Tectonophysics xxx (2015) xxx–xxx

TECTO-126523; No of Pages 13

Contents lists available at ScienceDirect

Tectonophysics

j ourna l homepage: www.e lsev ie r .com/ locate / tecto

Electrical conductivity structure of north-west Fennoscandia fromthree-dimensional inversion of magnetotelluric data

M. Cherevatova a,⁎, M.Yu. Smirnov a, T. Korja a, L.B. Pedersen b, J. Ebbing c, S. Gradmann d,M. Becken e, MaSca Working Groupa University of Oulu, Finlandb Department of Earth Sciences, Geophysics, Uppsala University, Uppsala, Swedenc Kiel University, Germanyd Geological Survey of Norway, Norwaye The University of Münster, Germany

⁎ Corresponding author.E-mail address: maria.cherevatova@oulu.fi (M. Cherev

http://dx.doi.org/10.1016/j.tecto.2015.01.0080040-1951/© 2015 Elsevier B.V. All rights reserved.

Please cite this article as: Cherevatova,M., etmagnetotelluric data, Tectonophysics (2015

a b s t r a c t

a r t i c l e i n f o

Article history:Received 16 July 2014Received in revised form 15 January 2015Accepted 19 January 2015Available online xxxx

Keywords:Magnetotellurics3-D inversionElectrical conductivityFennoscandian ShieldCaledonides

Newmagnetotelluric (MT) data in north-west Fennoscandia were acquired within the framework of the project“Magnetotellurics in the Scandes” (MaSca). The project focuses on the investigation of the crustal and uppermantle lithospheric structure in the transition zone from stable Precambrian cratonic interior to passive conti-nental margin beneath the Caledonian orogen and the Scandinavian Mountains in western Fennoscandia. Anarray of 59 simultaneous long period and 220 broad-band MT sites were occupied in the summers of 2011 to2013.The 3-D inversion of theMaSca datawas obtained using theModEM3-D code. The full impedance and tipper datawere used for the inversion. The rocks of Archaean and Proterozoic basement towards east and the Caledoniannappes towards west are modelled as resistive structures. In the central and southern parts, the whole crust isresistive and reflects the Trans-Scandinavian Igneous Belt granitoids. The middle to lower crust of theSvecofennian province is conductive. An uppermost crustal conductor is revealed in the Skellefteå Ore District.The south end of the Kittilä Greenstone Belt is seen in the models as a strong upper to middle crustal conductor.In the Caledonides, the highly conductive alum shales are observed along the Caledonian Thrust Front. A map ofthe crustal conductance for the north-west Fennoscandian Shield is presented.

© 2015 Elsevier B.V. All rights reserved.

1. Introduction

The “Magnetotellurics in the Scandes” (MaSca) project targets atdevelopment and application of the magnetotelluric (MT) method tostudy the Earth's structure in north-west Fennoscandia. The MaScaarray covers an area from E13 to E23 and N64 to N69. The area ofinvestigation consists of the Archaean Domain, northern Svecofennianvolcanic belt and the Scandinavian Caledonides. The array crosses twoimportant boundaries: the Archaean–Proterozoic boundary and theCaledonian Thrust Front (Fig. 1).

Prior to this project, there were no MT data acquired in westernFennoscandia, except for two recent MT profiles in southern Norway(Cherevatova et al., 2014) and across the central Scandinavian Moun-tains (Korja et al., 2008) (Fig. 1c). These studies have revealed highlyconductive alum shales between the resistive Proterozoic basementand the overlying Caledonian nappes. Airborne electromagnetic datafrom Finland indicate the alum shales formations extend to the north(Korja et al., 2002). Hence, it is likely that the Caledonides are underlain

atova).

al., Electrical conductivity stru), http://dx.doi.org/10.1016/j.

by the alum shales everywhere. A number ofMT surveys have been con-ducted in the rest of Fennoscandia (Agustsson, 1986; Hjelt et al., 2006;Korja, 2007; Korja et al., 1989, 2008; Lahti et al., 2005; Rasmussen,1988; Rasmussen et al., 1987) and its margins (Brasse et al., 2006;Jones, 1983; Smirnov and Pedersen, 2009). The Baltic ElectromagneticArray Research (BEAR) project was an international experiment fordeep electromagnetic sounding. Earlier studies together with theBEAR data allowed for compiling a map of integrated conductance ofthe Fennoscandian Shield (Korja et al., 2002). The map revealed largevariations in conductance ranging from a few Siemens to tens ofthousands of Siemens. The crust is generally resistive with a few highlyconducting belts: Skellefteå and Kittilä conductors (with the totalconductance of more than 1000 S) (Fig. 1). The MaSca array crossesonly parts of these conductors and further extension of the array isplanned to the north-east and south.

Unfortunately, north-west Fennoscandia is not well studied by seis-mic methods. Therefore, a comparison of the electric and velocitymodels is not currently possible. In Sweden, the European Geotraverseand related experiments provide information on the central part of theFennoscandian Shield (Guggisberg et al., 1991; Lund and Heikkinen,1987). The results show a thick crust below the low topography of the

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

Mo i Rana

Kiruna

Legend

Plutonic rocks and gneisses (3500-2500 Ma)

Plutonic rocks (1850-1660 Ma)Plutonic rocks (1960-1840 Ma)

Sedimentary and volcanic rocks (2500-1960 Ma)

Sandstone and shale

Sedimentary cover (Devonian)

Middle Allochthon (Mesoproterozoic - Cambrian)basement nappes of Baltic originUpper Allochthon (Neoproterozoic - Silurian)greenstones, gabbros & mica schistsfrom the Iapetus OceanUppermost Allochthon (Neoproterozoic - Ordovican)granites, schists & volcanic rocks of exotic originLower Allochthonsandstones, phyllites & mica schists from theperiphery of the Baltic basement Basement variably affected by the orogeny(Archaean and Proterozoic rocks within the Caledonides,basement windows)

Caledonian Thrust Front

14˚

14˚

16˚

16˚

18̊

18˚

20˚

20˚

22˚

22˚

24˚

24˚ 26˚

64˚

65˚

66˚

67˚

68˚

69˚

70˚

26˚

200 km

Sedimentary and volcanic rocks (1950-1800 Ma)(Vargfors group)

Archaean

Palaeoproterozoic

Neoproterozoic and Mesoproterozoic

Caledonides

Volcanic and intrusive rock belonging to TIB and Revsund-Sorsele suite (1650-1820 Ma)

Devonian extensional shear zones

Sagfjorden Shear Zone

Normal faults

Tromsø

Bodø

Lofote

n

Kiruna

Luleå

Bothnian

Bay

OuluArvidsjaur

Jokkmokk

Nattavaara

Skellefteå

10˚

10˚

20˚

20˚

30˚

30˚

40˚

40˚

56˚

60˚

64˚

68˚

500 km

4˚ 8˚ 12˚ 16˚ 20˚ 24˚ 28˚58˚ 58˚

60˚ 60˚

62˚ 62˚

64˚ 64˚

66˚ 66˚

68˚ 68˚

70˚ 70˚

500 km

0 500 1000 1500 2000Elevation

m

(a) (b)

(c)

Metavolcanic rock and mica shist, the Kittilä allochthon

Kittilä

Fig. 1.Main geological units of the Caledonian orogen and the Precambrian basement in thenorth-west Fennoscandia. (a) Simplifiedgeologymap. Themap is compiled and simplified after(Gorbatschev and Bogdanova, 1993; Ramberg et al., 2008) and (Ebbing et al., 2012). (b)Main tectonic domains in Fennoscandia (modified from (Koistinen and et al, 2001)). (c) Elevationmap: black circles—MaSca array, blue— Jämtland–TrøndelagMTprofile (Korja et al., 2008), green— ToScaMT profiles (Cherevatova et al., 2014), red line— seismic SCANLIPS experiment(Ebbing et al., 2012),magenta line— the BlueRoad seismic profile (Lund, 1979). (For interpretation of the references to colour in thisfigure legend, the reader is referred to theweb versionof this article.)

2 M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

central Fennoscandian Shield, which is not in agreement with the ob-served gravity signal (Ebbing et al., 2012). The Blue Road profile acrossthe northern Scandinavian Mountains (Lund, 1979) is located some100 km to the south from MaSca array (Fig. 1c). Lund (1979) statedthat there is evidence for a shear (S) wave low-velocity zone of a fewtens of kilometres in the uppermost mantle. Therefore, an extension oftheMaScameasurements is planned along the BlueRoad seismic profile.In addition, results from the SCANLIPS-2 seismic profile in northernNorway (Fig. 1c) will be available in the coming years and facilitatemodelling of the processes that shaped the topography of the Scandina-vian Mountains (Ebbing et al., 2012).

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

In this paper we present the first 3-DMTmodel of the crust beneaththe Archaean and Svecofennian Domains and the Caledonides. The MTmethod is sensitive to resistivity contrasts, therefore the model is veryinformative in mapping resistive basement features and conductingmineral phases. The possible causes of enhanced electrical conductivityin themiddle and lower crust are salinefluids, graphite or carbon grain-boundary films andpartialmelts. (Korja et al., 2008) described, that par-tial melts are unlikely to be present in the Fennoscandian middle andlower crust. Despite the advantages of 3-D inversion (see later inSection 2) we always undertake prior 2-D inversion. The detailed de-scription of the strike and dimensionality analysis of the MaSca data,

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

3M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

together with the 2-D inversion models is given in the parallel paper(Cherevatova et al, 2015). Here we present a 3-D model for the MaScaarray and compare the results with the 2-D interpretations.

