Interaction Between Bottom Currents and Slope Failure in the Late Cretaceous of the Southern Danish Central Graben, North Sea

By Esmerode, E V Lykke-Andersen, H; Surlyk, F

Abstract: The NW European Chalk Group was deposited in a deep epicontinental sea traditionally conceived as a quiet depositional setting that was only affected by redeposition along structural highs. The ‘chalk sea’ was, however, locally affected by powerful bottom currents during some periods. Three-dimensional seismic reflection data from the southern Central Graben show abundant intra- chalk morphological elements concentrated at two stratigraphic levels within the Turonian-Campanian sequence. Both levels are topped by extensive unconformities. The geometry of the elements produced by alongslope processes is interpreted as caused by SE- directed bottom currents flowing between the Bo-Jens and Adda ridges. The highest current intensities led to the formation of the two unconformities, judged by the occurrence of the largest current- formed channels and drifts at these two stratigraphic levels. This sea-floor topography was locally modified by large-scale downslope mass transport. Stratigraphical coincidence of the largest slumps and bottom-current erosion suggests a coupling between downslope and alongslope processes. Current erosion impinged upon the slopes during periods of highest current speeds, destabilizing the slope sediments and triggering slumping. The smoothing of the bathymetry following local relaxation of the tectonic inversion accounted for the coeval deactivation of alongslope and downslope processes during the succeeding Maastrichtian deposition.

The Upper Cretaceous-Danian Chalk Group covers a wide region in NW Europe. The chalk is a biogenic deposit mainly of pelagic origin, and therefore affected by variations in primary productivity (e.g. Hancock 1975; Kennedy 1987a; Surlyk et al. 2003). The traditional picture of deposition in the NW European ‘chalk sea’ is a quiet pelagic rain and drape of topographic elements. Redeposited chalk units have, however, also been identified in the Central North Sea adjacent to tectonically active areas and areas affected by halokinetic movements (Kennedy 1980; Watts et al. 1980; Hardman 1982; Brewster & Dangerfield 1984; Munns 1985; Harton 1986; Bromley & Ekdale 1987; Bramwell et al. 1999; Farmer & Barkved 1999). In some areas there is increasing evidence of a highly dynamic depositional context. Erosional channels, scars and associated mass-transport deposits are well known from the coastal outcrops of Haute Normandie, France (Kennedy & Juignet 1974; Quine & Bosence 1991). Winnowing and redeposition of chalk is revealed by thinning of the succession over structural highs and by detailed micropalaeontological analyses (Bergen & Sikora 1999; Sikora et al. 1999). Only little is known of Late Cretaceous oceanographic circulation patterns in the ‘chalk sea’, although information from outcrop and offshore seismic data indicates that bottom currents were locally of great importance (Quine & Bosence 1991; Evans & Hopson 2000; Evans et al. 2003). It has recently been demonstrated that powerful bottom currents played an important role in forming large drifts and moats in the Maastrichtian chalk of the Danish Basin (Lykke-Andersen & Surlyk 2004; Esmerode et al. 2007; Surlyk & Lykke-Andersen 2007). In this paper we provide examples of structures produced by bottom currents and mass-transport processes in the Upper Cretaceous chalk of the southern Central Graben in the North Sea (Fig. 1). Large mass-transport deposits have hitherto not been described in connection with chalk drifts. We aim to demonstrate the interplay between bottom currents and mass- transport processes in the area. Our interpretation is assisted by a reconstruction of the main topographic elements of the sea floor in the area, and aims to improve the understanding of the palaeocirculation pattern.

Geological setting

The study area covers the southern part of the Tail End Graben, the northern part of the Salt Dome Province, and the Poul Plateau in the Danish North Sea sector (Fig. 1). The NW-SE-trending Tail End Graben is a half-graben that is deepest against the eastern bounding Coffee Soil Fault, a major fault zone separating the Central Graben from the Ringkobing-Fyn High in the Danish North Sea sector (Fig. 2). Halokinetic influence in the Tail End Graben increases southwards towards the Salt Dome Province and northwards towards the Segne Basin (Vejbask 1986). The Poul Plateau is a faulted block that formed between Triassic and early Late Jurassic times (Meiler 1986). It is located where the Coffee Soil Fault changes its orientation from NNW-SSE to NW-SE and acts as a transition zone between the Tail End Graben and the Ringkobing-Fyn High (Fig. 1).

The Tail End Graben was subject to intense extension during a major Late Jurassic rift phase that affected the Central Graben region (Andersen et al. 1982; Gowers & Saeboe 1985). The depositional pattern of the Upper Cretaceous-Danian Chalk Group in the Central Graben resulted primarily from pelagic settling in a subsiding basin. Late Cretaceous regional subsidence was, however, punctuated by widespread inversion in the form of reversal of faults, reactivation of old fault trends, and folding of the basin infill (Vejbask & Andersen 1987, 2002). The Coffee Soil Fault underwent severe inversion in the southern part of the Tail End Graben, the Ringkebing-Fyn High locally becoming a chalk depocentre (Vejbaek & Andersen 2002). The fault-controlled Late Jurassic-Early Cretaceous depocentres, such as the Tail End Graben, formed the main areas subject to Late Cretaceous inversion. This is recorded by the development of inversion swells that constituted bathymetric elevations, and the formation of local depocentres in the intervening lows (Oakman & Partington 1998). The Bo-Jens and Adda ridges were syndepositional sea-floor topographic elements with a significant relief, reflected by thinning of the Upper Cretaceous succession over their crests (Fig. 2; Kennedy 19876; Farmer & Barkved 1999). The Adda Ridge, located north of the Poul Plateau, parallels the adjacent NW-SE-trending segment of the Coffee Soil Fault and resulted from inversion caused by the dextral movement of the fault (Fig. 1; Vejbaek & Andersen 2002). The structural development of the north-south-oriented Bo-Jens Ridge is not directly connected to the Coffee Soil Fault, but possibly to the presence of a system of extensional basement faults that influenced the movement of the Zechstein salt (Vejbaek & Andersen 2002). At the crest of the ridge there is evidence for post-Danian erosion as a result of a late Palaeocene-Eocene inversion episode. This episode also led to the formation of the NW-SE-trending Tyra-Igor Ridge, which crosses the Salt Dome Province and the southern Tail End Graben (Fig. 2; Vejbaek & Andersen 2002).

