, titanite

Titanite, zircon, apatite

Titanite, apatite, zircon

Apatite, jarosite

Apatite, sphene

Illitic clay

Illitic clay, z ­ ircon, apatite

Trace (SEM and XRD)

A.S. Azad et al

’Mica’ (mostly biotite) represents detrital/primary mica and ‘sericite’ represents secondary/authigenic mica, which mostly derived from alteration of K-feldspar.

Late post-impact sediments

Svodene–Ritlandsfjellet

Sedimentary Depositional units area

Table 2. Mineralogical composition (vol.%, from thin-section study) of the different facies types of Sandstone Unit–2 from Svodene–Ritlandsfjellet and conglomerate beds from the Dormålsknuten area (400 grains were counted in each thin section to determine the percentages). The quartz/feldspar ratios of the individual samples are also presented in a separate column. Traces of some minerals studied from SEM/thin section have been added in the column at the far right.

Fine sandstone

Fine to medium sandstone

50 NORWEGIAN JOURNAL OF GEOLOGY

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

An increased quartz/feldspar ratio (Fig. 11A) and decrease in feldspar, rock fragments, mica and matrix content in the fine- to medium-sandstone facies of Sandstone Unit–2 compared to that found in Sandstone Unit–1 (Table 1) suggest a gradual development of further reworking and weathering. The presence of authigenic pyrite, apatite and organic matter and the absence of calcite (Table 2) within the fine- to medium-grained sandstone facies (Sandstone Unit–2) indicate that sporadic, reducing, bottom-water conditions were established within the crater during the transitional stages.

Cruziana ichnofacies (here planolites, thalassinoides) are typical for the offshore transitional zones but have also been recorded in subtidal, poorly sorted, unconsolidated substrates (Pemberton et al., 1992). This ichnofacies is also found in the littoral to sublittoral parts of some estuaries, bays, lagoons and tidal flats. It generally represents low- to moderate-energy conditions at the fair-weather wave base, but in some cases may also be found in deeper, quieter waters (Pemberton et al., 1992). The ichnofauna appears at the upper surface (in silty layers) of the fine- to medium-grained sandstone beds (Sandstone Unit–2) as mostly simple, horizontal

51

and cylind­rical, sediment-filled burrows. They possibly represent fair-weather wave-base conditions with fairly low sedimentation rates or even periods of nondeposition. The crater was assumed to have been completely filled by seawater during the early postimpact stage (Azad et al., 2012); the presence of the Cruziana ichnofauna suggests that the water depth within the crater was in the order of 300 m. The 350 m of crater depth, a modelled rim height of about 115 m and an existing seawater depth of ≤100 m (Shuvalov et al., 2012) suggest that the crater rim was probably exposed to the atmosphere during the transitional stages of crater­ sedimentation. The occurrences of trace fossils in silty layers suggest a short-lived, quiet event of suspension deposition before the arrival of new turbidity currents.

The low abundance and diversity of these trace fossils precludes the drawing of any major conclusion with regard to further changes in deposition­al conditions.

Alternating layers of dark-grey, parallel-bedded, fine sandstone and silty-shale in the middle part of Sandstone Unit–2 (Fig. 5) may represent repetitive cycles of turbiditic and suspension deposition. The individual sandstone beds of this facies type also show a fining-upward trend,

Figure 10. Photomicrographs of Sandstone Unit–2. (A) Authigenic pyrite is commonly found both in the fine- to medium-grained sandstones and in the fine sandstones, replacing dissolved feldspar grains. (B) Discrete grains of apatite and titanite are more commonly found in the fine sand.

(C) Illitic clay forming a thin coating around feldspar grains. (D) Dark-grey to black illitic clay and organic material found as a major pore-filling component in the fine sand and silty shale facies.

52

A.S. Azad et al

NORWEGIAN JOURNAL OF GEOLOGY

Figure 11. Comparison of mineralogical characteristics of Sandstone Unit–1 and Sandstone Unit–2. (A) Quartz/feldspar ratio (thin-section study) shows an upward-increasing trend from Sandstone Unit–1 to Sandstone Unit–2. (B) A decrease in K-feldspar (XRD %) and a corresponding increase in the plagioclase feldspar observed in Sandstone Unit–2 compared to Sandstone Unit–1. (C) Sericitisation (thin-section study) shows an upward increase that corresponds to the decreased feldspar contents from Sandstone Unit–1 to Sandstone Unit–2. See Fig. 9A (Sandstone Unit–1) and Fig. 5 (Sandstone Unit–2) for the rock sample positions in the lithological column.

erosional bases to underlying beds and centimetre-thick, undulating/rippled, silty-shale cappings, possibly reflecting individual turbiditic events. The intervals between the turbidity flows were possibly longer during this stage of crater sedimentation as the fine-grained sandstone beds are separated by 0.5–2 cm-thick, silty-shale layers deposited from suspension settling (Fig. 5). The overall fining-upward trend in this part of the succession, without any evidence of syn-depositional deformational structures or water-escape structures, suggests a further reduced rate of sedimentation. The upward decrease in thickness of sandstone beds (Fig. 5) also indicates a reduced clastic input. Local, medium- to coarse-grained sandstone beds can reflect a sudden increase in the energy conditions, e.g., related to base-slope failures and renewed erosion probably triggered by episodic storms.

Increased thicknesses of the silty shales in the upper part of this facies (Fig. 5) indicate a progressive increase in seawater depth and establishment of suspension-dominated sedimentation. The increase in the quartz/feldspar ratio (Fig. 11A) and improved textural maturity of this facies compared to the underlying facies signify possible further reworking of the crater-filling sediments. The

dark-grey to black, illitic clayey matrix (Fig. 10D) with increased organic content, and fine-grained authigenic pyrite and apatite (Tables 2, 3) reflects more stable and partly anoxic conditions prevailing within the crater.

The dark-grey to black shales overlying the fine-sandstone facies (Fig. 5) reflect a complete establishment of almost anoxic bottom-water conditions within the crater. This flaky, thinly laminated shale attains a thickness of about 180 m in the central part of the crater, indicating that the crater existed as a depression beneath the Cambrian sea for millions of years and eventually was filled by suspension settling deposits. Sporadic occurrences of angular to subangular quartz and feldspar grains and granitic clasts in the lowermost parts of the shale beds suggest that although hemipelagic sedimentation was dominant within the crater, clastic input from the crater walls had not completely died out. The presence of pyrite (Table 3) and organic matter in the upper part of the succession (Table 2) and lack of any trace fossils are suggestive of reduced ventilation under partly sedimentstarved depositional conditions.

Late post-impact sediments

Depositional area

Dormåls­knuten

Svodene–Ritlandsfjellet

Sedimentary units

24

DOR–O4–02 DOR–O4–01

Shale

21

21

25

SVDX–01–02 DOR–O4–03

23 25

SVDX–01–06 SVDX–01–03

31 26

SVDX–01–11 SVDX–02–03

32 26

SVDX–02–07 SVDX–02–06

27 14

SVDX–01–16 SVDX–01–13

32 32

SVDX–02–13

22

SVDX–03–01 SVDX–02–10

26 25

SVDX–03–05 SVDX–01–20

Quartz

Sample no.

Clast congl.

Clast congl.

Fine to medium sandstone

Fine sandstone

Dark gray to black shale

Facies

61

38

48

47

56

47

46

50

51

43

60

50

51

58

21

0

24

2

29

21

25

10

25

22

0

8

8

2

11

0

0

21

3

24

PlagioK-feldspar clase

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Calcite

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Dolomite

0

0

0

2

3

1

1

0

0

0

1

0

0

0

4

1

3

Chlorite

16

12

6

2

6

4

5

16

13

13

23

12

17

10

31

71

20

Mica

0

0

0

0

0

0

0

0

2

1

0

0

0

0

1

0

3

Pyrite

0

0

0

0

0

0

0

3

0

3

0

0

0

0

0

0

0

Others

0.33

0.31

0.36

0.33

0.38

0.32

0.38

0.62

0.44

0.63

0.23

0.44

0.63

0.55

0.52

8.33

0.54

Quartz/ Felds.

Pyrite

Pyrite

Pyrite

Hematite

Hematite

Trace

Table 3. Mineralogical composition of the different facies types of Sandstone Unit–2 from Svodene–Ritlandsfjellet and conglomerate beds from Dormålsknuten area, obtained from XRD analysis (semiquantitative). Quartz/feldspar ratios of the individual samples are presented in a separate column. Peak values for quartz (d = 4.26 Å), K-feldspar (d = 3.24 Å, 3.25 Å), plagioclase (d = 3.18 Å, 3.19 Å), calcite (d = 3.03 Å), dolomite (d = 2.89 Å), mica (d = 10 Å), chlorite (d = 7 Å) and pyrite (d = 2.71 Å) were used for quantifying the minerals. Only the peak height has been considered for the semiquantitative analysis of these minerals ­(Morris et al., 2008).

NORWEGIAN JOURNAL OF GEOLOGY Late post-impact sedimentation in the Ritland impact structure, western Norway

53

54

A.S. Azad et al

Dormålsknuten area The clast-supported conglomerate beds in the Dormålsknuten area (Figs. 7, 8A) can be interpreted as submarine scree deposits that slid into unconsolidated, cohesive marine clays. The limited lateral and vertical extents of these possible scree beds (Fig. 7) compared to the extensive, early post-impact, breccia beds (Upper Breccia Unit) (Azad et al., 2012) suggest that the crater slopes were more stable and thus reduced the input of coarser clastics derived from the crater wall and rim during the later stages of crater infilling. An alternative interpretation of these beds is that they may represent debrites, possibly triggered by storm or precipitation-related, cohesive, debris-flow events.

‘Screes’ or ‘talus’ are downslope accumulations of rock fragments/rock falls produced by failure of cliffs encompassing the downward movement of individual, typically gravel-size clasts (Drew, 1873; Gardner, 1983).

Scree deposits are considered to have formed by a granular flow mechanism, and bedforms produced by scree processes represent coarsening-upward arrangements of the clasts. In contrast, cohesive debrisflow deposits show en masse depositional characteristics, explaining the chaotic arrangement of the deposits.

In debris flows, the cohesive strength of the matrix acts as the dominant clast-supporting mechanism; the matrix consists of fairly fine-grained sediments with a significant amount of clay-size particles (Costa & Williams, 1984; Mulder & Alexander, 2001; Gani, 2004). The diagnostic features of a cohesive debris-flow deposit include basal inverse grading (shear flow), an ungraded flow body (plug flow) and a sandy upper part (waterlain). The clast-supported conglomerate beds in the Dormålsknuten area do not show the well-developed basal reverse grading expected in cohesive debrites. The middle part of the conglomeratic beds is ungraded while the top part contains coarser clasts (Fig. 8B). Thin bands consisting of a sandy matrix have been found within the conglomeratic beds, but not at the upper surface of the bed. Consequently, the upward increase in clast size, poorly developed bedding, comparable clast and matrix composition (see below), and lens-like occurrence within the clay are more analogous to submarine scree deposits than to debris-flow deposits.

