During Leg 152, six sites were drilled on a transect at ~63ºN across the Southeast Greenland volcanic rifted margin (Larsen, Saunders, Clift, et al., 1994) (Fig. 2). Volcanic basement material was recovered at three of these sites; the most inland hole (Hole 917A) penetrated ~779 m of lava from the featheredge of the SDRS, and the more seaward holes (Holes 915A and 918D), placed within the main SDRS series, penetrated ~20 and ~110 m of lavas, respectively. The lava succession at Hole 917A has been subdivided into a Lower, Middle, and Upper Series on the basis of flow morphology, petrography, and geochemistry (Larsen, Saunders, Clift, et al., 1994). The Lower and Middle Series tend to consist of thick flows (>10 m) that display evolved basaltic to andesitic compositions with high Ce/Y, elevated ⁸⁷Sr/⁸⁶Sr, and/or low ¹⁴³Nd/¹⁴⁴Nd relative to Bulk Earth, pointing to substantial crustal contamination and fractional crystallization in large magma reservoirs at depth (Fitton et al., 1998a; Fitton et al., 1998b; Larsen, Saunders, Clift, et al., 1994; Saunders et al., 1998). We refer in the following to these lavas as continental type. The Upper Series, which is physically separated from the underlying Middle Series by a 70-cm-thick sandstone, records a marked change in the general eruption style from the thick flows in the Lower and Middle Series to thinner pahoehoe flows (Larsen, Saunders, Clift, et al., 1994). The lavas of the Upper Series are primitive olivine accumulative picrites and basalts, some originating from true high-MgO liquids, showing light rare earth element (REE) depletion and other compositional characters close to oceanic basalts (Fitton et al., 1998a; Fitton et al., 1998b; Larsen, Saunders, Clift, et al., 1994; Thy et al., 1998). In the following, these lavas are referred to as transitional-to-oceanic type. The more seawardly occurring lavas recovered at Holes 915A and 918D have restricted basaltic compositions with 6.8-8.4 wt% MgO and trace element compositions similar to light REE-depleted basalts in Iceland (Fitton et al., 1998b) and mid-ocean ridge basalts. We refer to these basalts as oceanic type. Many of the basalts from Holes 915A and 917A show isotopic evidence for slight crustal contamination, but the basalts at Hole 918D are unmodified by continental crust and have isotopic compositions akin to the most depleted basalts in Iceland (Fitton et al., 1998a). Although the degree of crustal contamination decreases drastically from the Lower and Middle Series to the Upper Series at Site 917, and decreases further at Site 915, it is evident that the basalts at these two sites erupted through continental crust (Fitton et al., 1998a; Larsen and Saunders, 1998; Saunders et al., 1998), which is consistent with the seismic crustal structure at this transect (Larsen et al., 1998). The temporal change in the amount of crustal contamination is accompanied by changes in trace element ratios that are sensitive to the mean depth and degree of mantle melting, but relatively insensitive to contamination. For example, a secular decrease in Zr/Sc and Hf/Lu from the Lower and Middle Series (continental type), through the Upper Series (transitional-to-oceanic type) to the lavas at Sites 915 and 918 (oceanic type) point to a decrease in the mean depth of mantle melting accompanied by an increase in the degree of melting related to the thinning and stretching of the continental lithosphere during continental breakup (Fitton et al., 1998b; Fram et al., 1998).

The age of the volcanic succession drilled during Leg 152 is constrained from radiometric ⁴⁰Ar-³⁹Ar dating, paleomagnetic data, and palynological data. The results of ⁴⁰Ar-³⁹Ar geochronology for samples seemingly uncontaminated with excess ⁴⁰Ar yield virtually indistinguishable ages between 61.4 ± 1.0 and 60.1 ± 0.2 Ma for six lavas and tuffs of the Lower and Middle Series, although two other samples of the Middle Series gave outlying ages of 62.4 ± 0.5 and 62.9 ± 0.4 Ma (Sinton and Duncan, 1998; Werner et al., 1998) (all previously published argon ages are adjusted to 27.84 Ma for the FCT-3 biotite monitor to be consistent with ages reported here and the time scale of Berggren et al., 1995). The lavas of the Lower and Middle Series are reversely magnetized but include two magnetic excursions to low negative or positive inclinations that are interpreted as cryptochrons, pointing to correlation with C26r (60.9-57.9 Ma), consistent with the majority of the argon ages (Larsen and Saunders, 1998; Vandamme and Ali, 1998). Reliable radiometric ages were not obtained for the reversely magnetized lavas of the Upper Series at Holes 917A, 915A, or 918D. The ages of these lavas, however, are constrained by (1) the age of the sediments overlying Site 918 dated to C23r (52.4-51.7 Ma) from palynological data (Jolley, 1998), (2) a concordant ⁴⁰Ar-³⁹Ar age of 52.5 ± 0.8 Ma for a trace element-enriched sill crosscutting the overlying sediments at Site 918 (Sinton and Duncan, 1998), and (3) the recognition of spreading Anomaly C24n.3n (53.3-52.9 Ma) <~50 km seaward of Site 918 (Larsen and Saunders, 1998). This proves that the main SDRS lavas represented by the Upper Series at Sites 917, 915, and 918 are of C24r age or older (i.e., older than ~53 Ma).

One of the main goals of Leg 163 was to return to the transect at 63ºN to retrieve additional material of the volcanic succession to complement the results of Leg 152. Site 990 was designed to recover material from the stratigraphic section between Hole 915A and the Upper Series of Hole 917A (Fig. 2) aiming, among other things, to obtain suitable material for ⁴⁰Ar-³⁹Ar dating. Site 989, the most westerly site at the 63ºN transect, was planned to be drilled down to the most landward volcanics within the featheredge of the SDRS and to reach the basement underlying the volcanics. According to the seismic stratigraphic interpretation (Fig. 2), the lava succession at Site 989 predates the Lower Series at Site 917 and, potentially, could represent the oldest volcanics at the 63ºN transect.