Geological Survey of Denmark and Greenland Bulletin 11, 185-204 185 Faults and fractures in central West Greenland: onshore expression of continental break-up and sea-floor spreading in the Labrador – Baffin Bay Sea Robert W. Wilson, Knud Erik S. Klint, Jeroen A.M. van Gool, Kenneth J.W. McCaffrey, Robert E. Holdsworth and James A. Chalmers The complex Ungava fault zone lies in the Davis Strait and separates failed spreading centres in the Labrador Sea and Baffin Bay. This study focuses on coastal exposures east of the fault-bound Sisimiut basin, where the onshore expressions of these fault systems and the influence of pre-existing basement are examined. Regional lineament studies identify five main systems: N–S, NNE–SSW, ENE–WSW, ESE–WNW and NNW–SSE. Field studies reveal that strike-slip movements predominate, and are consistent with a ~NNE–SSW-oriented sinistral wrench system. Extensional faults trending N–S and ENE–WSW (basement-parallel), and compressional faults trending E–W, were also identified. The relative ages of these fault systems have been interpreted using cross-cutting relationships and by correlation with previously identified structures. A two-phase model for fault development fits the development of both the onshore fault systems observed in this study and regional tectonic structures offshore. The conclusions from this study show that the fault patterns and sense of movement on faults onshore reflect the stress fields that govern the opening of the Labrador Sea – Davis Strait – Baffin Bay seaway, and that the wrench couple on the Ungava transform system played a dominant role in the development of the onshore fault patterns. Keywords: faults and fractures, extensional tectonics, wrench systems, sedimentary basins, basement reactivation, West Greenland _______________________________________________________________________________________________________________________________________ R.W.W., K.J.W.M. & R.E.H., Reactivation Research Group, Department of Earth Sciences, University of Durham, Durham DH1 3LE, UK. E-mail: robert.wilson@durham.ac.uk K.E.S.K., J.A.M.v.G. & J.A.C., Geological Survey of Denmark and Greenland, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark. Introduction Pre-existing heterogeneities in the continental crust, such as shear zones and terrain boundaries, have long been known to influence the structure and development of later deformation events (Butler et al. 1997; Holdsworth et al. 1997, and references therein). The sedimentary basins of the Labrador Sea – Baffin Bay region are situated west of Greenland (Fig. 1) and are early Cenozoic failed spreading centres (Chalmers & Pulvertaft 2001), separated by the Davis Strait. The orientation of the Davis Strait relative to the proposed spreading centres in the Labrador Sea and Baffin Bay is consistent with the geometry of an ‘exten- sional transform zone’ (Taylor et al. 1994). Steep base- ment fabrics of the Nagssugtoqidian orogen trend highly obliquely to these offshore structures (Fig. 1) and coin- cide with this ‘step-over zone’ in the Davis Strait. Fault systems fundamental to the development of sedimentary basins in the Davis Strait are exposed onshore in West Greenland. In this project, the onshore fault systems of central West Greenland were studied in order to improve © GEUS, 2006. Geological Survey of Denmark and Greenland Bulletin 11, 185–204. Available at: www.geus.dk/publications/bull 186 1000 200 km Baffin Island Disko Greenland Oceanic crust Transitional crust Mesozoic basin Unknown crust Exposed shallow continental basement ? ? ? ? 50°60° 70° 65° 65° 60° 70°70° 60° 50° Nuus sua q bas in Iti lli F Z D av is S tr ai t hi gh W est G reenland shelf N uu k ba si n C an ad a Greenland Nagssugtoqidian orogen Fig. 2 Baffin Bay Labrador Sea Border to Nagssugtoqidian orogenic belt Basement fabric Nuussuaq Nuuk Inland Ice Aasiaat Palaeogene volcanics basin Hecla high Ik er m iu t F Z D av is St ra it Sisimiut Maniitsoq high NISZ NSSZ Ikermiut FZ: Ikermiut fault zone Itilli FZ: Itilli fault zone NISZ: Nordre Isortoq shear zone NSSZ: Nordre Strømfjord shear zone Ubekendt Ejland Svartenhuk Halvø U ng av a fa ul t z on e Saglek basin Sis im iut lin eam ent Lady Franklin platform Sisimiut basin Nuuk Fig. 1. Regional tectonic map of the offshore geology of the Labrador Sea – Baffin Bay area between Canada and Greenland. Modified from Chalmers & Pulvertaft (2001). Facing page: Fig. 2. Geological map of the Nagssugtoqidian orogen of central West Greenland (modified from Escher & Pulvertaft 1995). A: Outline map of Greenland highlighting the region covered in Fig. 2B. B: Geological map of the Nagssugtoqidian orogen showing main lithological units and basement structures. C: Topographic contour map of the central coastal area showing the field camps chosen for this study (camps 0 to 4); black dashed lines highlight major topographic escarpments. D: 3-D model view of NNE-trending coastal escarpment, constructed in ArcGIS by draping a Landsat image onto a topographic model. Abbreviations used: SNO, CNO and NNO are the southern, central and northern Nagssug- toqidian orogen, respectively. ITZ, Ikertôq thrust zone; NISZ, Nordre Isortoq shear zone; NSSZ, Nordre Strømfjord shear zone. 187 N orth A m erican plate Greenland Iceland 10 km Hilly relief Planated relief 200–300 m > 1000 m Increasing height of topography Camp 0 Camp 1 D 50 km Disko Bugt Inland Ice Aasiaat NaternaqKangaatsiaq Attu Sisimiut Kangerlussuaq Sø nd re Str øm fjo rd Nordre Strømfjord Arfersiorfik 54° 67° 69° 68° 51° N ag ss ug to qi di an o ro ge n N or th A tla nt ic c ra to n C N O N N O SN O NISZ NSSZ ITZ Ussu it ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ Basement fabrics PROTEROZOIC ARCHAEAN Supracrustal rocks Surficial Quaternary deposits Metasedimentary rocks Anorthosite and ultrabasic rocks Granitic intrusions Orthogneisses Thrusts Calc-alkaline intrusions Arfersiorfik and Sisimiut suites Granitic intrusions Metasedimentary rocks Archaean gneisses reworked in the Palaeoproterozoic N N B C A Nordre Isort oq 67°10' 53° 54° Camp 4 Camp 2 Camp 3 Camp 0 Camp 1 Camp 4 Camp 2 Camp 3 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 Nordre Isortoq C 10 km Coastal flats Mountains Nord re Str øm fjor d N or dr e Is or to q Coastal escarpment Greenland plate Canada Mapping Mapping arearea a 67°20' 67°30' 67°10' 53° 67°20' 67°30' Contours m above sea level Mapping area Nord re Str øm fjo rd Nord re Str øm fjo rd Nord re Str øm fjo rd 67°40'67°40'67°40' 188 the understanding of the role played by basement reacti- vation in offshore basin development. In the summer of 2003 field work was carried out in an area along the coast stretching from Nordre Strøm- fjord (Nassuttooq) in the north, to Nordre Isortoq in the south (Figs 2, 3). This area was selected because it is be- lieved that two ENE-striking Palaeoproterozoic shear zones, the Nordre Strømfjord shear zone and the Nordre Isortoq shear zone, were reactivated and played an impor- tant role in the development of the Mesozoic to Tertiary sedimentary basins. The western projection of the Nordre Isortoq shear zone appears to coincide with the southern faulted boundary of the Sisimiut basin offshore (Figs 1, 2). In addition, the offshore basins occur close to the coast in this area, and the offshore extensional faults that drop the top of the basement down to 3 km below sea level, within 10 km west of the coast, are believed to correlate with fault escarpments onshore in this area. Tectonic and geological setting Offshore The Labrador Sea and Baffin Bay formed during diver- gent plate motion between Greenland and North America during the early Cenozoic (Chalmers & Pulvertaft 2001, and references therein). The extensional basins of Baffin Bay and the Labrador Sea are separated by a bathymetric high in the Davis Strait (Fig. 1). This transverse ridge is inter- preted as a complex sinistral-shear transform fault zone, known as the Ungava fault system (Fig. 1; Chalmers et al. 1995). Extensional faulting and tectonic subsidence are thought to have commenced in the early Cretaceous, at the same time as sea-floor spreading in the North Atlantic south of the Charlie Gibbs fracture zone. Opening started during the Paleocene (Chian & Louden 1994; Chalmers & Laursen 1995), and sea-floor spreading appears to have ceased by the Oligocene. Interpretation