The clast composition of these conglomerate beds is comparable to the general basement composition of the area (granitic/gneissic) (Azad et al., 2012) indicating that these sediments have been derived from the erosion of the adjacent crater wall or rim. The matrix consists of significant amounts of clay with mediumto coarse-grained, angular to subangular quartz and feldspar, and small granitic/gneissic clasts (Table 2).

The clay content of the matrix is probably sourced from pelagic sediments. The comparable clast and matrix mineralogical composition (granitic/gneissic) (Tables 2, 3) of the conglomerate beds suggests that the extra-crater sediment influx was not so significant in this later stage

NORWEGIAN JOURNAL OF GEOLOGY

of crater sedimentation, and that the sediments were mostly derived from crater-slope erosion and reworking of the crater infills. The average quartz/feldspar ratio (2.2) in the matrix of these conglomerate beds (Table 2) is higher compared to the quartz/feldspar ratio (0.3) in the matrix of the early post-impact, clast-supported breccias (Upper Breccia Unit) (Azad et al, 2012). This suggests a further reworking of these sediments during later stages of crater sedimentation. Traces of pyrite (Table 3) and the organic content (Table 2) within the matrix suggest restricted, less ventilated conditions in the crater basin at the time that these conglomerate beds were deposited.

The outsize clasts within the dark-grey to black shales can be interpreted as deriving from rock falls into marine clays (Fig. 8C). Evidence of smearing of the clays around the clasts (Fig. 8C) also supports a rockfall mechanism.

The brittle nature of the shale layers around the clasts and fracturing of the clasts (Fig. 8C) are probably related to compaction and Caledonian tectonics, which affected the upper parts of the crater infill. The thin (Fig. 8D) bed (at 10.6 m in the log) consisting of granitic clasts and coarse sand has probably resulted from crater-wall erosion and may represent a storm event, probably deposited from a hyperconcentrated density flow.

Depositional model synthesis A thin (0–30 cm thick) layer of basal conglomerate has been found overlying the sub-Cambrian peneplain in the southwestern part, outside the crater, probably deposited­ from reworking of weathering products of the Pre­ cambrian basement rocks (Riis et al., 2011; Setså, 2011).

This basal conglomerate, probably of Early Cambrian age, is overlain by a 10–20 m silty-shale layer (Setså, 2011) indicating the onset of the Early Cambrian marine transgression. This basal conglomerate and silty-shale unit represent the pre-impact sediments in the Ritland area, as evidenced by the overlying ejecta layer (Kalleson et al., 2012), suggesting a shallow-marine setting of the Ritland bolide.

The numerical modelling of the Ritland impact (Shuvalov et al., 2012), the distribution of the ejecta layers (Kalleson et al., 2012) and the sedimentary signatures within the crater infills (Azad et al., 2012) together indicate that the contemporary seawater depth during the Ritland impact event was ≤100 m. The depositional model in Fig. 12A represents crater sedimentation during the late syn-impact to early post-impact stage and has been discussed in detail in Azad et al. (2012). The depositional model presented in Fig. 12B represents crater sedimentation during the transitional stages (from early post-impact to late post-impact). A brief summary of the depositional processes during the late syn-impact to early post-impact interval is presented here to illustrate the link between the early post-impact and late postimpact stages of crater sedimentation.

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

The Ritland crater rim is considered to have formed subaerially in view of the crystalline target surface, shallow epeiric-sea setting and very thin, unconsolidated marineclay cover. With the rim probably above sea level (Shuvalov et al., 2012), the initial Ritland crater sedimentation was thus also likely to have been subaerial (Azad et al., 2012). Sedimentation in the crater was initiated by the collapse of the transient cavity and rim; debris avalanched down towards the crater centre depositing the lower part of the Lower Breccia Unit (Azad et al., 2012) (Fig. 12A). Debris-flow deposition dominated (Fig. 12A) immediately after the rock avalanches as sea­water resurged back into the crater. The numerical model­ ling (Shuvalov et al., 2012) and the sedimentary signatures of the lower part of the crater infills (Lower Breccia Unit) (Azad et al., 2012) suggest that the Ritland crater probably did not experience an instantaneous, powerful resurge of seawater. Water entered the crater through a breaching of the crater rim and eventually filled the crater. The crater rim remained exposed to the atmosphere (Fig. 12C) until it was inundated by later transgressions.

Turbidity flows dominated in the central part of the water-filled crater cavity. The Sandstone Unit–1 (Table 1) represents turbidites from a minor submarine-fan setting prograding on to the crater floor (Fig. 12A). Along the steep crater walls, deposition of the coarser clastics (Upper Breccia Unit) continued, developing fans from the crater walls, prograding towards the crater centre (Fig. 12A). Through time, the crater cavity became more stable and the site for suspension deposition of finegrained sediments (Sandstone Unit–2) (Fig.12B).

Sandstone Unit–2 was probably deposited during the retreating or abandonment phase (Fig.12B) of the earlier active, high-energy, minor submarine fans of the early post-impact stage (Sandstone Unit–1) (Azad et al., 2012) (Fig. 12A). The coarsening- and thickening-upward trend in the lower part and a rather homogeneous middle and upper part of the Sandstone Unit–1 (Fig. 9A) represent an upward transition from prograding-fan developments to a more stagnant situation during the later stages of early post-impact sedimentation. The fining- and thinning-upward trend (Figs. 5, 9B) of the succeeding transitional sandstones suggests that the active fans retreated during later stages of crater sedimentation (Fig.12B). The stratigraphical upward increase in the quartz/feldspar ratio (Fig. 11A) and sericite content (Fig. 11C) from early post-impact to the transitional sandstones, and a comparable mineral composition (Tables 2, 3) to that of the granitic and gneissic basement suggest that the reworking of the early post-impact craterwall sediments was the major source of sediment supply within this closed, sediment-starved, crater basin during the later stages of crater sedimentation.

The extrabasinal sediment supply during Sandstone Unit–2 crater sedimentation was probably insignificant. Outside the crater, the extensively peneplained, epicontinental sea platform was dominated by a series of

55

marine transgressions from the Early Cambrian onwards with little or no evidence of significant regressive phases before the end of the Cambrian (Nielsen & Schovsbo, 2011). Baltoscandia became extensively flooded and the clastic input was significantly reduced during this time. Similar conclusions were also suggested by Setså (2011) from studying the post-impact sediments outside the crater, e.g., the post-impact shale overlying the ejecta layer was characterised by a low quartz/feldspar ratio and greater amounts of pyrite, apatite and organic content, representing sediment-starved, anoxic ocean bottom conditions (Setså, 2011). The transitional sediments (Unit–2 time) inside the crater do not reveal any coarsening-upward trend, thus no regressive or prograding events were noted within the crater. A minimal supply of terrigenous clastics from outside the crater due to Early/Mid Cambrian, low-angle river profiles (Nielsen & Schovsbo, 2011) and reduced crater-slope erosion processes due to low-energy, depletive flow resulted in sediment-starved conditions within the crater. Fan building was probably abandoned in the early part of the late post-impact stage (Unit–2 time) and the crater was open mostly for suspension deposition over long time intervals. Increasing amounts of authigenic pyrite and organic matter, significant amounts of apatite, and an absence of trace fossils in the upper part of the succession support the interpretation of a gradual shift from high-energy, oxygenated (turbidity flows) to sediment-starved, lowenergy, stagnant, less-ventilated, bottom-water conditions (suspension deposition).

The thickness of the Alum Shale unit in the crater centre is estimated to be 180 m (Riis et al., 2011). The sedimentation rate of the Alum Shale was extremely low, calculated at 3–8 mm per 1000 years (Berger, 1974) with only minor variations across the Baltoscandian platform (Thickpenny, 1984). The lithological variations are small within these shales with no apparent breaks observed; thus, zone thickness could be proportional to depositional duration (Thickpenny, 1984). An estimated time period for deposition of a 80–100 m-thick column of Alum Shale in the Skåne area, Sweden, is approximately 23–25 million years (Thickpenny, 1984). The sedimentation rate of the Alum Shale unit in the Ritland crater basin may have been slightly higher since the basin functioned as a closed depression beneath the epicontinental sea and would have received some clay influx from outside the crater. Assuming an upper limit of the sedimentation rate of 10 mm per 1000 years within the crater, the 180 m-thick shale unit in Ritland would have accumulated in 37.5 million years, probably covering the whole of the Middle to Upper Cambrian epoch. Although uncertainties related to compaction of the overburden and the effects of Caledonian tectonics have to be considered, it can be assumed that the crater was completely filled with marine clays by Late Cambrian time. The thickness variations of the Alum Shale unit inside (180 m) and outside (15–20 m) the crater suggest that the crater was completely filled by marine clays during this time.

56

A.S. Azad et al

NORWEGIAN JOURNAL OF GEOLOGY

Figure 12. (A) Simplified depositional model (one half of the crater) representing the deposition of different sedimentary units in the Ritland crater during the late syn-impact and early post-impact stages (modified from Azad et al., 2012). (B) Model representing crater sedimentation during the transitional stage (early post-impact to late post-impact stage). The submarine fans were probably retreating (indicated by red arrow) due to a reduced clastic input from both inside and outside the crater. The channel abandonment processes were active, and the sediments were derived mostly from erosion and reworking of the crater-wall sediments. (C) Partially submerged crater within the Cambrian sea, with rims exposed to the atmosphere during the transitional stage.

The similar thickness (10 m) of the overlying Cambro– Ordovician sandstones (Knaust, 2004) inside and outside the crater confirms this observation.

The onlapping relations of the dark-grey shale on the early post-impact breccias exposed higher up in the northeastern crater wall in the Dormålsknuten area (Fig. 7) suggest a continued and prolonged interval of suspension deposition (Fig. 13). The clast-supported conglomerate beds in the northeastern part of the crater basin intercalated with the marine clays at an elevation of 700–720 m (Fig. 4) correspond to a depth of 50 m below the peneplain (the impact target surface), and 300 m above the base of the shale layer in the crater centre (Fig. 4). This implies that mass-flow processes, i.e., with development of screes, were active even in the very late stages of crater sedimentation (Fig. 13). The crater walls remained as escarpments beneath the epicontinental sea for millions of years.

