T41D-01
Thermal structure of ultraslow and transitional ridge axe
Geologists observe a transition at full spreading rates of 10-15 mm/yr between slow ridge axes with a normal thickness of oceanic crust (ca. 6 km) and ultraslow ridge axes with a thinner crust and outcropping mantle peridotites. I examine this transition with 2-D kinematic thermal models. The fate of the basaltic magma has a major effect on the temperature field. That is, the magma may erupt or intrude at shallow crustal depths or some of the magma may refreeze at mantle depths. In the latter case, the latent heat of the upwelling basaltic magma buffers the temperature of the upwelling mantle material. Two modes of crustal formation thus exist at the transition zone spreading rates. (1) Mantle and basalt upwell along a narrow axial zone. Only small volumes of basalt freeze below 6 km depth. Most of the remaining basalt freezes in the deep crust. That is, the crust forms as a tectonic mixture of intrusive rocks from the bottom up (Dijkstra and Cawood, 2004). The magmatic rocks unpin the ridge axis so that the mechanical boundary condition on the underlying mantle is the crust diverging from the axis at the nominal spreading rate. A well-defined moho separates the crust and the mantle. (2) Mantle and crust upwell over a broad zone. Basalt freezes at depth before it reaches normal crustal depths. The frozen tectonic mixture of gabbro and peridotite forms a lid above the upwelling. The moho is gradual. The lid extends laterally at a uniform strain rate so that the nominal spreading rate is reached well away from the axis. Both modes probably represent metastable situations in the transition region of spreading rates at actual ridge axes. Petrological models of ultraslow and transitional ridge axes need to distinguish melting that does not occur at depth because the upweilling material is too cold from freezing of magma near the base of a lid. This task is not straightforward as magma from deep sources arrives with superheat and it more likely to breach the lid and erupt at the surface where they can be sampled easily. Conversely, cooler melts from shallow sources freeze to form gabbro at depth. These effects may give the illusion that only deep sources actually melt at transitional ridge axes.
T41D-02
Life Cycle of Oceanic Core Complexes
Oceanic core complexes are the uplifted footwalls of very-large-offset low-angle normal faults that exhume lower crust and mantle rocks onto the seafloor at slow-spreading ridges. Numerical modelling has suggested that they form during periods of critically reduced magma supply, but this is yet to be confirmed by observation. Little is known about the mechanisms of their initiation and inactivation, nor why only certain normal fault systems develop into core complexes. We present results from a near-bottom sidescan sonar/bathymetric profiler survey and sampling study of the Mid-Atlantic Ridge near 13°N that allow us to infer the life cycle of oceanic core complexes and demonstrate the critical controls on their development and evolution. We show that core complex detachment faults initiate as high-angle (65°±10°) normal faults no different from surrounding valley-wall faults and, like them, rapidly flatten to dips of ~30° in response to flexural unloading. However, on some faults, displacement continues rather than jumping inward to a new normal fault, resulting in locally enhanced uplift of the footwall and further flattening of the fault to the horizontal or beyond. Active detachment faulting and core complex formation occurs - and only occurs - where active volcanism is locally absent from the axial valley, while detachment faults are terminated by renewed magmatism as neovolcanic ridges propagate laterally across them. Our observations strongly suggest that core complex development is primarily controlled by local waxing and waning of the magma supply near a critical threshold, supporting recent numerical models.
