Introduction

The opening of the North Atlantic Ocean between eastern Newfoundland and western Iberia–Europe started with continental rifting in the Late Triassic to Early Jurassic which led to continental break-up and the initiation of ocean crust in the Aptian–Albian (e.g. Welsink et al. 1989; Grant and McAlpine 1990; Bassi et al. 1993; Enachescu et al. 2005; Tucholke et al. 2007; Alves et al. 2009). Continental rifting was initially driven by an early NW–SE crustal extension between SW Newfoundland and Iberia in the Late Triassic–Early Jurassic and then propagated northward between NW Newfoundland and western Europe in the Middle Jurassic–Early Cretaceous (Fig. 1) (Tankard and Welsink 1989; Enachescu et al. 2005). Around the Aptian–Albian, continental break-up took place between Newfoundland and Iberia–Europe and involved major geodynamic changes in the evolution of the North Atlantic, as it led to the separation of the Grand Banks from Iberia, and the Flemish Cap from the Galicia Bank (Williams 1984; Tucholke et al. 2007). In the Late Cretaceous, extension propagated further north and initiated the opening of the Labrador Sea between Labrador and West Greenland (Chian et al. 1995; Chalmers et al. 1999; Enachescu 2006; Dickie et al. 2011).

Fig. 1
figure 1

Reconstruction of the central North Atlantic to Chron M0 (ca. 125 Ma) (modified from Tucholke and Sibuet 2012; Pe-Piper et al. 2013) showing the distribution of the major Early Cretaceous basins according to Welford et al. (2010)

The Late Triassic–Early Cretaceous crustal extension resulted in several major rift basins like the Porcupine, Rockall, and Hatton basins off the Irish margin (Fig. 1) and the Orphan Basin off the Newfoundland margin (Fig. 2) (Vogt et al. 1998; Mackenzie et al. 2002; Enachescu 2006; O’Reilly et al. 2006; Yang 2012; Dafoe et al. 2015; Lau et al. 2015). The Orphan Basin, which is the focus of this study, is a deep-water rift basin surrounded by the Newfoundland continental shelf in the west, the Grand Banks in the south, the Flemish Cap in the southeast, and the Orphan Knoll in the northeast (Fig. 2). Geophysical data from the basin show a highly thinned crust and a thick Mesozoic–Cenozoic sedimentary cover (e.g. Chian et al. 2001; Welford and Hall 2007; Lau et al. 2015). The East Orphan Basin appears to be older as it contains Jurassic syn-rift sediments, while the West Orphan Basin is younger as the first syn-rift sediments are assumed to be Lower Cretaceous (Enachescu et al. 2004, 2005). Despite the strong crustal thinning that is observed underneath the East and West Orphan depocentres, the Aptian–Albian continental break-up between the Newfoundland and the Irish margins occurred immediately east of the Orphan Knoll and the Flemish Cap (Figs. 1, 2) (Funck et al. 2003; Welford et al. 2012).

Fig. 2
figure 2

Bathymetric map of the Orphan Basin showing the seismic reflection profiles, plotted with solid lines, and the well data, plotted with white circles, considered in this work. A portion of the seismic profile C, highlighted in red, is displayed in detail in Fig. 11. Seismic line E is shown entirely and in detail in Figs. 6, 10, and 12. Seismic data courtesy of TGS. COB: interpreted continent–ocean boundary according to Funck et al. (2003)

Although several contributions have examined the large-scale structural and crustal architecture of the Orphan Basin (e.g. Enachescu et al. 2004, 2005; Welford et al. 2012; Watremez et al. 2015; Lau et al. 2015), only few tried to quantify its Mesozoic syn-rift history and the driving tectonic processes (Dafoe et al. 2015; Gouiza et al. 2015). In this contribution, (1) we examine the stratigraphic and tectonic evolution of the Orphan Basin using exploration wells and regional seismic lines, (2) we investigate its structural and crustal architectures to quantify amounts of crustal extension versus crustal thinning during rifting, and (3) we propose tectonic processes that can explain its tectono-stratigraphic evolution.

Data and methods

Well data

Several exploration wells, drilled offshore of the Newfoundland margin, are used in this work to characterize the Mesozoic sediments across the Orphan Basin (Figs. 2, 3). Four wells are located on the continental shelf (Hare Bay E-21, Baie Verte J-57, Bonavista C-99, and Sheridan J-87), two wells are found in the deep basin (Blue H-28 and Great Barasway F-66), and two wells in the Flemish Pass (Baccalieu I-78 and Gabriel C-60) (Figs. 2, 3). Other than the Great Barasway F-66 which was drilled on a roll-over structure (Gacal-Isler 2009), the wells in the Orphan Basin were drilled on structural highs characterized by a very reduced Mesozoic succession and, mostly, no syn-rift sediments (Fig. 3). We use the wells from the Flemish Pass Basin, which exhibit thicker Mesozoic series, to draw correlations into the deeper parts of the Orphan Basin.

Fig. 3
figure 3

Simplified stratigraphic logs of the exploration wells found in the Orphan Basin and used in this work. They show the main lithologies and the major stratigraphic unconformities correlated across the basin

2D seismic reflection data

Five deep marine multichannel seismic reflection lines (labelled A to E; Fig. 2) were acquired by TGS in the Orphan Basin in 2002. Lines A and B extend NW–SE over 617.9 and 677.54 km, respectively. Lines C and D run NE–SW over 593.07 and 369.36 km, respectively. Line E is oriented WNW–ESE and is 518.75 km long. All lines were recorded to a maximum two-way time (TWT) of 12 s and were provided to us (courtesy of TGS) in time-migrated SEG-Y format.