Dataset and methods

The 3D seismic reflection dataset used in this study forms part of the Contiguous Area data licence, a large 3D seismic volume acquired in 1994, and covers c. 900 km^sup 2^ of the southern Danish Central Graben (Fig. 1). The survey has inline and crossline intervals of 12.5 m; it was migrated with two-pass migration for the high-velocity intervals, and a phase rotation to zero phase was applied. In this area, the Chalk Group is buried to depths of 1500- 2700 m. The dominant frequency component at this depth interval is 50 Hz and the vertical resolution is estimated to be about 20 m. Seismic stratigraphie interpretation was conducted in the time domain and hence the maps generated are given in two-way travel time (TWT) units. Depth control is provided by the wells Boje-1, Bo-I, E- IX, E-3X, Adda-1, Adda-3, NW Adda-1 and Deep Adda-1, for which stratigraphie markers and petrophysical logs were available (Fig. 1). Estimates of the dimensions of intra-chalk structures are derived from the interval velocity records of the neighbouring wells. Time-structure, time-thickness and seismic attribute maps, as well as time slices and 3D seismic displays were the basic tools used for the seismic interpretation. The generation of restored maps was carried out by flattening a selected horizon to a datum surface. This process requires that the datum is a slightly younger surface assumed to have been nearly flat at the time of deposition. For the middle part of the Chalk Group the surface that best fulfils this condition is the Top Maastrichtian. This is justified by onlap analysis in the area, which shows the Maastrichtian unit to be of clear infilling character and internally rather conformable. However, it must be noted that in the northwestern corner of the area the Top Maastrichtian is slightly erosional, causing a local subtle uplift effect on the restored surfaces. The Palaeocene succession is characterized by a rather uniform thickness and parallel layering, thus the Top Palaeocene surface is used for restoring the upper part of the Chalk Group in the area. This type of restoration does not compensate for the tectonic activity occurring in the time span between the deposition of the restored and the datum surfaces. For visualization reasons, a qualitative correction is applied where synsedimentary subsidence is observed and this is annotated on the maps. A possible objection to the construction of restored maps in the time domain is that changes in both the thickness and the lateral velocity of the succession between datum and restored surfaces can lead to significant changes in the dip angle of the slopes. The steepness estimates of the slopes are therefore only tentative. However, the generated maps provide a means for identifying the location, strike and dip direction of the palaeoslopes, and thus allow interpretation of likely sediment transport directions. Unit subdivision

The Chalk Group appears in the cores as a homogeneous and pure limestone of various shades of white. Wireline logs show, however, variations of the geophysical properties through the succession related to lithological changes. These changes are primarily the result of fluctuations in the input of clay and in the formation of authigenic components such as glauconite and phosphatized nuclei. In terms of seismic facies and well log signatures the Chalk Group is in this study subdivided into five seismic units (Fig. 3). The units are generally bounded by strong reflections, which in some instances represent extensive unconformities. The Base Chalk is an easily recognized seismic reflection marking the transition from the marly Lower Cretaceous succession to the carbonatedominated Upper Cretaceous-Danian Chalk Group. A similar strong reflection of opposite polarity represents the Top Chalk and indicates a marked contrast in the acoustic impedance of the chalk and the overlying Selandian siliciclastic deposits.

Seismic Unit 1

Seismic Unit 1 (SUl) comprises the Cenomanian clay-rich chalk of the Hidra Formation and the lower Turanian mudstones of the Plenus Marl (Fig. 3; Deegan & Scull 1977; Munns 1985; Surlyk et al. 2003). The unit is delimited by a rather continuous Base Chalk and Top SU1 reflections at the base and top, respectively. In the areas where the unit is thickest it is possible to discern a parallel to slightly wavy internal reflectivity (Fig. 3). Areas of fault- controlled subsidence formed the main SU1 depocentres, the largest one defining a north-south trend south of the well Boje-1, and a minor NW-SE-oriented one being located around the well E-1X (Figs 4 and 5).

Seismic Unit 2

The lower part of the Hod Formation, consisting of white chalk with upward increasing clay content, is included in Seismic Unit 2 (SU2), which is of mid-Turonian to possibly Santonian age (Fig. 3; Surlyk et al. 2003). The unit is bounded at its base by the Top SUl reflection and at its top by a marked intra-Hod unconformity, the Top SU2, and is characterized by abundant reflection terminations and highly variable amplitudes (Fig. 3). The restored time- structure map of the Top SU2 surface shows the presence of two elevated areas, the north-south-trending Bo-Jens Ridge and the NW- SE-oriented Adda Ridge (Fig. 6a). The ridges are separated by an elongated NW-SE basin, which becomes narrower towards the NW, as the Adda and Bo-Jens ridges converge. Rather uniform amplitudes characterize the Top SU2 reflection (Fig. 6b). High-amplitude trends are normally found in areas of thin overlying units, such as the crest of the Adda Ridge, the greater amplitudes being due to tuning effects rather than changes in the nature of the Top SU2 surface itself. Minimum thicknesses of SU2 of about 40 m are found over the Adda Ridge and increase towards the basin, probably exceeding 250 m in the main north-south-elongated depocentre (Figs 4 and 6c).

Seismic Unit 3

Seismic Unit 3 (SU3) comprises the upper Hod Formation, consisting of cyclically bedded chalk with abundant allochthonous layers (Fig. 3; Surlyk et al. 2003). The unit is bounded by the Top SU2 at its base and the Top SU3 at its top, which represents a regional unconformity of latest Campanian to possibly early Maastrichtian age, in some areas developed as a hardground (Kennedy 1980; Brewster & Dangerfield 1984; Farmer & Barkved 1999; Surlyk et al. 2003). SU3 is characterized by good internal reflectivity with semi-parallel layering, and common downlaps or onlaps onto the Top SU2 throughout the area (Fig. 3). The restored Top SU3 map illustrates two main structural highs, the Bo-Jens Ridge in the western part of the area and the Adda Ridge in the eastern, separated by a NNW-SSE-trending basin (Fig. 7a). Concave-up and concave-down elongated structures are particularly abundant in the basin and on the flanks of the ridges. The amplitude map of the Top SU3 reflection shows two main high-amplitude trends, one in the areas corresponding to the flanks of channel-like structures and another along their bottoms (Fig. 7b). SU3 is about 160 m thick at the well Bqje-1 and thins over the Adda and Bo-Jens ridges, reaching minimum values of 12 m (Fig. 4). Northeast of the Adda Ridge the succession thickens dramatically, this pattern being opposite to that of the underlying units and showing that reversal of the Coffee Soil Fault was coeval with deposition of SU3 (Fig. 7c).