T41D-03 INVITED
The Depth of Detachment Faulting at Mid-Ocean Ridges : Evidence From Zircon Geo- and Thermochronometry
Pb/U and (U-Th)/He zircon ages determined from evolved samples of gabbroic crust exposed in the footwalls of large-offset, low-angle normal faults near the Atlantis and Fifteen-Twenty Transforms on the Mid-Atlantic Ridge (MAR; ODP Holes 1275D and 1270D, IODP Hole U1309D), provide new constraints on the depth of detachment faulting at mid-ocean ridges. Ti-in-zircon crystallization temperatures, taken with the closure temperature of the (U-Th)/He system in zircon bracket the acquisition temperature of magnetic remanence; collectively these three chronometers define a cooling history for footwall gabbro sections over the temperature range of ~900°-220° C. Time-averaged cooling rates over 900°- 220° C from all holes investigated range from 1025(+645, -330)° C/m.y. to 2110(+1600, -720)° C/m.y. Assuming the gabbroic footwall was denuded along a single, continuous fault system, the time interval defined by the difference in Pb/U and (U-Th)/He ages for zircon from rocks beneath the fault can be used to estimate the distance between the 900° and 200° C isotherms along the fault system, and therefore the length-scale of the fault system while it was active, if the fault slip rate is known. As these large-offset faults serve as the plate boundary, the fault slip rate is equivalent to the plate-spreading rate during formation of the footwall. During formation of the Atlantis Massif core complex (30° N, MAR), accretion was asymmetric, with spreading partitioned on the North American plate at a rate approaching the full spreading rate of 24 mm/yr. This rate, along with a cooling time interval of 0.42±0.09 Ma implies that a single, continuous fault system would have had a length of 10±2.3 km between the 900° and 200° C isotherms while active. Lengths of fault systems determined at ODP Holes 1275D and 1270D are 9.5±3.2 km and 5.0±2.9 km, respectively, assuming a fault slip rate equivalent to 65% of the full plate spreading rate (consistent with asymmetric accretion rates observed at other oceanic core complexes in the Atlantic and Indian Oceans) of 16 mm/yr. Previously published Pb/U and (U-Th)/He zircon ages from Atlantis Bank (SWIR) indicate a fault length of 7.7±3.8 km. Estimated length-scales for the active portion of several oceanic detachment faults therefore range from 5-10 km (between the 900° and 200° C isotherms). Using these calculated fault lengths, an estimated depth to the 200° C isotherm (1-2 km), and an assumed fault initiation dip of 50° (based on paleomagnetic data) we estimate that these faults were active to depths of up to 8 km (~8 km for Atlantis Massif and ODP site 1275, 5.5 km at ODP site 1270, and ~7 km for Atlantis Bank at ODP Hole 735B). These significant depths for active faulting in slow spreading environments are consistent with microseismicity studies, and the existence of high temperature fault rocks at many oceanic detachment faults.
T41D-04
Asymmetric Spreading, and the Construction of Oceanic Crust at the Kane Oceanic Core Complex
Detachment faulting at the Kane Oceanic Core Complex (OCC) on the Mid-Atlantic Ridge (23° N) has exposed a tectonic window through oceanic crust. Here we present fourteen Pb/U zircon SHRIMP ages of evolved lower crustal gabbro collected by ROV and dredging during R/V Knorr Cruise 180-2 from the Babel, Cain and Abel, and Adam domes of the Kane Oceanic Core complex (up to 46 km off axis), and from ODP Hole 923A, near the present day ridge axis. These data allow us to constrain the construction history of oceanic crust at Kane. Weighted average 206Pb/238U ages range from 2.45 +/-0.06 to 3.70+/-0.16 Ma for samples from the Kane OCC, and 0.86+/-0.14 Ma for Hole 923A. Age-distance plots yield a North American plate-spreading rate of 15.3+/-2.0 cm/yr during formation of the Kane OCC, with a time-averaged rate of 14.3+/-0.95 cm/yr determined from ODP 923A near the ridge axis to the western margin of the OCC, consistent with rates determined from magnetic anomaly data (Williams, 2007). Both the zircon and magnetic data show that the Kane OCC formed during a period of asymmetric spreading with over 60% of the total plate-motion accommodated by detachment faulting, requiring associated ridge migration. The average 206Pb/238U age is consistently ~~250,000 yr older than the magnetic age, implying acquisition of magnetic remanence roughly 4 km off-axis. Ten samples have ages within error of the calculated spreading rate. However, two samples are significantly older, and two significantly younger than those that define the spreading rate trend, implying variability in both depth and location of crustal accretion. One of the older samples (3.7 +/-0.16Ma) comes from the northern Babel Dome, adjacent to the Kane Transform Fault. This single age is 0.6 Ma older than samples from the Cain and Abel domes (15 km) to the south, and may suggest that the gabbros forming Babel Dome crystallized ~~5km deeper in cooler lithosphere adjacent the transform fault. Additional data is being acquired to test this hypothesis. The two anomalously young ages are about 0.5Ma younger than the expected age for their location. Both were collected from the Cain and Abel domes and are in close proximity to samples that give expected ages. We suggest that these young ages reflect the heterogeneous spatial distribution of magmatism within the axial valley. We envisage that the majority of the crust accreted beneath the detachment fault at a depth of ~~6-7 km, consistent with seismicity studies elsewhere, and with high-temperature fault rocks from the Kane OCC. The younger samples may represent later, more shallow magmatism (~~2km depth) that intruded earlier formed crust as it was denuded by the detachment fault. This interpretation is consistent with the observation that both gabbros and basaltic dikes cut recovered fault rocks. Taken together, all data suggest that any given piece of oceanic crust may record up to 0.5Ma of magmatic activity at the Kane OCC.