1.1. Geological background

The crust beneath theMaSca array consists of three distinct units viz.the Archaean Domain (Saamian orogeny 3.1–2.9 Ga, Lopian orogeny2.9–2.6 Ga) in the north-east, the Svecofennian Domain (2.0–1.75 Ga)in the east, and Caledonian orogen (0.6–0.4 Ga) in the west (Gaál andGorbatschev, 1987) (Fig. 1b).

The ArchaeanDomain is a typical Neoarchaean granitoid-greenstoneprovince consisting of granitoid gneiss complexes and supracrustalrocks ranging in age between 3.1–2.6 Ga. The Archaean rocks are over-lain by Palaeoproterozoic cover rocks deposited on Archaean basementsince 2.45 Ga. In the north-western part of the Archaean Domain(northern Finland and Sweden), large areas are intruded by 1.9–1.8 Ga old plutonic rocks (Korja et al., 2002).

The central part of the Fennoscandian Shield is occupied by theSvecofennian Domain of Palaeoproterozoic continental crust. NorthernSvecofennian volcanic belt (Gorbatschev and Bogdanova, 1993) locatedin the north-east, along the southern edge of the Archaean Domain. Thevolcanic belt consists of Paleoproterozoic supracrustal rocks of partlyrecycled Archaean rocks with an inferred Archaean basement (Gaáland Gorbatschev, 1987).

There is no distinct margin between the Archaean and Proterozoiccrust, but rather a combination of Archaean and Proterozoic materialin a belt some 100 km wide called Luleå–Jokkmokk Zone (Fig. 1b)(Juhlin et al., 2002). The area from the Luleå–Jokkmokk Zone toSkellefteå district is considered to represent the remnants of an islandarcwith a change from terrestrial lithologies in the north to shallowma-rine in the Skellefteå district. The final major tectonic event in the areawas the intrusion of 1.8 Ga Trans-Scandinavian Igneous Belt (TIB)Revsund granitoids. A more north–south-striking subduction zonefurther to thewestmay have provided the heat necessary for the gener-ation of these granitic melts (Juhlin et al., 2002).

The 40mWm−2 heat-flow isoline runs roughly parallel to the Luleå–Jokkmokk Zone, generally through Svecofennian crust (Nyblade andPollack, 1993).Within the SvecofennianDomain, the heat-flow increasesgradually south-westwards, from40mWm−2 in the north-east to nearly60mWm−2 in the south-west. The heat-flowmap shows a rather sharpincrease from ca. 40 to 50 mWm−2 going from Archaean toSvecofennian crust, whereas the heat-flow is rather stable around50 mWm−2 in the rest of the Svecofennian Domain. The Caledonidesyield an average heat-flow of 58 ± 9 mWm−2 (Slagstad et al., 2009).

The Scandinavian Caledonides (Fig. 1) were formed as the result of aclosure of the Iapetus Ocean and a continental collision of Baltica andLaurentia in the Late Silurian (Ramberg et al., 2008). In the Caledonianorogeny (540–400 Ma), the Precambrian rocks in the western marginof Baltica were thrust beneath Laurentia to ultra-high pressure depths.The underthrust rocks of the Baltica were heated, metamorphosed anddeformed (Andersen, 1998) whereas the rocks of the Neoproterozoicto Early Palaeozoic accretionary wedge were transported to the east/north-east over Baltica as the Caledonian nappes. The Caledoniannappes are generally divided into lower, middle, upper and uppermostallochthons (nappe series) (Ramberg et al., 2008). The major nappethrusting was completed between 400 and 405 million years ago.Today the remnants of the Caledonian orogen are preserved as a rela-tively thin cover (10–12 km, (England and Ebbing, 2012)) above thePrecambrian basement which stretches north-east from Ireland andScotland in the south to Svalbard in the north (Ramberg et al., 2008).

In the Devonian, the Caledonianmountainswere eroded. The exten-sional collapse of the ScandinavianCaledonides resulted in rapid tecton-ic denudation of the orogen, exhumation of high- to ultra-high-pressuremetamorphic rocks and provided a structural template for the forma-tion of a Devonian supra-detachment sedimentary basin (Andersen,

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

1998). The extensional shear zones and fabrics indicate W- to NW-di-rected translations (Gee et al., 2008). Extension-related structures inthe north are not as well studied as in central and southern Norway.No major extensional detachment zones have been documented innorthern Norway. The possible existence of a large extensionaldecollement zone west of the present coastline in the Lofoten–Vesterålen Archipelago (Lofoten, hereafter) has been suggested(Fossen, 2010). During the Cenozoic, Scandinavia was uplifted twiceand mountains were formed once more (the Scandinavian MountainRange— the Scandes). Hence, the Scandes represents the uplifted rem-nants of the Caledonides and the surface geology can be referred to asthe Caledonides (Chalmers et al., 2010; Ebbing et al., 2012).

The Precambrian rocks along the coast of northernNorway are prob-ably a westerly extension of Archaean and Palaeoproterozoic basementcomplexes. On themainland, the basement rocks are concealed beneaththe Caledonian nappes. The oldest dated rocks consist of tonalite andtonalitic gneiss occurring on islands north of Tromsø. These rocks arethe oldest rocks in Norway (2.88 Ga, (Ramberg et al., 2008)). Many ofthe well-known mountainous areas in northern Norway consist ofPrecambrian rocks which form tectonic windows in the Caledonides.The bedrock in thesewindowswasuplifted and locally deformedduringcrustal extension in the Devonian. The basement in the 160 km-longLofoten reaches far above sea level due to uplift in the period followingthe Caledonian orogeny along a series of faults mostly trending north-east–south-west. The landforms in Lofoten are controlled by faults relat-ed to later crustal extension. Narrow fjords and straits between theislands are oriented north-east–south-west to north–south and followfault zones where the bedrock is crushed. The rocks of Lofoten are ap-proximately as old as those in western Tromsø (Ramberg et al., 2008).

2. 3-D inversion

Three-dimensional inversion has rapidly developed during the lastyears (Egbert and Kelbert, 2012; Siripunvaraporn and Egbert, 2009).There are a fewmotivations for 3-D inversion instead of 2-D. The 3-D in-version provides more realistic images of the 3-D Earth structures and2-D is an approximation. The 3-D inversion does not require any as-sumptions about the strike and dimensionality of the data and all fourcomponents of the impedance tensor are used. We used the ModEMcode for 3-D electromagnetic modelling and inversion (Egbert andKelbert, 2012) in parallel implementations (Meqbel, 2009). The inver-sion algorithm inModEM is based onminimization of the penalty func-tional that seeks the smoothest model that fits the data at the specifiedlevel of the misfit. The advantage of the ModEM approach is that theJacobian computations are divided into several steps (data functionals,forward and adjoint solvers, model parameter mappings). Such tech-nique allows faster computation and makes ModEM suitable for largeinversions.

TheMaSca array covers 350 km (NW–SE) by 480 km (NE–SW) area.The 3-D inversion is very expensive in terms of computational time andmemory requirements. Thus, some compromises are required withregards to grid size. Particularly, the modelling mesh should be denseenough to accurately predict the model response. On the other hand,the computational costs increase with the number of the modelled pa-rameters. The grid size of the entire array inversion is coarse, with theminimumcell size of 10 km. Thismeans that the smaller features cannotbe resolved. To improve resolution properties of the 3-D inversion forsuch a large area we sub-divided the entire array into sub-arrays ofsmaller size. This allowed for making 3-D inversion with the finermesh with theminimum cell size of 4 km. In Fig. 2 red circles representsites used for 3-D inversion of the entire array and the blue ones showone of the selected sub-arrays in the Kiruna district.

The MaSca array is located close to the Atlantic Ocean, which causesthe distortion of the MT impedance tensor and vertical magnetic com-ponent. Particularly, near the coast line the vertical magnetic transferfunction is large which is caused by the coast effect (Parkinson, 1959).

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

14˚

14˚

16˚

16˚

18˚

18˚

20˚

20˚

22˚

22˚

24˚

24˚ 26˚

64˚

65˚

66˚ 66˚

67˚

68˚ 68˚

69˚ 69˚

70˚ 70˚

67˚

65˚

26˚

xy

3

45

67

Fig. 2. Location of magnetotelluric stations plotted on top of the geological map (Fig. 1).Crustal-scale 3-D inversion sites: grey circles — MaSca array stations (all), red circles (ontop of grey) — sites used for 3-D inversion of the entire array, green triangles — siteswith the vertical magnetic component, blue circles — sites used for 3-D inversion of theKiruna sub-array. The 2-D inversion profiles Crust 1—7 are marked with the red lines.Each line corresponds to 2-D model in Figs. 7a–g and 3-D cross-section in Fig. 8a–g.Black arrows show orientations of the coordinate systems, discussed in the text. For 3-Dinversion of the entire area, the coordinate system was rotated by 45 (rotated arrows onthe left). (For interpretation of the references to colour in this figure legend, the readeris referred to the web version of this article.)