In this study, we focus on seismic line E, which was interpreted initially in the time domain and then converted to depth. We used the unconformities identified in the wells (Figs. 3, 4, 5) as regional markers to correlate the lithostratigraphic successions along the seismic lines. Five seismo-stratigraphic units were identified, namely the Cenozoic, the Upper Cretaceous, the Lower Cretaceous, the Jurassic, and the pre-Mesozoic basement (Figs. 5, 6). After stratigraphic interpretation was accomplished by picking the different horizons in the time domain, seismic line E was depth-converted using interval velocities. Stacking velocities along the profile (Fig. 7), provided by TGS, were converted to interval velocities using the Dix formula (Dix 1955) and then used to infer an interval velocity function for each seismo-lithological unit (Fig. 8).

Fig. 4
figure 4

Simplified lithostratigraphic chart of Orphan Basin based on available exploration wells (Figs. 2, 3)

Fig. 5
figure 5

The Baie Vert J-57, Bonavista C-99, Blue H-28, and Baccalieu I-78 wells tied and calibrated with the seismic data in time domain. Light green, purple, dark green, blue, and red horizons represent the base Cenozoic unconformity (CU), the Santonian unconformity (SU), the Aptian–Albian unconformity (AU), the Tithonian unconformity (TU), and the Basement top (BU), respectively. Location of well and seismic data is shown in Fig. 2. Seismic data courtesy of TGS

Fig. 6
figure 6

a Interpreted seismic line E in the time domain illustrating the stratigraphic and structural architecture of the Orphan Basin. Interpretation of the Moho discontinuity is plotted with a solid yellow line. bd Illustrating in more detail the relationship between the different seismo-stratigraphic units and their distribution within the half-graben structures

Fig. 7
figure 7

Grid showing variation of stacking velocity with time (TWT) along seismic line E. Stacking velocity data courtesy of TGS

Fig. 8
figure 8

Cross-plots of [interval velocity, depth] pairs for each seismo-lithostratigraphic unit calculated from stacking velocities (Fig. 7) using the Dix formula (Dix 1955)

$$V_{\text{int}} = \, \left[ {{{\left( {t_{2} V_{\text{RMS2}}^{2} - t_{1} V_{\text{RMS1}}^{2} } \right)} \mathord{\left/ {\vphantom {{\left( {t_{2} V_{\text{RMS2}}^{2} - t_{1} V_{\text{RMS1}}^{2} } \right)} {\left( {t_{2} - t_{1} } \right)}}} \right. \kern-0pt} {\left( {t_{2} - t_{1} } \right)}}} \right]^{1/2} \quad \left( {{\text{Dix}}\;{\text{formula}}} \right)$$

where V int is interval velocity; t 1 is traveltime to the first reflector; t 2 is traveltime to the second reflector; V RMS1 is root-mean-square velocity to the first reflector; and V RMS2 is root-mean-square velocity to the second reflector.

Uncertainties related to the time-to-depth conversion produce minor thickness errors, of no consequence to this analysis, in the depth-converted lines when calibrated with the wells (Fig. 9).

Fig. 9
figure 9

Four seismic–well ties in the depth domain showing the correlation between the lithological unconformities from the well logs (in black) and the depth of the interpreted seismic horizons (in pink) (seismic data courtesy of TGS). J-57: Baie Verte well; C-99: Bonavista well; H-28: Blue well; I-78: Baccalieu well

Mesozoic sedimentary architecture in the Orphan Basin from well and 2D seismic reflection data

Seismo-stratigraphic units

Well data from the Orphan Basin, when correlated with the seismic reflection profiles, reveal the existence of four stratigraphic unconformities in the syn-rift and early post-rift sequences (Fig. 3). These are the Tithonian unconformity, the Aptian–Albian unconformity, the Santonian unconformity, and the base Cenozoic unconformity (Enachescu et al. 2005; Dafoe et al. 2015). Most of these unconformities form prominent seismic horizons that were picked and correlated as regional markers along the seismic profiles, except the Santonian unconformity which was only identified in the Baie Verte J-57 well. According to Dafoe et al. (2015), the Santonian unconformity is only recognized in proximal parts of the West Orphan Basin wells (e.g. Baie Verte J-57) and appears to change to paraconformity in the East Orphan Basin. Therefore, five major seismo-stratigraphic units were identified in the Orphan Basin defined by the major depositional and erosional unconformities and distinct seismic facies (Figs. 5, 6, 9):

  • The Cenozoic succession, characterized in the well logs by the Cenozoic unconformity at its base, is extremely thick on the continental shelf (3000–4000 m), but thins towards the east in the deep basin (1500 m) and towards the southeast in the Flemish Pass (600–1400 m). The Cenozoic succession is formed by the Banquereau Formation and consists mainly of shales, mudstones, and thin intercalations of carbonates. On seismic data, this unit is characterized by continuous high-amplitude reflections alternating with very weak amplitude reflections (Figs. 5, 6).

