T11G-01 08:00h
Lena Trough (Arctic Ocean): An oblique 'amagmatic' rift
Seafloor spreading is strongly controlled in many important ways by the geometry of the plate boundary. Lena Trough is a recent plate boundary in the Arctic Ocean with a highly oblique plate boundary configuration. It is also, significantly, the gateway for deep water circulation between the Arctic and North Atlantic oceans, and is the most recent and final event in the separation of the North American from the Eurasian continent. Models for the tectonic configuration of Lena Trough have until now differed only in the number and length of fracture zones and spreading segments thought to be present. Most authors admit however, that because of inadequate data coverage, the tectonics of Lena Trough have remained undefined. However, persistent ice cover has prevented any systematic mapping and sampling in most of Lena Trough until now. Here we report new mapping and sampling results from Lena Trough from the ARK XX-2 expedition of PFS Polarstern in summer 2004. Lena Trough is a deep fault-bounded basin with depths of 3800-4800m, and irregular, steep valley sides that are oblique to the spreading direction. Basement horst structures outcrop as sigmoidal ridges with steeply dipping sides project out of the valley floor, and are roughly parallel along flow lines to structures on either side. Ridge-orthogonal topography is simply absent (ie no segments trending parallel nor fracture zones perpendicular to Gakkel Ridge). Most faults trend approximately SSE-NNW, an obliquity with respect to Gakkel Ridge (SW-NE) of about $55\deg$. The basement ridges are composed nearly entirely of mantle peridotite, as are the valley walls. Only at the northern and southern extremities of Lena Trough do basalts appear at all. Peridotite compositions show increasing degrees of melt infiltration near the center of Lena Trough, suggesting magma generation and stagnation in the mantle that defines a first order segmentation at the scale of 100's of kilometers. Otherwise Lena Trough in unsegmented in any normal sense. These results show that Lena Trough belongs to the newly defined amagmatic class of mid-ocean ridge spreading centers, which via an oblique spreading mechanism exposes nearly exclusively ultramafic mantle rocks. At the same time, Lena Trough is a young ocean basin, having only about 9 million years of spreading history, the youngest ocean basin on Earth. Thus Lena Trough represents a transition form between a continental rift and an oceanic one. Without significant basaltic infilling of the nascent ocean basin, subsidence to very great depths (ca. 4500m) probably occurred rapidly after trans-tensional motion began on the plate boundary in the Miocene. The Lena Trough is also the only known modern analog of the Iberia Margin, the conjugate Newfoundland Margin, as well as the ophiolite complexes of the Western Alps.
T11G-02 08:15h
The Influence of Ridge Geometry at Ultraslow Spreading Rates
Ridges spreading at ultraslow rate less than 20 mm/yr have been identified as a unique class of ocean ridge as different from slow spreading as slow spreading are from fast 1. Ridge characteristics, such as the presence or absence of amagmatic accretionary segments, transform faults, axial valleys or axial rises, however, are not a simple function of spreading rate, and it is therefore difficult to define precisely ridge classes simply on this criterion. Ridge morphology, tectonics, and geochemistry are also largely a function of mantle thermal structure, upwelling rate, fertility, and ridge geometry. However, examination of ridge crustal structure with spreading rate clearly shows a sharp break, with seismic measurements of crustal thickness indicating highly variable, generally thin crust associated with spreading rates below 20 mm/yr. In contrast, crust formed at spreading rates greater than 20 mm/yr is generally thicker and less variable thickness, averaging between 6 and 7 km, without a clear relationship to spreading rate. The generally accepted explanation is the influence of conductive heat loss and the formation of a thick axial lithosphere due to slow mantle upwelling rates, thereby limiting melt production at ultraslow spreading rates 2. Comparatively, the influence of conductive heat loss at spreading rates greater than 20 mm/yr is likely negligible except near major large offset transforms. The latter effect is predicted by modeling to increase sharply with decreasing spreading rate below 20 mm/yr. Thus perturbations in ridge geometry that would otherwise have a negligible effect, can dramatically influence melt production and ridge tectonics at ultraslow spreading rates. Investigation of the SW Indian Ridge and along the Gakkel Ridge, for example, shows that where the effective spreading rate for mantle upwelling, which ridge obliquity, falls below ~12 mm/yr, long amagmatic accretionary ridge segments form and replace both magmatic accretionary ridge segments and transform faults. These amagmatic accretionary segments then link with magmatic accretionary segments to form stable plate boundaries on even the most oblique trending ridges. Magma geochemistry is strongly affected, with highly variable, incompatible and isotopically enriched MORB's and alkaline basalts appearing along ultraslow spreading ridges far from mantle hotspots, where the mantle upwelling rate is suppressed and the extent of mantle melting is low. The precise composition of such basalts varies considerably with location, indicating a heterogeneous upper mantle composition that varies both locally and on a global scale. References 1. Dick, H. J. B., Lin, J. & Schouten, H. An ultraslow spreading class of ocean ridge. Nature 426, 405-412 (2003). 2. Reid, I. & Jackson, H. R. Oceanic spreading rate and crustal thickness. Marine Geophysical Researches 5, 165-172 (1981).