4 M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

The previous BEAR experiment results (Varentsov et al., 2002) and geo-magnetic studies in Scandinavia by (Jones, 1981) showed therewas cer-tainly a large coast effect at all coastal stations. In our data, the inductionarrows reveal the coast effect at a period of 1024 s. This issue is ad-dressed in the parallel paper (Cherevatova et al, 2015). To tackle thisproblem, the ocean is included as a priori model. Coast effect dependson distance between coast and station. The Kiruna sub-array is located200 km away from the coast line and the ocean does not affect thedata. The effect of the Bothnian Bay can be neglected because it haslow salinity and consequently rather low conductivity of 0.1 S/m com-pare to 4 S/m for the ocean. Moreover, the depth of the Bothnian Bayis very shallow ranging from a few metres to a few tens of metres.

Four impedance componentswere inverted (Zxx, Zxy, Zyx, Zyy) togetherwith vertical transfer function (Hz, tippers). A total of 201 stations werecarefully chosen from a complete data set of 279, and inconsistent re-sponses were edited and bad stations removed. The MT responses foreach site were then reduced to 4 periods per decade with a total of 13periods in the range of 16–65,536 s for impedance tensor and from 16to 512 s for tipper. All together 67 stations had vertical transfer functionsamong sites used for 3-D inversion (green triangles in Fig. 2). For 3-Dinversion we set error floors 10% of the off-diagonal components for allimpedance elements and 5% for vertical component.

We aligned the 3-D numerical grid with the average geoelectricstrike of N45E (Cherevatova et al, 2015). The impedances were alsorotated to be consistentwith the coordinate system. See Fig. 2 for orien-tations of the x and y axes.

The mesh was discretized to be fairly regular in the centre, with cellwidths of approximately 10 km and comprises a total of 62 cells in the xdirection, 74 cells in the y direction, and 80 cells in the z direction. Thethickness of the uppermost layer was chosen to be 500 m. Layer

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

thickness was increased with depth by applying a scaling factor of 1.1between each layer. The total dimensions of the model domain were20,380 × 20,490 × 10,237 km. Several starting models, including lay-ered Earth model, different resistivity half-space models (100 Ωm and1000 Ωm) were compared. A 1000 Ωm half-space starting model(ocean included) resulted in geologically consistent models with rela-tively low root-mean square deviation (RMSD). Note, that the crust inthe studied area is very resistive as 2-D inversion models showed, thatis the 1000 Ωm is the best choice for the initial resistivity. The finalmodel, derived by inverting all four impedance tensor elements includ-ing the tippers, converged to an average RMSD of 4.13 (Fig. 3).

The crustal resistive layer (RL) is the largest unit in the 3-Dmodel. Inthe north, between 300 km and 400 km in the x direction, RL thickensfrom 20 km in the east to 40 km towards thewest (Fig. 3b). The thickestpart of the resistor RL is 80 km (0–100 km in the x direction) in thesouth and extends eastward by around 80 km. In the west, feature RLgradually thins to 10–20 km. The lower crust is more conductive withthe resistivity varying from 100 Ωm to a few hundreds of Ωm.

The upper crustal Kittilä conductor (KC) appears in the east at loca-tion 300–400 km in the x direction and 400 km in the y direction(Fig. 3b). The resistivity of the feature KC is around 10Ωm. The conduc-tor dips towards east from a few kilometres near surface to 20 km. Note,the conductor does not outcrop but is overlapped by a thin resistivelayer of 1–2 km thickness. The MaSca array crosses the southern endof the conductor KC and there are evidences that KC extends furtherto the north thickening up to 25 km. The overall size of the conductorKC is 50 km in the y direction and 70 km in the x direction.

A regional horizontal mid-crustal conductor HC is located between50 km and 300 km in the y direction and 200 km to 400 km in the xdirection. The resistivity of the conductor HC is around 100 Ωm. Thetop of the conductor is at a depth of 20 km and the lower boundary ataround 50 km. Another mid-crustal conductor C with the resistivity of10 Ωm is observed to the west of the insulator RL. The conductor isseen as a layer located between 100 km and 290 km in the x directionand between 20 km and 130 km in the y direction. The depth to thetop of the conductor C is 20 km and its thickness is 40–50 km (Fig. 3a).

A number of the upper crustal conductors can be observed in the 3-Dmodel. The first conductor CI at a distance of 100 km in the x directionstretches from thenorth to south around200km. The top of the conduc-tor is located at a depth of 2 km above the middle crustal conductor Cand its thickness is around 5 km. It is possible that CI outcrops to thesurface. The conductors C and CI are separated by a 10 km thick resistivelayer. In the south, conductor CI is shifted by 30 km in the y direction.The feature CI is a highly conductive unit with the resistivity of5–10 Ωm.

The upper crust in the Lofoten region is also conductive (LC). TheLofoten conductor LC is 120 km long in the x direction and 40 kmwide in the y direction. The feature LC is observed as a layer with thethickness between 2 km and 6 km and resistivity of 5–10 Ωm. Finally,the near-surface conductor in the Skellefteå district can be seen in thesouth (SC). The conductor slightly dips from the surface to 5 km east-ward. The resistivity of the conductor SC is around 5 Ωm.

2.1. 3-D inversion in the Kiruna district

In this section we present a 3-D inversion model obtained for theselected smaller area around Kiruna (Kiruna district). The data aremore densely spaced in this area and it is located far enough from theocean that the coast effect is negligible. The full impedance tensor wasinvertedwithout using the tipper data. A total of 59 stationswere select-ed in the area covering 190 km by 250 km in NE and SW directions, re-spectively. The MT responses for each site were then reduced to 2periods per decade with a total of 9 periods in the range 0.003–256 sfor impedance tensor. For 3-D inversion we set error floors 10% of theoff-diagonal elements for all impedance components. The model gridwas not rotated and the inversion model is presented in the measured

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

1 2 3lg resistivity

40

Distance

(km)

0

100

200

300

400

0

100

200

300

400

500

Distance (km)

Dep

th (

km) 0

40

80

x

y

z

NCaledonian Thrust Front

C

LC

HC

SC

(a)

(b)

Distance

(km)

0

100

200

300

400

0

100

200

300

400

500

Distance (km)

Dep

th (

km) 0

40

80

Skellefte

KirunaCI

CI

RL

RL

Caledonian Thrust Front LC

C

CICI

C

HC

SC

Kiruna

RL

RLKC

Fig. 3. Perspective view of 3-D inversionmodel, with NE (a) and NW (b) slices across themodelled space. Abbreviations: RL— resistive layer, HC— horizontalmid-crustal conductor, KC—

Kittilä conductor, SC— Skellefteå conductor, LC— Lofoten conductor, C—mid-crustal conductor, CI— upper crustal conductor. Continuous red line— Caledonian Thrust Front, thick greybox shows the Kiruna sub-array (another inversion for smaller sub-area, see inversion model in Fig. 4). (For interpretation of the references to colour in this figure legend, the reader isreferred to the web version of this article.)

5M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

coordinate system (black arrows on the right, Fig. 4). The mesh wasdiscretized to be fairly regular in the centre, with cell widths of 2 kmor 4 kmwith a total of 96 cells in the x direction, 116 cells in the y direc-tion, and 95 cells in the z direction. Themeshwas createdwith a verticalexpansion factor of 1.1with the top layer thickness being50m. The totaldimensions of themodel domainwere 8182 × 8232 × 4277 km. Severalstartingmodels were compared and 1000Ωmhalf-spacemodelwas se-lected. The final model, derived by inverting all four impedance tensorelements converged to an average RMSD of 2.25.

The obtained 3-D model of the Kiruna sub-array is consistent withthe previously described 3-D inversion of the entireMaSca array. Gener-ally, the same features appear in the 3-D inversion model: the regional-scale resistive layer (RL), extensive horizontal mid-crustal conductor(HC) and Kittilä conductor (KC). The insulator RL is shown in thesame form as in 3-D inversion of the entire array. In the north, the thick-ness of the resistor RL is around 40–50 kmand it thins to 20 km towards

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

south. Themiddle to lower crust in the south is more conductive due tothe conductor HC. The resistivity of the feature HC is lower in thismodelthan in the model of the entire array (Fig. 3) and characterized by thevalues of 10–100Ωm.However, the position and shape of the conductorHC is the same. The top of the conductor is at a depth of 20 km and thelower boundary at 60 km. The stable Kittilä conductor appears in the 3-D inversion model. The size of the conductor KC is (x, y) = (70, 40) km,which is nearly the same as it was previously found. The feature KC dipsfrom 10 km to 15 km towards the north.

3. Data fit in the 3-D inversion

We present the data fit for 3-D inversion of the entire MaSca array(Fig. 3) and the Kiruna sub-area (Fig. 4). The corresponding data fitplots are shown for four sites: B23 (BEAR array), M10, M15 and M36for (Fig. 5) and o04, e02, o16 and w12 for (Fig. 6).

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

1

2

3

lg resistivity

0

4

0 50 100 150 200 250

Distance (km)

0

50

100

150

200

Dis

tanc

e (k

m)

020406080100

Dep

th (

km)

KC

Kiruna

x

y

z

N

HC

HC

RLRL

RL

Fig. 4. Perspective view of 3-D inversionmodel in the Kiruna sub-area, with NW–SE slices across themodelled space. Abbreviations: RL— resistive layer, HC— horizontal mid-crustal con-ductor, KC — Kittilä conductor.