Seismic Unit 4

Seismic Unit 4 (SU4) roughly corresponds to the homogeneous white chalk of the Maastrichtian Tor Formation (Fig. 3; Deegan & Scull 1977; Surlyk et al. 2003). SU4 is bounded at its base by the Top SU3 and at its top by the Top SU4 surface. Parallel reflections with good continuity dominate in SU4 and show clear onlaps onto the topography of the Top SU3 surface. The restored map of Top SU4 shows a rather smooth relief, the only elevation being the northern part of the Bo-Jens Ridge, characterized by subtle, low-angle erosion of the top of SU4 (Fig. 8a). The few noteworthy erosional features resulted from post-Danian erosion and there are essentially no structures that could represent primary topographic sea-floor elements. Comparison between the restored map of Top SU3 and the time-thickness map of SU4 indicates that SU4 is mainly an infill succession (Figs 7a and 8b). Minimum thicknesses of around 20 m are found over the Bo-Jens Ridge, whereas thicknesses of about 10 m characterize the Adda Ridge (Figs 4 and 8b). Maximum thickness values, exceeding 200 m, are found in the southern part of the area.

Seismic Unit 5

The Top SU4 and Top Chalk reflections define the base and top of the Seismic Unit 5 (SU5), which comprises the commonly reworked white and marly chalk of the Danian Ekofisk Formation (Fig. 3; Deegan & Scull 1977; Brewster & Dangerfield 1984; Surlyk et al. 2003). The unit is very thin and its upper part is strongly eroded, precluding identification of seismic facies (Figs 3 and 4). The restored Top Chalk surface shows only minor irregularities (Fig. 9a). The former Adda Ridge has been overprinted by smaller topographic elements in the form of NW-SE-oriented parallel erosional channels and interchannel ridges. Otherwise, the general palaeo-topography is similar to the Top SU4 surface. SU5 is absent in the northwestern part of the study area, whereas wells near the Coffee Soil Fault zone show SU5 thicknesses of about 40 m (Figs 4 and 9b).

Seismic depositional structures

Structure 1

Description. A system of NW-SE-oriented ridges, channel-like structures and infill units in the Coniacian-Campanian succession is mapped in the area of convergence of the Bo-Jens and Adda ridges (Fig. 10). The Top SU2 surface has an asymmetrical channel-like morphology, c. 1300 m wide and up to 80 m deep. Its relief is significantly more accentuated than the Top SU1 and Base Chalk surfaces partly as a result of its elevated northeastern margin, which forms a coherent system of aggrading concave-down layers (Fig. 10b). The channel-like structure observed in the Top SU2 surface was filled in by SU3 in the form of a NW-SE-trending ridge flanked by 0.5-1.5 km wide channel-like structures (Fig. 10b and c). The ridge is c. 1 km wide and 3 km long, its maximum height is about 60 m and the real maximum dip of the flanks may have reached 3-4[degrees]. Internally, the ridge is characterized by a well-layered parallel set of reflections, which commonly onlap or downlap onto the walls of the channel formed by the Top SU2 surface (Fig. 10b). The channel- like element flanking the southwestern side of the ridge is of erosional nature, as seen by the truncations observed in its SW margin and high reflection amplitudes in both margins (Fig. 7b). No clear signs of erosion are seen in the eastern margin of the ridge, although reduced deposition is inferred from the apparent pinching out of single layers (Fig. 10b).

Interpretation. The pronounced relief of the channel-like feature observed in the Top SU2 surface is due to the mainly aggradational growth along its NE margin, whereas its position is clearly linked to synsedimentary relative subsidence between the BoJens and Adda highs (Fig. 10b). Structural restriction and progressive subsidence are probably the two factors explaining the stable location of the channel through time and the predominance of aggradational over progradational growth. The thickening of SU2 along the NE margin of the structure and the absence of reflection truncations indicate preferential deposition and formation of a levee, whereas sedimentation rates were lower in the bottom and SW margin of the channel. This type of depositional pattern suggests formation by a channelized flow, and the asymmetrical geometry in cross-section of the system suggests a SE-directed flow. This is based on the deflection and intensification of the flow towards the right exerted by the Coriolis force, which prevented deposition in the bottom and on the SW margin of the channel (i.e. to the right of the flow direction) and allowed levee build-up on the lee side to the NE (i.e. to the left; Fig. 1Od). Turbidity currents are often invoked to explain the formation of this type of channels. When levees no longer confine the flow, turbidity currents lose their capacity for transport and deposit their load as frontal splays. Such deposits are not observed at either end of the channel-levee system, suggesting that it was formed by a more continuous and more dilute bottom current. The coherent internal layering of the channel infill and its ridge-like geometry, of mainly depositional nature, are signs of continuous sediment supply and growth of a drift (Fig. 10b). The two channels flanking the ridge in the Top SU3 surface appear as areas of erosion and reduced deposition, mimicking moats produced by contour currents (Faugeres et al. 1999). A bottom- current origin is also supported by the sharp peak observed in the sonic and density logs of the well Boje-1, drilled through the SW flank of the ridge, which suggests the presence of a hardground formed during a period of nondeposition or even erosion (Fig. 10a). The greater relief of the southwestern channel and the polarized erosion, which is strongest along the SW margin, represent an asymmetry similar to that seen in the Top SU2 surface. This corroborates the interpretation of the ridge-channel system being formed by a bottom current flowing towards the SE. A final supporting piece of evidence is the presence of a spit-like feature at the southern end of Structure 1, probably formed where the emergence of the confined flow led to a drop in the transport capacity of the current (Fig. 10c). Structure 2

Description. A system comprising a NNE-SSW-trending ridge and flanking channel-like structures is observed south of Structure 1 (Fig. 10c). A seismic view of the succession shows that the Top SU2 surface is marked by two erosional concave-up structures, each c. 1700 m wide, truncating the upper part of SU2 (Fig. Ila). The western structure is strongly asymmetrical and has a maximum relief of 75-8Om, whereas the eastern feature has a relief of about 40 m (Fig. 1 lb). The latter extends northwards towards the upper slope of the Adda Ridge, ending as a C-shaped submarine incision with downdip concave contours (Fig. 10c).

The Top SU3 surface is characterized by a c. 2.5 km wide and 6.5 km long ridge flanked by two channel-like structures, each c. 1.5 km wide (Figs 10c and lia). The ridge presents parallel low-amplitude reflections in its western side grading to more irregular reflections in the eastern. The eastern flanking channel-like feature is up to about 85 m deep and is slightly asymmetrical, showing reflection truncations in the eastern margin (Fig. 1 lb). The western one is close to symmetrical, has a maximum depth of 110 m, and shows evidence of pronounced erosion of SU3 in both margins.

