Wireline log data recorded in subaerial flood basalt constructions show a characteristic cyclic pattern, with large variations in elastic parameters within individual basalt units (Planke, 1994; Delius et al., 1995; Planke and Cambray, 1998). Typically, the P-wave velocity (V_p) is as low as 2-3 km/s in the basalt top, but increases to 5.5-6 km/s in the interior of lavas thicker than 5-10 m and decreases rapidly near the lava base. This three-part flow zonation is very similar to observations of textural characteristics revealed by field studies of inflated flood basalt lavas, with an upper vesicular crust, a central massive and fractured core, and a thin basal vesicular crust (Self et al., 1997, 1998).

Continental breakup is frequently associated with voluminous volcanism. Seismic data are essential for mapping the aerial distribution of the volcanic constructions along rifted margins. Seismic imaging of intra- and sub-basalt features is generally difficult. The large variations in seismic properties in single subaerial lava flows give rise to numerous high-impedance boundaries within the flood basalt constructions. Because the dominant wavelength of 50-100 m used in conventional seismic experiments is much larger than the typical thickness of a flood basalt unit, a reflected wave field in these terrains will consist of strong reverberations and tuned waves (Planke and Eldholm, 1994). However, well-defined intra- and sub-basalt reflectivity is imaged on modern seismic reflection profiles (e.g., Symonds et al., 1998). Improved understanding of the seismic properties of individual basalt flows may help to improve the seismic imaging capability in volcanic terrains.

During Leg 163, Hole 990A was drilled 131 m into basaltic basement, penetrating 14 subaerial flood basalt units with an excellent average recovery of 74% (Fig. 1) (Duncan, Larsen, Allan, et al., 1996). The recovery is biased toward sampling the massive flow interiors as shown by (1) a recovery of 30%-60% in many of the cores penetrating unit boundaries, compared with the 74% average recovery and (2) only 2 of 14 unit boundaries were recovered. P-wave velocity, gamma-ray attenuation porosity evaluator (GRAPE) density and magnetic susceptibility were measured shipboard at 2-15-cm core intervals (Fig. 2). Unfortunately, the hole was not logged because of bad weather. However, the high recovery and closely spaced shipboard measurements allow detailed studies of systematic intra-basalt physical property variations.

This paper focuses on how the interplay of primary emplacement structures (such as vesicularity and fractures) and subsequent alteration determine the large variations in physical properties observed in the top part of the basalt flows. We investigated in detail the changing textural, mineralogical, and physical property variations in the top part of three representative and well-sampled basalt flows. Cerney and Carlson (Chap. 3, this volume) describe the velocity variations in the massive part of the lavas from modeling of high-pressure P- and S-wave minicore velocity measurements. They show that the velocity distribution can be explained by variations in porosity and a common pore aspect ratio distribution.