T33B-1347
From Basalt to Dacite: Examining Magmatic Evolution at the 9° 03'N Overlapping Spreading Center, East Pacific Rise
Over 280 rock samples were collected in the spring of 2007 from the 9°03' N Overlapping Spreading Center (OSC) using the ROV Jason II. This was done in combination with an extensive DSL 120A side-scan sonar survey, providing one of the most detailed sampling suites from an OSC. The site was also the focus of past geophysical studies that indicated the presence of an extensive magma lens in the region. Lava geochemistry at 9°03' N covers a wide range of compositions (Mg# from 17 to 56), including evolved MORB, FeTi basalts, andesites, and dacites (~67 wt% SiO2). The diversity in composition at the OSC appears to be related to geologic setting of the erupted lavas, shallow magma chamber processes, and variations in magma supply. The 9°03' N OSC can be divided into three main sections: the eastern limb propagating southward, the western receding limb, and an overlap basin separating the two limbs. Lavas erupted within the neovolcanic zone on the eastern limb span the range of compositions erupted at the OSC (7.29 to 0.67 wt% MgO), while those erupted over the northern portion of the overlap basin are more limited in composition (6.47 to 7.23 wt% MgO). Preliminary results suggest that the on-axis major element variations may be due to mixing of evolved MORB and a silica-rich melt beneath eastern limb axis. Magma mixing is also supported by petrographic and macroscopic textural observations. Eruptions on the southern tip of the east limb, which is propagating into cold crust, are more primitive in composition and contain large phenocrysts suggesting that both thermal conditions in the crust and low magma supply may be important factors in the magmatic evolution. The phenocryst abundance and composition of lavas erupted on the dying western limb may also be related to similar factors.
T33B-1348
Petrography of Lava Flows from the 9 North Overlapping Spreading Center, East Pacific Rise
The region of the 9 deg 03" N overlapping spreading center (OSC) along the East Pacific Rise was sampled extensively (~300 lavas) by the ROV Jason II in spring 2007 (AT 15-17). Sampling locations included both the east and west limbs of the OSC, as well as off-axis areas of interest, and within the overlap basin. Diverse lava compositions were recovered, ranging from basalts to dacites. Here we report initial petrographic results. Approximately 85% of samples are either aphyric or sparsely phyric, but the remainder (42 samples) contain more than 3% phenocrysts by volume. The most porphyritic rocks (11 samples) are comprised of up to 20% phenocrysts. The highest silica rocks contain few phenocrysts, of which most are plagioclase (with minor clinopyroxene) <1 mm in length. The dominant crystal phases in evolved basaltic rocks are plagioclase and clinopyroxene, up to 4 mm in length. The groundmass of most samples exhibit a variolitic texture. Microphenocrysts are predominantly plagioclase and clinopyroxene. Cotectic intergrowth of these two phases is common in rocks retrieved from the east limb, and may either aggregate into microlitic dendrites, or surround larger xenocrysts. Equilibrium-phase skeletal olivines can only be identified in samples from the west limb, though highly resorbed olivines exist in samples from both limbs. Phenocrysts of plagioclase, clinopyroxene, olivine, and rarely orthopyroxene are seen in samples from the east limb. Many of these exhibit oscillatory (plag, cpx) or sector zoning (cpx), or are highly resorbed, suggesting an extended history of multistage residence in the magma chamber. The complex phase chemistry coupled with macroscopic flow banding common in some higher silica samples suggest mixing between basaltic and high silica magmas. These observations of phenocryst populations and crystal growth patterns can be used to elucidate the physical and chemical processes that occurred in the magma reservoir system prior to eruption, and which lead to the great diversity of magma compositions sampled. Particularly when combined with the major and trace element analyses of the quenched glasses, the study of these crystals allows us to gain a more complete picture of the recent history of magmatism at this OSC, including processes such as fractional crystallization, magma mixing, and influxes of new melt into the magma reservoirs.
T33B-1349
Deep-tow Investigation of Crustal Magnetization at the 9N Overlapping Spreading Centers on the East Pacific Rise
Previous estimations of crustal thickness at OSCs have been determined by using seismic velocity models to determine the location of seismic layer 2A in the subsurface. These estimations may be inaccurate due to accretionary processes and alteration of the rocks at the OSC which place the base of layer 2A at depths that do not coincide with the base of the extrusive volcanic layer. A more accurate estimation of the true crustal thickness can be made by measuring the magnetic fields present in the extrusive rock layer of the region. Determination of thickness of the extrusive layer beneath both limbs and overlap basin of the OSC can aid in understanding the geomorphology and ongoing processes of these ridge axis discontinuities. A recent research expedition was conducted in Spring 2007 to study the OSC at 9°03'N on the East Pacific Rise. A fluxgate magnetometer was attached to the deep-towed DSL-120A towfish to measure the magnetic field at a 110 meter altitude above the sea floor. A total of 10 track lines were run in a north-south orientation covering the 20 km length of the OSC with an average spacing of 1 km. Computational inversions were performed on the magnetic fields using the Fourier approach of Parker and Huestis to solve for magnetism of the rocks and to constrain for the thickness of the underlying extrusive volcanic rock layer. Results of the inversions show magnetic anomalies of up to 600 nT near the tip of the southward propagating eastern limb. The ROV Jason 2 collected over 200 rock samples at the OSC of which 25 were subsequently analyzed for NRM, susceptibility and stability. Results of the rock sample analyses show consistently high stability from both limbs and central basin and lack of secondary alteration. NRM and susceptibility measurements show two distinct zones of high anomalies on the eastern limb located between 9°05'N and 9°08'N and on the western limb between 9°00'N and 9°01'N.
T33B-1350
Photographic analysis of seafloor geologic features at the 9 N Overlapping Spreading Center, East Pacific Rise
In March and April 2007, a cruise to the overlapping spreading center (OSC) at 9 deg 03' N on the East Pacific Rise collected geological, geochemical, and hydrothermal activity data along both limbs of the OSC and within the overlap basin to explore linkages between the melt lenses identified by Kent et al. (2000) and surficial geological processes. Here we report preliminary results of photographic data obtained by the ROV Jason II (4 lowerings; covering ~ 20.4 km; ~7000 digital photographs; 213 hrs continuous video for each of the three video streams) and the WHOI TowCam (7 lowerings; covering 29.3 km; ~ 10,000 digital photographs). The majority of our work focused on and adjacent to the eastern (propagating) limb of the OSC. Along the east limb, lavas (including pillow/lobate and sheet/hackly flows) with the freshest and most abundant glass, least sediment cover, and most delicate ornamentation are present along the bathymetrically elevated portion of the ridge between ~9 deg 7'- 10'N. This area also includes lava lake-style collapse structures with remnant pillars, as well as the only hydrothermal vent site observed in this study (at 9 deg 8.3' N). Further south along the east limb, to 9 deg 0' N, sediment cover and tectonic features (predominantly ~N-S-trending fissures) gradually increase, only lobate flows are observed, and glassy surfaces on the lavas are less abundant. Along the southernmost portion of the east limb, the melt lens plunges and cuts east across the ridge axis fabric. Above the location of the plunging melt lens, heavily sedimented lavas are observed with no evidence of recent volcanism. Photographs collected west of the bathymetrically robust portion of the east limb, overlying the wide, off-axis melt lens, generally show volcanism older than that found on-axis at the same latitude, although some areas associated with an elevated NW-trending ridge appear younger than surrounding terrane. Our study of the west (dying) limb of the OSC extends from ~ 9 deg 3' N to 8 deg 57 ' N. In this area, only pillow/lobate flows with variably fresh glass were observed and sediment cover varied, with some areas having a light dusting while others were completely blanketed. Tectonism, in the form of ~N-S fissures and fault scarps (some with talus accumulations) were also observed along the west limb. At the southern end of our study of the west limb, an area of lightly sedimented large pillows with delicate ornamentation was observed. Thorough analysis of our photographic data, in combination with information on the geochemical characteristics of collected samples and U-series age constraints will lead to a better understanding of the magma plumbing system beneath the limbs of the OSC and its relationship to the melt lenses imaged at depth.
