T32A-01
Formation of the Rio Grande Rift and Upper Mantle Seismic Wave Velocity Anomaly
The Rio Grande rift extends from central Colorado through New Mexico where it merges with the southern Basin and Range province, to Chihuahua in Mexico. Rifting started in the Oligocene-early Miocene, and strain rates were probably very modest since the Late Miocene. The rift is currently active (strain rates ~1-2 mm/yr) but rifting is not accompanied by seismic activity. The Rio Grande rift was formed close to a step in lithosphere thickness between the Great Plains and Colorado Plateau. The La RISTRA passive seismic experiment reveals a shallow low seismic wave velocity perturbation under the Rio Grande rift, extending to a depth of about 100-180 km. Prior explanations for this observed upper mantle heterogeneity include a broad region of lithosphere thinning and asthenospheric upwelling, compositional variations, presence of partial melt in the upper mantle, thermal anomalies, or a combination of these factors. There is no evidence for deep mantle upwelling beneath the rift and therefore active rifting of the Rio Grande rift is not supported. Here we present a model for formation of this continental rift system and the low seismic wave velocity anomaly in the upper mantle based on results of a lithosphere-upper mantle finite element model. The numerical results suggest that formation of the Rio Grande rift took place west of the Great Plains at the location where the lithosphere was thermally weakened by small scale convection. The rift zone was formed at the time that the lithosphere of the south-western US was no longer under compression. Localized extension at the rift zone further thinned the mantle part of the lithosphere and asthenosphere material welled up under the rift, replacing the colder lithosphere. From the temperature structure predicted by the model we calculated the expected seismic wave velocity perturbation in the upper mantle. Comparison with tomographic results shows a good agreement, and we conclude that the low seismic wave velocity anomaly is the result of upward flow of asthenospheric material below the rift zone.
T32A-02
Comparison of Pure-Shear and Upwelling-Divergent Flow Models of Breakup Continental Lithosphere Thinning for the N. Iberian, N. Newfoundland and N. Angolan Rifted Margins
Many rifted continental margins show a region > 200 km wide of highly thinned continental crust with thickness 10-15 km. Two different models of continental lithosphere thinning leading to continental breakup prior to sea- floor spreading initiation have been investigated and compared. One model assumes that continental lithosphere thinning is achieved by depth-uniform (pure-shear) lithosphere stretching. The second model assumes that continental lithosphere thinning is achieved by a combination of upwelling divergent flow within continental lithosphere and asthenosphere and pure-shear deformation. Both models have been applied to the N. Iberia - N. Newfoundland conjugate rifted margins and the N. Angolan rifted margin, and used to predict margin crustal thickness and lithosphere temperature. The isostatic response to crustal thinning, lithosphere geotherm perturbation and loading by observed sediment thickness have been determined. For both models the final breakup rupture of thinned continental lithosphere of thickness ~10-15 km is assumed to occur by normal faulting evolving into extensional detachments. Model predictions have been constrained using observed bathymetry and free air gravity anomaly. Both the pure-shear and upwelling divergent flow models of breakup lithosphere thinning are able to satisfactorily predict observed bathymetry and gravity anomaly. The best fit pure-shear model requires ~235 km of lithosphere extension for the N. Iberian - N. Newfoundland margin and 315 km for the N. Angolan margin. The best fit upwelling divergent flow model requires ~ 80 km of pure-shear extension for the N. Iberian - N. Newfoundland margin and ~25 km for the N. Angolan margin. Cross-plots of lithosphere and upper crustal thinning have been computed for both models. The upwelling-divergent flow model predicts depth- dependent lithosphere thinning and stretching which is particularly strong within the wide regions of highly thinned continental crust. Observations of depth-dependent stretching are compared with model predictions. For the N. Iberian - N. Newfoundland margins, the pure-shear breakup lithosphere thinning model predicts that the onset of melt generation occurs prior to breakup rupture of the continental crust for normal mantle temperature and chemical composition. In contrast the upwelling divergent flow model predicts the onset of melt generation after continental crust rupture leading to ~ 100 km mantle exhumation on each margin.
T32A-03
How warm are passive continental margins? A 3D case study from the Norwegian margin
Though being of increasing economic relevance, the heat flow regime at passive continental margins is debated. This is due to different hypotheses on the thermal structure below continents and oceans which, in turn, is related to the difference in thickness of the lithosphere and in thermal properties between both areas. Accordingly, the lithosphere-scale thermal field from the continent over the margin to the ocean is poorly understood. Here we show how far detailed knowledge of the crustal part of the system provides constraints for the thermal field of the lithosphere. We calculate the 3D conductive thermal field for a detailed structural model of the Norwegian passive margin and evaluate the lithospheric configuration consistent with two independent observables: temperature and gravity. We find that the thickness of a 55 Ma year old oceanic lithosphere controls to a large degree the conductive thermal field of the entire margin and needs to be less than 70 km. Our results support thickness estimates from seismology and disagree with predictions derived from cooling models assuming an equilibrium thickness of the oceanic lithosphere of 125 km. We obtain higher upper mantle temperatures and heat flows beneath the ocean than beneath the continent that can explain the reduced p-wave velocities and the smaller densities derived for the oceanic lithospheric mantle compared to the continent. The lateral temperature variation across the margin is ~100°K at 5 km depth and increases to ~400°K at 50 km depth. Thus, knowing the lithosphere configuration appears to be essential to correctly assess the present temperature conditions at passive margins and the related impact on the maturity of organic matter for petroleum exploration.
