A regional grid of high-quality MCS data on the Møre Margin has provided more detailed structural information of the extrusive breakup complex here than presently available offshore elsewhere in the northeast Atlantic (Alvestad, 1997). However, volcanics have never been drilled on the margin. The drilling results from elsewhere in the northeast Atlantic are important for constraining the seismic volcanostratigraphic interpretation of the Møre Margin profiles, whereas the high-quality data from the Møre Margin can provide constraints for understanding the volcanic history of less well-imaged margin segments. In addition, the Møre Margin and EG63 transects are situated at a similar distance from the Iceland plume trail (Fig. 1). Previous studies have suggested that there is an overall relationship between melt production and the distance to the main plume source (Barton and White, 1997b), providing an additional reason for comparing the two margin transects.

Several distinct seismic facies units are identified on the Jan Mayen Ridge and Møre Margin profiles (Fig. 9B)-the SDR, Outer High, and Landward Flows are interpreted as being on the Southeast Greenland-Hatton Bank transect. However, Lava Delta, Inner Flows, and Volcanic Basin units are also identified below and landward of the SDR. Finally, a set of segmented, steeply landward-dipping reflectors are interpreted below the Outer High and within the main part of the SDR.

The seismic facies units on the Møre Margin have been interpreted in terms of volcanic eruption and emplacement environment (Table 1) (Alvestad, 1997). Landward Flows are penetrated at the deep Sites 917 and 642, where they represent subaerially erupted and emplaced volcanics. The Lava Delta unit has not been penetrated by any drill holes but is interpreted to represent a volcaniclastic lava delta formed as subaerially erupted lavas reached a paleo-shoreline. This interpretation is compatible with the hypothesis that the Faeroe-Shetland Escarpment is a coastal feature (Smythe et al., 1983). Hydroclastite lava delta constructions are found exposed, for example, on Nuussuaq Island off western Greenland (Pedersen et al., 1996), and are well developed along the coastline of Hawaii (Moore et al., 1973). The Inner Flows are interpreted as dominantly hydroclastite deposits consisting of a mixture of massive and fragmented basalts. Together, the three facies units form a coarse-grained Gilbert-type clastic delta, where the Landward Flows, Lava Delta, and Inner Flows represent the top-set, fore-set, and bottom-set facies of the delta.

Two SDR units are identified on the transect across the Møre Margin, where the seaward one is overlaid by an internally stratified Outer High. Based on the drilling results elsewhere in the northeast Atlantic, the SDRs are interpreted as dominantly subaerially emplaced and erupted basalts, while the stratified Outer High is interpreted as consisting of shallow-marine volcaniclastics. The deeper landward-dipping reflectors are not drilled anywhere but are interpreted as feeder dikes that likely followed weakness zones such as fault planes. The dikes may originally have been emplaced as near-vertical feeders but have acquired their landward dip by synconstructional margin flexuring and subsidence.

On the Jan Mayen Ridge, the conjugate seismic profile is located near the southernmost part of the mapped SDR (Fig. 9B) (Åkermoen, 1989). A thin SDR unit is identified here, terminating near a broad Outer High. The apex of the SDR was later faulted, probably occurring when the Jan Mayen Ridge broke off East Greenland during the mid-Tertiary (Gudlaugsson et al., 1988; Eldholm et al., 1990). Landward-dipping reflector segments below the SDR, similar to those found on the Møre Margin, have been interpreted as a feeder dike system on the Jan Mayen Ridge (Åkermoen, 1989).

The nature of the intrabasement reflectivity on volcanic rifted margins is largely unconstrained (Planke and Eldholm, 1994; Barton and White, 1997b; Smallwood et al., 1998). Currently, only the Landward Flows unit has been sampled by several deep drill holes. Only a few, very shallow holes have been drilled into the SDR, and only one hole has recovered material from the Outer High. None of the other seismic facies units have been sampled. Moreover, none of the northeast Atlantic drill holes have penetrated any well-defined intrabasalt reflectors, with the exception of reflector K, representing the lower boundary of the SDR on the Vøring Margin.

Internal reflectors in the SDR may represent ponded or thick lava units, interference phenomena between several flows of similar thickness, or very fragmented or hydroclastic units deposited in a shallow-marine environment. It has further been suggested that deep marine-emplaced flood-basalt constructions may be imaged as SDR (Planke et al., 1995), but this hypothesis cannot be addressed by the existing borehole data. Our preferred interpretation of the nature and emplacement environments for the main seismic facies units is summarized in Table 1. However, the interpretational freedom in these volcanic provinces is large. Field observations and data from other volcanic provinces are therefore essential information necessary to integrate into the interpretation of volcanic complexes on rifted margins.

