Previous   |   Next

doi:10.2204/iodp.proc.301.102.2005

Results and discussion

The discussions that follow refer to several regions that include and extend beyond Leg 168 and Expedition 301 work areas (Figs. F1, F2). Expedition 301 worked mainly in the Second Ridge area, above the same buried basement high where Site 1026 is located. Secondary drilling sites included the First Ridge area, close to Sites 1030 and 1031, and the Deep Ridge area, east of Second Ridge, but these were never occupied during Expedition 301. We also discuss (briefly) data from the Southern outcrop area, located 30–55 km south of the Second Ridge area, and an additional area of bathymetric highs located ~50 km north of the Second Ridge sites. More extensive presentation and discussion of data from these two areas will appear elsewhere.

Swath mapping

The regional swath map illustrates several notable features (Fig. F2). The JFR is clearly visible at the western edge of the survey area, with the Cobb Offset near 47°40′N being particularly prominent. Much of the seafloor east of the ridge is relatively flat, resulting from a thick accumulation of sediment, but care must be taken in interpreting the regional map because the depth scale emphasizes ridge bathymetry rather than flank bathymetry. Local maps reveal considerably greater detail.

One of the most important goals of swath mapping in this region was to assess whether there were basement outcrops in the survey area in addition to those mapped previously. Outcrops are hydrogeologically important because they provide permeable pathways between basaltic oceanic crust and the overlying ocean, allowing both recharging and discharging fluids to bypass low-permeability sediments (e.g., Davis et al., 1992; Fisher et al., 2003a; Johnson et al., 1993; Villinger et al., 2002). Regional swath coverage (Fig. F2) confirms that the basement outcrops in the Second Ridge area are separated from the next nearest outcrops by tens of kilometers. There remain gaps in swath coverage north of the Second Ridge outcrops, and a few small patches to the west, hopefully to be filled by future surveys.

Grizzly Bare outcrop is the largest outcrop found on this part of the ridge flank, located near the southeastern edge of the survey area (Figs. F2, F3). This feature is 3.5 km across and rises 450 m above the surrounding sediment plain. Because regional sediment thickness is 500 m, the total height of the edifice above the basaltic crust is 950 m. The next largest outcrop is Grinnin' Bare outcrop, 33 km north-northwest of Grizzly Bare outcrop. Unlike Grizzly Bare outcrop, which is symmetric and conical, Grinnin' Bare outcrop appears to have lost about one-fourth of its exposed mass along a steeply dipping failure surface on the eastern side of the edifice. The missing material is not visible at the seafloor and must have been buried by subsequent sedimentation. Grinnin' Bare outcrop is 2 km across at its widest point and rises 300 m above the surrounding seafloor and 800 m above the surrounding basaltic crust.

North-northeast of Grizzly Bare outcrop are the three basement outcrops that originally attracted attention on this ridge flank: Papa Bare outcrop, Mama Bare outcrop, and Baby Bare outcrop (Davis et al., 1992; Mottl et al., 1998; Wheat and Mottl, 1994). These three outcrops are high points along two adjacent, otherwise buried basement ridges, and all are elongate from the south-southwest to the north-northeast, subparallel to the underlying structure of the upper crust and the active spreading center to the west. Papa Bare outcrop is the largest of the three, extending 3 km in the long dimension, 1.5 km across the short dimension, and rising 250 m above the surrounding sediment. Mama Bare and Baby Bare outcrops are located above another basement ridge ~7 km to the west. Mama Bare outcrop is 8 km west-southwest of Papa Bare outcrop and is 2.0 km long and 0.7 km wide. It rises 150 m above the surrounding sediments. Baby Bare outcrop is the smallest of the three outcrops—1.0 km long, 0.5 km wide, and rising 70 m above the surrounding seafloor. Baby Bare outcrop is located 15 km south of Mama Bare outcrop, and Site 1026 is located about halfway between these two features (Fig. F3). Baby Bare and Mama Bare outcrops are located on crust dated by seafloor magnetic anomalies at 3.5 Ma, whereas Papa Bare outcrop is located on 3.8 Ma seafloor (Becker et al., 2000; Wilson, 1993).

The relationship between the Second Ridge outcrops and outcrops to the south is unclear. Grizzly Bare outcrop is located south-southwest of Baby Bare and Mama Bare outcrops, along roughly the same heading as that between the latter outcrops and consistent with the strike of basement topography and magnetic anomalies (Shipboard Scientific Party, 1997a; Wilson, 1993). This suggests that Grizzly Bare, Baby Bare, and Mama Bare outcrops might be located on the same buried abyssal hill. Thermal and chemical data suggest that there is fluid flow in basement between Grizzly Bare and Baby Bare outcrops, perhaps in part because of enhanced basement permeability in the along-strike direction (Fisher et al., 2003a; Wheat et al., 2000). However, seismic lines run between Grizzly Bare and Baby Bare outcrops show that the buried ridge below Baby Bare and Mama Bare outcrops dies out to the south, reappearing immediately north of Grizzly Bare outcrop. In addition, regional marine magnetic anomalies (Wilson, 1993) indicate that Grizzly Bare outcrop may be located on crust that is 100–200 k.y. younger than that below Baby Bare and Mama Bare outcrops. The age of Grizzly Bare outcrop is unknown; Baby Bare outcrop is thought to have formed from an off-axis eruption, being as much as 800 k.y. younger than the crust on which it sits (Becker et al., 2000; Karsten et al., 1998).