T33B-1351
Lava Flows, Faults and Geomorphic Provinces: GIS-based Terrain Modeling at the 9N EPR Overlapping Spreading Centers
How are the mounds, fissures, lava morphology and scarps we see at an overlapping spreading center related to the presence, and for that matter the depth and size, of the melt lens? Using DSL-120A, we obtained 100% side- scan (2m resolution) and 50% bathymetric (4m resolution) coverage at the 9°N OSC within an area 9 km wide between 9.18°N and 8.92°N. The 9°N OSC consists of a propagating eastern limb and a western retreating limb separated by an ~4 km wide overlap basin which is 1 km deeper than the adjacent limbs. We edited and then merged the bathymetry from the SM2000 and DSL-120A data along with EM300 and ARAD hydrosweep data to create a high-resolution bathymetry grid (<30m) covering the entire OSC. We run the NOAA Benthic Terrain Modeler on this merged bathymetry which gives us slope, roughness and shape of the terrain. We use output from the Benthic Terrain Modeler and the provinces described by Sempere and Macdonald (1986) to establish bounds on geomorphic provinces of the OSC. Geomorphic provinces each have a distinct terrain pattern. Features observed within each province from near-bottom photography and side-scan sonar are used to understand the origin of the geomorphic differences within the provinces. Photographs from the Jason 2 ROV and WHOI TowCam are analyzed and categorized by flow type (pillows, lobate, sheet, hackley and curtain flows), thickness of sediment cover and the presence of glass for each province and the whole OSC. Fissures and pillow mounds are manually digitized from DSL-120A side-scan sonar collected over the OSC. This allows us to determine how the distribution of mounds, fissures, lava morphologies and scarps differ between provinces and how changes in the location of the melt lens as well as its chemical make up and size may influence the seafloor morphology.
T33B-1352 INVITED
Implications for Magma Supply to Overlapping Spreading Centers at 9N East Pacific Rise from Widespread Low Effusion Rate Eruptions for the Past 2 M.Y.
Zero-age seafloor along the crest of mid-ocean ridges deepens in the vicinity of over 95% of well-mapped ridge axis discontinuities such as transform faults and overlapping spreading centers. Ever since the discovery that the ridge axis deepens at ridge axis discontinuities, there has been vigorous debate about whether ridge axis depth and shape reflects magma supply. The East Pacific Rise (EPR) is an ideal location to investigate the processes taking place at ridge discontinuities because the fast-spreading rate combined with long spreading segments and relatively little off-axis deformation provides a clear record of the history of a discontinuity by preserving the seafloor on the ridge flank. Here we report evidence, from one of the most thoroughly studied ridge discontinuities on a fast-spreading mid-ocean ridge, that volcanic eruption effusion rates have been low relative to the mid- segment regions for at least the past 2 million years since the discontinuity began. This is inferred from the distribution of fields of mounds which are much more abundant in the off-axis discordant zone of the EPR 9° N overlapping spreading centers than on typical EPR seafloor. These mounds are constructed primarily by pillow lava which is erupted at lower effusion rates along the EPR than lobate and sheet lavas. This new observation, combined with a host of previously reported measurements ranging from the distribution of hydrothermal faunal communities to crustal magnetization and geochemistry, supports the hypothesis that segmentation of fast-spreading mid-ocean ridges reflects the pattern of melt generation and supply from the underlying mantle. These new observations indicate that this pattern was sustained on the EPR at 9-10° N for at least 2 My.
T33B-1353
Lava Dome Formation Along a Superfast Overlapping Spreading Center, 28-32 S on the East Pacific Rise
Deep-towed DSL-120 bathymetric data reveal the pattern of axial lava dome formation along a superfast spreading portion of the southern East Pacific Rise (EPR). The data span the overlapping limbs of the giant (~120 km offset) propagator system near 30 S. Domes are identified using a closed-contour algorithm, with auto- detections manually edited to account for the bathymetric gaps associated with the sonar system's nadir region and topographic shadowing. Along non-overlapping portions of the rift system with a full spreading rate of ~150 mm/yr, lava domes are rare and exhibit low relief relative to those imaged in equivalent datasets along the 17-19 S portions of the EPR [White et al., 2000 JGR]. As the axial depth increases and spreading rate decreases along the overlapping portion of the rift, the density of lava domes increases abruptly. Similarly, near-bottom photographic data show that pillow lava morphologies are restricted to these overlapping sections of the ridge crest. Both observations are consistent with the eruption of lower effusion rate lavas near the ends of these first- order segments. The occurrence of higher lava dome density near the ends of finer-scale segments, however, is less evident from our observations. http://web.mac.com/drbohnen/iWeb/MG_NCSU/Welcome.html
T33B-1354
AUV Mapping of Axial Seamount, Juan de Fuca Ridge: The Southern Caldera Floor and Upper South Rift
During September 2006 and August 2007 NOAA NeMO cruises, we conducted 7 high-resolution near-bottom seafloor mapping surveys of Axial Seamount using the MBARI Mapping AUV D. Allan B. The 200 kHz multibeam and 110 kHz sidescan surveys of the south caldera and upper south rift, conducted at 50 m vehicle altitude, achieved sub-meter resolution bathymetry and sidescan imagery. Numerous previous and concurrent submersible or ROV dives provide ground-truth of what the maps depict. A companion poster presents the AUV surveys of the north caldera and northeastern caldera rim. The southern caldera wall is buried beneath at least 5 voluminous lava flows, including the 1998 flow, each erupted from fissures extending along the southeastern edge of the caldera roughly parallel to the upper south rift zone. The caldera wall here was not as tall as on the southwest, north, and northeast of the caldera, and may have been as low as 35 m tall before it was buried. Active and inactive hydrothermal vents are generally located along the inferred buried caldera wall. Eruptive fissures are characterized by series of depressions aligned along each fissure; no ramparts or other constructional edifices were constructed along them. The aligned depressions suggest that lava drained back down the fissures at the end of the eruptions. The fissure eruptions were large volume and had large effusion rates as seen by their interwoven channels and the extent of the flows. Most of these flows have central channels of lineated sheet flows, bordered by folded and then jumbled sheet flows, surrounded by lobate flows with lava pillars and collapse structures and pillowed flow margins. As an example, the 1998 eruption in and near the caldera issued from 5 en echelon fissures extending at least 3 km. The largest flow lobe extending to the south was mapped along its entire western boundary using JASON II, but the flow extends to the southeast beyond the mapped region. An unusual km-across feature was mapped near the center of the caldera. Appelgate and Embley (1992, Bull. Volc. 54, 447-458) inferred that this was a large tumulus based on partial imaging using sidescan sonars. The feature is clearly formed by inflation beneath a rigid lava crust, but does not appear to be above a tube system nor pressurized by hydrostatic head produced as lava flowed downhill. Instead, this feature appears to be a low-relief dome formed of fluid lava. To produce such a feature atop a vent, the eruption rate must have been very low and the thick crust prevented the flow from advancing. As it domed upward, the thick brittle crust cracked and separated, producing fissures as observed on tumuli.