of seismic reflection data has revealed the existence of a number of sedimentary basins offshore western Greenland (Chalmers et al. 1995; Whittaker 1995). One such basin is the deep Sisimiut basin, located in the Davis Strait to the west of the Nordre Strømfjord region (Fig. 1). At about 10 km west of the coast, the top of the basement is at c. 3 km depth, while there is no cover preserved on top of basement exposed onshore. Therefore, the eastern border of the basin must be a major fault, but it is located too close to the coast to have been surveyed by a seismic experiment. The orientation of this bounding fault is likely to follow the NNE–SSW trend of the coastline. The western margin of the Sisimiut basin is the NNE–SSW-trending Ikermiut fault zone (Fig. 1), a transpressional flower structure developed along the trans- form fault (Ungava fault zone) between the North Amer- ican and Greenland plates (Fig. 2) formed during the Pal- aeogene (i.e. it cuts early Eocene strata; Chalmers & Pul- vertaft 2001). The Nukik platform lies to the south of the Sisimiut basin and is separated from it by a line of ENE- and E-trending faults that coincide with the offshore ex- tension of the Nordre Isortoq shear zone onshore (Figs 1, 2). It has therefore been proposed that the faults at the southern margin of the basin developed by reactivation of basement shear zone structures in the central Nagssugto- qidian orogen. These faults affect Mesozoic sediments and are overstepped by Paleocene sediments, so that the latest significant movement on them must have been prior to the end of the Paleocene. Onshore Onshore exposures in central West Greenland, from Søndre Strømfjord in the south to Disko Bugt in the north, com- prise high-grade gneisses of the Palaeoproterozoic Nags- sugtoqidian orogen (Fig. 2; Ramberg 1949; van Gool et al. 2002). The Nagssugtoqidian orogen is a 300 km wide belt of predominantly Archaean orthogneisses, Palaeopro- terozoic paragneisses and intrusive rocks, that were rewor- ked during Palaeoproterozoic orogenesis (van Gool et al. 2002). These basement rocks form ENE-trending linear belts of steeply dipping gneisses, some of which are crus- tal-scale shear zones (i.e. the Nordre Strømfjord and Nordre Isortoq shear zones, Fig. 2), which alternate with zones dominated by kilometre-scale fold structures (van Gool et al. 2002). The Nagssugtoqidian orogen is divided into three tectonic segments: the southern, central and northern Nagssugtoqidian orogen (Fig. 2; Marker et al. 1995). The onshore research in this study lies entirely with- in the granulite facies orthogneisses of the central Nags- sugtoqidian orogen, which is bound to the north by the Facing page: Fig. 3. Lineament map of the main study area derived from lineament mapping of a Landsat TM image at 1:100 000 scale (total 1284 line- aments), using ArcGIS. The main rose diagram (bottom right) shows the distribution of lineaments for this map, while smaller rose dia- grams (left) show results from aerial photograph analyses (at 1:20 000 scale) for each field camp. Green, system 1; blue, system 2; red, sys- tem 3; yellow, system 4; purple, system 5. Red box shows the posi- tion of Fig. 4. 189 Nordre Iso rtoq Nord re S trø mfjor d Fig. 4 n = 295 n = 784 n = 320n = 320 n = 821n = 821 n = 295 n = 784 n = 320 n = 821 n = 1284n = 1284 N N N N 67°10´N67°10´N Landsat lineaments (1:100 000)Landsat lineaments (1:100 000) Aerial photograph lineamentsAerial photograph lineaments (1:20 000) 000) Aerial photograph lineaments (1:20 000) ESE–WNWESE–WNW SE–NWSE–NW Lineaments:Lineaments: 6 km6 km NE–SWNE–SW ENE–WSWENE–WSW E–WE–W NNE–SSWNNE–SSW N–SN–S ESE–WNW SE–NW NE–SW ENE–WSW E–W NNE–SSW N–S N N N N Camp 0Camp 0 Camp 1Camp 1 Camp 2Camp 2 Cam p 4 Cam p 4 67°10´N NNW–SSENNW–SSENNW–SSE Lineaments: 6 km N N N N N Camp 0 Camp 1 Camp 2 Cam p 4 6 km N 190 Nordre Strømfjord shear zone (Marker et al. 1995; van Gool et al. 2002) and to the south by the Ikertôq thrust zone (Fig. 2). Indirect topographic evidence from geomorphological investigations has suggested that late (Mesozoic or Ceno- zoic) onshore fault movements may have occurred (Bonow 2004). For example, in some of the larger inlets and val- leys, characteristic recent shelly marine sands can be observed up to 30 m above sea level. These are likely to have been uplifted due to isostatic rebound following glacial retreat. However, variations in elevation of these palaeoshorelines may also result from differential vertical fault movements or differential unloading. On the larger scale, the Nordre Strømfjord shear zone marks a major change between two landscape types (Japsen et al. 2002). South of the Nordre Strømfjord shear zone topography is planated, with flat mountain tops forming a plateau that gradually increases in height southwards from 500 to 1000 m (Fig. 2C, D). In contrast, north of the Nordre Strømfjord shear zone, the land has a hilly relief with a relatively flat and low- lying topography with isolated hills up to 300 m high. Locally, the change in landscape type occurs across a more than 500 m high, ENE–WSW-oriented escarpment that drops down to the north (Fig. 2). There is also a pronounced NNE–SSW-oriented escarp- ment almost 1 km high that drops down to the west between Nordre Strømfjord and Nordre Isortoq (Fig. 2). This major escarpment separates low-lying (50–150 m high) coastal flats to the west from the much higher (500 m+) mountains to the east (Fig. 2) and can be traced for over 80 km, from Sisimiut in the south to Nordre Strøm- fjord in the north. Similar escarpments can be observed in the near offshore on both bathymetric and horizon maps of depth to basement, thus supporting the theory that onshore structures reflect those offshore. Methods The present study combines regional to outcrop-scale mapping and regional studies of remotely sensed data to determine fault-fracture geometries, distribution, relative timing and kinematics in selected key areas of central West Greenland. Regional studies comprised satellite image and aerial photograph analysis at a variety of scales (1:500 000; 1:100 000; 1:20 000) in order to identify lineaments and other geological structures (e.g. variations in lithology, fabric intensity, faults, fractures). Field investigations were carried out in the well-exposed Precambrian basement rocks in key areas of interest that were identified during aerial photograph analysis prior to departure. A number of field camps were used during the mapping (Fig. 2C). Camp 0 was located at Inussuk, a site visited previously in 2002 (Japsen et al. 2002). Camps 1 and 2 were located on the north and south shores of Nordre Strømfjord, while camp was on the north shore of Nordre Isortoq. Camp 3 was farther inland, and is not analysed further in this study. During field work, fault and fracture systems were mapped, and the following structural data were collected for statistical/structural analysis: • Fault attributes including: orientation; kinematics; fault surface characteristics; mineralisation. • Relative age relationships. • Structural/statistical analyses to determine kinematic patterns. Over 200 pseudotachylite and mica-bearing fault-rock samples were also collected from different fault sets at var- ious localities in order to date the fault movements using 40Ar/39Ar geochronology (results to be discussed elsewhere). All field data were geospatially located (to 5 m resolution) using Global Positioning System (GPS) waypoint collec- tion, and were subsequently stored in a computer database with links to Geographic Information System (GIS) based maps. Fault and fracture characterisation Fractures include all brittle structures such as joints, fis- sures, cracks, veins, etc. that are not faults, bedding or cleavage surfaces, and are larger than the grain size of the rock. In general, fractures are defined as dominantly ten- sile (mode I) cracks, and as such, they are associated with characteristic stress, strain and displacement fields. They are distinguished from small faults by distinctive surface textures and lack of shear displacements. Faults are mapped where distinct offsets have been identified, often with a development of slip striae on the surface (slickenlines). Criteria for determining the sense of movement were based on methods outlined in McClay (1987) and Petit (1987). In the present study