Bathymetric data from the Second Ridge area show that the seafloor above the ridge is slightly deeper than nearby areas to the west and east (Fig. F3; deeper blue in color immediately above the buried basement ridge). The seafloor continues to deepen to the south toward Grizzly Bare outcrop (Fig. F3; blue grades to purple). A sediment transport channel is readily apparent east of Grizzly Bare outcrop. Its surface expression is considerably less pronounced to the north, and the channel disappears within the limits of hydrosweep resolution (~5–10 m) at ~47°30′N. However, the channel can be traced farther to the north in seismic data (described below).

Bathymetric data from the First Ridge area reveal several notable features (Fig. F4). The seafloor above the largest buried basement high is broadly elevated. Another basement high, which is less pronounced bathymetrically, strikes from northwest to southeast and merges with the first basement high in the south. A local, triangular topographic high indicates where turbidities coming in from the north are trapped between these basement peaks, forming a thick sediment package (e.g., Spiess et al., 2001; Spinelli et al., 2004; Underwood et al., 2005). Small distributary channels are visible on the southwestern edge of this topographic high. There is a bathymetric depression on the western flank of the main basement high associated with the location of a small, buried basement ridge, as revealed by seismic data. The strike of this feature is oblique to the dominant structural trend in the First Ridge area.

There are three prominent bathymetric highs north and northwest of the First Ridge area, at ~48°–48°10′N (Fig. F4). One is located northeast of the First Ridge area, along the dominant structural trend of the buried basement high, and two others are located to the west, 11–15 km north of the Leg 168 drilling transect. Careful examination of swath data and seismic data indicate that these bathymetric highs are basement outcrops. Two of these features were crossed by seismic lines during the 1996 Sonne survey (Rosenberger et al., 2000). There is at least one additional small bathymetric high that may include areas of basement exposure at 47°47′N, 128°37′W. Seismic and heat flow data were collected across this small feature as part of the ImageFlux and RetroFlux surveys, but no data or samples were collected near the outcrops north of the First Ridge area.

In addition to mapping the seafloor around primary and secondary drilling targets at the Second Ridge and First Ridge areas, two additional bathymetric highs were examined ~50 km north of the Second Ridge area (Figs. F2, F5). These were discovered fortuitously by swath-mapping during a transit by the Sonne in 1996 and were examined in greater detail during the 2000 ImageFlux and RetroFlux expeditions. The smaller feature is elongate southwest to northeast and is known as Zona Bare outcrop. It is 2 km long, 1 km wide, and rises 30–40 m above the surrounding seafloor. The southwestern edge of the outcrop is steeper than the other sides, whereas the top is virtually flat. To the northwest is a larger area of elevated seafloor known as Rattlesnake Ridge. The seafloor here is as much as 100 m shallower than the surrounding area, but the slopes along the sides of the feature are gentle and there does not seem to be basement exposed. Considerable time was spent coring and collecting heat flow data around these two features during the RetroFlux expedition, but there is only limited seismic coverage in this area; more detailed analysis of these features will be presented elsewhere.

Seismic surveys

Second Ridge area

The main seismic grid across the Second Ridge area consists of numerous parallel lines (average length = ~19 km) that were oriented across structural strike and are spaced ~2 km apart (Fig. F6). The area between the southernmost of these lines (GeoB00-192) and the northernmost line (GeoB00-218) extends roughly between 47°40′N and 47°52′N and includes Papa Bare, Mama Bare, and Baby Bare outcrops. Within a portion of this larger grid, approximately between Leg 168 Hole 1026C and a location 4.6 km north, line spacing was decreased to 0.5 km to investigate a buried basement high known as Wuzza Bare subcrop (Fig. F3). A grid of closely spaced profiles was later added in the vicinity of Leg 168 Sites 1026 and 1027 and Site U1301. This grid consists of 26 100 m spaced lines ~5.4 km long. Another set of six parallel lines, spaced ~600 m apart, was located west of the main grid to study a buried basement high called Isita Bare subcrop (Fig. F3). A small number of seismic lines oriented south-southwest to north-northeast were run as well, the most important of which are GeoB00-490, close to drill sites above the peak of the buried basement ridge, and GeoB00-459, connecting Baby Bare and Mama Bare outcrops to newly recognized Zona Bare outcrop.