6 M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

A total RMS misfit for 3-D inversion of the entire array is 4.13, thusmost of the data are fit as can be seen in Fig. 5, where observed and pre-dicted curves for both off-diagonal and diagonal impedance compo-nents are shown for the selected sites. In Fig. 5, the real and imaginaryparts of the components of the impedance tensor are scaled by

ffiffiffi

Tp

,where T is period in seconds. This compensates for the reduction in im-pedance amplitudes at long periods (Patro and Egbert, 2011). The datafit is reasonably good for both components. For site B23 a discrepancybetween measured and modelled imaginary part is observed for Zyy,similar poor fit is observed at a few other sites (not shown).

The data fit for the 3-D inversion of the Kiruna sub-array is shown inFig. 6. The triangles represent the predicted data from the inversion ofthe entire array, for comparison. The final RMSD of the model is 2.25,thus most of the data are predicted. The site o04 shows poor fit of thediagonal components for periods less than 10 s. Similarly, for site e02the diagonal components are not fitted at shorter periods (b1 s). Atsite o16, Zxy and Zxx components are not fitted by the inversion of theentire array, but have a good fit for the Kiruna sub-array inversion.

4. Comparison of the 2-D and 3-D inversion results

In the following we present a comparison of the 2-D and 3-D inver-sion models. The strike and dimensionality analysis together with thedetailed description of the 2-D inversion, resolution tests and interpre-tation are given in a parallel paper (Cherevatova et al, 2015). Therefore,in this section we provide a brief overview of the obtained results. The2-D crustal-scale inversions were performed for seven profiles (seeFig. 2 for location) with the azimuth of N135E. The average distancebetween the profiles is 50 km. There are gaps in site spacing of up to80 km, mostly because of high topography. The strike and dimensional-ity analysis by (Caldwell et al., 2004; Zhang et al., 1987) revealed 3-Dbehaviour of the data. Therefore, the inversion of the determinant wasundertaken as an appropriate technique for 2-D inversion of 3-D data(Cherevatova et al., 2014; Pedersen and Engels, 2005). The final modelswere obtained using the Rebocc algorithm (Siripunvaraporn andEgbert,2000). A homogeneous half-space initial model and an ocean a priori

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

model were used for all inversions. The a priori model was obtainedfrom the map of integrated conductance (S-map) of Fennoscandia(Korja et al., 2002). Due to the expected 3-D effects and possible staticshifts, higher error floors for apparent resistivity were assigned thanfor impedance phases. The final inversion models are presented inFig. 7, the sharp boundaries at a depth are the result of the resolutiontests, described in (Cherevatova et al, 2015).

We extracted 2-D cross-sections for all profiles to compare the 3-Dand 2-D inversion models: Fig. 8a corresponds to profile Crust 1 andFig. 8b–g to profiles Crust 2–7, respectively. Same abbreviations wereused for both models. Note that discretization of the 3-D and 2-Dmodels are different, particularly in 3-D the minimal cell size is 10 kmwhile in 2-D — 2 km. Therefore, the small features in the 3-D modelare hardly resolved. In addition, a slightly different data set was usedin 2-D and 3-D cases. The most extensive units are in good consistency:the resistive layer (RL), the middle to lower crustal conductor in thewest (C) and horizontal conductor in the east (HC). Some smallerfeatures have disappeared (C7-4) or changed their shape (Cc). In thefollowing we give the detailed comparison of the 2-D and 3-D models.

The resistive layer (RL) appears to be less resistive in 3-D than in2-D:mainly 103Ωmand 104Ωm in someplaces, compare to 104Ωmal-most everywhere in 2-D (Fig. 8). Generally, the shape of the insulator RLis the same as in 2-D: thinner in the east and in thewest and significant-ly thicker in the central part. The lower crust is more conductive withthe resistivity of 100 Ωm which is consistent with the 2-D models.

The upper crustal Kittilä conductor (KC) is well presented in bothmodels. The south end of the conductor KC is clearly seen in profileCrust 2 in 2-D model (Fig. 7b). It is not obvious from the 3-D modelthat the conductor KC extends to the profile (b), therefore it is markedas KC? (Fig. 8b). Further to the south, the horizontal conductor HC canbe seen in profiles (c), (d) and (f) in 2-D. The resistivity of the conductoris the same in both cases – 100Ωm, but its shape differs. The top of theconductor HC in the 3-Dmodel is placed deeper (20–25 km) than in 2-D(10 km). Such difference is surprising, because the top of the conductoris alwayswell resolved by theMTmethod. Another difference is that thefeature HC is more narrow in 3-D and terminated by the resistor at aprofile distance of 350 km. In the 3-D model, the conductor HC is

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

−20

−10

0

10

20

imag

inar

y

101 102 103 104 105

Period (s)

−20

−10

0

10

20

real

101 102 103 104 105

−80

−40

0

40

80

imag

inar

y

101 102 103 104 105

Period (s)

−30

−20

−10

0

10

20

30

real

101 102 103 104 105

−40

−20

0

20

40

101 102 103 104 105

Period (s)

−40

−20

0

20

40

101 102 103 104 105

−60

−40

−20

0

20

40

60

101 102 103 104 105

Period (s)

−40

−20

0

20

40

101 102 103 104 105

−100

−50

0

50

100

101 102 103 104 105

Period (s)

−100

−50

0

50

100

101 102 103 104 105

−100

−50

0

50

100

101 102 103 104 105

Period (s)

−100

−50

0

50

100

101 102 103 104 105

−40

−20

0

20

40

101 102 103 104 105

Period (s)

−40

−20

0

20

40

101 102 103 104 105

−100

−50

0

50

100

101 102 103 104 105

Period (s)

−100

−50

0

50

100

101 102 103 104 105

Zxx

·√Τ

and

Zyy

·√Τ

Zxy

·√Τ

and

Zyx

·√Τ

Blue−YX, Red−XY

Blue−XX, Red−YY

B23 M10 M15 M36

Fig. 5. Observed (circle) and predicted (line) responses of Zxx, Zyy (bottom) and Zxy, Zyx (top) scaled byffiffiffi

Tp

for 3-D inversion results of the entire array (Fig. 3). RMSD= 4.13.

7M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

presented as a homogeneous layer contrary to 2-D inversions, wherethe smaller conductive units C3-3, C3-5, C5-3 and C5-5 are observedwithin HC. These conductors can be observed in 3-D model, but theyare located closer to the surface, at a depth of 2–3 km embedded inthe resistive layer RL. In the south-east (Fig. 8f and g), the 2-D and 3-Dinversion models are different. The 2-D inversion suggests separatedmid-crustal conductors C6-2, C6-3, C6-4 and C7-3 and highly conduc-tive C7-4. While in 3-D, the lower crust is presented as the conductivelayer with the resistivity of 10–100Ωm that is more geologically mean-ingful. In addition, the Skellefteå conductor SC is seen as upper crustallayer, which is in good agreement with the previous studies in thisregion, suggesting SC at a depth of 4 km (Hübert et al., 2009). Contraryto this, the strong conductor C7-4 in the middle to lower crust wasfound in the 2-D model.

The regional conductive structure C is seen to the west of RL in 2-Dand 3-Dmodels. The shape of C is not clear from 2-Dmodels and differsfrom 3-D. In the upper crust, another conductor CI is located above C(Fig. 8). The conductors C and CI are separated by the resistive layer. Itis possible that CI represents the same conductive unit Cc in the 2-Dmodels, for example below site a07 in (d) slice. In the profiles (f) and(g), the feature CI coincides with the conductors C6-1 and C7-1 in 2-Dmodels. The upper crustal Lofoten conductor (LC) is also seen in 2-Dmodels as a conductive spot below sites a13, M30, M14 and C5-1 in(e) profile.

Some differences between 2-D and 3-D models can be identified inthe upper crust. For example, in the 3-D model there are upper crustalconductors within RL: below sites j11 (a), P06 and o11 (b), M11 (c),

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

z08 (e), q29 (f), S06 (g). These conductors are not present in 2-D inver-sionmodels. The near-surface conductors Cs1, Cs2, which are stable fea-tures in 2-D models, do not appear in 3-D inversion. There is someevidence of C1 conductor below site k19 in the 3-D inversion model.The strong conductor C7-2 in (g) 2-D profile disappeared in 3-Dmodel.

The comparison of twomodels (2-D and 3-D) proves that 2-D inver-sion of the determinant is reliable in case of 3-D data. It reproduces thesame features with nearly the same characteristics as in 3-D case. Sincethe regional structure is generally 3-D, we prefer 3-D inversion modelfor interpretation. However, the smaller near-surface features are betterpresented in 2-D models.