The natural gamma-radiation measurements in the lower part of the dark-grey to black shale unit in the Svodene– Ritlandsfjellet area varies from 200 to 270 cps (Fig. 5), while in the upper middle part (Dormålsknuten area) the values range from 210 to 280 cps. These Alum Shale values are much lower than those in the fine- to mediumsandstone facies (350–415 cps) (Fig. 5) of Sandstone Unit–2 (Svodene–Ritlandsfjellet area). The higher gamma-radiation value within the fine- to mediumsandstone facies could be related to the higher average content of K-feldspar found in thin sections and XRD analyses (21%, Table 3). The surprisingly low gamma values of the dark-grey to black shales do not correspond with the typical highly radioactive characteristics of Alum Shale. Particularly the upper part of the Upper Cambrian Alum Shale in the Mjøsa region (Norway) and Øland (Sweden) is remarkable for its high (400–800 cps and 500–1000 cps, respectively) natural gamma activity due to high uranium concentrations (Dypvik, 1993). The lower gamma measurements in the shales of the Ritland

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

crater could probably be related to the lithological heterogeneity and variations in geochemical reactions due to the differences in oxygen and organic contents at the ocean bottom (Dypvik, 1993).

Late post-impact crater sedimentation dominated by suspension settling with a transitional (early post-impact to late post-impact) episode of turbidity flows has been observed in several terrestrial shallow-marine, impact craters. The Gardnos crater, assumed to have been roughly contemporaneous with the Ritland structure (Riis et al., 2011), is marked by a succession (25 m thick in the Branden core) of interbedded sandstones, siltstones and shales in the uppermost part of the craterinfilling succession, representing alternating episodes of turbidity flows and suspension deposition (Kalleson et al., 2008). In the Estonian Kärdla crater, a thick (>200 m) succession of carbonate sediments marks the late post-impact stage, deposited from suspension settling under sediment-starved, stable crater conditions (Ainsaar et al., 2002). A comparable geological setting in the Swedish Lockne crater also resulted in a thick sequence (>90 m) of carbonate sediments (Frisk & Ormö, 2007) during the late post-impact stage of crater sedimentation. In Virginia, USA, the Chikahominy Formation (94 m in the Eyreville core) consisting of finely laminated clays and silts with abundant sandsize foraminifera, shell fragments, glauconite and traces of pyrite and mica, represents the late post-impact sedimentation in the Chesapeake Bay impact structure.

The Chikahominy Formation has been interpreted to have been deposited from suspension settling in a restricted offshore environment (Browning et al., 2009).

The late post-impact sedimentation in the Flynn Creek crater, Tennessee, is marked by the deposition of black shales (up to 55 m thick), deposited over several hundred thousand years by alternating episodes of fine-grained turbidites and pelagic suspension settling (Schieber & Over, 2005).

The thickness of the late post-impact successions within these various craters is commonly many times higher compared to the coeval late post-impact succession outside the craters. The thickness of the Alum Shales in the central part of the Ritland structure is almost ten times higher compared to the thickness (20 m) outside the crater (Knaust, 2004). In the Kärdla crater, the late post-impact, Upper Ordovician limestone encountered in the crater centre is at least four times thicker than the equivalent limestone outside the crater (Ainsaar et al., 2002). The thickness of the Dalby Limestone in the Lockne crater also exceeds by several times that of the coeval Dalby Limestone outside the crater (Frisk & Ormö, 2007). The Flynn Creek crater has a thickness of black shale which is at least five times greater than the coeval Upper Devonian Chattanooga Shale (Schieber & Over, 2005). The Chickahominy Formation in the Eyreville core, drilled approximately in the crater centre, is also thicker compared to any other core drilled

57

through this formation outside the crater (Edwards et al., 2009).

Summary The Sandstone Unit–2 in the Svodene–Ritlandsfjellet area clearly represents a transitional phase between the early post-impact and late post-impact stages of crater sedimentation. The Ritland crater sedimentation during the late syn-impact and early post-impact stages was mainly controlled by gravity and mass flows, e.g., rock avalanches, screes, debris and turbidity flows (Azad et al., 2012), whereas during the late-post impact stage sedimentation was mostly suspension-dominated, interrupted by minor episodes of turbidity flows during the basal transitional phase.

The fine- to medium-sandstone facies of the transitional sandstones (Unit–2) is texturally more mature compared with the early post-impact sandstones (Unit–1) and interpreted to be derived from base-slope erosion and further reworking of the earlier crater infills. These sediments were probably deposited by turbidity currents during the retreating or abandonment stage of submarine fans that developed during the early post-impact stage.

Through time, reduced rates of both crater-slope erosion and sediment supply from outside the crater resulted in a gradual, progressive establishment of suspensiondominated sedimentation. The dark-grey to black clayey matrix, with increased amounts of authigenic pyrite, apatite and organic matter within a fine-sandstone facies, represents more stable, calm and less-ventilated, bottomwater conditions in the crater basin. An upward transition of this facies into thick, dark-grey to black shales

Figure 13. Simplified depositional model (one half of the crater) representing depositional processes during the late post-impact stage. The clast-supported conglomerate beds in the Dormålsknuten area represent small-scale scree deposits (orange) sliding into the marine clay (green) as a result of occasional slope failures in the later stages of crater sedimentation.

58

A.S. Azad et al

suggests a complete establishment of anoxic to hypoxic conditions where sediments were deposited during a protracted (millions of years) and quiet period of suspension deposition.

The scree deposits found in the Dormålsknuten area probably resulted from sporadic slope failures during the later stages of crater sedimentation. The small-scale scree beds, the metre-size clasts within the marine clay, and a few thin sandstone beds suggest that submarine slides, rock falls and reworking of the crater-wall sediments were active processes for a long time after the impact, occasionally depositing coarser materials between the long, quiet periods of suspension deposition.

Acknowledgments. We express our sincere gratitude to the other Ritland project-group members for their kind assistance and worthy discussion. Constructive and thoughtful comments of Drs. Kevin Evans and David Jolley helped to improve the manuscript. Adrian Read worked on improving the language, we appreciate his contribution. We are also grateful to the Sven Egil and Kari Sørensen family for their hospitality during the field days at Ritland. The Research Council of Norway generously supported the three-year Ritland project; we highly appreciate this. Finally, the laboratory technicians at the Department of Geosciences, University of Oslo, and the people of Ritland, are acknowledged for their support and assistance.

References Ainsaar, L., Suuroja, K. & Semidor, M. 2002: Long-term effect of the Kärdla crater (Hiiumaa, Estonia) on Late Ordovician carbonate sedimentation. Deep Research II: Topical studies in Oceanography 49, 1145–1155.

Azad, A.S., Dypvik, H., Tomczyk, M., Kalleson, E. & Riis, F. 2012: Late syn-impact and early post-impact sedimentation in the Ritland impact structure, western Norway, Norwegian Journal of Geology 92, 405–431.

Berger, W.H. 1974: Deep sea sedimentation. In Burk, C.A. & Drake, C.L. (eds.): The geology of continental margins, Springer-Verlag, New York, pp. 231–241.

Bergström, J. & Gee, D.G. 1985: The Cambrian of Scandinavia. In Gee, D.G. & Sturt, B.A. (eds.): The Caledonide Orogen – Scandinavia and related areas, John Wiley & Sons, Chichester, pp. 247–271.

Bouma, A.H. 1962: Sedimentology of some flysch deposits: a graphic approach to facies interpretation. Elsevier, Amsterdam, 168 pp.

Bruton, D.L. & Harper, D.A.T. 2000: A mid-Cambrian shelly fauna from Ritland, western Norway and its palaeogeographical implications. Bulletin of the Geological Society of Denmark 47, 29–51.

Bruton, D.L., Harper, D.A.T. & Repetski, J.E. 1989: The stratigraphy and faunas of the Parautochthon and Lower Allochthon of southern Norway. In Gayer, R.A. (ed.): The Caledonide geology of Scandinavia, Graham & Trotman, London, pp. 231–241.

Browning, J.V., Miller, K.G., McLaughlin Jr., P.P., Edwards, L.E., Kulpecz, A.A., Powars, D.S., Wade, B.S., Feigenson, M.D. & Wright, J.D. 2009: Integrated sequence stratigraphy of the postimpact sediments from the Eyreville core holes, Chesapeake Bay impact structure inner basin. In Gohn, G.S., Koeberl, K., Miller, K.G. & Reimold, W.U. (eds.): The ICDP–USGS deep drilling project in the Chesapeake Bay impact structure: Results from the Eyreville core holes, Geological Society of America Special Paper 458, pp. 775–810.

Costa, J.E. & Williams, G.P. 1984: Debris flow dynamics. U.S. Geological Survey, Open-file Report 84–606 (videotape).

Drew, F. 1873: Alluvial and lacustrine deposits and glacial records of

NORWEGIAN JOURNAL OF GEOLOGY

the upper Indus Basin. Geological Society of London Quaternary Journal 29, 441–471.

Dypvik, H. 1993: Natural gamma activity – a possible aid in sedimentological framework. Norsk Geologisk Tidsskrift 73, 58–73.

Dypvik, H. & Kalleson, E. 2010: Mechanisms of late synimpact to early postimpact crater sedimentation in marine-target impact structures.

Geological Society of America Special Papers 465, 301–318.

Edwards, L.E., Powars, D.S., Gohn, G.S., & Dypvik, H. 2009: Geologic columns for the ICDP–USGS Eyreville A and B cores, Chesapeake Bay impact structure: Sediment-clast breccias, 1096 to 444 m depth, In Gohn, G.S., Koeberl, K., Miller, K.G. & Reimold, W.U. (eds.): The ICDP–USGS deep drilling project in the Chesapeake Bay impact structure: Results from the Eyreville core holes, Geological Society of America Special Paper 458, pp. 51–90.

Frisk Å.M. & Ormö J. 2007: Facies distribution of post-impact sediments in the Ordovician Lockne and Tvären impact craters: Indications for unique impact-generated environments. Meteoritics and Planetary Science 42, 1971–1984.

Folk, R.L. 1974: Petrology of Sedimentary Rocks. Hemphill Publishing, Texas, 182 pp.

Gani, M.R. 2004: From turbid to lucid: A straightforward approach to sediment gravity flows and their deposits. The Sedimentary Record 2, 4–8.

Gardner, J.S. 1983: Accretion rates on some debris slopes in the Mt. Rae area, Canadian Rocky Mountains. Earth Surface Processes and Landforms 8, 347–355.

Gee, D.G., Fossen, H., Henriksen, N. & Higgins, A.K. 2008: From the early Paleozoic platforms of Baltica and Laurentia to the Caledonide orogen of Scandinavia and Greenland. Episodes 31, 44–51.

Henningsmoen, G. 1952: Early Middle Cambrian fauna from Rogaland, SW Norway. Norsk Geologisk Tidsskrift 30, 13–31.

Kalleson, E., Dypvik, H. & Naterstad, J. 2008: Post-impact sediments in the Gardnos impact structure, Norway. In Evans, K., Horton Jr., J.W., King Jr., D.T & Morrow J.R. (eds.): The Sedimentary Record of Meteorite Impacts, Geological Society of America Special Paper 437, pp. 19–41.

Kalleson, E., Riis, F., Setså, R. & Dypvik, H. 2012: Ejecta Distribution and Stratigraphy – Field Evidence from the Ritland Impact Structure. 43rd Lunar and Planetary Science Conference, Texas, U.S.A. Abstract # 1351.

