Geophysical Studies of the Upper Crust of the Central Swedish Caledonides in Relation to the COSC Scientific Drilling Project

Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1281

Geophysical studies of the upper crust of the central Swedish Caledonides in relation to the COSC scientific drilling project

PETER HEDIN

ACTA UNIVERSITATIS UPSALIENSIS ISSN 1651-6214 ISBN 978-91-554-9320-2 UPPSALA urn:nbn:se:uu:diva-261112 2015 Dissertation presented at Uppsala University to be publicly examined in Hambergsalen, Geocentrum, Villavägen 16, Uppsala, Friday, 16 October 2015 at 10:00 for the degree of Doctor of Philosophy. The examination will be conducted in English. Faculty examiner: Associate Professor Charles Hurich (Department of Earth Sciences, Memorial University of Newfoundland, St. Johns, Canada).

Abstract Hedin, P. 2015. Geophysical studies of the upper crust of the central Swedish Caledonides in relation to the COSC scientific drilling project. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1281. 87 pp. Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-554-9320-2.

The Collisional Orogeny in the Scandinavian Caledonides (COSC) project aims to provide a deeper understanding of mountain belt dynamics through scientific deep drilling in the central parts of the mountain belt of western Sweden. The main targets include a subduction related allochthon, the basal orogenic detachment and the underlying partially subducted Precambrian basement. Research covered by this thesis, focusing primarily on reflection seismic data, was done within the framework of the COSC project. The 55 km long composite COSC Seismic Profile (CSP) images the upper crust in high resolution and established the basis for the selection of the optimum location for the two 2.5 km deep COSC boreholes. Together with potential field and magnetotelluric data, these profiles allowed the construction of a constrained regional interpretation of the major tectonic units. Non-conventional pseudo 3D processing techniques were applied to the 2D data prior to the drilling of the first borehole, COSC#1, to provide predictions about the 3D geometry of subsurface structures and potential zones of interest for the sampling programs. COSC-1 was drilled in 2014 and reached the targeted depth with nearly complete core recovery. A continuous geological section and a wealth of information from on-site and off- site scientific investigations were obtained. A major post-drilling seismic survey was conducted in and around the borehole and included a 3D reflection seismic experiment. The structurally and lithologically complex Lower Seve Nappe proved difficult to image in detail using standard processing techniques, but its basal mylonite zone and underlying structures are well resolved. The 3D data, from the surface down to the total drilled depth, show good correlation with the initial mapping of the COSC-1 core as well as with preliminary results from on-core and downhole logging. Good correlation is also observed between the 2D and 3D reflection seismic datasets. These will provide a strong link between the two boreholes and a means to extrapolate the results from the cores and boreholes into the surrounding rock. Ultimately, they will contribute to the deeper understanding of the tectonic evolution of the region, the Scandinavian Caledonides and the formation of major orogens.

Keywords: reflection seismic, collisional orogeny, Scandinavian Caledonides, COSC, scientific drilling, geophysical logging, gravity, magnetics

Peter Hedin, Department of Earth Sciences, Geophysics, Villav. 16, Uppsala University, SE-75236 Uppsala, Sweden.

ISSN 1651-6214 ISBN 978-91-554-9320-2 urn:nbn:se:uu:diva-261112 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-261112)

Dedicated to my beloved Marilyn and wonderful son Samuel

Supervisors & Committee

Supervisor Professor Christopher Juhlin Department of Earth Sciences – Geophysics, Uppsala University, Uppsala, Sweden

Assistant Supervisor Associate Professor Alireza Malehmir Department of Earth Sciences – Geophysics, Uppsala University, Uppsala, Sweden

Dr. Bjarne Almqvist Department of Earth Sciences – Geophysics, Uppsala University, Uppsala, Sweden

Faculty Opponent Associate Professor Charles Hurich Department of Earth Sciences, Memorial University of Newfoundland, St. Johns, Canada

Examination Committee Dr. Joaquina Álvarez Marrón Department of Earth’s Structure and Dynamics and Crystallography – Earth’s structure and Dynamics, Institute of Earth Sciences Jaume Almera, Barcelona, Spain

Dr. Cédric Schmelzbach Department of Earth Sciences – Geophysics, ETH Zürich, Zürich, Switzer- land

Dr. Ari Tryggvason Department of Earth Sciences – Geophysics, Uppsala University, Uppsala, Sweden

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Hedin, P., Juhlin, C., Gee, D. G. (2012) Seismic imaging of the Scandinavian Caledonides to define ICDP drilling sites. Tectonophysics, 554-557:30–41 II Hedin, P., Malehmir, A., Gee, D. G., Juhlin, C., Dyrelius, D. (2013) 3D interpretation by integrating seismic and potential field data in the vicinity of the proposed COSC-1 drill site, central Swedish Caledonides. Geological Society, London, Special Publications, 390:301–319 III Hedin, P., Almqvist, B., Berthet, T., Juhlin, C., Buske, S., Si- mon, H., Giese, R., Krauß, F., Rosberg, J-E., Alm, P-G. (2015) 3D reflection seismic imaging at the 2.5 km deep COSC-1 scientific borehole, central Scandinavian Caledonides. Tectonophysics, manuscript under review IV Juhlin, C., Hedin, P., Gee, D. G. (2015) Seismic imaging of the eastern Scandinavian Caledonides: Siting the 2.5 km deep COSC-2 borehole. Solid Earth, manuscript to be submitted

Reprints were made with permission from the respective publishers.

Additional contributions to conference proceeding written during my Ph. D. which are not included in this thesis:

 Hedin, P., Malehmir, A., Gee, D. G., Juhlin, C., Dyrelius, D. (2013) COSC geophysical and geological site investigations. 75th EAGE Con- ference & Exhibition, London, UK, Extended Abstract 16529, TuSP1- 07, doi: 10.3997/2214-4609.20131097  Ahmadi, O., Hedin, P., Malehmir, A., Juhlin. C. (2013) 3D Seismic In- terpretation and Forward Modeling – an approach to providing reliable results from 2D seismic data. In: Johnson, E. (Ed.), Mineral Deposit Re- search for a High-Tech World, vols 1-4, p 50-53, 12th SGA Biennial Meeting, Uppsala, Sweden.  Juhlin, C., Hedin, P. (2014) 3D Seismic Processing of Crooked Line 2D Data in the Vicinity of the COSC 2.5 Km Deep Scientific borehole. 76th

EAGE Conference & Exhibition, Amsterdam, The Netherlands, Extend- ed Abstract 21889, WS5-P11, doi: 10.3997/2214-4609.20140522

Contributions

The papers included in this thesis are the result of collaboration with several authors. The individual contributions of the author of this thesis are summarized below. I I participated in the seismic acquisition and performed the data decoding, processing and analysis. I participated in the discussion and interpretation and then wrote the first draft, with input from co-authors on geology. Worked with the co-authors to re- fine the manuscript. II I constructed the 3D geological model. Participated in the discussion and interpretation and wrote a draft of the manuscript, with input from co-authors on the inverse modeling. I then made improvements to the manuscript based on feedback and guidance from the co-authors. III I participated in the acquisition, processing and analysis of core logging data as well as the processing and analysis of downhole logging data. I participated in the seismic acquisition and performed the decoding, processing and analysis. I wrote the first draft, with input from co-authors on geology, and then improved the manuscript after helpful feedback. IV I participated in the seismic acquisition in 2010 and 2011 and processed the data from these two surveys. Participated in the discussion and interpretation of the full seismic profile and contributed with some parts of the draft as well as refinement of the final manuscript.

Contents

1 Introduction ...... 13 2 Collisional Orogeny in the Scandinavian Caledonides ...... 16 2.1 The Caledonian geology of central Sweden ...... 17 2.2 The Seve Nappe Complex ...... 21 2.3 Geophysical Background ...... 23 2.4 Scientific Drilling in the Scandinavian Caledonides ...... 26 3 Methods ...... 29 3.1 The reflection seismic method and data acquisition ...... 29 3.1.1 2D crooked line acquisition ...... 30 3.1.2 Limited 3D acquisition ...... 31 3.2 Reflection seismic processing ...... 32 3.2.1 Crossdip analysis ...... 33 3.2.2 Swath 3D imaging ...... 37 3.3 Drilling of COSC-1 scientific borehole and logging of geophysical rock parameters ...... 42 4 Summary of Papers ...... 46 4.1 Paper I: Seismic Imaging of the Scandinavian Caledonides to define ICDP drilling sites ...... 46 4.1.1 Summary ...... 46 4.1.2 Conclusions ...... 50 4.2 Paper II: 3D interpretation by integrating seismic and potential field data in the vicinity of the proposed COSC-1 drill site, central Swedish Caledonides ...... 50 4.2.1 Description of data ...... 50 4.2.2 3D interpretation ...... 52 4.2.3 Forward and inverse modeling ...... 53 4.2.4 Conclusions ...... 57 4.3 Paper III: 3D reflection seismic imaging at the 2.5 km deep COSC-1 scientific borehole, central Scandinavian Caledonides...... 57 4.3.1 Summary ...... 58 4.3.2 Conclusions ...... 61 4.4 Paper IV: Seismic imaging of the eastern Scandinavian Caledonides: Siting the 2.5 km deep COSC-2 borehole ...... 64 4.4.1 Summary ...... 65 4.4.2 Conclusions ...... 69

5 Conclusions ...... 71 6 Outlook ...... 74 7 Summary in Swedish ...... 75 Acknowledgements ...... 78 References ...... 81

Abbreviations

1D One Dimensional 2D Two Dimensional 3D Three Dimensional BL Byxtjärn-Liten 2D reflection seismic profile CABLES Caledonian And Bothnian Lithosphere Elucidated by Seismics CCT Central Caledonian Transect CDMO Crossdip Moveout CDP Common Depthpoint CISP Concentric Impact Structures in the Paleozoic CMP Common Midpoint COSC Collisional Orogeny in the Scandinavian Caledonides CSP COSC Seismic Profile DAFNE Drilling into Active Faults in Northern Europe DGRF Definitive International Geomagnetic Reference Field DH Dammån-Hallen 2D reflection seismic profile DMO Dip Moveout FX Frequency-Space domain ICDP International Continental Scientific Drilling Program IGCP International Geological Correlation Programme IGSN International Geo Sample Number KF Kallsjön-Fröå 2D reflection seismic profile LD Liten-Dammån 2D reflection seismic profile NMO Normal Moveout MT Magnetotelluric MSCL Multi Sensor Core Logger OSG ICDP Operational Support Group PaMVAS Paleoproterozoic Mineralized Volcanic Arc Systems SCANLIPS Scandinavian Lithosphere P and S-wave experiment S Sällsjö 2D reflection seismic profile SFDZ Sveconorwegian Frontal Deformation Zone SGU Geological Survey of Sweden SIST Swept Impact Seismic Technique SNC Seve Nappe Complex S/N Signal-to-Noise ratio SSDP Swedish Scientific Drilling Program

TD Total Depth of drilling TIB Transscandinavian Igneous Belt vp Compressional-wave velocity VSP Vertical Seismic Profile XRF X-ray Fluorescence Hz Hertz mGal milliGal mm millimeter cm centimeter m meter km kilometer ms milliseconds s seconds Ma Million years ago Ga Billion years ago kg kilogram

1 Introduction

More than a century has passed since the first mapping of the mountains in Scandinavia indicated that some of the rock units had been transported far from the west before being emplaced in their current location. Törnebohm (1888) suggested that the Åreskutan metamorphic rock had been transported at least 100 km, sparking a heated debate at the end of the 19th century. Fol- lowing the dawn of modern plate tectonic theory in the 1960's (Vine and Matthews, 1963; McKenzie and Parker, 1967; Morgan, 1968), the mountain belts of Scandinavia, Greenland and Scotland were early recognized to have been parts of a major Paleozoic orogen (Dewey, 1969), the Caledonides (Figure 1.1). In the 1970's, the International Geological Correlation Program (IGCP) project 27 was initiated to study the mountain belts surrounding the North Atlantic in detail. This resulted in major breakthroughs in the understanding of mountain building processes and, in particular, the Caledonide Orogen (Gee and Sturt, 1985). The northern parts of the Caledonides are the product of the collision of the two continents Baltica and Laurentia that occurred more than 400 million years ago, with partial subduction of the former beneath the latter (Gee et al., 2008). Nappe emplacement along the Laurentian margin in Greenland involved at least 200 km of transport to the west (Higgins and Leslie, 2000); along the Baltica margin, the eastwards displacement was at least twice this amount (Gee, 1978). The mountain belt was, by the end of its formation, in many respects comparable with the present day active Himalaya-Tibet orogen (Andersen, 1998; Gee et al., 2010; Labrousse et al., 2010). Post- orogenic collapse (Andersen, 1998) followed by erosion, uplift and extension has over time brought down the surface to a level that cuts through the core of the paleo mountain belt, revealing the internal architecture of the nappes. The well preserved remnants of key orogenic features in the Caledonides that are found at, or just beneath, the presently exposed surface, presents a superb environment to study the processes of thrusting (Törnebohm, 1888; Asklund, 1960; Gee, 1975b; Hossack and Cooper, 1986) and extensional tectonics (Andersen, 1998; Fossen, 2000). The past few decades have seen improved knowledge of the composition, configuration and pre-Caledonian origin of the tectonic units as well as the timing of events (Corfu et al., 2014a).

Figure 1.1. The configuration of the North Atlantic Caledonides prior to early Ceno- zoic opening North Atlantic Ocean. (Modified from Lorenz et al., 2015a).

In recent years there has been an increased focus on questions regarding the formation and transportation of these allochthons during the active formation of the mountain belt (e.g. Grimmer et al., 2015; Majka et al., 2014a). Some of these questions require quantitative, high resolution investigation of specific targets that are inaccessible without the use of sophisticated methods. The Collisional Orogeny in the Scandinavian Caledonides (COSC) project was designed to shed light on some of the features that appear to be central to the formation of the Caledonides through scientific deep drilling (Gee et al., 2010; Lorenz et al., 2011). The project is divided into two phases, each involving a fully cored 2.5 km deep borehole supported by geological and geophysical investigations in the area. The targets are located in the province of Jämtland in west central Sweden, one of the most extensively studied regions of the Scandinavian Caledonides. The first borehole, COSC-1, was drilled near the town of Åre in 2014 to investigate the formation and transportation of a hot allochthon, the Seve Nappe Complex (SNC, Lorenz et al., 2015a). The second borehole, COSC-2, is currently in the planning stages and will study the nature of deformation in the underlying allochthons along the basal detachment (the décollement) and in the underlying basement (Gee et al., 2010). The COSC project constitutes the framework of this PhD thesis, with over five years dedicated to mainly reflection seismic surveys related to the project and the drilling of COSC-1. A main reflection seismic profile was acquired in 2010 and subsequently extended in 2011 and 2014 with the aim of locating the most suitable sites for drilling the two specific targets and to characterize their tectonic setting in the area. The COSC-1 borehole was

14 successfully drilled during the spring and summer of 2014 with nearly complete core recovery. The drilling was supported by comprehensive down hole logging campaigns and followed by a complex multi-component seismic survey in and around the borehole, including a 3D reflection seismic experiment. This thesis is divided into two parts; a comprehensive summary and a collection of four papers. The summary consists of seven chapters. Following this general introduction, chapter 2 gives a more comprehensive introduction to the COSC project and the geology and geophysics of the two drilling targets. Chapters 3 and 4 constitute the major part of this work. The first of these takes a look at the acquisition and processing of reflection seismic data and some results, while the second comprises summaries of the four papers. This is followed by conclusions and outlook in chapters 5 and 6 and finally a summary in Swedish in chapter 7.

15 2 Collisional Orogeny in the Scandinavian Caledonides

The Caledonide Orogen is understood to have resulted from the closure of an ocean (Iapetus) and the collision of two continents, Baltica and Laurentia. The Iapetus Ocean began to close in late Cambrian or possibly early Ordovi- cian time, with subduction occurring along the margins of both continents. The convergence of the two plates culminated with Scandian continent- continent collision that began around 445 Ma with Baltica being partially subducted beneath Laurentia (Ladenberger et al., 2014). The ocean-derived nappes were thrust at least 400 km eastwards onto the Baltoscandian platform (Gee, 1978). By contrast, the nappes that were thrust westwards onto the continental platform of Laurentia, now exposed in northeastern Green- land, comprise only continental crust and involved only about 200 km of crustal shortening (Higgins and Leslie, 2000; Gasser, 2014). Orogenic con- traction lasted for about 50 million years (Gee et al., 2008). Post-orogenic collapse and extension took over towards the end of the early Devonian and continued into the late Devonian (Andersen, 1998). This was followed by a few hundred million years of erosion before rifting and early Cenozoic opening of the North Atlantic Ocean (Mosar, 2003). Prior to this continental break-up, the northern Caledonides were perhaps 1000 km across and spanned a distance of about 3000 km, from today’s British Isles to the Sval- bard Archipelago (Figure 1.1; Gee, 2015). Today, the remnant of the Caledonian mountain belt in Scandinavian, the Scandes, dominate the geology and topography of Norway and western Sweden (Figure 2.1) and extend a distance of about 1800 km from southern Norway to the Barents Sea in the north. A geotraverse through the central Scandes, stretching nearly 300 km from the orogenic thrust front in the province of Jämtland, central Sweden, to the North Atlantic coast in the province of Tröndelag, Norway (Figure 2.1a), has become one of the best studied regions across the mountain belt (Törnebohm, 1888; Högbom, 1909; Asklund, 1938; Gee, 1975b; Dyrelius et al., 1980). Over the past few decades, the geological mapping in this region has also received support from several large geophysical surveys, including the reflection seismic profiling along the Central Caledonian Transect (CCT) (Hurich et al., 1989; Palm et al., 1991; Hurich, 1996; Juhojuntti et al., 2001). Western Jämtland in Swe- den was therefore a natural choice for scientific deep drilling in the context

16 of the International Continental Scientific Drilling Program (ICDP) and the COSC project (Gee et al., 2010; Lorenz et al., 2011). Therefore, this thesis is also limited to this part of the Scandinavian Caledonides with a focus on the geological formations of central to western Jämtland (Figure 2.2). This chapter aims to give a geological and geophysical background to the study area (Figure 2.2) with emphasis on the Seve Nappe Complex (SNC), which was the target of the COSC-1 borehole, and to introduce the COSC project in more detail.

2.1 The Caledonian geology of central Sweden To provide a general overview of the structure of the orogen, Gee et al. (1985) constructed a tectonic map of the Scandinavian Caledonides. The orogen was subdivided into for major complexes that were, for convenience, referred to as the Lower, Middle, Upper and Uppermost Allochthons, all overlying the Precambrian autochthonous basement. Although this grouping of nappes is not uncontroversial (Corfu et al., 2014b), it has become widely accepted that the Lower Allochthon is derived from the Baltica platform (continental shelf) and foreland basin, the Middle Allochthon is dominated by units that originated from the Baltoscandian rifted margin and continent- ocean transition zone, the Upper Allochthon is composed of igneous suites and sedimentary formations of the Iapetus oceanic domain, and the Upper- most Allochthon comprises fragments of the Laurentian continental margin (Figure 2.1a). In Jämtland (Figure 2.2), the Lower, Middle and Upper Allochthons are well developed and distinct and the province hosts the type localities to many of the nappes. Along the orogenic front, autochthonous Cambrian sedimentary rocks including black alum shales rest unconformably on top of the Precambrian crystalline basement. These underlie the major detachment which, due to a comprehensive mineral prospecting drilling program in the Caledonian front in the late 70’s (Figure 2.2; Gee et al., 1978, 1982), has been shown to dip westwards at about 1-2°. The CCT reflection seismic profile (Figure 2.2; Palm et al., 1991; Juhojuntti et al., 2001), demonstrated that this frontal décollement continues westwards and probably reaches the Swedish-Norwegian border at a depth of about 6 km. The kerogen-rich alum shales are thought to have acted as a lubricant along this basal detachment, to facilitate the shallow angle thrust emplacement of Caledonian nappes over distances of several hundred kilometers.