5. Interpretation

Fig. 9 displays the logarithmof the integrated conductance (S) distri-bution for the 3-D inversion of the entire MaSca array for the upper(0–20 km) and lower (20–40 km) crust. The conductance is the best re-solved andmost stable parameter of themodels. The conductance givesa limit within which the thickness and conductivity of a layer can vary.This implies that a thickmoderately conducting layer can be replaced bya thinner but a more conducting layer or by a set of thin conductors aslong as the conductance remains constant. The resolution propertiesof the MT method together with Occam-type inversions often suggestthe presence of large conductive regions in the Precambrian environ-ment in Fennoscandia. However, near-surface geological and geophysi-cal mapping (Airo, 2005; Koistinen and et al, 2001; Korja et al., 1996)shows inmany cases that these conductors are caused by volumetrically

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

−100

−50

0

50

100

imag

inar

y

10−310−210−1100 101 102 103 104

Period (s)

−100

−50

0

50

100

real

10−310−210−1100 101 102 103 104

−300

−200

−100

0

100

200

300

imag

inar

y

10−310−210−1100 101 102 103 104

Period (s)

−300

−200

−100

0

100

200

300re

al

10−310−210−1100 101 102 103 104

−60

−30

0

30

60

10−310−210−1100 101 102 103 104

Period (s)

−60

−30

0

30

60

10−310−210−1100 101 102 103 104

−100

−50

0

50

100

10−310−210−1100 101 102 103 104

Period (s)

−100

−50

0

50

100

10−310−210−1100 101 102 103 104

−100

−50

0

50

100

10−310−210−1100 101 102 103 104

Period (s)

−100

−50

0

50

100

10−310−210−1100 101 102 103 104

−200

−100

0

100

200

10−310−210−1100 101 102 103 104

Period (s)

−200

−100

0

100

200

10−310−210−1100 101 102 103 104

−100

−50

0

50

100

10−310−210−1100 101 102 103 104

Period (s)

−100

−50

0

50

100

10−310−210−1100 101 102 103 104

−400

−200

0

200

400

10−310−210−1100 101 102 103 104

Period (s)

−400

−200

0

200

400

10−310−210−1100 101 102 103 104

Zxx

·√Τ

and

Zyy

·√Τ

Zxy

·√Τ

and

Zyx

·√Τ

Blue−YX, Red−XY

Blue−XX, Red−YY

o04 e02 o16 w12

Fig. 6.Observed (circle) and predicted (line) responses of Zxx, Zyy (bottom) and Zxy, Zyx (top) scaled byffiffiffi

Tp

for 3-D inversion results of the Kiruna sub-array (Fig. 4). The triangles mark thepredicted responses from the 3-D inversion of the entire MaSca array. The final RMSD is 2.25.

8 M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

minor amounts of highly conducting rocks within otherwise resistiverocks. Typical examples are conductivefillings of shear zones (sulphides,graphite) or shear-enhanced connectivity of carbon (graphite)-bearingrock assemblages (Korja et al., 1996). Consequently, the interpretationshould take into account that thick conducting layers might also becaused by the superimposition of thin conducting layers/formations.The conductance slices clearly demonstrate the difference betweenupper and lower crust. The upper crustal conductance is generallybelow 10 S (Fig. 9a), whereas the lower crustal conductance (Fig. 9b)is above 100 S and reaches values of 104 S. The upper crust is generallyresistive with the presence of the high conductance of 104 S. Most ofthe conductance is concentrated in the lower crust to the south of theCaledonian Thrust Front. An extensive Precambrian conductor is locatedat 250 km of the x direction and between 200–400 km in the y directionnear Kiruna (Fig. 9b). The Kittilä conductor to the east at 400 km in the xdirection is also seen in the lower crust. The MaSca array crosses thesouthern edge of the Kittilä Greenstone Belt and suggests that its con-ductance is from 1000 S to a few thousands of S, which agrees with(Korja et al., 2002) estimates.

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

In the Skellefteå district, geological evidence (Nironen, 1997) andMTdata (Korja et al., 2002) indicate gently dipping structures. Skellefteåconductor is dipping north-eastward and is modelled with a constanttotal conductance everywhere (4000 S, (Rasmussen et al., 1987)). Inour measurements, the Skellefteå anomaly is presented only in thesouthernmost end (450–500 km of the y direction, Fig. 9a and b). TheMaSca array only partly covers the Skellefteå anomaly.

The resistive layer (RL) in the inversion models (for example,Fig. 3) mostly represents the Precambrian basement of the northernSvecofennian volcanic belt. In the upper crust, the SvecofennianDomainappears as a highly resistive unit with the resistivity of several thou-sands of Ωm. In the northernmost part (profiles Crust 1, 2 in Fig. 7aand b or in Fig. 8a and b), the resistive rocks cross the south-westernedge of the Archaean Domain associated with the Archaean gneisses.Since the southward edge of the Archaean Domain appears to havebeen formed by rifting (Gaál and Gorbatschev, 1987), detached frag-ments of Archaean crust could possibly occur in the Svecofennian Do-main (Section 1.1). The Svecofennian resistive rocks extend to thelower crust under the Caledonian Thrust Front, where the resistive

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

01020304050

Dep

th(k

m)

01020304050

Dep

th(k

m)

150 200 250 300 350 400 450

Distance(km)

n15

S17

n14

S15

S14

o52

o51

S13

S11

S10

q43

S09

q37

S08

g61

q30

S07

S06

B15

S05

q16

S04

q12

S03

q06

S02

q03

C7−1 C7−2 RLC7−3 C7−4Moho

NW SE

01020304050

01020304050

50 100 150 200 250 300 350 400

d43

a15

o65

n20

o62

n18

o58

n17

n08

q42

q50

q41

q40

q52

q31

q29

q25

q22

q17

q13

C

C6−1 RLC6−2 C6−3

C6−4Moho

NW SE

01020304050

01020304050

50 100 150 200 250 300 350 400 450

a19

a20

d13

a16

M19

o73

n21

o65

o62

M42

M18

z05

z04

B16

M41

z08

z07

z13

q14

q15

B22

C5−1C

CcRL

C5−3 HCC5−5

Moho

NW SE

01020304050

01020304050

50 100 150 200 250 300 350

M14

a18

a11

a07

a08

M42

M18

M17

z01

M38

z08

M15

C Cc RLHCMoho

NW SE

01020304050

01020304050

0 50 100 150 200 250 300 350

M30

a21

a18

M11

a09

a08

w10

w07

w06

w02

w01

c06a

x02

x01

x05

M39

B23

j28

c16

rr01

M15

C RL C3−3 C3−4 C3−5HCMoho

NW SE

01020304050

01020304050

100 150 200 250 300 350 400

a13

o31

a12

y05

P09

M34

P08

P10

o21

o18

P06

o17

o27

o26

o25

o15

M31

P03

o11

o14

o13

P01

M37

e04

o04

o06

o05

B31

Cs C2 RLKC

Moho

NW SE

01020304050

Dep

th(k

m)

01020304050

Dep

th(k

m)

150 200 250 300 350 400

k19

k16

k13

k11

k10

k09

k08

k07

k06

k05

k03

k02

k01

GT

3

j12

M32

j13

P04

o13

P01

e02

o01

o02

o04

o03

C1 RLKC

MohoNW SE

Caledonian Thrust Front

0

1

2

3

4lg(ρ, Ωm)

(a); RMSD = 1.2

(b); 1.7

(c); 1.7

(d); 1.3

(e); 1.0

(f); 1.1

(g); 1.1

Fig. 7. Final 2-D crustal-scale inversion models obtained by inverting the determinant of the impedance tensor using the Rebocc algorithm (Siripunvaraporn and Egbert, 2000) for sevenprofiles (a)–(g): Crust 1–7. Abbreviations: RL — resistive layer, KC — Kittilä conductor, HC— horizontal crustal conductor, C — other crustal conductors. The black line shows the Mohodepth for Fennoscandia from Grad et al. (2009).

9M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

crust is thickening to at least 50 km. In the south the crustal resistiveregion can be associated with the Trans-Scandinavian granitoids, whichextends to great depth (Olesen et al., 2002). The aeromagnetic datashow that the Precambrian igneous rocks of the Fennoscandian Shieldcan be followed below the Caledonian nappes to the tectonic windowsin northern Norway (Olesen et al., 2010a). To the west of the CaledonianThrust Front, the resistive layer is thinning to 20 km. The resistive rocksin Norway are associated with the Caledonian nappes, underlain byPrecambrian basement.

Themiddle and lower crust in northern Svecofennian volcanic belt ismore conductive (Fig. 3). The integrated conductance of themiddle andlower crust varies from 100 S to 104 S (Fig. 9b). At a crustal scale, in par-ticular in the Precambrian Shield areas, the only viable candidates to en-hance the conductivity are graphite or carbon in grain-boundary films

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

(Duba et al., 1988; Frost et al., 1989; Jödicke, 1992; Raab et al.,1998), graphite in the metamorphosed deposits of black shales(Boerner et al., 1996; Korja and Koivukoski, 1994; Korja et al.,1996), sulphides (Jones et al., 1997; Korja et al., 1996) and oxides(Duba et al., 1994). Elsewhere, in Fennoscandia, similar conductorsusually represent metamorphosed carbon- and sulphide-bearing sedi-mentary rocks, transported into deep crustal levels by tectonic process-es (Hjelt et al., 2006; Korja et al., 1996, 2002, 2008).