faults were classified as: • Normal (extensional dip / oblique-slip fault). • Reverse (compressional dip / oblique-slip fault). • Strike-slip faults (dextral or sinistral). The classification of faults and fractures into systems was primarily based on the orientation of the structures (i.e. 191 trend of lineament or strike of plane). Structures with dif- ferent orientations can reflect different deformation phases, but in complex fault zones developed in three-dimensional (3-D) strain fields, multiple fault and fracture orienta- tions may develop during a single event (see e.g. De Paola et al. 2005). Therefore further classification needs to be applied, either through systematic fracture properties (such as surface type or mineralisation), or through kinematic studies, in order to determine if only one or several phases of deformation are apparent. Fault and fracture measurement technique In order to accurately classify the fault and fracture sys- tems, populations of at least 50 fractures/faults were meas- ured at most localities. Faults were classified according to type (normal/reverse dip-slip faults or dextral/sinistral strike-slip faults). Fault orientations and the directions of slickenlines (when observed) were measured, and the fol- lowing characteristics recorded: • Surface shape: The overall fracture shape (metre-scale) was described as listric, planar, undulating or irregular. • Surface roughness character: The surface roughness char- acter (millimetre scale) was described as smooth, rough or slickenside (striae). • Other features: Some fractures/faults have a filling of iron oxide precipitates, quartz crystals, epidote, or pref- erential growth of other crystals on the surface show- ing the slip direction. Special types of fractures such as conjugate shear fractures, en échelon fractures, plumose jointing etc. were noted if present. = + + + or System 1 System 4 (yellow) System 2 System 3 YoungestOldest 1 km N Fig. 4. Age relationships interpreted from cross-cutting relationships of lineament systems derived from aerial photographs for camp 1. Four dominant lineament trends are apparent: N–S (system 2), NNE–SSW (system 4), ENE–WSW (system 1), and NNW–SSE (system 3). Through cross- cutting relationships a relative order of fault development is apparent, as indicated across the bottom of the image. 192 Lineament mapping Lineament maps for the central Nagssugtoqidian orogen (Fig. 3) were plotted from Landsat TM images and aerial photographs at a variety of scales (1:500 000 and 1:100 000 for Landsat images and 1:20 000 for aerial photographs). Images were georeferenced and displayed in a GIS envi- ronment and the lineaments picked by hand. After inter- pretation, lineaments were then refined using digital ter- rain model (DTM) analysis and compared to pre-existing geological maps of the region (e.g. Henriksen et al. 2000). As the data are stored in a GIS, attribute data for each lineament (i.e. trend; length; offset; other features) were also measured or calculated and stored. Spatial analysis and rose plotting tools in ArcView GIS were used to analyse the ori- entation (Fig. 3) and spatial distribution of these structures. A more detailed analysis of selected areas was then car- ried out at 1:20 000 scale using aerial photographs. As image resolutions are much higher in aerial photographs (2 m pixel size), particular attention was paid to how the lineaments interact with topography (e.g. V-ing into val- leys, etc.) to gain a better understanding of their overall geometry. Generally all lineaments picked appear to have a steep dip as only minor interactions with topography were observed. Attention was also paid to cross-cutting relationships between lineaments in an attempt to deter- mine the relative timing of structures (Fig. 4). Lineament systems In total 1284 lineaments have been identified from Land- sat TM images (pixel size 30 m) at 1:100 000 scale (Fig. 3). Lineaments derived from both Landsat and aerial pho- tographs have been grouped into systems based on their orientation. Five main lineament systems (N–S, NNE– SSW, ENE–WSW, ESE–WNW, and NNW–SSE) have been identified (Fig. 3; Table 1). System 1 structures (green; Figs 3, 4) are oriented ENE– WSW (~060–090° trend), and are pervasively distributed across the region. This system has a trend similar to the Nordre Strømfjord and Nordre Isortoq fjords, and lies parallel to the regional basement fabric (foliation, gneis- sic banding, and shear zones; van Gool et al. 2002). Note that as these lineaments may represent either faults or basement fabrics, care must be taken when analysing these quantitatively. In an attempt to minimise the amount of oversampling, only the most pronounced lineaments (e.g. most weathered out) that mark a distinct change in struc- ture were mapped, while those that are clearly basement fabrics (i.e. those showing ductile features such as folds) were not. System 2 lineaments (blue; Figs 3, 4) are N–S oriented (trend ~350–010°), and often show sinistral offsets of pre- existing structures (basement fabrics). This system can be traced from Nordre Isortoq to the northernmost part of the investigated area (Fig. 3), and previous investigations indicated that they may continue as far north as Aasiaat, Disko and Nuussuaq (Japsen et al. 2002). The fault zones are closely spaced (100–500 m), and strike-slip separations of up to 30 m have been observed. System 3 lineaments (red; Figs 3, 4) are NNW–SSE oriented (trend ~140–170°). They are closely spaced (50– 100 m), and offsets of marble beds show net dextral sepa- rations in the order of 20–40 m (Fig. 4). This system is most pronounced in the Nordre Strømfjord shear zone, and less dominant in the Nordre Isortoq shear zone (see rose diagrams in Fig. 3). System 4 lineaments (yellow; Figs 3, 4) are oriented NNE– SSW (trend ~010–040°) and are strongly developed in the Nordre Strømfjord shear zone region (camps 0 and 1, Fig. 3). These structures show apparent sinistral strike- slip separations of up to 400 m in the westernmost part of the study area. The spacing between them increases from approximately 500 m at the coast, to approximately 2 km farther inland. The same lineament directions were encoun- tered at camps 2 and 4, south of Nordre Strømfjord, where these structures are shorter and more discontinuous, pos- sibly due to differences in rock type and fabric between the two areas. System 5 lineaments (purple; Fig. 3) consist of E–W to ESE–WNW (trend ~090–120°) -oriented structures. The valleys that distinguish this system are generally 10–30 m wide and have a curved trend. This system appears to be mostly localised into two specific areas: the first of these lies in the fold belt south of Nordre Strømfjord (Fig. 3), and the second is located south of Sisimiut (Fig. 1). Relative timings Figure 4 shows an aerial photograph of an area around camp 1 where an apparent order of lineament develop- ment can be deduced. The oldest structures appear to be system 1 (green), and in this area these structures appear to be basement fabrics in the form of alternating layers of semipelite and marble up to 100 m thick (Henriksen et al. 2000). System 2 (blue) structures show sinistral offsets of these lithological layers, while system 3 structures (red) show dextral displacements. In Fig. 4, system 3 structures appear to dominantly cross-cut system 2 structures, but this is not always the case as in some areas the reverse is true (system 2 cross-cutting/displacing system 3). As there 193 is evidence for systems 2 and 3 mutually cross-cutting each other, and also because apparent movements are compat- ible with a conjugate system of strike-slip faults, it is pos- sible that they are contemporaneous. Cross-cutting all other systems are the NNE-trending system 4 lineaments, sug- gesting that they are likely to be the youngest structures, or at least have experienced the most recent movements. Other areas show a similar pattern of events, although some system 1 structures show evidence for younger move- ments (reactivation?), especially in areas around camps 2 and 4. System 5 is not represented in Fig. 4 as it was not observed at camp 1. This system is marked by quite wide (30–50 m) valleys, thus making its displacements diffi- cult to determine; however, as it is quite pronounced, it may be a more recent system. Field observations