T33B-1355
AUV Mapping of Axial Seamount, Juan de Fuca Ridge: The Northern Caldera Floor and Northeast Rim
During September 2006 and August 2007 NOAA NeMO cruises, we conducted 7 high-resolution near-bottom seafloor mapping surveys of Axial Seamount using the MBARI Mapping AUV D. Allan B. The north caldera floor and northeastern caldera rim were surveyed with 200 kHz multibeam and 110 kHz sidescan from altitudes of 50 and 90 m, respectively, achieving bathymetry and sidescan imagery lateral resolution from < 1 m to 1.5 m. Numerous previous and concurrent submersible or ROV dives provide ground-truth of what the maps depict. A companion poster presents the AUV surveys of the south caldera and south rift zone. Numerous young fissures extend north from the center of the north caldera wall along the north rift zone and northward from the northeastern corner of the caldera. One large north rift fissure cuts the caldera wall, extending 300 m south across the caldera floor and containing the CASM hydrothermal vent site. A few narrow fissures, in line with the north rift, crosscut older pillowed flows near 45°57.6`N. The east rim of the caldera is cut by numerous fissures, which are continuations of the south rift zone. These fissures extend as far north as 45°59.4`N and form a subdued horst-and-graben terrain. Some of these fissures erupted voluminous pillow flows on top of 2 m of volcaniclastic sediment on the northeastern rim. The sediment overlies channelized fluid flows predating caldera formation. The youngest north caldera flow erupted from the CASM fissure, partially covering two other large flows erupted from the same region. The CASM flows have central channels of lineated sheet flows and jumbled sheet flows, flow margins of lobate flows that grade outward to pillowed margins, and only rare collapse structures. Along channel margins in the jumbled sheet flows, crustal fragments have stacked up into lava spires several m tall. One of two large flows erupted near the eastern caldera wall flowed N and the other flowed W until it encountered the western caldera wall, then turned N and finally NE. The distal end of this flow has a peculiar mottled pattern in the bathymetric data; it consists of jumbled sheet flows with small ridges of broken crusts. A 150-m across pillow cone with a 32 m deep central crater is apparently the oldest volcanic feature in the north caldera floor. The southern flank of this cone is truncated by a 070° scarp. Along the northeast caldera wall there is a smooth lava pond with a jumbled sheet flow surface. The pond has a flat bottom (1580 m depth) and is bordered to the east and south by a 4 m high lava highstand (bath-tub ring). Inside the pond, we observe 12 mounds consisting of uptilted slabs of pond crust draped over large pillars. A second small region of flat pond surface, located just SE of the cone, is at the same depth and may have once been contiguous with the eastern pond. Adjacent to the small pond is a segmented ring of spatter levees with channels cut through in all directions.
T33B-1356
Mechanical Interaction between volcanoes in the East Volcanic Zone of Iceland
When assessing volcanic hazards, it is important to know if volcanoes within a given area interact mechanically, that is, is unrest in one volcano likely to trigger unrest in one or more nearby volcanoes. Here we report a study of the mechanical interaction between central volcanoes within a part of the Neovolcanic Zone of Iceland. The Neovolcanic Zone is the area confining the majority of volcanism in Iceland today, containing rocks belonging to the Bruhnes magnetic epoch (<0.8Ma). The Neovolcanic Zone can be divided into 3 subzones: the North Volcanic Zone (NVZ), the West Volcanic Zone (NVZ) and the East Volcanic Zone (EVZ). All the volcanoes that we have studied belong to the EVZ. The prominent geological features of the Neovolcanic Zone are the volcanic systems, of which there are 30. 19 of these systems hold 23 central volcanoes, indicating that some systems hold more than one central volcano. A central volcano, by definition, has a shallow crustal magma chamber. The presence of a central volcano within a volcanic system indicates a level of maturity within it; for example, the Vestmannaeyjar Volcanic System, the southernmost system in Iceland, is thought to be developing a central volcano, and thus is a young volcanic system. The central volcanoes in the EVZ were modeled as two-dimensional structures, namely as circular holes or inclusions in an elastic crust, under a tensile stress of 5MPa in a direction parallel with the spreading vector. The modeling shows several results which allow an insight into the characteristically different volcanic activity within the EVZ. The first main result is that the modeling clearly indicates the presence of two clusters of central volcanoes: one at the northern end of the EVZ, under the Vatnajökull ice cap and over the mantle plume; the second at the southern end of the EVZ, and possibly influenced by the propagation of the rift-zone front of the EVZ to the southwest. The second main result is that the extent of mechanical interaction between the central volcanoes, as indicated by local stress fields, is governed by distance. There are no interactions that span the distance between the two clusters at the northern and southern end of the EVZ. Remarkably, this part of the EVZ, namely the part located between the central-volcano clusters, is the site of the largest Holocene fissure eruptions in Iceland. Whilst these models do not offer any answers to the question of why these large fissure eruptions take place in this part of the EVZ, the models suggest that the mechanical conditions for large fissure eruptions are most favorable away from clusters of central volcanoes.
T33B-1357
Structure and segmentation of the eastern Gulf of Aden basin and the Sheba ridge from gravity, bathymetric and magnetic anomalies: implications for accretion processes
The eastern Gulf of Aden is a key place for investigating seafloor spreading processes and the evolution in space and time of the margin and ridge segmentation. The rifting of the Gulf that separated Arabia from Somalia started around 35 Ma ago followed by oceanic accretion from at least17.6 Ma. Bathymetric, gravity and magnetic data from the Encens-Sheba cruise are used to study the structure and segmentation of the eastern part of the basin and ridge, which have strong implications for accretion processes. The segmentation of the first oceanic spreading centre, which is dated at least 17.6 Ma by the magnetic anomaly (A5d) identification, seems to be directly related to the structural geometry of the margins. Then, magmatic processes governed the evolution of the segmentation. The segmentation of the oceanic crust evolved, by eastward propagation of the western segment, from three segments (from an5d to an5) to two segments (from an5). At 6 Ma (an3a) a third segment appeared by duplication of the Socotra transform fault, maybe due to a regional kinematics change. The Encens-Sheba oceanic domain is divided in two distinct areas trending NE-SW perpendicular to the Sheba ridge. (1) The Eastern area is characterized by a shorter wavelength variation of the axial segmentation with two spreading segments 30 to 40 km long, and by a thin crust particularly on the northern and southern ends of its flanks. (2) The Western zone, whose axial segment is more than 120 km long, is characterized by a thick crust and/or a hot mantle and no axial rift valley. This abnormal volcanic activity for a slow spreading ridge is emphasized by bathymetric highs with 5-10 km wide volcanic edifices, and by a negative anomaly of the MBA. These different results support the presence of an off-axis thermal anomaly located below the southern flank of the Sheba ridge. The magnetic anomalies and spreading asymmetry reveal that the location of this thermal anomaly might be relatively recent (~ 10 Ma). We propose that at 10 Ma the ridge jumped southwards as attracted by this thermal anomaly, which explains the asymmetry observed between the 5c and 5 anomalies (16 to 10 Ma).
T33B-1358
Shear velocities in the oceanic crust at the East Pacific Rise 9° 18' N to 10° 30' N from compliance measurements
Compliance was measured at 21 sites along the East Pacific Rise (EPR) from 9° 18' N to 10° 30' N during the MADCAP (Melt And Diking from Compliance And Pressure) experiment on the R/V Atlantis from February 13 to March 19, 2007. Measurements at 10° 30' N across the ridge segment 22 km north of the Clipperton transform fault show a stiff lower crust, which suggests that there is little crustal melt. This is consistent with previous descriptions of this segment as "magmatically starved" based on its morphology. Most of the compliance measurements were made on the EPR segment south of the Clipperton transform fault. At the northern end of this ridge segment, a compliance transect at 10° 2' N spans the ridge axis and continues to a seamount 16 km east. These measurements indicate that shear velocities are low beneath the ridge axis but increase rapidly off axis to the east, suggesting no magmatic connection between the ridge axis and the Watchstander seamount chain. Shear velocities beneath the nearest (and most recently active) seamount are similar to other off axis sites, suggesting that there is little or no crustal melt there. A 26 km long compliance transect across the ridge axis near 9° 20' N suggests that the region of low crustal shear velocities is constrained to within 3-4 km of the ridge axis. The compliance measurements preclude the existence of a melt body 18-20 km east of the ridge axis as had been inferred from an apparent mid-crustal reflector observed in a recent OBS experiment. The compliance over that site show low shear velocities only in the uppermost crust associated with a thick layer 2A. A final compliance transect stretches northward along the ridge axis and across the Clipperton ridge-transform intersection (RTI). Measurements made north of where the ridge crosses the inferred location of the RTI show lower crustal shear velocities than normal for off-axis crust, but this observation is consistent with previous refraction work that suggested extensive alteration of the mid-crust and a thick layer 2A north of the transform. Within 9 km south of the RTI, lower crustal shear velocities on axis decrease to values comparable to other sites on-axis along this ridge segment suggesting the robust magma supply extends to within 9 km of the RTI.