In the Skellefteå district, the top of the conductor is located at adepth of 20 km and the lower boundary at 30 km (marked as C6-3and C7-3 & C7-4 in Fig. 8f, g). The Skellefteå district is a relic of a Prote-rozoic island arc that separates Archaean Domain to the north fromSvecofennian rocks to the south (see Section 1.1). Thus, the middle tolower crustal conductor in the Skellefteå region can be an ancient suture

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

01020304050

Dep

th(k

m)

01020304050

Dep

th(k

m)

100 150 200 250 300 350 400 450

Distance(km)

S02

S03

S04

S05

S06

S07

q03

q06

q12

q16

q20

q30

q31

q40

q41

q42

q51

q52

n06

n08

n15

n17

n18

o51

o54

01020304050

01020304050

0 50 100 150 200 250 300 350 400

d43

q13

q17

q22

q25

q29

q31

q40

q42

q50

q51

z05

z06

n08

n15

n17

n18

n20

o67

01020304050

01020304050

0 50 100 150 200 250 300 350 400 450

d13

M19

M20

M28

M41

M42

a19

a20

q15

z01

z04

z07

z08

z13

z15

o73

o74

B16

B22

01020304050

01020304050

0 50 100 150 200 250 300 350

M11

M14

M15

b16

M17

M38

a07

a08

a11

a18

w10

01020304050

01020304050

0 50 100 150 200 250 300 350

M01

M02

M11

M15

M30

a09

a10

c16

j28

w02

w06

w10

x01

rr01

B23

01020304050

01020304050

100 150 200 250 300 350 400

M09

M31

M34

M37

a13

e04

y04

y07

o05

o06

o11

o15

o18

o25

o27

o31

o32

B31

P02

P06

P10

01020304050

Dep

th(k

m)

01020304050

Dep

th(k

m)

100 150 200 250 300 350 400

M22

M32

e02

j11

j13

k01

k02

k05

k06

k08

k10

k12

k16

k19

o01

o02

o03

o12

GT

3

ki14

P01

P04NW SE

0

1

2

3

4lg(ρ, Ωm)

(a)

(b)

(c)

(d)

(e)

(f)

(g)

RL

RL

RL

RL

RL

RL

RL

KC

KC?

LC

LC

LC

LC

HC

HC

C

C

C

C

C

C

CI

CI

CI

CI

CI

C5-5C5-3

C3-5C3-3

C6-4

C6-3

C7-3&C7-4

SC

x05

Fig. 8. Profile cross-sections of the 3-D model (Fig. 3). The cross-sections are aligned with the 2-D inversion profiles (a)–(g) Crust 1–7 for comparison. RMSD = 4.13.

10 M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

zone, but the dipping of the conductor towards the north-east(Rasmussen et al., 1987) is not supported neither by our 2-D nor 3-Dmodels. (Hübert et al., 2009) interpreted the highly conductive zonewith interconnected graphite from the Bothnian Basin. The Skellefteåregion is also characterized by the shallow strong conductivity anoma-lies at a depth of 2–5 km (SC in Fig. 3). These conductive features arerelated to the ore bearing Skellefteå Group (1.89–1.88 Ga, (Skyttäet al., 2011)) and the overlaying Vargfors Group (1.88–1.87 Ga,(Billstrom and Weihed, 1996)), comprising turbiditic sedimentaryrocks, shales, conglomerates and local intercalations of volcanic rocks(Allen et al., 1996; Årebäck et al., 2005; Weihed et al., 1992) (Fig. 1).Similar anomalies at the same depth have been observed in this region90 km further to the south (Hübert et al., 2009). These anomalies have

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

been interpreted as caused by the graphite in black shales in thecontact between metasediments and metavolcanics (Hübert et al.,2013).

In the north-east, the region of the enhanced conductivity of a fewthousands of Siemens is observed in the upper and middle crust (KCin Fig. 4, 3). The survey area is located in the western and northernparts of the Central Lapland Greenstone Belt (CLGB) that is one of thelargest Proterozoic greenstone belts in the world. The CLGB consists ofa Palaeoproterozoic (2.5–1.97 Ga) volcanic and sedimentary coverthat was deposited on the Archaean (N2.5 Ga) basement. The highestconductivities are related toNS elongated graphite- and sulphide-bearingschists of the CLGB, which are visible also in the airborne electromagneticdata of the study area (Lahti et al., 2012).

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

0

100

200

300

400

Distance

x (km

)

0

100

200

300

400

500

Distance y (km)

0

100

200

300

400

500

Distance y(km) −1

0

1

2

3

4lg (S)

(a) 0−20 km (b) 20−40 km

N

KittiläKittilä

Skellefteå Skellefteå

Kiruna Kiruna

Fig. 9. Interpretation of the crustal conductivity obtained from 3-D inversion of the entire MaSca array (Fig. 3). Integrated crustal conductances for depth ranges: (a) 0–20 km,(b) 20–40 km. Red line marks Caledonian Thrust Front. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

11M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

To the west of the Caledonian Thrust Front the uppermost resistivecrust most likely is composed of the Caledonian nappes. In the Jämtlandregion, it was found in theMT study by (Korja et al., 2008) that resistiv-ity of the Caledonian allochthons varies by one order of magnitude, butare generally resistive. The resistivity of the middle/upper allochthonsvaries from a few thousands to 105 Ωm while the lower allochthon ismore conductive having resistivity from a few hundreds to a few thou-sands Ωm.

Extensive drilling in the Jämtland region (Andersson et al., 1985) hasshown that a fewmetres to tens of metres thick alum shale layers existbelow the Caledonian rocks and may enhance the conductivity in theuppermost crust. There are highly conductive upper crustal units inthe west, below the Caledonian Thrust Front. For instance, the conduc-tors C2 and C7-2 on the profiles Crust 2 and Crust 7 (Fig. 7b and g), re-spectively. These conductors can be associated with the alum shales.Alum shales, the autochthonous Cambrian carbon-bearing black shaleson the top of the Precambrian basement and subordinate conglomer-ates and sandstones, crop out at the Caledonian Front. Alum shales areknown to be electrically highly conducting (Gee, 1972). It is also possible,that conductor C1 on the profile Crust 1 represents the alum shales over-lain by the resistive rocks of the Caledonian nappes. The alum shales arenot modelled on the other profiles, because of the larger gaps in sitespacing and low resolution of the grid in 3-D case (the minimal cellsize is 10 km).

There is other highly conductive upper crustal unit CI in the 3-D in-version model, extending from the north to the south. The conductormay represent the alum shales buried towards west from the Caledo-nian Thrust Front, as it was interpreted in the Jämtland–Trøndelag re-gion (Korja et al., 2008) or in southern Norway (Cherevatova et al.,2014). The other possible interpretation is themica schists of the upper-most allochthon, which may contain the highly conductive material.These rocks of exotic origin belong to the eastern margin of Laurentiaor to amicro-continent between Laurentia and Baltica. Another possibleexplanation is that the conductor is caused by the presence of the salinefluids (Becken and Ritter, 2012), infiltrated in the Sagfjorden shear zone(Fig. 1). The Sagfjorden shear zone may have governed the location ofthe large-scale, Mesozoic, normal fault zones that bound the sides ofthe Lofoten and Utrø st Ridges (Olesen et al., 2002). It should be notedthat EM methods are extremely sensitive to the presence of a minoramount of conducting phases distributed in large volumes, such asfluids or carbon.

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

In the westernmost margin of the Fennoscandian Shield, underLofoten, the highly conductive layer can be observed at a depth of afew kilometres (LC in Fig. 3). The Lofoten consists of the Archaean andProterozoic rocks of the ancient Baltica, exposed during Devonian ex-tension (Section 1.1). Thus, it is unlikely that the rocks of the uppermostallochthon extend below the Lofoten. The only possible explanation ofthe enhanced conductivity below the Lofoten is the presence of thesaline fluids in complicated fault system within the Lofoten. Note, thatthe topology of the Lofoten ridge and the coast line is not included inthe inversion, thus the conductors may be caused by the presence ofthe salt water within the fjords.

The basement below the Caledonides supposed to belong to thenorthern Svecofennian volcanic belt as the Precambrian rocks of Balticawere underthrust beneath Caledonian rocks during Caledonian oroge-ny. There are no other major differences between thewestern and east-ern parts than the distance from the Caledonian orogen and from thewestern margin of Fennoscandia, therefore it is logical to associate theenhancement of the average crustal conductivity (C in Fig. 3) withCaledonian related processes or later opening of the Atlantic Ocean,which also might have affected the lower crust. A similar middle tolower crustal conductor has been identified in the Jämtland-Trø ndelagregion (Korja et al., 2008) further to the south. This explanation hasbeen accepted as a possible cause of the enhanced conductivity. Wealso suggest that the middle to lower crustal conductor identified inour 3-D and 2-D models, can be a northern extension of the Jämtland-Trø ndelag conductor.

6. Conclusions

We have conducted magnetotelluric measurement within the“Magnetotellurics in the Scandes” (MaSca) project, during the summersof 2011 to 2013. All together data at 220 broad-band and 59 long periodsites were measured. The array crosses the major tectonic structureswithin the Precambrian and younger Caledonian rocks.

• We performed 3-D inversion of the entire MaSca array and com-pared this result with the inversion of the smaller sub-areaaround Kiruna. The sub-area inversion allowed improvement ofthe resolution properties of the inversion by increasing the gridresolution and consequently also the accuracy of the forwardcomputations.

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

12 M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

• The 3-D inversion model was compared with the 2-D inversions forseven profiles. The 2-D models were obtained by inverting the deter-minant of the impedance tensor. The comparison proves the reliabilityof the 2-D inversion of the determinant for 3-D data and demonstratesthe good consistency of the 2-D and 3-D models.

• In the north, the resistive upper crust is composedmostly of Archaeangneisses. In the east, the upper crust consists of the resistive rocks ofthe Svecofennian volcanic belt. In the central and southern parts, thewhole crust is resistive and composed of the granitoidal rocks of theTrans-Scandinavian Igneous Belt. In the Caledonides the resistiverocks are presented as a thin cover of Caledonian nappes togetherwith the exposed northern Svecofennian volcanic basement.

• The conductive middle to lower crust is observed in the northernSvecofennianvolcanic province. The enhanced conductance canbe ex-plained by the presence of the graphite-bearing rocks and sulphides.