The main goal of the site survey at Second Ridge was to map basement morphology to allow for a more complete geological interpretation and better selection of locations for IODP drilling. A basement relief map created by interpolation between basement picks from the seismic data shows several parallel buried basement ridges and troughs, one of which connects Baby Bare and Mama Bare outcrops and, possibly, Zona Bare outcrop to the north (Fig. F7B). A more detailed basement map (Fig. F7C) based on the grid of the most closely spaced seismic lines indicates that the ridge in the vicinity of Sites 1026 and U1301 splits into two parts with a small depression in between. Sites 1026 and U1301 are located on the eastern part of the ridge, where basement elevations are greatest. Basement elevations around the drill sites are considerably lower than elevations north and south along the ridge.

Baby Bare and Mama Bare outcrops are volcanic constructions built on top of a ridge-flank abyssal hill (Becker et al., 2000). The northern part of the basement ridge, including Mama Bare outcrop, is 150–200 m higher in elevation than the shallower part. The peak of the basement ridge plunges toward the south just south of Mama Bare outcrop; the basement slope becomes gentler across Sites 1026 and U1301. Baby Bare outcrop is located at the southern edge of a region characterized by very high basement relief. The expression of the abyssal hill is more subdued south of Baby Bare outcrop. The relatively flat summit and low elevation of the buried basement ridge around Sites 1026 and U1301 suggest that they are located beyond the limits of volcanic construction that created Mama Bare and Baby Bare outcrops (Fig. F3).

The deepest band of coherent reflections in seismic lines across the Second Ridge area mark the sediment/basement interface (Fig. F8). Basement morphology is very similar between Line GeoB00-483, at the northern end of the area of dense coverage on Second Ridge, and Line GeoB00-466, to the southern end of this area, crossing Site U1301 (Figs. F6, F8A, F8B, F8C, F8D, F8E). This consistency is also apparent in the basement relief map (Fig. F7) and in crossing Line GeoB00-490 (Figs. F6, F8F).

The upper part of the sediment section is dominated by a large sediment transport channel, known from drilling experience at Sites 1026, 1027, and U1301 to be filled mainly with sandy turbidites transported from the north (Davis et al., 1997; Underwood et al., 2005). The channel is defined by prominent seismic reflectors that extend to ~200 ms below the seafloor above the local basement topographic low (Fig. F8A, F8B, F8C, F8D, F8E). The deepest part of the channel is located approximately halfway between the buried basement ridge upon which Papa Bare outcrop is located and the basement ridge upon which Mama Bare and Baby Bare outcrops and Sites 1026 and U1301 are located. The channel deposits are thinner above the buried basement high around Sites 1026 and U1301, but the channel can be traced 1–2 km west of the buried basement peak (Fig. F8).

Sediments recovered from near the base of these channel deposits at Site 1027, 200 ms below the seafloor, are ~0.28 m.y. old on the basis of microfossil stratigraphy (Shipboard Scientific Party, 1997b; Underwood et al., 2005). The strong correlation between channel geometry and basement relief suggests that the distribution of turbidite sedimentation in the Second Ridge area has been influenced significantly by basement structure. Some influence of basement relief on sedimentation patterns may continue today, as indicated by small variations in bathymetry (Fig. F8). Drilling into these coarse, poorly consolidated sediments is not difficult, but recovery tends to be low when using rotary coring systems. The sediments are easily washed away during drilling and casing operations, which can lead to difficulties in stabilizing reentry cone and casing systems, as experienced during both Leg 168 and Expedition 301.

The deepest part of the sediment column in this area is characterized by lower reflection amplitudes than the upper part, probably due to seismic attenuation and the lack of coarse-grained turbidites within the oldest part of the sediment column (e.g., Shipboard Scientific Party, 1997b; Underwood et al., 2005). Normal growth faults showing upwardly decreasing offsets are commonly observed in seismic images from this area (Fig. F8). The nature of off-axis tectonic activity, and its influence on sediment deformation around Second Ridge, are currently being investigated.

Wuzza Bare subcrop is located southwest of Papa Bare outcrop and is covered by a few tens of meters of sediment, much like Isita Bare subcrop to the west (Figs. F3, F7, F8G). Basement at Wuzza Bare subcrop comes closest to the sediment surface on Line GeoB00-208 (Fig. F8G). On the same line, Parasound data show the most pronounced expression in subsurface reflectivity (Fig. F8G). An acoustic "washout" zone of ~200 m width is observed ~10 m below seafloor but does not reach the surface as a coherent unit. A similar feature is observed on parallel line GeoB00-205, 200 m to the north, but the top of the washout is 10–20 m deeper. This suggests that the center of the anomaly in Parasound reflectivity (and probably the center of the basement high) may be located close to Line GeoB00-208. This interpretation is consistent with heat flow data collected along both of these seismic lines, as discussed later.

A thin veneer of acoustically transparent sediments is seen in Parasound profiles crossing this area (Fig. F8G). At Wuzza Bare subcrop and to the east, this layer may be <1 m thick, but it thickens westward to ~2 m at 1.5 km distance from the peak of the buried basement high (Fig. F8G). The veneer appears to be thickest within the large sediment transport channel described above, and it may represent the last stage of localized sediment distribution in this area, following nearly complete infilling of the original bathymetric (basement) low.