T41D-05
Evidence for Footwall Rotation in an Oceanic Core Complex From IODP Core Samples Reoriented Using Borehole Wall Imagery
Oceanic core complexes expose lower crustal and upper mantle rocks on the seafloor by tectonic unroofing in the footwalls of large-slip detachment faults. The common occurrence of these structures in slow and ultra- slow spreading oceanic crust suggests they accommodate a significant component of plate divergence. However, the sub-surface geometry of oceanic detachment faults remains unclear. Competing models involve either: (a) displacement on planar, low-angle faults with little tectonic rotation; or (b) progressive shallowing by rotation of initially steeply dipping faults as a result of flexural unloading (the "rolling-hinge" model). We resolve this debate using paleomagnetic remanences as a marker for tectonic rotation within a unique 1.4 km long footwall section of gabbroic rocks recovered by Integrated Ocean Drilling Program (IODP) sampling at Atlantis Massif oceanic core complex on the Mid Atlantic Ridge (MAR). For the first time we have independently reoriented initially azimuthally-unconstrained drill-core samples of lower crustal gabbros to a true geographic reference frame by correlating structures in individual core pieces with those identified from oriented imagery of the borehole wall. This allows reorientation of paleomagnetic data and subsequent tectonic interpretation without the need for a priori assumptions on the azimuth of the rotation axis. Results indicate a 34°±7° counterclockwise rotation of the footwall around a MAR-parallel horizontal axis trending 007°±9°. This provides unequivocal confirmation of the key prediction of flexural, rolling-hinge models for oceanic core complexes, whereby faults initiate at higher dips and rotate to their present day low angle geometries.
T41D-06
Seismotectonic analysis of active detachment faulting at the TAG segment of the Mid- Atlantic Ridge, 26°N
Detachment faulting appears to accommodate a high percentage of extension on slow-spreading ridges. Detachments exhume lower crustal and mantle rocks to the seafloor, and are capable of providing long-term, high-permeability pathways for high-temperature hydrothermal circulation. Recent results have elucidated the morphological characteristics of detachment faults, but the nature of extension on these fault systems is not well understood. Here we present a detailed analysis of the space-time behavior of 20,730 well-constrained microearthquakes observed during a 245-day period at the TAG segment of the Mid-Atlantic Ridge (26°N). We observe a high, continuous rate of seismicity throughout the deployment that lacks the mainshock-aftershock sequences characteristic of most active faults. Frequency-magnitude analysis yields a b-value of ~1.5 long the detachment, which is intermediate between standard values for tectonic (~0.8-1.2) and volcanic (~1.8-2.0) environments. Clustering analyses using waveform cross- correlation techniques to quantify similarity suggest that seismicity is characterized by repeated slip on small fault patches. The recurrence interval between earthquakes within each patch ranges from 1 to 3 days. The moment release rate for detachment fault earthquakes observed during our study is roughly 5×1021 dyn-cm/yr, which is approximately an order of magnitude lower than the rate observed during a previous 3-week microearthquake survey from the same area in 1985, indicating that seismic slip rates are highly variable on decadal time scales. Both of these moment rates are orders of magnitude less than the nominal rate expected based on a simple model of extension (~1024 dyn-cm/yr), indicating that either slip along the detachment is predominately aseismic, or extension rates during both observation periods were significantly less than the long-term geological average. The peculiar nature of seismicity on the detachment fault (high-rates of small events without mainshock-aftershock sequences) may result from the composition (e.g., altered mafic/ultra-mafic rocks) or the stress state (e.g., high pore pressure) of the fault surface.