Four key areas were chosen for detailed fracture and fault analysis in the field (Fig. 3), based on their structural in- terest (i.e. their potential to enable all systems to be ana- lysed) and accessibility. The first objective of the field work was identification of the lineaments picked from the aer- ial photographs. In most cases field observations proved that the lineaments correspond to major fault structures, many of which are weathered out to leave gorges and river valleys (Fig. 5A–C). However, whilst many of the ENE– WSW-oriented system 1 structures are faults, others also correspond to basement fabric features, such as strongly foliated zones, lithological contacts and shear zones (Fig. 5D). Therefore care must be taken in any quantitative geometric or spatial analysis of this system. After a regional reconnaissance from each field camp, detailed structural analysis was carried out. Ninety out- crop locations were investigated in the four camps, dis- tributed along the coast between Nordre Isortoq and just north of Nordre Strømfjord (Fig. 2). In total c. 1700 faults and fractures were measured and described. Fault geometries A wide range of fracture orientations were observed (Fig. 6A), with dominantly N–S and NNW–SSE strikes and an overall mean fracture plane of 167/89E. Various fault orientations can be separated out in the field (dominant trends are N–S and ENE–WSW), showing a range of slip movements and shear senses (Fig. 6B–F). Dominant fault movements appear to be strike-slip (71% of faults recor- ded show strike-slip movements), although extensional and compressional faults were also apparent. • Basement-parallel to subparallel • Multiple phases of movement • Closely spaced (100–500 m) • Displacements range between 0.2 and 30 m for individual faults • May show an en échelon to irregular trend • Dominant fracture/joint trend is associated with this system • Closely spaced (50–100 m) • Marble layers show displacements in the order of 20–40 m • Major subvertical faults and fault zones • Generally associated with wide (> 50 m) valleys • Exposed fault cores show complex fracture sets associated with strike-slip movements • Localised to the fold belt south of Nordre Strømfjord and north of the Nordre Isortoq shear zone • Prominent structures at regional scale (i.e. from Landsat and aerial photos) but not at outcrop • Spatially associated with compressional faults (i.e. system 1 reverse faults) • Normal (dip-slip) • Reverse (dextral oblique-slip) • Dextral and sinistral strike-slip • Sinistral strike-slip • Normal (dip-slip) • Dextral strike-slip • Normal (dip-slip and oblique-slip) • Sinistral strike-slip • Dextral strike-slip System 1 System 2 System 3 System 4 System 5 ENE–WSW N–S NNW–SSE NNE–SSW E–W to ESE–WNW Comments Sense of movementLineament system Orientation Table 1. Main characteristics for each fault system, identified from remote sensing and outcrop studies 194 A set of ENE–WSW-oriented faults (green planes/mean poles in Fig. 6) appear to reactivate strong basement fab- rics in the Nordre Strømfjord and Nordre Isortoq shear zones. These faults correspond to system 1 lineaments and exhibit various forms of fault movement (e.g. extension- al, compressional and strike-slip; Fig. 6B–F). Faults cor- responding to systems 2 (N–S, blue), 3 (NNW–SSE, red) and 4 (NNE–SSW, yellow) can also be easily distinguished from the fault data in Fig. 6. However, lineament system 5 (E–W to ESE–WNW, purple) is not apparent. As pre- viously discussed this system appears to be a more geo- graphically localised system (i.e. local to areas south of the Nordre Strømfjord and Nordre Isortoq shear zones), and correspond to zones dominated by reverse fault move- ments (Fig. 6B) and a small number of ESE-trending dex- tral strike-slip faults (Fig. 6E). Fault systems corresponding to lineament systems 2, 3 and 4 appear to consist of parallel fault zones separated by non-faulted, but generally strongly fractured rock. The fault zones range typically between 1 and 50 m in width (Fig. 5B) and consist of multiple parallel faults with vari- able spacing. These zones are commonly located in pro- nounced valleys and gorges (Fig. 5A), so characterisation of fault planes was often difficult as the valley floors are generally covered by recent sediment and vegetation. BB Lunate fracture indicatingLunate fracture indicating dextral strike-slip movements dextral strike-slip movements RM structuresRM structures indicating dextralindicating dextral strike-slip strike-slip WW E A CC Raised beachRaised beach D System 1 System 1 System 4System 4 Sinistral strike-slip movementsSinistral strike-slip movements (inferred from ramp-step geometry) (inferred from ramp-step geometry) Steeply dippingSteeply dipping marble unit marble unit NNENNESSWSSW Raised beach System 1 System 4 Steeply dipping marble unit NNE N S SSW NNESSW EW A C D B Lunate fracture indicating dextral strike-slip movements Sinistral strike-slip movements (inferred from ramp-step geometry) RM structures indicating dextral strike-slip Fig. 5. Field identification of lineaments picked from Landsat and aerial photographs. A: Two major gorges/valleys trending ENE and NNE near camp 1; these correspond to major lineament systems 1 and 4, respectively. Fault core exposed within the NNE-trending stream bed (system 4) shows evidence for sinistral strike-slip faulting (Fig. 5B); fault movements on the ENE-trending basement-parallel valley (system 1) were not identified. B: Photograph of subhorizontal, sinistral strike-slip slickenlines observed within the fault core of the NNE-trending fault (system 4) identified in Fig. 5A. C: Some ENE-trending (system 1) lineaments correspond to basement fabrics such as steeply dipping (and tightly folded) marble units. D: Other basement parallel lineaments, however, do show evidence for brittle fault movement, as identified in this ENE-trending fault core (fault movement criteria defined by secondary fracture indicators, i.e. RM and lunate fractures, outlined in Petit 1987). 195 Fault kinematics In addition to the characterisation of the faults and frac- tures in terms of their trend and distribution, they can also be described according to their movement patterns (see Table 1). The nature and timing of tectonic events that are responsible for the formation of these fault-fracture systems is quite complex. Multiple directions of slicken- lines on several fault surfaces indicate that many faults were either reactivated or that individual faults exhibit curved movement trajectories consistent with complex strain histories. Strike-slip faulting Strike-slip slickenlines account for 71% of those observed and were observed on all main fault geometries or sys- tems. Multiple orientations of strike-slip faulting are com- mon in wrench-dominated fault systems due to the devel- opment of Riedel, P and X shears (e.g. Woodcock & Schu- bert 1994). Basement parallel faults (system 1, green) show both dextral and sinistral movements (note, RM structures – Petit 1987 – associated with R-shears suggest dextral move- ments on basement faults in Fig. 5D). NNW-trending (system 3, red) faults appear to correspond to dextral movements. N-trending faults (system 2, blue) appear to Poles to planes Slickenlines A B C FED N N N NNN Fractures n = 983 Mean fracture orientation Reverse faults n = 80 Normal faults n = 141 Strike-slip faults, n = 339 (shear sense uncertain or unassigned) Dextral faults n = 96 Sinistral faults n = 107 E +2S +4S +6S +8S +10S +12S +14S +16S +18S +20S +22S +24S E +2S +4S +6S +8S +10S +12S +14S E +2S +4S +6S E +2S +4S +6S +8S 10S E +2S +4S +6S +8S E +2S +4S +6S Fig. 6. Lower hemisphere, equal area stereographic projections of all fault and fracture data collected at camps 1, 2 and 4 (total number of measurements = 1746). A: Fractures (i.e. planes showing no evidence for slip). B: Reverse faults. C: Normal faults. D: Strike-slip faults with undetermined sense of movement. E: Dextral strike-slip faults. F: Sinistral strike-slip faults. Black dots, poles to planes of fault and fracture surfaces. Red dots, slickenlines. Mean fault planes are also shown, coloured according to lineament/fault systems identified in Fig. 3. Poles to fault and fracture surfaces are contoured using a Gaussian weighting function; n, number of measurements for each plot. In the labels, E corresponds to the background value (calculated as number of points/100), while S = standard deviations above this value. 