Knaust, D. 2004: Cambro–Ordovician trace fossils from the SW-Norwegian Caledonides. Geological Journal 39, 1–24.

Melosh, H.J. 1989: Impact cratering: a geological process. Oxford University Press, New York, 245 pp.

Morris, C., Sieve, B.J. & Bullen, H.A. 2008: E-Learning Module: Introduction to X-ray Diffraction. http://www.asdlib.org/ onlineArticles/ecourseware/Bullen_XRD/XRDModule_index.htm (accessed 01.05.12).

Mulder, T. & Alexander, J. 2001: The physical character of subaqueous sedimentary density flows and their deposits. Sedimentology 48, 269–299.

Nielsen, A.T. & Schovsbo, N.H. 2011: The Lower Cambrian of Scandinavia: Depositional Environment, Sequence Stratigraphy and Palaeogeography. Earth Science Reviews 107, 207–310.

Pemberton, S.G., MacEachern, J.A. & Frey, R.W. 1992: Trace fossil facies models: Environmental and allostratigraphic significance.

In Walker, R.G. & James, N.P. (eds.): Facies Models: Response to Sea Level Change, Geological Society of Canada, pp. 47–72.

Rey, P., Burg, J-P. & Casey, M. 1997: The Scandinavian Caledonides and their relationship to the Variscan belt. In Burg, J-P. & Ford, M. (eds.): Orogeny through time, Geological Society of London Special Publication 121, pp. 179–200.

Riis, F., Kalleson, E., Dypvik, H., Krøgli, S.O. & Nilsen, O. 2011: The Ritland impact structure, southwestern Norway. Meteoritics and Planetary Science 46, 748–761.

Schieber, J. & Over, D.J. 2005: Sedimentary fill of the Late Devonian Flynn Creek crater: a hard target marine impact. In Over, D.J.,

NORWEGIAN JOURNAL OF GEOLOGY

Late post-impact sedimentation in the Ritland impact structure, western Norway

Morrow, J.R. & Wignall P.B. (eds.): Understanding Late Devonian and Permian–Triassic Biotic and Climatic Events: Towards an Integrated Approach, Developments in Palaeontology and Stratigraphy 20, pp. 51–69.

Setså, R. 2011: The Ritland impact structure: characteristics and distribution of the ejecta layer and associated Lower Paleozoic sedimentary succession. MSc thesis, University of Oslo, 111 pp.

Shuvalov, V., Dypvik, H., Kalleson, E. & Setså, R. 2012: Modelling the 2.7 km, shallow marine Ritland impact structure. Earth Moon Planets 108, 175–188.

Thickpenny, A. 1984: The sedimentology of the Swedish Alum Shales.

In Stow, D.A.V. & Piper, D.J.W. (eds.): Fine-grained Sediments: Deepwater Processes and Facies, Geological Society of London Special Publication 15, pp. 511–525.

Wentworth, C.K. 1922: A scale of grade and class terms for clastic sediments. Journal of Geology 30, 377–392.

59

60

A.S. Azad et al

NORWEGIAN JOURNAL OF GEOLOGY

NORWEGIAN JOURNAL OF GEOLOGY

Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway

61

Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway Rannveig Øvrevik Skoglund & Stein-Erik Lauritzen Skoglund, R.Ø. & Lauritzen, S.-E.: Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway. Norwegian­ Journal of Geology, Vol 93, pp 61–73, Trondheim 2013. ISSN 029-196X. The present-day Grønli–Seter karst aquifer in Rana is fed by a surface stream that invades a relict cave system, so that the present hydrology is constrained by pre-existing geometry. This type of karst aquifer is quite common in the glacially sculptured landscape of Norway, where speleo­ genesis displays a strong imprint from glacier-ice contact. The Grønli–Seter cave system is situated in the hillside of Røvassdalen, with passage morph­ology and palaeoflow marks indicating development below the watertable and by ascending water flow in the presence of a glacier filling up the valley. The aquifer has a mean annual discharge of about 0.13 m3 s-1, which makes up about half of the total runoff from Strokbekken from where it is sourced. Hydrograph and quantitative dye-tracer experiments demonstrated that the aquifer is totally dominated by conduits and comprises three sections with distinctly different cross-sectional areas. The lower part of the traced aquifer comprises essentially phreatic (submerged) conduits with a mean cross-sectional area of about 4.5 m2 and a static volume of minimum 4000 m3. In the upper sections, the ratio of submerged, water-filled passages is lower, and the cross-sectional area of the water flow is smaller, but with a greater increase during floods. The maximum estimate of the total aquifer volume is 10,000 m3 (at a discharge of about 0.2 m3 s-1). The total hardness (the concentration of Ca2+ and Mg2+ in the water) has been measured in the low range of 15 to 36 mg CaCO3 equivalents per litre, which is normal for settings like this. The highest transport rates of calcite leaving the aquifer are related to the highest discharges. The evolution of invasion aquifers like this one is therefore suggested to accele­rate when the recharge conditions are changed, i.e., due to glacial erosion, so that the entire surface stream is captured by the stream sink.

Rannveig Øvrevik Skoglund, Department of Geography, University of Bergen, Fosswinckelsgt. 6, 5007 Bergen, Norway. Stein-Erik Lauritzen, Department of Earth Science, University of Bergen, Allégaten 41, 5007 Bergen, Norway.

E-mail corresponding author (Rannveig Øvrevik Skoglund): rannveig.skoglund@geog.uib.no

Introduction Complex, multiphase cave systems develop where there have been shifts in the hydraulic function of the cave during development (e.g., Ford & Williams, 2007; Skoglund & Lauritzen, 2010). During the Quaternary, high variability in climate was accompanied by a strong variability in the ice-sheet cover (e.g., Sejrup et al., 2000; Olsen et al., 2002), and accordingly a high variability in the karst hydrological regime as the ice redirected hydraulic gradients, water chemistry and topography (e.g., Ford & Williams, 2007; Benn & Evans, 2010). Karst systems throughout Norway are considered to have been exposed to such large changes during their existence and development (Lauritzen & Skoglund, 2013).

In the glacially sculptured landscape of Norway, karst caves commonly appear in a hanging position in the valley walls and their entrances appear to have been truncated by subsequent glacial erosion (Lauritzen, 2001; Lauritzen & Skoglund, 2013). These caves are in essence relict, i.e., they are no longer active, and their origin and development cannot be explained by their present position in the terrain. In contrast, active caves are either juvenile, due to the short time of development (only

about 10,000 years since the deglaciation), or they are invasion caves where the postglacial water flow has been superimposed on a pre-existing cave developed under different conditions.

Karst aquifers may comprise three types of porosity: pores, fractures and conduits. The main karstic rock in Norway is marble, which is a metamorphic carbonate rock essentially devoid of primary porosity related to sedimentary and diagenetic processes (Lauritzen, 2001). Accordingly, karst groundwater circulation and cave development depend on secondary porosity such as fractures, faults and tectonised lithological contacts.

When slightly acidic water flows through fractures in carbonate rocks the fracture walls are dissolved, slowly widening the fracture into a conduit. Below the watertable (i.e., the phreatic zone), corrosion attacks the rock along the entire perimeter and radially away from the original flow path, forming conduits with a subcircular, lenticular or rift-shaped cross section (e.g., Lauritzen & Lundberg, 2000). Above the watertable (i.e., in the vadose zone) where fractures and conduits are only partly water filled, corrosion attacks the floor forming gravity-driven, canyon-shaped passages. The conduit water flow resembles underground rivers, which

62

R.Ø. Skoglund. & S.-E. Lauritzen

NORWEGIAN JOURNAL OF GEOLOGY

makes underground water flow through karst aquifers contrast the water flow in porous and fracture aquifers.

Objectives Few studies have been published on karst water and karst aquifers in Norway (Lauritzen, 1981, 1986; Bakalowicz, 1984; Lauritzen et al., 1985; Øvstedal & Lauritzen, 1989; Knutson, 2008; Skoglund & Lauritzen, 2010). The purpose of this paper is to give a physical and chemical characterisation of a marble karst aquifer, the Grønli– Seter aquifer. It is associated with two well-known tourist caves in Norway, Grønligrotta and Setergrotta (Oxaal, 1914; Horn, 1947; St. Pierre, 1988). The Grønli–Seter cave system consists of more than 8 km of relict, essentially phreatic conduits (Øvrevik, 2002; Skutlaberg, 2003; Hestangen, 2005), although the caves are not yet linked by human exploration and survey. Passage morphology and palaeoflow marks (Curl, 1974) in the cave walls demonstrate that the relict conduits were essentially formed under phreatic (water-filled) conditions by an ascending water flow (Skutlaberg, 2003). The contemporary cave stream has invaded the cave system and connects the two caves through conduits of inexplorable size. We may hypothesise that the aquifer structure and conduit pattern are inherited from a previous hydraulic regime, presumably in response to a glacier occupying the present valley.

Through hydrological monitoring data and analyses of water samples, the chemical state of the aquifer and some considerations about the transport of total hardness and corrosion will be addressed. The main focus in this paper is on the structure and volume of the aquifer as revealed by a series of tracing experiments through different sections of the aquifer. The characteristics of the studied aquifer and implications for its evolution are discussed in a wider context of other invasion aquifers in Norway.

Description of the field area The drainage area The Grønli–Seter cave system (66°25’N, 14°15’E) is located about 10 km north of Mo i Rana, northern Norway (Fig. 1). The cave system and the aquifer are situated in the eastern hillside of Rødvassdalen valley.

The Grønli–Seter aquifer receives recharge from a stream sink in the Strokbekken stream, approximately 250 m a.s.l. (Fig. 2). Several smaller stream sinks exist farther upstream, but these have been blocked artificially during the recent decades. The Strokbekken stream has a catchment area of about 7 km2, which consists mainly of mica schist (Søvegjarto, et al., 1989), i.e., the Grønli–Seter aquifer has allogenic drainage (Ford & Williams, 2007).

The drainage area is sparsely vegetated, partly situated above the tree line, and with widespread bogs. The karst aquifer itself is situated below the tree line.

Only a part of the discharge from the catchment area of

Figure 1. Location of the field area. The rectangle shows the location of the map in Fig. 2.

Strokbekken is captured into the Grønli–Seter aquifer, the remaining water continuing along a surface stream course towards the Røvassåga river in the valley bottom.

Water from the aquifer re-emerges in a spring pool in the glaciofluvial deposits in the valley bottom, approximately 48 m a.s.l. (Fig. 1), forming a dammed spring caused by the valley aggradation (Ford & Williams, 2007).

Climate The climate is subarctic oceanic with mild winters and cool summers. The mean annual temperature at sea level in the inner part of Ranafjorden (Båsmoen) is 3.0°C (Det norske meteorologiske institiutt, DNMI, 1991–2001). The field area is most likely slightly cooler due to its elevation and distance from the fjord. During the hydrological year 2000/2001, the mean annual air temperature at the spring gauging station was 1.4°C (48 m a.s.l.).