Interpretation. The western concave-up structure observed at the Top SU2 surface forms the northern section of a curved indentation clearly seen on the dip magnitude map (Fig. lie). Its outline in map view and its geometry in cross-section closely resemble a listric slide (Fig. lib). The observed concave-up elongated element was formed in association with the rollover of the slide unit, whose formation and growth along a listric detachment surface was probably aided by coeval steepening of the basin margins during inversion. The displacement of the slide lasted beyond deposition of SU3 and influenced the location of the overlying ridge and western channel- like structure. In view of the similarity in geometry and the adjacent location of the ridge-channel systems in Structure 2 and Structure 1, it is tempting to suggest that the two systems are parts of the same complex and that the processes accounting for their formation were analogous. A SE-flowing bottom current, deflected towards the slope of the Bo-Jens Ridge, is interpreted to have formed the north-south-trending western channel or moat and the genetically related ridge in Structure 2. A plausible explanation for the coincidence in location of the moat and the underling listric slide is that the synsedimentary displacement of the slide influenced the course of the current and, in turn, the location of the moat (Fig. 1 lb). A bottom-current origin is also in agreement with the orientation of the system, parallel to the palaeo- contours, and with the parallel internal layering that characterizes the western side of the ridge. However, the more chaotic seismic reflection of the eastern side and the formation of the eastern flanking channel-like structure is the result of a north-south slump event in the flank of the Adda Ridge prior to deposition of SU4 (Fig. lie and d). Figure lib shows a cross-section of the mass- transport detachment surface, characteristically with steep sidewalls or trenches and flat floor (Top SU2 surface). This slump event masked the original geometry of the current-produced ridge- channel system, creating a topographic element of mixed nature.

Structure 3

Description. Two channel-like structures trending NW-SE, i.e. perpendicular to the palaeo-contours of the Adda Ridge, are mapped in the SE part of the study area (Fig. 12). The channel shapes are associated with the apparent truncation of SU2 and SU3, and their maximum widths and depths are 2550 m and 110-120 m for the southern structure and 2250 m and 70-80 m for the northern one, respectively (Fig. 12a). The head of the northern channel-like structure forms a soft incision in the Top SU2 surface and a more marked one in the Top SU3 surface towards the SE; the two incisions do not seem to be vertically connected (Fig. 12b). The northern and southern channel- like structures differ in their geometry mainly in dip section. Unlike the northern structure, the incisions observed in the Top SU3 and Top SU2 surfaces in the southern channel-like shape seem to be connected (Fig. 12c).

Interpretation. The geometry of the channel-like elements of Structure 3 resembles that of a mass-transport system (Fig. 12d). This is consistent with their orientation perpendicular to the palaeo-contours of the Adda Ridge and with the irregular seismic signature of the fill, characteristic of the redeposited unit (Figs 6a and 12e). In the northern structure, the SE-dipping incision at the Top SU2 surface represents a low-angle scar formed prior to deposition of SU3 (Fig. 12b). The more marked scar observed in the Top SU3 surface seems to have the same origin, although the lack of physical connection between them indicates that they were generated during different events. In the case of SU3, the slide sheet was not transported far from the scar and a cross-section of the proximal part of the scar shows a smoother relief for the Top SU3 surface than for the Top SU2 surface, where the allochthonous material was transported further downslope (Fig. 12a). The presence of a single, well-defined detachment surface and the coherent seismic facies of the mass-transport deposits in SU3 point to a single slide event. The southern channel-like structure was formed by a single slope failure event, which took place prior to deposition of SU4 and affected both SU2 and SU3 (Fig. 12c). The observed channel-like structure in the Top SU2 surface therefore never formed an open conduit, but represents the basal detachment surface of the mass- transport complex.

Structure 4

Description. A persistent channel-like feature trending WNW-ESE parallel to the palaeo-contours of the Adda Ridge is situated adjacent to the well E-1X (Fig. 13). The structure overlies an asymmetrical incision in the Top SU2 surface associated with a low- angle erosional surface affecting the upper part of SU2 in the updip direction (Fig. 13b). Erosion is also reflected by the thinned SU2 compared with the downdip section. At the Top SU3 surface, a much more accentuated (c. 2 km wide and 60-65 m deep) channel-like structure is observed (Fig. 6b). The channel has a low sinuosity and is of depositional origin, as suggested by the lack of truncation along its margins and bottom.

Interpretation. The gradual downdip thinning of SU2 from the upper slope of the Adda Ridge followed by a sharp thickness increase is interpreted as caused by sediment removal from a scar zone by mass-transport processes and downslope stacking of the slumped mass (Fig. 6c). Together with the presence of a clear detachment surface this suggests that the asymmetrical incision in the Top SU2 surface was formed by downslope redeposition (Fig. 13b). The later formation of the low-sinuosity channel observed in SU3, characterized by well- layered levee deposits with hardly any erosion involved, is a sign of local organized deposition for a relatively long period of time. Both bottom currents and turbidity currents are known to produce similar channel-levee systems. The orientation, parallel to the palaeoisobaths, is indicative but not diagnostic of an alongslope mechanism (Fig. 13c). However, the straight outline of the channel and the absence of a frontal lobe at its end support a bottom- current origin. The irregular sea floor produced by slumping of the upper part of SU2 possibly intercepted and locally modified the course of the bottom current, generating a turbulent flow in the near-bottom water column responsible for the inception of the channel-levee system. This explains the apparent vertical linkage between the scar at the Top SU2 surface and the bottom-current channel at the Top SU3 surface.

The abrupt termination of the channel towards the NW needs, however, additional explanation (Fig. 13c). The restored and amplitude maps of the Top SU3 surface show an elongated concave-up structure further towards the NW, which resembles the channel with regard to orientation and width (Figs 7b and 13c). Between the two structures the palaeo-isobaths of the flank of the Adda Ridge bend updip on the upper slope, whereas they are more separated and bend downdip on the lower slope (Fig. 13c). This pattern differs from the adjacent sections of the slope and possibly resulted from large- scale downslope resedimentation by gravity flows (Fig. 13d). The channel is probably the relict of a larger system extending further northwestwards, part of which was obliterated by large-scale slope failure.