T33B-1359
Is Skew of Mantle Upwelling Beneath Spreading Centers Globally Significant? Observations (EPR), Hints (MAR) and Predictions (Juan de Fuca)
One popular and long-held view for the origin of spreading center segmentation is the magma supply hypothesis, wherein magma is centrally injected at mantle depths into a ridge segment bounded by long-lived tectonic discontinuities and ridge parallel flow of magma governs segment-scale variations in magmatic and tectonic processes. An alternative hypothesis is that tectonic forces transmitted through oceanic plates are responsible for the creation and migration of ridge axis offsets. These two hypotheses are often considered as the end- member alternatives for the origin of ridge crest offsets. We have recently proposed a new hypothesis - that does not fit neatly between the two end-member alternatives noted above - for the origin of segmentation along fast-spreading ridges. In our model, segmentation of the East Pacific Rise (EPR) is a result of a skew between the axes of mantle upwelling and plate spreading. As a result of this skew, mantle upwellings organize into en echelon segments and large-offset overlapping spreading centers (OSCs) provide an accommodation zone that moves the lithospheric plate boundary from one segment of mantle upwelling to another. Between transforms and OSCs, the intrasegment variability in ridge crest processes is related to the cross-axis distance to a center of mantle upwelling: Where mantle upwelling intersects the plate boundary, volcanic eruptions are frequent, seafloor hydrothermalism is intense and persistent and the average lava composition is more primitive. By contrast, where mantle upwelling is offset from the axis of spreading the seafloor is more tectonized, eruptions are less frequent and the average lava composition is more evolved. Here we address the following question: Is skew of mantle upwelling a feature peculiar to the EPR near 9°N, or is there evidence that it is a global phenomenon? A key observation supporting our hypothesis is the azimuthal rotation of seismic anisotropy, and by inference mantle flow, beneath the spreading axis. We present models supporting our interpretation that the azimuth of mantle divergence controls the observed seismic anisotropy. We then reinterpret results from the Mid-Atlantic Ridge (MAR) that also reveal an azimuthal rotation of seismic anisotropy beneath the ridge axis. A synthesis of results from the MAR further suggests that the relation of skewed mantle flow to tectonics is similar to that observed along the EPR and that variations in lithospheric thickness augment focusing of melt toward the segment center. Lastly, we make predictions for the orientation of mantle anisotropy beneath the Endeavour segment of the Juan de Fuca ridge. An upcoming seismic experiment to test these predictions is scheduled for 2008.
T33B-1360
Tectonic and Magmatic Modification of Ocean Crustal Bulk Physical Properties: A Global Perspective
We compiled crustal physical property data such as porosity, density, and velocity from several decades of ocean drilling to compare normal ocean crust with crust influenced by fracture zones, hotspots, hydrothermal alteration, and other magmatic or tectonic influences. Cores and holes from IODP Expeditions and ODP and DSDP Legs that have penetrated 100 m or more into basement have been divided into the following geologic settings: a) normal crust, b) tectonic windows, c) off-axis seamounts, d) mid-ocean ridge seamounts e) off-axis plateaus and f) areas of arc volcanism. This study aims not only to provide a more global perspective of the ocean crust with respect to factors such as geologic setting, age, and rock type from many seafloor holes globally that have yet to be compared, but also to focus more specifically on how normal crustal properties compare with those of crust which has interacted with other tectonic and magmatic zones. Preliminary analyses of the different geologic settings with respect to their average physical properties without regard for differences in rock type, depth, age, or composition show some interesting trends. Although the average crustal velocity is highest in normal oceanic crust, we find no significant difference between normal crust and any other setting. All the settings exhibit variability in the upper 200 m of crust, but below 200 m the velocity of normal oceanic crust never drops below 4 km/s. The average porosity of normal crust is similar to that of tectonic windows, but much lower than the average of any of the other geologic settings. We find that the average bulk density is again highest in normal oceanic crust and similar to the average of tectonic windows, but normal crust is not significantly different from the other settings. Average grain density is also highest in normal crust. The present data suggest a comprehensive model of both normal crust and crust that has been influenced by tectonic and magmatic processes. Using physical properties and the identifying the controlling factors such as hydrothermal alteration, mechanical collapse, and faulting, we can synthesize the global datasets provided by ocean drilling during the last four decades.
T33B-1361
Segment-Scale Variations in Crustal Thickness Along the Southeast Indian Ridge From 2-D Body Wave Tomography
The Southeast Indian Ridge (SEIR) west of the Australian Antarctic Discordance (AAD) exhibits a nearly constant intermediate spreading rate (72-76 mm/yr) and dredge sample basalt geochemistry despite systematic changes observed in the depth and morphology of the ridge axis. These unique conditions make it possible to investigate how variations in melt production (or equivalently, crustal thickness) correlate with axial morphology along the SEIR. A total of six seismic refraction experiments were conducted along SEIR segments P1, P2, S1, and T as part of R/V Maurice Ewing cruise EW0114 (Dec 2001-Jan 2002). The 100-125 km lines were positioned above zero- age crust for all four segments as well as 20 km south of the axis (~525 ka) for segments P1 and P2. In each survey, four ocean bottom hydrophones deployed 15-20 km apart recorded shots fired by the 8480 in3 20-gun airgun array. P wave travel times for crustal refractions (Pg) and Moho reflections (PmP) were hand-picked and assigned error values ranging from 15-150 ms based on the trace signal-to-noise ratio. Starting with 1-D crustal velocity-depth profiles, we constructed an overparameterized 2-D inverse problem and derived minimum structure solutions by regularizing the inversion with predefined damping and smoothness constraints. The forward step calculated ray travel times by combining shortest path (graph) method results with a beta-spline ray- bending refinement scheme. We applied a top-down methodology to the inversion, including picks from progressively larger instrument offsets with each iteration to account for the dominant influence of upper crustal heterogeneity on travel time variations. Our preferred models show a consistent thinning of the crust toward the AAD coincident with the transition from a robust axial high (P1) to a well-developed axial valley (T). We performed a direct assessment of model resolution using checkerboard tests, nonlinear Monte Carlo analysis, and a variable depth kernel weighting strategy. These tests indicate our results are robust, which suggests the relationship between melt production and ridge character behaves in a stepwise fashion where changes in melt supply and crustal thickness are expressed by abrupt shifts in the depth and morphology of the ridge axis.
