Six basement-penetrating drill sites are located along the EG63 transect on the Southeast Greenland Margin (Fig. 1). Five sites are located near the landward termination of the volcanic complex, whereas Site 918 is located within the main SDR unit. Both conventional MCS and high-resolution seismic data are available on the margin (Fig. 4, Fig. 5) (Larsen and Jakobsdóttir, 1988; Larsen and Saunders, 1998).

Two well-defined SDR units are imaged on profile GGU-81-08, separated by a 20-km-wide zone with more disrupted intrabasement reflectivity (Fig. 4). Distinct reflectors are typically 5 to 10 km long, with a vertical extent of up to 1.5 s. The reflectors terminate upward at, or merge with, the top-basement reflector. Site 918 was drilled 121 m into the volcanic basement, recovering 21 basaltic lava flows (Larsen, Saunders, Clift, et al., 1994). Deep, subaerial weathering affected the top three units, while several recovered red flow tops within the entire basement interval indicate persistent subaerial conditions during construction of the lava pile.

The top-basement reflector can be traced landward from the SDR until it intersects the seafloor near the western end of profile GGU-81-08 (Fig. 4). Segmented, low-frequency reflectors characterize the interval below the top-basement reflector landward of the SDR. Both the seafloor and the top-basement represent high acoustic impedance contrast boundaries, causing numerous short- and long-path multiples and converted waves in the shallow-water part of the profile. Much of the intrabasement reflectivity is therefore interpreted as nonprimary energy. The relatively smooth, high-amplitude nature of the top-basement reflector and its position landward of the SDR lead us to define this seismic facies unit as the Landward Flows.

High-resolution seismic data along the inner part of the EG63 transect show characteristics similar to the conventional MCS data (Fig. 5). A smooth, high-amplitude top-basement reflector is easily identified, showing characteristic irregular down-to-northwest stepping at the seafloor and between Sites 989 and 917. Deeper intrabasement reflectivity is segmented and irregular, with an overall seaward dip. A weak basal reflector can be identified in some places but is largely obscured by strong seafloor multiples. The reflector was drilled at Site 917 where it was found to be the base of the volcanic complex (Larsen, Saunders, Clift, et al., 1994). The basement is divided into two slightly different units southeast of Site 990, with the upper unit having a somewhat better-defined internal reflector pattern. Landward, the reflector pattern clearly changes below and northwest of Site 989, where well-defined, steeply landward-dipping reflectors are imaged down to about 0.9 s.

Site 917 penetrated 779 m of subaerially emplaced basalts and dacites of a late Paleocene age and terminated in a steeply inclined metamorphosed sedimentary section of possibly Paleocene age (Larsen, Saunders, Clift, et al., 1994; Larsen and Saunders, 1998). Wireline log data reveal a similar characteristic impedance structure as on Site 642, with low-velocity and low-density lava tops (Vp ~2.5-3 km/s, ~2.2-2.4 g/cm³) and high-velocity lava interiors (Vp ~5-6 km/s, ~2.8-3.0 g/cm³) (Planke and Cambray, 1998). At Site 917, the top-basement reflector clearly represents the top of the lava pile, while the southeastward-dipping reflector at 1.13 s is interpreted as the base of the volcanics (Fig. 5). Few coherent reflectors are identified within the volcanic sequence, and no primary reflectors can be interpreted with confidence.

The four shallow basement sites along the inner part of the EG63 transect (Sites 915, 916, 989, and 990) all terminated within the volcanic pile, and numerous subaerially emplaced basaltic lava flows were recovered from them. No intrabasalt reflectors were penetrated at any of these sites. Site 989 almost reached a reflector at 0.68 s depth (Fig. 6) but was terminated at 84 mbsf. Two flow units were drilled, the upper was a 69-m-thick, compound, massive basalt, while the lower was >11-m-thick massive lava. The negative-polarity reflector just below the termination of the hole is interpreted to represent the base of this flow unit (Fig. 6). The nature of the underlying rocks is undetermined but may correspond to low-velocity, thin lava flows, or pre-breakup sediments as drilled at Site 917 and interpreted to be present farther west. The great thickness of the composite flow unit may be related to lava ponding. No weathered top was recovered from this unit. It was probably removed by glacial erosion, leaving the more resistant interior. The southwest end of the flow unit is clearly identified at shotpoint (SP) 450 (Fig. 6) as a 30-m-high erosional escarpment. The northwestward stepping of the top-basement reflector on Fig. 5 is interpreted to represent similar erosional features, forming on of the characteristic "trapps" (stairs) frequently found in flood-basalt provinces. These staircase-like erosional features are typically developed in interlayered hard and soft rocks, such as flood-basalt constructions with soft lava tops and hard lava interiors.

