Preliminary drill sites were selected from air gun seismic Line DLC9724 (Fig. F2) acquired by the Dana cruise (Hopper et al., 1997). Final site selection was made after examination of shipboard sparker profiles following the DLC9724 navigational track. The apparent dip of imaged seaward-dipping reflectors along the proximal portion of DLC9724 is between 15° and 26°, whereas at the seaward end of the line, where the volcanic basement dips below the sediment wedge, the apparent dip decreases to between 10° and 15° (Figs. F3, F4). We interpret the trend of seaward-decreasing dip as reflecting a deeper erosional level within the inner part of seismic Line DLC9724 and a shallower erosional level along the outer part of the line, consistent with the crustal accretion model of Pálmason (1980). Assuming an average dip of 18° for the lavas exposed along the ~30-km-long DLC9724 seismic line implies that as much as 10 km of volcanic stratigraphy may be accessible at the seabed between the coast and the outer sediment wedge. Normal faulting may result in stratigraphic repetition, but large offset faults are not recognized in the seismic data. Nielsen (1978) noted from field mapping that although major normal faults cutting the onshore lava succession were absent, closely spaced small-offset normal faults were recognized in the flexure zone along the coast leading to accentuated dips and, likewise, stratigraphic thickness. Therefore, we consider our estimate of stratigraphic thickness exposed at the seabed as an upper limit, with the additional caveat that if the Pálmason (1980) model applies, the shingled structure of the seaward-dipping reflectors precludes the lava pile from ever reaching a vertical thickness equal to the stratigraphic thickness.
Air gun seismic Line DLC9724 (Fig. F2) and sparker seismic Line DLC99-2 (Fig. F3) show potential drilling sites close to the coast. Two sites selected from the inner area were drilled (Sites SEG07 and SEG08). Both sites failed to recover basement. Onboard processing of the sparker seismic data improved the resolution at these sites, and it was concluded that the inner shelf had a considerable cover of glaciomarine sediments, possibly between 5 and 10 m thick in places. Based on our re-evaluation of the seismic data, we concluded that further drilling in the inner area would be futile using British Geological Society's Rockdrill fitted with a 3-m drill string.
The location of the outer area along the DLC99-1 sparker seismic line is shown in Figure F4. A series of topographic highs, presumably of basement, were selected as targets for drilling. The highs typically have gentle slopes dipping south-southeast and more steeply north-northwest–dipping escarpments facing the opposite direction toward the coast. The latter slopes often cause strong diffraction hyperbolae in the seismic data, which hide the true reflection from the north-northwest–facing steep slopes and suggest that they are made up of acoustically competent material. Some sediment infill, probably glaciomarine, is seen between the highs. This is visible on the DLC9724 air gun line and in particular on the DLC99-1 sparker line (Fig. F4). On the sparker line, the maximum thickness of the sediment is ~40 m, although it is generally much less. Clear reflections are rarely seen between the glaciomarine cover and the underlying basalt. The sediments, therefore, are primarily recognized because of the shape and internal seismic structure of the fills. In practice, we were only able to resolve sediment covers that exceeded 3 m from the sparker seismic data. This permitted us to eliminate many target areas that had excessively thick glaciomarine covers, but it did not guarantee sediment-free conditions at a site judged most promising based on the sparker data.
The more gently south-southeast–dipping slopes of the basement highs dip at 4° to 6° and are interpreted as representing more or less stratigraphically controlled erosional surfaces. These most likely are flow boundaries along which glacial erosion has carved out the less competent flow tops and thin interbedded sediments or soil horizons. Similarly, the north-northwest–facing escarpments are interpreted as "trap" morphology with steep erosional surfaces following original subvertical fractures. The sparker seismic data also show that at least some of the steep landward escarpments are blanketed with glaciomarine sediment, as are the flat terraces showing strong seismic diffractions.
A large number of possible drill sites on basement highs were selected on the sparker seismic Line DLC99-1 within a ~10-km-wide region ~40 km from the coast. Of these, 16 sites were drilled within a 5-km-wide zone (Fig. F4). Two marked basement highs of the type described above with long, gentle south-southeast–dipping slopes and steep escarpments to the north-northwest are centrally located within the drilling region.