T32A-04
The dynamics of continental breakup-related magmatism on the Norwegian volcanic margin
The Vøring margin off mid-Norway was initiated during the earliest Eocene (~54 Ma), and large volumes of magmatic rocks were emplaced during and after continental breakup. In 2003, an ocean bottom seismometer survey was acquired on the Norwegian margin to constrain continental breakup and early seafloor spreading processes. The profile P-wave model described here crosses the northern part of the Vøring Plateau. Maximum igneous crustal thickness was found to be 18 km, decreasing to ~6.5 km over ~6 M.y. after continental breakup. Both the volume and the duration of excess magmatism after breakup is about twice of what is observed off the Møre Margin south of the Jan Mayen Fracture Zone, which offsets the margin segments by ~170 km. A similar reduction in magmatism occurs to the north over an along-margin distance of ~100 km to the Lofoten margin, but without a margin offset. There is a strong correlation between magma productivity and early plate spreading rate, which are highest just after breakup, falling with time. This is seen both at the Møre and the Vøring margin segments, suggesting a common cause. A model for the breakup- related magmatism should be able to (1) explain this correlation, (2) the magma production peak at breakup, and (3) the magmatic segmentation. Proposed end-member hypotheses are elevated upper-mantle temperatures caused by a hot mantle plume, or edge-driven small-scale convection fluxing mantle rocks through the melt zone. Both the average P-wave velocity and the major-element data at the Vøring margin indicate a low degree of melting consistent with convection. However, small scale convection does not easily explain the issues listed above. An elaboration of the mantle plume model by N. Sleep, in which buoyant plume material fills the rift-topography at the base of the lithosphere, can explain these: When the continents break apart, the buoyant plume-material flows up into the rift zone, causing excess magmatism by both elevated temperature and excess flux, and magmatism dies off as this rift-restricted material is spent. The buoyancy of the plume-material also elevates the plate boundaries and enhances plate spreading forces initially. The rapid drop in magma productivity to the north correlates with the northern boundary of the wide and deep Cretaceous Vøring Basin, thus less plume material was accommodated off Lofoten. This model predicts that the magma segmentation will show little variation in the geochemical signature.
T32A-05
Elevated Passive Continental Margins may form much Later than the time of Rifting
Many current models of the development of elevated passive continental margins assume that they are either the remains of foot-wall uplift at the time of rifting or due to underplating by magma from a plume or other mantle source. We have studied the rift and post-rift history of such a passive margin in West and South Greenland and have concluded that the present-day elevations developed 25-60 million years after cessation of rifting and local volcanism, suggesting that additional factors need to be considered when modelling such margins. The morphology of West Greenland is similar to that of other elevated passive margins ion many parts of the world. There are high-level, large-scale, quasi-planar landscapes (planation surfaces) at altitudes of 1-2 km cut by deeply incised valleys. The gradient from the highest ground to the coast is much steeper than that away from the coast. We combined analysis of the morphology of the landscape with studies of fission tracks and the preserved stratigraphic record both on- and off-shore. Rifting and the commencement of sea-floor spreading in the Early Paleogene was accompanied by voluminous high-temperature volcanism. Kilometer-scale uplift at the time of rifting was followed shortly afterwards by kilometer-scale subsidence and possibly by transgression of marine sediments across the rift margin. The present elevated margin formed during three episodes of uplift during the Neogene, 25-60 million years after the cessation of rifting and local volcanism. The quasi-planar planation surfaces presently at 1-2 km altitude are the end-products of denudation to near sea-level in the mid- and late Cenozoic and these surfaces were uplifted to their present altitudes during the Neogene events. Rivers then incised the summit surface to form valleys that were further enlarged and deepened by glaciers. Similar elevated margins exist all around the northern North Atlantic and in many other parts of the world; eastern North America, on both sides of the South Atlantic, western India, eastern Australia, and possibly in Antarctica. Our results show that we cannot simply assume that these elevations were produced either at the time of rifting or as underplating at the time of plume impact. There is, however, no general agreement as to what caused them and we suggest that the history of these margins need to be re-assessed in the light of our results.