• The upper crustal conductor in the Skellefteå district is interpreted asshallow conductive graphitic shales within the Vargfors group. In thenorth, another upper crustal conductor is seen as the southern endof theKittilä greenstone belt. The enhanced conductivity there is relat-ed to graphite- and sulphide-bearing schists, which are visible also inthe airborne electromagnetic data of the study area.

• In the Caledonides, the presence of the highly conductive alum shalesis observed along the Caledonian Thrust Front. The regional highlyconductive upper crustal units are closely located along the upper-most allochthon mica shists, which can also contain the conductivematerial.

• A highly conductive structure is revealed beneath the Lofoten penin-sula. This is surprising, as the Lofoten is generally composed of resis-tive Precambrian basement rocks. The enhanced conductivity can becaused by the presence of saline fluids within cracks and faults inthe basement.

• Themiddle to lower crustal conductor under the Caledoniannappes isassociated with Caledonian related processes or later opening of theAtlanticOcean,which alsomight have affected the lower crust. A similarmiddle to lower crustal conductor has been identified in the Jämtland–Trøndelag region (Korja et al., 2008) further to the south.

Acknowledgements

Wewould like to thank Lars Dynesius for help in preparing the fieldwork. Financial support was provided by the Academy of Finland(272912 and 136345).

References

Agustsson, K., 1986. A magnetotelluric pilot study in the Scandes. Geol. Fören. Stockh.Förh. 108 (3), 258–261.

Airo, M.-L., 2005. Aerogeophysics in Finland 1972–2004: Methods, System Characteristicsand Applications. Geological Survey of Finland Special Paper 39.

Allen, R., Weihed, P., Svenson, S., 1996. Setting of Zn–Cu–Au–Ag massive sulfide depositsin the evolution and facies architecture of a 1.9 Ga marine volcanic arc, Skelleftedistrict, Sweden. Econ. Geol. Bull. Soc. Econ. Geol. 91 (6), 1022–1053.

Andersen, T.B., 1998. Extensional tectonics in the Caledonides of southern Norway, anoverview. Tectonophysics 285, 333–351. http://dx.doi.org/10.1016/S0040-1951(97)00277-1.

Andersson, A., Dahlman, B., Gee, D.G., Snäll, S., 1985. The Scandinavian alum shales,Sveriges Geologiska Undersökning (SGU). Serie Ca NR 56, 1–49.

Årebäck, H., Barrett, T., Abrahamsson, S., Fagerstrom, P., 2005. The PalaeoproterozoicKristineberg VMS deposit, Skellefte district, northern Sweden, part I. Geol. Miner.Depos. 40 (4), 351–367.

Becken, M., Ritter, O., 2012. Magnetotelluric studies at the San Andreas Fault Zone: impli-cations for the role of fluids. Surv. Geophys. 33 (1), 65–105. http://dx.doi.org/10.1007/s10712-011-9144-0.

Billstrom, K., Weihed, P., 1996. Age and provenance of host rocks and ores in thePaleoproterozoic Skellefte District, northern Sweden. Econ. Geol. 91 (6), 1054–1072.http://dx.doi.org/10.2113/gsecongeo.91.6.1054.

Boerner, D.E., Kurtz, R.D., Craven, J.A., 1996. Electrical conductivity and PALEO-Proterozoicforedeeps. J. Geophys. Res. 101 (B6), 13775–13791.

Brasse, H., Kreutzmann, A., Cerv, V., Ernst, T., Jankowski, J., Jozwiak, W., Neska, A.,Pedersen, L.B., Smirnov, M., Schwarz, G., Sokolova, E., Varentsov, I.M., Hoffmann, N.,

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

Palshin, N., Korja, T., 2006. Probing electrical conductivity of the Trans-European Su-ture Zone. Eos Trans. Am. Geophys. Union 87 (29), 281.

Caldwell, T.G., Bibby, H.M., Brown, C., 2004. The magnetotelluric phase tensor. Geophys.J. Int. 158 (2), 457–469. http://dx.doi.org/10.1111/j.1365-246X.2004.02281.x.

Chalmers, J., Green, P., Japsen, P., Rasmussen, E., 2010. The Scandinavian mountains havenot persisted since the Caledonian orogeny. A comment on Nielsen et al. (2009a).J. Geodyn. 50 (2), 94–101. http://dx.doi.org/10.1016/j.jog.2010.02.001.

Cherevatova, M., Smirnov, M., Korja, T., Kaikkonen, P., Pedersen, L., Hübert, J., Kamm, J.,Kalscheuer, T., 2014. Crustal structure beneath southern Norway imaged bymagnetotellurics. Tectonophysics 628 (0), 55–70. http://dx.doi.org/10.1016/j.tecto.2014.04.036.

Cherevatova, M., Smirnov, M.Y., Jones, A.G., Pedersen, L.B., 2015. Magnetotelluric arraydata analysis from north-west Fennoscandia. Tectonophysics http://dx.doi.org/10.1016/j.tecto.2014.12.023.

Duba, A., Huengest, E., Nover, G., Will, G., Jödicke, H., 1988. Impedance of black shale fromMönsterland i borehole: an anomalously good conductor? Geophys. J. 94 (3),413–419. http://dx.doi.org/10.1111/j.1365-246X.1988.tb02264.x.

Duba, A., Heikamp, S., Meurer, W., Mover, G., Will, G., 1994. Evidence from borehole sam-ples for the role of accessory minerals in lower-crustal conductivity. Nature 367(6458), 59–61.

Ebbing, J., England, R., Korja, T., Lauritsen, T., Olesen, O., Stratford,W.,Weidle, C., 2012. Struc-ture of the Scandes lithosphere from surface to depth. Tectonophysics 536–537, 1–24.http://dx.doi.org/10.1016/j.tecto.2012.02.016.

Egbert, G.D., Kelbert, A., 2012. Computational recipes for electromagnetic inverse prob-lems. Geophys. J. Int. 189 (1), 251–267. http://dx.doi.org/10.1111/j.1365-246X.2011.05347.x.

England, R.W., Ebbing, J., 2012. Crustal structure of central Norway and Sweden from in-tegrated modelling of teleseismic receiver functions and the gravity anomaly.Geophys. J. Int. 191 (1), 1–11. http://dx.doi.org/10.1111/j.1365-246X.2012.05607.x.

Fossen, H., 2010. Extensional tectonics in the North Atlantic Caledonides: a regional view.Geol. Soc. Lond., Spec. Publ. 335 (1), 767–793. http://dx.doi.org/10.1144/SP335.31.

Frost, B.R., Fyfe, W.S., Tazaki, K., Chan, T., 1989. Grain-boundary graphite in rocks andimplications for high electrical conductivity in the lower crust. Nature 340,134–136.

Gaál, G., Gorbatschev, R., 1987. An outline of the Precambrian evolution of the Balticshield. Precambrian Res. 35, 15–52. http://dx.doi.org/10.1016/0301-9268(87)90044-1.

Gee, D.G., 1972. The regional geological context of the Tåsjö Uranium Project, CaledonianFront, Central Sweden, Sveriges Geologiska Undersökning (SGU). Serie C (671), 1–36.

Gee, D.G., Fossen, H., Henriksen, N., Higgins, A.K., 2008. From the Early Paleozoic Platformsof Baltica and Laurentia to the Caledonide Orogen of Scandinavia and Greenland.Episodes 31, 44–51.

Gorbatschev, R., Bogdanova, S., 1993. Frontiers in the Baltic Shield. Precambrian Res. 64,3–21.

Grad, M., Tiira, T., Group, E.W., 2009. The Moho depth map of the European Plate.Geophys. J. Int. 176 (1), 279–292. http://dx.doi.org/10.1111/j.1365-246X.2008.03919.x.

Guggisberg, B., Kaminski, W., Prodehl, C., 1991. Crustal structure of the Fennoscandianshield: a traveltime interpretation of the long-range FENNOLORA seismic refractionprofile. Tectonophysics 195 (2-4), 105–137. http://dx.doi.org/10.1016/0040-1951(91)90208-A.

Hjelt, S.E., Korja, T., Kozlovskaya, E., Lahti, I., Yliniemi, J., SVEKALAPKO SeismicTomography, W.G., BEAR, W.G., 2006. Electrical conductivity and seismic velocitystructures of the lithosphere beneath the Fennoscandian Shield. Geol. Soc. Lond.Mem. 32, 541–559.

Hübert, J., Malehmir, A., Smirnov, M., Tryggvason, A., Pedersen, L.B., 2009. MT measure-ments in the western part of the paleoproterozoic Skellefte Ore District, NorthernSweden: a contribution to an integrated geophysical study. Tectonophysics 475 (3–4),493–502. http://dx.doi.org/10.1016/j.tecto.2009.06.011 (doi:http://dx.doi.org/10.1016/j.tecto.2009.06.011).

Hübert, J., M. d. l., Juanatey, A.G., Malehmir, A., Tryggvason, A., Pedersen, L.B., 2013. Theupper crustal 3-D resistivity structure of the Kristineberg area, Skellefte district,northern Sweden revealed by magnetotelluric data. Geophys. J. Int. 192, 500–513.http://dx.doi.org/10.1093/gji/ggs044.

Jödicke, H., 1992. Water and graphite in the Earth's crust-an approach to interpreta-tion of conductivity models. Surv. Geophys. 13, 381–407. http://dx.doi.org/10.1007/BF01903484.