17 a) 10° 20° b) 10° 20° N N 70° 70° W E W E S S

0 100km 68° 0 100km 68°

Norwegian Norwegian Sea Sea CF 66° CF 66°

64° 64° Trondheim Östersund Trondheim Östersund Jämt- Jämt- land land Tröndelag Gulf of 62° Tröndelag Gulf of 62° Dalarna Bothnia Dalarna Bothnia Siljan ring Siljan ring Oslo 60° Oslo 60° Stockholm

Stockholm SFDZ 10° 20° 10° 20° Oslo Permian Rift Pre-Svecofennian rocks (>1.96 Ga) Svecofennian supracrustal and granitoid rocks Devonian - Old Red Sandstones (1.95-1.86 Ga) and granites (1.82-1.75 Ga) Uppermost Allochthon Transscandinavian Igneous Belt (TIB) rocks Laurentian margin Småland-Värmland granitoids (1.85-1.65 Ga) Outboard Mafic TIB Plutonics (1.85-1.65) Upper Allochthon Terranes Iapetus-derived Mafic-intermediate TIB volcanics (1.8-1.7 Ga) Köli Nappe Complex Felsic TIB volcanics (1.8-1.7 Ga) Dala granitoids (1.81-1.68) Middle Allochthon Dala sedimentary rocks (1.8-1.7 Ga) Outermost Baltica margin - Seve & related nappes Rätan Batholith (1.71-1.68 Ga) Revsund granitoid suite (1.86-1.80) Outer margin of Baltica Subjotnian rapakivi complexes (1.58-1.5 Ga) Lower Allochthon Jotnian sedimentary formations (~1.5-1.25 Ga) Inner margin of Baltica Postjotnian dolerites (~1.25 Ga) Precambrian in windows SW Scandinavian domain (1.7 - 0.9 Ga) with Sveconorwegian reworking (1.2 - 0.9 Ga) Autochthon Phanerozoic cover rocks Sedimetary cover Precambrian basement COSC-1 drill site Figure 2.1. a) Interpretation of the tectonostratigraphy of the Scandinavian Caledon- ides (modified from Gee et al., 1985). b) Geology of the Baltoscandian Platform which forms the basement underneath the Caledonian cover, emphasizing formations related to the TIB and the Sweconorwegian orogeny (based on the bedrock geological map of Sweden, © Geological Survey of Sweden [I2014/00601] and Högdahl et al., 2004). The Rätan granite and Dala sandstones and granites are strongly correlated with the magnetic and gravity anomalies shown in Figure 2.4 and are thus thought to continue NNW underneath the Caledonian cover.

Olden-Oviksfjällen Antiform Offerdal Synform

Skardöra Antiform Skardöra Tännforsen Synform Tännforsen a) Mullfjället Antiform Åre Synform 01020kmKöli Nappe Complex Kallsjön calcareous phyllite phyllites, greywacke gabbro, peridotite, serpentinite

Caledonian Front Caledonian Seve Nappe Complex

Norway Sweden migmatitic gneiss, pyroxene granulites

Åre marble, calc-silicate gneiss, 100 500 calc-phyllite Storlien 900 Järpen amphibolites, metadolerites, 1300 Undersåker Liten Mörsil peridotite, serpentinite metapsammite, gneisses, Kallsjön 1700 b) 2100 mica aschicsts, marbles 2500 2900 3300 Storsjön Särv Nappe Sällsjö 3700 Åreskutan 4100 Neoproterozoic sedimentary 4500 Hallen Östersund formations -50 4900 Marby 0 +50 Fröå Offerdal Nappe 5300 +100 Åre Neoproterozoic sedimentary +150 formations +200 +250 Lower Allochthon N Åresjön Myrviken Silurian formations, +300 undifferentiated WE Hackås Ordovician formations, undifferentiated 012km S Cambrian Alum shales Neoproterozoic sedimentary COSC 2D Reflection Seismic profiles COSC-1 3D Reflection Seismic survey Drill sites formations COSC Byxtjärn-Liten, BL (2010) Explosive source points (2015) COSC-1 drill site Baltoscandian basement, COSC Kallsjön-Fröå, KF (2010) VIBSIST source points (2015) Alum Shale Project drill sites undifferentiated COSC Liten-Dammån, LD (2011) Receiver points (2015) Autochthon COSC Sällsjö, S (2014) Depth to basement Cambrian Alum shales COSC Dammån-Hallen, DH (2014) Other 2D Reflection Seismic profiles Isolines (m above sea level) Baltoscandian basement, COSC Seismic Profile CMP line CCT (1987-92) undifferentiated Figure 2.2. a) Regional bedrock geological map along the Swedish part of the Central Caledonian Transect, central Jämtland (based on the bedrock geological map of Sweden, © Geological Survey of Sweden [I2014/00601] and Strömberg et al., 1984). The 2D reflection seismic profiles are shown as well as the location of the drill sites for the Alum Shale project which took place in the late 1970’s. b) Local bedrock geology around the COSC-1 drill site, with the acquisition geometry of the limited 3D reflection seismic survey that took place after drilling was completed.

19 Thrust over this detachment are the Jämtlandian Nappes of the Lower Allochthon which comprise sedimentary successions of low metamorphic grade (sub-greenschist to low greenschist facies). These show a transition westwards from Ordovician and early Silurian carbonate-rich rocks deposited on the Baltica platform, into greywackes (turbidites) deposited in the Cal- edonian foreland basin (Gee, 1975a). As seen in a window in the core of the Mullfjället Antiform, basement-derived late Paleoproterozoic acid volcanic rocks are increasingly becoming incorporated into the base of the Lower Allochthon towards the west. The Middle Allochthon is thrust over the Lower Allochthon with a major mylonite zone separating the two. The lowermost units of the Middle Allochthon comprise highly deformed Precambrian granitic gneisses, such as the Tännäs Augen Gneiss Nappe and the Vemån Nappe found to the south of the study area. Overlying these are the psammitic formations of the Offerdal Nappe and the Särv Nappe, the former highly strained and the latter intruded by 600 Ma rift-related dolerites that are locally dominating the formations. Within thin zones of high strain and ductile deformation under high greenschist facies conditions, these dolerites have been rotated into becoming parallel with the dominant foliation (Gilotti and Kumpulainen, 1986). Thrust over the Särv Nappe is the Seve Nappe Complex (SNC), described in more detail below. This complex was originally included in the basal units of the Upper Allochthon but is today treated as the uppermost unit of the Mid- dle Allochthon. The Middle Allochthon is characterized by a metamorphic grade that increases upwards through the stratigraphy to granulite and locally eclogite facies in the central parts of the SNC, before abruptly decreasing to amphibolite facies at the top. The Köli Nappe Complex (mostly in greenschist facies) and related nappes of the Upper Allochthon were originally thrust on top of the Seve Nappe Complex (although this contact is now, in many places, a normal fault). These include greenschist facies metamor- phosed ophiolites, island-arc and back-arc assemblages and fossiliferous sedimentary rocks originating from the Iapetus Ocean floor (Gee et al., 2008; Gee, 2015). The entire assemblage of nappes is folded by major N-trending antiforms and synforms (Figure 2.2) that are related to late orogenic deformation. Some of the antiforms (e.g. the Skardöra antiform) are evidently built up by imbricate stacks of thrust sheets (Hurich et al., 1989) and the basement exposed in the core of Mullfjället antiformal window is likely to be allochthonous (Palm et al., 1991). These fold structures (e.g. the Tännfors synform hosting the Köli and Seve Nappe Complexes (Bergman and Sjöström, 1997)) also have clear signatures of superimposed syn- and post tectonic upper crustal extension. The nature of the autochthonous basement underneath the Caledonian cover in Jämtland is debated and one of the targets for the COSC project (Gee et al., 2010; Lorenz et al., 2011). Windows through the Caledonian

20 cover of the central Scandes, such as those at Mullfjället, reveal deformed and allochthonous basement units (mostly c. 1.7 Ga felsic volcanic rocks) that have been transported some distance from their original location. To the east of the Caledonian front (Figure 2.1b) is the Fennoscandian shield hosting a diverse collection of Paleoproterozoic magmatic units collectively called the Transscandinavian Igneous Belt (TIB) (Högdahl et al., 2004). East of the front, from north to south in the provinces of Jämtland, Härjedalen and Dalarna, are the Revsund granites (c. 1.8 Ga), the massive Rätan batholith (c. 1.68-1.7 Ga) and the Dala Province which includes a suite of granites, volcanites and associated sedimentary formations with ages ranging from 1.68 to 1.81 Ga (Högdahl et al., 2004). The latter is bounded to the south- west by the Sveconorwegian Frontal Deformation Zone (SFDZ, Fig- ure 2.1b). Mafic intrusions are found throughout the Fennoscandian shield with ages between 0.95 and 1.6 Ga (Söderlund et al., 2005). The Central Scandinavian Dolerite Group, with ages of about 1.2 Ga and often occurring as up to 100 meter thick sills, are exposed east of the Caledonian front from northern Dalarna and further northwards in Sweden and Finland (Söderlund et al., 2006). The Blekinge-Dalarna Dolerites, tracing the eastern side of the SFDZ from southern Sweden to Dalarna, intruded the Svecofennian crust during and after the main compressional phase of the Sveconorwegian Orogeny (c. 1.0 Ga) (Söderlund et al., 2005). The Rätan batholith and Dala granites and sandstone appear to extend northwest beneath the Caledonian cover and are considered the most likely components of the deep autochthonous basement in the study area (Dyrelius, 1985).

2.2 The Seve Nappe Complex The Seve Nappe Complex can be followed for at least 1000 km along the Scandinavian Caledonides and has been mapped for 200 km across the central parts of the mountain belt in Jämtland and Tröndelag Figure 2.1a. It is characterized by deformation under amphibolite to granulite, and locally eclogite facies, metamorphic conditions with partial migmatization and is thought to have been emplaced by ductile extrusion at mid-crustal levels while still at high temperatures (Gee et al., 2010). The formation and emplacement of this hot allochthon (Gee et al., 2012; Majka et al., 2014a; Grimmer et al., 2015) is central to understanding mountain building processes in the Caledonides (Gee et al., 2010) and other active orogens such as the Himalayas (e.g. Law et al., 2006) and the western Pacific including the Izu- Bonin-Mariana arc system (Tamura et al., 2010) and the Taiwan orogen (Malavieille and Trullenque, 2009). The SNC is usually subdivided into the Lower, Middle and Upper Seve Nappes. The contact between the Särv Nappe and the Lower Seve Nappe is

21 marked by an increase in metamorphic grade to amphibolite facies which is usually abrupt, but in some places more gradual (Gee, 2015). In the Åre Synform (Figure 2.2), this boundary is marked by a thick zone of phyllonite and mylonite (Arnbom, 1980; Lorenz et al., 2015a). The Lower Seve Nappe is evidently of a similar protolith as the underlying Särv, but ductilely deformed under amphibolite (locally eclogite) facies conditions (Arnbom, 1980; Lorenz et al., 2015a), indicating a higher pressure tectonic evolution than the Särv Nappe. Felsic, amphibole and calc-silicate gneisses dominate the lithology of the Lower Seve Nappe in the Åre area. Amphibolites are common with thick- nesses on the order of cm to several tens of meters and occur as lenses or boudins (Figure 2.3). Mica schists (sometimes garnet-bearing) become a more important component in the lower parts and marbles, pegmatite dykes, metagabbros and mylonites are also present (Arnbom, 1980; Lorenz et al., 2015a). The recent coring of a 2.5 km vertical section through the Lower Seve Nappe by the COSC-1 borehole revealed the mylonite zone to be unexpectedly thick. It is present in nearly 800 m of the lowermost part of the core and was not fully penetrated at the drilled Total Depth (TD) (Lorenz et al., 2015a; Paper III). The thickness and frequency of mylonite bands (often rich in garnet) increase through the upper 400 m of this zone. Below this, mylonites become the dominant component. About 150 m above TD, there is a transition into lower-grade metasediments. The tectonostratigraphic affinity of these are, however, yet to be determined and mylonites (often garnet- bearing) are still a dominant component in the lithology (Lorenz et al., 2015a). The Särv Nappe has been mapped to underlie the Lower Seve Nappe on the western flank of the Åre Synform (Strömberg et al., 1984) while only minor traces are found in the eastern flank (whether due to its absence or poor exposure is not clear) and it may thus pinch out below the synform (Paper III).

Figure 2.3. Amphibolite boudins surrounded by felsic gneisses in a quarry located about 4 km SE of the COSC-1 drill site. (Photo by P. Hedin)

22 The Middle Seve Nappe is represented in the study area by the well exposed Åreskutan Nappe, situated in the hinge of the Åre Synform as a klippen (Figure 2.2; Arnbom, 1980). A ductile shear zone which has been mapped to be roughly 50 m thick separates it from the Lower Seve Nappe (Majka et al., 2012). The unit is dominated by granulite facies migmatites, pelitic gneiss, metabasites and leucogranites. Discordant pegmatite intrusions cut the foliation within the Middle Seve Nappe but are deformed in the basal shear zone (Arnbom, 1980; Klonowska et al., 2014; Ladenberger et al., 2014). Eclogites occur locally in the Middle Seve Nappe in other nearby parts of Jämtland (e.g. Majka et al., 2014b). Recent discoveries of microdiamond inclusions in garnets within paragneisses at both Åreskutan (Klonowska et al., 2015) and Snasahögarna (Majka et al., 2014a) witness to the ultrahigh pressure metamorphism and subduction of the Middle Seve Nappe to depths exceeding 100 km. Dating of the Middle Seve Nappe at Åreskutan shows an evolutionary sequence of metamorphism during the emplacement of the SNC. The oldest monazite ages from Åreskutan of ~455 Ma may be related to prograde metamorphism corresponding to peak pressures (Majka et al., 2012) of mantle depths. A slightly younger age of ~440 Ma within the migmatites is interpreted as peak temperature metamorphism and decompression melting during rapid exhumation to mid-crustal levels (Klonowska et al., 2014; Ladenberger et al., 2014). The c. 430 Ma age of pegmatite intrusion in the lower part of the Middle Seve Nappe is evidence of cooling and transition from a ductile to a more brittle regime (Ladenberger et al., 2014). Samples from sheared migmatite at the base of the Middle Seve Nappe with ages of ~424 Ma contain a new generation of garnets within the basal shear zone are related to the final phase of thrust emplacement through the crust (Majka et al., 2012; Ladenberger et al., 2014). The Upper Seve Nappe is found within the Tännfors Synform farther north in the Scandes (Gee et al., 2012). The transition from the Middle to Upper Seve Nappe is marked by a decrease in metamorphic grade to amphibolite facies. Amphibolites, mica schists and metasandstones are the major components of this unit which is tectonostratigraphically bounded at the top by the greenschist facies outboard terranes in the Köli Nappe Complex of the Upper Allochthon (Bergman and Sjöström, 1997).

2.3 Geophysical Background A wide range of geophysical studies has been conducted in the Scandinavian Caledonides over the past few decades to complement the surface geological mapping and provide information on structures at depth. Several surveys were part of the Swedish contribution to the IGCP project number 27, including refraction seismic profiling (Palm, 1984), magnetometry (Dyrelius,

23 1980), gravimetry (Dyrelius, 1985) and petrophysical sampling (Elming, 1980). The first magnetic and gravity maps over Scandinavia resulting from this work were produced and published along with a geophysical review (Dyrelius, 1985) of the Scandinavian Caledonides in the compilation by Gee and Sturt (1985). Much focus has been along the well studied profile crossing the central Scandes in the provinces of Tröndelag and Jämtland with several comprehensive projects. In the late 1970’s, prospection of the black alum shales in Jämtland was accomplished by extensive drilling through a thick (up to 180 m) alum shale formation in the Lower Allochthon and reaching into basement, to map the shales at the basal thrust. As a result, the orogenic décollement was defined south of lake Storsjön to dip westwards at 1-2 degrees 30 km west from the front (Figure 2.2; Gee et al., 1978). Modeling of refraction seismic experiments (Palm, 1984) as well as aeromagnetic (Dyrelius, 1980) and gravity (Dyrelius, 1985; Elming, 1988) measurements supported and extended this gently west dipping geometry of the easternmost Caledonide sole thrust to other parts of Jämtland. The magnetic measurements revealed the vast Jämtland magnetic anomaly (Figure 2.4) that has been linked to the Rätan granite and the Dala granites of the TIB east of the orogenic front (Fig- ure 2.1b). Modeling of potential field data suggest that these TIB units extend towards the northwest underneath the Caledonian cover in western Jämtland with a thickness of at least 10 km (Dyrelius, 1980, 1985, 1986; Elming, 1988; Pascal et al., 2007; Ebbing et al., 2012). In the years from 1987 to 1992, ambitious reflection seismic surveys were undertaken along the so called Central Caledonian Transect (CCT) from the Atlantic coast in Tröndelag, to east of the Caledonian front in Jämtland (Hurich et al., 1989; Palm et al., 1991; Hurich, 1996; Juhojuntti et al., 2001). These were designed to image the entire crust down to Moho and revealed a highly reflective crust, throughout the entire section, down to depths of about 15 km. Although the uppermost few hundred meters to a kilometer were not imaged properly, the sections show remarkable correlation with both geological mapping of the Caledonian allochthons as well as with the previous geophysical results (Palm et al., 1991; Gee et al., 2010). The reflection seismic sections clearly showed that synforms hosted the higher allochthonous units and while some of the antiforms (e.g. Skardöra and Olden-Oviksfjällen, Figure 2.2) were clearly lifted by imbricate stacks, others are likely allochthonous basement derived units (e.g. Mullfjället, situated above continuous subhorizontal to west dipping reflections) (Palm et al., 1991; Hurich, 1996). This agrees with potential field modeling (Elming, 1988) which, additionally, provides a means to extend the interpretation of these major structures away from the reflection seismic sections (Dyrelius, 1985).

Figure 2.4. a) Topography/bathymetry of the north-western Scandinavia Caledonides. b) Istostatic residual anomaly gravity map of the Scan- dinavian Caledonides (based on Olesen et al., 2010c). The map is produced by application of Bouguer correction (using a standard density of 2200 kg/m3 and 2670 kg/m3 at sea and land, respectively) and Airy isostatic correction (assuming Moho depth of 30 km and mantle-crust density contrast of 300 kg/m3). The effects of topography and hypothetical crustal thickness (based on the observed topography) across the mountain belt are thus removed which enhances crustal anomalies such as the Jämtland anomaly. c) Total field magnetic anomaly map of the Scandinavia Caledonides (based on Olesen et al., 2010a). The Definitive International Geomagnetic Reference Field has been removed. The potential field maps are explained in further detail by Olesen et al. (2010b).