T11G-03 INVITED 08:30h
Spreading Geometry and Melt Supply at the Ultraslow-Spreading Southwest Indian Ridge
The Southwest Indian Ridge (SWIR) stretches 7900 km and is among the world's slowest spreading ridges with a full rate of only 1.4-1.5cm/yr. Regional axial depths vary between 3100m near Marion Island, and 4800m east of the Melville transform. Basalt Na8.0 contents increase correlatively and are consistent with variations of the ridge's melt supply corresponding with a magmatic crustal thickness of a little more than 6 km near Marion, to about 3 km east of the Melville transform. This along-axis change in melt supply is not correlated with ridge obliquity as would be the case for a homogeneous mantle rising passively beneath the ridge (upwelling velocities equal to effective half spreading rates). This shows that melt supply along the SWIR is not a simple function of spreading geometry. Instead, mantle temperature and/or mantle chemistry appear to be the principal controls on melt supply at the regional scale. We further investigate this issue at the smaller scale of the low melt supply end-member, easternmost portion of the SWIR between 61\deg and 67\deg E. We use the most extensive set of off-axis bathymetry, gravity and magnetic data available to date for an ultraslow spreading environment. We find that spreading rate and ridge geometry have changed little over the past 26myrs, with an average rate of 14.5cm/yr, and two broad sub-regions in terms of ridge obliquity: 35\deg in a 400km-long western region, and 7\deg in a 250km-long eastern region. We find no evidence for a systematic change in melt supply between these two sub-regions. Instead, we find evidence in both sub-regions for transient melt focusing events, and for longer-lasting periods of tectonically-dominated spreading, with large-offset normal faults and lower melt supply. We use available on-axis basalt chemistry data to propose that such short-lived and apparently randomly distributed pulses in axial melt delivery could be related to short-scale heterogeneities in the composition of the sub-axial mantle.
T11G-04 INVITED 08:45h
The Transition from Initial Rifting to Ultra-Slow Seafloor Spreading within Endeavor Deep
Endeavor Deep is a NW-SE trending, 3 km-deep rift basin located along the divergent portion of the Nazca/Juan Fernandez plate boundary. The rift basin is the result of the propagation of the East Ridge toward the northwest with relative motion across the ridge defined by a rapidly rotating (5.5 degrees/myr) Euler Pole located ~100 km to the northwest. The close proximity of Endeavor Deep to this Euler Pole results in a rapidly varying velocity field along the length of the deep and represents a unique location to study the effect of varying divergence rates on initial crustal extension. Recently collected EM300 bathymetry, DSL120 sidescan, surface-towed magnetics and JASON II observations have documented 4 distinct stages of rifting along the 70 km length of Endeavor Deep. These stages include (from NW to SE): amagmatic rifting, distributed initial volcanism, centralized waxing volcanism, and crustal formation by ultra-slow seafloor spreading. Amagmatic extension, evolving to rifting, occurs at spreading rates less than 13 km/myr and is characterized by rapidly deepening rift depths from NW to SE with an overall increase in depth of about 2.5 km. Extension is accommodated over a width of about 10-15 km and some flexural uplift of the defining scarps is observed. Distributed initial volcanism occurs at spreading rates from 13-14 km/myr and is characterized by coalesced volcanic constructs (100-200 m-high, 1-2 km-wide) across the width of the rift floor. The depth of the rift basin becomes fairly constant, but the cross-sectional area of the deep continues to increase. Centralized waxing volcanism occurs at spreading rates from 14-17 km/myr and is characterized by pillow ridges and tectonic lineations along the central portion of the rift floor which are oriented parallel to the long axis of the rift basin (orthogonal to the direction of extension). The floor of the rift basin begins to shoal and the cross-sectional area of the deep decreases initially and then maintains a constant area. At spreading rates greater than ~17 km/myr, true seafloor spreading begins and is characterized by a well-formed central magnetic anomaly and zero-age depths comparable to mid-ocean ridges spreading at similar rates.