Discussion

The main Early Cretaceous depocentre in the area was located in the Tail End Graben, thickening from the centre towards the margins, and recording localized subsidence along the edges of the graben (Fig. 14a). Inversion folding of the graben infill started in early Late Cretaceous times, as indicated by erosion of the upper Lower Cretaceous succession. Consolidation of new structural elements such as the Bo-Jens and Adda ridges contributed to enhancing the sea- floor relief during the Late Cretaceous, creating mainly north- south- and NW-SE-trending slopes (Fig. 14b). As inversion of the ridges progressed, an intervening basin possibly linked to a deep fault was formed, trending north-south and widening southwards (Figs 6a and 7a). Subsidence was more intense in the south and the basin was continuously supplied with sediment from the ridges, forming an important depocentre for SU2, SU3 and SU4. The northern part of the basin, in the area of convergence of the Bo-Jens and Adda ridges, established a narrow sea-floor passage during deposition of SU2 (Fig. 6). The passageway was later partially infilled by SU3 sediments and vanished during early deposition of SU4 (Figs 7 and 8). The arrangement and evolution of the chalk depocentres around the Coffee Soil Fault zone indicate that the fault underwent reversal coeval with deposition of SU3 and possibly represented a gentle NE-dipping slope towards the East North Sea High (Figs 6a and 14b). Palaeocene flexuring of the central part of the Tail End Graben led to the complete reversal of the basin between the inversion swells and to the formation of the Tyra-Igor Ridge (Fig. 14c). Tectonic inversion seems to have continued until the Miocene, enhancing the asymmetrical uplift of the western side of the graben with respect to its eastern side (Fig. 14d). Formation of the Top SU2 surface

The Top SU2 surface is recognized in the area as a prominent unconformity that resulted from pervasive erosion by bottom currents and mass-transport processes. The majority of the erosional features associated with the unconformity are concentrated along the slopes of the Bo-Jens and Adda ridges, and in the intervening basin. Based on the analysis of individual structures, a strong bottom current is interpreted to have flowed through a narrow sea-floor pathway or channel in the northern part of the basin, which was kept open by the sediment removal effected by the current (Fig. 15a). The asymmetrical aggradational growth of the channel (see Structure 1) is used as a criterion for suggesting a southeastwards direction of the flow, in accord with the northern hemisphere Coriolis deflection. Part of the sediment supplied from the flanks of the ridges was transported down-current, and part was redistributed mainly to the left of the core of the current, depositing a levee.

In the areas further towards the south the effect of bottom currents becomes less clear. This may be accounted for partly by a decrease in current velocity in the less confined southern area, and partly by the abundant sediment supplied from the ridges. Synsedimentary inversion of the ridges and ensuing oversteepening of their flanks destabilized the slope sediments and promoted mass- wasting events such as those seen in Structures 2, 3 and 4 (Figs 11- 13). The formation of mass-transport complexes was geologically instantaneous, whereas bottom current-produced channel-levee systems resulted from continuous erosion and sedimentation over long time spans. Large-scale slumping mainly occurred at the Top SU2 surface, locally overprinting the earlier relief and introducing isobath components oblique to the slopes.

Formation of the Top SU3 surface

Large-scale structures similar to those observed in SU2 are widely recognized in SU3 along the margins of the basin (Fig. 15b). The ridge-channel systems in Structures 1 and 2 gradually filled in both the bottom-current channel and the slide scar mapped in the Top SU2 surface (Figs 10 and 11). The apparently SE-dipping axis of the intervening basin makes it difficult to distinguish between turbidity and bottom-current channels. However, the channels follow the contours of the ridges rather than the axial dip direction of the basin, indicating an alongslope mechanism. Some of the channels described in SU3 have highly variable depths. This is a common feature of bottom-current systems, reflecting localized current scouring along the moats. The system described here resembles modern confined drifts produced by contour currents related to deep and narrow passageways (e.g. Carter & McCave 1994). Confined drifts have moats along both flanks and are normally formed by a single unit, which in cross-section is characterized by bilateral sigmoidal growth bounded by topographic confinements (Faugeres et al. 1999). A turbidity current origin of the ridge-channel system is thus considered unlikely and formation by bottom currents is preferred. The asymmetrical cross-section of the ridge and side channels, and the presence of a spit-like feature at the southern end of Structure 1 are in agreement with a SE-flowing current similar to the one that swept the area during deposition of SU2. Flow velocity probably decreased after the Top SU2, as inferred from the more depositional nature of the structures in SU3. The abundance of reflection truncations in the Top SU3 indicates that local current velocities peaked again during the late Campanian. The SE-flowing current system fanned through the basin, and its velocity declined when it reached a less confined area towards the south. This was probably accompanied by a loss in erosive capacity, although it was still capable of sediment reworking as seen by the formation of the depositional channel in Structure 4 (Fig. 13).

Some of the major structures identified in the upper part of SU3 are linked to gravity-driven resedimentation, which seems to have been widespread at the SU3-SU4 boundary (Figs 11 and 12). Tectonic inversion coeval with deposition of SU3 initiated and contributed to the maintenance of unstable slopes, triggering large downslope mass movements that strongly disrupted the shape and internal architecture of the ridge-channel system present in the northern part of the basin (Figs 11 and 15b). This was also the case for the depositional channel in Structure 4, which was partly obliterated by a large slump (Fig. 15b). Signs of erosion produced by downslope mass flows are, however, almost absent along the margins of the southern part of the basin. The significant thickening of SU3 suggests that mass transport at a scale below seismic resolution could have contributed to the infill of the fault-controlled intervening basin.

Coupling of slumping and bottom currents

The interplay between downslope and alongslope sediment transport mechanisms is a common phenomenon (Faugeres & Stow 1993). A large number of studies have focused on the effects of bottom currents on the geometry of submarine fans (e.g. Locker & Laine 1992; Stanley 1993; Carter & McCave 1994; Masse et al. 1998). Downslope and alongslope processes seem also to have coexisted in the study area, leading to the formation of structures of mixed nature in SU2 and SU3. Elements produced by bottom currents attained maximum relief at the Top SU2 and Top SU3 surfaces, and downslope mass-transport events occurred repeatedly at these two stratigraphic levels. Both bottom-current deposits and large-scale slide or slump units are almost absent in SU4, being replaced by a draping and well-layered type of deposition that filled in and smoothed the topography of the Top SU3 surface. The stratigraphic concurrence of slumps and the largest bottom-current drifts and channels, and their concomitant disappearance in SU4, suggests some type of coupling between the two sets of features. When bottom currents are sufficiently competent to erode the sea bed, erosion and channel formation take place predominantly along the slopes (Faugeres et al. 1993). This destabilizes the slope sediments by increasing the slope inclination, triggering gravity-driven redeposition and emplacement of allochthonous units in the basin (e.g. Hansen et al. 2004). The bottom currents seem to have been most vigorous at the Top SU2 and Top STJ3, as indicated by the most intense erosion. Erosion of the slopes by the SE-flowing current was probably directly involved in triggering mass-wasting events combined with destabilization produced by coeval tectonic inversion. Downslope redeposition appears to post-date the formation of bottom-current elements and to have locally overwritten the previous sea-floor topography, filling in and removing parts of the channel-drift systems. Modern analogues are known from the West Shetland Drift, part of which was affected by large slide events (Knutz & Cartwright 2003). The scars produced by mass-transport events were probably opportunistically used by the bottom currents as new paths in their course, determining in a feedback loop the locations of further erosion and slope failure.