  • The Upper Cretaceous succession is defined by the Cenozoic unconformity at the top and the Aptian–Albian unconformity at the base. The latter unconformity was initially interpreted as base Cenomanian (Enachescu et al. 2005), but new palynological analyses yielded a more reliable age at the Aptian–Albian boundary (Dafoe et al. 2015). The Upper Cretaceous succession is drilled in the continental shelf (650 m thick in Baie Verte J-57) and the deep Orphan Basin where it becomes significantly thin towards the east (10 m thick in Great Barasway F-66). It mainly consists of the Cenomanian–Campanian Dawson Canyon Formation and contains predominantly medium- to coarse-grained siliciclastic sediments (siltstones and sandstones), mudstones, and marls in the south (Sheridan J-87). The Upper Cretaceous unit shows two different seismic signatures on seismic profiles. It is mainly characterized by continuous low-amplitude and low-frequency reflections in the eastern part of the basin and overlain by continuous high-amplitude and high-frequency reflections in the western part of the basin (Figs. 5, 6). This sequence roughly corresponds to the breakup sequence recognized in Northwest Iberia by Soares et al. (2012).

  • The Lower Cretaceous succession is defined by an Aptian–Albian unconformity at the top and the Tithonian unconformity at the base. It is found only in the East Orphan Basin (Great Barasway F-66) and the Flemish Pass wells (Baccalieu I-78 and Gabriel C-60). It includes mainly the Nautilus Formation and the Whiterose Formation, formed by shale-dominated series with occasional siltstone and sandstone interbeds. This unit has a recognizable seismic signature along the profiles consisting of discontinuous reflections with medium to low amplitudes and high to medium frequencies (Figs. 5, 6).

  • The Jurassic succession, capped by the Tithonian unconformity, occurs only in the Great Barasway F-66 and the Baccalieu I-78 wells in the East Orphan Basin and Flemish Pass, respectively. It includes the Tithonian Jeanne d’Arc Formation and the Callovian–Kimmeridgian Rankin Formation, which consist of shales, siltstones, sandstones, and minor limestone interbeds. The base of this succession was never drilled in this area, and the age of the first Mesozoic sediments deposited in the Orphan Basin, in relation to the Atlantic rifting, has yet to be determined. However, well data in the Jeanne d’Arc Basin to the south indicate the presence of Upper Triassic clastics lying unconformably over the Palaeozoic basement (McAlpine 1990). This configuration is similar to West Iberia (Alves et al. 2003), West France (Roberts and Bally 2012), and West Ireland (Murphy and Ainsworth 1991). In this work, the Jurassic succession includes the sediments between the Tithonian unconformity and the Palaeozoic basement, which may or may not include Upper Triassic deposits. The Jurassic unit is very thin and scattered in the western part of the basin where it is represented by discontinuous high-amplitude reflections. It is much thicker in the eastern part of the basin and characterized by medium- to very-low-amplitude reflections (Figs. 5, 6). The major uncertainty in our seismic interpretation is the Jurassic deposits drilled only in the East Orphan Basin and for which there is no direct evidence in the west.

  • The oldest succession, encountered in wells drilled on structural highs, is the pre-Mesozoic basement probably of Palaeozoic to Neoproterozoic age. It comprises low-grade metamorphosed sediments and igneous granitic rocks. The basement is characterized in seismic data by chaotic and discontinuous reflections with medium to high amplitudes (Figs. 5, 6).

Structural and stratigraphic architecture imaged by the 2D seismic data

Seismic line E (Figs. 6, 10) reveals that rifting and crustal stretching in the Orphan Basin resulted in two extensional systems separated by a horst structure. In the East Orphan sub-basin, extension was accommodated by wide half-graben bounded mainly by east-dipping listric normal faults (Figs. 6, 10). In the West Orphan sub-basin, extension occurred along steeper normal faults that dip towards the west and define narrower half-graben and eastward tilted blocks with pronounced ridges (Figs. 6, 10, 11). The sedimentary infill of the extensional structures is primarily composed of Jurassic and Lower Cretaceous successions.

Fig. 10
figure 10

Upper crustal portion of the depth-converted seismic line E. The upper panel shows the stratigraphic and structural interpretation of the seismic reflection data. The lower panel displays the seismic data without interpretation (seismic data courtesy of TGS)

Fig. 11
figure 11

Portion of the seismic reflection profile C in the time domain showing extensional systems defined by west-dipping normal faults, characteristic of the Early Cretaceous rifting phase localized in the West Orphan Basin (seismic data courtesy of TGS). Faults are plotted with solid black lines. Colours of the stratigraphic horizons are described in the caption of Fig. 5. Location is shown in Fig. 2

The Jurassic sediments are found mainly in the eastern domain of the basin to the east of the horst structure, where they show large thickness variations related to east-dipping normal faults. They show growth strata and thickening towards the faults (Fig. 6a–c), a character indicating a syn-tectonic deposition during crustal stretching. The Jurassic deposits get much thinner and scattered as we move to the west of the horst structure, and are absent on the continental shelf. They are characterized by sub-parallel reflections with no evidence of syn-depositional tectonics in the western part of the Orphan Basin (Figs. 6a, d, 11). The Jurassic succession is capped by an erosional unconformity (Tithonian unconformity) across the basin illustrated on the seismic data by truncated reflections (Fig. 6).