T33B-1362
ASTERICS (= Geophysical Studies near the Ascension Transform: Evolution of Ridge Segmentation and Crustal Structure) - Segment A0/A1
The ASTERICS project investigates the crustal and upper mantle structure in the area of the Ascension Transform (AT) at the Mid-Atlantic-Ridge (MAR). The AT is a "double transform fault" consisting of two parallel transform fault/fracture zone systems sandwiching a very short segment. The project investigates the tectonics and structure of the double transform as well as the segments south of it. Here four spreading segments can be identified. Segment A1 is formed by a typical axial ridge with a distinct graben. To the South segment A2 continues with an intermediate shaped seafloor topography, while the segment A3, is marked by a well-defined axial high. Subject of this poster are seismic models derived 1.) from data collected along a 200 km long wide–angle seismic profile (OBS), which has been recorded along the center of the SAFZ and 2.) from data collected along a 85 km long profile across the well pronounced MAR of segment A1. Slightly offset to the East an axial ridge like graben is observed between the two transforms at the western end of the AT. The velocity model derived from a first arrival tomography after Korenaga et al. (2000) of the MAR in segment A1 shows a high of the mantle 15 km east of the axial graben. A tongue of reduced velocities (6-6.5 km/s) indents from West underneath it. Together with the elevated seafloor on the Eastern flank of this MAR segment this setting forms the typical structure expected with Inside-Corner-Highs (ICH) and Outside-Corners (OC). From available bathymetry no corrugated surface has been identified, arguing for a young stage of development. http://www.ifm- geomar.de/index.php?id=asterics&L=1
T33B-1363
ASTERICS (= Geophysical Studies near the Ascension Transform: Evolution of Ridge Segmentation and Crustal Structure) - Segment A3
The project ASTERICS deals with the investigation of the crustal structure in the area of the Ascension Transform (AT) at the Mid-Atlantic-Ridge (MAR). The AT is a 'double transform fault' consisting of two parallel transform fracture zone systems sandwiching a very short segment. The aim of the seismic studies carried out within ASTERICS is to investigate the tectonics and structure of the double transform and to constrain the processes of segmentation occuring on the MAR just south of the South Ascension Fracture Zone (SAFZ). South of the SAFZ four spreading segments can be identified. Segment A1 is formed by a typical axial ridge with a distinct graben. To the south segment A2 continues with an intermediate shaped seafloor topography. Segment A3, the topic of the poster, is marked by a well-defined axial high that might correspond to a site of excess magmatism. The MBA (Mantel-Bouger-Anomaly) of earlier measurements shows a well-developed bulls-eye gravity low here consistent with either thickened crust and/or active mantle upwelling. Wide angle studies with Ocean-Bottom- Seismometers (OBS) and deep towed multichannel seismic streamer profiles are used to investigate the crustal fabric in order to distinguish between the two models. Forward modelling with MacRay was applied to inline traveltime observations from a grid of five profiles. The resulting models show a high of the uppermost crustal layers in the centre of each profile. Further model development using the First Arrival Seismic Tomography (FAST) code confirm this. Within the inversion models lower crustal velocities (> 7 km/s) are found to spread almost horizontal across the model. Mantel velocities (> 7.5 km/s) are found at about 10 km depth. Exact position of the MOHO will be confirmed by modelling of PMP reflection events. First tests in MacRay expect a flat MOHO at a depth of about 10 km. These observations argue for an increased crustal thickness of about 8 km in the centre of the high, compared to 6 km - 7 km at the sides in the segment A3. Which is much thicker than observed along segment A1 (about 4 km). 3-D inversion of OBS data recorded within a grid of 12 profiles will be accomplished with FAST to further image the crust with more detail. http://www.ifm- geomar.de/index.php?id=asterics
T33B-1364
Layer 2A thickness variations and upper crustal emplacement on the Lucky Strike segment (37N), Mid-Atlantic Ridge
We present seismic reflection images from a grid of 29 "2D" shot lines spaced approximately 2.5 km apart collected along and across the Lucky Strike segment of the Mid-Atlantic Ridge during the 2005 SISMOMAR cruise, part of the MOMAR (MOnitoring the Mid Atlanic Ridge) program. These shot lines provide reflection data out to the segment end and up to 30 km off-axis. The air guns were tuned to generate a single bubble and had an array volume of 8410 cubic inches for 25 of the lines (150 m shot spacing) and 5658 cubic inches for the remaining four lines (75 m shot spacing). Data were recorded on a 360 channel, 4.5 km long digital streamer. We use a 2D data processing approach for each line, applying frequency filtering, normal move out correction, stacking and 2D poststack depth migration. Each line was stacked using the same 1D subseafloor velocity profile to assure line- to-line comparability. The velocity profiles for the depth migration were derived from a coincident 3D seismic refraction experiment using 25 ocean bottom seismometers. The central along-axis reflection line reveals that the layer 2A reflector depth is nearly constant over the middle 44 km of the approximately 65 km long segment. In this region, the 2A reflector and the 4.5 km/s isovelocity contour lie about 0.6 km beneath the seafloor, except within the central 4 km of the segment, where the reflector shallows to about 0.45 km. At both segment ends the 2A reflector shallows whereas the isovelocity contour deepens. If the 2A reflector marks the transition between extrusive and intrusive volcanics, the extrusive crustal thickness is nearly constant along two thirds of the segment length, far beyond the extents of the central magma chamber. The thinning of layer 2A and thickening of the isovelocity contour beyond this zone suggests a transition from a magmatic/diking regime to a tectonic regime, in which the extrusive layer marked by layer 2A reflection is formed by surface lava flows from the magmatic/diking zone or by occasional horizontal or vertical magma injections, whereas rock fracturation has lowered the upper crustal velocities. We present a 30x40 km image of layer 2A thickness over the central part of the segment. The 4.5 km/s isovelocity surface varies little along axis, except for some thickness anomalies at the central volcano and its northwest ridge, which we interpret as sites of recent magmatic accretion. The isovelocity surface varies more across axis, thinning significantly off-axes at the valley bounding faults. We will correlate these variations with variations in the reflector depth and interpret the results in terms of emplacement processes and hydrothermal and mechanical alteration.
T33B-1365
Crustal Structure of a Propagating Ridge Segment from Seismic Refraction and Wide-angle Data
The Mid-Atlantic Ridge at 21.5° N shows a very clear structure of a propagating ridge segment in both bathymetric and satellite altimetry-based gravity. This active ridge system in a median valley environment started its southwards propagation roughly five million years ago, after a stable transform fault had existed there for several million years. A linear slightly asymmetric v-shaped wake has been formed by a propagation rate of approx. 16 mm/y with an average half-spreading rate of approx. 13 mm/y. The experiment comprises five active seismic profiles, with ocean bottom hydrophones deployed at 80 stations, shot during the cruise M60-2 in 2003/2004. One profile has been placed along the ridge axis while the other profiles cross the ridge segment in its different zones to estimate the velocity structures for these areas. Results from the active part of the experiment indicate a strong variation in the crustal thickness along the ridge. While the crust has unusual thinning towards the north in the deeper region, it becomes thicker at the doomed ridge. The small scale seismic activity, recorded by a temporal NOAA/PMEL hydrophone network, is confined to the northern deeper region, where mantle rocks outcrop at the seafloor. The segment centre with its dome-like topography and the south of the segment, the ridge-tip, correlate with gaps in seismic activity. The aseismic zone of the ridge at the topographical high hints at a thin lithosphere at the centre in contrast to a thicker lithosphere beneath the basin with minimal volcanism. The model along the ridge axis corresponds to the idea of focussed melt supply and lateral redistribution at crustal levels which results in a thick crust at the segment centre with thinning towards the segment ends. In the process of segment propagation apparently a fraction of the melts rising at the centre of the segment are redistributed laterally before emplacement. However, upto 8 km thick crust in the segment centre suggest that a major part is used locally for crustal formation, while crust at the segment ends is 5 to 6 km thick.
T33B-1366
Partial melting and mantle dynamics at the Fifteen-Twenty Fracture Zone (Mid-Atlantic Ridge)
During ODP Leg 209, a magma-starved portion of the Mid-Atlantic Ridge (MAR) was drilled in the vicinity of the Fifteen-Twenty Fracture Zone (FZ). Mantle peridotites sampled at Sites 1272 and 1274 (to the south and the north of the FZ, respectively) represent the most depleted peridotites ever sampled at a slow spreading ridge, Site1272 representing the most depleted end-member. They were formed by high degrees of melting (>17%). In addition to melting, Site1274 peridotites show evidence for late melt transport and melt freezing reactions. Sites 1270 and 1271, drilled in the vicinity of Site 1272, sampled highly reacted and impregnated peridotites with fertile compositions similar to peridotites from the mantle-crust transition in ophiolites. In the Fifteen-Twenty FZ area, the compositional variability of the mantle does not relate to different degrees of partial melting on each side of the FZ but to variable degrees of melt-rock interaction. The formation of highly refractory peridotites with, locally, high melt focussing areas implies a more actively convecting mantle than generally inferred below slow spreading centers. We propose different scenarios to explain our observations: (i) melting during a period of high magmatic activity followed by present-day exhumation during amagmatic rifting or (ii) formation of shallow micro-convection cells close to FZ-ridge intersections. In addition, the complex and dynamic ridge system that characterizes the Fifteen-Twenty FZ area has probably favoured the sampling of melts formed at different depths and exposure of parts of the upwelling mantle rarely exposed in oceanic mantle windows elsewhere.
T33B-1367
Thermally induced brittle deformation in oceanic lithosphere and the spacing of fracture zones
Brittle deformation of oceanic lithosphere due to thermal stress is explored with a numerical model, with an emphasis on the spacing of fracture zones. We use a parallel FEM code, SNAC, which explicitly solves the momentum balance and heat energy equations. The problem domain is 500 km long and 50 km deep. The 2-D plane strain condition is imposed such that the domain represents a vertical ridge-parallel cross-section. Effects of cooling on momentum balance are two-fold: Viscosity is temperature-dependent and volumetric strain has a contribution from thermal contraction. Brittle deformation is represented by localized plastic strain within a material having an elasto-visco-plastic rheology with strain softening. Fracture zones are brittle zones that extend to the surface creating topographic features. Crustal thickness, creep strength, and the plastic flow rule are shown to control the formation of fracture zones. There is a threshold crustal thickness such that fracture zones are created when crust is thinner than the threshold. Creep strength shifts the threshold in such a way that lowered creep strength can completely prohibit the formation of fracture zones even if crustal thickness remains the same. Through finite versus zero dilatation in plastic strain, associated and non-associated flow rule results in nearly vertical and V-shaped primary cracks, respectively. Changes in the tectonic environment of a ridge system can be reflected in variation in crustal thickness, and thus related to brittle deformation. The slow- and oblique-spreading Reykjanes ridge is free of transform faults and fracture zones. It is known to have a uniform and larger-than-average crustal thickness (≥10km) due to the vigorous magma supply from the Iceland hotspot. The Australian-Antarctic Discordance is an anomalous portion of the intermediate-spreading Southeast Indian Ridge (SEIR), spanning from 120°E to 128°E. Compared to the normal parts of the SEIR, its notable characteristics include highly rugged topography, deeper residual depth, increased number of fracture zones, and thin crust (~4km). These syntheses are consistent with increased brittle deformation of oceanic lithosphere when the crust is thin and vice versa.
