T22E-01 INVITED
Black Smoker Vents and Oceanic Detachment Faults
It is generally assumed that the seawater-derived fluids that feed black smoker vent fields on the seafloor are discharged vertically from depths of around 1 to 3 km. Yet several Atlantic vent fields are located several km away from the neovolcanic zone, and the TAG field is underlain by a zone of seismicity dipping toward the ridge crest. Fossil detachment faults at both 15o 45' N and 30o N are underlain by gabbro bodies surrounded by serpentinized peridotite and are intruded by syntectonic basaltic sills. Both faults show evidence for high fluid flux at black smoker temperatures in the form of metasomatic talc-tremolite-chlorite assemblages replacing ultramafic rocks, and intense alteration of oxygen and strontium isotopes. Fluids in equilibrium with this assemblage would be similar to ultramafic-hosted black smoker fluids. We present a new model in which hydrothermal circulation around detachment faults evolves from hangingwall basalt-hosted (TAG-type) to footwall ultramafic-hosted (Rainbow-type) to low temperature ultramafic-hosted (Lost City type) as displacement on the detachment increases. Key features of our model are the intrusion of gabbro bodies immediately below the detachment to provide a heat source for circulation, and focusing of fluid flow into the detachment fault to allow venting away from the neovolcanic axis.
T22E-02 INVITED
Do Fault-Controlled Hydrothermal Systems Control the Thermal Evolution of the Lower Crust at Fast-Spreading Ridges?
Some geophysical data (e.g. seismic, compliance, bathymetry) suggest that the lower oceanic crust loses heat far more rapidly in the near-axis region that is predicted by conductive cooling models. Hydrothermal circulation through the lower oceanic crust provides one potential mechanism to cool the lower crust in the near axis region. However, there is scant evidence in samples of the lower oceanic crust for extensive pervasive fluid flow occurring over a wide temperature range. This suggests that any large-volume fluid flow must be channelised, most likely in fault-zones. Observations in the Oman ophiolite demonstrate that faults in the plutonic complex can form at high-temperatures. The Sr-isotopic composition of rocks within these fault zones are as elevated as those of the sheeted dike complex (~0.7045) requiring large fluid fluxes (Coogan et al., 2006). Thermal modeling suggests that fluid flow in faults in the lower crust could lead to much more heterogeneous cooling than if the permeability is constant. If fluid flow in the lower crust is focused in faults, with heat conducted through the wall-rock to the fluid in a fault, then the separation of faults and how far from the axis they form are critical parameters in controlling the impact of fluid flow on the axial thermal structure. The best estimate of the separation of lower crustal faults in the Oman ophiolite is approximately 1km but in the Troodos ophiolite, where this is better constrained by seismic data, it is approximately 7 km (Mackenzie et al., 2006). We currently have little data to support the existence of deeply penetrating faults in modern lower oceanic crust close to the ridge axis. Future off-axis micro-seismic surveys, and geological observations of lower crust exposed at tectonic windows, provide perhaps our best opportunities to better understand if such faults exist and, if so, what role they play in controlling the hydrology of the lower oceanic crust. References: Coogan et al, 2006, Am. J. Sci. v306, p389; Mackenzie et al, 2006, Geophys. J. Int. v167, p1385.
T22E-03
Complex Tectonics of an Oceanic Transform Fault: Segmentation of the Blanco Transform Fault Zone from Earthquake Analysis
The Blanco Transform Fault Zone (BTFZ), northeast Pacific Ocean, forms the ~350 km-long Pacific-Juan de Fuca plate boundary between the Gorda and Juan de Fuca ridges. Near-by broadband seismic networks provide the unique framework for a detailed, long-term seismotectonic study of an entire oceanic transform fault (OTF) system. We use regional waveforms to determine 125 earthquake source parameters; combined with 28 Harvard moment-tensors, they represent the largest waveform derived OTF source parameter dataset. Joint epicenter determination removes the northeasterly routine location bias. Projecting seismicity onto the BTFZ, we map along-fault seismic slip rate variations. Earthquake source parameters and morphology indicate several transform segments separated by extensional step-overs. The eastern segment from Gorda Ridge to Gorda Depression is a pull-apart basin. The longest transform (~150 km) following Blanco Ridge from the Gorda to Cascadia depression is seismically very active, fully coupled, has a wider seismic zone (~9 km) than other transform segments and accommodates the largest (Mw 6.4-6.5) BTFZ earthquakes. Interpretation of Cascadia Depression as spreading ridge is supported by plate motion parallel normal faulting T-axes. Spreading is currently tectonic; 9 km deep earthquakes indicate a deep source for intermittent intrusives and rapid post- emplacement cooling. A short transform connects to the pull-apart Surveyor Depression. Widely spread seismicity along the western BTFZ reflects complex morphology indicating ongoing plate boundary reorganization along short, narrow-width sub-parallel faults that limit earthquake size to Mw ≤ 6.2. Seismic coupling is low in extensional (≤15%) compared to transform areas (35-100%) implying different mechanical properties. Centroid depth variations are consistent with seismic slip cut-off near 600°C.