The Lower Cretaceous sediments unconformably overlay the Jurassic and show more lateral continuity across the deep basin and the continental shelf (Figs. 6, 10). Their thickness is relatively constant across the continental shelf, but shows substantial variations in the deeper domains of the basin where the normal faulting appears to have controlled sedimentation. This is particularly the case in the western part of the deep basin where the Lower Cretaceous package is characterized by significant thickening inside the half-graben and strata growth patterns towards the faults (Figs. 6d, 11), a character yet again similar to West Iberia’s Lusitanian Basin (Alves et al. 2003, 2006, 2009) and offshore Ireland (Shannon et al. 2001). East of the horst structure, the Lower Cretaceous succession is also affected by syn-rift normal faults but displays smaller offsets than in the west. The topmost seismic reflections of this sedimentary unit exhibit truncation patterns related to the Aptian–Albian erosional unconformity (Figs. 6, 10).

The Upper Cretaceous succession seals most of the syn-rift structures in the eastern part of the basin (Figs. 6b, c, 10), where it is characterized by sub-parallel reflections onlapping the underlying Aptian–Albian unconformity. It is much thinner in the western part of the basin (Figs. 6d, 11) between the continental shelf and the horst structure where the west-dipping faults were still active and the tilted block ridges remained emerged during Late Cretaceous time.

The Cenozoic succession has a very consistent character across the entire Orphan Basin. Its base shows reflections sub-parallel to the underlying Upper Cretaceous with no evidence of erosional or angular unconformity. This succession is extremely thick above the edge of the continental shelf and thins considerably as we move to the east of the basin (Figs. 6, 10). Its internal architecture is well imaged along the different seismic profiles but will not be discussed in detail in this work as the main focus is on the syn-rift evolution.

Crustal structure of the Orphan Basin

Moho and crustal thickness along seismic line E

Our Moho interpretation corresponds to the base of high-amplitude reflections located in the deeper part of the seismic section, below moderately reflective layering above and lower reflectivity below (Fig. 12). These reflections are shallow and continuous underneath the faulted crystalline basement, but become diffuse and discontinuous as they deepen below the rigid crustal blocks (Fig. 12). While the Moho ‘pick’ is somewhat uncertain in the seismic reflection imaging, there is generally a reasonable correlation with wide-angle seismic results and gravity modelling (see below). Furthermore, the crustal seismic velocities used for the depth conversion range between 5790 m/s in the shallowest part of the crust and 6750 m/s in deepest part of the crust (Fig. 8). This is consistent with crustal velocity estimates from wide-angle seismic data in the basin (Watremez et al. 2015; Lau et al. 2015).

Fig. 12
figure 12

Depth-converted seismic reflection line E showing the entire crustal architecture across the Orphan Basin. Profile location is shown in Fig. 2. The crustal domains from west to east are: the continental shelf, the western sub-basin, the structural high, and the eastern sub-basin. Interpretation of the Moho discontinuity is plotted with a solid yellow line and is shown in more detail in a to c (seismic data courtesy of TGS)

The crustal architecture of the Orphan Basin, illustrated on seismic line E (Fig. 12) crossing the Orphan Basin in a WNW–ESE direction, clearly shows the existence of four distinct crustal domains. The continental shelf domain to the west is characterized by the thickest post-rift succession (up to 7000 m), a very thin syn-rift (<2000 m), and no syn-rift faulting affecting the basement. The crust is 28 km thick in the west where the Moho is 30 km deep, but thins progressively and considerably towards the east where the Moho is 15 km deep and the crust is only 7 km thick. The western sub-basin domain displays a thick post-rift succession in the west (<5500 m), which is much thinner in the east (<2500 m), and isolated syn-rift depocentres in the half-graben filled mainly by Lower Cretaceous sediments (<3000 m thick). The crust is affected by a series of west-dipping faults, which bound the half-graben and sole out at a mid-crustal detachment ~14 to 15 km deep (Fig. 12). The crust is 9 km thick in the west and 16 km thick in the east. The crust thickens as the Moho deepens towards the east but thins locally under the half-graben where it may be only 5 km thick. The eastern sub-basin domain is separated from the western sub-basin by a 150-km-wide structural high, showing very reduced syn- and post-rift successions (<3000 m thick) and a 17-km-thick crust with a 21- to 22-km-deep Moho. The eastern sub-basin is covered by a 2.5- to 4-km-thick post-rift succession, overlying the Jurassic–Lower Cretaceous syn-rift sediments (<5000 m thick), which fill the half-graben. This domain is affected by major east-dipping listric normal faults soling out at the Moho level. The crust is 12 to 15 km thick in the west but only 4 km thick in the east. The Moho underneath the sub-basin shallows rapidly towards the east, reaches a minimum depth of 12 km in the centre, and remains about 14 km deep in the east (Fig. 12).

Figure 13 displays a comparison between our interpretation of the basement top and the Moho along seismic line E with the basement top and Moho predicted by Welford et al. (2012) in their 3D gravity inversions. Although the first-order variations in the basement topography compare favourably, it is 1 to 2 km deeper in the continental shelf domain and 2 to 3 km shallower in the eastern sub-basin, especially underneath the half-grabens. However, the Moho topography predicted from the gravity data is smoother and varies between 22 and 15 km depth, while the seismic data suggest a Moho ranging between 30 and 12 km depth with more abrupt changes, notably beneath the continental shelf and the eastern part of the basin. The mismatches are likely due to the basement depth constraints, the lower-frequency content of the satellite gravity data, and the density bounds used in the inversion (Welford et al. 2012).