First Ridge area

Parallel seismic lines across strike in the First Ridge area cover a region between 47°47′N and 47°57′N, with the closest lines being only 25 m apart (Figs. F4, F9). Five additional lines were shot along strike, two of them extending to the southern end of First Ridge survey area. The seafloor across the First Ridge area is relatively smooth compared to the rugged basement below (Fig. F10). Although relief is not truly 2-D, structural variations on scales of a few tens of meters (at both the seafloor and at the sediment/basement interface) tend to occur across strike (northwest–southeast) rather than along strike (northeast–southwest). The data suggest that the First Ridge area can be roughly divided into three parts: northern and southern parts with moderate basement topography (Fig. F10B, F10C, F10D, F10F) and a center part between Site 1030 and a location ~800 m south of Site 1031, where basement is rougher (Fig. F10E). Seismic data in the southern part show a second buried ridge that strikes southeast and eventually merges with main basement high in the south. Turbiditic sediments coming in from the north appear to be trapped between the two ridges, forming a local topographic high (Fig. F10F). The dense grid of seismic profiles, located in the northern part of the First Ridge area, reveals a third buried basement high that is smaller and slightly oblique to the dominant basement trend. The seafloor between the buried ridges shows a linear depression at the western flank of the main basement high (Fig. F10A, F10B, F10C, F10D, F10E).

The area covered by the densest seismic grid can be divided into three smaller parts. At the northern end, the basement high is characterized by a single peak structure (Fig. F10D, F10E). Older sedimentary layers are clearly separated by the basement ridge, whereas the turbidites associated with younger layers can be traced across the ridge. The layers are folded, fractured, or faulted and show vertical offsets above the basement peak. The folding and faulting seen within the sediments in this area are much less pronounced than the folding and faulting found at the southern end of the dense profile grid, where the presence of two buried basement highs results in greater deformation (Fig. F10A).

Layer inclination at the southern edge of the dense seismic grid is larger than that to the north, and vertical offsets of individual horizons are also greater. There is a transition between the northern and southern parts of the dense seismic grid where the basement ridge splits into two parts (Fig. F10B, F10C). The continuity of some layers suggests that overspilling occurs across the basement high (i.e., layers associated with turbidites coming from one side of the ridge thin out on the other side with increasing distance from the ridge). The stratigraphic pattern of turbiditic sedimentation across the buried basement highs is complex and warrants greater study.

Narrow zones of decreased reflection amplitudes (washouts) are observed in many profiles across the First Ridge area (Fig. F10). These washouts are most pronounced in the high-resolution Parasound data but are also present in GI gun data (Zühlsdorff et al., 1999). Mapping on the basis of Parasound data revealed that zones of lower reflectivity form a coherent pattern that can be traced along much of the buried basement high. Comparison of the Parasound and GI gun data further shows that zones of decreased amplitudes are well correlated with local basement peaks or related to pervasive distortion or disruptions of sedimentary layering (Zühlsdorff and Spiess, submitted). As in the Second Ridge area, the correlation on seismic profiles between these distinct sedimentary features and underlying basement relief suggests that there is a strong control of basement structure on sedimentary processes. Other studies have explored possible connections to fluid seepage, local lithologic variability, and postdepositional deformation (Giambalvo et al., 2000; Spinelli et al., 2004; Zühlsdorff and Spiess, 2001, submitted; Zühlsdorff et al., 1999). It is possible that similar patterns in the seismic profiles result from different causes, but the occurrence of these features across a broad geographic area suggest that they may result from common mechanisms. Future studies in this area may address this issue through additional measurements and sampling.

Southern outcrop area

At the southeastern edge of the survey area, seismic lines approach and cross Grizzly Bare and Grinnin' Bare outcrops. These seismic lines are 37–56 km long and separated by 6–11 km. In addition to parallel lines oriented across strike, two long crossing lines were run, forming a large triangle with Grizzly Bare outcrop at the apex to the south and Grinnin' Bare outcrop at the northwestern corner (Figs. F2, F3). As a result, both outcrops were covered by at least three crossing lines. No Expedition 301 drilling was planned in the Southern outcrop area, so we discuss only general mapping and seismic results in this area. Part of one seismic line is shown in the discussion of heat flow results.

Grizzly Bare outcrop is circular in plan view from its base ~500 m below the sediment surface and up to its summit. This seamount rises >600 ms TWT (~450 m) above the surrounding sediments. Significant reflection energy is observed on top of Grizzly Bare outcrop in Parasound data, and there may be a thin cap of hemipelagic or pelagic sediment, as seen above Baby Bare outcrop to the north (Becker et al., 2000).