25 Subhorizontal to slightly west dipping reflections, strong and continuous over large distances, were traced from the orogenic front to a depth of 6 km at the Swedish-Norwegian border and interpreted to correspond to the basal thrust. Magnetotelluric measurements carried out along the Swedish part of the CCT revealed a strongly conductive layer which was attributed to the carbon rich autochthonous alum shales and shows good correlation with the reflections interpreted as the orogenic décollement (Korja et al., 2008). The basement hosts a remarkable suite of strong enigmatic reflectors. Their origin is unknown but they may be related to fault or shear zones from deformation during the Caledonian or perhaps Sveconorwegian orogenies. Another possibility is the abundant dolerite intrusions of the Fennoscandian shield. The c. 1.2 Ga Central Scandinavian Dolerite Group are found to out- crop just east of the front in Jämtland (Söderlund et al., 2006), where strong reflections project to the surface (Juhojuntti et al., 2001). The c. 1.0 Ga Blekinge-Dalarna Dolerites are found in the Dala granites and sandstones. The latter were found to produce a similarly strong seismic response in the Siljan ring area after correlation of subhorizontal reflectors with deep drilling results (Juhlin, 1990). Additional seismic sounding (CABLES, Schmidt, 2000) and passive seismics (SCANLIPS, England and Ebbing, 2012) in the same area that have helped to improve the interpretation of the large scale structures in the crust beneath the Caledonides of central Scandinavia. A new era has begun with the COSC project with several high resolution surveys including reflection seismic (Figure 2.2, Papers I, III and IV; Krauß et al., 2015; Simon et al., 2015), magnetotelluric (Yan et al., 2015) and potential field data (Paper II and unpublished data) in addition to the core analysis and downhole logging from the finished COSC-1 and planned COSC-2 boreholes.

2.4 Scientific Drilling in the Scandinavian Caledonides The Swedish Scientific Drilling Program (SSDP) was initiated in 2007 (originally named the Swedish Deep Drilling Program, SDDP) and successfully promoted Swedish membership in the International Continental Scientific Drilling Program (ICDP) in the year after (Lorenz, 2010). In 2009, a grant was received from the Swedish Research Council (VR) to develop and im- plement a new national infrastructure for scientific drilling, Riksriggen. An Atlas Copco CT20C mobile rig was purchased in 2012, capable of diamond wireline core-drilling in the common P, H and N sizes (hole/core diameters of 123/85 mm, 96/63 mm and 76/48 mm) down to 2.5 km depth. Mounted on crawlers and only requiring about 500 m2 of space during operation makes it very versatile and ideal for operation in remote areas with minimal impact to the environment (Rosberg and Lorenz, 2012).

26 The purpose of SSDP is the study of fundamental questions in Earth science that are of global importance, but unique to Scandinavia and require deep drilling. Several projects were initially proposed to study phenomena such as orogenic processes (e.g. COSC, Gee et al., 2010; Lorenz et al., 2011), active post glacial faults (DAFNE, originally PFDP, Kukkonen et al., 2010, 2011), impact structures (CISP, Högström et al., 2010) and ore genesis (PaMVAS, Weihed, 2010). The target specific projects of SSDP are bridged through societal relevance and several common scientific goals such as development of drilling and logging technology and methodology, and studies of heat flow and geothermal potential as well as the deep hydrosphere and biosphere (Lorenz, 2010). The COSC project is designed to investigate the processes that were central to the formation of orogenic belts and particularly the Scandinavian Caledonides. Comparison with modern analogues such as the Himalaya- Tibet mountain belt, may increase the understanding of mountain building processes in general and aid the interpretation of presently active orogens (Gee et al., 2010). The main goals are several and may be summarized as follows:

1. An increased understanding of mountain building processes with a focus on processes and conditions governing the vast horizontal displacements; 2. Calibrate surface geology and geophysics in the area and reveal the nature of the sources for the observed strong reflectivity, conductivity and vast magnetic anomaly. 3. Investigation of heat flow and the geothermal gradient as well as the influence of major paleoclimatic variations; 4. Investigate the current hydrogeological and hydrochemical state of the mountain belt and deep water circulation; 5. Probe the various geological formations for the presence of a deep biosphere and possibly analyze its diversity and function.

The increased understanding of, for example, the deep hydrogeological environment and water circulation, the deep conditions for underground engi- neering and infrastructure, ore genesis/mineralization, and the geothermal potential may have direct implications and benefit for industry and society. The science of the COSC project involves a large number of international research groups and has been divided among six main working groups; geology/thermochronology, geophysics, geothermics, hydrogeology, microbiol- ogy and drilling management and technology. The project involves two fully cored c. 2.5 km deep boreholes located at different tectonostratigraphic levels within the orogen. This will essentially create a 5 km combined section that starts in the Lower Seve Nappe of the Middle Allochthon and passes through the underlying lower grade

27 allochthons before penetrating the basal décollement and reaching into the Precambrian Baltica basement. The first borehole, COSC-1 (ICDP drill site 5054-1-A, IGSN: ICDP5054EEW1001 (Figure 2.2; Lorenz et al., 2015b)), is focused on investigating the formation and ductile emplacement of the SNC (described in detail above) and its influence on adjacent units. It focuses on the degree, type and timing of deformation through the evolution of the Lower Seve Nappe (Gee et al., 2010). The second borehole, COSC-2, has several primary objectives. It will first drill through the Lower Allochthon to study the character of the deformation and the metamorphic gradient through these thrust emplaced units. It will then study the nature of the basal décollement and the autochthonous Cam- brian alum shales that have facilitated the vast horizontal transportation of the overthrusting allochthons. Finally, the borehole will penetrate into the subducted Precambrian basement and cut at least one of the enigmatic subhorizontal to slightly west dipping basement reflectors. This will both give conclusive insights about the basement lithology and the sources for the reflections and allow investigation of the deformation and metamorphism during Caledonian orogeny and possibly remnants of older phenomena such as the Sveconorwegian orogeny (Gee et al., 2010). Defining the subsurface location and geometry of the specific targets is essential for identifying the most suitable sites for deep drilling. 2D reflection seismic profiles were therefore acquired in 2010 for the purpose of imaging the allochthons and basement in western Jämtland in high resolution (Figure 2.2a, Paper I). These 2D reflection seismic data were also used together with potential field data (Paper II) and for pseudo 3D processing (sections 3.2.2 and 3.2.3) to extract more information about the targeted SNC. After the drilling of COSC-1 in 2014, with nearly complete core recovery, a wealth of data from on-core and downhole measurements were obtained (Lorenz et al., 2015a, 2015b). This was followed by further high resolution 2D and 3D seismic surveys in and around the COSC-1 borehole (Fig- ure 2.2b, Paper III; Krauß et al., 2015; Simon et al., 2015). The optimum location for COSC-2 could not be defined from the original 2D reflection seismic profile, because the slightly west dipping basal décollement was interpreted to become within drillable depth only at the very eastern edge of the section (Paper I). The 2D reflection seismic profile was therefore subsequently extended towards the southeast in 2011 and 2014 (Figure 2.2a, Paper IV). This entire c. 55 km long main seismic profile, the COSC Seismic Profile (CSP), was complemented by magnetotelluric measurements in 2013 (Yan et al., 2015). Two potential sites for COSC-2 have been proposed based on these new data (Paper IV) and this second major phase of the COSC project will be planned in detail during and following an international science workshop that will be held in October 2015.

28 3 Methods

Geological targets of interest, whether for basic research or prospecting for economic resources, are often situated in geological settings which are difficult to reach for direct sampling (Yilmaz, 2001). For a study of large and complex formations, and the processes that may have formed them, it is desired to obtain a detailed geological section through them. In the Swedish Caledonides, accessible outcrops of exposed bedrock are sparse and key areas are not exposed at all. In this case, scientific drilling is an excellent way to acquire continuous datasets with a wealth of information, including geological, geophysical, geochemical, petrophysical, hydrogeological and geothermal (Lorenz et al., 2015a). Drilling is, however, an expensive procedure and will give information which is, although very detailed, limited to a 1-dimensional line intersecting 3-dimensional space. Therefore it is important to have a geometrically well defined target prior to drilling, as well as the means to extrapolate results from the borehole into the surrounding formation. This is done by combining geological mapping with surface geophysical techniques, such as reflection seismic, potential field and electromagnetic surveying. The work of this thesis focuses primarily on the reflection seismic data acquired 1) to define the optimum location of two drill sites for scientific deep boreholes with the specific targets of the COSC project (Papers I, II and IV), and 2) to image the geological structure around the COSC-1 borehole (papers II and III).

3.1 The reflection seismic method and data acquisition The reflection seismic technique is based on the reflection of elastic energy incident on an interface associated with significant contrast in impedance (the product of velocity and density). In practice, a wave is first generated by a controlled source, for example explosives, vibrators, hydraulic hammers or weight drops. The wave then spreads out spherically from the source location and travels through the ground to a receiver which produces a continuous record of ground motion with time. The resulting seismogram, or seismic trace, is a record of direct waves, refractions, reflections, diffractions and scattering from the source generated signal as well as ambient noise (e.g.

29 natural seismicity, wind, rain, interference from electrical equipment and human activities). The acquisition can be done in 2D with sources and receivers along the same line (as in the case of the CCT, Palm et al., 1991; Juhojuntti et al., 2001), in 3D with several source and receiver lines (commonly orthogonal) covering a surface grid (e.g. Milkereit et al., 2000), or in semi 3D with source and receiver lines at an angle with each other (e.g. multi-azimuthal walkaway Vertical Seismic Profiles (VSP) and cross-profile acquisition (Rodriguez-Tablante et al., 2007; Malehmir et al., 2011)). The type of acquisition to use depends on the target and its geological setting and the questions one aims to answer. Most often this choice is also severely restricted by logistical considerations, permissions from land owners and state agencies, as well as available funding and equipment. Multifold acquisition on land is designed to increase the signal-to-noise (S/N) ratio of reflected signal and involves multiple receivers (geophones) planted along lines that simultaneously record the incoming seismic signal from multiple consecutive source locations. Assuming horizontal reflectors, the signal recorded by a receiver contains reflections below the source- receiver midpoint. By binning seismic traces which have nearly the same midpoint location and stacking horizontally onto a Common Midpoint (CMP) line, the S/N ratio can be significantly improved (Yilmaz, 2001). Note that the term Common Depth point (CDP) is often used interchangea- bly with CMP, as has been done in some of the papers that are a part of this thesis. They are, however, not the same; the CDP refers to the true location of the reflection, which only coincides with the CMP if all surfaces are horizontal. When structures are dipping, the reflection point will not be at the source- receiver midpoint. The acquisition line is therefore commonly chosen to be in the expected dip direction so that reflections occur in the vertical plane below the stacking line. The geometry of structures which have a considerable cross profile dip (crossdip) component, resulting in out-of-the-plane reflections, cannot be accurately reproduced in a 2D survey (Larner et al., 1979; Wu et al., 1995; Nedimović and West, 2003a).

3.1.1 2D crooked line acquisition In the case of rough terrain and thick vegetation, the accessibility with vehicle mounted seismic sources (e.g. vibrators, hydraulic hammers and weight drops) is often severely limited and acquiring seismic data along straight lines is not feasible. The acquisition is therefore often restricted to follow available roads and paths along a crooked line. This is the case for western Jämtland and all of the 2D reflection seismic profiles for the COSC project were acquired along crooked lines (Figure 2.2, Papers I, II and IV) using the vehicle mounted hydraulic hammer shown in Figure 3.1, the VIBSIST.

30 2D crooked line acquisition introduces several challenges to subsequent processing related to uneven fold distribution, irregular representation of offsets and deviation from a straight line. The resulting source-receiver midpoints will be spread out over the surface and have a perpendicular offset from the stacking line. In the presence of structures that dip in a direction other than the stacking line, the reflections will stack out of phase and reduce the quality of the image. On the other hand, such a dataset with midpoint traces laterally offset from the stacking line contain information about the 3D geometry of the reflectors that may be extracted by various techniques (Larner et al., 1979; Wu et al., 1995; Nedimović and West, 2003a, 2003b; Malehmir et al., 2009; Lundberg and Juhlin, 2011).

Figure 3.1. The vehicle mounted hydraulic hammer (the VIBSIST) that was used as the seismic source for reflection seismic acquisition in 2010, 2011 and 2014. The recording truck is seen in the background. (Photo by P. Hedin)

3.1.2 Limited 3D acquisition For 3D reflection seismic acquisition in rough terrain, where sources are restricted to roads and trails, source points may be located along several crooked lines crossing the area. Geophones are more easily deployed in the terrain and a regular grid of receiver lines may still be used. In severe terrain, such as mountainous areas, also the receivers may have to be deployed in an irregular or even random fashion. Although a large 3D volume of reflection seismic data can be obtained, the acquired data may suffer from substantial unsampled regions and be severely limited in the representation of offsets and azimuths. These limitations in the acquisition geometry may leave an acquisition footprint (Marfurt et al., 1998) that propagates into the data and generate artifacts that are hazardous to subsequent processing and interpretation (Cheraghi et al., 2012).

31 The major seismic survey that was conducted in and around the COSC-1 borehole after completion included a limited 3D reflection seismic experiment (paper III). In this setting, receivers were planted in a square geometry centered on the drill site and covering a surface area of roughly 1.5 km2 (Figure 2.2). A combination of explosives and a mechanical source (Figure 3.1) were excited along roads through the area, forming a star pattern of crooked lines radiating outwards from the drill site.

3.2 Reflection seismic processing In raw reflection seismic data, the desired reflection signal is obscured by refracted and scattered signal, as well as ambient noise. Furthermore, an original impulsive source signal is altered into an extended wave-train due to varying subsurface conditions along the travel path. A variety of seismic processing techniques are therefore applied to suppress unwanted signal and noise and to produce sharp reflections with properties and geometries that represent the true subsurface structures (Yilmaz, 2001). The three fundamental steps are 1) Deconvolution (collapsing the waveform of the recorded signal into an impulse like signal) to enhance temporal (vertical) resolution; 2) stacking (horizontal summation of seismic traces to a common midpoint (CMP) line) to enhance the S/N ratio through constructive and destructive interference of signal and noise, respectively; and 3) migration to move dipping reflections to their correct position, collapse diffractions and enhance the spatial (horizontal) resolution. Additionally, various filtering techniques to suppress noise, the application of amplitude gain functions and trace balancing are commonly applied to enhance the quality of the final image. More detail about the processing steps and different techniques may be found in Yilmaz (2001). Variations in the subsurface (e.g. structural, lithological, topographical, and the characteristics of the near surface) along the path from source to receiver, cause time delays that vary from trace to trace. Misalignment of signal causes destructive interference of reflected signal and loss of information. Therefore, the application of vertical time shift corrections prior to stacking is a central part of the processing. Refraction and residual static corrections are applied to adjust for irregular time delays caused by varying topography and near surface characteristics. Time delays of a specific reflection caused by varying source-receiver offsets are hyperbolic in nature and are remedied with Normal Moveout (NMO) corrections. Dip Moveout (DMO) corrections can be applied to correct for the effect of conflicting dips, and can serve as an alternative partial pre-stack migration which is less sensitive to the velocity model (Yilmaz, 2001). In conventional 2D processing, it is assumed that the reflected signal is contained within the vertical plane between source and receiver, i.e. there are

32 no changes in velocity and structure in the perpendicular direction. This is seldom the case in reality and the central assumptions for CMP stacking are incorrect. This is especially true for data acquired along 2D crooked lines which are often associated with a significant spread in source-receiver azimuth as well as large variations in fold and offset distribution along the line (Larner et al., 1979; Wu, 1996; Nedimović and West, 2003a, 2003b). This may lead to transparent zones and amplification of coherent noise. Reflec- tions from structures with a dip which is oblique to the CMP line do not stack in phase, ultimately resulting in lower resolution and loss of information. Structural and lithological variations which are 3D in nature require 3D acquisition and processing to reproduce an accurate geometry in the final image. Despite the challenges with processing 2D crooked line data, its pseudo 3D nature can, however, be exploited to gain information about the 3D geometry of reflectors through various techniques such as crossdip analysis and moveout corrections (Larner et al., 1979; Wu et al., 1995; Nedimović and West, 2003a; Rodriguez-Tablante et al., 2007; Lundberg and Juhlin, 2011), azimuthal binning (Lundberg and Juhlin, 2011) and swath 3D imaging (Nedimović and West, 2003b; Malehmir et al., 2009, 2011). As part of the preparations for the COSC-1 drilling, a cross dip analysis and swath 3D imaging was done on the 2D reflection seismic profiles in the vicinity of the proposed COSC-1 drill site to extract more information about the structures. Knowledge of lithological and structural variations along the planned borehole trajectory is of importance not only for the planning of the drilling activities, but also the sampling and logging programs by identifying potential zones of interest.

3.2.1 Crossdip analysis An analysis of cross-dip can be used as a pseudo 3D technique to extract more information from a 2D survey over complex structures. Correcting for the Crossdip Moveout (CDMO) can greatly improve the coherency of reflected energy, resulting in a much higher quality image (Larner et al., 1979; Wu et al., 1995; Nedimović and West, 2003a; Rodriguez-Tablante et al., 2007; Lundberg and Juhlin, 2011). Knowledge of the inline dip and crossdip component allows estimation of true dip and strike. In the case of a crooked line acquisition, midpoints have an offset perpendicular to the stacking line as shown in Figure 3.2. The delay in travel time for a given dipping reflector varies as a function of offset, but is neither hyperbolic nor surface consistent and cannot be corrected for by normal moveout or static corrections. The traveltime delay, Δtij, of energy in a midpoint trace with perpendicular offset, yij, from the CMP line, can be calculated as

33 2 ∆ where φ is the crossdip angle, v is the velocity above the reflector and indices i and j refer to the jth trace of the ith CMP bin. By calculating the Crossdip Moveout (CDMO) correction for a range of crossdip angles on NMO corrected CMP gathers for a constant velocity (Rodriguez-Tablante et al., 2007; Lundberg and Juhlin, 2011), a rough estimate for the crossdip component of a particular reflector may be attained. A straight CMP line is used to simplify the calculations (Wu et al., 1995).

Figure 3.2. Schematic figure illustrating the binning geometry of a crooked line 2D acquisition profile and the parameters used in crossdip analysis (based on Nedimović and West, 2003a). Δtij is the Crossdip Moveout (CDMO) correction, ϕ is the crossline dip angle, v is the average velocity, yij is the perpendicular offset to the stacking line and indices i and j refer to the jth trace of the ith CMP bin.

The structures within the SNC, although highly reflective, were not clearly imaged after regular 2D processing (Paper I). The nature and limited continuity of the reflections observed within the unit suggested a complex 3D geometry of the structures which was not handled properly. The maximum perpendicular offset of midpoint traces included in the stack was 500 m, which assumes that no variations occur over this distance. A crossdip analysis was therefore performed during the preparations for the COSC-1 drilling in an attempt to gain more structural information prior to drilling. The chosen location for the COSC-1 borehole is close to where the BL and KF profiles meet and in an area where both profiles are very crooked (Figure 3.3). Furthermore, this location is in the middle of an almost 2 km long part of BL where the source point spacing had been intentionally reduced to 10 m (Paper I). These factors combined result in a large lateral spread of midpoints (Figure 3.3). The crossdip analysis was performed assuming a constant velocity of 5800 m/s and CMP bins were defined to have

34 a half width of 600 m in the perpendicular direction from the straight CMP line (Figure 3.3). The midpoint spread varies from 1 km in width and is symmetrical about the stacking line (around CMP 300 in Figure 3.3), to 100 m in width with about a 500 m asymmetric offset (around CMP 700 in Figure 3.3). Figure 3.4 shows a selected region of the BL-profile after application of CDMO for crossdip angles of -24° (towards NE) to +24° (towards SW). Application of CDMO improves the coherency of reflections related to major boundaries outside of the SNC (marked by the dashed line at 0° crossdip in Figure 3.4). Several strong subhorizontal reflections below the SNC, such as R2 and R3, become more coherent by a positive crossdip correction that is increasing from +6° in the southeast to +12° beneath the SNC. These structures thus have a slightly more westwards dip than indicated in the standard 2D processing of the BL profile. Within the SNC, reflection coherency is improved locally over short distances at varying CDMO over the entire range of crossdip angles. This indicates a complex geometry of structures within the unit, with conflicting dips and smaller lateral extent than the used bin-size. Crossdip analysis of the KF profile (not shown here) show the same indications of complex structures within the SNC.