The drastic decline in the number of slumps and the abandonment of the bottom-current elements of SU4 probably records tectonic quiescence coupled with passive drape and infill of the earlier topography. The general onlap of SU4 onto the Top SU3 surface is consistent with a temporal waning of inversion uplift, in concordance with the absence of elevation in the Adda Ridge area in the restored map of the Top SU4 surface (Fig. 8a). This led to the gradual infill of the intervening basin, reducing the slope angles and consequently decreasing the number and importance of slope failure events. Lowering of the relief of the structural elements also affected bottom-current velocity, which became slower in the less confined setting. As the currents lost their erosional capacity, removal of sediment from the slopes and ensuing destabilization drastically diminished, leading to the decrease in large slump events.

Contour current deposits are sparsely identified in the geological record and those described are often questioned (Bein & Weiler 1976; Duan et al. 1993; Stow et al. 1998, and references therein). Contourite systems are typically formed on the lower slope of passive continental margins and abyssal plains and have a low preservation potential because of later subduction. The present study provides examples of relicts of large channels, moats and ridge-like drift systems produced by bottom currents in a deep epeiric sea located above a system of Jurassic grabens. The formation of these systems preceded two episodes marked by widespread slope failure at the Top SU2 and Top SU3 surfaces. Large- scale mass-transport processes seem to have been capable of modifying the existing topography, in some areas removing all signs of bottom-current deposition. Areas affected by active tectonism or halokinetic movements are prone to large-scale downslope mass transport. A similar effect is produced by high bottom-current velocities. Therefore, bottom-current structures formed under moderate current velocities and in tectonically quiet settings have a greater chance of being preserved in the geological record. Late Cretaceous bottom-current circulation

The present study shows that the chalk sea floor in the southern Tail End Graben had a pronounced relief during the late Turonian- Campanian caused by strong bottom currents flowing towards the SE along tectonically controlled slopes. Moats, erosional and depositional channels, levees and confined drifts represent some of the elements of bottom-current origin. Conventional contourite systems are formed along continental margins as a result of slow growth under the action of contour currents, occasionally interrupted by important hydrological events, leading to the formation of extensive unconformities (Faugeres et al. 1998, 1999). The alongslope chalk features described here are similar to modern contourites, although they were not formed in a continental margin or oceanic floor setting. The extensive epeiric ‘chalk sea’ was relatively deep and covered most of present-day NW Europe, connecting the Tethyan, Boreal and a young Atlantic realm during the Late Cretaceous-Early Palaeocene highstand. As oceanic conditions extended onto the craton, the exchange of water masses between the three basins was probably important. The bottom circulation was particularly intense at passageways formed between the numerous submerged structural highs. Overall changes in the velocity of the bottom currents affecting the ‘chalk sea’ may have been caused by eustatic sea-level changes (e.g. Surlyk & Lykke-Andersen 2007). The Top SU2 and SU3 surfaces represent regional submarine unconformities, and the latter can probably be correlated with an erosional surface in the eastern part of the Danish Basin (Esmerode et al. 2007). The geographical extent of the unconformities supports a global eustatic origin, and correlation to sealevel curves indicates that they were formed during periods of falling sea level (e.g. Haq et al. 1988). The bottom currents flowing through the ‘chalk sea’ may therefore have formed part of the global circulation system. Tidal effects cannot, however, be disregarded and the overall geostrophic circulation in the epeiric sea was probably affected by tidal currents. Eustasy was also often concealed by the tectonic activity, as reflected by the absence of current-produced structures in the Maastrichtian succession, following cessation of inversion tectonism in the area. This is clearly a local effect, as abundant evidence of bottom-current activity has been reported from the Maastrichtian of the Danish Basin (Lykke-Andersen & Surlyk 2004; Esmerode et al. 2007; Surlyk & Lykke-Andersen 2007).

The bottom currents that swept the southern Tail End Graben are interpreted to have flowed towards the SE. This is opposite to the current direction in the Danish Basin, which is interpreted to have been towards the NW along the inverted Sorgenfrei-Tornquist Zone as part of a large circulation system from the Tethys Ocean to the Boreal Sea (Lykke-Andersen & Surlyk 2004; Surlyk & Lykke-Andersen 2007). The apparently opposite current directions cannot be explained in a straightforward way. The bottom current in the central North Sea may have been part of an anticlockwise gyre of a current flowing from the Tethys Ocean to the Boreal Sea, it may have belonged to a circulation system from the Boreal Sea to the Tethys Ocean, or it may have been part of a much more complex circulation system. Continuing analyses in neighbouring regions of bottom- current directions during deposition of the NW European chalk will probably throw light on this question.

Conclusions

(1) Five seismic units, SU1-SU5, are recognized in the study area: the Cenomanian-Turonian SUl (Hidra and Plenus Marl Formations), the middle Turonian-Santonian SU2 (lower Hod Formation), the Campanian SU3 (upper Hod Formation), the Maastrichtian SU4 (Tor Formation) and the Danian SU5 (Ekofisk Formation).

(2) The inception of the Late Cretaceous Bo-Jens and Adda inversion ridges resulted in the generation of an important sea- floor relief with isobaths trending predominantly north-south and NW- SE. Intra-chalk elements of erosional and depositional nature dominate in SU2 and SU3, attaining maximum relief at two major unconformities: the Top SU2 (an intra-Hod unconformity) and the Top SU3 surfaces (Base Tor Formation). These structures are interpreted as produced by downslope, alongslope and mixed processes.

(3) Bottom-current drifts and channel fills have coherent internal reflectivity and in most cases show growth patterns in accord with Coriolis deflection. During deposition of SU2, the area was swept by a strong SE-directed flow confined in a tectonically controlled narrow passage in the northern part and fanning in a downcurrent direction. Strongest current velocities occurred at the Top SU2 surface, as inferred from the presence of the most pronounced erosional channels at this level.

(4) A SE-flowing bottom-current system seems to have persisted across the area during deposition of SU3, although with lower velocity judging by the more depositional nature of the intra-chalk structures. Current velocities probably peaked again during deposition of the uppermost SU3, as indicated by the largest relief of the erosional structures associated with the Top SU3 unconformity.

(5) At the Top SU2 and Top SU3 unconformities, the existing sea- floor topography was modified by abundant large-scale mass-waste processes, which concealed and in some instances obliterated the topographic elements produced by bottom currents. The continuing inversion of the Bo-Jens and Adda ridges initiated and maintained the instability of the slope material, triggering downslope slumping and sliding.