Precipitation falls throughout the year with May and June as the driest months. During winter, the precipitation falls largely as snow, but rainstorm events may occur throughout the year. The mean annual precipitation in Grønlia is 1680 mm (1988–2001). During the hydrological year 2000/2001, the total precipitation in Grønlia was 1351.4 mm and the maximum snow depth was 108 cm. The annual runoff from the Strokbekken drainage area, assuming an evapotranspiration of 250 mm yr-1, is about 7.7 ×106 m3, equal to a mean annual discharge of 240 l s-1, or a drainage intensity of 34 l s-1 km-2, in good accordance with other, regional estimates of 20–80 l s-1 km-2 (NVE, 2002).

NORWEGIAN JOURNAL OF GEOLOGY

Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway

63

Description of the explored part of the aquifer The stream sink is small and not of enterable size. The cave stream first appears in a siphon in a large, subcircular, phreatic tube (about 2 m in diameter) at the uppermost end of Grønligrotta (Fig. 2). Divers have explored the system upstream for a short distance before the conduit becomes too restricted. The stream has incised a vadose canyon in the subcircular, phreatic tube of the main passage (Fig. 3A). The stream flows below the maze of Grønligrotta through a river gorge in which water flows down the dip of strata along the shortest route of interconnected fractures. The river gorge ends in a rift passage after travelling a horizontal distance of about 300 m and dropping 70 m (Fig. 4). The rift here is too narrow to enter, though the sound of flowing water indicates that the water flow is vadose for some distance beyond.

In Setergrotta, the cave stream appears in four different vadose sections (passages), indicating a series of submerged conduits (Figs. 2, 4). The upper, southern part of Setergrotta has a 330 m-long vadose section with a vertical drop of about 60 m. The next section of the aquifer is phreatic, but this lower flow route has low flow capacity. The re-emergence of the cave stream at site 2 is shown in Fig. 3B. During floods, a perched sump (Palmer, 2007) develops upstream of site 2 (Fig. 4). The water level may rise up to 11 m and form a 200 m-long floodwater path where water eventually flows into Marmorgangen or down to site 2.

Figure 2. Map over the field area with cave maps and bedrock. The hydrological system is illustrated in accordance with the legend.

Bedrock The bedrock in the field area comprises Caledonian thrust-nappe complexes and forms part of the Rødingsfjell Nappe Complex of the Uppermost Allochthon (Stephens et al., 1985; Søvegjarto et al., 1988, 1989). The cave system is developed in a calcite and dolomite marble within the Ramnålia Nappe. The rocks are folded in a recumbent, anticlinal structure with a marble core confined beneath massive strata of mica schist where the local fold axis trends E–W (Søvegjarto et al., 1989) (Fig. 2). At the cave location, the Rødvassdalen valley cuts across the strike of the rock units.

Cave divers have confirmed the connection between the sump in Marmorgangen and the upstream sump at site 1 (Figs. 2, 4), though the central part has not yet been surveyed (R. Nielsen, pers. comm., 2011). The deepest part of this loop reaches sea level, which is far deeper than the present valley bottom (48 m a.s.l.). A shaft side-passage in the upstream direction has also been identified, probably leading to the sump at site 2 (R. Nielsen, op.cit., 2011). The sump in Marmorgangen appears not to be an active part of the conduit volume determined at low discharges. Sites 1 and 2 and the sump in Marmorgangen are only slightly elevated above the base level of the system (spring pool).

The estimated length of the conduit aquifer is based on survey data along the cave stream and the shortest or direct distance across the inaccessible segments. The direct distance between two unconnected segments of the cave stream is multiplied by a sinuosity factor to obtain a more reliable estimate of the actual length of the flow path.

Based on data from the surveyed cave stream, a sinuosity factor of 1.5 seems to be reasonable. The total length of the aquifer (from the stream sink (site 5) to the spring (site 0)) is estimated to 2750 m. Estimated lengths, hydraulic gradients and phreatic/vadose ratios of the five cave segments are presented in Table 1. The average gradient of the underground flow path is 0.07 (7%). However, the average channel slope of the different sections ranges from 0.18 (site 4–3) to 0.005 (site 2–1–0) (Fig. 4).

64

R.Ø. Skoglund. & S.-E. Lauritzen

NORWEGIAN JOURNAL OF GEOLOGY

Figure 3. Left: The cave stream in the main passage in Grønligrotta at base flow. Right: The reemergence of the cave stream just upstream of site 2 in Setergrotta (base flow).

Figure 4. Vertical profile of the Grønli–Seter conduit aquifer and cave system.

Table 1. Distance, hydraulic gradient and vadose ratio of the aquifer segments between the sites for hydrological monitoring, water sampling and trac er experiments, based on survey data. Lengths are rounded off to the nearest 50 m.

Site section

Survey distance, s (m)

5–4

Estimated distance, s + 1.5x (m)

Hydraulic gradient

Vadose proportion of segment

Distance to spring (m)

260

400

<0.025

a

2750

500

0.18

100%

2350

0.10

37%

1850

<0.005

<10%

4–3

500

3–2

330

240 + 140

900

2–1

230

220

550

260

400

1–0 a

Direct distance, x (m)

950 400

Probably <10%

Material and methods Time series of hydrological and physicochemical para­ meters Six sites are numbered along the flow path, with increasing number showing increasing distance from the spring (Figs. 2, 4). The sites are used for various purposes as

explained below. In the cave inlet in Grønligrotta (site 4) and in the spring pool (site 0), water level, water temperature and electrical conductivity were recorded at one-hour intervals over a period of one year using permanently installed digital data loggers (Aanderaa Instruments). In addition, discharge was measured at both sites during fieldwork using the standard salt-dilution method (Hongve, 1987). Correlation of discharge and

NORWEGIAN JOURNAL OF GEOLOGY

Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway

water level yielded a stage–discharge relationship, which was employed to convert the automatically recorded water level to discharge and establish a hydrograph that shows the variation in discharge with time. Hydrograph analysis can reveal characteristic properties of the aquifer (Palmer, 2007). Discharge data are essential during tracing experiments and are combined with water-chemistry data to estimate how much calcite is transported out of the system. According to ‘Murphy’s law’, there are inherent logistical problems in instrumentations like this, causing some breaks and lacunas in the record during the total period of observation.

The electrical conductivity (EC) is directly related to the total concentration of ions in the water. In karst aquifers, EC is dominated by ions related to calcium carbonate equilibria (e.g., Krawczyk & Ford, 2005; Massei et al., 2007). The main idea behind logging EC at the aquifer inlet and outlet was to evaluate how the EC changes through the aquifer and how the change in EC responds to recharge events.

However, some problems occurred because of sensor drift in the logger’s conductometer, as revealed through checking against a hand-held field instrument. The EC sensor at the spring broke down after three months of logging.

Water samples Water samples were collected at sites 0, 2 and 4 (Fig. 2) at intervals of a few hours, making up 14 batches (only 12 from site 2). Water samples at sites 0 and 4 were collected simultaneously with the discharge measurements.

Simultaneous water samples from three sites along the cave stream made it possible to evaluate the influence of the underground water flow on water chemistry. The water samples were analysed in the laboratory by an ion chromatograph (Waters Ion/Liquid Chromatograph ILC–1) to determine ion concentrations. Alkalinity, the estimate of bicarbonate HCO3–, was determined through conductometric titration with diluted hydrochloric acid, HCl. Total hardness (TH) is the sum of Ca2+ and Mg2+ expressed as mg CaCO3 equivalents per litre, which is conventional in the karst literature as it is easily transformed to the denudation rate of limestone (e.g., Ford & Williams, 2007). The saturation index of calcite (SIc) was calculated using the PCWATEQ computer program (Rollins, 1987). SIc = log (IAP/Kc), where IAP is the ion activity product and Kc is equilibrium constant of calcite. SIc is a measure of how fast the calcite is corroded. Positive values indicate that the water is supersaturated with respect to calcite and precipitation may occur, while negative values indicate undersaturation and that corrosion occurs. Hydrological monitoring data are combined with water-chemistry data from sporadic water samples to make a rough estimate of how much calcium carbonate is transported out of the aquifer.

Tracing experiments Quantitative dye-tracing experiments provide valuable information about the aquifer structure and conduit

65

volume (e.g., Field, 2002; Palmer, 2007). Two series of a total of 14 tracing experiments were performed in the aquifer using Rhodamine WT. Tracer was injected at five different locations along the aquifer length (sites 1–5, Fig. 2). Detection was made in the spring (site 0) with the withdrawal of water samples at one-hour intervals by a Manning Automatic Water Sampler. The water samples were analysed by a Turner Designs Field Fluorometer Model 10– AU–005. The instrument was calibrated before each batch of samples. Tracer breakthrough curves were analysed by the QTRACER program (Field, 2002), which interpolates curves to make them smoother and extrapolates missing recession. Breakthrough curves are expressed as equivalent to a 100 g injection. During the first series of seven experiments, the spring discharge started at less than 100 l s-1 and decreased to about 50 l s-1. During the second series, discharge started at about 100 l s-1 and increased to about 200 l s-1, then a rainstorm increased the discharge to about 400 l s-1 with a subsequent decrease to about 150 l s-1.

Tracer injection at five different sites along the aquifer Formulas length made it possible to evaluate different segments of the underground water system. The amount of retrieved tracer mass in the spring, Mout, is given by (Field, 2002): 

 =  

Formulas

(1) (1)

where C(t) is the concentration of tracer mass in the  water and Q(t) is the discharge during that same  = samples ∙100% (2)   =  (1)  interval. time Recovery, R, is reported as the percentage of recovered tracer mass with respect to the injected tracer mass, Min, (Field, 2002): Formulas 

 =  ∙ 100% 

 Formulas

Formulas

(3) (2) (2)



 (1) Calculation of recovery requires that the entire  =    breakthrough curve is detected.

Good recovery estimates    (3)(1)  =   good discharge estimates. depend oncorrespondingly  =  Combining tracer-recovery and -discharge data may =   (1) provide about additional tributaries and  = information ∙ 100% (2)  outlets that are not monitored, and if subsystems (4) are  =      connected to the main drainage line (Field, 2002).  =  ∙ 100% (2)    =  ∙ 100% -1 (2)  Flowvelocity, u (m h ), can be calculated from the tracer  = time, t, and the distance between the injection and (3) travel  =  (5)  =   sites, d (Field, 2002): (4) detection = (3)  (3) = (3)  T =Q (6)  spring (TH out − TH in ) A= crude estimate of the active conduit volume, V (m  (5)3),  canalso be calculated from the tracer data (Field, 2002):

  =   (4) (4) T = Q spring (TH out − TH in ) (6) where  =   is the mean discharge during the tracer(4) 3 -1 travel time (in(4) s). experiment  =   (in m s ) and t is the tracer 2 ), of the cave stream, The mean cross-sectional area, A (m = (5) 

=  = T = Q spring (TH out − TH in )

(5)

(5) (6)

 66

R.Ø. Skoglund. & S.-E. Lauritzen

 =   (4) or the water-filled conduit in the phreatic sections, is (Field, 2002):

 = (5) (5)

Results

T = Q spring (TH out − TH in ) The hydrographs and chemographs

(6)

The hydrographs from the cave inlet in Grønligrotta and the spring (Fig. 5E) show a perennial base-flow hydrological regime dominated by snowmelt during late spring and early summer, and frequent rainstorm floods during autumn and winter, in accordance with a subarctic, oceanic climate (Fig. 5A, E). The rainfall response is rapid due to the small, low-permeability catchment. The lowest flow rates (base flow) occur during winter, when the ground is frozen and the Strokbekken stream is fed mainly by bogs and snowmelt.