196 AA BBB CCC AA 1st order fractures Normal faults Associated slickenlines 1st order fractures 2nd order fractures Strike-slip faults (sinistral) WSW ENE 2 m2 m2 m 0.5 m0.5 m Pseudotachylite En échelon footwall joints (subvertical dip) A 0.5 m ND n = 56 Fig. 7. A: Panoramic photograph showing exposures of a series of parallel ENE-dipping extensional faults, in the vicinity of camp 1. B and C: close-up photographs showing en échelon fracturing on the footwall of normal faults and pseudotachylite fault exposure in more detail. D: Lower hemisphere, equal area stereographic projection of poles to planes, and associated slickenlines, for faults and fractures observed at the outcrop of Fig. 7A. Fault orientations and kinematics suggest ENE–WSW extension as indicated by stress arrows (red); n, number of measurements. 197 show both dextral and sinistral movements, while NNE- trending faults (system 4, yellow) are sinistral structures (Fig. 6D–F). Extensional faulting Although all five fault systems show dominantly strike- slip movements, normal and oblique-slip components of displacement were also recorded on some sets. These faults appear to have two dominant orientations, N–S (system 2) and ENE–WSW (system 1; Fig. 6C). In some areas NNW–SSE (system 3) -oriented structures also appear to be normal faults (Fig. 7), but these are not the dominant orientations in bulk analyses (Fig. 6). Field observations suggest that strike-slip movements post-date dip-slip. Compressional faulting Reverse faults (Figs 6B, 8) appear to be confined to areas close to camps 2 and 4, and to be spatially associated with system 5 lineaments. These faults strike parallel or sub- parallel to basement structures (ENE–WSW to E–W, Fig. 6B) and exhibit dextral-oblique slickenlines, which plunge towards the ESE (Figs 6B, 8). As these compressional or thrust faults are only found on the southern shore of Nordre Strømfjord and the north- ern shore of Nordre Isortoq (i.e. abutting against major basement shear zones) it is possible that these structures are the result of local transpressional thrust faulting, which may be linked to steps in the en échelon sinistral fault system (system 4). Field observations suggest these com- pressional faults post-date most other fault and fractures. However, at camp 4, a N–S sinistral strike-slip fault ap- pears to cross-cut these thrusts (Fig. 8B). Reverse faults Sinistral strike-slip Dextral strike-slip Strike-slip fault (move- ment undetermined) n = 91 N Reverse faults Basement-parallel dextral faults Mean slickenlines orientation on reverse faults 1 m 1 m ENE ENE AA C WSW WSW 48-120 078 / 50 S 078 / 50 S 1 m N S WSW B A 078 / 50 S 078 / 50 S 48-120 48-120 C ENE Fig. 8. A: Exposure of a localised set of reverse faults near camp 4. B: Lower hemisphere, equal area stereographic projections of poles to structures observed at the locality of Fig. 8A. Three dominant fault sets are apparent: basement parallel – i.e. ENE-trending – reverse and dextral strike-slip faults, and a set of sinistral faults oriented N–S (e.g. along the large rock face in shadow). Slickenline orientations and relative fault movements suggest ESE–WNW compression as indicated by stress arrows (red); n, number of measurements. C: Photograph of surface of reverse fault, showing dextral-oblique slickenlines (mean slickenline orientation 45/120, see stereonet). Note that faults coloured in red here highlight thrust faults and do not refer to system 3 faults as in other figures. 198 Fractures and joints A diverse array of fracture orientations was recorded. Dom- inant orientations vary from NW–SE through to NNE– SSW. The orientation of the mean plane is NNW–SSE. All fractures recorded showed no evidence for shear move- ment (i.e. rough surfaces and with no apparent offsets) and are thus interpreted as opening mode 1 fractures and suggesting extension directions varied from c. E–W to NE– SW. Interpretation and discussion An overall summary of each of the fault systems identi- fied through remote sensing (i.e. lineament mapping) and field studies is presented in Table 1. In this section we discuss the possible interpretations and implications of these observations. Fault development System 1 (ENE–WSW) faults and fractures appear to be the oldest structures, however multiple slip vectors and apparent fault movements suggest that there has been ac- tivity on this system during later tectonic episodes (note that system 1 faults are apparent in all stereoplots for all fault types, Fig. 6B–F). These lie parallel to the pre-exist- ing Nagssugtoqidian basement fabric, which dates at c. 1.8 Ga (van Gool et al. 2002). Cross-cutting relationships interpreted from analysis of aerial photographs suggest that the next systems to devel- op were systems 2 and 3 (Fig. 5). It is difficult to deter- mine if one of these systems predates the other as mutual- ly cross-cutting relationships can be seen; however, it does appear that system 3 is the more pervasive system and thus may be more recent. Strike-slip movements and offsets on systems 2 (N–S) and 3 (NNW–SSE) suggest that, if active at the same time, these would represent a strike-slip conjugate system. In such cases the inferred extension vector would trend ENE– WSW, subparallel to system 1 foliation-parallel faults. This extension vector is also consistent with the dip-slip fault movements seen locally on these same fault systems. These strike-slip movements appear to be preceded by dip-slip extensional movements. System 2 (N–S) is the dominant extensional fault orientation in the area (Fig. 6C), while some localities showed small populations where NNW–SSE-oriented extensional faults represent the pre- ferred trend (e.g. Fig. 7). These faults are indicative of E– W to ENE–WSW extension. This extension cannot, how- ever, explain the apparent basement-parallel (system 1) extensional faults, which suggest an apparent NNW–SSE extension. These ENE-trending normal faults have also been observed in seismic interpretations and are thus im- portant structures regionally. These two extensional fault sets show a quadrimodal fault distribution, i.e. four sets of fault planes (Fig. 6C). If regarded as two separate fault sets, this geometry would suggest two separate extension directions (E–W and NNW–SSE); however an alternative to this is that these faults formed contemporaneously under three-dimensional strain (Reches 1983; Nieto-Samaniego & Alaniz-Alvarez 1997). As one set of normal faults trends parallel to a pre- existing plane of weakness (e.g. basement fabric) it is like- ly that basement reactivation played a role in the develop- ment of these faults, and that this has lead to formation of extensional faults oblique to the regional extension. The regional tectonic setting would fit with 3-D strain as the area borders the transfer zone between two extensional basins, i.e. is a transtensional deformation zone (Dewey 2002; De Paola et al. 2005). Geoffroy et al. (1998) recorded similar fault geometries and kinematics farther north in Disko and Nuussuaq (Fig. 1). Their interpretation is that strike-slip and dip-slip fault- ing developed during a single tectonic episode of WSW– ENE extension, which is in agreement with a model of 3-D transtensional strain. As the stereoplots in Fig. 6 show, strike-slip faults are the dominant fault type in the area (as mentioned above, 71% of all faults measures are strike-slip) and these ap- pear to post-date extensional movements. All fault sys- tems show evidence for strike-slip movements. System 1 (ENE–WSW) exhibits both dextral and sinistral senses of shear, systems 3 (NNW–SSE) and 5 (ESE–WNW) show dextral shear, while systems 2 (N–S) and 4 (NNE–SSW) are dominated by sinistral shear movements. System 4 faults (NNE–SSW) appear to cross-cut all other fault sets, and are characterised by major fault zones (Fig. 5). These major sinistral strike-slip structures lie subparallel to the sinistral Ikermiut and Ungava fault zones that dominate the Davis Strait offshore (Fig. 1). Assuming a ~NNE- trending sinistral wrench system for the study area, strike- slip fault movements on each system appear to correlate closely with synthetic (R) and antithetic (R’) Reidel shears, and also with synthetic P and antithetic X