Two large sediment transport channels are observed at the seafloor in this area, a narrower one passing east of Grizzly Bare outcrop (clearly visible on Fig. F3) and a broader one passing between Grizzly Bare and Grinnin' Bare outcrops and trending to the southwest. Both of these channels may extend from far to the north across the survey area, based on correlations between adjacent seismic profiles. The easternmost channel has a more pronounced seafloor expression south of 47°30′N and appears to be the southward continuation of the large filled channel observed at Second Ridge that passes across Site 1027 (Figs. F3, F8). The westernmost channel may pass west of the Second Ridge area, explaining why it is not observed in seismic lines from that region.

Deep Ridge area

Seismic reflection images formed by processing MCS data collected on the EW0207 expedition were used to select locations for the Deep Ridge sites, ~110–140 km east of the active spreading center (Figs. F2, F11). Both Deep Ridge sites are located on basement highs to minimize drilling time to basement and because basement highs in this region tend to be overpressured and thus may eventually provide good basement fluid and microbiological samples once thick sediments are penetrated.

The thickness of the sedimentary cover in the Deep Ridge area varies from 750 to >1000 m (Fig. F11). A series of small-throw normal faults are imaged within the sediment section, possibly indicating differential compaction or recent tectonic activity outside the axial region. Fault offsets gradually diminish upsection with sediment age, from older to younger, possibly suggesting growth faulting caused by long-term slip within basement. The interface between the sediments and igneous basement has the strongest acoustic impedance, although this is not always evident in the reflection image because of the applied automatic gain control and because, where the basement topography is rough, much of the acoustic energy is scattered and not returned. A Layer 2A event, found ~200 ms below the top of the igneous basement, is imaged across part of this area. Based on estimates of upper crustal P-wave velocities from wireline logs at Site U1301 during Expedition 301, this event corresponds to a subbasement depth of ~500 m, roughly consistent with earlier interpretations based on regional MCS data (Rohr, 1994).

Heat flow data

Regional values and references for subsequent analyses

Researchers have made ~1700 multipenetration heat flow measurements on 0.1–5.0 Ma seafloor on the eastern flank of the JFR since 1978, with most of these measurements colocated along seismic reflection profiles (Fig. F2). The 2000 RetroFlux program included acquisition of 437 heat flow values. All of the multipenetration measurements made in this region suggest that conditions within shallow sediments are dominantly conductive. Submersible measurements made on Baby Bare outcrop (Becker et al., 2000; Wheat et al., 2004) are excluded from the totals cited above, and from the following presentation and discussion, because those measurements are influenced strongly by local patterns of fluid and heat transport associated with the basaltic edifice.

Measurements made away from the hydrothermal influence of outcrops and basement relief allow assessment of the "background" thermal state of the upper oceanic crust. Standard lithospheric cooling models generally give a (time)^–1/2 relationship between heat flow and plate age (e.g., Davis and Lister, 1974; Parsons and Sclater, 1977; Stein and Stein, 1994). These models were calibrated mainly using bathymetric data and secondarily using thermal data; thermal data from young seafloor were excluded because of the confounding influence of hydrothermal circulation. Global data compilations tend to exhibit considerable variability, particularly at young ages, and this is generally attributed to hydrothermal circulation. Circulation can advectively mine heat from the crust but can also redistribute heat locally in basement even if there is little or no advective heat loss. The ~1700 heat flow measurements presented in this compilation are not uniformly distributed (Figs. F2, F3, F4, F5); most data have been collected near features of hydrogeologic interest, including areas close to buried and exposed basement highs. Given this bias, it is not possible to determine unambiguously either the lithospheric heat input across this ridge flank or the magnitude of local and regional anomalies associated with advection. Nevertheless, this young, heavily sedimented ridge flank has the greatest density of high-quality multipenetration heat flow measurements in the world, and it is worth considering what the data may indicate about both regional heat input and the extent of thermal anomalies.

A histogram of multipenetration heat flow measurements across this area, including seafloor aged 0.1–5.0 Ma, shows what appears to be a Poisson distribution, with a mode at ~200 mW/m² and a long tail extending to values ~4 W/m² (Fig. F12A). The highest values are associated with local basement highs, some of which are buried and some of which penetrate the sediment as outcrops, and the lowest values are generally found close to areas having extensive basement exposure, where fluid recharge is thought to occur (e.g., Davis et al., 1992; Fisher et al., 2003a; Wheat and Mottl, 1994).

Davis et al. (1999) showed a long transect of seismic data and calculated heat flow across this ridge flank, extending from 20 to 100 km east of the spreading center and passing through or near Leg 168 drilling sites. Observed and calculated heat flow rises rapidly from west to east along the western 20 km of this transect, across a hydrothermal transition (HT) area (Fig. F4), where thermal conditions respond to increasing sediment thickness and associated hydrogeologic isolation of the basement aquifer. Despite considerable local variability, the rest of the calculated heat flow profile tends to follow standard lithospheric cooling curves. The calculated heat flow at the eastern end of the transect, across Sites 1026 and 1027, varies locally as a function of depth to basement but is typically ~250–260 mW/m² after correcting for sedimentation effects (fig. 10 from Davis et al., 1999).

