Drilling at the oldest two Leg 168 sites targeted a buried basement ridge and adjacent valley, where it had been inferred that buoyancy-driven fluid circulation was sufficiently vigorous so as to have thermally homogenized the sediment/basement contact. Temperatures measured close to the sediment/basement contact during Leg 168 supported this hypothesis, and these results were confirmed by seafloor observatories (CORKs) installed during Leg 168 that permitted basement fluid temperatures and pressures to be determined long after drilling disturbances had dissipated (Davis and Becker, 1998). Basement temperatures at the two sites (1026 and 1027) were found to be virtually identical (Fig. 1), despite a contrast in sediment thickness of 2.5:1. Permeabilities required to sustain such small temperature differences were inferred to be very high, consistent with borehole observations based on packer pumping experiments (Becker and Fisher, 2000) and open-hole flow-rate determinations (Fisher et al., 1997). Local rates of fluid flow also must be high, on the order of meters per year or more, as inferred for the area to the west.
Basement fluids appear to be compositionally homogeneous on a lateral scale of kilometers, but are different within distinct regions along the drilling transect. Basement water collected from Site 1026 is nearly identical in composition to basement waters collected from basement edifices roughly 7 km north and south of the site (Mottl et al., 1998; Wheat et al., in press). A subtle lateral compositional gradient between the outcrops suggests that along-strike flow, not resolved by the cross-strike drilling transect, also may be important in this area (Wheat et al., in press). Basement pore fluids are extensively altered but relatively young within the basement below Sites 1026, 1030, and 1031. This observation suggests that there may be different sources of recharge along the Leg 168 transect and that a simple model of fluid flow from west to east can not explain all observations.
Experience gained through Leg 168 drilling and postdrilling observations have led to new opportunities for crustal fluid sampling. For example, a clear relationship has emerged between hydrologic structure and local fluid pressure: fluid pressures in buried basement edifices consistently exceed local hydrostatic pressure (Davis and Becker, 1998), allowing large volumes of formation water to be produced at the seafloor (Fisher et al., 1997). This phenomenon has been exploited for both geochemical and biological sampling (e.g., Mottl et al., 1998; Mottl et al., Chap. 9, this volume; Elderfield et al., 1999; Wheat et al., in press). Average formation pressures generated through thermal buoyancy are relatively small, and in some instances, tidally generated pressure differentials create occasional overpressures even where the average formation pressure is subhydrostatic (Davis et al., 1999). And finally, in cases where fluids cannot be naturally produced, in situ fluid samplers, a technology recently developed and successfully tested in Leg 168 holes, can provide continuous time-series samples of crustal fluids.