shears typical of a plane strain wrench tectonic regime (Fig. 9; Wood- cock & Schubert 1994). Compressional faults appear to be relatively late struc- tures (although cross-cut by ~N–S-trending sinistral faults) and are localised in areas of strong basement fabric (i.e. 199 shear zones). These faults strike parallel to basement fab- rics and indicate an oblique compression (from the ESE or SE; Fig. 8). Offshore there is evidence for thrusting in a similar orientation along the Ikermiut fault zone (Fig. 1). Positive flower structures have been identified (Chalm- ers & Pulvertaft 2001) and are interpreted as inversion structures formed at a restraining bend during sinistral strike-slip along the Ungava transform fault, during the early Eocene (c. 54–49 Ma, Chalmers & Pulvertaft 2001). If trends of basement shear zones (outlined in Fig. 2) are continued along strike offshore they appear to coincide with these transpression zones within the Ikermiut fault zone. It is possible that thrusts observed onshore have formed in a similar way to those offshore with basement shear zones acting as restraining bend structures, thus leading to localised compressional zones. Furthermore, slicken- lines on the reverse faults suggest a compression from the ESE or SE (Figs 6B, 8), which is consistent with the com- pressional axis for a sinistral wrench system (i.e. NE–SW extension and NW–SE compression; Fig. 9). Regional comparison and implications A key prerequisite for building tectono-stratigraphic mod- els is being able to date each tectonic event. As all onshore exposures in this part of West Greenland are in Precam- brian basement rocks, there are no stratigraphic markers for constraining the timing of Phanerozoic tectonic events. Relative timing has been inferred from various cross-cut- ting relationships in the field and from lineament analy- sis, but is open to different interpretations. In the absence of age data that constrain the absolute age(s) of fault acti- vity, comparisons with offshore models and with data col- lected in other onshore areas are used here to infer ages for events in our tectonic model (see Table 2 for a summary). Regional onshore correlations Farther north in the region of Disko and Nuussuaq, onshore faulting episodes can be dated relative to the deposition of basaltic lavas and the sedimentary systems during Cam- panian to Eocene times (Geoffroy et al. 1998; Storey et al. 1998; Chalmers et al. 1999; Dam et al. 2000). Dam & Sønderholm (1998), Dam et al. (2000) and Dam (2002) document at least three phases of faulting recorded in the sedimentary record prior to Paleocene volcanism. Creta- ceous–Paleocene sediments on Nuussuaq show distinctun- conformities, with incised valleys and submarine canyons, reflecting disturbances in early Campanian, Maastricht- ian and early Paleocene times. These unconformities and channels are thought to have formed in response to struc- tural movements associated with regional NE–SW rifting (and also the arrival of the North Atlantic plume in the latter two cases). 40Ar/39Ar dating has revealed that volcanism com- menced in West Greenland between 60.9 and 61.3 Ma and that 80% of the Paleocene lava pile was erupted in less than 1 Ma (Storey et al. 1998). These lavas show a distinct coastal flexure (Geoffroy et al. 1998, 2001; Larsen & Pulvertaft 2000), presently expressed by seaward dip- ping basalt lavas. This flexure has an arcuate course, strik- ing NW–SE in southern Svartenhuk Halvø and northern Ubekendt Ejland, turning through N–S in south-west Ubekendt Ejland to NE–SW in north-west Nuussuaq and finally to N–S in north-west Disko (Fig. 1; Geoffroy et al. 1998, 2001; Larsen & Pulvertaft 2000). Numerous dykes cut these lavas (e.g. see figs 4 and 5 in Larsen & Pulvertaft 2000), and so do various fault sets (Geoffroy et al. 1998). The timing of the various phases of volcanic eruption, dyke emplacement, and block faulting relative to one another is still a matter of debate. Geoffroy et al. (1998) presented detailed structural evidence suggesting that fault and dyke intrusion on Disko took place during tilting of lava sys- R` R P X X R P R` A B NNE–SSW (10–15°) wrench system: 55–60° extension, 145–150°compression Sy st em 4 System 1 System 5 System 3 System 2 NFig. 9. A: Diagram showing fault systems and their corresponding movements. B: Strain ellipse for a NNE–SSW (~010–190°) -oriented sinistral wrench system, showing Riedel (R and R’), P and X shears (Wood- cock & Schubert 1994). Also shown are the regional stress vectors (σσσσσ1 and σσσσσ3). Systems 2 (sinistral) and 5 (dextral) correspond to R’ and R shears, while systems 1 (dextral) and 4 (sinistral) correspond to P and X shears. System 3 corresponds to normal fault sets in Fig. 9B; however, dominant movements on this system were dextral. 200 tems. Geoffroy et al. (2001) then stated that NW–SE- oriented, flexure-parallel dykes in southern Svartenhuk yield dates of around 54.6 ± 0.6 Ma. This suggests that most of the coastal flexure is of Eocene age (or later?). Further evidence for this comes from north-west Nuus- suaq where tilted lavas (Larsen & Pulvertaft 2000) have been dated at ~53 Ma (Storey et al. 1998). Systems 2 (N–S) and 3 (NNW–SSE) extensional faults in this study appear comparable to faults that cross-cut these Paleocene basalt lavas (Geoffroy et al. 1998). Taking the dates outlined above for dyke emplacement which is believed to be associated with faulting, it would appear that our system 2 and 3 faults were active during Eocene times. Taking all these onshore tectonic timings into account it would then appear that the faults observed may have been active from late Cretaceous (Maastrichtian) through to Eocene times (Dam & Sønderholm 1998; Geoffroy et al. 1998, 2001; Chalmers et al. 1999; Dam et al. 2000; Larsen & Pulvertaft 2000; Dam 2002). Normal fault orientations similar to systems 2 (N–S) and 3 (NNW–SSE) occur in and around the Nuussuaq basin (Fig. 1; Geoffroy et al. 1998; Chalmers et al. 1999) and are consistent with either ENE–WSW (Geoffroy et al. 1998) or E–W (Chalmers et al. 1999) extension. Chalmers et al. (1999) proposed that the N–S faults formed by E–W-oriented crustal extension, while associ- ated ESE–WNW faults formed as a consequence of reac- tivation of shear zones in the underlying basement. Nor- mal faults are dominantly N–S onshore (Fig. 6C), fitting with this model proposed by Chalmers et al. (1999). How- ever, this implies that while Baffin Bay in the north and Labrador Sea in the south were undergoing ENE–WSW extension (deduced from dominant fault trends and earli- est magnetic anomaly trends, Chalmers & Pulvertaft 2001), southern West Greenland and the Nuussuaq basin were undergoing E–W extension. A better explanation is that the Davis Strait at this time (i.e. prior to the onset of sea- floor spreading) lay in a transfer/step-over zone between two extensional basins, and that it was strongly influenced by basement fabrics such that this region experienced com- plex 3-D strain associated with regional ENE–WSW ex- tension (Fig. 10). Onshore normal fault sets form a quad- rimodal fault distribution (four sets of fault planes, Fig. 6C) consistent with 3-D strain. This faulting is then subsequently dissected by N–S (system 2) and NNE–SSW (system 4) -oriented faults during the Eocene (Chalmers et al. 1999). The Itilli fault zone (Fig. 1) is one such NNE–SSW-oriented structure cutting through north-west Nuussuaq. This fault zone appears to be a left-lateral splay from the northern exten- sion of the Ungava fault zone in the Davis Strait (Chal- mers et al. 1999). • Uplift? • Possible reactivation of systems 1–4 as normal faults • Faults consistent with NNE-oriented sinistral wrench system: NNE-trending (system 4) faults and basement-parallel (system 1) faults active as antithetic X and synthetic P shears • Systems 2 (reactivation) and 5 active as Reidel shears • Local transpressional thrust faulting observed near camps 2 and 4, associated with steep basement fabrics • Systems 1, 2 (and 3?) faults all active as extensional faults during NE–SW to ENE–WSW extension (3D strain)? • Uncertain? • Possible system 1 ENE–WSW foliation-parallel faulting prior to late Mesozoic? • Subsidence • N- to NNE-trending sinistral transverse fault system (Ungava fault zone). • Local transpressional and transtensional faulting (e.g. Ikermiut fault zone) • N-trending faults in Davis Strait (E–W extension) leading to formation of the Sisimiut Basin offshore • NW-trending normal faults in the Labrador Sea (SW–NE extension) • WSW–ENE faults bordering the Sisimiut basin to the south were active offshore during events 2 and/or 3 • Uncertain? 