Fig. 13
figure 13

Comparison between basement top and Moho from our seismic interpretation of Line E and from 3D gravity inversion after Welford et al. (2012)

Recent results from traveltime tomography and dense wide-angle reflection data along the OBWAVE line (Lau et al. 2015; Watremez et al. 2015), which is slightly oblique to our seismic line E (Fig. 14a), show a crustal structure comparable to our interpretation. They demonstrate the existence of two sub-basins with extremely thinned crust (<10 km) and separated by a narrow domain (80 km wide) with a 20-km-thick crust. Nevertheless, they estimate a much thicker crust underneath the Flemish Cap (32 km thick) and a much steeper Moho shallowing at the transition between the continental shelf and the western Orphan sub-basin.

Fig. 14
figure 14

a Free-air gravity anomalies over the Orphan Basin and surrounding areas with tracks of seismic line E in black and the OBWAVE line in white (Watremez et al. 2015; Lau et al. 2015). b Gravity model along line E: upper panel shows the observed and calculated gravity anomalies along line E; lower panel displays the density model with the predicted mid-crust boundary (solid purple line). Calculated densities are indicated in red. For comparison, the mid-crust and Moho boundaries obtained along OBWAVE line are projected along our line E (Lau et al. 2015)

2D gravity model along Line E

We performed 2D forward gravity modelling along line E, using the GM-SYS profile modelling module in Geosoft Oasis Montaj software, to assess the consistency of our interpretation of the Moho and the resulting crustal structure (Fig. 14). We use the satellite altimetry gravity anomalies (Fig. 14a) (Sandwell and Smith 2009), which can be downloaded from Scripps Institution of Oceanography (https://scripps.ucsd.edu/). The free-air gravity anomalies in the Orphan Basin are overall negative and surrounded by strong gravity highs over the continental shelf, the Flemish Cap, and the Orphan Knoll (Fig. 14a). A narrow N–S elongated gravity high is also shown within the Orphan Basin and corresponds to the structural high domain, which is documented along the seismic lines between the eastern and western sub-basins.

The depth-converted seismic line E was simplified for gravity modelling purposes. It was subdivided into six sub-horizontal layers: sea water, post-rift sediments, syn-rift sediments, upper crust, lower crust, and upper mantle, and each layer was given an average initial density based on the density analysis shown in Fig. 15. The latter figure illustrates density variation along line E derived from interval velocities using Gardner’s equation, which describes the relationship between P-wave velocity and density (Gardner et al. 1974). It shows that density ranges between 2100 and 2600 kg/m3 in syn- and post-rift sediments, 2400–2700 kg/m3 in the upper crust, and 2700–2900 kg/m3 in the lower crust. The forward modelling was constrained by fixing the thickness of the sedimentary layers and the crust, and the density of the sea water. The densities of the other layers and the mid-crustal boundary were adjusted to achieve the best fit between calculated and observed gravity anomalies along the profile.

Fig. 15
figure 15

Density grid along line E, calculated from interval velocities (Fig. 7) according to Gardner et al. (1974)

The gravity model, presented in Fig. 14b, shows a good agreement between calculated and observed gravity values with a low RMS error of 1.6 mGal, which supports our interpretation of the Moho underneath line E. The good fit is achieved by assuming densities of 2180, 2400, 2600, 2990, and 3340 kg/m3 for syn-rift sediments, post-rift sediments, upper crust, lower crust, and mantle lithosphere, respectively. These density values are comparable with the density model proposed by Lau et al. (2015) who assume slightly denser upper crust (2700 kg/m3) and mantle lithosphere (3350 kg/m3) along the OBWAVE line. The obtained mid-crust boundary (Fig. 14b) indicates strong upper and lower crustal thinning below the continental shelf and the eastern sub-basin. The upper crust is extremely thinned between 60 and 100 km and between 420 and 480 km, while the lower crust completely disappears beneath the edge of the continental shelf between 140 and 160 km and the eastern sub-basin between 370 and 420 km. Figure 16 shows that the mid-crustal boundary revealed by the gravity model is very consistent with the syn-rift structures inferred from the interpretation of seismic line E. The variations at the base of the upper crust somewhat coincide with the deep syn-rift normal faults, especially in the eastern Orphan sub-basin where the normal faults sole out at the Moho level and in the western Orphan sub-basin where the extensional faults die within the crust along a discontinuity which is roughly coincides with the predicted mid-crustal boundary (Fig. 16). However, the continental shelf at the western end of the section shows a sharp shallowing of the mid-crust boundary that is not related to any faulting (Fig. 16).

Fig. 16
figure 16

Crustal section along line E showing correlation between the predicted mid-crust boundary and the syn-rift structures

Reconstruction of the tectono-stratigraphic evolution of the Orphan Basin

To reconstruct the tectonic evolution of the Orphan Basin and quantify the amount of crustal stretching during Mesozoic rifting, we restore crustal section E (Fig. 12) using the MOVE® software (http://www.mve.com/). The 2D restoration algorithm assumes two fundamental rules which are bed length conservation and bed area conservation. The former is based on the assumption that a given sedimentary bed keeps the same length before and after deformation. The latter rule assumes that deformation does not change the cross-sectional area of sedimentary units. The kinematic restoration is based on the relationship between fault geometry and deformation patterns in the hanging walls of the normal faults, as documented along the seismic profile (Figs. 6, 10). Therefore, a simple shear model is used for most of the faults and the shear angle is defined based on the hanging-wall geometries (i.e. rollover structures, sedimentary growth, and flexural slip). Most of the normal faults show an antithetic shear angle ranging between 90 and 60°. During the restoration, decompaction is applied to the different lithostratigraphic successions using surface porosities and rates of decay with depth according to Sclater and Christie (1980). Decompaction requires palaeowater depth to be constrained along the profile. We assume that the basin remained at a very shallow position near sea level during the entire syn-rift period (i.e. from Jurassic to Late Cretaceous times). This is supported by the terrigenous nature of the Upper Jurassic and Lower Cretaceous sediments and by the emerged position of the basement highs (i.e. horsts and tilted block ridges) during that time and until the Late Cretaceous. We also assume that the basin started to deepen in the Palaeocene and that the water depth increased linearly with time until it reached the present-day bathymetry. Local Airy isostasy is used to account for the isostatic response to sedimentary unloading during decompaction. We adopt average values for sediment (2300 kg/m3), crust (2800 kg/m3), and mantle (3340 kg/m3) densities, which are inferred from the density model described above.