NORWEGIAN JOURNAL OF GEOLOGY

Mean annual discharge through the aquifer was 130 ± 50 l s-1 during the hydrological year 2000/2001.

This is about half of the annual runoff from the Strokbekken catchment. Base flow in a karst aquifer is generally supplied by water leaking from storage in less permeable parts of the aquifer, i.e., pores and fractures (Palmer, 2007). However, the hydrograph from the cave inlet shows that the surface stream feeds the aquifer throughout the year. The hydrograph from the spring seems to be more sensitive to variation during base flow and detects even lower discharges during long-lasting, dry periods. This implies that leakage from water stored in pores and fractures is negligible. Accordingly, the karst aquifer is totally dominated by conduit flow.

Transient storm-induced pulses are characterised by a rapid increase in discharge followed by a more gradual recession. Rapid changes in discharge are accompanied by rapid changes in water temperature (Fig 5C, E). The direction of the change is controlled by the surface

Date 2/1/01

4/1/01

6/1/01

8/1/01

0

40

40

80

80

120

120 30

B

C

Water temperature (°C)

10

D

-10 20

-30

10

0

60 40 20

10000

Discharge (l s-1)

E

0

1000 100 10 Cave inlet (site 4) Spring (site 0)

1 8/1/00

10/1/00

12/1/00

2/1/01

Date

4/1/01

6/1/01

8/1/01

Snow depth (cm)

12/1/00

Air temperature (°C)

Precipitation (mm)

10/1/00

Electric conductivity (µS cm-1)

0

A

8/1/00

Figure 5. Meteorological and hydro­ logi­ cal data from the Grønli–Seter aquifer. Data from the hydrological monitoring stations in the cave inlet, site 4 (red), and the spring, site 0 (black). (A) Precipitation data (black bars) and snow depth (grey bars) from Grønlia (Det norske meteoro­ logiske institutt, DNMI). (B) Air tempera­ture from the gauging station at the spring.

Water temperature (C), electric conductivity (D) and discharge (E) for each site. Breaks in curves­indicate periods of lost data (due to equipment malfunction).

NORWEGIAN JOURNAL OF GEOLOGY

Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway

from the cave inlet to the spring (one-tailed, paired t-test, p = 0), which is conspicuous in the cumulative-frequency plots of EC (Fig. 6). Comparison of the three-month and one-year cumulative curves from the cave inlet shows that the longer time span makes the curve less steep, i.e., the standard deviation is higher (6.4 and 8.1, respectively) while the mean value is only slightly higher for the entire year (24.1 and 24.9 μS cm-1, respectively).

100

Cumulative frequency

80 60 40 20

Site 4 - h yd .yr.

Site 4 - 3 m onths Spring - 3 m onths

0 0

10

20

30

40

50

60

Electric conductivity (µS cm-1)

Figure 6. Cumulative-frequency plot of the electric conductivity chemographs during a three-month autumn period at the cave inlet (dotted) and spring (dashed), and during one year at the cave inlet (solid).

temperature with an increase in water temperature during summer floods and a decrease during winter. As expected, the water temperatures in the cave inlet are more extreme than in the spring and display a stronger variability (Fig. 5C). During winter base flow, the aquifer raises water temperature from close to 0°C, to about 3°C, i.e., the mean annual air temperature. Slow water flow allows the water to reach thermal equilibrium with the subsurface environment.

The EC ranges between 5 and 56 μS cm-1 (Fig. 5D), as is typical for high latitudes with sparse vegetation (Lauritzen, 1981). The EC range is rather narrow, which is probably explained by the single-source water with limited solution capacity. There is a significant rise in EC

Both EC chemographs (Fig. 5D) show distinct responses to flood events by a marked drop in EC. Rainwater or meltwater in the recharge pulses have lower ion concentrations than in the base flow, as can be seen from both EC chemographs. During flood recession and base flow, the EC gradually increases.

Water chemistry Calcium, Ca2+, is the most abundant cation and bicarbonate, HCO3–, is the most abundant anion, as is to be expected in a carbonate karst aquifer (Fig. 7A, B).

All ions related to dissolution of marble (Ca2+, Mg2+ and HCO3–) show a significant rise in concentration through the aquifer, due to dissolution of calcite marble (positive, one-sample t-tests of the difference in concentration between sites 4 and 0, p<0.05). The Ca2+ concentration from the intermediate site (site 2) is significantly higher than in the cave inlet (site 4) and significantly lower than in the spring (site 0) (positive, one-sample t-tests of the difference, p<0.05; Fig. 8). This implies that dissolution occurs along the entire flow path. Ca2+, Mg2+ and HCO3– are all positively correlated with EC (Fig. 7A, B) (linear correlation, one-tailed, paired t-test, p<0.001). The total hardness, [Ca2+] + [Mg2+], varies in the water samples between 15 and 36 mg CaCO3 equivalents per litre.

These values are quite low, but normal for karst waters in Norway (Lauritzen, 1981). However, (autogenic) karst springs below the tree line tend to have higher values,

B

12

Ca2+

Anion concentration (ppm)

Cation concentration (ppm)

A

67

Mg2+ Na+ 8

4

0

40

HCO3 –

Cl –

SO42–

30

20

10

0 0

10

20

30

40

Electric conductivity (µS cm-1)

50

0

10

20

30

40

Electric conductivity (µS cm-1)

50

Figure 7. Cation (A) and anion (B) concentrations as a function of electric conductivity. Linear regression lines are shown for Ca2+, Mg2+ and HCO3–; they are all significant at a level of 0.05.

Formulas

68

R.Ø. Skoglund. & S.-E. Lauritzen

Ca2+ concentration (ppm)

12

8

4 Site Spring (0) Setergrotta (2) Cave inlet, Grønligrotta (4)

0 0

2

4

6

8

10

Set of water samples

12

14

Figure 8. Ca2+ concentrations of all 14 sets of water samples show how the concentration changes through the aquifer.

Discharge (l s-1) 10

100

1000

Saturation Index calcite

0

10000

Site 0 - spring

Site 4 - cave inlet -1

-2

-3

-4

Figure 9. Saturation index of calcite as a function of discharge.

 Transport of calcite

The transport rate of calcite out of the aquifer may serve as an estimate of when and how fast dissolution (4)  =  All sets of water samples except one showed occurs. an increase in TH from the cave inlet to the spring. No statistical relationship between spring discharge and TH is  detectable in this small sample set (Fig. 10A).  = instantaneous transport rates of calcite out of (5) The the aquifer, T (mg s-1), for each set of water samples can be estimated:

T = Q spring (TH out − TH in )

2

0

(6) (6)

where (THout–THin) is the increase in total hardness from the cave inlet to the spring, and Qspring is the spring discharge. The instantaneous transport rates of calcite out of the aquifer range between 7 and 2088 mg s-1 (Fig. 10B). A crude estimate for the transport rate corresponding to the mean annual discharge (~130 l s-1) would be in the range of 100 to 500 mg s-1, which is equivalent to 3 to 15 tons yr-1. The transport rates increase with increasing discharge because the discharge dominates this relationship despite the lack of a statistical relationship between change in TH and discharge. Under

Transport of calcite (mg s-1)

4

(1)

but as the (allogenic) catchment area of this system extends above the tree line and is sparsely vegetated, the contribution of organic-derived CO2 is probably low.   =  ∙ 100% (2)  In accordance with the low TH, water is always strongly undersaturated with respect to calcite though it is significantly less in the spring than in the cave inlet  (Fig. = 9). The saturation index decreases with increasing (3) discharge, in accordance with theory; a shorter contact time between water and rock under high flow rates should result in less saturated water at the effluent end.

B

6

-1

Increase in TH (mg CaCO3 l )

A

NORWEGIAN JOURNAL OF GEOLOGY



 =  

2500 2000 1500 1000 500 0

0

400

800

1200

Discharge (l s-1)

1600

0

400

800

Discharge (l s-1)

1200

1600

Figure 10. (A) The increase of total hardness (TH) between the cave inlet and the spring as a function of spring discharge. (B) Transport of calcite as a function of spring discharge.

NORWEGIAN JOURNAL OF GEOLOGY

Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway

higher discharge, more calcite is transported out of the aquifer, which implies that the aquifer is more efficiently corroded during flooding.

Tracing experiment All tracer experiments reveal single-peaked break­ through curves (e.g., Fig. 11A) which suggests that the flow path is not branched (Field, 2002; Palmer, 2007).

From the hydrological gauging stations, the discharge through the system has been evaluated to be roughly the same. Tracer-mass recovery has been estimated for all tracing experiments where the entire breakthrough curve was detected. The tracer recovery ranged from 54 to 90% (Table 2). Injections from the upper sites and at high discharge resulted in the highest recoveries, more than 80%. This implies that the traced sections represent the main flow path from the stream sink to the spring pool.

A

50

Site

In the lower section of the aquifer and at low discharge, the tracer recoveries were lower (50–70%). This indicates that part of the water may drain through one or several smaller outlets. Since the recovery is lower at lower discharge, the additional outlet is suggested to be a baseflow outlet of low flow capacity, i.e., with a low discharge under all flow conditions. As the recoveries from sites 1 and 2 are similar, the diversion towards the additional outlet is most likely located between sites 1 and 0. This outlet(s) may either be diffuse injection into the valley deposits and/or small concentrated springs. This may also explain why the base-flow hydrograph from the spring is lower than in the cave inlet.