5 (Youngest) 4 3 2 1 (Oldest) Pliocene to Pleistocene tilting Eocene Labrador sea-floor spreading (Ungava system) Late Cretaceous to early Paleocene extension Early or middle to late Cretaceous extension and thermal subsidence Proterozoic and later localised reactivation Onshore tectonic structures Offshore tectonic structuresEvent # Timing and event Table 2. Proposed event stratigraphic model and apparent correlation with offshore events 201 Fig. 10. Proposed two-stage tectonic model for the tectonic evolution of upper Mesozoic – Cenozoic extension within the Nagssugtoqidian orogen. Stage 1(A): N- and ENE-trending normal faults and dextral basement reactivation due to ENE–WSW extension. Stage 2 (B): N- and NNE-trending sinistral strike-slip faulting, and associated strike-slip wrench tectonic systems, with compressional structures (reverse faults) forming in zones of basement anisotropy (e.g. shear zones). Block diagrams show schematic cartoons outlining fault patterns observed onshore, while maps show the regional context, based on correlations between onshore and offshore fault structures. Labrador Sea Baffin Bay Possible oceanic crust Ikermiut fault zone (Pop-up flower structure) N 200 km B NNE-trending sinistral wrench system Labrador Sea Ba se m en t a nis otro py Ba se men t a niso tro py Base ment an isotropy Base ment an isotropy Ba se m en t a ni so tr op y N Nordre Strømfjord Nordre Isortoq SisimiuteSisimiute BasinBasin 3D strain zone: Quadrimodal fault patterns (strain partitioning?) N 200 km U ng av a F Z U ng av a F Z 2D strain zone: Simple Andersonian faulting Baffin Bay ? Nagssugtoqidian orogen D av is St ra it STAGE 1: Late Cretaceous–Paleocene extension (NE–SW to ENE–SSW). Pre-existing basement anisotropy (Nagssugtoqidian orogenic belt) appears to influence fault patterns in the Davis Strait. Possible strain partitioning between basement- parallel and N–S-trending normal faults. STAGE 2: NNE-trending Eocene sinistral wrench system (NE to ENE extension, SE to SSE compression) reactivating earlier extensional faults and basement-parallel structures (e.g. thrust faults at camp 4, and the Ikermiut fault zone). N Nordre Strømfjord Nordre Isortoq A B A N–S normal faults Basement- parallel faults Fractures 202 Correlation with fault structures offshore Fault patterns offshore in the Davis Strait exhibit similar c. N–S and ENE–WSW dominant orientations. The south and east bounding faults to the Sisimiut basin strike ENE– WSW and N–S respectively (Fig. 1). Significantly, the southern margin of the Sisimiut basin is coincident both in orientation and location with a major basement shear zone (the Nordre Isortoq shear zone), and is likely to have exerted a similar structural control to that interpreted for onshore. Block faulting has been dated via drilling as hav- ing taken place between the late Campanian and late Pale- ocene (Christiansen et al. 2001; Dalhoff et al. 2003), indi- cating that these extensional faults are of similar age to those associated with valley incision on Nuussuaq and north Disko (Dam & Sønderholm 1998; Dam et al. 2000; Dam 2002). The Ungava fault zone with its associated fault systems (e.g. the Ikermiut fault zone) is the most prominent struc- ture in the Davis Strait (Chalmers & Pulvertaft 2001). This NNE–SSW-oriented structure is interpreted as a transform fault zone showing sinistral shear, and has been linked to the Itilli fault zone (Fig. 1; Chalmers et al. 1999). System 4 (NNE–SSW) faulting onshore, around Nordre Strømfjord, is consistent with late sinistral strike-slip move- ments, and it is reasonable to suggest that this system is of similar age. As already discussed, offshore evidence for sini- stral strike-slip movements can be seen in the Ikermiut fault zone on the western margin of the Sisimiut basin (Chal- mers & Pulvertaft 2001) where transpressional thrusts (similar to compressional flower structures modelled in Dooley et al. 1999) appear to have formed in the restrain- ing bend of a strike-slip fault (see fig. 6 of Chalmers & Pulvertaft 2001). These thrusts cut early Eocene mud- stones, but are overlain by late Eocene sediments, provid- ing further evidence for timing of these movements. This sinistral shear is thus a consequence of left-lateral move- ment of the Canadian plate relative to the Greenland plate along the Ungava transform system during sea-floor spreading in the Labrador Sea (Fig. 1). Evidence for neotectonic faulting? Chalmers (2000) presented evidence for Neogene uplift in offshore areas of central West Greenland, while recent onshore topographic and apatite fission-track data analy- sis has identified similar Neogene activity (Japsen et al. 2002, 2005). A common observation in this field area is the presence of raised beaches and palaeoshorelines, up to elevations 30 m above present sea levels (Fig. 5A). They are probably the result of isostatic readjustment following the removal of Pleistocene ice load. Topographically, the region appears to be divided into blocks, split by ENE- trending fjords and NNE-trending escarpments (Fig. 2C). These must be quite recent features as they have not been eroded during glacial activity (and may in fact be a conse- quence of it). The trend of system 4 faults (and also local- ly those of system 2) is generally parallel to the pronounced NNE-trending escarpment from Nordre Isortoq to Nor- dre Strømfjord onshore, and also to a similarly trending scarp near offshore (identified in bathymetry maps), and it is possible that these faults have been reactivated as nor- mal faults during a recent tectonic event. This conjecture still needs to be verified, as the main escarpments were not studied in detail during our field work. Summary The observed fault and fracture systems reflect a brittle tectonic history that is ultimately related to far-field plate movements, uplift and basin formation. The development of Mesozoic to Cenozoic basins offshore West Greenland appears to be strongly controlled by faults. Therefore, knowledge of the fracture systems in the exposed Precam- brian basement provides a valuable insight into fault ge- ometries and kinematics during the development of off- shore basins and potential hydrocarbon reservoirs. It also provides insights into the possible influence of basement reactivation. Several possible tectonic-event models may be con- structed for this region given the lack of definite ages for structures. Table 2 shows a basic summary of the relative timings of fault systems identified in this study relative to regional offshore tectonic models, while Fig. 10 presents a model for fault development based on the observations and correlations made in this study. The absolute timing of the fault activity onshore, as deduced from correlation to other fault systems with known ages, needs to be tested by dating of fault rock samples. A simple two-stage model has been outlined to explain the complex fault patterns exhibited in onshore exposures of the central Nagssugto- qidian orogen (Fig. 10). The brittle tectonic evolution of the region appears to be dominated by NE–SW extension, which is consistent with the opening of the Labrador Sea and Baffin Bay. Only slight variations in the regional stress field are required to account for the diversity of fault orientations. According to Chalmers & Pulvertaft (2001) there was a 15° counter 203 clockwise rotation in spreading direction between the Pale- ocene and the Eocene in the Labrador Sea as opening start- ed between Greenland and Europe, which is consistent with the two-stage model outlined in Fig. 10. In the early stages of opening, faulting was dominated by extensional structures (under 3-D strain conditions), favouring an E– W to ENE–WSW extension (Fig. 10A); however, as the Ungava transform fault developed, faulting became more wrench dominated (2-D plane strain), and suggests NE– SW extension (Fig. 10B). Variations in fault geometry re- flect these changes in the regional stress field. However, the influence of basement structure also appears to have played an important