The current compilation allows examination of heat flow north and south of the Leg 168 transect. We focused this effort on the Second Ridge area, where there is the greatest distribution of high-quality, colocated heat flow and seismic data across similarly aged seafloor. There are distinct, local thermal anomalies associated with both discharging and recharging outcrops in this area (e.g., Davis et al., 1992; Fisher et al., 2003a; Thomson et al., 1995; Wheat et al., 2004), but their influences appear to extend only a few kilometers from individual basement features.

In order to determine the background heat flow for this part of the ridge flank, we examined measurements on 3.5 ± 0.1 Ma seafloor and hand-filtered the data to exclude all values that are (1) not colocated with seismic data that image basement or (2) located within 2 km of significant basement relief, including outcrops. The remaining 300 values come from locations extending across nearly 60 km from north of Mama Bare outcrop to just north of Grizzly Bare outcrop (Fig. F3). The filtered data show a remarkably narrow Gaussian distribution (Fig. F12B), as might be expected if variations in heat flow result mainly from conductive refraction and local redistribution of heat by hydrothermal circulation. The mean of the filtered values is 181 ± 16 mW/m².

Even if one assumes the lowest standard lithospheric reference of 477 × (age)^–1/2 (Parsons and Sclater, 1977), a value of 251–259 mW/m² is predicted for 3.5 ± 0.1 Ma seafloor and other reference curves suggest lithospheric values as great as 277 mW/m² (e.g., Stein and Stein, 1994). The typical sediment thickness associated with the 300 filtered values is 500–600 ms, equivalent to 450–550 m, requiring a sedimentation correction of 15%–16% (Davis et al., 1999), but this correction brings the filtered mean up to only ~215 mW/m², a deviation from the reference curves of 15%–20%. This result differs from that determined across Sites 1026 and 1027 by Davis et al. (1999); the present analysis suggests that the transect across Sites 1026 and 1027 is somewhat "warmer" than most of the seafloor having a similar age in this area.

The discrepancy between the filtered, sediment-corrected, mean heat flow on 3.5 ± 0.1 Ma seafloor in this area (~215 mW/m²) and that predicted by standard lithospheric cooling models (~250–280 mW/m²) is greater than uncertainties associated with either the heat flow measurements or with sedimentation or other reasonable corrections. There are several possible explanations, none of which is completely satisfying. One possibility is that the lithospheric cooling curve appropriate for 3.5 ± 0.1 Ma seafloor on this ridge flank falls below standard global models by 15%–20%. Although we might not be surprised to find regional differences in lithospheric cooling trends (and this might help to explain some of the variability seen in global compilations), the possibility of sampling bias or advective heat loss from basement cannot be dismissed.

A second explanation is that the filtered values do not represent mean conductive conditions but show the influence of ~15%–20% advective heat loss. The problem with this explanation is that there are very few basement outcrops in this area, and driving forces are much too small to move significant fluxes of crustal fluid through thick sediments overlying basement (e.g., Fisher et al., 2003a; Giambalvo et al., 2000; Spinelli et al., 2004; Wheat and Mottl, 1994). Examination of thermal transects near both recharging and discharging basement outcrops show that typical upper basement temperatures are ~65°C within a few kilometers of the outcrops. It is this basement temperature and the typical sediment thickness, accumulation rate, and properties in the area that define the characteristic heat flow of ~180 mW/m². The total heat output of Baby Bare outcrop is 1–3 MW (Mottl et al., 1998; Thomson et al., 1995; Wheat et al., 2004); this much heat could be accounted for by suppressing seafloor heat flow by 40 mW/m² to a radial distance of 4–5 km from the outcrop, but low values extend far beyond this distance.

Other areas where seafloor heat flow is suppressed by 15%–20% as a result of lateral advection in basement (requiring fluid flow rates on the order of meters per year or more) typically show spatial trends related to patterns of fluid flow: lower values are found in areas of recharge (or deepening basement) and higher values are found in areas of discharge (or shoaling basement) (e.g., Davis et al., 1999; Fisher et al., 2003b; Langseth and Herman, 1981). In the filtered data set from this area, there is no spatial trend to the data. If the observed mean of ~180 mW/m² results from advective extraction of heat from basement, this would have to occur through a mechanism that extracts heat remarkably evenly at distances ≥10 km from points of recharge or discharge.

A final explanation for the low background heat flow value is that there is a bias in data locations toward lower values. If anything, there is a bias in measurements toward higher values because basement highs are common targets within the survey area and isothermality of shallow basement tends to increase heat flow above these features. The data set was filtered specifically to avoid areas having significant basement topography, either local highs or lows. We expanded the filter to exclude additional areas within 4 km of the nearest basement high, reducing the number of filtered measurements to 70, and calculated essentially the same result. Even areas that are 10 km from the nearest known basement high are consistent with a background value of ~180 mW/m², sedimentation corrected to ~215 mW/m². It will be challenging to resolve this conundrum given the available data because one must assume a reference in order to quantify the magnitude of a thermal anomaly. It would be ideal to collect additional thermal data from 3.4–3.6 Ma seafloor in this region at locations tens of kilometers from outcrops—for example between the Second Ridge area and the southern outcrops (Fig. F3), where there are high-quality seismic data and little basement relief.