T22E-04 INVITED
The effect of fault segmentation on the dynamics of fast-slipping oceanic transform faults
The majority of oceanic transform faults (OTFs) systems along the fast-spreading equatorial East Pacific Rise are segmented into two or more strike-slip fault strands as a result of plate motion reorganization. Fresh basaltic lava sampled from these locations (e.g., the Siqueiros and Garrett OTFs) indicate that active crustal accretion is occurring within these transform systems. New Residual Mantle Bouguer gravity Anomalies (RMBA) calculated along fast-slipping OTFs are found to be more negative than the RMBA values along adjacent ridge segments. One possible explanation for these observations is enhanced magmatic upwelling and crustal accretion at intra- transform spreading centers (ITSC) and within the transform valley of the fast-slipping OTFs. In this study, we examine two end-member 3-D thermal models (constant viscosity rheology versus visco-plastic rheology) to explore mantle flow and melt extraction beneath segmented transform faults. Melt fraction is calculated using the parameterized fractional melting model of Kinzler and Grove (JGR, 1992a, 1992b, and 1993), and the fractional crystallization models of Yang et al. (Cont. Min. Pet., 1996). We evaluate the sensitivity of these models to various parameters including transform fault geometry, mantle potential temperature, and initial mantle composition. Preliminary results for a 100 km-long transform fault, slipping at 100 mm/yr, segmented by a single 10 km-long ITSC indicate that incorporating a visco-plastic rheology results in an approximately 35 percent decrease in the brittle fault area (< 600°C isotherm) compared to a constant viscosity model. Assuming upward melt migration along the base of the lithosphere, we find that crustal production is enhanced at ITSC by 11.5 km compared to the adjacent ridge segments. However, crustal thickness variations are sensitive to transform fault geometry and assumptions made about the pooled melt region. For example, if melt migration is not permitted across the transform fault, the calculated crustal thickness of the ITSC is approximately 5 km less than the crustal thicknesses predicted at the adjacent ridge segments. We apply these modeling techniques to Siqueiros transform fault and make direct comparisons between predicted crustal thickness and melt composition with gravity-derived crustal thickness variations and major element analyses of glass samples recovered from the Siqueiros transform fault domain.
T22E-05
Diving surveys of small seamounts on the outer rise of the Japan Trench, and replacement of benchmarks for seafloor geodesy
Geodetic experiments and diving surveys of the outer rise of the Japan Trench off Northeastern Japan were carried out during the Kairei KR07-07 cruise in June 2007 using the ROV KAIKO 7000 II. The cruise aimed at three objectives. The first was diving surveys of small seamounts on the outer rise. A new type of seafloor volcanism called gpetit spoth was recently found near the Japan Trench (Hirano et al., 2006). Since each of diving surveys to three seamounts found fresh basalts, the survey area can be a potential locality for another petit spot volcanic field (Abe et al., this meeting). The second objective was diving surveys of the seafloor in the source region of the 2005 M7.1 outer rise earthquake. It was the largest outer rise earthquake recorded off the Japan Trench since the 1933 M8.4 Sanriku- oki tsunami earthquake. Although we could not find any indication of deformation on the seafloor during the survey, we are interested in the observation that the aftershock distribution (Hino et al., this meeting) was overlapped with the area of the small seamounts. The seamounts may have resulted from young volcanism like gpetit spoth caused by oceanic plate flexure (Hirano et al., 2006). Bathymetric maps show graben structures near the small seamounts sub-parallel to the trench axis suggesting normal faults in the outer rise region. We can suppose that the normal faults can be another mechanism for the young volcanism near the Japan Trench, or that the gpetit spoth volcanism may have induced the large intra-plate normal fault earthquake. Anyway the seamounts can be related to intra-plate earthquakes. The third objective was renewal of acoustic seafloor benchmarks deployed on the outer rise. Three precision acoustic transponders (PXPs) were deployed in 2002 to observe the motion of the pacific plate near the subduction plate boundary, and somehow exhausted the batteries after a few short observations. Each of two PXPs was replaced with a new one after cm-order observation of the relative positions. A new PXP was deployed from the sea surface near the third PXP. We plan to combine the geodetic observations with the old PXP net with future works based on the data of precise relative positioning.