Figure 3.3. Bedrock geology map showing the acquisition geometry of the BL and KF profiles near the COSC-1 drill site (based on the bedrock geological map of Sweden, © Geological Survey of Sweden [I2014/00601] and Strömberg et al., 1984). The resulting spread in midpoint traces from the BL profile is shown, color coded by source-receiver offset, together with the straight stacking line used for the crossdip analysis.

Figure 3.4. Selected region of the BL profile with application of CDMO correction for crossdip angles of -24° (dipping towards NE) to +24° (dipping towards SW). The dashed line at 0° crossdip is the base of the SNC according to the interpretation in Paper I. CDMO correction from a wide range of crossdip angles improve reflector coherency locally over short distances within the SNC. CDMO correction for +6° to +12° crossdip improves the coherency reflections below the SNC.

36 The estimated crossdip angles for the reflectors should, for several rea- sons, be viewed with caution in any interpretation. The assumed constant velocity may be a valid approximation for the crystalline environments in many parts of Scandinavia (e.g. Lundberg and Juhlin, 2011), but may be less accurate for the complex Lower Seve Nappe (Paper III). The CDMO correction is more sensitive to changes in angle than velocity, especially at shallow dips (Lundberg, 2014). For the large variations in the SNC, even crossdip angles of 10° may be associated with uncertainties of several degrees (Lundberg, 2014). In the calculation of strike of the reflectors, these uncertainties are amplified for shallow dipping structures (García Juanatey et al., 2013). Furthermore, these calculations assume interfaces to be planes that extend without change in geometry over the region included in the binning, which is not the case for structures in the area, especially within the SNC. The analysis thus confirms an important 3D geometry of structures within the SNC but does not allow precise determination of these geometries. Fur- thermore, the CMP line both in the regular 2D processing and this crossdip analysis are some distance away from the COSC-1 drill site and the geometries are too complex to draw any conclusions about the geometry of the structure along the borehole trajectory. Therefore, actually imaging of the reflectors in 3D would have been needed to give more relevant information for the preparations of the drilling and sampling programs.

3.2.2 Swath 3D imaging Because of the often considerable areal coverage of midpoints resulting from 2D crooked line acquisition, the data can essentially be treated as a sparse 3D dataset and produce a swath 3D image (Malehmir et al., 2009, 2011). This swath 3D seismic volume is less affected by lateral variations in structure as binning is done over much smaller distances and it potentially allows delineation of the 3D geometry of structures. The datasets are however often restricted by several limitations; 1) significant unsampled regions; 2) low and variable fold; 3) narrow azimuth coverage, limited to the direction of the acquisition lines; and 4) varying representation of offsets. Far offset data, which is required for imaging dipping reflectors, are only available in the direction of the acquisition line and therefore reflectors with considerable cross dip components will not be clearly resolved (Cheraghi et al., 2012). Near offset data, needed for imaging the near surface, are only available in the vicinity of the acquisition line.

Figure 3.5. a) Bedrock geology map around the COSC-1 drill site showing the acquisition geometry (based on the bedrock geological map of Sweden, © Geological Survey of Sweden [I2014/00601] and Strömberg et al., 1984). b) The processing geometry with inlines roughly aligned with the BL acquisition profile. c) The offset- azimuth distribution is irregular and shows the limited representation of far offsets to a narrow band of azimuths.

The areal distribution of trace midpoints of the combined data from the BL and KF profiles is large around the COSC-1 drill site due to the acquisition geometry (Figure 3.5). After merging the two datasets, a processing geometry was defined with the inline direction aligned with the general direction of the BL acquisition line as shown in Figure 3.5b. The inlines are thus parallel in their western part with the CMP line of the BL-profile, allowing for better comparison between the two. The midpoints were binned into 20 m × 20 m CMP bins, resulting in a maximum fold of c. 200 although the fold distribution is highly varying and low on average. Figure 3.5c shows the offset-azimuth distribution of this geometry. As seen in Figure 3.5b and 3.5c, the data severely suffer from all of the above mentioned limitations.

38 Doing a full 3D pre-stack processing of this 2D dataset was deemed un- necessary and the already pre-stack processed data (as in Paper I) were first updated with the new geometry and then used as input to a standard 3D processing sequence. Velocity analysis was done in several iterations to find the optimal stacking velocities. Coherency filtering (FX-Deconvolution) was applied after stacking to remove noise and artifacts introduced during the pre-stack filtering and stacking process. Partial pre-stack migration with DMO corrections in 3D were tried, but dismissed as they did not give any improvement on reflection coherency within the SNC (as was also the case in the 2D processing of this dataset, Paper I). Although a full pre-stack migration may have given a better result (given an accurate velocity model), a post-stack time migration routine based on a finite difference algorithm was used because of its considerably lower computational requirements. Furthermore, only semi 3D migration was performed by doing a 2D migration first in the inline direction and then in the crossline direction. This could potentially introduce artifacts in the migrated sections, especially surrounding initially empty regions. Therefore CMP bins within the volume that were initially empty were muted after each pass of the 2D migration to limit the creation of such artifacts. Finally, depth conversion was done on the seismic 3D cube to simplify the interpretation process. Inline 1038 from the migrated swath 3D seismic volume is shown in Fig- ure 3.6a next to an extracted section from the BL-profile (Figure 3.6b) for comparison. Deep reflections, especially below 1 s, are not imaged as clearly in the 3D inline as in the 2D section owing to the reduced fold. Near surface information is also missing because of the lack of near offset traces away from the acquisition profile. On the other hand, reflections around 0.4-0.5 s are more pronounced in the 3D data, possibly because of the smaller bin-size and less destructive interference from out-of-the-plane signal.

Figure 3.6. A comparison of inline 1038 of the Swath 3D (a) and an extracted section from the BL 2D profile (b), roughly parallel on the interval inline 1000-1175 and CDP 329-700. Above 0.2 s and below 1 s, the image quality in Swath 3D is much lower due to underrepresentation of near offset trace midpoints and the lower fold. Reflections around 0.4 to 0.5 s are however more pronounced and continuous in the Swath 3D, likely owing to less destructive interference from out-of-the-plane reflections. The vertical black line shows a horizontal projection of the approximate planned borehole trajectory.

The location of the COSC-1 drill site relative to the seismic profiles (both governed by road access) means that the borehole is at the very edge of the midpoint distribution. The upper few hundred meters near the borehole are not well imaged and could host interfaces with significant impedance contrasts that cannot be predicted from the seismic data. The low fold and unsampled regions surrounding the borehole limits the imaging capabilities along the trajectory (Figure 3.7). However, as seen in Figure 3.8, clear reflected features can be traced from the more clearly imaged central part of the seismic volume out towards the borehole trajectory despite a highly variable fold. Distinct reflections are seen at depths of about 900 m below the drill site (Figure 3.7) and are interpreted to be due to a mafic lens within the generally gneissic Lower Seve Nappe. Gently east dipping reflections observed between 2.1 to 2.3 km depth (Figure 3.7), followed by a significant decrease in reflectivity were interpreted to represent the base of the Lower Seve Nappe and underlying thrust zone with a transition into lower grade metamorphic rocks of the underlying allochthons. Although limited in several aspects, the migrated 3D volume gave clues to the 3D geometry of some of the major structures expected to intersect the borehole. 3D processing of crooked line 2D data thus allowed better predictions to be made prior to drilling of the COSC-1 borehole.

Figure 3.7. Inline 1022, which crosses the COSC-1 drill site, is shown after migration and depth conversion. The vertical black line shows a horizontal projection of the approximate planned borehole trajectory.

Figure 3.8. 3D view of the swath seismic volume showing inline 1043 and crossline 1034 as well as two depth slices at 950 m and 2250 m below the surface, respectively. Lateral continuity of reflections from the well imaged center of the volume out to the edge allows interpretation at the actual planned trajectory of the COSC-1 borehole.

41 3.3 Drilling of COSC-1 scientific borehole and logging of geophysical rock parameters Based on the site investigations it was concluded that the area around the abandoned copper mine at Fröå in western Jämtland meets the scientific criteria of the COSC-1 borehole, to obtain continuous sampling through an as thick as possible section of the Lower Seve Nappe. The final location was selected on the basis of technical and practical considerations, including site accessibility and supply of electrical power and water, as well as permissions from the landowner and a positive community. Preparations began with the construction of the drill site in August 2013. In September the first 103 meters of the COSC-1 borehole were drilled and a conductor casing was installed. Two additional holes, 50 m and 100 m deep, respectively, were drilled and equipped with seismometers for passive seismic monitoring during the drilling operations. By late April 2014, the drill site was fully mobilized with the Riksriggen drill rig and the technical, scientific and common facilities all contained within an area of 30 m by 35 m. Drilling operations began on the 28th of April, with the first core retrieved on the 1st of May and a total drilled depth (TD) of 2495.8 m was reached on the 26th of August. The photo in Figure 3.9 was taken about half way through the project. The lithology encountered in the core may be roughly subdivided in two parts. The Lower Seve Nappe, comprising alternating layers of felsic calc- silicate/gneisses and amphibolites (Figure 3.10a,b), is interpreted to extend down to 1710 m depth. The presence of thick units of amphibolite at depths of around 900 m fits well with the strong reflectivity observed in the swath 3D imaging (section 3.2.2). Below 1 km, mafic units are still common, but are generally much thinner than above. Below 1710 m depth there is an obvious increase in strain with depth. Mylonite bands increase in thickness and frequency to a depth of about 2100 m, below which mylonites are the dominant component. A core section from the mylonite zone is shown in Figure 3.10c. The deepest occurring mafic rocks were encountered at about 2314 m and a transition from gneissic to lower-grade metasedimentary rocks occurs around 2350 m (Lorenz et al., 2015a). This correlates well with the decreased reflectivity observed at this depth. Although metasandstones are a prominent component in the lowermost c. 150 meters, their tectonostratigraphic affinity have not yet been determined and the mylonites are still dominating the lithology at TD. Thus, this mylonite zone, the thrust zone that separates the SNC from underlying allochthons, was not fully penetrated by the borehole and is thus more than 800 m thick at the COSC-1 location.

Figure 3.9. The “Riksriggen” drill rig and about 1400 m of drill pipe at the COSC-1 drill site. (Photo by H. Lorenz)

Figure 3.10. 360° images of three core sections from the COSC-1 borehole. a) The typical sharp boundary between felsic gneisses and amphibolite at depth of about 745 m (sections 5054-1-A-228Z-1WR (IGSN: ICDP5054ESVT001) and 5054-1-A- 228Z-2WR (IGSN: ICDP5054ESWT001)). b) A section of the COSC-1 core from around 790 m showing some of the thin layering of felsic and mafic rocks (sections 5054-1-A-243Z-2WR (IGSN: ICDP5054ESUV001) and 5054-1-A-243Z-3WR (IGSN: ICDP5054ES1W001)). c) A section from the mylonite zone at a depth of about 2340 m (section 5054-1-A-671Z-4WR (IGSN: ICDP5054ESSL201). The red marker line on the core is on the right when looking towards the top.

43 Aside from the continuous geological section through the Lower Seve Nappe obtained with the core, on-site and off-site logging of the entire core, downhole logging, and monitoring of the circulating drilling fluid and gasses emerging from the borehole, produced continuous datasets for a large number of geophysical, petrophysical, geochemical, hydrogeological and geothermal parameters. As part of the on-site science, the core was logged using a Multi Sensor Core Logger (Geotek MSCL-S, Figure 3.11, provided by the ICDP) which measured the core diameter deviation, P-wave travel time, gamma attenuation and magnetic susceptibility. Due to problems with the sensor for diameter deviation (drifting) and P-wave travel time (suggesting unreasonably low velocities), reliable datasets were, in the end, obtained for rock density (calculated from a constant core diameter and gamma attenuation) and magnetic susceptibility. The core was also later XRF-scanned off- site for geochemical data and the content of 18 minerals and trace elements.

Figure 3.11. The Multi Sensor Core Logger (MSCL) measuring gamma attenuation, core diameter deviation, P-wave travel time and magnetic susceptibility on a core section at the COSC-1 drill site. (Photo by P. Hedin)

On several occasions, during natural breaks in the drilling, the hole was logged by personnel from Lund University, to secure data in case of hole loss. After TD was reached, more comprehensive logging campaigns were conducted by Lund University and the ICDP Operational Support Group (OSG). A full suite of logs were obtained, including full waveform sonic, sidewall density, temperature, electrical rock resistivity, fluid conductivity, magnetic susceptibility and acoustic televiewer. However, instrument limitations and technical issues with the equipment, prevented some sondes from reaching the full depth of the borehole. The downhole logs of sonic velocity and density, most relevant for correlation with the reflection seismic images, for example, reached a depth of only about 1605 m.

44 All of the downhole sondes had natural gamma sensors and a continuous natural gamma log from the surface to TD obtained from the post drilling OSG logging campaign was used as reference for depth matching all other logs. The data from the MSCL were depth matched with the downhole logs using the magnetic susceptibility log, thus providing density measurements down to TD. In the uppermost 1600 m, the downhole and MSCL logs of density show good correlation, although the MSCL measurements are slightly noisier due to the many transitions between core sections, and have a slight but constant shift towards lower densities. A large contrast in both density and P-wave velocity (vp) exists between the felsic and mafic rocks. This is visualized in the cross plot of density ver- sus vp (Figure 3.12) where two distinct distributions are observed; a broad 3 distribution around densities and vp of 2800 kg/m and 5700 m/s, respectively, and a more concentrated distribution around densities and vp of about 3000 kg/m3 and 6000 m/s. The sharp interface between felsic and mafic rocks (Figure 3.10b) is believed to be capable of producing strong reflections under favorable conditions. Unfortunately, no reliable sonic velocity data exist from within the interpreted mylonite zone. Correlation of the geology and logs of vp and density with the seismic data is done in Paper III. For details on the preparations, drilling, on-site science and logging, see Lorenz et al. (2015b).

Figure 3.12. The crossplot of density vs. vp shows a broad distribution, attributed to 3 felsic rocks, is observed around density and vp of 2800 kg/m and 5700 m/s, respectively. A second, more concentrated distribution, observed around density and vp of 3000 kg/m3 and 6000 m/s, respectively, is attributed to mafic rocks. Color coded by the number of data points normalized by the maximum number of data points in one bin. Plotted in grey are the isolines of acoustic impedance showing that the impedance contrast between these two major lithological groups is about 2×106 kg/m2s.

45 4 Summary of Papers

4.1 Paper I: Seismic Imaging of the Scandinavian Caledonides to define ICDP drilling sites Two targets had been previously identified for the COSC project (Gee et al., 2010). The aim with the first of the two 2.5 km deep scientific boreholes, COSC-1, was to study the lower unit of the Seve Nappe Complex (SNC, described in detail in section 2.2 above) near the town of Åre in western Jämtland. The second borehole, COSC-2, will be placed further to the east and target the basal detachment which separates the lowermost thrust nappes of the Caledonian cover from the underlying autochthonous basement. COSC-2 will also try to penetrate and investigate the enigmatic strong basement reflectors (Gee et al., 2010). High resolution 2D reflection seismic profiles were acquired in 2010 to identify the most suitable locations for the two scientific boreholes. This paper focuses on the acquisition, processing and interpretation of the two profiles in relation to the objectives of the COSC project.

4.1.1 Summary A nearly 36 km long main profile, stretching from lake Byxtjärn in the west to the southern tip of lake Liten in the east (the BL-profile), was designed to image the uppermost part of the crust and to be used as a base for placing the two COSC boreholes. A second north-south directed profile, about 7 km long, was acquired between lake Kallsjön and the old abandoned copper mine at Fröå (the KF-profile) to link the main profile with the previously acquired reflection seismic section along the CCT. For both of these profiles, a rock breaking hydraulic hammer mounted on a construction vehicle (VIBSIST, Figure 3.1) was used as a source, hitting the ground repeatedly with increasing frequency over a period of time. Source locations were activated every 20 m and normally involved 3-4 sweeps with more than 100 hits each. The signal was recorded using 300-360 active channels with 28 Hz geophones. Along the BL profile, receivers were placed at 20 m separation and the split spread was continuously rolled along with the source. A fixed spread with a receiver spacing of 25 m was used for the KF profile.

Figure 4.1. a) The stacked and b) migrated and depth converted section of the Byxtjärn-Liten (BL) profile. In c), our interpretation is plotted on top of the migrated and depth converted section in (b). The SNC, the target of the COSC-1 borehole, is shown in red together with an approximate borehole trajectory at the proposed drill site. The west dipping basal décollement, with associated alum shale, is represented by the Cambrian cover in green color. It comes to within a depth of about 2.3 km below the surface at the eastern limit of the section and thus the interpreted section does not reveal a suitable location for achieving the objectives with the COSC-2 borehole.

The VIBSIST data were decoded following the Swept Impact Seismic Technique (SIST) (Park, 1996; Cosma and Enescu, 2001) down to 3 s (corresponding to c. 9 km depth) and then manually edited to remove low quality data (e.g. due to bad weather conditions, environmental noise or instrument issues). Processing followed a standard sequence including static corrections, deconvolution, spherical divergence compensation, filtering, and moveout corrections before stacking to a crooked CDP line (Figure 4.1a). The stacked section was then migrated using a finite difference algorithm

47 and finally depth converted for the purpose of interpretation. The migrated and depth converted section is shown in Figure 4.1b and, with the interpretation overlain in Figure 4.1c. The processing of the KF profile differed slightly from that of the BL profile in that spectral equalization was used in the pre-stack processing while predictive Wiener deconvolution and DMO corrections were not performed (Figure 4.2).

Figure 4.2. a) The stacked and b) migrated and depth converted section of the Kallsjön-Fröå (KF) profile.

The seismic sections revealed a very reflective upper crust in much finer detail than previously available, especially in the near surface. Subhorizontal reflections are easily traced from the main BL profile, through the KF profile and into the older regional CCT sections (Figure 4.3a). Three regions of distinctly different reflectivity patterns are observed (Figure 4.1b). A highly reflective unit dominates the upper few km in the westernmost part of the BL profile as well as the entire KF profile. As seen in Figure 4.3b, this correlates nicely with the surface geological mapping of the SNC. Scattering of seismic energy within this unit (comprising banded gneisses and amphibolites) severely decreases the strength of the signal from underlying reflectors. To the east of this, in the BL profile, a nearly transparent near surface overlies a highly reflective package between about 250 m and 1.5 km depth (Figure 4.1b). These patterns are attributed to tightly folded Silurian and Ordovician turbidites on top of thrust sheets comprising Cambrian quartzites and shales and Mesoproterozoic acid volcanites, all a part of the Lower Allochthon (Figure 4.1c). None of these reflections project to the surface along the profile and their interpretation is based on basement windows to the south of Undersåker and in the Mullfjället antiform to the west.

Figure 4.3. a) Correlation of the reflection seismic sections. Reflections can be followed from the older CCT on the right, through the newly acquired connection profile, KF, and into the main COSC profile, BL, on the left. b) Correlation of the western part of the BL profile with the surface geology. MA – Middle Allochthon, LA – Lower Allochthon.

Below these two regions, albeit less clearly imaged below the highly scattering SNC, continuous reflections span the entire length of the profile (Fig- ure 4.1b). The uppermost of these, dipping westwards from about 2.5 km depth in the east to nearly 4 km in the west, was interpreted as the basal décollement (Figure 4.1c). Underlying reflections were attributed to reflections within the crystalline basement of possibly Rätan type granite. This is found further southeast and indicated by magnetic and gravity modeling to extend far in beneath the Caledonian cover of Jämtland. The deep basement reflectors might either be mafic intrusions similar to those found to the east (1.2 Ga; Söderlund et al., 2006) and south (1.0 Ga; Juhlin, 1990) of the orogenic front, or be shear zones related to deformation during the Caledoni- an or Sweconorwegian orogenies.