For the purposes of subsequent analyses in the Second Ridge area, we use a background heat flow value of 180 mW/m² as a boundary condition at the base of finite-element models. For the First Ridge and other areas closer to the active spreading center to the west, we use a conventional lithospheric cooling curve having the form

q (mW/m²) = 500 × [age]^–1/2 (Ma).

This is justified in younger areas where deviations from lithospheric cooling curves are large; rebound from cooling at the ridge may be incomplete (e.g., Fisher, 2003), and crustal heat extraction is clearly dominated by advection. Because they are so clearly influenced by advective heat loss, it is unlikely that observations from these younger areas will ever be used to constrain lithospheric cooling curves.

Example heat flow profiles

In this section, we show examples of colocated heat flow and seismic profiles that illustrate characteristics typical of several environments on this ridge flank, beginning with the near-axis area and then moving to older sites. In each of the associated figures, we show three panels. The top panel shows measured heat flow and values calculated using conductive, 2-D numerical models based on the geometries and crustal properties indicated by survey and drilling data. Because these models do not include hydrothermal circulation, differences between basal heat input and output at the seafloor result entirely from conductive refraction as a result of basement and seafloor relief and variations in thermal properties. This approach is conceptually similar to that used by Davis et al. (1999), except that it takes into account 2-D (nonvertical) heat conduction and uses a heat flow (rather than constant temperature) lower boundary condition. The middle panels show uppermost basement temperatures calculated by downward continuation of seafloor heat flow values, assuming dominantly vertical heat conduction. These calculations include uncertainty bars based on the accumulation of potential errors, as discussed earlier. Upper basement temperatures calculated with the 2-D conductive models are also shown. The lower panels are displays of seismic data used to generate the other parts of the figures, with superimposed isotherms.

Three profiles illustrate characteristic features on 0.1–1.5 Ma seafloor close to the active spreading center, from north to south across the HT and First Ridge areas (profile locations are shown in Fig. F4). Figure F13 shows heat flow along seismic Line 950802g in the northern HT area. Basement is generally flat below ~180 m of sediment, and seafloor elevation varies by <10–20 m. Heat flow is 202–240 mW/m², ~50% of that predicted by standard lithospheric cooling models. Extrapolated temperatures along the sediment/basement interface remain uniform at 32°–38°C. This segment of data is typical of locations where the seafloor is flat and there is little buried basement relief; regional heat flow suppression results from rapid lateral fluid flow in basement (e.g., Davis et al., 1992, 1999; Stein and Fisher, 2003).

Figure F14 shows a 23 km long profile extending from the edge of basement exposure west of the HT area to several kilometers east of the First Ridge area. Similar profiles were shown in earlier studies, illustrating how crust in this area warms with increasing age (e.g., Davis et al., 1992, 1997); we show this example because it allows comparison of earlier and newly acquired seismic data. Measured heat flow increases systematically along the transect, from ~100 mW/m² near the edge of basement exposure to >600 mW/m² above the first buried basement ridge, and upper basement temperatures increase systematically from <10° to ~45°C. Heat flow is suppressed by >80% relative to lithospheric predictions near the exposed basement in the west but achieves predicted values where basement shoals. Basement isotherms at the western end of the transect are strongly depressed. A local ~50 mW/m² increase in heat flow west of common depth point (CDP) 175, Line 950804e (Fig. F14, immediately adjacent to the area of exposed basalt) may result in part from thermal refraction associated with basement and seafloor relief and in part from the thinning of the conductive sediment layer as locally isothermal basement rises to the seafloor (Fig. F14A).

Heat flow values across the First Ridge area along this transect are locally variable, but on average, they are consistent with lithospheric cooling models. Basement temperatures remain essentially isothermal at 45°C across most of seismic Line GeoB00-252 (Fig. F14B). This uniformity in upper basement temperatures, despite differences in basement elevation of 100 m, is consistent with vigorous, local hydrothermal circulation in basement.

Figure F15 shows heat flow measurements colocated along seismic Line GeoB00-240 crossing the eastern part of the HT and First Ridge areas. Measured seafloor heat flow across the HT area is generally ~320 mW/m², well below the lithospheric prediction, but rises abruptly above the buried basement high to >800 mW/m². Upper basement temperatures remain remarkably uniform at ~25°C across the entire profile. These values are somewhat lower than those seen along Line GeoB00-252 to the north (Figs. F4, F14B), largely because the sediment cover becomes thinner to the south. Basement temperatures near the eastern end of the transect are also somewhat greater than calculated numerically with a 2-D conductive model, suggesting that heat is redistributed locally by fluid circulation in basement.