T11G-05 09:00h
A Small Ocean Rift Leads to a New View of the Galapagos Microplate: the Incipient Rift at 2 Degrees North, East of the East Pacific Rise
The Galapagos microplate (GMP) shares a complex plate boundary configuration with the surrounding Cocos, Nazca and Pacific plates. While the configuration of the microplate's southern boundary with the Nazca plate is relatively well understood, the nature of its northern boundary with the Cocos plate has remained more elusive. Work by Lonsdale and co-workers (1988; 1992) identified an "incipient" spreading center east of and orthogonal to the East Pacific Rise (EPR) at 2deg40minN, forming a portion of the northern boundary of the GMP. They described this spreading center as a slowly diverging, westward propagating rift, verging toward the EPR. In 2002, we mapped and sampled a broad region centered on the IR, extending ~140 km east of the EPR. Various geological, geophysical and rock sampling tools were used, including Seabeam2000 bathymetry and side-scan (amplitude); towed magnetometer; bottom photography (14 camera tows using the WHOI camera system); water column hydrothermal surveying and rock sampling. Our recent analysis of the bathymetric, side-scan and magnetic data suggests that the IR forms a lozenge-shaped feature, which tapers both westward toward the EPR and eastward, away from the EPR. Within the eastern half of the lozenge, there is a sinuous trough that trends ~100° to the southeast, along which reflection amplitudes indicate high-reflectivity consistent with relatively sparse sediment cover. The eastern end of this trough clearly cuts north-south-oriented abyssal hills and then dies out in the vicinity of 101deg29minW. The photographic results show sparsely sedimented lavas, often with delicate ornamentation and basaltic glass, with local in-filling of sediment between pillows. In a number of photographs, particularly along the eastern portion of the IR, lavas appear to emanate from local fissures with southeast-northwest orientations. Taken together, these data suggest that the IR is magmatically active along its length and has propagated eastward to the current location of its eastern rift tip. We suggest that continued spreading leads to pivoting of lithosphere about this rift tip, with important implications for the integrity and kinematics of the GMP to its south (see abstract by Schouten et al).
T11G-06 09:15h
Counter-Rotating Microplates at the Galapagos Triple Junction, Eastern Equatorial Pacific Ocean
We recently mapped and sampled a broad region of ocean floor centered on the Incipient Rift (IR), an east-west-trending spreading center that extends eastward from the East Pacific Rise (EPR) at 2deg40minN. The IR forms a portion of the boundary of the complex Cocos-Nazca-Pacific triple junction (TJ). Based on bathymetric, side-scan (amplitude), magnetic, photographic and sampling data, we conclude that the IR is magmatically active along its length, has rifted eastward, and as it spreads, pivots about its eastern terminus (see abstract by Klein et al. this session). If the IR opens about a pivot at its eastern end, it follows that lithosphere immediately to the south of the IR rotates in a counter-clockwise direction about this pivot. This counter-clockwise rotation contrasts with the known clockwise rotation of the Galapagos microplate (GMP) to the south. It follows, then, that there must be two separate, adjacent microplates in this region: the northern microplate, herein called the northern Galapagos microplate (NGMP), and the GMP. We estimate the kinematics of the GMP and NGMP by drawing upon the concepts of edge-driven microplate mechanisms. In this approach, the rotation of a microplate is driven by a shear couple between a pair of bounding plates moving in opposite directions. The two points of coupling between the microplate and bounding plates are represented by two instantaneous relative rotation axes (IRRAs). In edge-driven microplate systems like Easter and Juan Fernandez, these axes commonly lie ahead of the tips of the microplate bounding rifts. We identify 3 such IRRAs. A flat Earth approximation yields instantaneous rotation rates of 13 deg/myr for NGMP and 21 deg/myr for the GMP relative to translating major plates. Since