Site 989 is located 23 nmi east of the East Greenland coast. It is one of the three drill sites planned for the southern drilling transect EG63 (Figs. 2, 4). Drilling at Leg 152 Sites 915 and 917 had penetrated a thick lava sequence that recorded development from an early continental crust-contaminated volcanism, through transitional picritic and tholeiitic volcanism contemporaneous with breakup, into steady-state oceanic volcanism. Site 989 was selected to penetrate and sample the very oldest lavas of the SDRSs that overlie the breakup unconformity and underlying, layered prerift crust. The primary drilling objectives at this site were to (1) determine the stratigraphy, composition, age, and eruption environment of the volcanic rocks above the breakup unconformity; (2) determine the nature and age of the breakup unconformity; and (3) determine the nature and deformation of the continental basement and/or prerift sediments beneath the volcanic sequence.
Lithologic Unit I is a thin layer (0-4 mbsf) of Quaternary(?) glaciomarine sediments unconformably overlying basaltic basement (igneous Units 1 and 2). The only material recovered consists of discrete rock fragments, including gneiss, aphyric basalts/metabasalts, quartzite, and dolerite. The lithologies of these clasts are consistent with an ice-rafted origin, even though no finer-grained matrix was recovered. The relatively weak nature of the sediments recovered at Site 989 suggests that these are glaciomarine deposits, rather than overcompacted glacial tills.
Two igneous flow units were recognized in the core recovered from the interval 4-84 mbsf (Hole 989B). From seismic data, these are interpreted to lie stratigraphically below the lavas drilled at Site 917 and represent the oldest part of the SDRS. Igneous Unit 1 is at least 69 m thick, the thickest lava flow yet reported from an SDRS. It is notable for its constant grain size, constant vesicularity, high mesostasis content, and repeated bands showing quench textures. These features indicate rapid cooling during solidification throughout the lava flow. We interpret Unit 1 as a compound lava flow consisting of numerous individual flow units 0.1-10 m thick. The large number of thin flow units, together with the absence of sharp flow contacts, may indicate both (1) proximity to the eruptive vent and (2) rapid eruption of the entire lava flow. The observed decrease in maximum flow unit thickness upward in the lava may reflect an exponentially diminishing eruption rate with time.
Unit 1 is essentially aphyric. The groundmass consists of plagioclase, augite, magnetite, trace olivine, and mesostasis. Clay alteration is total for both mesostasis and olivines whereas plagioclases and augites are generally fresh. The textures vary between two extremes: (1) a very fine grained rock with quench textures of spherulitic, acicular, and skeletal plagioclases (and sporadic augites) within a vesicular and mesostasis-rich matrix and (2) a "normal" fine-grained intersertal, intergranular to variolitic and sub-ophitic rock with large disseminated vesicles (up to 20% and up to 4 mm across). The transition between quenched and normal textures may be sharp (internal flow boundaries) or gradational.
Igneous Unit 2 is porphyritic with phenocrysts of plagioclase, augite, and trace olivine in a very fine grained matrix. Olivine phenocrysts occur as individual disseminated grains that are now totally altered to clay. Plagioclase and augite phenocrysts are fresh, commonly strongly zoned and resorbed (plagioclase), and in glomerocrystic clusters. The groundmass has a seriate texture defined by microphenocrysts of very elongate plagioclases (4%-5% of groundmass). Stubby olivines (trace to 2%) and anhedral augites (<1%) also form microphenocrysts. The groundmass (up to 0.2-mm grains) consists of plagioclase laths, equant augites, euhedral to skeletal magnetite, and mesostasis in an intersertal/intergranular to variolitic texture.
The two units recovered at Site 989 are both strongly depleted in a number of incompatible elements such as Zr, Nb, Ti, and P and presumably melted from a depleted mantle source (Fig. 5). Both lava flows are composed of evolved basalt, which implies storage in a magma chamber underlying this part of the volcanic succession. Similar crustal magma chambers were invoked for the lavas in the Lower Series at Site 917, which have assimilated a Sr- and Ba-rich crustal component. The low Sr and Ba contents in the Site 989 lavas preclude a direct correlation with the Lower Series in Site 917. In contrast, the lavas at Site 989 have either escaped crustal contamination or have assimilated a crustal component very different from that which contaminated the lavas in Site 917.
Physical property measurements (P-wave velocity, bulk and grain densities, porosity) of Unit 1 are quite constant with depth and correlate well between the holes. The transition from Unit 1 to Unit 2 is clearly recognized, with average P-wave velocity increasing from 5.2 to 6.0 km/s, density increasing from 2.8 to 3.0 g/cm3, and porosity decreasing from 12% to 2%. The 3% average reduction between P-wave velocity measurements performed on minicores vs. half-rounds may be due to drilling-induced fracturing of the outer edges of the cores.
Deformation of the cored sequence at Site 989 is principally in the form of brittle fracturing, manifest as veining and jointing. Veining is commonly present as two conjugate sets, one postdating the other, but both infilled with a combination of green clays and zeolites. Measured dips within Unit 1 for flow banding and other forms of textural variation, such as vesicular layers, are scattered but are concentrated between 15û and 45û. These features are interpreted to be chilled surfaces of flow units within a compound flow.
Unit 1 recovered in both holes appears to carry a normal magnetic polarity with a mean inclination of 68.4°. If confirmed, this will be the first flow of normal polarity reported from the East Greenland margin, and current stratigraphic evidence correlates this normal event with Chron C27n. Unit 2 appears to contain both normal (top) and reversed polarity. The top part of the flow was possibly remagnetized during the emplacement of the normally magnetized Unit 1. All discrete samples from Holes 989A and 989B sections, demagnetized to 80 mT and measured on the cryogenic magnetometer, carry normal polarity. Discrete samples from the lower part of Unit 2 contain reversed polarity. Confirmation of the magnetic polarity must await further alternating field and thermal demagnetization studies on shore.