Post-rift period

Restoring the post-rift strata consisted mainly of removing the water column and the Upper Cretaceous-Cenozoic successions (Fig. 17a, b), as most fault movements ended before the Late Cretaceous. The resulting decompaction affects primarily the western part of the profile (i.e. the continental shelf and the western sub-basin domains) where the thickest (up to 5 km) post-rift sediments are found. The structural architecture of the basin remained the same, except for minor changes in the angle of the normal faults in the western sub-basin domain which resulted in tilting of the faulted blocks by 2–4° further towards the east. The isostatic rebound resulted in a shallower Moho, especially beneath the sub-basins.

Fig. 17
figure 17

Kinematic and structural restoration of crustal line E, showing the evolution of the Orphan Basin. Each section is obtained by applying decompaction and fault restoration, starting from a the present-day profile, going through, b the Early Cretaceous, c the Late Jurassic, d the Middle Jurassic, and ending at e the pre-rift time. The restoration is accomplished using Move software (academic licence provided by Midland Valley Exploration)

Syn-rift period

Syn-rift restoration was carried out in two steps. The first step consisted of removing the Lower Cretaceous succession, decompacting the Jurassic sediments, and restoring the normal faults to the end of Jurassic time (Fig. 17c). This step (Fig. 17c) substantially affected the western sub-basin domain where Early Cretaceous deformation is more pronounced. Extension is distributed along west-dipping normal faults, and tilting of the faulted blocks is compensated at the mid-crustal level. The obtained profile (Fig. 17c) shows a wide and continuous Jurassic basin in the east and isolated Jurassic troughs in the crustal high and western sub-basin domains. In contrast to the Jurassic sediments in the eastern sub-basin, which show syn-depositional tectonic structures (Fig. 10), the scattered Jurassic deposits to the west along the profile show no evidence of fault-controlled deposition (Fig. 11). Therefore, we assume that any (modest?) strata deposited in the western sub-basin early during the Jurassic syn-rift phase were eroded sometime before the initiation of the second rifting phase in the Early Cretaceous (Fig. 17d). The second syn-rift restoration step (Fig. 17e) involved removing the Jurassic succession and restoring the basement rocks. It mainly affected the eastern sub-basin, which was the locus of crustal stretching during the Jurassic.

Pre-rift crust

The restored pre-rift profile in Fig. 17e displays the crustal architecture prior to rifting, assuming that the Jurassic to Early Cretaceous extension was entirely accommodated by brittle deformation along the faults. It shows that the total extension is on the order of 110 km; 50 km of extension occurred during the Jurassic and was primarily accommodated by the east-dipping normal faults in the eastern sub-basin domain (Fig. 17c), while 60 km of extension is related to the Early Cretaceous rifting phase accommodated predominantly by the west-dipping normal faults (Fig. 17b). The restored profile (Fig. 17e) shows a 25- to 28-km-thick crust beneath the western part of the shelf, the western sub-basin and the crustal high domains, and a 10- to 12-km-thick crust beneath the eastern portion of the continental shelf and the eastern sub-basin domains. This is reflected in the topography of the restored mid-crust boundary (Fig. 17e), which displays two highs in the continental shelf domain and in the east sub-basin domain.

Similarly, seismic data and kinematic restoration of highly attenuated crustal domains found on the Iberian rifted margin (Manatschal et al. 2001; Manatschal 2004) indicate that the pre-faulting crustal thickness does not exceed 12 km. The authors argue that the pre-faulting stage does not correspond to a pre-rift stage, for which an equilibrated crust (i.e. 30 km thick) is assumed.

Discussion

Tectono-stratigraphic evolution of the Orphan Basin during rifting

Data presented in this work indicate that rifting in the Orphan Basin occurred in at least two phases. The initial Jurassic rifting phase affected mainly the eastern part of the basin where continental extension is documented by east-dipping faults and thick Jurassic sediments, preserved in the hanging wall of the half-graben (Figs. 6, 10). Seismic reflections within the Jurassic succession show growth structures (Fig. 6), an indication that the normal faulting was coeval with, and controlled, sedimentary deposition. The Jurassic succession in the West Orphan Basin is substantially thinner and discontinuous with no evidence of syn-depositional tectonics (Figs. 6, 11). The major uncertainty in our seismic interpretation relates to the Jurassic deposits drilled only in the East Orphan Basin and for which there is no direct evidence in the west. However, the Deep-Sea Drilling Project (DSDP) Site 111, drilled on the Orphan Knoll, intersected coarse sandstone and shales at the base of the well that were determined to be Middle Jurassic (Bajocian) in age and above the Palaeozoic basement (Laughton et al. 1972). The compositions of these Middle Jurassic clastics indicate they were deposited in a coastal plain environment (Laughton et al. 1972), which we believe extended over most of the West Orphan Basin and may be part of the continental shelf.