NGMP and GMP rotate in an opposite sense, the NGMP-GMP rate is the sum, or, 34 deg/myr. The IRRAs and their respective rotation rates predict reasonable velocities at the two ridge-ridge-ridge triple junctions, e.g., Cocos-NGMP velocity at the 2deg40minN TJ of 15 km/my (008 deg) closely matches previous estimates. The GMP-Nazca velocity at the 1deg10minN TJ of 55 km/myr (323 deg) provides a better match to the ~050deg direction of the GMP-Nazca boundary at the 1deg10minN TJ than the previous estimate of 40 km/myr (337 deg). If our new model is correct we speculate that it may be applicable to other triple junctions. In the specific case of the Cocos-Nazca-Pacific triple junction, we think that the dual microplate system acts to control the location and configuration of the Hess Deep Rift and the stability of the Cocos-Nazca-Pacific triple junction. Further work is needed to understand the evolution of this triple junction, and the nature of triple junctions and their stability in general.
T11G-07 09:30h
A plate kinematic explanation for the magmatic segmentation of mid-ocean ridges
Along fast and intermediate spreading mid-ocean ridges (MOR), a direct relationship is found between magmatic segmentation inferred from ridge morphology and the geometry and migration velocity of the plate boundary. All MOR are migrating in the "fixed" hotspot reference frame. Analysis of bathymetric undulations along 9,500 km of ridge crest reveal that segments which are offset in the direction of ridge migration (leading) across transform and non-transform ($>$5km) discontinuities overwhelmingly correspond with shallow, magmatically robust segments. Differences in ridge elevation are similar (10-500 m) in spite of order of magnitude differences in discontinuity length, ruling out cold edge effects to explain observed changes in ridge morphology. Furthermore, morphological contrasts between adjacent segments diminish for the longest transform faults ($> $~150 km), with offset lengths comparable to the half width of the mantle upwelling zone predicted from passive mantle flow models. We attribute these relationships to asymmetric mantle upwelling and melt production beneath migrating ridges, with entrainment of melts generated in the upwelling zone of adjacent segments across discontinuities. Where transform offsets are less than the width of the mantle melt generation zone, this model predicts leading segments can tap a greater portion of melts generated beneath adjacent trailing segments giving rise to differences in magma delivery with plate geometry. In addition to bathymetric changes, geochemical differences in erupted lavas are commonly observed across discontinuities, which indicate different mantle source compositions often attributed to small-scale mantle heterogeneities. Our model provides a mechanism whereby early melting heterogeneities and melts originating deeper in the melting column from the upwelling zone of an adjacent advancing plate could be preferentially entrained beneath leading ridge segments. Along the NEPR 8-18degN, the most enriched and diverse lavas as indicated by incompatible trace element ratios, as well as those derived from deepest melting are found along leading segments, consistent with the notion that geochemical variability along this ridge may reflect the influence of plate kinematics on mantle melt production.
T11G-08 09:45h
Ridge migration, asthenospheric flow and the origin of magmatic segmentation in the global mid-ocean ridge system
Global observations of mid-ocean ridge (MOR) bathymetry demonstrate an asymmetry in axial depth across ridge offsets that is correlated with the direction of ridge migration. Motivated by these observations, we have developed two-dimensional numerical models of asthenospheric flow and melting beneath a migrating MOR. The modification of the flow pattern produced by ridge migration leads to an asymmetry in melt production rates on either side of the ridge. By coupling a simple parametric model of three dimensional melt focusing to our simulations, we generate predictions of axial depth differences across offsets in the MOR. These predictions are quantitatively consistent with the observed asymmetry.
http://www.ldeo.columbia.edu/~katz/publications/Katz_et_al_GRL04.pdf