49 4.1.2 Conclusions The high resolution images from the new reflection seismic sections allow an interpretation of the uppermost few kilometers of the central Caledonides in western Jämtland (Figure 4.1c). The Lower Seve Nappe (which was the target of the first borehole, COSC-1) stands out as a highly reflective unit with a synformal base that was interpreted to extend down to less than 2.5 km below the surface at its maximum depth. The COSC-1 borehole was placed nearby the abandoned Fröå mine where the seismic data suggested that the base of the SNC would be penetrated at a depth of about 2.3 km. This location was also suitable for a number of other practical considerations with regards to the drilling operations. The interpretation of the basal orogenic detachment is less certain. It is primarily based on correlation with the previous interpretation along the CCT profile where magnetotelluric data link the conductive alum shales found at the frontal thrust with subhorizontal reflections in the lower resolution CCT reflection seismic section. The reflections attributed to the décollement only come within drillable depth at the very eastern edge of the interpreted seismic section (Figure 4.1c). Thus, the second objective of COSC-2 of penetrating basement reflectors would not be possible in the area. No location was therefore proposed for the second phase of the COSC project. Instead, an extension of seismic profiling towards the east was proposed before a suitable location for COSC-2 could be selected.

4.2 Paper II: 3D interpretation by integrating seismic and potential field data in the vicinity of the proposed COSC-1 drill site, central Swedish Caledonides After a location had been proposed for the first 2.5 km deep scientific borehole of the COSC project, COSC-1, a better understanding of the 3D geometry of the large scale structures within the Caledonian cover nappes was desired, with a special focus on the Seve Nappe Complex (SNC). In this paper, surface geology, petrophysical data, reflection seismic sections and potential field data are integrated to extend the interpretation of the allochthonous units and basal décollement into three dimensions and to test its validity.

4.2.1 Description of data The data used in this paper include the reflection seismic sections from the recent acquisition (Paper I) and the older CCT (Palm et al., 1991; Juhojuntti et al., 2001), petrophysical data (Elming, 1980), Bouguer gravity data (c. 300 point measurements over an area of 17000 km2), total field aeromag-

50 netic data (acquired with a point and line spacing of 7 m and 200 m, respectively) and a bedrock geological map. The magnetic data were acquired by the Geological Survey of Sweden (SGU) in 2011 in a survey dedicated to the area encompassing the new seismic profiles (Figure 4.4a). Bouguer gravity data (reduced assuming a standard density of 2670 kg/m3, provided by SGU) were used to exclude the effect of topography (Figure 4.4c).

Figure 4.4. a) Total field aeromagnetic data. b) The data in (a) after removal of DGRF. c) Bouguer gravity anomaly data. d) Approximated 2nd order polynomial regional gravity field. e) Residual gravity anomaly map after removal of (d) from (c). Data are courtesy of the Geological Survey of Sweden.

Regional field separation was performed to remove the influence of deeper structures and regional trends, and attain datasets more representative of

51 upper crustal structures. A regional gravity field, approximated from the gravity data as a 2nd order polynomial, was removed to produce the residual Bouguer gravity anomaly map shown in Figure 4.4e. The residual magnetic anomaly map in Figure 4.4b was produced by subtracting the Definitive International Geomagnetic Reference Field (DGRF), which has an average value of 50550 nT in this area. The study area lies within the vast Jämtland anomaly (positive magnetic and negative gravity) that is thought to be linked with Rätan and/or Dala granites in the underlying basement (Figure 1.2). Superimposed on this are north-south trending anomalies that can be linked with the regional antiforms and synforms. The high density SNC in the Åre and Tännfors synforms (underlying the Köli nappes in the latter) are associated with positive anomalies while the Mullfjället antiformal window in between is associated with a gravity low. The Jämtland magnetic anomaly is too small to be removed with the DGRF and is still dominating the residual magnetic anomaly map. The Caledonian cover, including the allochthonous basement units at e.g. Mullfjället, is generally weakly magnetized (Dyrelius, 1980) and many of the observed features (such as the decrease at the eastern limb of the Åre synform in Figure 4.4b) are likely related to undulations in the basement.

4.2.2 3D interpretation To obtain a 3D interpretation of the major nappes in the area around the COSC-1 drill site, migrated and depth converted reflection seismic sections were imported into 3D modeling software. A bedrock geology map was incorporated and projected to a digital elevation model. Boundaries separating major tectonic units were traced in the surface geology (Figure 4.5a). After this, horizons were picked on the seismic sections along major reflections that, according to the 2D interpretation, separate these tectonic units (Fig- ure 4.5b). Finally, surfaces roughly tracing the boundaries between adjacent units were created by linking picked horizons with the boundaries in the surface geology and extending laterally. The maps of geology and potential field anomalies were used as indicators for the extension and varying thickness of the units away from the seismic lines (Figure 4.5c). The 3D interpretation was kept as simple as possible, despite the complex reflectivity observed in several places. Figure 4.5d shows the internal configuration of these lithological boundaries at the COSC-1 location.

Figure 4.5. a) Bedrock geological map projected to a digital elevation model (based on the bedrock geological map of Sweden, © Geological Survey of Sweden [I2014/00601] and Strömberg et al., 1984). b) Reflection seismic sections with picked horizons. c) Resulting model of tectonostratigraphic boundaries. d) Internal configuration of tectonic units at the location of COSC-1. UA – Upper Allochthon, MA – Middle Allochthon, LA – Lower Allochthon, A – Autochthon (basement).

4.2.3 Forward and inverse modeling To test the consistency of this interpretation, gravity and aeromagnetic data were first used in 2.5D forward modeling and later in 3D inverse modeling. Two key points for validation of the modeling were defined; 1) the 3D geometry of the central SNC with high density and magnetic susceptibilities and 2) whether the SNC is separated from an assumed basement of Rätan type granites (high density and high magnetic susceptibility) by units of the Lower Allochthon (low density and magnetic susceptibility). The 2.5D forward modeling was performed along the BL profile on both gravity and aeromagnetic data. A model was created by extending the 2D interpretation laterally in the direction perpendicular to the seismic profile and assigning petrophysical properties from surface measurements in the area (Elming, 1980). The calculated gravity and magnetic response from the model is in good agreement with the observed data along some parts of the section. Other parts, however, show a large discrepancy between the observed and modeled data (Figure 4.6). One important contributing factor to this misfit is the true 3D geometry of structures which was not taken into consideration in this 2.5D modeling.

Figure 4.6. a) Interpretation of the 2D reflection seismic sections from Paper I with average petrophysical data from Elming (1980). b) Observed and calculated magnetic and gravity data along the reflection seismic profile. c) Cross section from the 3D inverse model parallel with the reflection seismic profile.

3D inverse modeling was performed on the residual Bouguer gravity data covering the area around the drill site. An initial cubical model was constructed containing rough volumetric shapes based on the 3D geological interpretation (Figure 4.7) and with smoothness constraints imposed at the interfaces. The model was populated with rectangular cells which were as- signed density contrasts from a background of 2700 kg/m3 (the average density used for the basement). 3D inversion of potential field data is highly ambiguous and can fully be explained by an infinite number of models. To avoid the natural tendency to concentrate density variations at the surface, a depth weighting function was used to give higher contribution to deeper cells (Li and Oldenburg, 1998).

Figure 4.7. The inversion algorithm.

Iterative inversion was performed with small perturbations to the previous model to prevent unreasonable deviation from the geometrical constraints from reflection seismics and surface geology (Figure 4.7). Inversion continued until the calculated response from the 3D inversion model reproduced the observed residual gravity field within an error misfit of 0.2 mGal. Only minor discrepancies are observed, the majority of which are partly introduced by the inversion and partly due to the forced geometrical constraints of the initial model. The magnetic data were not considered for 3D inversion because of its limited surface coverage with respect to both the location of the COSC-1 drill site and the Åre synform hosting the SNC. The central high density unit in the 3D inversion model stands out clearly and corresponds to the SNC. It appears to be at its thickest just west of the drill site and becomes shallower both towards the north and south (Fig- ure 4.8). Importing the shape of the SNC from the 3D geological interpretation for comparison shows that the 3D geometry of the high density unit in the inversion model converges on the same shape (Figure 4.9). Furthermore, it is clear that low density material is required underneath the high density SNC in order to generate an accurate response from the inversion model (Figures 4.6c and 4.8). The inversion does, however, fail to resolve the

55 thickness of this layer, and hence the depth to basement. To the east of the Åre synform, no a priori geometrical constraints were imposed and density contrasts are much smaller. Therefore the inversion does not resolve any structure in this area (Figures 4.6c, 4.8 and 4.9).

Figure 4.8. Depth slices through the 3D inverse model at 200, 700, 1700 and 2700 m depth.

Figure 4.9. a) 3D view of the inverse model with vertical slices intersecting near the COSC-1 drill site. b) The same as (a) but with the 3D geometric shape of the SNC from the 3D interpretation in Figure 4.5.

56 4.2.4 Conclusions Integration of various geophysical techniques with geological and petrophysical information is essential to produce an accurate interpretation of subsurface structures. Here 2D reflection seismic data, with excellent depth resolution along a vertical cross section, was combined with potential field data, which lacks the depth resolution but adds the 3rd dimension to the model. The presented 3D lithological interpretation, based on surface geology and 2D reflection seismic data, provided a better understanding of the local geometry of major tectonic units around the proposed COSC-1 drill site. Constrained 3D inverse gravity modeling over the area supports the interpretation as it converges on the same 3D geometrical shape. It also shows that material of lower density, such as the lower units of the Middle or Lower Allochthon, separates the SNC from an assumed high density basement. No conclusions about the deeper parts of the interpretation, or to the sides of the central Åre synform, could however be drawn from the inversion. The interpretation and inverse modeling further indicated that a borehole at the proposed location for COSC-1, will potentially give the longest possible cored section through the SNC.

4.3 Paper III: 3D reflection seismic imaging at the 2.5 km deep COSC-1 scientific borehole, central Scandinavian Caledonides Drilling of the 2.5 km deep scientific COSC-1 borehole began in early may 2014 and was successfully completed by late August the same year. Nearly complete core recovery supported by downhole logging campaigns have resulted in an impressive continuous dataset of a wide range of parameters through the targeted lower Seve Nappe and into the underlying mylonites (Lorenz et al., 2015a). Extrapolation of the results from analysis of the core and borehole into the rock surrounding the borehole requires high resolution imaging and control of the geometry of structures. Therefore, a major seismic survey was conducted in and around the borehole in September 2014, comprising three separate experiments that simultaneously recorded the signal from the same source points. The two first ones were a high resolution zero-offset Vertical Seismic Profile (VSP) (Krauß et al., 2015) and a multi-azimuthal walk-away VSP with long offset surface recordings (Simon et al., 2015). These used a combination of a receiver chain containing 15 three component geophones at 10 m separation and 180 three component geophones placed on the surface along three 2D profiles centered on the COSC-1 drill site. The third experiment was the first ever,

57 albeit limited, 3D reflection seismic acquisition on land in Scandinavia to target the Caledonian nappe structures. Acquisition, processing and interpretation of the latter are the focus of Paper III.

4.3.1 Summary The 3D reflection seismic acquisition used a stationary spread that was kept active for the entire survey. It involved 429 single component 10 Hz geophones that were deployed at 20 m separation along seven lines spaced 200 m apart. This square of receivers covers an area of about 1.5 km2 and was centered on the COSC-1 drill site (Figure 4.10a). Source locations were constrained by rough terrain to follow available roads in the area, forming a star pattern of crooked lines radiating out from the drill site (Figure 4.10a). A mechanical source (the same VIBSIST system as in the 2010 acquisition, Figure 3.1) was used at near offsets from the borehole with a source point separation of 20 m. At larger offsets, explosive charges of 0.5 kg were placed in holes of 3.5 to 5 m depth and fired every 80 m. As the receiver chain of the simultaneously running VSP experiment was moved between seven different levels to cover the full length of the borehole, each source point was activated on several occasions. Every VIBSIST source point was excited up to five times with each normally gen- erating three sweeps. 73 out of 128 explosive shot points (57 %) could be reused for a second activation. The resulting spread of midpoints was binned into rectangular bins of 20 m by 60 m, with the short side parallel with the east-west directed receiver lines. The limitation in the acquisition geometry results in large unsampled regions, highly varying fold (Figure 4.10b) and an irregular offset-azimuth distribution (Figure 4.10c), where far offset data are contained within narrow azimuth bands. This acquisition footprint may affect the lateral resolution and the capability to resolve dipping reflectors. Several other factors further complicate the processing of this dataset and an optimum solution for the calculation and application of residual statics and DMO in particular have not yet been found. These complicating factors include the large variation in elevation between both receivers (<100 m) and source points (<250 m), the varying surface conditions and the complex geometry of structures within the SNC. In addition to this, results from the VSP experiments (Simon and Krauß, per. Comm., 2015) as well as lab measurements on rocks in the area (Czaplinska et al., 2015) and on core from the COSC-1 borehole (Wenning et al., 2015) suggests that seismic anisotropy is an important factor in this unit, ranging from 5 % in felsic rock to 12 % in amphibolites. Furthermore, the core (Figure 3.10) and outcrops (Figure 2.3) suggest that mafic boudins within felsic gneisses/calc-silicates may be the origin of the reflectivity and that these may be too small to create laterally continuous reflections.

Figure 4.10. a) Bedrock geology map around the COSC-1 drill site showing the acquisition geometry of the 3D reflection seismic experiment around the COSC-1 borehole (based on the bedrock geological map of Sweden, © Geological Survey of Sweden [I2014/00601] and Strömberg et al., 1984). b) The processing geometry with the inlines in the west-east direction. c) The offset-azimuth distribution is irregular and shows the limited representation of far offsets to narrow azimuth bands.

Processing followed a standard sequence with static corrections, deconvolution, spectral equalization, spherical divergence compensation, filtering, Normal Moveout (NMO) corrections and CMP stacking before migration and time-to-depth conversion. A 2.5D post-stack time migration using a finite difference algorithm was performed by essentially migrating first in the inline direction and then in the crossline direction. Considering the complex 3D geometry and large variation in velocities, a full 3D pre- stack migration may have given a better result (assuming that an accurate velocity model could be obtained).

59 In Figure 4.11, inline 1045 from the migrated and depth converted seismic volume is compared with a simplified geological section from on-site mapping of the COSC-1 core, as well as the first preliminary results from logging of P-wave velocity (vp) and density from both core and borehole (introduced in section 3.3). An acoustic impedance log down to 1605 m was calculated from the downhole logging datasets. The sampling rate of these logs was 0.01 m, while the resolution of the reflection seismic data is on the order of a few tens of meters. A running average with a window size of 5 m was therefore applied to the presented logs.

Figure 4.11. a) On the left is a simplified lithological section based on the initial on- site mapping of the retrieved core. For the purpose of visualization, lithologies have been roughly partitioned by the dominant component as 1) felsic rocks (gneisses, calc-silicates, migmatites, marble and metasandstone), 2) mafic rocks (amphibolite and metagabbro), 3) gneisses with a large amount of mica schists and mylonite bands, and 4) mylonites. vp and sidewall density are available from borehole logging between 105 and 1605 m depth while rock density after core retrieval is available for the entire core (below 103 m). The impedance log is calculated from the borehole data. The geophysical logs are shown after application of a 5 m running average and are scaled vertically to match the true location along the borehole trajectory shown in (b). b) Inline 1045 from the migrated and depth converted seismic volume is shown with major reflections and the interpreted Mylonite Zone. The COSC-1 borehole trajectory deviates a maximum 100 m eastwards. Depth is referenced to the surface elevation at COSC-1 (523 m above sea level).

60 The impedance log shows an abundance of sharp contrasts of about 2×106 kg/m2s in the Lower Seve Nappe, down to the maximum logging depth (Figure 4.11). These can be correlated with the transition between felsic and mafic rocks (section 3.3) and are likely capable of producing strong reflectivity under favorable conditions. A complex pattern of reflections with limited lateral continuity is observed within the Lower Seve Nappe (Figure 4.11). Although not clearly imaged, several reflective pack- ages can be followed through the volume and correlated with the lithological section and logs (reflections S1-S8 in Figure 4.11). Density variation in the deeper parts of the core, together with the velocity profile obtained from the zero offset VSP experiment, suggests that significant impedance contrasts are equally strong but less abundant in the Mylo- nite Zone down to about 2.3 km. In the seismic data, the Mylonite Zone is characterized by fewer reflections (than in the overlying Lower Seve Nappe), but those present are more laterally continuous over larger distances and all dip more or less uniformly towards the southeast (Figure 4.11). The reflections in the Mylonite Zone link nicely with a similarly reflective package in the previous 2D reflection seismic data which appears to turn into a westward dipping zone and project up to the surface at the mapped transition between the SNC and the Lower Allochthon (Figure 4.12). Reflectivity below the Mylonite Zone suggests that this zone may continue to a maximum depth of about 2700 m below the COSC-1 drill site. At this depth and just below, strong and continuous reflections have a different character with a more eastwards dip direction. These reflections are termi- nated by a west dipping reflection that could indicate the pinching out of lithological units belonging to the lowermost Middle Allochthon (found as thick units to the west of the Åre synform, but only as thin slivers on the eastern side) between the SNC and Lower Allochthon. Deeper reflections are similarly strong and continuous down to the nearly 9 km of decoded data and have dips that only slightly deviate from the east-west direction.

4.3.2 Conclusions The Lower Seve Nappe exhibits a complex pattern of strong and frequent reflections of conflicting dips. Static and moveout corrections are not opti- mized at present to image these reflections in detail and finding a proper solution may improve the coherency of reflections within the geometrically complex formation. Several strong reflective zones can, however, still be correlated with lithological boundaries and logs of vp, density and impedance.

Figure 4.12. A comparison of inline 1045 from the new 3D data with the 2D seismic section of the BL profile (Paper I) where they intersect around crossline 1093 (vertical dashed line), viewed from SSW. Reflections within the SNC, although weaker and less continuous in the 3D seismic volume due to non-optimal processing, show good correlation across the two datasets and the 3D seismic data allow constraining the geometry of reflectors in 3D. The good match between the datasets is clearly seen on the larger scale through the reflections at depths between 2.5 and 4 km depth. At greater depths, down to 9 km, reflections nearly indistinguishable in the 2D section can be traced laterally through most of the seismic cube. Depth is referenced to the surface elevation at COSC-1 (523 m above sea level).

The unexpectedly thick Mylonite Zone, constituting the lowermost 800 m of the COSC-1 borehole, is imaged as a package of continuous reflections, uniformly dipping towards the southeast at the location of the borehole (between MZ and U1 in Figure 4.13). The base of the Mylonite Zone was not penetrated by drilling and the comparably weak reflectivity at the bottom of the hole and below (previously interpreted to be below the basal thrust zone of the SNC) may be related to small impedance contrast between mylonites and metasandstones. A complete transition into the underlying units is thought to be first at a couple hundred meters below the bottom of the COSC-1 borehole, where continuous reflections have a significantly different character than the Mylonite Zone.

Figure 4.13. Inline 1052 and crossline 1104 are shown in 3D with picked horizons. The top of the Mylonite Zone (M.Z.) was picked along reflections intersecting the borehole at about 1700 m in accordance with the present interpretation of the core although reflections within 200 m directly above have similar apparent geometry. A distinct difference can be observed in the reflectivity pattern above M.Z., with limited lateral extension and conflicting dips, compared to within and below M.Z., with better defined geometry and lateral continuity. Whereas the reflections within the interpreted Mylonite Zone dip towards the Southeast, reflection U1 and U2 only dip towards east and underlying interfaces (U3 and below) are subhorizontal to slightly west dipping. Depth is referenced to the surface elevation at COSC-1 (523 m above sea level).