Mapping of seismic facies units on margin segments with relatively dense profile coverage shows that the intrabasement reflectors are continuous for several kilometers along strike (e.g., Barton and White, 1997a) and that the facies units are clearly three-dimensional features (Alvestad, 1997). The Outer High is commonly only mappable several tens of kilometers along strike, and the individual SDR seismic facies units are often discontinuous along a margin segment. It is therefore difficult to assess how representative a margin transect is without a relatively dense seismic grid. The Jan Mayen Ridge and Møre Margin transects are the best constrained conjugate margin pair, but the seismic facies interpretation of the Hatton Bank and EG63 transects are also constrained by nearby profiles (Fig. 9).

The conjugate margin transects are symmetric in the sense that SDR are present on both sides of the ocean, but the structure of the volcanic complexes is quite different. On the Jan Mayen Ridge profile, a relatively small extrusive complex is identified, consisting mainly of one SDR, whereas an extensive volcanic complex is interpreted on the Møre Margin profile, consisting of two SDR units and various other volcanic seismic facies units (Fig. 9B). A somewhat different asymmetry is found on the EG63-Hatton Bank conjugate transects. Here, the thickness and width of the SDR is much larger on the EG63 transect than on the Hatton Bank Margin. In contrast, the total width of the volcanic zone is much greater on the Rockall Margin extending almost 1000 km landward of the SDR (e.g., Joppen and White, 1990).

Several different models have been proposed for the formation of the breakup-related volcanic extrusive complexes. These models can be divided into two main categories-those that relate the construction of the volcanic complex to infilling and capping of a rifted terrain (e.g., Hinz, 1981; Eldholm, Thiede, Taylor, et al., 1989; Planke and Eldholm, 1994) and those that are focused on the volcanics being formed during a phase of subaerial seafloor spreading (e.g., Mutter et al., 1982; Larsen and Saunders, 1998). The drilling data have not been able to distinguish between these end-member models, and both processes may actually form constructions imaged as SDR. Sites 642 and 917 provide direct evidence that the Landward Flows are underlain by extended continental crust (Eldholm, Thiede, Taylor, et al., 1989; Larsen and Saunders, 1998). The seaward continuation of the deep reflector patterns below the SDR suggests that at least the inner parts of the SDR are underlain by continental material (Figure 9B). However, the Site 918 results suggest that the entire crust below the Inner SDR on the Southeast Greenland Margin is of igneous nature (Larsen and Saunders, 1998). Volcanic margin models need to satisfactorily explain the conjugate margin asymmetry and lateral variation in volume and reflection characteristics of the breakup volcanic complexes as revealed by Figure 9. Models further must explain the consistent seaward dip of the SDR packages, even in provinces where ridge-axis jumps are inferred such as on the Vøring Margin (e.g., Skogseid and Eldholm, 1987) and the discontinuous, segmented magnetic anomaly pattern along the margin. We believe that the variations in the seismic reflection pattern of the volcanic complexes point toward a structural control for the emplacement of a major part of the extrusive volcanics. However, voluminous volcanic episodes during sea-floor spreading can locally form subaerial or deep marine basaltic sheet-flow constructions also imaged as SDR.

The conjugate margin seismic reflection profiles provide little direct evidence about the nature of the deep crust and the location of the continent-ocean boundary as a reflection Moho is not identified and almost no deep crustal reflectors are present. Magnetic data are of limited use in determining the exact location of the continent-ocean boundary as the oldest magnetic lineations are discontinuous and segmented in the vicinity of the transects and located seaward of the main breakup-related volcanic complex (Fig. 1, Fig. 9). On the conjugate margin transects, the oldest magnetic anomalies, 24A and 24B, are both located seaward of the SDR on the Jan Mayen Ridge and Møre Margin profiles (Fig. 1, Fig. 9). Similarly, Anomaly 24n.3n is located near the seaward termination of the SDR on transect EG63, whereas the outer part of the SDR is questionably related to Anomaly 23 on the Hatton Bank Margin (Fig. 9A). Magnetic modeling on the Vøring Margin shows that a 4- to 5-km-thick volcanic cover is sufficient to explain the magnitude of the magnetic anomalies (Schreckenberger, 1997). Magnetic data, thus, may not provide any direct information about the nature of the deeper crust, which may be entirely igneous or extended and intruded continental fragments.