The second rifting phase took place in the Early Cretaceous. During this phase, deformation spread towards the west of the Orphan Basin and resulted in half-graben bounded by west-dipping faults and filled by thick Lower Cretaceous deposits (Figs. 6, 11). In the east, some of the Jurassic east-dipping normal faults were reactivated and new sedimentary depocentres were developed (Fig. 6). Two-phase rifting in the Orphan Basin was already mentioned in previous work by Enachescu et al. (2005) who proposed an initial NW–SE extension in the Jurassic localized in the eastern part of the basin, and a late E-W extension which mainly affected the West Orphan Basin, but can also be documented in the east, during the Early Cretaceous. Seismic and well data indicate that rifting in the Orphan Basin was interrupted by two erosional events corresponding to the Tithonian and the Aptian–Albian truncation unconformities. The resulting angular unconformities between Jurassic and Lower Cretaceous strata (Tithonian unconformity), and Lower Cretaceous and Upper Cretaceous strata (Albian–Aptian unconformity) suggest that the basin was first uplifted then eroded. However, the duration of these uplift/erosion events remains difficult to constrain.

Late Cretaceous time marks the beginning of the post-rift stage as lithospheric break-up of the Newfoundland–Iberia/Europe rift occurred east of the Orphan Basin and the Grand Banks. Mantle lithosphere rocks were exhumed before the initiation of sea-floor spreading which occurred near the Aptian–Albian boundary (Tucholke et al. 2007). In the Orphan Basin, the Upper Cretaceous succession sealed most of the syn-rift structures but did not cover the structural highs (i.e. horsts and tilted block ridges), which remained above sea level at that time (Figs. 6, 10, 11). Some of the west-dipping faults in the western part of the basin appear to have been reactivated during the Late Cretaceous (Figs. 6, 10, 11), probably in relation to the ongoing rifting to the north between Labrador and Greenland. It is only in the Palaeocene that the Orphan Basin evolves in a clearly post-rift setting, with a strong subsidence resulting in deposition of a thick Cenozoic succession over the rift structures.

Crustal extension and thinning in the Orphan Basin

The Moho and basement top in our interpretation of seismic line E (Fig. 12) indicate substantial variations in crustal thickness beneath the Orphan Basin. The crust is revealed to be significantly thin beneath the east and west sub-basins, and much thicker beneath the westernmost part of the continental shelf and the structural high separating the two sub-basins. The gravity modelling in this paper supports our interpretation of the seismic Moho as it shows that the observed free-air gravity anomalies can be explained by assuming a mid-crust boundary separating a low-density upper crust from a high-density lower crust (Fig. 14b). The obtained mid-crust boundary is consistent with the syn-rift structures imaged on the seismic data (Fig. 16). The syn-rift faults appear to be deeply rooted in the crust and sole out at the mid-crust boundary in West Orphan Basin and reach the Moho discontinuity in the East Orphan Basin where the crust is much thinner.

Our structural restoration (Fig. 17) shows that the observed brittle extension during rifting can explain only 60 % of the total thinning in the Orphan crust (i.e. in 2D along line E), assuming an initial pre-rift crustal thickness of 28 km—the maximum thickness observed beneath the continental shelf. The remaining crustal thinning is unlikely to be localized further offshore and should occur within the Orphan Basin because the crustal blocks surrounding the basin appear to be relatively undeformed—i.e. the Orphan Knoll and, especially, the Flemish Cap.

The discrepancy between recorded brittle extension and crustal thinning in the Orphan Basin is in accordance with the results of subsidence analysis. The latter shows that there is a large syn-rift subsidence deficit when we consider the amount of crustal thinning observed across the basin (Dafoe et al. 2015; Gouiza et al. 2015). According to Couiza et al. (2015), the syn-rift succession should be at least 1.5–2.5 km thicker in the Orphan Basin. This may suggest the existence of, in addition to the preserved brittle extension, another extension phase that took place early during the rift process (Early Jurassic or even Late Triassic?) which is not easily recognizable on seismic data. Reston (2007, 2009) attributes the extension/thinning discrepancy to unrecognized polyphase faulting which is very difficult to identify in seismic data due to limitations in seismic resolution with depth and structural overprinting. However, fault geometries, syn-rift stratigraphic architecture, and basement structures imaged in the examined seismic data do not provide solid evidence that polyphase rifting occurred in the Orphan Basin. Furthermore, Reston (2007) reckons that polyphase faulting is expected to occur in rifted margins where extension exceeds 100 % rather than in rift basins where crust is often less stretched (only 25 % of crustal extension in the Orphan Basin, along Line E). Alternatively, brittle extension in the Orphan Basin could be only partially preserved due to important crustal flexure, uplift, and erosion (Burov and Cloetingh 1997; Burov and Poliakov 2001). Well and seismic data presented above clearly indicate sub-aerial exposure and truncation of the syn-rift successions and tilted blocks (Enachescu et al. 2004; Dafoe et al. 2015). The existence of discontinuous pockets of Jurassic sediments in the West Orphan Basin might even be the remnants of a more extensive Jurassic or even Late Triassic (McAlpine 1990; Driscoll et al. 1995; Lau et al. 2006; Van Avendonk et al. 2009) syn-rift basin that have been eroded during the ensuing rift phase.