There are currently two alternative interpretations of the underlying reflectors. The first, shown in Figure 4.13, follows the interpretation in Paper I but modified to include the possibility of the lower units of the Middle Allochthon, Särv and Offerdal, to underlie the Mylonite Zone in the west (between U1 and U3 in Figure 4.13) and pinch out at a depth of 3 km below COSC-1. The underlying Lower Allochthon may comprise Cambrian quartzites and shales from the Baltica platform and allochthonous acid volcanic units derived from the basement. Reflections interpreted as the décollement in Paper I are found at a depth of 4.3 km (U8). Deeper reflections occur within the autochthonous basement and may be related to orogenic deformation of dolerite sills.

63 In light of extended reflection seismic profiling (Paper IV) and new constraints from magnetotelluric (MT) data (Yan et al., 2015) and aeromagnetic data, an alternative interpretation with a much shallower basal detachment has been proposed (see Paper IV for more details). The reflections in the 2D seismic sections that are attributed to the main detachment in this alternative interpretation can be followed into the 3D reflection seismic volume in Fig- ure 4.13 and be correlated with the well-defined and continuous reflection U4 dipping westwards from a depth of about 3.3 km at the eastern edge of the seismic volume. In this case, the underlying reflectors would all be in the autochthonous basement.

4.4 Paper IV: Seismic imaging of the eastern Scandinavian Caledonides: Siting the 2.5 km deep COSC-2 borehole The interpretation of the 2D reflection seismic data acquired in 2010 to identify the most suitable sites for the two COSC boreholes, did not reveal an obvious location for the second borehole, COSC-2 (Paper I). The objectives with this borehole are several; to drill through the Lower Allochthon, penetrate the basal décollement and to intersect at least one of the major enigmatic basement reflectors. The west dipping décollement was interpreted to approach the maximum drilling depth of 2.5 km at the eastern edge of the seismic section. Thus reaching the basement reflectors would be impossible and it was suggested that the seismic profile should be extended towards the east for the purpose of finding an optimum site for COSC-2. The nearly 36 km long Byxtjärn-Liten profile (BL) profile from the 2010 acquisition, was extended in 2011 by the 17 km long Liten-Dammån profile (LD), with a 1 km overlap with BL. Later, in 2014, this was further extended by the 14 km long Dammån-Hallen profile (DH), also with 1 km overlap. The LD profile of 2011 suffered a 4.5 km gap in acquisition from missing road permissions and therefore a fourth profile, the Sällsjö profile (S), was acquired to bridge this gap, taking a 16 km long detour to the south before again joining with the LD profile further east. The resulting composite main profile, the COSC Seismic Profile (CSP), spans a distance of nearly 55 km from the Lower Seve Nappe in the northwest, into units of the Lower Allochthon in the area south of Hallen in the southeast. All the 2D reflection seismic acquisition lines are shown in Figure 2.2. The fourth paper of this thesis deals with the acquisition and processing of the new reflection seismic data. Two interpretations of the composite CSP are presented, linking the seismic sections with previous drilling for prospection of the alum shales near the eastern edge of the seismic profile, and using constraints from aeromagnetic data (acquired by the Geological Survey of

64 Sweden in 2011) and magnetotelluric surveying (Yan et al., 2015). Based on these interpretations, two potential locations for the COSC-2 borehole are proposed.

4.4.1 Summary The acquisition of the extensions to the main profile in 2011 and 2014 are, like the BL profile of 2010, 2D crooked line profiles (Figure 2.2). However, they differ slightly due to terrain, road permissions and updates to recording and source equipment. The LD profile (2011) used the same mechanical source, the VIBSIST (Figure 3.1), and 28 Hz geophones as along the BL profile (Paper I). The gap in acquisition was partially bridged by deploying wireless units on the western side of the gap (coinciding with the last 1 km of the BL profile) and wired units on the eastern side to undershoot as much as possible. Complete undershooting was not possible and the uppermost 1 km was not imaged at all because of the lack of near offsets. For the 2014 acquisition of the DH extension and the S detour, a new source (a 400 kg weight drop source mounted on a Bobcat excavator) and new receivers (10 Hz geophones) were introduced. The VIBSIST data were, as before, decoded following Cosma and Enescu (2001) and Park (1996), and the individual activations with the weight drop were stacked to generate seismograms with high signal to noise ratio. The individual profiles were processed separately, in general following the standard processing sequence in Paper I, and then merged and stacked to a single CDP line. The structures in the area of the LD and S profiles are subhorizontal or dip in the general NW direction of profiling. Stacking the midpoint traces from the S profile onto the same CDP line as the LD profile, despite large perpendicular offsets, produced a seismic section with coherent reflections that represent the general structure in the area and eliminates the gap from the LD acquisition. Migration was done using the Stolt algorithm and time-to-depth conversion was done to allow geological interpretation. The migrated and depth converted section is shown in Figure 4.14, together with magnetic and MT data along the profile. Both the VIBSIST and the weight drop are shown to generate sufficient seismic energy to generally image the upper 9 km of crust. The southeastern- most part of the CSP is, however, not well imaged (Figure 4.14) due to the up to 60 m thick unconsolidated sediments at the surface. However, a subhorizontal reflection at about 500 m depth can be traced to the southeastern edge of the section. Projected eastwards, this reflection links nicely with the basal detachment in a geological section (Figure 4.15) based on the results from the drilling program in the late 70’s targeting the alum shale (Gee et al., 1978), only a few kilometers southeast of the CSP.

Figure 4.14. a) Total magnetic field anomaly along the CSP. The anomalies at about CDP 1800, 3100 and 4100 can be interpreted as due to variations in the magnetic basement at depths of 1.3 km, 1.3 km and 1.0 km, respectively. b) The composite CSP after migration and depth conversion, shown at a vertical exaggeration of 2:1. The black line marks the depth to the highly conductive feature in the MT data as mapped by Yan et al. (2015). This conductor and the base of the uppermost seismically transparent zone are in excellent agreement. Therefore, the onset of reflectivity below the transparent zone is interpreted to represent the uppermost alum shale. Magnetic data are courtesy of the Geolog- ical Survey of Sweden (SGU).

Figure 4.15. Geological cross section through the Myrviken area boreholes based on the SGU report on alum shales by Gee et al. (1982) with vertical exaggeration 10:1.

A magnetotelluric (MT) survey was recently acquired along the entire CSP (Yan et al., 2015). This shows a very strong conductive signature attributed to the alum shales that can be correlated with continuous subhorizontal reflections at a depth of around 1 km throughout the entire 2D seismic profile to the east of the SNC (Figure 4.14). All along the orogenic front, and in basement windows, Cambrian alum shales are found to rest unconformably on top of the Precambrian basement and may have acted as a lubricant to help facilitate the shallow angle thrusting along the basal detachment. The same conductor is found along the CCT and can be interpreted as related to the basal décollement. Deeper features below a strong conductor are, however, difficult to resolve with MT and a second, deeper seat- ed, conductive layer of alum shales cannot be ruled out. In this case the shallow conductor may represent imbricated layers of alum shales within the Lower Allochthon. Aeromagnetic data along the BL profile (Figure 4.14) show north-south trending anomalies (see Paper II) that may be interpreted using Peter’s Half- Slope method. This constrains the depth to an assumed highly magnetic Pre- cambrian basement of Rätan type granitic rocks at CDP’s 1800, 3100 and 4100 (Figure 4.14) to depths of around 1.3 km, 1.3 km and 1.0 km, respectively, but may also represent allochthonous basement-derived units. Thus, magnetic and MT data, albeit ambiguous, both indicate a basal detachment at about 1 km depth. Presented in Figure 4.16 are two alternative interpretations of the structures beneath the Caledonian cover in Jämtland. The first (Figure 4.16a) extrapolates the interpretation of Paper I (based primarily on the geology in windows in nearby areas) into the extended seismic section, with a deep décollement that may be penetrated first to the east of CDP 3600. The second interpretation (Figure 4.16b) considers a shallow basal detachment more in line with constraints from all other geophysical data. Intermediate interpretations with more or less basement involvement are also possible and the only way to resolve this ambiguity is by deep drilling.

Figure 4.16. Two possible interpretations along the CSP. In (a) the interpretation west of CDP 2800 is the same as in Hedin et al. (2012) with a deep basal detachment and significant basement involved thrusting. In (b) the basal detachment is much shallower in the west and follows a continuous reflector that lies only a few hundred meters below the top of the alum shales as interpreted from the CSP and the MT data in Fig- ure 4.14.

4.4.2 Conclusions Two potential locations for the COSC-2 are identified based on the presented interpretations. At the first location (around CDP 4100, Figure 4.17a) both presented interpretations are in agreement. A borehole here would reach the shallowest alum shales at a depth of about 400 m (as indicated by the MT data) and the main detachment would be penetrated at a depth of 800 m. Reflections at 1.7 km and 2.2-2.3 km depth could represent basement features where the latter is seen to extend down to depths of at least 7 km. Al- ternatively, the reflection at 1.7 km could be the main detachment (in case the interpretation of a deeper detachment of Hedin et al. (2012) is correct) with quartzites and shales being stacked upon one another above this reflector east of CDP 3400. Assuming the interpretation in Figure 4.16b is correct, a borehole south of the lake Liten (around CDP 2200, Figure 4.17b) would give the longest possible section through the Lower Allochthon, from the youngest stratigraphic successions of Silurian turbidites, through underlying Ordovician turbidites, and a complex of Cambrian shales and quartzites between 800 m and 1.2 km where the main detachment would be encountered. Below this, four basement reflections would be penetrated between depths of 1.4 to 2.2 km. The two deeper ones can be traced to a depth of at least 7 km. The main disad- vantages of this option are the possibility that the deeper interpreted main detachment is correct as well as logistical considerations that may restrict the drill site to be moved some distance away from the scientifically optimum location.

Figure 4.17. The two proposed locations for the COSC-2 borehole with the conductivity profiles shown at the borehole location. (a) At alternative A, the two interpretations in Figure 4.16 agree. Here, the borehole would penetrate the uppermost alum shale at about 400 m depth and the basal detachment at about 800 m depth as interpreted in Figure 4.16a. Potentially, the strong reflection at about 1.7 km depth repre- sents the detachment. If so, the rock between 800 m and 1.7 km could consist of duplex structures. (b) At alternative B, the two interpretations in Figure 4.16 differ significantly. Here, the uppermost alum shale would be penetrated at about 900 m depth and, if the interpretation in Figure 4.16b is correct, the basal detachment would be drilled at about 1.3 km depth. Logistically, it is easier to place the borehole about 1 km to the east. Even at this location, two or three Precambrian reflectors would be penetrated if the interpretation in Figure 4.16b is correct.

70 5 Conclusions

This thesis presents the outcome of primarily reflection seismic investigations in relation to the major COSC scientific drilling project in the central Scandinavian Caledonides. In addition to standard and non-conventional reflection seismic processing techniques, forward and inverse modeling of gravity data and constraints from aeromagnetic and magnetotelluric data have been used to allow an integrated interpretation of the major tectonic units of the Caledonian orogen. The thesis also incorporates results from on- core measurements and downhole logging from the COSC-1 borehole which was drilled during the spring and summer of 2014. Paper I and IV focus on the acquisition, processing and interpretation of several 2D reflection seismic profiles which collectively make up the composite c. 55 km long COSC Seismic Profile (CSP). The aim was to identify the most suitable locations for the two COSC boreholes. Standard processing of these 2D crooked line profiles revealed impressively reflective structures within the Caledonian allochthons and underlying basement in the uppermost 9 km of crust. While the COSC-1 drill hole was successfully placed on the basis of the interpretation of the main profile acquired in 2010 (Paper I), extensions of this profile in 2011 and 2014, and additional constraints from magnetotelluric data (acquired in 2013 (Yan et al., 2015)) and aeromagnetic data (acquired in 2011 by the Geological Survey of Sweden), were used to reveal two candidate sites for the upcoming COSC-2 borehole (Paper IV). The SNC was studied in detail prior to the drilling of COSC-1 to gain more information on its tectonic setting and the 3D geometry of the internal structures, and to allow better predictions for the drilling and sampling programs. In Paper II the available 2D reflection seismic profiles in the area were combined with surface geological information and potential field data (primarily Bouguer gravity data) to create a constrained 3D lithological model of the Åre synform that hosts the SNC. In sections 3.2.1 and 3.2.2 of this thesis, previously unpublished results from non-conventional pseudo 3D processing of the 2D crooked line data in the vicinity of the COSC-1 borehole are presented. Crossdip analysis confirmed the complex geometry of structures within the SNC as well as the 3D characteristics of underlying reflections. Swath 3D imaging allowed structural predictions to be made for the planned trajectory of the COSC-1 borehole. The COSC-1 borehole successfully reached a final driller’s depth of 2495.8 m. Nearly complete recovery of core and comprehensive logging of

71 both core and borehole resulted in a continuous geological section and datasets of a wide range of parameters through the Lower Seve Nappe and into the underlying thrust zone. A major post-drilling seismic survey was conducted in and around the COSC-1 borehole with multiple experiments running simultaneously. One of the experiments involved a 1.5 km2 array of surface receivers to obtain a 3D reflection seismic dataset which is the topic of Paper III. The Lower Seve Nappe is shown from drilling, logging and seismic data, to be geometrically and lithologically very complex. Amphibolites and metagabbro occur as thin layers, lenses or boudins (on the order of cm to tens of meters thick) within mainly felsic and calc-silicate gneisses. Despite diffi- culties with imaging continuous structures within the Lower Seve Nappe using standard processing techniques, reflective zones can be correlated with initial geological mapping of the core and logs of density and P-wave velocity. A surprisingly thick mylonite zone was encountered in the COSC-1 borehole, constituting the lowermost 800 m of the core and was not fully penetrated. The first mylonites of this zone are observed at a depth of about 1700 m and which then increase in frequency and thickness down to about 2100 m depth. Below this depth they dominate large portions of the core down to total depth. This zone is imaged as a package of continuous reflections, uniformly dipping towards the south-east. The deepest mafic rock was encountered just below 2314 m depth and a transition from gneisses to lower-grade metasediments occur at a depth of around 2350 m. Although the tectonostratigraphic affinity of the deepest interval hasn’t been determined, this zone correlates with a significant decrease in reflectivity and may indicate a transition from the Lower Seve Nappe into underlying units. To the east of the SNC, metasediments of the Lower Allochthon appear seismically transparent from the surface down to an undulating reflective boundary at depths varying from a few hundred meters just east of the Åre synform, to about 1.2 km between the Åre Synform and Olden-Oviksfjäll antiform. None of the underlying reflections project to the surface and in Paper I, the structures beneath the SNC and the sedimentary formations to the east, were interpreted based on windows through the Caledonian cover to the south and west of the study area. In these locations, Cambrian quartzites and shales from the Baltoscandian platform overlie allochthonous basement units of acid volcanites. Recently acquired MT data constrains the depth to the highly conductive Cambrian black alum shales (associated with the basal detachment) to about 1 km. Aeromagnetic data further locally constrain the depth to highly magnetic formations, interpreted as Rätan type granites found in the basement, to depths of between 1 and 1.3 km. In light of these observations (although slightly ambiguous), Paper IV, provides two alternative interpretations of the deeper structures, where one extends the previous interpretation and the other relies on the geophysical evidence that indicate a much shallower basal

72 detachment. In either case, underlying the basal detachment is believed to be an extension of the TIB granites in the basement, and the observed reflectivity has been inferred in this study and previous ones, to originate from either shear zones from Caledonian or older orogenic deformation, or 1.0 to 1.2 Ga dolerite intrusions. Finally, the high resolution reflection seismic profiling across the Caledo- nian cover and underlying basement is once more a witness of the remarkable seismic response in this area. The resulting seismic images allow extrapolation of results from analysis of core and borehole data into the surrounding rock. The COSC 2D reflection seismic profile, CSP, will link the two boreholes and, through integration with other geophysical data such as MT, gravity and magnetic, it will be possible to expand the local understanding gained from the drilling into a regional scale.

73 6 Outlook

A new era of investigating and understanding deep orogenic processes has begun with the COSC scientific deep drilling project. Studies of the targeted units, keys to the tectonic evolution of the central Caledonides, will test current hypotheses and may contribute vital pieces for a deeper understanding of mountain belt dynamics and the formation of major orogens worldwide. The drilling of COSC-1 was successful beyond all expectations, reaching the targeted depth with nearly complete core recovery despite several technical issues. From a technical point of view it was a great achievement since it was, to some extent, the first of its kind. The impressive collection of information and high quality data obtained through continuous coring and logging are available to the scientific community for analysis. At the same time, the COSC-1 borehole was left open and uncased and available for additional logging campaigns and borehole-related experiments in the future. Several projects have been initiated on core-samples, log data and seismic data, but much is left to do and the science of COSC will keep researchers busy for many years to come. Presently, the planning of COSC-2 has started and an international science workshop will be held in October 2015. In relation to the COSC project, regional mapping of the geology as well as the acquisition of reflection seismic, magnetotelluric, magnetic and gravity data in the area around the boreholes, either has been done, or is planned for the near future. These regional datasets will allow the results from the drilling to be expanded into regional scale and to form a basis for a greater understanding of the Scandinavian Caledonides and underlying basement. The reflection seismic work presented in this thesis is a part of the initial stages of the COSC project; site surveying, placing the boreholes in a regional context, and forming a basis for future interpretation. There are several paths to explore as an extension to the presented work. These include, for example, improvements to the 3D reflection seismic imaging around the COSC-1 borehole using conventional and non-conventional techniques, integration of the 3D surface seismic data with 2D surface seismic and VSP data for imaging and studies of seismic anisotropy and structures in the Lower Seve Nappe, and integration of seismic data with other geophysical data such as MT, potential fields and downhole and on-core logging for modeling and interpretation on both the local and regional scale.