T33B-1368
Thermal regime of young seafloor : how important are the fluids?
The statistical analysis of the heat-flow data base has been often used to infer general outcomes on the heat transfers and processes throughout continental and oceanic lithospheres. The most important outcome was the relationship between the magnitude of hydrothermal heat-loss and the age of the ocean floor. It caused obviously a declining interest to make new measurements, at least at a funding level, since heat-flow was believed critically perturbed by fluids circulations. However, several recent local or regional studies are not totally consistent with such outcomes. For instance, we recently carried out two marine heat flow surveys on young oceanic sea-floor, one on the mid Atlantic ridge near Azores Islands (Luckyflux cruise, 2003) and the other in the Gulf of Aden (Encensflux cruise, 2006), near the ocean-continent transition and we got surprising results after having processed more than 330 heat flow measurements collected with the POGO method. Off-axis (6-11 Ma) heat flow values in the Atlantic are not all affected by important fluids circulations and almost half of them are quite consistent with conductive cooling models for that age. In the Gulf of Aden, most of the values obtained on the oceanic domain (6-11 Ma) can be considered as not thermally perturbed by fluids. Our results also show that where perturbations exist, basement exposures are present, but not all of these exposures are associated with thermal perturbation. These observations are not only important for the understanding of hydrothermal processes, but also to promote new heat-flow surveys for geodynamics purposes.
T33B-1369
Li Isotopes as Tracers of Fluid-Rock Interactions in Oceanic Hydrothermal Systems: Evidence From the Hess Deep Rift
Li isotopes have the potential to be powerful tracers of fluid-rock interactions at mid-ocean ridge hydrothermal systems due to the large isotopic difference between seawater (δ7Li = 31‰) and fresh MORB (δ7Li = 1.5 to 5.6 ‰). Sampling programs along tectonic escarpments at Hess Deep provide an ideal suite to examine the spatial variability of Li mobility and Li isotopic content within young (1 Ma) lavas and sheeted dikes formed at the fast-spreading East Pacific Rise towards the end of a segment. Previous work has shown that the lavas are relatively fresh, with minor alteration to clay minerals and Fe oxyhydroxides. Sheeted dikes are variably altered to amphibole-dominated assemblages, with localized zones where chlorite dominates. Sr and O isotope data correlate with these regional patterns. Preliminary data, collected by Thermo X-series quadrupole ICP-MS, show that the lavas have similar Li concentrations to fresh MORB (5 to 6 ppm) and are slightly enriched in δ7Li (4.1 to 7.7 ‰). There is a positive correlation between Li content and δ7Li within the lavas, however, the trend is not a simple mixing line between seawater and fresh MORB, being offset to lower δ7Li in the altered component. Similar to previous studies, these relationships support the prediction that 6Li is preferentially retained in low temperature clay minerals. Sheeted dykes are depleted in Li (0.8-4.63 ppm) and most samples are enriched in δ7Li (6.7-15.4 ‰) relative to fresh MORB. These samples show an inverse relationship between decreasing concentration and increasing isotopic enrichment. This implies that 6Li is not simply preferentially retained in the replacement mineral assemblages. The degree of isotopic enrichment appears to increase with increasing clinopyroxene alteration, and is greatest for amphibolite dominated assemblages. A broad positive correlation between Li concentration and δ18O is observed in the dikes. Thus Li isotopes are sensitive indicators of the nature and extent of hydrothermal interactions for both high and low temperature processes.
T33B-1370
Correlations Between Hydrothermal Venting and Axial Magma Chamber Characteristics
The principal power sources for high-temperature hydrothermal venting have long been thought to be heat mined from solidified magma ("hot rock"), as by a propagating cracking front, and heat supplied conductively from an unsolidified magma body ("axial magma chamber"). However, nearly all high-temperature vents have been found at, or near, locations where seismic studies have imaged an AMC. This association seems especially true on fast-spreading ridges, where lengthy seismic transects have been conducted on several ridge sections. To test the robustness of this association, I here review data from all six multi-segment ridge sections, ranging from 170- 560 km in length, where detailed surveys of both the AMC and hydrothermal plumes have been conducted. The sections include the southern East Pacific Rise (14°-18° S) (560 km), Galapagos Spreading Center (380 km), Juan de Fuca Ridge (365 km), Valu Fa/Eastern Lau Spreading Center (360 km), northern EPR (9° - 12° N) (307 km), and northern EPR (15.3° -16.8° N) (170 km). These sections include a total of 21 2nd or 3rd order tectonic segments. At the section scale, there is no significant correlation (r2 = 0.05) for a least-squares fit between percent AMC coverage and percent plume coverage (ph). Removing the hotspot-affected GSC, however, increases r2 to 0.61. This finding is consistent with evidence showing that hydrothermal activity on hotspot-affected ridges is markedly less than normal ridges of similar spreading rate. The correlation between mean AMC depth below the seafloor and ph for all six sections is high (r2 = 0.58), and improves to r2 = 0.87 if the Valu Fa/ELSC section is excluded. The AMC is unusually deep and the crustal structure along this section is unique along among midocean ridges, perhaps because of its proximity (40-100 km) to the Kermadec volcanic arc. At the individual segment scale, the same correlations are much less robust. AMC coverage vs. ph decreases to r2 = 0.15 whether or not the Galapagos data are included. AMC depth vs. ph decreases to r2 = 0.3 whether or not the Valu Fa/ELSC data are included. This analysis demonstrates that the correlation between the spatial density of venting and the spatial extent or depth of an AMC is weak on the small segment scale but significant at the multi-segment scale. This pattern is the same as found between ph and spreading rate on the small and multi-segment segment scales. In contrast, at both scales venting is more sensitive to AMC depth than AMC extent, a correlation likely linked to the observation that the AMC shallows as spreading rate increases. It may be that ph increases with spreading rate not simply because of a higher magma budget, but also because heat extraction becomes more efficient as AMC depth decreases.
T33B-1371
Hydrothermal Heat Output from a Convecting, Crystallizing, Replenished Sub-Axial Magma Chamber
Models of high-temperature seafloor hydrothermal systems require that heat is transferred from an underlying magma body across a conducting boundary layer to the hydrothermal system. Because magma is typically at or near its liquidus, heat transfer will result in crystallization and cooling of the magma itself. In previous models of magma cooling and crystallization, solidification was assumed to occur from the top downwards. Consequently, the conducting thermal boundary layer between the hydrothermal system and the magma body rapidly thickened, resulting in a concomitant decay in the hydrothermal heat output and vent temperature. We present a simple time-dependent model of heat transfer between a turbulently convecting crystallizing thin basaltic magma lens and the overlying hydrothermal circulation. Two different crystallization scenarios are considered—crystals in suspension and crystals settling. In either case, we assume that large-scale convection within the magma chamber is homogenous and can be parameterized by its Rayleigh and Nusselt numbers. Also, the effect of crystallinity-dependent magmatic viscosity is considered. The simulation results show that without magma replenishment, the heat output and the hydrothermal temperature decay markedly on a decadal time scale, though decay occurs more slowly for the crystal settling model. The rapid decay of heat transfer from a crystallizing magma lens leads us to consider the effects of magma replenishment. To investigate the heat output from a growing magma chamber, we consider replenishment at both a constant rate at an exponentially declining rate, respectively. Simulation results show that magma heat flux approaches a steady state on decadal time scales, provided the magma volume doubles during a replenishment episode of approximately 20 years duration.