A hidden early rift phase, either unrecognized (i.e. polyphase faulting) or removed (i.e. uplifted and eroded), is not sufficient to explain the extension discrepancy in the Orphan Basin. Comparing lower and upper crustal thinning against whole crustal thinning (Fig. 18) shows that upper crustal thinning factors are higher in the continental shelf and the Flemish Cap (i.e. domains with the thickest crust; Fig. 12), while lower crustal thinning factors are higher in the eastern and western sub-basins (i.e. domains with the thinnest crust; Fig. 12). This is consistent with the findings of Lau et al. (2015) and Watremez et al. (2015) who constrained the Orphan Basin crust along the OBWAVE line (Fig. 14) using wide-angle seismic data and traveltime tomography data, respectively. These observations are clear evidence of depth-dependent thinning within the crust and can indicate additional processes of ductile flow within the lower crust, mainly underneath domains of localized upper crustal brittle deformation, namely the East and West Orphan sub-basins.

Fig. 18
figure 18

Cross-plot showing variation of crustal, upper crustal, and lower crustal thinning factors (γ = 1 − 1/δ, where δ = initial thickness divided by final thickness) along line E

Another source of uncertainty in quantifying the pre-rift crust is inheritance. The Mesozoic North Atlantic rift formed on remnants of the Palaeozoic Caledonian–Appalachian Orogeny, where the closure of the Iapetus Ocean led to continental collision, crustal and probably lithospheric thickening (imbrication of multiple terranes), crustal uplift, and erosion (Dewey and Kidd 1974; Williams 1984, 1995; McKerrow et al. 2000). The existence of four distinct crustal domains (namely the East Orphan sub-basin, the structural high, the West Orphan sub-basin, and the continental shelf), showing different deformation styles and migration of rifting, suggests that lateral heterogeneities—variations in thickness and/or rheology inherited from the pre-Mesozoic Caledonian–Appalachian Orogeny—may have played a crucial role in the tectonic evolution of the Orphan Basin during, and following rifting (Smith and Mosley 1993; Vauchez et al. 1998; Huismans and Beaumont 2007; Gouiza et al. 2015; Chenin et al. 2015).

Conjugate Irish margin

Palaeoreconstructions of the North Atlantic Ocean to a time prior to lithospheric break-up and oceanic accretion (Chron M0; Fig. 1) show that the East and West Orphan sub-basins were juxtaposed against the Porcupine and the Rockall Basins, respectively, which are presently located on the conjugate Irish margin (Knott et al. 1993; Louden et al. 2004). These three basins probably evolved as one large extensional system during the Mesozoic rifting (Lundin and Doré 2011; Lau et al. 2015) and should exhibit comparable stratigraphic, structural, and crustal syn-rift evolution.

Seismic data from the Porcupine and Rockall Basins indicate the presence of thick Late Palaeozoic to Cenozoic sedimentary successions (up to 10 km) overlying faulted blocks caused by extensive crustal stretching and suggest a multiphase Mesozoic syn-rift history (Mackenzie et al. 2002; Morewood et al. 2005; O’Reilly et al. 2006), similar to our observations from the Orphan Basin. The Porcupine Basin, where the syn-rift succession is well defined, shows three rift episodes of Permo-Triassic, Late Jurassic, and mid-Cretaceous age which were separated by periods of uplift and erosion (Morewood et al. 2005). The major episode of brittle extension is believed to be Jurassic, which is consistent with westward migration of the loci of rifting described in the Orphan Basin. Like in the Orphan Basin, the deep structures of the Porcupine and Rockall Basins are characterized by thick crust (25–30 km) beneath the basin-bounding highs and highly thinned crust below the basins (2–10 km) (Morewood et al. 2005; O’Reilly et al. 2006). The Moho beneath the two basins shows a very pronounced asymmetry that is less evident in the Orphan Basin, but we do observe a lateral offset between areas of main upper crustal faulting and areas of extreme lower crustal thinning, especially in the western sub-basin (Fig. 12). The major difference between the two conjugate margins is revealed by the various velocity models which show strong evidence of low-velocity serpentinized mantle beneath the highly thinned crust in the Porcupine and Rockall Basins, but not in the Orphan Basin (Chian et al. 2001; Morewood et al. 2005; O’Reilly et al. 2006; Watremez et al. 2015; Lau et al. 2015).

Conclusions

The Orphan Basin, located offshore of the Newfoundland rifted margin, is a broad deepwater rift basin, characterized by thick Mesozoic to Cenozoic sedimentary successions which overlie a highly thinned continental crust.

The deep seismic reflection data used in this work clearly demonstrate the existence of strong brittle deformation coupled with a high degree of crustal thinning. Sedimentary successions, preserved in the Orphan Basin, suggests a two-phase rifting during the Jurassic and the Early Cretaceous, respectively, separated by periods of uplift and erosion. Upper crustal extension initiated in the east sub-basin in the Jurassic and spread to the west sub-basin in the Early Cretaceous. The latter experienced a late phase of deformation that reactivated the syn-rift structures and that was likely related to the contemporaneous rifting to the north in the Labrador Sea.

However, the interpreted structural and crustal architectures in the Orphan Basin suggest that the observed brittle extension does not fully match the quantified crustal thinning. An Early Jurassic, or even Late Triassic, brittle rift phase and localized thinning by ductile flow in the lower crust are proposed as additional processes that could have contributed to continental deformation in the basin.