Jones, A.G., 1981. Geomagnetic induction studies in Scandinavia. J. Geophys. 50, 23–36.Jones, A.G., 1983. The electrical structure of the lithosphere and asthenosphere beneath

the Fennoscandian Shield. J. Geomagn. Geoelectr. 35, 811–827.Jones, A.G., Katsube, T.J., Schwann, P., 1997. The longest conductivity anomaly in the

world explained: sulphides in fold Hinges causing very high electrical anisotropy.J. Geomagn. Geoelectr. 49, 1619–1629.

Juhlin, C., Elming, S., Mellqvist, C., Öhlander, B.,Weihed, P.,Wikström, A., 2002. Crustal reflec-tivity near the Archaean–Proterozoic boundary in northern Sweden and implications forthe tectonic evolution of the area. Geophys. J. Int. 150 (1), 180–197. http://dx.doi.org/10.1046/j.1365-246X.2002.01706.x.

T. Koistinen, M. B. Stephens, V. Bogatchev, O. Nordgulen, M. Wennerström, J. Korhonen,2001. Geological map of the Fennoscandian Shield, scale 1:2 000 000, Espoo: Trond-heim: Uppsala: Moscow: Geological Survey of Finland: Geological Survey of Sweden:Geological Survey of Norway: Ministry of Natural Resources of Russia.

Korja, T., 2007. How is the European lithosphere imaged by magnetotellurics? Surv.Geophys. 28, 239–272. http://dx.doi.org/10.1007/s10712-007-9024-9.

Korja, T., Koivukoski, K., 1994. Crustal conductors along the SVEKA profile in theFennoscandian (Baltic) Shield, Finland. Geophys. J. Int. 116 (1), 173–197. http://dx.doi.org/10.1111/j.1365-246X.1994.tb02135.x.

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008

13M. Cherevatova et al. / Tectonophysics xxx (2015) xxx–xxx

Korja, T., Hjelt, S.E., Kaikkonen, P., Koivukoski, K., Rasmussen, T.M., Roberts, R.G., 1989. Thegeoelectric model of the POLAR Profile, Northern Finland. Tectonophysics 162,113–133. http://dx.doi.org/10.1016/0040-1951(89)90359-4.

Korja, T., Tuisku, P., Pernu, T., Karhu, J., 1996. Lapland Granulite Belt — implications forproperties and evolution of deep continental crust. Terra Nova 8, 48–58.

Korja, T., Engels, M., Zhamaletdinov, A., Kovtun, A., Palshin, N., Tokarev, A., Smirnov, M.,Asming, V., Vanyan, L., Vardaniants, I., BEAR, W.G., 2002. Crustal conductivity inFennoscandia — a compilation of a database on crustal conductance in theFennoscandian Shield. Earth Planets Space 54, 535–558.

Korja, T., Smirnov, M., Pedersen, L., Gharibi, M., 2008. Structure of the Central Scandina-vian Caledonides and the underlying Precambrian basement, new constraints frommagnetotellurics. Geophys. J. Int. 175, 55–69.

Lahti, I., Korja, T., Kaikkonen, P., Vaittinen, K., BEAR,W.G., 2005. Decomposition analysis ofthe BEARmagnetotelluric data: implications for the upper mantle conductivity in theFennoscandian Shield. Geophys. J. Int. 163, 900–914.

Lahti, I., Korja, T., Smirnov, M., Vaittinen, K., Sandgren, E., Niiranen, T., Nykänen, V., 2012.3-D imaging of the Central Lapland Greenstone Belt using magnetotelluric and seismicdata. Geophysical Research Abstracts. vol. 14.

Lund, C.-E., 1979. The fine structure of the lower lithosphere underneath the Blue Roadprofile in northern Scandinavia. Tectonophysics 56, 111–122. http://dx.doi.org/10.1016/0040-1951(79)90017-9.

Lund, C.-E., Heikkinen, P., 1987. Reflection measurements along the EGT POLAR-profile,northern Baltic Shield. Geophys. J. R. Astron. Soc. 89 (1), 361–364. http://dx.doi.org/10.1111/j.1365-246X.1987.tb04431.x.

Meqbel, N.M., 2009. The electrical conductivity structure of theDead Sea Basin derived from2D and 3D inversion of magnetotelluric data. (Ph.D. thesis). Freie Universität Berlin.

Nironen, M., 1997. The Svecofennian Orogen: a tectonic model. Precambrian Res. 86,21–44.

Nyblade, A.A., Pollack, H.N., 1993. A global analysis of heat flow from Precambrian ter-rains: implications for the thermal structure of Archean and Proterozoic lithosphere.J. Geophys. Res. Solid Earth 98 (B7), 12207–12218. http://dx.doi.org/10.1029/93JB00521.

Olesen, O., Lundin, E., Nordgulen, O., Osmundsen, P.T., Skilbrei, J.R., Smethurst, M.A., Solli,A., Bugge, T., Fichler, C., 2002. Bridging the gap between onshore and offshore geologyin Nordland, northern Norway. Nor. J. Geol. 82, 243–262.

Olesen, O., Brönner, M., Ebbing, J., Gellein, J., Gernigon, L., Koziel, J., Lauritsen, T.,Myklebust, R., Sand, M., Solheim, D., Usov, S., 2010. New aeromagnetic and gravitycompilations from Norway and adjacent areas—methods and applications. PetroleumGeology Conference series 7, pp. 559–586.

Parkinson, W.D., 1959. Directions of rapid geomagnetic fluctuations. Geophys. J. R.Astron. Soc. 2 (1), 1–14. http://dx.doi.org/10.1111/j.1365-246X.1959.tb05776.x.

Please cite this article as: Cherevatova,M., et al., Electrical conductivity strumagnetotelluric data, Tectonophysics (2015), http://dx.doi.org/10.1016/j.

Patro, P.K., Egbert, G.D., 2011. Application of 3D inversion to magnetotelluric profile datafrom the Deccan Volcanic Province ofWestern India. Phys. Earth Planet. Inter. 187 (1-2),33–46. http://dx.doi.org/10.1016/j.pepi.2011.04.005.

Pedersen, L.B., Engels, M., 2005. Routine 2D inversion of magnetotelluric data using thedeterminant of the impedance tensor. Geophysics 70 (2), G33–G41.

Raab, S., Hoth, P., Huenges, E., Müller, H.J., 1998. Role of sulfur and carbon in the electricalconductivity of the middle crust. J. Geophys. Res. 103 (B5), 9681–9689.

Ramberg, I.B., Bryhni, I., Nøttvedt, A., Rangnes, K. (Eds.), 2008. The Making of a Land—

Geology of Norway. Norwegian Geological Association, Trondheim.Rasmussen, T.M., 1988. Magnetotellurics in southwestern Sweden: evidence for electrical

anisotropy in the lower crust? J. Geophys. Res. Solid Earth 93 (B7), 7897–7907.http://dx.doi.org/10.1029/JB093iB07p07897.

Rasmussen, T.M., Roberts, R.G., Pedersen, L.B., 1987. Magnetotellurics along theFennoscandian long range profile. Geophys. J. R. Astron. Soc. 89, 799–820.

Siripunvaraporn, W., Egbert, G., 2000. An efficient data-subspace inversion method for 2Dmagnetotelluric data. Geophysics 65 (3), 791–803.

Siripunvaraporn, W., Egbert, G., 2009. WSINV3DMT: vertical magnetic field transfer func-tion inversion and parallel implementation. Phys. Earth Planet. Inter. 173 (3-4),317–329. http://dx.doi.org/10.1016/j.pepi.2009.01.013.

Skyttä, P., Hermansson, T., Andersson, J., Weihed, P., 2011. New zircon data supportingmodels of short-lived igneous activity at 1.89Ga in thewestern SkellefteDistrict, centralFennoscandian Shield. Solid Earth 3 (1), 355–383.

Slagstad, T., Balling, N., Elvebakk, H., Midttømme, K., Olesen, O., Olsen, L., Pascal, C., 2009.Heat-flow measurements in Late Palaeoproterozoic to Permian geological provincesin south and central Norway and a new heat-flow map of Fennoscandia and theNorwegian-Greenland Sea. Tectonophysics 473, 341–361. http://dx.doi.org/10.1016/j.tecto.2009.03.007.

Smirnov,M.Y., Pedersen, L.B., 2009. Magnetotelluric measurements across the Sorgenfrei–Tornquist Zone in southern Sweden and Denmark. Geophys. J. Int. 176, 443–456.

Varentsov, I.M., Engels, M., Korja, T., Yu, Smirnov M., the BEAR Working Group, 2002. Ageneralized geoelectric model of Fennoscandia: a challenging database for long-periods 3D modelling studies within the Baltic Electromagnetic Array Research(BEAR) project, Izvestya. Phys. Solid Earth 38 (10), 855–896.

Weihed, P., Bergman, J., Bergström, U., 1992. Metallogeny and tectonic evolution of theEarly Proterozoic Skellefte district, northern Sweden, Precambrian Research. Precambri-an Metallogeny Related to Plate. Tectonics 58 (1–4), 143–167.

Zhang, P., Roberts, R.G., Pedersen, L.B., 1987. Magnetotelluric strike rules. Geophysics 52(3), 267–278.

cture of north-west Fennoscandia from three-dimensional inversion oftecto.2015.01.008