T33B-1372
Biological Sulfate Reduction Rates in Hydrothermal Recharge Zones
We develop a model to determine the rate of removal of seawater sulfate in the recharge regions of deep-sea hydrothermal systems as a result of biogenic sulfate reduction. The rate of sulfate reduction as a function of temperature derived from laboratory measurements on cores from the Guaymas Basin in Mexico [Jorgensen et al., 1992] is incorporated into a steady state 1-D advection-diffusion temperature equation, and a 1-D, steady- state, advection dominated conservation of solute equation. The diffusivity of sulfate in seawater is on the order of ~ 10-10 m2/s, and unless the flow speeds are < 10-12 m/s, the effects of diffusion are negligible, except within thin diffusive boundary layers. This model is then compared with a model that utilizes Gibbs free energy to quantify biogenic sulfate reduction [Bach and Edwards, 2003] in the upper oceanic crust of aging lithosphere. Using the high rates determined by Jorgensen et al. [1992], our model indicates that biological activity would reduce all seawater sulfate transported into the system within the upper 10 meters or less of the crust, which is inconsistent with the estimates of Bach and Edwards [2003]. Sulfate concentrations from ODP borehole Legs 64 and 168, at the sedimented Guaymas Basin and Juan de Fuca Ridge, respectively, show that most of the seawater sulfate is removed in the upper 100 meters. If the sulfate is assumed to all be reduced biogenically, the sulfate reduction rates at the ODP sites are at least 2 orders of magnitude less than the laboratory estimates of Jorgenson et al. [1992]. Finally, we compare the rate of seawater sulfate removal as a result of the precipitation of anhydrite, with the rate of biogenic sulfate reduction. We find that if hydrothermal recharge occurs rapidly through highly permeable faults, that biogenic sulfate reduction is negligible and that anhydrite precipitation would rapidly clog the recharge zone [Lowell and Yao, 2002]. If recharge occurs through broad zones of slow downwelling (u <ƒn 10-9 m/s); however, anhydrite precipitation would seal pore on the order of thousands of years even in the absence of biogenic sulfate reduction. At these slower flow speeds, the biogenic sulfate reduction may provide an important mechanism for the removal of seawater sulfate from the deeper parts of the reaction zone. Bach, W. and K.J. Edwards (2003), Iron and sulfide oxidation within the basaltic ocean crust: Implications for chemolithoautotrophic microbial biomass production, Geochim.Cosmochim. Acta, 67, 3871-3887. Jorgensen, B.B., M.F. Isaksen and H.W. Jannasch (1992), Bacterial sulfate reduction above 100 degrees C in deep-sea hydrothermal vent sediments, Science, 258, 1756-1757. Lowell, R.P. and Y. Yao (2002), Anhydrite precipitation and the extent of hydrothermal recharge zones at ocean ridge crests, J. Geophys. Res., 107(B9), 2183, doi:10.1029/2001JB001289.
T33B-1373
Quantitative Analysis of a Hydrographic Transient Above Salty Dawg Hydrothermal Field, Juan de Fuca Ridge in Summer 2005
During systematic repeat hydrography cruises to the Endeavour segment of the Juan de Fuca Ridge in the summers of 2004, 2005 and 2006, we encountered a transient increase in the water column heat content above the Salty Dawg hydrothermal field. First observed in July 2005 and mapped in greater detail in August 2005, this feature was not a typical event or megaplume as potential temperature anomalies were elevated from the plume top to the seafloor. An order of magnitude increase to the volume flux from Salty Dawg would be required to generate a neutrally buoyant plume of this size based on scaling analyses. The mode of geological forcing is unknown, making this observation unique in that no earthquakes have been detected in the time frame of this increased heat flux, though unanalyzed seismic data are available. The transient was sustained over a time period longer than is expected with a typical dike intrusion. Possible forcing mechanisms for this transient include: an increase of heat in an underlying magma chamber or advancement of a cracking front, a small scale dike intrusion or aseismic crustal movement. The transient disappeared before our return in August 2006, likely due to thermal expansion of shallow host rock decreasing the permeability. Results from simple models help constrain the contribution of each hypothesized mechanism.
T33B-1374
Seismicity Along the Endeavour Segment of the Juan de Fuca Ridge: Automated Event Locations for an Ocean-Bottom Seismometer Network
From 2003-2006, the W.M. Keck Foundation supported the operation of a network of eight ocean-bottom seismometers (OBSs) that were deployed with a remotely operated vehicle along the central portion of the Endeavour Segment of the Juan de Fuca mid-ocean ridge as part of a multidisciplinary prototype NEPTUNE experiment. Data from 2003-2004 were initially analyzed during a research apprenticeship class at the University of Washington's Friday Harbor Laboratories. Eight student analysts located ~13,000 earthquakes along the Endeavour Segment. Analysis of data from 2004-2005 has to date been limited to locating ~6,000 earthquakes associated with a swarm in February-March 2005 near the northern end of the Endeavour Segment. The remaining data includes several significant swarms and it is anticipated that tens of thousands of earthquakes still need to be located. In order to efficiently obtain a complete catalog of high-quality locations for the 3-year experiment, we are developing an automatic method for earthquake location. We first apply a 5-Hz high-pass filter and identify triggers when the ratio of the root-mean square (RMS) amplitudes in short- and long- term windows exceeds a specified threshold. We search for events that are characterized by triggers within a short time interval on the majority of stations and use the signal spectra to eliminate events that are the result of 20-Hz Fin and Blue whale vocalizations. An autoregressive technique is applied to a short time window centered on the trigger time to pick P-wave times on each station's vertical channel. We locate the earthquake with these picks and either attempt to repick or eliminate arrivals with unacceptable residuals. Preliminary S-wave picks are then made on the horizontal channels by applying a 5-12 Hz bandpass filter, identifying the peak RMS amplitude for a short running window, and making a pick at the time the RMS amplitude rises above 50% of this value. The picks are refined using the autoregressive technique. We then locate earthquakes using the P- and S-wave picks and either repick or discard S-wave picks with unacceptable residuals before obtaining the final location. We have tested the method using data segments from 2003-2004 and find the catalog completeness and quality of locations is comparable to that obtained with the student analysts.