T22E-06
Seismic Reflection Images of Deep Lithospheric Faults and Thin Crust at the Actively Deforming Indo-Australian Plate Boundary in the Indian Ocean
Recently, we acquired deep seismic reflection data using a state-of-the-art technology of Schlumberger having a powerful source (10,000 cubic inch) and a 12 km long streamer along a 250 km long trench parallel line offshore Sumatra in the Indian Ocean deformation zone that provides seismic reflection image down to 40 km depth over the old oceanic lithosphere formed at Wharton spreading centre about 55-57 Ma ago. We observe deep penetrating faults that go down to 37 km depth (~24 km in the oceanic mantle), providing the first direct evidence for full lithospheric-scale deformation in an intra-plate oceanic domain. These faults dip NE and have dips between 25 and 40 degrees. The majority of faults are present in the mantle and are spaced at about 5 km, and do not seem cut through the Moho. We have also imaged active strike-slip fault zones that seem to be associated with the re-activation of ancient fracture zones, which is consistent with previous seismological and seafloor observations. The geometries of the deep penetrating faults neither seem to correspond to faulting associated with the plate bending at the subduction front nor with the re-activation of fracture zone that initiated about 7.5 Ma ago, and therefore, we suggest that these deep mantle faults were formed due to compressive stress at the beginning of the hard collision between India and Eurasia, soon after the cessation of seafloor spreading in the Wharton basin. We also find that the crust generated at the fast Wharton spreading centre 55-57 Ma ago is only 3.5-4.5 km thick, the thinnest crust ever observed in a fast spreading environment. We suggest that this extremely thin crust is due to 40-50°C lower than normal mantle temperature in this part of the Indian Ocean during its formation.
T22E-07
Tecto-Magmatic Cycles, Faulting, and Morphology of Mid-Ocean Ridges
Numerical models of brittle-ductile deformation of the crust and mantle are used to study the dynamical causes of the range of axial morphology observed at mid-ocean ridges including faulted rift valleys, relatively unfaulted axial highs, and transitional morphologies with highly faulted topography but no distinct axial valley or high. Tecto- magmatic cycles are simulated by varying the rheology and stress in a 2-km-wide magma accretion zone centered between two diverging lithospheric plates. During tectonic episodes, the whole region near the ridge axis deforms elasto-plastically (i.e., Mohr-Coulomb yielding) so that faults form and grow, and the ridge axis deepens below the level of isostatic equilibrium. During magmatic times, the center of the magmatic zone is allowed to open freely (i.e., stress is lithostatic) to simulate dike injections. Also, the magmatic zone deforms viscously with viscosity not exceeding an imposed maximum, ηM. This viscous deformation causes the axis to rise toward the level of isostatic equilibrium thus simulating topographic accretion by magmatism with a timescale τ. Varying the fraction, FM, of the magmatic time period relative to the duration of the whole tecto- magmatic (magmatic+amagmatic) cycle primarily affects fault characteristics, with faults spacing and heave decreasing with increasing FM. The major factor controlling whether an axial valley or axial high forms is the timescale τ of axial topographic accretion by magmatism. The full range of morphologies from deep (>1 km) axial valleys, to faulted flat topography, to faulted highs, to smooth highs is predicted, simply by varying τ from much less than, to ~5 times greater than the length of magmatic periods. FM and lithospheric thickness effect τ but the dominant effect is the limiting viscosity ηM. Thus, in contrast to previous studies, which have argued that lithospheric thickness controls faulting and axial morphology, we show that it is the timescale of topographic growth by magmatism that is dominant factor. It is differences in the mechanics of magma storage and transport in the lithosphere that cause the wide range of morphologies at mid-ocean ridge as well as the abyssal hill fabric on the seafloor created. http://www.soest.hawaii.edu/GG/FACULTY/ITO/
T22E-08
Do Lava Flows Drive Caldera Collapse at Spreading Centers?
Many fast and intermediate rate spreading centers with axial highs show sharp topographic relief on or near the spreading axis indicating offset of lithospheric blocks. These include the ~100 m deep "axial summit troughs" seen at many fast spreading segments and the ~300 m offsets of "split volcanoes" and "rifted axial highs" seen at some intermediate spreading ridges. These features were first thought to result from normal fault offset during amagmatic periods. However, seismic imaging of axial magma chambers (AMCs) under such segments indicates that magma is present. This magma should allow dikes to open and relieve stresses before significant normal fault offset. These topographic features are also hard to explain by the kind of magmatic collapse that occurs when a dike breaks out of the side of a steep-sided volcano. Topographic relief at these ridges is insufficient to drive the magma pressure drop needed to cause such collapse. The correlation of topographic relief with magma chamber depth may indicate that collapse is driven by loads related to magmatic accretion. Magma intrudes to a depth that is controlled by the magma density then solidification causes the density of the intruded dike to increase. Local isostatic compensation of this density increase would lead to a topographic drop that is proportional to the axial lithospheric thickness. Effectively local isostasy could result when collapse events occur. Magma accretion/solidification loads would bend down the lithosphere above the AMC, but collapse-forming breaks could only occur when magma allows the bent region to fail as a tension crack. Load-related bending should put the base of the axial lithosphere in tension at its center, thereby triggering dikes fed by the underlying AMC. The same bending would put the surface in tension at the sides of the axial lithosphere above the AMC. When loading has reached a sufficient level the areas of surface tension could start to crack, but those cracks would not propagate through the lithosphere unless lavas pour into them. Collapse could occur when these fluid-filled cracks open and so lose all shear strength. Simple numerical models including flexure of the axis and a parameterized magma sources are used to track the development of the observed range of collapse related relief.