Tracer travel time The time the tracer mass requires to traverse the aquifer is an essential parameter in estimating the aquifer

B

30 20

2750 m (5) 2350 m (4) 1850 m (3) 950 m (2) 400 m (1)

10

10

1 0

10

400

20

Time (h)

30

40

D

Distance to spring (site) 2750 m (5) 2350 m (4) 1850 m (3) 950 m (2) 400 m (1)

300

10

100

1000

Discharge (l s-1)

12000 10000

u = 0.94 Q r 2 = 0.98

Volume (m3)

Velocity (m h-1)

Distance to spring (site)

3 2 1

0

C

100

Time (h)

Concentration (ppb)

40

69

200

100

8000 6000 Distance to spring (site)

4000

2750 m (5) 2350 m (4) 1850 m (3) 950 m (2) 400 m (1)

2000 0

0 0

100

200

Discharge (l s-1)

300

400

0

100

200

300

400

Discharge (l s ) -1

Figure 11. Results from tracing experiments. (A) Example of tracer breakthrough curves from three different experiments with injections at sites 1, 2 and 3, with discharge in the range of 100–200 l s-1, and detection made at the spring (site 0). (B) Tracer peak time versus spring discharge in logarithmic scales. Results are shown for each site, symbols are similar in figure B–D. (C) Tracer velocity (based on tracer peak time) versus discharge. (D) Aquifer volume as a function of discharge. The dashed ellipses show the two groups of experiments upon which the volume estimates in Table 3 are based.

70

R.Ø. Skoglund. & S.-E. Lauritzen

NORWEGIAN JOURNAL OF GEOLOGY

Table 2. Results from tracer experiments.

a b

Site

Distance to spring (m)

Discharge (l s-1)

Peak time (h)

Volume (m3)

Recovery (%)

Tracer velocity (m h-1)

5

2750

70

25.7

6500a

54

110

5

2750

210

13.3

10100

88

210

4

2350

40

37.8

5400a

4

2350

330

7.3

8700

3

1850

70

25.6

6500

70

3

1850

50

35.2

6300

50

3

1850

180

3

1850

160

60 87

320

13.0

8400

b

90

140

12.1

7000

71

150

63

2

950

70

14.8

3700

2

950

60

19.4

4200

50

60

2

950

120

9.0

3900

110

2

950

140

9.4

4700b

62

100

1

400

60

8.7

1900

65

50

1

400

110

4.0

1600

59

100

Comparing tracer peak time with discharge suggest that the discharge estimate is too low, which underestimates the volume.

Comparing tracer peak time with discharge suggest that the discharge estimate is too high, which overestimates the volume.

similar flow velocities under the same discharge, which implies that the velocity is similar in the vadose and phreatic segments, at least for discharges below 350 l s-1.

volume and mean cross-sectional area. The tracer breakthrough time is the time of the first arrival of the tracer mass in the spring (Fig. 11A). A better estimate of the overall water residence time is peak time, which is the time of arrival of the highest concentration of tracer mass. This is closer to the mean travel time of the tracer mass but more robust than the mean tracer residence time based on arrival of the centroid of the tracer mass (Field, 2002). The centroid of the tracer mass depends on detection of the entire breakthrough curve where the decrease is often unreliable. The tracer travel time is a function of distance and discharge. Shorter distance and higher discharge give shorter flow times (Table 2). In the discharge–tracer-time diagram, the overall trend in the plot shows this inverse relationship (Fig. 11B). Ideally, the plots should lie along negative trend lines lying above each other as the distance to the spring increases. Based on tracer peak times, we may suggest that some of the discharge estimates are too high (site 2: Q = 140 l s-1, and site 3: Q = 180 l s-1), while others are too low (site 4: Q = 40 l s-1 and site 5: Q = 70 l s-1).

Volume and mean cross-sectional area have been estimated for each aquifer section (Fig. 11D, Table 3). Tracing experiments from sites 1 and 2 were only performed at low discharge (less than mean annual discharge). The mean volume estimate for this section (sites 2–0) is 4100 m3, which corresponds to a mean cross-sectional area of about 4.3 m2. However, the tracer recoveries were low (around 60%) in these experiments.

If only part of the water flows towards site 0, the volume estimates only account for the volume of the active water flow towards this site. Consequently, the actual conduit size of this section, which comprises essentially phreatic conduits (Table 1), may be as much as 40% larger. Accordingly, the mean cross-sectional area of the conduits may possibly reach about 7 m2.

Average flow velocities, u, estimated from peak time are in the range of 100 to 200 m h-1 for flow rates in the range of the mean annual flow (0.1–0.2 m3 s-1; Table 2).

The average flow velocity is closely related to discharge (Fig. 11C). The outliers in the discharge–velocity plot represent experiments where discharge was poorly estimated (see previous section). Tracing experiments over different sections of the aquifer (i.e., with different ratios of vadose and phreatic conditions) seem to have

Under the base-flow regime, the volume estimates from sites 3 and 5 are similar. As previously mentioned, the discharges during the base-flow tracing experiments from sites 4 and 5 are suspected to be underestimated, which means that the volume estimates are also too low.

Accordingly, no volume estimate exists for the upper part of the aquifer (sites 5–3) under base-flow conditions, though we may assume that the volume is quite small.

Under these conditions, the volume estimate for the

Aquifer volume

NORWEGIAN JOURNAL OF GEOLOGY

Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway

71

Table 3. Estimated volume and cross-sectional area of the aquifer sections.

Discharge <100 l s-1 Site

Distance (m)

5–4

400

4–3

500

3–2

900

2–1 1–0

Volume (m3)

Discharge 100 – 250 l s-1

Cross-sectional area (m2)

Volume (m3)

Cross-sectional area (m2)

a

2400b

2.7

2400

2.7

3400

3.7

550

2200

4.0

400

1800

4.5

4300

4.5

a

Using the volume estimate of 6500 m3 from site 5 during low discharges would give a cross-sectional area of 1.4 m2 for section 5–2, which probably is an underestimate. b Maximum volume estimate for section 5–3 is used.

middle part of the aquifer (sites 3–2) is 2400 m3 with a corresponding mean cross-sectional area of 2.7 m2 (Table 3).

The aquifer volume increases with increasing discharge (Fig. 11D). The largest volume estimate of the entire aquifer is about 10,000 m3 (discharge ~0.2 m3 s-1). In the lower section between sites 2 and 0, the volume increase seems to be moderate. The middle section shows a stronger increase in volume and a corresponding mean cross-sectional area. Comparison of the three aquifer sections under moderate discharge (0.1–0.25 m3 s-1) demonstrates that the volume and mean cross-sectional area increase in the downstream direction in accordance with the increasing phreatic component.

Discussion and conclusion Aquifer structure and aquifer volume The Grønli–Seter aquifer is an invasion aquifer where most of the aquifer structure and volume is inherited.

Large phreatic conduits below the present valley bottom and watertable (between sites 2 and 0) give the aquifer a large static volume. The aquifer consists essentially of one single, unbranched flow path, fed by a major stream sink, and with water leaving through a major spring pool.

Tracer recovery data and base-flow hydrographs suggest that another outlet exists between site 1 and the spring at site 0. This might be a perennial, base-flow spring of low flow capacity, making most impact during low discharge.

The floodwater zone in Setergrotta has a maximum amplitude of 11 m. During flood recession the water level in the flood-water passages in Setergrotta is gradually reduced. Tracing experiments from sites 3, 4 and 5 performed under higher flow rates are expected to show some loss of tracer mass to the perched sump that fills up under rising discharge. However, these tracing experiments have the highest mass recoveries, which are hard to explain unless the main flow path has a poor

connection to the sump. The recoveries are below 100%, so one possible interpretation is that the perched sump fills up slowly when the flow rates are in the range of a few hundred litres per second, which results in a small loss of tracer mass. No release of tracer mass at a later time has been detected in the breakthrough curves (i.e., the systemes annexes effect, Ford & Williams, 2007).

Chemical state The low values of total hardness in the cave waters are in accordance with data from other Norwegian resurgences in similar settings (Lauritzen, 1981).

Stripe karst is an extreme end-member of contact karst (Lauritzen, 2001). Allogenic water and a short residence time between the water and the karstic rock causes low total hardness values characteristic of the Norwegian stripe karst, even below the tree line, where soil gives the water a higher solution potential through its higher carbonic-acid content (Lauritzen, 1981). The water is undersaturated with respect to calcite throughout the aquifer, and simultaneous water samples from various sites along the aquifer length (sites 4, 2 and 0) reveal that the Ca2+ concentration increases through the aquifer, demonstrating that dissolution occurs in both the upper and the lower section of the aquifer, without regard to the different volume and flow conditions.

Instantaneous rates of calcite transport show that even though there is no statistical relationship between increase in TH through the aquifer and spring discharge, the transport of calcium carbonate out of the aquifer will increase with increasing discharge, making higher discharges more efficient in transporting calcium carbonate out of the aquifer. The transport rate of calcium carbonate is the result of dissolution and wall retreat within the aquifer and thus indicates which discharges are the most corrosive. Our result is in accordance with previous studies from the area (Lauritzen, 1989), where the highest rates of wall retreat occur during flooding with a duration of less than 20%

72

R.Ø. Skoglund. & S.-E. Lauritzen

of the time, and the scallop-dominant discharge (i.e., the discharge that dissolves the flow marks in the cave walls) is related to floods with a duration between 2 and 15% of the time. Studies with high-resolution flow and chemical data also show that intense, short-duration events are more important in enlargement of the karst aquifer (Groves & Meiman, 2005).

Invasion aquifers The Grønli–Seter invasion aquifer serves as a representative for many Norwegian karst aquifers, and possibly other karst aquifers in formerly glaciated areas.

The Okshola cave system in Fauske is a multiphase cave system that shows evidence of development through several glacial–interglacial cycles (Skoglund & Lauritzen, 2010). Abandoned, vadose passages at different elevations up in the hillside, and a large, active canyon (10–30 m deep) below the marine limit, show that the Okshola cave has developed through several interglacials and interstadials. In the Grønli–Seter cave system, on the other hand, the vadose canyon is quite small and the active water path between the caves seems to be immature. The difference between the Okshola and the Grønli–Seter aquifers is related to the size of their drainage areas, the discharge and how long the marble outcrop has been exposed in a position where it captures channel water flow. The Grønli–Seter aquifer is situated in a smaller catchment where it only captures part of the runoff during higher flow rates. It resembles the upper, small and abandoned, vadose levels of Okshola that are likely to have captured only part of the surface-stream flow, in contrast to the contemporary underground streamway which captures the entire surface stream. The present study shows that the instantaneous transport rates of calcium carbonate out of the Grønli–Seter aquifer are quite low, which means that widening of the underground flow path is quite slow, and that the highest rates occur during the highest floods. This has also been shown in several other studies (e.g., Lauritzen, 1989; Groves & Meiman, 2005). Groves & Meiman (2005) suggested that karst areas where rainfall is concentrated evolve faster than areas with similar precipitation rates, but where rainfall is more evenly spread throughout the year. Likewise, we may suggest that karst aquifers where the stream sink captures the entire surface flow (and thus the full extent of floods) evolve more quickly than aquifers with a similar mean discharge but fed only by part of the flow of a surface stream (thus missing the peak floods). Karst systems that capture only a part of the water flow from a surface stream are likely to miss a larger proportion of the flood discharges than the base and middle flows. Since the Grønli–Seter invasion aquifer captures only a part of the surface channel flow, it misses much of the flow during flooding when corrosion (and also abrasion) is highest. This probably leads to a slower evolution of the underground flow paths that capture only part of a surface stream flow in contrast to those that capture the entire surface flow, like the contemporary

NORWEGIAN JOURNAL OF GEOLOGY

streamway in the Okshola cave. In conclusion, we suggest that the evolution of invasion aquifers in settings like this is quite slow, but is likely to accelerate if the recharge settings are changed (e.g., by glacial erosion) so that the entire surface stream is captured.