74 7 Summary in Swedish

Den Skandinaviska bergskedjan, Skanderna, dominerar idag geologin och topografin i Norge och västra Sverige. Mer än etthundra år har gått sedan Törnebohm (1888) föreslog att Åreskutan skulle ha förflyttats minst 100 km till sin nuvarande plats genom överskjutning vilket skapade het debatt om bergens uppkomst. Idag vet vi att bergskedjan är resultatet av den kollision som ägde rum mellan kontinenterna Laurentia (dagens Nordamerika och Grönland) och Baltika (dagens Skandinavien, västra Ryssland och Baltikum) för ungefär 445 miljoner år sedan. Den bergsbildande kollisionen, eller oro- genesen, varade i 50 miljoner år och skapade de norra Kaledoniderna som var jämförbara med dagens bergskedja Himalaya-Tibet. Det har visats att transporten av bergsblock (skållor) omfattar överskjutningar om minst 400 km österut, ovanpå dagens Skandinavien, och 200 km västerut, ovanpå dagens Grönland. Under de 400 miljoner år som följt, har post-orogenisk kol- laps, erosion, extension och öppnandet av Nordatlanten sakta men säkert gjort att dagens markyta är på en nivå där den skär igenom det som en gång var kärnan av bergskedjan, i mellersta jordskorpan. Även om förståelsen för bergskedjans formation klarnade väsentligt under de senaste dryga hundra åren, står en rad frågor fortfarande obesvarade. Många av dessa frågor kräver kvantitativa och kvalitativa undersökningar av specifika mål som är svåråtkomliga utan sofistikerade metoder. ”Collisional Orogeny in the Scandinavian Caledonides” är ett tvärvetenskapligt djupborr- ningsprojekt som skapades för att undersöka bergskedjans uppbyggnad och speciellt några av formationerna som tros ha spelat en nyckelroll under den aktiva bergsbildningen. Kärnborrning av två hål, vardera 2,5 km djupa, ska genomföras för att i detalj studera 1) Seveskållan, som har tryckts ner till ett djup av mer än 100 km innan den transporterats som en partiellt uppsmält enhet, 2) den basala skjuvzonen över vilken alla de Kaledoniska skållorna transporterats, och 3) den underliggande kristallina berggrunden som upp- levde partiell subduktion under kontinentkollisionen (Gee et al., 2010). Vi- dare kommer borrprojektet medföra utveckling av teknologi och metoder för borrning och borrhålsmätningar samt kunskap om de rådande geotermiska, geohydrologiska och geokemiska förhållandena i berget som har direkt koppling till industrin och samhället. Även möjligheten att mikroorganismer eventuellt existerar djupt nere i den kristallina berggrunden undersöks. I arbetet som ligger som grund för denna avhandling har reflektionsseis- mik använts dels för att definiera de mest lämpliga borrplatserna för att upp-

75 nå COSCs vetenskapliga mål, och dels för att kunna extrapolera resultaten från borrningen till intill liggande bergsformationer. Artikel I och IV be- handlar de reflektionsseismiska 2D profiler som insamlades under 2010, 2011 och 2014. Tillsammans bildar dessa den nästan 55 km långa ”COSC Seismic Profile”, CSP, som sträcker sig från Fröå strax öster om Åreskutan, till området strax söder om Hallen. Denna profil visar strukturerna ner till 9 km djup i den mycket reflektiva berggrunden. Tillsammans med resultat från magnetiska och elektromagnetiska mätningar har CSP skapat en bild av sammansättningen av de olika Kaledoniska skållorna och den underliggande berggrunden (artiklar I och IV). Tolkningen av dessa 2D profiler har legat till grund för placeringen av det första borrhålet, COSC-1, vid Fröå gruva (artikel I), samt för de två före- slagna platser för det andra hålet, COSC-2, som nu planeras längre österut (artikel IV). Innan borrningen av COSC-1 användes dessa reflektionsseismiska data tillsammans med mätningar av magnetfältet och tyngdkraften för att skapa en 3D modell över de storskaliga strukturerna runt COSC-1 (artikel II). Senare användes mer okonventionella processeringsmetoder för att, från de 2D seismiska data, utvinna information om 3D geometrin på struktu- rer runt det planerade borrhålet (sektion 3.2.1 och 3.2.2 i denna avhandling) och på så sätt hjälpa planeringen av borrning, mätningar och provtagningar. COSC-1 borrades under våren och sommaren 2014 med resultat som överträffade allas förväntningar. Ett totalt djup av 2 495,8 m uppnåddes med minimal kärnförlust, trotts att hålet var det första av sitt slag och att ett flertal problem uppstod under tiden. Utöver den kontinuerliga geologiska sektionen (den upptagna kärnan), insamlades en stor mängd information från en rad mätningar både på kärnan och i borrhålet, samt på gaserna och den cirkule- rande vätskan i borrhålet. Efter avslutad borrning genomfördes en stor seis- misk undersökning med tre experiment där seismiska mätinstrument utplace- rades nere i hålet, längs 2D profiler som korsade borrplatsen, och över en drygt 1,5 km2 stor yta runt hålet. Data från det senare av dessa användes i artikel III för att avbilda formationerna runt COSC-1 i 3D. Trots svårigheter att tydligt avbilda de komplexa strukturerna i den undre Seveskållan med konventionella processeringsmetoder, kan klara samband ses mellan reflek- tionerna, den första geologiska kartläggningen av kärnan, och preliminära resultat från loggningen av kärna och borrhål. Borrningen har visat att den undre Seveskållan består av mestadels gneis- ser med förekommande linser eller boudinage av amfiboliter och metagabbro (med tjocklekar på cm till tiotals meter). En av överraskningarna med borrningen var den mer än 800 m tjocka mylonitzonen som antagligen skiljer undre Seveskållan från underliggande skållor. De första myloniterna påträf- fades på 1 700 m djup och ökade i omfattning ner till 2 100 m. Under detta djup dominerar den litologin ner till botten av COSC-1 borrhålet. Den sista mafiska enheten påträffades vid drygt 2 300 m djup och vid runt 2 350 m djup ses en övergång från i huvudsak gneiss till metasediment av lägre

76 metamorf grad. Dessa djup överensstämmer med en markant gräns i reflek- tionsseismiken. Mellan ca 1 700 m djup och 2 300 m djup kan ett paket av starka kontinuerliga reflektioner följas genom hela den seismiska volymen, med en nästan konstant lutning mot sydost. Dessa kan tydligt följas in i 2D profilerna där de böjer av uppåt och kommer till ytan i närheten av gränsen mellan Seveskållan och underliggande skållor. Även om den tektonostrati- grafiska tillhörigheten av dessa djupaste metasediment inte har klarlagts, och att myloniter fortfarande dominerar kärnan på dessa djup, så kan detta indi- kera övergången från Seve till underliggande enheter. Öster om Seveskållan består bergarterna av metasediment som visar sig vara seismiskt transparenta från ytan ner till en undulerande yta på ett djup av några hundra meter strax öster om Åresynformen, till ungefär 1,2 km mellan Åresynformen och Olden-Oviksfjällantiformen. Ingen av underliggande reflektioner når ytan eftersom de är täckta av Kaledoniska skållor. I artikel I baserades därför tolkningen på observationer som gjordes i fönster genom dessa skållor till söder och väster om området, alltså i områden där skållorna eroderats bort. I artikel IV tas hänsyn till nyligen insamlade magnetotelluriska data som kopplar den väldigt konduktiva Alunskiffern (som associeras med den basala skjuvzonen) till reflektioner som ligger på ett djup av drygt 500 m i östra delen av den seismiska profilen, och ca 1 km strax öster om Seveskållan. Även anomalier i flygmagnetiska data har använts för att begränsa djupet till Baltikas urberg under Kaledoniska skållarna (som tros bestå av den väldigt magnetiska Rätan graniten som finns sydöst om Kaledoniska fronten) till djup mellan 1 till 1,3 km längs profilen öster om Seveskållan. Detta tyder på att den basala skjuvzonen ligger på ett djup av ungefär 1 km och att djupare reflektioner kommer från skjuvzoner från Kaledonisk eller äldre tektonisk deformation i urberget, eller mafiska intrusioner (likt de som återfinns öster och söder om Kaledoniska fronten) i den Baltiska kontinentalplattan. Där- igenom har en ny bild av skållberggrunden och underliggande äldre berg- grund skapats. En ny era har börjat i och med COSC projektet. En enorm mängd högkva- litativa data från COSC-1 väntar på att analyseras av forskargrupper från hela världen, samtidigt som planeringen av COSC-2 börjar under hösten 2015. Ett flertal geofysiska mätningar i området och geologiska undersök- ningar kommer ligga till grund för att skapa en djupare förståelse för den tektoniska utvecklingen i området.

77 Acknowledgements

Looking back at the nearly five and a half years that I’ve spent with the reflection seismic group at the Department of Earth Sciences, Uppsala Univer- sity, I can say that time has truly flown by and it feels as though it’s brought me a remarkable distance from where I began to where I am today. When doing my M. Sc. in space science, aiming for the stars with a master thesis in astronomy, I didn’t expect that I would end up a few kilometers under ground. Today, I can’t imagine a better place to be. I am incredibly grateful for everything that I’ve learned and all that I have experienced, for all the memories that will last a lifetime and for all the people that I have had the pleasure of working with and meeting. The best thing about this job is that it involves such a diverse group of people and the great variety of tasks. Underlying all is a positive energy and a common drive to move forward and explore. It’s a wonderful thing to get to switch the indoors office environment and get out in nature now and then and nearly seven months of my time has been spent in the field. While most of this time has been spent in the wonderful nature of western Jämtland, I’m fortunate to have been given the opportunity to work in many different places. Perhaps the most memorable of these additional fieldworks are my very first fieldwork in Kevitsa in northern Finland and the 40 days and nights spent by the mining town of Solwezi in northern Zambia. To come this far would never have been possible without the people I have been surrounded by and I would like to extend my warmest gratitude to all of you. First of all, I want to thank my second half Marilyn and my son Samuel who stood by me all this time, through the good times and the tough, through weeks of absence for fieldwork and especially the last few very stressful months. It’s truly amazing to come home and be greeted with smiles and hugs, no matter if it’s after weeks in the field or a regular day at the office. You are my biggest source of inspiration and motivation. I’m also thankful for the constant support and encouragement from my parents Per-Eric and Ingrid and my brother Jonas and his wife Jenny as well as my in-laws, the Pascual family (especially Joan and Maricel for all the help during the last few months). I’m very thankful to my main supervisor Christopher Juhlin who took a chance on me and has believed in me from beginning to end. I am often amazed at what you are able to accomplish in minimal time and how you

78 sometimes perform magic on seismic data. I want to thank Alireza Malehmir who has been my co-supervisor since the beginning and who is always quick to help and guide. Over the last year of my Ph. D., Bjarne Almqvist also came in as a co-supervisor and working together with you on the COSC- project was a pleasure. I admire all three of you for your knowledge, enthusiasm, drive and positive energy, which you all have but express in very different ways. You are excellent role-models to any aspiring geoscientist. Thank you also for all the opportunities I have been given, for the responsi- bilities that you have entrusted me with, and for giving me the chance to take part in interesting fieldwork and conferences and visit many incredible places during my Ph. D. studies. I also want to thank David Gee for your unbelievable patience with my sometimes seemingly constant state of confusion when it comes to geology. You are like an interactive encyclopedia of Caledonian geology. Your incredible enthusiasm and undying passion for the geosciences and the Caledonides is a great inspiration. I am grateful for all the time I’ve spent in the field with Hasse Palm and for all the experience and knowledge about seismic fieldwork that I have gained from you. I also want to thank Sverker Olsson and Artëm Kashubin who were my first mentors in “the White house” during seismic acquisition. Thank you Dan Dyrelius for all the interesting conversations and for all the time you spent on providing incredibly informative answers to my questions and e-mails about gravity and magnetic surveying when I was trying to figure this out. Thanks also to Théo Berthet who came to Uppsala during the last year and a half and also has been patient with my questions about potential fields and other things. It’s incredible how you can be interested and ready to dive head first in to absolutely anything. The great diversity of people at the department of Earth Sciences creates a great environment to learn and develop, both in the office and in the field. And not only about geosciences, but also from the many interesting discussions around the coffee table and at field dinners that can cover just about any imaginable topic. Thanks to every single one of you, past and present, who have made this department a special place to be over the past 5 and a half years. Over these years I have had the pleasure of sharing office with Can Yang, Erzhad Gholamrezaie and Fei Huang. Thanks for all the interesting discussions about Chinese and Persian traditions, history and politics and for putting up with my sometimes constantly tapping fingers and feet. I want to thank everyone involved with COSC for making it such a great project. There are too many of you to mention by name; the reflection seismic field crews, the COSC-1 on-site science team, engineers and drill crew, the people at the ICDP, and the many researchers across the globe. A per- sonal thanks though to Henning Lorenz for all your hard work and for helping me out with many COSC related things over the years.

79 I’ve had many great teachers who have been very important to me when I’ve been trying to find my way through new territory; thank you Christo- pher Juhlin, Laust Pedersen, Dan Dyrelius, Hemin Koyi and Håkan Sjöström. I have also been involved in teaching several courses during this time and I never imagined that I would enjoy it as much as I did. It’s been a great learning experience for me in many ways (I believe I’ve learned a lot more than I’ve taught), both from the teaching itself and from working together with excellent teachers such as Laust Pedersen, Dan Dyrelius, Chris- tophe Hieronymous, Bjarne Almqvist, Theó Berthet, Håkan Sjöström, Karin Högdahl and Lars Holmer. I also thank Peter Schmidt whose excellent course material I got to inherit when I started teaching. I’ve been lucky to meet many talented students whom I wish the best of luck in the future. During the years I have collaborated with several people on manuscripts and abstracts whom I haven’t mentioned yet, and it’s been a great pleasure working with all of you. Stefan Buske and Helge Simon from TU Bergakademie Freiberg, Germany, Rüdiger Giese and Felix Krauß from GFZ Potsdam, Germany, Jan-Erik Rosberg and Per-Gunnar Alm from Lund University, Sweden, and Omid Ahmadi from down the hall. Christopher Juhlin, Alireza Malehmir, Bjarne Almqvist, David Gee, Hen- ning Lorenz, Emil Lundberg, Mahdieh Azita Dehghannejad, María García, Magnus Andersson, Karin Högdahl and Jonas Hedin all contributed with constructive criticism and feedback on parts or the full thesis and helped to improve it for which I am very grateful. I gratefully acknowledge the Swedish Research Council for partially funding this Ph. D. and Värmlands Nation for rewarding scholarships to attend conferences. The Geological Survey of Sweden is thanked for providing gravity and magnetic data. Last but not least, to all the people around Fröå with whom I’ve had the pleasure to interact and have interesting conversations on many occasions. Thanks for being positive and interested in the fieldwork (even helping out with fieldwork), for your patience during the noisy seismic acquisition, and for allowing me to run through the forest around your homes with strange equipment strapped to my body.

Peter Hedin, Uppsala, August 2015

80 References

Andersen, T. B., 1998. Extensional tectonics in the Caledonides of southern Nor- way, an overview. Tectonophysics 285 (3-4), 333–351, doi:10.1016/S0040- 1951(97)00277-1. Arnbom, J.-O., 1980. Metamorphism of the Seve Nappes at Åreskutan, Swedish Caledonides. GFF 102 (4), 359–371, doi:10.1080/11035898009454493. Asklund, B., 1938. Hauptzüge der Tektonik und Stratigraphie der mittleren Kaledo- niden in Schweden. Sveriges Geologiska Undersökning, C 417. Asklund, B., 1960. The Geology of the Caledonian Mountain Chain and of adjacent areas in Sweden. Sveriges Geologiska Undersökning, Ba 16. Bergman, S. and Sjöström, H., 1997. Accretion and lateral extension in an orogenic wedge: evidence from a segment of the Seve-Köli terrane boundary, central Scandinavian Caledonides. J. Struct. Geol. 19 (8), 1073–1091, doi:10.1016/S0191-8141(97)00028-X. Cheraghi, S., Malehmir, A. and Bellefleur, G., 2012. 3D imaging challenges in steeply dipping mining structures: New lights on acquisition geometry and processing from the Brunswick no. 6 seismic data, Canada. Geophysics 77 (5), WC109-WC122, doi:10.1190/geo2011-0475.1. Corfu, F., Gasser, D. and Chew, D. M., (Eds.) 2014a. New Perspectives on the Caledonides of Scandinavia and Related Areas, Geol. Soc. London, Spec. Publ. 390, ISBN: 978-1-86239-377-6. Corfu, F., Andersen, T. B. and Gasser, D., 2014b. The Scandinavian Caledonides: main features, conceptual advances and critical questions. in: Corfu, F., Gasser, D., Chew, D.M. (Eds.), New Perspectives on the Caledonides and Related Are- as. Geol. Soc. London, Spec. Publ. 390, 9–43, doi:10.1144/SP390.25. Cosma, C. and Enescu, N., 2001. Characterization of fractured rock in the vicinity of tunnels by the swept impact seismic technique. Int. J. Rock Mech. Min. Sci. 38 (6), 815–821, doi:10.1016/S1365-1609(01)00046-6. Czaplinska, D., Piazolo, S. and Almqvist, B., 2015. Seismic anisotropy in the lower crust: The link between rock composition, microstructure, texture and seismic properties. Geophysical Research Abstracts 17, EGU2015-1666. Dewey, J. F., 1969. Evolution of the Appalachian/Caledonian Orogen. Nature 222 (5189), 124–129, doi:10.1038/222124a0. Dyrelius, D., 1980. Aeromagnetic interpretation in a geotraverse area across the central Scandinavian Caledonides. GFF 102 (4), 421–438, doi:10.1080/11035898009454498. Dyrelius, D., 1985. A Geophysical Perspective of the Scandinavian Caledonides, in: Gee, D. G. and Sturt, B. A., (Eds.) 1985. The Caledonide Orogen - Scandinavia and related areas, 185–194, John Wiley & Sons Ltd. Dyrelius, D., 1986. Gravity and magnetics in the central Scandes. GFF 108 (3), 278– 280, doi:10.1080/11035898609454705.

81 Dyrelius, D., Gee, D. G., Gorbatschev, R., Ramberg, H. and Zachrisson, E., 1980. A profile through the central Scandinavian Caledonides. Tectonophysics 69 (3-4), 247–284, doi:10.1016/0040-1951(80)90213-9. Ebbing, J., England, R. W., Korja, T., Lauritsen, T., Olesen, O., Stratford, W. and Weidle, C., 2012. Structure of the Scandes lithosphere from surface to depth. Tectonophysics 536-537, 1–24, doi:10.1016/j.tecto.2012.02.016. Elming, S.-Å., 1980. Density and magnetic properties of rocks in the Caledonides of Jämtland, Sweden. GFF 102 (4), 439–453, doi:10.1080/11035898009454499. Elming, S.-Å., 1988. Geological modelling based on gravity data from the central part of the Swedish Caledonides. GFF 110 (4), 317–327, doi:10.1080/11035898809452666. England, R. W. and Ebbing, J., 2012. Crustal structure of central Norway and Swe- den from integrated modelling of teleseismic receiver functions and the gravity anomaly. Geophys. J. Int. 191 (1), 1–11, doi:10.1111/j.1365- 246X.2012.05607.x. Fossen, H., 2000. Extensional tectonics in the Caledonides: Synorogenic or postorogenic? Tectonics 19 (2), 213–224, doi:10.1029/1999TC900066. García Juanatey, M. Á., Tryggvason, A., Juhlin, C., Bergström, U., Hübert, J. and Pedersen, L. B., 2013. MT and reflection seismics in northwestern Skellefte Ore district, Sweden. Geophysics 78 (2), B65–B76, doi:10.1190/GEO2012-0169.1. Gasser, D., 2014. The Caledonides of Greenland, Svalbard and other Arctic areas: status of research and open questions. in: Corfu, F., Gasser, D., Chew, D.M. (Eds.), New Perspectives on the Caledonides and Related Areas. Geol. Soc. London, Spec. Publ. 390, 93–129, doi:10.1144/SP390.17. Gee, D. G., 1975a. A Geotraverse through the Scandinavian Caledonides – Östersund to Trondheim. Sveriges Geologiska Undersökning, C 717. Gee, D. G., 1975b. A tectonic model for the central part of the Scandinavian Caledonides. Am. J. Sci. A275, 468–515. Gee, D. G., 1978. Nappe displacement in the Scandinavian Caledonides. Tectonophysics 47 (3-4), 393–419, doi:10.1016/0040-1951(78)90040-9. Gee, D. G., 2015. Caledonides of Scandinavia, Greenland, and Svalbard. Reference Module in Earth Systems and Environmental Science, Elsevier Inc., 15 pp., doi:10.1016/B978-0-12-409548-9.09133-8 Gee, D. G. and Sturt, B. A., (Eds.) 1985. The Caledonide Orogen - Scandinavia and related areas. John Wiley & Sons Ltd., Chichester, Great Britain. ISBN: 0-471- 10504-X Gee, D. G., Kumpulainen, R. and Thelander, T., 1978. The Tåsjön Décollement, Central Swedish Caledonides, Sveriges Geologiska Undersökning, Cq 742. Gee, D.G., Snäll, S., Stejskal, V., 1982. Alunskiffer projektet – Fjällen. Undersök- ningar mellan Östersund och Svenstavik. Rapport till NSG. BRAP 82502, Geo- logical Survey of Sweden, unpublished report Gee, D. G., Kumpulainen, R., Roberts, D., Stephens, M. B., Zachrisson, E. and Thon, A., 1985. Scandinavian Caledonides - Tectonostratigraphic Map, Scale 1:2 000 000, Sveriges Geologiska Undersökning, Ba 35. Gee, D. G., Fossen, H., Henriksen, N. and Higgins, A. K., 2008. From the early Paleozoic platforms of Baltica and Laurentia to the Caledonide Orogen of Scan- dinavia and Greenland. Episodes 31 (1), 44–51. Gee, D. G., Juhlin, C., Pascal, C. and Robinson, P., 2010. Collisional Orogeny in the Scandinavian Caledonides (COSC). GFF 132 (1), 29–44, doi:10.1080/11035891003759188.