role throughout (e.g. extensional faults not normal to the extension direction, and the ap- parent localised compressional zones associated with in- tense basement fabrics). Although most faults observed onshore trend highly obliquely to basement fabrics, fault patterns do appear to vary in areas of intense pre-existing structure (such as the Nordre Strømfjord and Nordre Iso- rtoq shear zones) which suggest that the fabrics within the Nagssugtoqidian orogen may have had some influ- ence on the fault complexity of the Davis Strait. The conclusions from this study show that the fault patterns and sense of movement on faults onshore reflect the stress fields that govern the opening of the Labrador Sea – Davis Strait – Baffin Bay seaway, and that the wrench couple on the Ungava transform system played a domi- nant role in the development of the onshore fault pat- terns. Acknowledgements The authors would like to thank BP (Norway) and Sta- toil (UK) for providing additional funding for this field research, and to NERC for funding R.W.W.’s Ph.D. re- search (NER/S/S/2001/06740). Thorough and insight- ful reviews from S. Bergh and T.C.R. Pulvertaft and addi- tional help from the latter concerning numerous details are greatly appreciated. The bulletin editors, A.A. Garde and F. Kalsbeek, are also thanked for their helpful and encouraging comments. References Bonow, J.M. 2004: Palaeosurfaces and palaeovalleys on North Atlantic previously glaciated passive margins – reference forms for conclu- sions on uplift and erosion. Ph.D. thesis. Thesis in Geography with Emphasis on Physical Geography 30, 17 pp. + 4 articles. Stockholm University, Sweden. Butler, R.W.H., Holdsworth, R.E. & Lloyd, G.E. 1997: The role of basement reactivation in continental deformation. Journal of the Geological Society (London) 154, 69–71. Chalmers, J.A. 2000: Offshore evidence for Neogene uplift in central West Greenland. Global and Planetary Change 24, 311–318. Chalmers, J.A. & Laursen, K.H. 1995: Labrador Sea: the extent of con- tinental crust and the timing of the start of sea floor spreading. Marine and Petroleum Geology 12, 205–217. Chalmers, J.A. & Pulvertaft, T.C.R. 2001: Development of the conti- nental margins of the Labrador Sea: a review. In: Wilson, R.C.L. et al. (eds): Non-volcanic rifting of continental margins: a comparison of evidence from land and sea. Geological Society Special Publica- tion (London) 187, 77–105. Chalmers, J.A., Dahl-Jensen, T., Bate, K.J. & Whittaker, R.C. 1995: Geology and petroleum prospectivity of the region offshore south- ern West Greenland. Rapport Grønlands Geologiske Undersøgelse 165, 13–21. Chalmers, J.A., Pulvertaft, T.C.R., Marcussen, C. & Pedersen, A.K. 1999: New insight into the structure of the Nuussuaq basin, central West Greenland. Marine and Petroleum Geology 16, 197–224. Chian, D. & Louden, K.E. 1994: The continent-ocean crustal transi- tion across the southwest Greenland margin. Journal of Geophysi- cal Research 99, 9117–9135. Christiansen, F.G., Bojesen-Koefoed, J.A. & Chalmers, J.A. 2001: Pe- troleum geological activities in West Greenland in 2000. Geology of Greenland Survey Bulletin 189, 24–33. Dalhoff, F., Chalmers, J.A., Gregersen, U., Nøhr-Hansen, H., Rasmus- sen, J.A. & Sheldon, E. 2003: Mapping and facies analysis of Pale- ocene – Mid-Eocene seismic sequences, offshore southern West Greenland. Marine and Petroleum Geology 20, 935–986. Dam, G. 2002: Sedimentology of magmatically and structurally con- trolled outburst valleys along rifted volcanic margins: examples from the Nuussuaq basin, West Greenland. Sedimentology 49, 505–532. Dam, G. & Sønderholm, M. 1998: Sedimentological evolution of a fault-controlled early Paleocene incised-valley system, Nuussuaq Basin, West Greenland. In: Shanley, K.W. & McCabe, P.J. (eds): Relative role of eustasy, climate, and tectonism in continental rocks. Society of Economic Paleontologists and Mineralogists Special Pub- lication 59, 109–121. Dam, G., Nøhr-Hansen, H., Pedersen, G.K. & Sønderholm, M. 2000: Sedimentary and structural evidence of a new early Campanian rift phase in the Nuussuaq Basin, West Greenland. Cretaceous Research 21, 127–154. De Paola, N., Holdsworth, R.E., McCaffrey, K.J.W. & Barchi, M.R. 2005: Partitioned transtension: an alternative to basin inversion models. Journal of Structural Geology 27, 607–625. Dewey, J.F. 2002: Transtension in arcs and orogens. International Ge- ology Review 44, 402–439. 204 Dooley, T., McClay, K. & Bonora, M. 1999: 4D evolution of segment- ed strike-slip fault systems: applications to NW Europe. In: Fleet, A.J. & Boldy, S.A.R. (eds): Petroleum geology of Northwest Eu- rope, Proceedings of the 5th Conference, 215–225. London: Geo- logical Society. Escher, J.C. & Pulvertaft, T.C.R. 1995: Geological map of Greenland, 1:2 500 000. Copenhagen: Geological Survey of Greenland. Geoffroy, J., Gélard, J.P., Lepvrier, C. & Olivier, P. 1998: The coastal flexure of Disko (West Greenland), onshore expression of the ‘ob- lique reflectors’. Journal of the Geological Society (London) 155, 463–473. Geoffroy, J. et al. 2001: Southeast Baffin volcanic margin and the North American – Greenland plate separation. Tectonics 20, 566–584. Henriksen, N., Higgins, A.K., Kalsbeek, F. & Pulvertaft, T.C.R. 2000: Greenland from Archaean to Quaternary. Descriptive text to the geological map of Greenland, 1:2 500 000. Geology of Greenland Survey Bulletin 185, 93 pp. Holdsworth, R.E., Butler, C.A. & Roberts, A.M. 1997: The recogni- tion of reactivation during continental deformation. Journal of the Geological Society (London) 154, 73–78. Japsen, P., Bonow, J., Klint, K.E.S. & Jensen, F.K. 2002: Neogene up- lift, erosion and resedimentation in West Greenland. Field report summer 2002. Danmarks og Grønlands Geologiske Undersøgelse Rapport 2002/71, 114 pp. Japsen, P., Green, P.F. & Chalmers, J.A. 2005: Separation of Palaeogene and Neogene uplift on Nuussuaq, West Greenland. Journal of the Geological Society (London) 162, 299–314. Larsen, J.G. & Pulvertaft, T.C.R. 2000: The structure of the Creta- ceous–Palaeogene sedimentary-volcanic area of Svartenhuk Halvø, central West Greenland. Geology of Greenland Survey Bulletin 188, 40 pp. Marker, M., Mengel, F., van Gool, J. and field party 1995: Evolution of the Palaeoproterozoic Nagssugtoqidian orogen: DLC investigations in West Greenland. Rapport Grønlands Geologiske Undersøgelse 165, 100–105. McClay, K.R. 1987: The mapping of geological structures (Geological Society of London handbook), 161 pp. London: John Wiley and Sons. Nieto-Samaniego, A.F. & Alaniz-Alvarez, S.A. 1997: Origin and tec- tonic interpretation of multiple fault patterns. Tectonophysics 279, 197–206. Petit, J.-P. 1987: Criteria for the sense of movement on fault surfaces in brittle rocks. Journal of Structural Geology 9, 597–608. Ramberg, H. 1949: On the petrogenesis of the gneiss complexes between Sukkertoppen and Christianshaab, West Greenland. Meddelelser fra Dansk Geologisk Forening 11, 312–327. Reches, Z. 1983: Faulting of rocks in three-dimensional strain fields. II. Theoretical analysis. Tectonophysics 95, 133–156. Storey, M., Duncan, R.A., Pedersen, A.K., Larsen, L.M. & Larsen, H.C. 1998: 40Ar/39Ar geochronology of the West Greenland Tertiary vol- canic province. Earth and Planetary Science Letters 160, 569–586. Taylor, B., Crook, K. & Sinton, J. 1994: Extensional transform zones and oblique spreading centres. Journal of Geophysical Research 99(B10), 19707–19718. van Gool, J.A.M., Connelly, J.N., Marker, M. & Mengel, F.C. 2002: The Nagssugtoqidian orogen of West Greenland: tectonic evolution and regional correlations from a West Greenland perspective. Cana- dian Journal of Earth Sciences 39, 665–686. Whittaker, R.C. 1995: A preliminary assessment of the structure, basin development and petroleum potential offshore central West Green- land. Open File Series Grønlands Geologiske Undersøgelse 95/9, 33 pp., 6 maps. Woodcock, N.H. & Schubert, C. 1994: Continental strike-slip tecton- ics. In: Hancock, P.L. (ed.): Continental deformation, 251–263. Oxford: Pergamon Press. __________________________________________________________________________________________________________________________________________________________________________________________________________________________ Manuscript received 26 October 2004; revision accepted 1 November 2005