Previous   |   Next

doi:10.2204/iodp.proc.301.206.2009

Regional studies, prior drilling, and modeling of the Expedition 301 area

Earlier studies summarize the geological, survey, and drilling history of the Expedition 301 area and surrounding region in detail (Davis et al., 1992, 1997; Davis and Currie, 1993; Hutnak et al., 2006; Rosenberger et al., 2000; Zühlsdorff et al.). The eastern flank of the Juan de Fuca Ridge near 48°N has some features common to ridge flanks in general, including extrusive igneous basement overlain by sediments that thicken with crustal age and abyssal hill topography bounded by high-angle faults, forming linear structural trends that run subparallel to the spreading ridge to the west (Fig. F1). However, turbidites that flowed from the nearby North American continent in the Pleistocene blanketed the crust in the Expedition 301 area with thick sediments at an unusually young age. Basement rocks remain exposed over large areas mainly close to the active spreading center, and seamounts and smaller basement outcrops also occur up to 100 km east of the spreading center, especially to the north and south of the Expedition 301 work area (Fig. F1).

Operations during Ocean Drilling Program (ODP) Leg 168 completed a drilling transect of eight sites across 0.9 to 3.6 Ma seafloor; collected sediment, rock, and fluid samples; determined thermal, geochemical, and hydrogeologic conditions in basement; and installed a series of CORK observatories in the upper crust (Davis, Fisher, Firth, et al., 1997). Two of the Leg 168 observatories were placed in 3.5–3.6 Ma seafloor near the eastern end of the drilling transect, in Holes 1026B and 1027C (Fig. F1). Expedition 301 returned to this area and drilled deeper into basement; sampled additional sediment, basalt, and microbiological materials; replaced the borehole observatory in Hole 1026B; and established two multilevel observatories at Site U1301 for use in long-term, three-dimensional hydrogeologic experiments (see the "Expedition 301 summary" chapter).

Several studies prior to Expedition 301 explored locations where hydrothermal fluids discharge, recharge, or flow laterally through basement. Before Leg 168, there was a mostly two-dimensional view of the dominant fluid circulation pathways in this area, with recharge occurring across large areas of basement exposure close to the ridge (near the western end of the Leg 168 transect) (Fig. F1A), then flowing toward the east. Some results from Leg 168 were consistent with this view, including seafloor heat flow and basement temperatures that increased and basement fluids that were warmer and more altered with greater distance to the east along the drilling transect (e.g., Davis et al., 1992, 1999; Davis, Fisher, Firth, et al., 1997; Elderfield et al., 1999). But there were inconsistencies with this conceptual model of large-scale hydrogeologic flow, especially after Leg 168 results were considered. For example, although basement fluids warmed and aged along the western end of the Leg 168 drilling transect with increasing distance from the ridge (from Site 1023 to Site 1025), fluids were younger with respect to ¹⁴C at the next nearest site to the east (Site 1031) and younger still farther to the east (at Site 1026) (Fig. F1A), despite being warmer and more altered (Elderfield et al., 1999). In addition, reexamination of existing bathymetric data and collection of additional data along western end the Leg 168 transect showed basement outcrops to the north and south that could allow hydrothermal fluids to recharge and discharge, with flow occurring nearly perpendicular to the transect (Hutnak et al., 2006). There was also the vexing problem of explaining where fluids flowing toward the east at the western end of the Leg 168 transect might exit the crust (e.g., Davis et al., 1999). It is not possible for large volumes of fluid to recharge the crust, flow laterally across tens of kilometers, and then be stored indefinitely. In fact, discharge is required as a complement to recharge in order for a flow system such as this to be self-sustaining; it is the difference between pressures at the base of recharging and discharging columns of crustal fluid that creates a "hydrothermal siphon" capable of operating at a crustal scale (e.g., Stein and Fisher, 2003; Fisher et al., 2003).