82 Gee, D. G., Janák, M., Majka, J., Robinson, P. and van Roermund, H., 2012. Subduction along and within the Baltoscandian margin during closing of the Iapetus Ocean and Baltica-Laurentia collision. Lithosphere 5 (2), 169–178, doi:10.1130/L220.1. Gilotti, J. A. and Kumpulainen, R., 1986. Strain softening induced ductile flow in the Särv thrust sheet, Scandinavian Caledonides. J. Struct. Geol. 8 (3-4), 441– 455, doi:10.1016/0191-8141(86)90062-3. Grimmer, J. C., Glodny, J., Drüppel, K., Greiling, R. O. and Kontny, A., 2015. Ear- ly- to mid-Silurian extrusion wedge tectonics in the central Scandinavian Caledonides. Geology 43 (4), 347–350, doi:10.1130/G36433.1. Higgins, A. K. and Leslie, A. G., 2000. Restoring thrusting in the East Greenland Caledonides. Geology 28 (11), 1019–1022, doi:10.1130/0091- 7613(2000)28<1019:RTITEG>2.0.CO;2. Hossack, J. R. and Cooper, M. A., 1986. Collision tectonics in the Scandinavian Caledonides, in: Coward, M.P. and Ries, A.C. (Eds.), Collision Tectonics. Geol. Soc. London, Spec. Publ. 19, 285–304, doi:10.1144/GSL.SP.1986.019.01.16. Hurich, C. A., 1996. Kinematic evolution of the lower plate during intracontinental subduction: An example from the Scandinavian Caledonides. Tectonics 15 (6), 1248–1263, doi:10.1029/96TC00828. Hurich, C. A., Palm, H., Dyrelius, D. and Kristoffersen, Y., 1989. Deformation of the Baltic continental crust during Caledonide intracontinental subduction: views from seismic reflection data. Geology 17 (5), 423–425, doi:10.1130/0091- 7613(1989)017<0423:DOTBCC>2.3.CO;2. Högbom, A. G., 1909. Studies in the post-Silurian thrust region of Jämtland. GFF 31 (5), 289–346, doi:10.1080/11035890909445606. Högdahl, K., Andersson, U. B. and Eklund, O., (Eds.) 2004. The Trans Scandinavi- an Igneous Belt (TIB) in Sweden: A review of its character and evolution, Geo- logical Survey of Finland, Special Paper 37, Espoo, Finland, ISBN:951-690- 889-6. Högström, A. E. S., Sturkell, E., Ebbestad, J. O. R., Lindström, M. and Ormö, J., 2010. Concentric impact structures in the Palaeozoic of Sweden – the Lockne and Siljan craters. GFF 132 (1), 65–70, doi:10.1080/11035890903469971. Juhlin, C., 1990. Interpretation of the reflections in the Siljan Ring area based on results from the Gravberg-1 borehole. Tectonophysics 173 (1-4), 345–360, doi:10.1016/0040-1951(90)90229-2. Juhojuntti, N., Juhlin, C. and Dyrelius, D., 2001. Crustal reflectivity underneath the Central Scandinavian Caledonides. Tectonophysics 334 (3-4), 191–210, doi:10.1016/S0040-1951(00)00292-4. Klonowska, I., Majka, J., Janák, M., Gee, D. G. and Ladenberger, A., 2014. Pres- sure-temperature evolution of a kyanite-garnet pelitic gneiss from Areskutan: evidence of ultra-high-pressure metamorphism of the Seve Nappe Complex, west-central Jamtland, Swedish Caledonides. in: Corfu, F., Gasser, D., Chew, D.M. (Eds.), New Perspectives on the Caledonides and Related Areas. Geol. Soc. London, Spec. Publ. 390, 321–336, doi:10.1144/SP390.7. Klonowska, I., Janák, M., Majka, J., Froitzheim, N. and Gee, D. G., 2015. Discovery of microdiamond in the Åreskutan Nappe of the Seve Nappe Complex, overlying the COSC-1 drillhole. Geophysical Research Abstracts 17, EGU2015- 12046-1. Korja, T., Smirnov, M., Pedersen, L. B. and Gharibi, M., 2008. Structure of the central Scandinavian Caledonides and the underlying Precambrian basement, new constraints from magnetotellurics. Geophys. J. Int. 175 (1), 55–69, doi:10.1111/j.1365-246X.2008.03913.x.

83 Krauß, F., Simon, H., Giese, R., Buske, S., Hedin, P. and Juhlin, C., 2015. Zero- Offset VSP in the COSC-1 borehole, Geophysical Research Abstracts 17, EGU2015-3255. Kukkonen, I. T., Olesen, O., Ask, M. V. S. and the PFDP WORKING GROUP, 2010. Postglacial Faults in Fennoscandia: Targets for scientific drilling. GFF 132 (1), 71–81, doi:10.1080/11035891003692934. Kukkonen, I. T., Ask, M. V. S. and Olesen, O., 2011. Postglacial fault drilling in Northern Europe: Workshop in Skokloster, Sweden. Sci. Drill. 11, 56–59, doi:10.2204/iodp.sd.11.08.2011. Labrousse, L., Hetényi, G., Raimbourg, H., Jolivet, L. and Andersen, T. B., 2010. Initiation of crustal-scale thrusts triggered by metamorphic reactions at depth: Insights from a comparison between the Himalayas and Scandinavian Caledonides. Tectonics 29 (5), 1–14, doi:10.1029/2009TC002602. Ladenberger, A., Be’eri-Shlevin, Y., Claesson, S., Gee, D. G., Majka, J. and Romanova, I. V., 2014. Tectonometamorphic evolution of the Åreskutan Nappe - Caledonian history revealed by SIMS U-Pb zircon geochronology. in: Corfu, F., Gasser, D., Chew, D.M. (Eds.), New Perspectives on the Caledonides and Related Areas. Geol. Soc. London, Spec. Publ. 390, 337–368, doi:10.1144/SP390.10. Larner, K. L., Gibson, B. R., Chambers, R. and Wiggins, R. A., 1979. Simultaneous estimation of residual static and crossdip corrections. Geophysics 44 (7), 1175– 1192, doi:10.1190/1.1441001. Law, R. D., Searle, M. P. and Godin, L., (Eds.) 2006. Channel Flow, Ductile Extru- sion and Exhumation in Continental Collision Zones, Geol. Soc. London, Spec. Publ. 268, ISBN:1-86239-209-9. Li, Y. and Oldenburg, D. W., 1998. 3-D inversion of gravity data. Geophysics 63 (1), 109–119, doi:10.1190/1.1444302. Lorenz, H., 2010. The Swedish Deep Drilling Program: For Science and Society. GFF 132 (1), 25–27, doi:10.1080/11035891003763354. Lorenz, H., Gee, D. G. and Juhlin, C., 2011. The Scandinavian Caledonides- scientific drilling at mid-crustal level in a Palaeozoic major collisional orogen. Sci. Drill. 11, 60–63, doi:10.2204/iodp.sd.11.10.2011. Lorenz, H., Rosberg, J.-E., Juhlin, C., Bjelm, L., Almqvist, B. S. G., Berthet, T., Conze, R., Gee, D. G., Klonowska, I., Pascal, C., Pedersen, K., Roberts, N. M. W. and Tsang, C.-F., 2015a. COSC-1 – drilling of a subduction-related allochthon in the Palaeozoic Caledonide orogen of Scandinavia. Sci. Drill. 19, 1–11, doi:10.5194/sd-19-1-2015. Lorenz, H., Rosberg, J.-E., Juhlin, C., Bjelm, L., Almqvist, B. S. G., Berthet, T., Conze, R., Gee, D. G., Klonowska, I., Pascal, C., Pedersen, K., Roberts, N. M. W. and Tsang, C.-F., 2015b. Operational Report about Phase 1 of the Collision- al Orogeny in the Scandinavian Caledonides scientific drilling project (COSC-1), ICDP Operational Report, GFZ German Research Centre for Geo- sciences, doi:10.2312/ICDP.2015.002. Lundberg, E., 2014. 2D and 3D Reflection Seismic Studies over Scandinavian De- formation Zones. Acta Universitatis Upsaliensis, Digital Comprehensive Sum- mary of Uppsala Dissertations from the Faculty of Science and Technology 1102, Uppsala University, Uppsala, Sweden. Lundberg, E. and Juhlin, C., 2011. High resolution reflection seismic imaging of the Ullared Deformation Zone, southern Sweden. Precambrian Res. 190 (1-4), 25– 34, doi:10.1016/j.precamres.2011.07.012.

84 Majka, J., Be’eri-Shlevin, Y., Gee, D. G., Ladenberger, A., Claesson, S., Konečny, P. and Klonowska, I., 2012. Multiple monazite growth in the Åreskutan migmatite: Evidence for a polymetamorphic Late Ordovician to Late Silurian evolution in the Seve Nappe Complex of west-central Jämtland, Sweden. J. Geosci. 57 (1), 3–23, doi:10.3190/jgeosci.112. Majka, J., Rosén, Å., Janák, M., Froitzheim, N., Klonowska, I., Manecki, M., Sasinkova, V. and Yoshida, K., 2014a. Microdiamond discovered in the Seve Nappe (Scandinavian Caledonides) and its exhumation by the “vacuum-cleaner” mechanism. Geology 42 (12), 1107–1110, doi:10.1130/G36108.1. Majka, J., Janák, M., Andersson, B., Klonowska, I., Gee, D. G., Rosén, Å. and Kosminska, K., 2014b. Pressure-temperature estimates on the Tjeliken eclogite: new insights into the (ultra)-high-pressure evolution of the Seve Nappe Com- plex in the Scandinavian Caledonides. in: Corfu, F., Gasser, D., Chew, D.M. (Eds.), New Perspectives on the Caledonides and Related Areas. Geol. Soc. London, Spec. Publ. 390, 369–384, doi:10.1144/SP390.14. Malavieille, J. and Trullenque, G., 2009. Consequences of continental subduction on forearc basin and accretionary wedge deformation in SE Taiwan: Insights from analogue modeling. Tectonophysics 466 (3-4), 377–394, doi:10.1016/j.tecto.2007.11.016. Malehmir, A., Schmelzbach, C., Bongajum, E., Bellefleur, G., Juhlin, C. and Tryggvason, A., 2009. 3D constraints on a possible deep >2.5 km massive sul- phide mineralization from 2D crooked-line seismic reflection data in the Kristineberg mining area, northern Sweden. Tectonophysics 479 (3-4), 223– 240, doi:10.1016/j.tecto.2009.08.013. Malehmir, A., Dahlin, P., Lundberg, E., Juhlin, C., Sjöström, H. and Högdahl, K., 2011. Reflection seismic investigations in the Dannemora area, central Sweden: Insights into the geometry of polyphase deformation zones and magnetite-skarn deposits. J. Geophys. Res. 116 (11), 1–21, doi:10.1029/2011JB008643. Marfurt, K. J., Scheet, R. M., Sharp, J. A. and Harper, M. G., 1998. Suppression of the acquisition footprint for seismic sequence attribute mapping. Geophysics 63 (3), 1024–1035, doi:10.1190/1.1444380. McKenzie, D. P. and Parker, R. L., 1967. The North Pacific: an example of Tecton- ics on a Sphere. Nature 216 (5122), 1276–1280, doi:10.1038/2161276a0. Milkereit, B., Berrer, E. K., King, A. R., Watts, A. H., Roberts, B., Adam, E., Eaton, D. W., Wu, J. and Salisbury, M. H., 2000. Development of 3-D seismic explora- tion technology for deep nickel-copper deposits – A case history from the Sud- bury basin, Canada. Geophysics 65 (6), 1890, doi:10.1190/1.1444873. Morgan, W. J., 1968. Rises, trenches, great faults, and crustal blocks. J. Geophys. Res. 73 (6), 1959–1982, doi:10.1029/JB073i006p01959. Mosar, J., 2003. Scandinavia’s North Atlantic passive margin. J. Geophys. Res. 108 (B8), 2360, doi:10.1029/2002JB002134. Nedimović, M. R. and West, G. F., 2003a. Crooked‐line 2D seismic reflection imaging in crystalline terrains: Part 1, data processing. Geophysics 68 (1), 274–285, doi:10.1190/1.1543213. Nedimović, M. R. and West, G. F., 2003b. Crooked-line 2D seismic reflection imaging in crystalline terrains: Part 2, migration. Geophysics 68 (1), 286–296, doi:10.1190/1.1543214. Olesen, O., Gellein, J., Gernigon, L., Kihle, O., Koziel, J., Lauritsen, T., Mogaard, J. O., Myklebust, R., Skilbrei, J. R. and Usov, S., 2010a. Magnetisk Anomalikart, Norge og tilgrensende områder/Magnetic anomaly map, Norway and adjacent areas. Målestokk/Scale 1:3 million, Norges Geologiska Undersøkelse/Geological Survey of Norway, Trondheim.

85 Olesen, O., Brönner, M., Ebbing, J., Gellein, J., Gernigon, L., Koziel, J., Lauritsen, T., Myklebust, R., Pascal, C., Sand, M., Solheim, D. and Usov, S., 2010b. New aeromagnetic and gravity compilations from Norway and adjacent areas: methods and applications. In: Vining, B. A. and Pickering, S. C. (Eds.) Petroleum Geology: From Mature Basins to New Frontiers – Proceedings of the 7th Petro- leum Geology Conference. Geol. Soc. London, 559–586, doi:10.1144/0070559 Olesen, O., Ebbing, J., Gellein, J., Kihle, O., Myklebust, R., Sand, M., Skilbrei, J. R., Solheim, D. and Usov, S., 2010c. Tyngdeanomalikart, Norge og tilgrensende områder/Gravity anomaly map, Norway and adjacent areas. Målestokk/Scale 1:3 million, Norges Geologiska Undersøkelse/Geological Survey of Norway, Trondheim. Palm, H., 1984. Time-delay interpretation of seismic refraction data in the Caledoni- an front, Jämtland, central Scandinavian Caledonides. GFF 106 (1), 1–14, doi:10.1080/11035898409454597. Palm, H., Gee, D. G., Dyrelius, D. and Björklund, L., 1991. A reflection seismic image of Caledonian structure in Central Sweden. Sveriges Geologiska Undersökning, Ca 75. Park, C. B., 1996. Swept impact seismic technique (SIST). Geophysics 61 (6), 1789–1803, doi:10.1190/1.1444095. Pascal, C., Ebbing, J. and Skilbrei, J. R., 2007. Interplay between the Scandes and the Trans-Scandinavian Igneous Belt: Integrated thermo-rheological and potential field modelling of the Central Scandes profile. Nor. Geol. Tidsskr. 87 (1-2), 3–12. Rodriguez-Tablante, J., Tryggvason, A., Malehmir, A., Juhlin, C. and Palm, H., 2007. Cross-profile acquisition and cross-dip analysis for extracting 3D information from 2D surveys, a case study from the western Skellefte District, northern Sweden. J. Appl. Geophys. 63 (1), 1–12, doi:10.1016/j.jappgeo.2007.03.001. Rosberg, J.-E. and Lorenz, H., 2012. A new scientific drilling infrastructure in Swe- den. Geophysical Research Abstracts 14, EGU2015-7379. Schmidt, J., 2000. Deep seismic studies in the western part of the Baltic Shield. Acta Universitatis Upsaliensis, Uppsala Dissertations from the Faculty of Science and Technology 24, Uppsala University, Uppsala, Sweden. Simon, H., Krauß, F., Hedin, P., Buske, S., Giese, R. and Juhlin, C., 2015. A combined surface and borehole seismic survey at the COSC-1 borehole. Geophysi- cal Research Abstracts 17, EGU2015-4554. Strömberg, A., Karis, L., Zachrisson, E., Sjöstrand, T., Skoglund, R., Lundegårdh, P. H. and Gorbatschev, R., 1984. Berggrundskarta över Jämtlands län utom förut- varande Fjällsjö kommun, scale 1:200 000. Sveriges Geologiska Undersökning, Ca 53, Uppsala, Sweden. Söderlund, U., Isachsen, C. E., Bylund, G., Heaman, L. M., Patchett, P. J., Vervoort, J. D. and Andersson, U. B., 2005. U-Pb baddeleyite ages and Hf, Nd isotope chemistry constraining repeated mafic magmatism in the Fennoscandian Shield from 1.6 to 0.9 Ga. Contrib. to Mineral. Petrol. 150 (2), 174–194, doi:10.1007/s00410-005-0011-1. Söderlund, U., Elming, S.-Å., Ernst, R. E. and Schissel, D., 2006. The Central Scan- dinavian Dolerite Group-Protracted hotspot activity or back-arc magmatism?. Constraints from U-Pb baddeleyite geochronology and Hf isotopic data. Pre- cambrian Res. 150 (3-4), 136–152, doi:10.1016/j.precamres.2006.07.004.

86 Tamura, Y., Ishizuka, O., Aoike, K., Kawate, S., Kawabata, H., Chang, Q., Saito, S., Tatsumi, Y., Arima, M., Takahashi, M., Kanamaru, T., Kodaira, S. and Fiske, R. S., 2010. Missing oligocene crust of the Izu-Bonin Arc: Consumed or rejuvenat- ed during collision? J. Petrol. 51 (4), 823–846, doi:10.1093/petrology/egq002. Törnebohm, A. E., 1888. Om fjällproblemet. GFF 10 (5), 328–336, doi:10.1080/11035898809444211. Weihed, P., 2010. Palaeoproterozoic mineralized volcanic arc systems and tectonic evolution of the Fennoscandian Shield: Skellefte District Sweden. GFF 132 (1), 83–91, doi:10.1080/11035890903477743. Wenning, Q., Almquist, B., Ask, M. V. S., Schmitt, D. R. and Zappone, A., 2015. Elastic anisotropy and borehole stress estimation in the Seve Nappe Complex from the COSC-1 well, Åre, Sweden. Geophysical Research Abstracts 17, EGU2015-6594. Vine, F. J. and Matthews, D. H., 1963. Magnetic Anomalies Over Oceanic Ridges. Nature 199 (4897), 947–949, doi:10.1038/199947a0. Wu, J., 1996. Potential pitfalls of crooked-line seismic reflection surveys. Geophys- ics 61 (1), 277–281, doi:10.1190/1.1443949. Wu, J., Milkereit, B. and Boerner, D. E., 1995. Seismic imaging of the enigmatic Sudbury Structure. J. Geophys. Res. 100 (B3), 4117–4130, doi:10.1029/94JB02647. Yan, P., García Juanatey, M. Á., Kalscheuer, T. and Juhlin, C., 2015. The upper crust of the Scandinavian Caledonides as seen by Magnetotellurics. Geophysical Research Abstracts 17, EGU2015-11587. Yilmaz, Ö., 2001. Seismic Data Analysis: Processing, Inversion, and Interpretation of Seismic Data, 2nd ed. Investigations in Geophysics 10, Society of Explora- tion Geophysicists, 2065 pp. ISBN: 978-1-56080-158-0.

87 Acta Universitatis Upsaliensis Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1281 Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science and Technology, Uppsala University, is usually a summary of a number of papers. A few copies of the complete dissertation are kept at major Swedish research libraries, while the summary alone is distributed internationally through the series Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology. (Prior to January, 2005, the series was published under the title “Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology”.)

ACTA UNIVERSITATIS UPSALIENSIS Distribution: publications.uu.se UPPSALA urn:nbn:se:uu:diva-261112 2015