T41D-07
Seismic Constraints on Upper Crustal Processes in the Lucky Strike Segment of the Mid- Atlantic Ridge: Interplay Between Magmatic, Tectonic and Hydrothermal Processes
The 2005 SISMOMAR survey studied the structure of the slow-spreading Lucky Strike segment of the Mid- Atlantic Ridge using seismic reflection and refraction measurements. The rift valley is 15 to 20 km wide and bounded by large faults (approximately 800 m total vertical offset). The segment center is dominated by the 7 km wide Lucky Strike volcano hosting a hydrothermal vent field. We present results from a segment scale seismic reflection study and seismic refraction studies at two different scales, covering the Lucky Strike segment and volcano. We stack the seismic waves turning in the layer 2A/2B transition zone to image spatial variations of layer 2A and use travel time tomography to constrain the crustal velocity structure. The combination of the three seismic studies allows us to infer the interplay between magmatic, tectonic and hydrothermal processes shaping the upper crust of the Lucky Strike segment. The three-dimensional velocity models show an axis-parallel low velocity anomaly, which is limited laterally by the median valley bounding faults. Seismic velocities increase by 0.5 km/s over a distance of 10 km from the center of the anomaly. This increase coincides with an abrupt off-axis thinning of seismic layer 2A by 150 ms to 200 ms or approximately 200 m. The thinning of seismic layer 2A and the increase in seismic velocities indicate a porosity decrease across the median valley bounding faults. This porosity decrease is linked to hydrothermal circulation and the aging of the crust. Within the median valley, the seismic velocities and layer 2A thicknesses are rather uniform. The lowest seismic velocities, approximately 1 km/s lower than in the surrounding rock, are observed inside the median valley underneath recent volcanic edifices like the Lucky Strike volcano. These low velocities can be explained by elevated porosities and a thicker extrusive layer. However, the regions of decreased velocities do not systematically coincide with regions of a thickened layer 2A. This allows for the possibility that the layer 2A/2B boundary in the Lucky Strike segment does not correspond to the extrusive/intrusive transition, but is defined by pore space collapse at depth caused by weight of the rocks above.
T41D-08
Detailed distribution and rapid degradation of small seamounts on the MAR axial volcanic ridge, 45°30'N
In May-June 2008, James Cook cruise 24 conducted a detailed geophysical survey and geological sampling of a single axial volcanic ridge on the Mid-Atlantic Ridge axis at 45°30'N using the TOBI deep-towed instrument and Isis ROV. TOBI sidescan and bathymetry define the AVR morphology in unprecedented detail, while TOBI and Isis magnetic field measurements help define the younger parts of the ridge (all within the 0.8 Ma Brunhes chron). Over 180 samples were recovered on 11 Isis dives for geochemical analysis and radiometric dating. Here we concentrate on a detailed 2 km-square survey of part of the AVR crest and flank. This area contains ~100 identifiable volcanic cones ranging from 250 m to ~20 m diameter, both conical and flat-topped. There is no very clear relationship between neighbouring cones, though they tend to follow broad trends both parallel and oblique to the AVR axis that may reflect deep-seated tectonic control. Inferred crustal magnetisation peaks follow similar trends. Magnetisation varies significantly within the AVR flank, does not fall monotonically away from the axis, and displays the highest values 1800 m off- axis, corresponding to an age of 164 ka if accretion was linear at the regional accretion rate of ll km/Ma. Many of the volcanic cones, including some on the AVR axis, have already undergone significant degradation by flank failure and mass wasting, leaving scars that cut deep into the cones and are extremely steep (often vertical). This evidence suggests that the volcanic landscape of the oceanic crust is susceptible to significant tectonic degradation virtually from the time it is formed.