Regional site surveys in preparation for Expedition 301 focused on and near basement outcrops that could be fluid entry and exit points to and from the crust, allowing hydrothermal flows to bypass generally thick and impermeable sediments (Fisher et al., 2003; Hutnak et al., 2006; Zühlsdorff et al.). Thermal data suggest a significant component of south to north (ridge-parallel) fluid flow in basement at the eastern end of the Leg 168 transect, an interpretation consistent with geochemical studies (Walker et al., 2007; Wheat et al., 2000). Consideration of bathymetric, sediment thickness, and heat flow data near the western end of the Leg 168 transect suggests that there may be a component of north to south fluid flow in basement in this area as well (Hutnak et al., 2006).

Two sets of numerical studies were completed to assess the regional thermal and hydrologic state of the upper crust in the Expedition 301 area, and to help place survey and drilling results in context. One set of numerical models was crafted to estimate basement properties consistent with inferred patterns and rates of fluid circulation between recharge and discharge sites separated by 50 km (the approximate distance between Grizzly Bare and Baby Bare outcrops) (Fig. F1B). These models complemented analytical calculations based on the same system geometry that suggested basement permeability in excess of 10^–12 m² was required to allow formation of a self-sustaining hydrothermal siphon (Fisher et al., 2003). Numerical results suggested that outcrop to outcrop circulation can be sustained when basement permeability is ≥10^–12 m² (Hutnak et al., 2006). At lower permeabilities, too much energy is lost during lateral fluid transport for circulation to continue without forcing, given the limited driving pressure difference at the base of recharging and discharging fluid columns.

These numerical models also showed that fluid temperatures in upper basement are highly sensitive to modeled permeability, providing a critical constraint on regional basement properties (Fig. F2A). When crustal permeability is too high (10^–10 to 10^–9), fluid circulation is so rapid that basement is chilled to temperatures below those seen regionally (modeled values of 20°–50°C). A good match is achieved to observed upper basement temperatures of 60°–65°C (Davis et al., 1992; Hutnak et al., 2006; Shipboard Scientific Party, 1997; also see the "Site U1301" chapter) when lateral basement permeability is 10^–11 m². These models were run with a two-dimensional domain; fully three-dimensional models are likely to require somewhat higher basement permeability to sustain conditions consistent with observations.

Another set of regional thermal models was created to evaluate the significance of "background" heat flow around the Expedition 301 work area. These models were prompted by a key observation following numerous regional thermal studies: heat flow along a 100 km long swath of 3.4–3.6 Ma seafloor, extending from 50 km north to 50 km south of ODP Sites 1026 and 1027, is lower by 15%–20% than predicted by standard lithospheric conductive cooling models, even after correcting for rapid sedimentation rates during the Pleistocene (an additional 12%–18% correction) (Davis et al., 1999; Hutnak et al., 2006; Zühlsdorff et al.).

Three explanations for the observed seafloor heat flow in this area were considered initially: (1) lithospheric heat flow is regionally low by 15%–20% in comparison to predictions from global lithospheric cooling models, (2) large-scale advective heat loss presently affects a 100 km long swath of seafloor (or perhaps an even large region), or (3) there is bias in the distribution of heat flow measurements. The first explanation seems especially ad hoc and would require anomalous lithospheric temperatures and/or structure, neither of which has been inferred from earlier studies. The second explanation is inconsistent with the lack of a spatial trend in heat flow values ≥5 km from outcrops and buried basement highs. Where advection has been inferred to explain heat flow suppression on the order of 15%–20% in other locations, there is generally lower heat flow near areas of recharge (e.g., Langseth and Herman, 1981; Stein and Fisher, 2003). The low values adjacent to Grizzly Bare outcrop extend only a few kilometers from the outcrop, and background heat flow away from this feature is identical to that away from Baby Bare and other outcrops to the north (Zühlsdorff et al.). The third explanation (sampling bias) remains possible but seems unlikely given the number and spatial coverage of available data and the consistency of background measurements (Hutnak et al., 2006).