One well-defined SDR unit is imaged on the composite EG66 transect (Fig. 7). The unit is ~50 km wide, with a high-amplitude, undulatory top-basement reflector. The top-basement reflector is locally onlapping or interfering with intrabasement reflectors (e.g., near SP 2000; Fig. 7). Seaward, the top-basement reflector becomes more segmented and chaotic, forming a buildup near the southeastern termination of the profile. A basal reflector is identified below the inner part of the SDR (GGU-82-02, SP 1000-1400; Fig. 7). Internal reflectors are characterized by an overall divergent/arcuate pattern with short, segmented reflectors continuing down to about 1.5 s below the top-basement. The SDR unit is clearly not continuous along the strike of the margin, as no well-defined SDR can be identified on the seaward continuation of profile GGU-82-01 ~50 km to the north (Fig. 1, Fig. 7).

A wide Landward Flows unit is primarily identified based on the nature of the top-basement reflector (Fig. 7). This reflector is a high-amplitude, smooth event identified landward of the SDR. The reflector becomes rougher and more segmented just before intersecting the seafloor near the northwest termination of profile GGU-82-01. The top-basement reflector identified on high-resolution data (Fig. 8) shows a down-to-northwest stepping pattern interpreted as erosional escarpments (trapps) along the EG63 transect. The intrabasement reflectivity is dominated by nonprimary energy related to short- and long-path multiples and converted waves, and hardly any coherent events can confidently be interpreted as primary reflectors. The base of the facies unit is only identified near the SDR/Landward Flow transition (GGU-82-02, SP 1500-1800) and below the region where the unit is subcropping at the seafloor (Fig. 7).

The Rockall Margin is conjugate to the Southeast Greenland Margin, as shown by the well-defined symmetric magnetic lineation pattern to at least Anomaly 22 time on both sides of the spreading axis (Fig. 1) (Skogseid et al., in press). The seismic data from the Edoras Bank and Hatton Bank margins reveal characteristic SDR on most profiles (White et al., 1987; Barton and White, 1997a, 1997b). Additionally, a several-hundred-kilometer-wide zone landward of the SDR is covered by breakup-related volcanic extrusive and shallow intrusive rocks (Joppen and White, 1990; Neish, 1993).

The Hatton Bank Margin profiles are approximately conjugate to the Southeast Greenland Margin EG63 transect (Fig. 1). Interpretative line drawings show several major similarities and differences on the conjugate margins (Fig. 9A). The Hatton Bank Margin profile shows two SDR units separated by an Outer High (Fig. 9A). Two SDR units are also identified on EG63. The characteristic Outer High is not identified (Fig. 4) but likely corresponds to the hummocky-type basement mapped seaward of magnetic Anomaly 24B by Larsen and Jakobsdóttir (1988). The external shape and internal reflector characteristics of the SDR are quite different. The Hatton Bank Margin profile is dominated by a wedge-shaped unit with divergent reflector segments. In contrast, the EG63 transect reveals a more rhombic-shaped unit, where the arcuate intrabasement reflector segments are dominantly subparallel.

Landward, reflection data on both margins have high noise levels related to multiple and converted energy in relatively shallow-water environments. The subsequent post-breakup margin development has further influenced the ability to image intrabasement structures. In particular, the deposition of prograding glacial sediments on the Southeast Greenland Margin has locally degraded the intrabasement images (e.g., near SP 750 on profile GGU-81-08; Fig. 4). Further, erosion has reduced the initial landward extent of the volcanics off East Greenland.