(6) Stratigraphic coincidence of large slumps and the largest bottom-current structures resulted from coupling of downslope and alongslope processes. Erosion produced by contour currents at the slopes decreased the stability of an area already destabilized by tectonic inversion, triggering further slope failure. This study provides outstanding examples of how rapid large-scale downslope redeposition affects the preservation of slowly formed bottom- current deposits.

(7) The draping, depositional nature of SU4 marks the change to more tranquil local sea bottom conditions as the result of a relaxation of inversion in the area, followed by infilling of the basin. This reduced the number and importance of mass-transport events, as it led to a decrease in slope angles and to a drop of the erosional capacity of the bottom currents in a less confined bathymetric setting.

(8) It is proposed that the bottom currents flowing through the southern Tail End Graben during the Late Cretaceous belonged to a geostrophic circulation system between the Tethys Ocean and the Boreal Sea. Periods of intensified bottom-current circulation probably followed eustatic sea-level falls and were recorded by the formation of unconformities of regional extent.

This study was funded by the Danish Natural Science Research Council. Maersk Oil & Gas is gratefully acknowledged for suggesting the study area as a research location, for placing seismic and well data at our disposal, and for providing work station facilities and technical support. Thanks go to C. Andersen and F. Jakobsen for their contribution to the regional geology section. H. Tirsgard, A. Uldall, L. Clausen, P Knutz and J. Cartwright are thanked for discussion and advice. We thank M. Huuse and F. J. Hernandez Molina for critical reading of the manuscript.

References

ANDERSEN, C, OLSEN, J.C, MICHELSEN, O. & NYGAARD, E. 1982. Structural outline and development. In: MICHELSEN, O. (ed.) Geology of the Danish Central Graben. Geological Survey of Denmark, Series B, 8, 9-25.

BEIN, A. & WEILER, Y. 1976. The Cretaceous Taime Yafe Formation; a contour current shaped sedimentary prism of calcareous detritus at the continental margin of the Arabian Craton. Sedimentology, 23, 511- 532.

BERGEN, J.A. & SIKORA, P.J. 1999. Microfossil diachronism in southern Norwegian North Sea chalks: Valhall and Hod fields. In: Jones, R.W. & Simmons, M.D. (eds) Biostratigraphy in Production and Development Geology. Geological Society, London, Special Publications, 152, 85-111.

BRAMWELL, N.P., CAILLET, G., MECIANI, L., JUDGE, N., GREEN, M. & ADAM, P. 1999. Chalk exploration, search for a subtle trap. In: FLEET, A.J. & BOLDY, S.A.R. (eds) Petroleum Geology of Northwest Europe: Proceedings of the 5th Conference. Geological Society, London, 911-938.

BREWSTER, J. & DANGERFIELD, J.A. 1984. Chalk fields along the Lindesnes Ridge, Eldfisk. Marine and Petroleum Geology, 1, 239-278.

BROMLEY, R.G. & EKDALE, A.A. 1987. Mass transport in European Cretaceous chalk; fabric criteria for its recognition. Sedimentology, 34, 1079-1092.

CARTER, L. & MCCAVE, I.N. 1994. Development of sediment drifts approaching an active plate margin under the SW Pacific Deep Western Boundary Current. Paleoceanography, 9, 1061-1085. DEEGAN, C.E. & SCULL, B.J. 1977. A proposed standard lithostratigraphic nomenclature for the Mesozoic of the Central and Northern North Sea. Report of the Institute of Geological Sciences, 77/25; Bulletin of the Norwegian Petroleum Directorate, 1.

DUAN, T., GAO, Z., ZENG, Y. & STOW, D.A.V. 1993. A fossil carbonate contourite drift on the Lower Ordovician palaeocontinental margin of the middle Yangtze Terrane, Jiuxi, northern Hunan, southern China. Sedimentary Geology, 82, 271-284.

ESMERODE, E.V., LYKKE-ANDERSEN, H. & SURLYK, F. 2007. Ridge and valley systems in the Upper Cretaceous chalk of the Danish Basin: contourites in an epeiric sea. In: VIANA, A.R. & REBESCO, M. (eds) Economic and Palaeoceanographic Significance of Contourite Deposits. Geological Society, London, Special Publications, 276, 265-282.

EVANS, D.J. & HOPSON, P.M. 2000. The seismic expression of synsedimentary channel features within the Chalk of southern England. Proceedings of the Geologists’ Association, 111, 219-230.

EVANS, D.J., HOPSON, P.M., KIRBY, G.A. & BRISTOW, C.R. 2003. The development and seismic expression of synsedimentary features within the Chalk of southern England. Journal of the Geological Society, London, 160, 797-813.

FARMER, C.L. & BARKVED, O.I. 1999. Influence of syn-depositional faulting on thickness variations in chalk reservoirs; Valhall and Hod fields. In: FLEET, A.J. & BOLDY, S.A.R. (eds) Petroleum Geology of Northwest Europe: Proceedings of the 5th Conference. Geological Society, London, 949-957.

FAUGERES, J.C. & STOW, D.A.V. 1993. Bottom-current-controlled sedimentation-a synthesis of the contourite problem. Sedimentary Geology, 82, 287-297.

FAUGERES, J.C, MEZERAIS, M.L. & STOW, D.A.V. 1993. Contourite drift types and their distribution in the North and South Atlantic Ocean basins. Sedimentary Geology, 82, 189-203.

FAUGERES, J.C, IMBERT, P., MEZERAIS, M.L. & CREMER, M. 1998. Seismic patterns of a muddy contourite fan (Verna Channel, South Brazilian Basin) and a sandy distal turbidite deep-sea fan (Cap Ferret system, Bay of Biscay): a comparison. Sedimentary Geology, 115, 81-110.

FAUGERES, J.C, STOW, D.A.V., IMBERT, P. & VIANA, A. 1999. Seismic features diagnostic of contourite drifts. Marine Geology, 162, 1- 38.

GOWERS, M.B. & SAEBOE, A. 1985. On the structural evolution of the Central Trough in the Norwegian and Danish sectors of the North Sea. Marine and Petroleum Geology, 1, 298-318.

HANCOCK, J.M. 1975. The petrology of the Chalk. Proceedings of the Geologists’ Association, 86, 499-535.

HANSEN, J.P.V., CLAUSEN, O.R. & HUUSE, M. 2004. 3D seismic analysis reveals the origin of ambiguous erosional features at a major sequence boundary in the eastern North Sea: near top Oligocene. In: Davies, R.J., CARTWRIGHT, J.A., LAPPIN, M. & UNDERHILL, J.R. (eds) 3D Seismic Technology: Application to the Exploration of Sedimentary Basins. Geological Society, London, Memoirs, 29, 83-89.