Acknowledgments. This paper is based on the Cand. scient. thesis written by R. Ø. Skoglund (Øvrevik, 2002). The fieldwork, cave survey and drawing of the cave maps were carried out together with Sara Skutlaberg and Hilde Hestangen who also wrote their Cand. scient. theses on other aspects of the Grønli–Seter cave systems (Skutlaberg, 2003; Hestangen, 2005). We would like to thank the owners of Grønligrotta, Bjarne and Milda Pedersen, for free accommodation during the fieldwork, and Setergrotta AS and Per Gunnar Hjorthen for making a free car available to us. Some financial support was provided

by Bergen Myrdyrkningsforenings fond. We also wish to thank the cave divers, Raymond Nielsen and Hallvard Winterseth, for survey data and description of submerged passages. Several local cavers and family members assisted during the cave survey and we are grateful to all of them. Chris Smart and an anonymous referee are thanked for constructive comments that strong ly improved the manuscript.

References Bakalowicz, M. 1984: Water chemistry of some karst environments in Norway. Norsk Geografisk Tidsskrift 38, 209–214.

Benn, D.I. & Evans, D.J.A. 2010: Glaciers and glaciations. Hodder Education, London, 802 pp.

Curl, R.L. 1974: Deducing flow velocity in cave conduits from scallops.

National Speleological Society Bulletin 36, 1–5.

Field, M.S. 2002: The QTRACER2 program for tracer-breakthrough curve analysis for tracer tests in karstic aquifers and other hydrologic systems. U.S. Environmental Protection Agency, Washington DC, 179 pp.

Ford, D.C. & Williams, P. 2007: Karst hydrogeology and geomorphology.

John Wiley & Sons Ltd, Chichester, 562 pp.

Groves, C. & Meiman, J. 2005: Weathering, geomorphic work, and karst landscape evolution in the Cave City groundwater basin, Mammoth Cave, Kentucky. Geomorphology 67, 115–126.

Hestangen, H. 2005. The sedimentology of the Grønli–Setergrotta cave system, Mo i Rana, Nordland, Norway. Cand. scient. thesis, University of Bergen, 190 pp.

Hongve, D. 1987: A revised procedure for discharge measurement by means of the salt dilution method. Hydrological Processes 1, 267– 270.

Horn, G. 1947: Karsthuler i Nordland. Norges geologiske undersøkelse 165, 77 pp.

Knutson, G. 2008: Hydrogeology of the Nordic countries. Episodes 31, 148–154.

Krawczyk, W.E. & Ford, D.C. 2005: Correlating specific conductivity with total hardness in limestone and dolomite karst waters. Earth surface processes and landforms 31, 221–234.

Lauritzen, S.-E. 1981: A study of some karst waters in Norway. Spatial variation in solute concentrations and equilibrium parameters in limestone dissolution. Norsk Geografisk Tidsskrift 35, 1–19.

Lauritzen, S.-E. 1986: Hydraulics and dissolution kinetics of a phreatic conduit. In Proceedings of the 9th International Speleological Congress, Barcelona, pp. 20–22.

Lauritzen, S.-E. 1989: Scallop dominant discharge. In Proceedings of the 10th International Speleological Congress, Budapest, pp. 123– 124.

Lauritzen, S.-E. 2001: Marble stripe karst of the Scandinavian Caledonides: An end-member in the contact karst spectrum. Acta Carsologica 30, 47–79.

NORWEGIAN JOURNAL OF GEOLOGY

Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway

Lauritzen, S.-E. & Lundberg, J. 2000: Solutional and erosional morphology of caves. In Klimchouk, A., Ford, D.C. & Palmer, A.N. (eds.): Speleogenesis: Evolution of Karst Aquifers, National Speleological Society, Huntsville, Alabama, pp. 406–426.

Lauritzen, S.-E. & Skoglund, R.Ø. 2013: Glacier ice-contact speleogenesis in marble stripe karst. In Frumkin, A. (ed.): Karst Geomorphology. In Shroder, J. (ed.): Treatise on Geomorphology 6, AcademicPress, San Diego, pp. 363–396.

Lauritzen, S.-E., Abbott, J., Arnesen, R., Crossley, G., Grepperud, D., Ive, A. & Johnson, S. 1985: Morphology and hydraulics of an active phreatic conduit. Cave Science 12, 139–146.

Massei, N., Mahler, B.J., Bakalowicz, M., Fournier, M. & Dupont, J.P. 2007: Quantitative interpretation of specific conductance frequency distributions in karst. Ground Water 45, 288–293.

NVE, 2002: Runoff map of Norway, sheet 5, scale 1:500,000. Norges vassdrags- og energidirektorat.

Olsen, L., Sveian, H., van der Borg, K., Bergstrom, B. & Broekmans, M. 2002: Rapid and rhythmic ice sheet fluctuations in western Scandinavia 15–40 Kya—a review. Polar Research 21, 235–242.

Oxaal, J. 1914: Kalkstenshuler i Ranen. Norges geologiske undersøkelse Aarbok 1914 II, 1–41.

Palmer, A. 2007: Cave geology. Cave Books, Dayton, Ohio, 454 pp.

Rollins, L. 1987: PCWATEQ. Computer program. Woodland, California, Shadoware.

Sejrup, H.P., Larsen, E., Landvik, J., King, E.L., Haflidason, H. & Nesje, A. 2000: Quaternary glaciations in southern Fennoscandia: evidence from southwestern Norway and the northern North Sea region. Quaternary Science Reviews 19, 667–685.

Skoglund, R.Ø. & Lauritzen, S.E. 2010: Morphology and speleogenesis of Okshola (Fauske, northern Norway): example of a multi-stage network cave in a glacial landscape. Norwegian Journal of Geology 90, 123–137.

Skutlaberg, S. 2003. Paleohydrogeologi, bruddgeometri og litostratigrafi i Grønli–Setergrottesystemet, Mo i Rana. Cand. scient. thesis, University of Bergen, 188 pp.

St. Pierre, S. 1988: Morphology and sediments of the Grønli–Seter Caves, Norway. Cave Science 15, 109–116.

Stephens, M.B., Gustavson, M., Ramberg, I.B. & Zachrisson, E. 1985: The Caledonides of central-north Scandinavia–a tectonostratigraphic overview. In Gee, D.G. & Sturt, B.A. (eds.): The Caledonide Orogen–Scandinavia and Related Areas, John Wiley & Sons Ltd., pp. 135–162.

Søvegjarto, U., Marker, M., Graversen, O. & Gjelle, S. 1988: Mo i Rana, bedrock geology map 1927 I, scale 1:50,000, Norges geologiske undersøkelse.

Søvegjarto, U., Marker, M. & Gjelle, S. 1989: Storforshei, bedrock geology map 2027 IV, scale 1:50,000, Norges geologiske undersøkelse. Øvrevik, R. 2002. Hydrogeologi og karstkorrosjon i Grønli– Seterakviferen, Mo i Rana. Cand. scient. thesis, University of Bergen, 169 pp. Øvstedal, J. & Lauritzen, S.-E. 1989: The Sirijorda karst aquifer, Nordland, Northern Norway. In Proceedings of the 10th International Speleological Congress, Budapest, pp. 121–122.

73

74

NORWEGIAN JOURNAL OF GEOLOGY

Norwegian Journal of Geology Norsk Geologisk Tidsskrift The main journal for Norwegian geological research, distributed to all members of the Norwegian Geological Society. First published in 1905, this geological journal is issued quarterly by the Norwegian Geological Society (Norsk Geologisk Forening). The journal publishes research articles, review articles, papers, notes and discussions relevant to the varied and complex geology of Norway, the Arctic regions, and adjacent offshore areas.

Scientific papers from all geological, geophysical and geo­chemical disciplines are considered for publication.

The contributors­, referees and readership are widely international.

Instructions to authors Instructions to authors are found on www.geologi.no/njg

Indexed in

Trond Slagstad, Trondheim

Biological Abstracts (-1985); British Geological Literature; Chemical Abstracts; Current Contents/ Physical, Chemical and Earth Sciences; Energy Research Abstracts; Geological Abstracts Economic Geology; Geographical AstractsI/ Physical Geography; GeoREF (Online Database for Bibliography and Index of Geology); Life Sciences Collection; Norske Tidsskriftartikler; Metal Abstracts; Mineralogical Abstracts; Petroleum Abstracts; Science Citation Index; World Aluminum Abstracts

Editorial Board

Subscription price

Silje Berg and Christine Fichler, Statoil, Terje Bjerkgård, NGU, Henriette Linge, UiB, Sverre Planke, VBPR, Anders Solheim, NGI.

Subscription rates 2013: EUR 280 In Scandinavia: NOK 1600

Editor

Manuscript address Manuscripts shall be addressed to: Editor of Norwegian Journal of Geology N-7491 Trondheim, Norway Tel: +47 73 90 40 00, Fax: +47 73 92 16 20 E-mail: trond.slagstad@ngu.no

P

RI

NT

R

241

731

Graphic production by Skipnes AS

T E D M AT

E

Front cover The cave stream in the main passage in Grønligrotta at base flow.

Photo by Stein-Erik Lauritzen.

B-Economique B-Blad 2013 Number 1, Volume 93

Return: NGF c/o NGU NO–7491 Trondheim NORWAY

Norwegian Journal of Geology

Per Inge Myhre, Fernando Corfu, Steffen G. Bergh & Kåre Kullerud

www.geologi.no/njg

P.P. Norge/Norvège

No. 1, Vol. 93

NORWEGIAN JOURNAL OF GEOLOGY

Porto betalt ved innleveringen

U–Pb geochronology along an Archaean geotransect in the West Troms Basement Complex, North Norway ................................. 1

Raymond S. Eilertsen, Nils R.B. Olsen, Nils Rüther & Peggy Zinke Channel-bed changes in distributaries of the lake Øyeren delta, southern Norway, revealed by interferometric sidescan sonar ............ 25

Abdus Samad Azad, Henning Dypvik, Fridtjof Riis & Elin Kalleson Late post-impact sedimentation in the Ritland impact structure, western Norway .................................................................... 37

Rannveig Øvrevik Skoglund & Stein-Erik Lauritzen Characterisation of a post-glacial invasion aquifer: the Grønli–Seter karst system, northern Norway .............................. 61

ISBN 978-82-92-39481-6

ISSN 0029-196X ISBN 978-82-92-39481-6

9 788292 394816

2013 Number 1 Volume 93

NORWEGIAN JOURNAL OF GEOLOGY Norsk Geologisk Tidsskrift