Modeling allowed a new hypothesis to be explored: that the Expedition 301 work area is currently undergoing "thermal rebound" following the cessation of a long period of regionally efficient, advective heat extraction from the crust. A coupled model of heat transport and sedimentation was developed to assess the timing of rebound following the cessation of hydrothermal cooling and to check the extent of heat flow suppression likely to result from rapid sedimentation (Hutnak and Fisher, 2007). Rapid sedimentation tends to lower measured seafloor heat flow values by cooling near-seafloor sediments. The new model was an extension of earlier sedimentation models used for this area (Davis et al., 1999; Wang and Davis, 1992). The new model allows simulation of multiple basement layers beneath accumulating sediments. Heat sinks can be distributed within these layers as a proxy for advective heat loss caused by hydrothermal circulation, with the efficiency of the heat sinks representing the extent of hydrothermal heat extraction. Heat sinks are deactivated (either abruptly or gradually) to replicate the end or reduction of hydrothermal heat extraction, and the basement aquifer and overlying sediments are allowed to recover (rebound) conductively as sedimentation continues. This approach differs from that used in earlier analytical studies that assessed hydrothermal rebound (e.g., Benfield, 1949; Hobart et al., 1985) because the finite thickness of the basement aquifer in the newer models acts as a heat capacitor and delays recovery to lithospheric conditions. The model also can incorporate the influence of local convection within basement before and after the cessation of advective heat loss by using a high Nusselt number approximation.

Application of this model to the Expedition 301 field area suggests that the region may still be recovering from an earlier period of relatively efficient advective heat extraction from basement prior to the most recent period of Pleistocene sedimentation (Fig. F2B). Sedimentation rates were on the order of 250–440 m/m.y. during the last 1 m.y. around Sites 1026, 1027, and U1301 (Shipboard Scientific Party, 1997). Basement relief and sediment thickness maps in this area show numerous locations where sediment thickness is presently ≤100 m (Hutnak et al., 2006; Zühlsdorff et al.). Many basement areas now covered by thin sediments would have been exposed at the seafloor prior to the last several hundred thousand years of sedimentation, and areas of current basement exposure (e.g., Baby Bare outcrop) would have been larger (Fig. F1C). Larger areas of basement exposure and the greater spatial distribution of these areas would have been conducive to more efficient regional advective heat loss, as is currently seen at the western end of the Leg 168 transect (Davis et al., 1992; Hutnak et al., 2006), where measured heat flow is ~20% of lithospheric predictions, and on other ridge flanks where basement outcrops are more common (e.g., Hutnak et al., 2008; Lucazeau et al., 2006; Villinger et al., 2002).

We do not know the detailed history of hydrothermal heat extraction around the Expedition 301 field area during the Pleistocene, but we can assess the relative timing and magnitude of rebound following a reduction in advective heat loss (Fig. F2). If the efficiency of advective heat loss was initially 80%, as currently observed at the western end of the Leg 168 transect, then the abrupt cessation of advective heat loss in the last several hundred thousand years could result in regional heat flow at the seafloor being lower than the present lithospheric value by 15%–40% (taking into account the sedimentation history of this area, which should have lowered seafloor heat flow by 12%–18%). Assuming a lower initial efficiency for advective heat extraction would reduce the magnitude of remaining rebound, but if the advective extraction of lithospheric heat were reduced gradually as basement outcrops were buried (instead of ending abruptly), then rebound would be delayed even more (Hutnak and Fisher, 2007).

This explanation for suppressed regional heat flow around the Expedition 301 field area would mean that the thermal and hydrogeologic state of the crust is recovering slowly from an earlier period of more vigorous low-temperature hydrothermal circulation. Thus upper basement temperatures around Sites 1026 and U1301, typically 60°–65°C, will continue to warm because of hydrothermal rebound and sedimentation, even while lithospheric heat flow becomes lower as the plate ages. This result implies that the decline in efficiency of regional advective heat extraction from the crust does not require a significant reduction in basement permeability. Instead, the reduction in the extent of closely spaced basement outcrops may be the primary explanation for anomalously warm basement conditions seen in this area at present, relative to similarly aged seafloor in other settings.

Top of page   |   Previous   |   Next