Three additional heat flow and seismic transects illustrate characteristic conditions in the Second Ridge area (profile locations shown in Fig. F3). Heat flow data are colocated along seismic Line GeoB00-194, south of Baby Bare outcrop and Site U1301 (Fig. F16). The seafloor in this region is generally flat, but basement elevation varies by hundreds of meters. Sediment thins to ~280–300 m above the ~1500 m wide basement peak along Line GeoB00-194 (between CDP 675 and 825; Fig. F16), and basement continues to shoal toward the north, eventually emerging as Baby Bare outcrop 4.5 km away. Seafloor heat flow varies between 150 and 280 mW/m² along this profile, with higher values corresponding to measurements made directly above the basement high. Upper basement temperatures are projected to dip slightly over the peak of the basement high but are generally near 60°–70°C despite significant differences in basement elevation. Upper basement temperatures are ~20°C warmer than predicted on the basis of conductive models, assuming a background heat flow for this area of 180 mW/m². As described in earlier studies, the homogeneity in basement temperatures in this area, which would vary by ~50°C if heat flow were entirely conductive, requires vigorous local hydrothermal circulation (e.g., Davis and Becker, 2004; Davis et al., 1997b; Spinelli et al., 2004).

Seismic Line 92-1 (Becker et al., 2000) crosses Second Ridge ~3 km north of Baby Bare outcrop and 2 km south of Site U1301 (Figs. F3, F17). As along Line GeoB00-194 7.5 km to the south, heat flow is elevated immediately above the buried basement high, but once again calculated temperatures in uppermost basement are consistent at ~60°–70°C. This profile shows heat flow returning to a background value of ~180 mW/m² within several kilometers of the main basement high.

Seismic Lines Geob00-205 and 208 pass across the Wuzza Bare subcrop, a former basement outcrop that is presently buried ~5 km north-northeast of Site U1301 (Figs. F3, F7, F18). This buried basement high is not directly associated with abyssal hill topography like Baby Bare or Mama Bare outcrops but appears to be a more isolated volcanic edifice (Fig. F7). Wuzza Bare subcrop rises >450 m above surrounding basement to within ~40–50 m of the seafloor, perhaps closer because the nearest seismic line does not necessarily pass immediately above the buried summit. Seafloor heat flow increases above Wuzza Bare subcrop, from ~150 to >2500 mW/m² along Line GeoB00-208. Values are somewhat lower along Line GeoB00-205, ~200 m to the north. Making heat flow measurements in this area was challenging because coarse-grained turbidites and debris flow deposits near the seafloor made penetration difficult and led to uneven frictional heating and cooling of thermistors in the heat flow probe. In addition, picking the depth of basement from seismic profiles is challenging across Wuzza Bare subcrop because the sediment/basement interface is steeply sloping and sometimes indistinct. As a result, uncertainties in seafloor heat flow and upper basement temperatures along this transect are greater than those on most other parts of the ridge flank (Fig. F18). Estimated basement temperatures are generally 50°–100°C along Line GeoB00-208. Estimated temperatures are lower along Line GeoB00-205, but this must result, at least in part, from the 1-D and 2-D assumptions inherent in these calculations and from the seismic line not crossing the buried summit.

A final example of seismic and thermal data illustrates extreme conditions that develop near relatively isolated basement outcrops in locations where hydrothermal fluids recharge and discharge (Fig. F19). Baby Bare and Grizzly Bare outcrops are located 52 km apart and both are surrounded by flat seafloor below which there is 500–600 m of sediment. Fluids exiting from Baby Bare outcrop are known to be very young, despite being highly altered, and geochemical considerations preclude the possibility that these fluids recharged through nearby sediment or basement rocks (Mottl et al., 1998; Wheat et al., 2000; Wheat and Mottl, 2000). Instead, recharge of fluids that eventually exit through Baby Bare outcrop is thought to occur through Grizzly Bare outcrop (Fisher et al., 2003a). Thermal surveys of these outcrops show strongly contrasting conditions in basement: temperature contours are swept upward by venting fluids at Baby Bare outcrop, but are swept downward by recharging fluids at Grizzly Bare outcrop (Fig. F19). The thermal influence of both recharging and discharging outcrops appears to extend only a few kilometers from the limit of basement exposure. Considering the forces available to drive recharge, discharge, and transport between the two outcrops (the difference in fluid pressure at the base of recharging and discharging columns of fluid) and the volume flux of fluid exiting Baby Bare outcrop (Mottl et al., 1998; Thomson et al., 1995; Wheat et al., 2004), the bulk permeability of the basement rocks along the flow path must be on the order of 10^–12 to 10^–9 m² (Fisher et al., 2003a). These values are consistent with independent estimates made from packer and open hole thermal tests and results of numerical models of coupled heat and fluid flow in basement (e.g., Becker and Davis, 2003; Becker and Fisher, 2000; Davis et al., 1997b; Fisher et al., 1997; Spinelli and Fisher, 2004).

Top of page   |   Previous   |   Next