T32A-06
Lithospheric Controls on the Rifting of Continents at Slow Rates of Extension
The North Atlantic Igneous Province (NAIP) has escaped a simple explanation for its vast size and thickness. The region has a complicated history with many sedimentary basins that may or may not contain oceanic lithosphere and that pre-date the NAIP by between 20 and 100 Ma. Should these regions of pre-thinned lithosphere be ignored however in models that hope to explain the presence of the NAIP? We use the major and rare earth composition of melts generated within our dynamic model of extension of the lithosphere to predict seismic velocities within the emplaced igneous material. Previous models of the North Atlantic make the case clear for the presence of an exhaustible thermal anomaly that lay under the lithosphere. We test models with a 200 m°C, 50 km thick thermal anomaly beneath 125 km thick lithosphere to the sensitivity to regions of pre-thinning. The mantle potential temperature is 1325 m°C. The aim is to assess the importance of such failed rift basins. Extensional events that have stretching factors more than 3 to 4; pre-date the rift by 20 to 40 Myrs; and are less than 150 km from the centre of extension of a successful rift, have the effect of reducing the viscosity of the upper lithosphere, creating a region where the flow of mantle material can reach shallower depths. For an exhaustible thermal anomaly to have an affect, it must be advected above the solidus before it loses significant amounts of heat to the surroundings. There are two ways to achieve this, have a fast rate of rifting or introduce a pre-thinned region. We suggest that for enhanced melting leading to the formation of the NAIP, on or off axis pre-thinning is likely requirement, given the half rates of extension are slow. We suggest that the pre- thinning caused by the formation of the Hatton-Rockall Trough is enough, when combined with a 200 m °C thermal anomaly, to explain the magmatism of the Southeast Greenland Hatton Bank conjugate system. Such pre-thinning leads to asymmetric melt regions that crystallise to form underplate with seismic velocities of more than 7 km s-1.
T32A-07
Crustal structure of the northern Seychelles margin: Evidence for restricted magmatism during breakup
We present a model of the structure of the northern Seychelles continental margin derived from wide-angle seismic travel-time inversion, teleseismic receiver functions and potential field data. Break-up occurred 2.5 Ma after, and 1000 km to the west of, the emplacement of the Deccan Traps. However, despite this spatial and temporal closeness, we do not find the typical set of geophysical characteristics reported at other volcanic rifted margins associated with continental flood basalts, such as those of the north Atlantic. The oceanic crust formed during the first 3 Ma of seafloor spreading is just 5 km thick, less than half that seen at some other volcanic margins. The continent-ocean transition itself is narrow and whilst two packages of seaward-dipping-reflectors are imaged they are weakly developed. Within the transitional lower crust there is a region of high-velocity material (7.5-7.7 km/s) directly below the inner seaward dipping reflector package that we interpret as an intruded body of mafic material, but it is small, just 10 km wide by 4 km thick. However a much more extensive area of probable mafic material (130 km wide and 6 km thick) is present beneath the entire stretched continental crust and part of the unstretched Seychelles Plateau. A similar pattern of underplated material beneath the stretched continental portions, but limited magmatism in the actual break-up zone and thin earliest oceanic crust, is also seen on the conjugate-Indian (Laxmi Ridge) margin. Together with the known ages of volcanic products from the Seychelles Plateau, we conclude that the underplating of the continental portions pre-dates the rifting, and that break-up between India and the Seychelles was characterised by a lack of excess volcanism. We suggest that the underplating event dehydrated, and hence increased the solidus of, the underlying mantle, such that the subsequent Seychelles-India separation, despite elevated temperatures did not generate excess volcanism. We conclude that tectonic history as well as mantle temperature can determine the degree of magmatism during continental break-up
T32A-08
Geodyamic Models of the Tectonomagmatic Evolution of the West Antarctic Rift System
Finite element geodynamic models reproduce the general aspects of the style and timing of extension in the West Antarctic Rift System (WARS). In particular, the models simulate a prolonged period of diffuse extension throughout the WARS during the Cretaceous and early Cenozoic followed by later focused extension in the Victoria Land Basin during the middle Paleogene. Two families of models are identified that are consistent with the Paleogene onset of magmatism in the West Antarctic Rift System under both normal and warm (70-120º hotter than global average) mantle potential temperatures. The two families of models produce distinctly different melt distribution patterns across the WARS. In models invoking normal mantle temperatures the greatest thickness of melt is confined to the Victoria Land Basin region. In models invoking warmer mantle temperatures, the greatest thickness of melt is widely distributed in the region currently underlaying the West Antarctic Ice Sheet. The models indicate that 1) the onset of Cenozoic magmatism in the Victoria Land Basin can be explained as rift induced without requiring the impingement of a plume or a change in plate motion, 2) if rifting is associated with a plume then the magmatic rocks under the West Antarctic Ice Sheet should be late Cretaceous to early Cenozoic in age, and 3) if the magmatic rocks under the West Antarctic Ice Sheet are older or younger than the main stage of rifting (ca. 105 Ma to 35 Ma), a syn-extensional plume could not have been present.