HAQ, B.U., HARDENBOL, J. & VAIL, P.R. 1988. Mesozoic and Cenozoic chronostratigraphy and cycles of sea-level change. In: WILGUS, CK., HASTINGS, BJ., POSAMENTIER, H., VAN WAGONER, J.C, ROSS, CA. & KENDALL, CG.ST.C. (eds) Sea-level Changes-An Integrated Approach. SEPM Special Publications, 42, 71-108.

HARDMAN, R.F.P. 1982. Chalk reservoirs of the North Sea. Bulletin of the Geological Society of Denmark, 30, 119-137.

HATTON, I.R. 1986. Geometry of allochthonous Chalk Group members, Central Trough, North Sea. Marine and Petroleum Geology, 3, 79-98.

KENNEDY, W.J. 1980. Aspects of chalk sedimentation the southern Norwegian offshore. In: The Sedimentation of the North Sea Reservoir Rocks. Norsk Petroleumsforening, Geilo, 1-29.

KENNEDY, W.J. 1987a. Late Cretaceous and early Palaeocene Chalk Group sedimentation in the greater Ekoftsk area, North Sea Central Graben. Bulletin des Centres de Recherches Exploration-Production Elf-Aquitaine, 11, 91-126.

KENNEDY, W.J. 1987b. Sedimentology of Late Cretaceous-Paleocene chalk reservoirs, North Sea Central Graben. In: BROOKS, J. & GLENNIE, K.W. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 469-481.

KENNEDY, W.J. & JUIGNET, P. 1974. Carbonate banks and slump beds in Upper Cretaceous (Upper Turonian-Santonian) of Haute-Normandie, France. Sedimentology, 21, 1-42.

KNUTZ, P.C. & CARTWRIGHT, J.A. 2003. Seismic stratigraphy of the West Shetland Drift: implications for late Neogene pal[a]eocirculation in the Faeroe-Shetland gateway. Paleoceanography, 18, 1-17.

LOCKER, S.D. & LAINE, E.P. 1992. Paleogene-Neogene depositional history of the middle United States Atlantic continental rise-mixed turbidite and contourite depositional systems. Marine Geology, 103, 137-164.

LYKKE-ANDERSEN, H. & SURLYK, F. 2004. The Cretaceous-Palaeogene boundary at Stevns Klint, Denmark: inversion tectonics or sea-floor topography? Journal of the Geological Society, London, 161, 343- 352.

MASSE, L., FAUGERES, J.C. & HROVATIN, V. 1998. The interplay between turbidity and contour current processes on the Columbia Channel fan drift, southern Brazil Basin. Sedimentary Geology, 115, 111-132.

MOLLER, J.J. 1986. Seismic structural mapping of the Middle and Upper Jurassic in the Danish Central Trough. Geological Survey of Denmark, Series A, 13.

MOLLER, J.J. & RASMUSSEN, E.S. 2003. Middle Jurassic-Early Cretaceous rifting of the Danish Central Graben. In: INESON, J.R. & SURLYK, F. (eds) The Jurassic of Denmark and Greenland. Danmarks og Gronlands Geologiske Undersogelse (GEUS), Copenhagen, 247-264.

MUNNS, J.W. 1985. The Valhall Field; a geological overview. Marine and Petroleum Geology, 2, 23-43.

OAKMAN, C.D. & PARTINGTON, M.A. 1998. Cretaceous. In: GLENNIE, K.W. (ed.) Petroleum Geology of the North Sea; Basic Concepts and Recent Advances. Blackwell Scientific, Oxford, 294-349.

QUINE, M. & BOSENCE, D. 1991. Strafai geometries, facies and sea- floor erosion in Upper Cretaceous chalk, Normandy, France. Sedimentology, 38, 1113-1152.

SIKORA, P.J., BERGEN, J.A. & FARMER, C.L. 1999. Chalk palaeoenvironments and depositional model, Valhall-Hod fields, southern Norwegian North Sea. In: JONES, R.W. & SIMMONS, M.D. (eds) Biostratigraphy in Production and Development Geology. Geological Society, London, Special Publications, 152, 113-137.

STANLEY, D.J. 1993. Model for turbidite-to-contourite continuum and multiple process transport in deep marine settings; examples in the rock record. Sedimentary Geology, 82, 241-255.

STOW, D.A.V., FAUGERES, J.C, VIANA, A.R. & GONTHIER, E. 1998. Fossil contourites; a critical review. Sedimentary Geology, 115, 3- 31.

SURLYK, F. & LYKKE-ANDERSEN, H. 2007. Contourite drifts, moats and channels in the Late Cretaceous chalk of the Danish Basin. Sedimentology, 54,405-422.

SURLYK, F., DONS, T., CLAUSEN, C.K. & HIGHAM, J. 2003. Upper Cretaceous. In: EVANS, D., GRAHAM, C, ARMOUR, A. & BATHURST, P. (eds) The Millennium Atlas: Petroleum Geology of the Central and Northern North Sea. Geological Society, London, 213-233.

VEJBAEK, O.V. 1986. Seismic stratigraphic and tectonic evolution of the Lower Cretaceous in the Danish Central Trough. Geological Survey of Denmark, Series A, 11.

VEJBAEK, O.V. & ANDERSEN, C. 1987. Cretaceous-early Tertiary inversion tectonism in the Danish Central Trough. Tectonophysics, 137, 221-238.

VEJBAEK, O.V. & ANDERSEN, C. 2002. Post mid-Cretaceous inversion tectonics in the Danish Central Graben-regionally synchronous tectonic events? Bulletin of the Geological Society of Denmark, 49, 129-144.

WATTS, N.L., LAPRE, J.E, VAN SCHUNDEL-GOESTER, F.S. & FORD, A. 1980. Upper Cretaceous and lower Tertiary chalks of the Albuskjell area, North Sea; Deposition in a slope and a base-of-slope environment. Geology, 8, 217-221.

Received 11 September 2006; revised typescript accepted 16 May 2007.

Scientific editing by Gary Hampson

E. V. ESMERODE1, H. LYKKE-ANDERSEN2 & F. SURLYK3

1 DONG Energy E&P, Agent Alle 24-26, DK-2970 Hersholm, Denmark (e- mail: esesm@dongenergy.dk)

2 Department of Earth Sciences, University of Aarhus, Hoegh Guldbergs Gade 2, DK-8000 Arhus C, Denmark

3 Department of Geography and Geology, Geology Section, University of Copenhagen, Oster Voldgade 10, DK-1350

Copenhagen K, Denmark

Copyright Geological Society Publishing House Jan 2008

(c) 2008 Journal of the Geological Society. Provided by ProQuest Information and Learning. All rights Reserved.