T33B-1375
Regional hydroacoustic perspective of the January 2006 eruption at the East Pacific Rise, 9º- 50'N
New information on the seismic activity from the January 2006 eruption at East Pacific Rise, 9°-50'N [Tolstoy et al., 2006; Cowen et al., 2007] was provided by the December 2006 recovery of OSU/NOAA autonomous hydrophones deployed along the EPR from 8°N-8°S and at 110°W and 95°W longitude. The hydrophones record the acoustic energy of seafloor earthquakes that propagate along the ocean sound channel with little attenuation over large distances. Frequency-magnitude relationships indicate the hydrophone catalogs are complete to body-wave magnitude 2.5 at the EPR. The hydrophones recorded data from December 2004 to 2006 and successfully recorded the seafloor spreading event and eruption that occurred on 22 January 2006. At this time the hydrophone data from 29 June 2005 through 30 January 2006 has been reviewed to compile an event count and locate earthquakes using their regional T- wave records. The hydrophones detected 255 earthquakes from the vicinity of 9°-50'N in the 3 weeks prior to 22 January (the inferred day of the eruption), then 252 events on 22 January, after which earthquake counts rapidly declined to background levels in less than 3 days. As Tolstoy et al (2006) observed, the peak seismic activity (and likely the diking event) occurred within 5 hours on 22 January, during which there is a two-hour period where events rates exceed 3 to 7 times the levels of other parts of the day. The vast majority of the detected events, however, were low magnitude and it was possible to locate only 20 earthquakes, 18 of which occurred during the 1 hour of peak earthquake activity on 22 January. These earthquakes are all located within a 35-km long area to the north and south of 9°-50'N, extending several kilometers past the section of ridge-crest where recent lava flows were observed. The earthquakes do not indicate migration occurred, although admittedly the dataset is limited in number. However the seismicity does take place in discrete time clusters, occurring in succession from being centered at 9°-50'N, south ~20 km to 9°-40'N, then back to back to as far north as 9°-59'N within the 1 hr time period. Roughly 15 hrs after the last earthquake located at the eruption site occurred a sequence of 17 earthquakes began along the western Clipperton Transform, and possibly on the eastern flank of the ridge segment to the north of 9°- 50'N, likely in response to the magma intrusion at the ridge crest. Tolstoy, M, et al., A seafloor spreading event captured by seismometers, Science, 1920-1922, 2006. Cowen, J.P., et al., Volcanic Eruptions at East Pacific Rise Near 9°-50′N, Eos, Trans. Amer. Geophys. Un., 88, No. 7, 81, 2007
T33B-1376
Tidal Influences on Tremor Activity at East Pacific Rise 9-50 N Eruption Site
Seismic activity along the East Pacific Rise (EPR) at 9-50 N was monitored by an Ocean Bottom Seismometer (OBS) array continuously from October 2003 to March 2006. These data are characterized by both discrete microearthquakes and short-period tremor signals with fundamental frequency in the 5-8 Hz band. The RMS amplitude of the velocity records within this 5-8 Hz band shows strong tidal periodicity with a spectral peak at 2 cycles per day. Early in the monitoring period, tremors commonly were initiated following discrete microearthquakes. However, closer to the January 2006 eruption, tremor activity becomes more sustained and intense, with less earthquake dependence. During this time, tremor signals also occurred preferentially during periods of positive extensional tidal strain rate (i.e., between times of maximum compression and maximum extension in the crust) and are highly concentrated along the ridge axis. Minor changes in the frequency of the tremor signals also are observed through the monitoring period. In February 2004, tremor was focused around 7 Hz with two harmonics. Precursory tremor activity in January 2006, however, showed tremor concentrated near 5.5 Hz with only one weak harmonic. Although the source of these tremor signals is not well understood, temporal changes in their properties may provide an indication of impending eruptive activity, enhancing our ability for short-term prediction.
T33B-1377
Variability in Tidal Triggering of Microearthquakes at 9˚50'N East Pacific Rise
In October 2003 through April 2004, an ocean bottom seismometer (OBS) array deployed between 9°49'N and 9°51'N East Pacific Rise (EPR) recorded over 7,500 double difference relocated microearthquakes (ML < 2.0). The deployment represents the longest period of continuously recorded microearthquake data analyzed for tidal triggering at a fast-spreading (~110 mm/yr) mid-ocean ridge. At this location, high temperature hydrothermal vents and a shallow (~1.4 km depth) axial magma chamber create a critically stressed upper crust that is susceptible to tidal influence. Although the amplitude of semidiurnal tidal stress changes at this location are small (< 2.5 kPa), seismicity is found to be tidally triggered during times of peak extensional stress associated with the solid Earth tide (Stroup et al., 2007). Furthermore, the number of earthquakes that occur during times of encouraging (extensional) stress is linearly dependent on the amplitude of peak-to-peak tidal stress. When the frequency of seismicity is compared with the combined Earth and ocean tide stresses, microearthquake activity at times of maximum extension increases to ~5 times the background frequency for stress amplitudes > 2 kPa. However, preliminary analysis of microearthquakes with along-axis depth and magnitude suggests that tidal triggering is not uniform. The along- axis variations in tidal triggering may reflect differences in the background state of stress associated with the axial magma chamber and hydrothermal vent circulation.
T33B-1378
Investigation of Deep Microearthquakes at 9 deg 50' N on the East Pacific Rise
An array of nine closely-spaced ocean bottom seismometers (OBSs) recorded microseismicity at 9 deg 50' N on the East Pacific Rise (EPR) between October 2003 and April 2004. Precise double-difference earthquake locations for more than 7,200 events were calculated from analyst-reviewed arrivals. Important insights have been gained about the anatomy of this fast spreading ridge and the implications for hydrothermal circulation above the axial magma chamber (AMC) located at approximately 1.5 km depth. Nearly 10% of the events, however, were located below the AMC and require further investigation, as they are counter to models suggesting that brittle fracture below the melt zone is not possible. Still, other data point to a solid floor to the AMC with high P-wave velocities (Singh et al., 1999). We evaluate the robustness of deep-located events on the 9 deg 50' N segment by reviewing their waveforms for excessive noise and poor phase picks, spatial distribution, time difference in P and S arrivals, event magnitude, and spectral content. Although S arrivals can be difficult to discern and easily mispicked, many events have clear S arrivals with consistent picks. Deep-located events show a similar along-axis distribution to shallow events, predominantly located near the center of the OBS array. The difference between S- and P-wave arrival times increases steadily with earthquake depth, supporting accurate locations even at greater depth. Further, the magnitude of events beneath the melt lens decreases with depth, consistent with a warming trend. Our preliminary analysis does not reveal drastic differences in spectral slope, although we might expect less high frequency content for deep events due to waves passing through partial melt or because the earthquakes originate in semi-brittle material. However, since the AMC at this location is only 500 m wide and most of the OBSs are off-axis, the seismic waves of deep events may not travel through melt at all.
T33B-1379
U-Series Disequilibria From the East Pacific Rise (9N, 10N and 11N): evidence for off-axis magmatism along fast spreading ridges?
Lavas erupted along Mid-Ocean Ridges provide important information on melt formation and movement beneath the oceanic lithosphere. Despite the fact that the majority of lavas are erupted along the spreading axis itself, it has been proposed that a small quantity of lavas are erupted off-axis at distances >5 km [1]. The East Pacific Rise has a fast (5.5 cm/yr) half spreading rate and so the age of lavas sampled off-axis are well constrained assuming an on-axis origin. Zou et al. [2] tested this assumption by comparing U-series disequilibria in lavas sampled away from the ridge with that observed at the present day axis. They found larger U-Th disequilibria than predicted by decay of the ridge signal in a number of lavas and interpreted this to indicate that these lavas were erupted off-axis. These are important results and so we have analysed U-Th disequilibria in lavas from three traverses along the East Pacific Rise (9°3 N, 10°4N and 11°5N) extending to a maximum distance of 31 km East and West from the ridge axis (corresponding to an age ≤550 ka), including reanalysing the critical samples analysed by Zou et al. [2]. Our analyses of four samples analysed by Zou et al. [2] confirm greater U-Th disequilibria than predicted by decay of the axis signal. However, we could not reproduce the U-excesses in two lavas. Relative to a calculated decay curve assuming an initial (230Th/238U) of 1.23 at the ridge, all of our analyses from 9°3N have greater Th excesses (2-6%) than predicted. Similarly, the lavas from the traverse at 10°4N have (230Th/238U) ratios up to 1.067 (6% Th excess) some 13 km from the ridge axis which is comparable to ratios of 1.042 (4% Th excess) 23 km distant from the ridge at 9°3N. Thus, lavas in a zone 13-30 km off- axis to both sides of the axis have greater Th excess than predicted by decay from the ridge. In contrast, we have found U excesses in lavas from the traverse at 11°5N. 226Ra data will be presented for those samples with disequilibria in excess of that predicted by decay of the axis signature. If the observed Th excess result from melting at the ridge axis, anomalously high initial (230Th/238U) ratios are required at the ridge (≥2). This would require unrealistically small degrees of partial melting and very slow melting and inferred upwelling rates. Dynamic melting models show that even <2% partial melting at an upwelling rate of 0.5cm/a produces initial (230Th/238U) ratios that are too low (~1.6) to explain the off-axis disequilibria. Moreover, the inferred upwelling rate is too low to be consistent with a half spreading rate of 5.5 cm/a. Small degree melts may be consistent with the off-axis eruption and be largely independent from the melting regime beneath the ridge axis. The frequent occurrence of off-axis 230Th excess at two different traverses at similar distance from the axis (5-20 km) may provide evidence that bending stresses open up tensile cracks to provide ascent paths for melts [1]. [1]R.A. Sohn, K.W.W. Sims, Bending as a mechanism for triggering off-axis volcanism on the East Pacific Rise, Geology (Boulder) 33(2005) 93-96. [2]H. Zou, A. Zindler, Y. Niu, Constraints on melt movement beneath the East Pacific Rise from 230Th-238U disequilibrium, Science 295(2002) 107-110.