V31F-01
Explaining the Morphological Range of Hotspot Tracks Formed by Plume-Ridge Interactions: Towards a 3D Crustal Flow Model for the Oceanic Crust
Hotspot tracks formed by plume-ridge interactions show a large morphological diversity. The Greenland-Iceland- Faroes Ridge (GIFR) forms a sharp-edged plateau, with width of up to 700 km. The observed plateau which forms the GIFR is not a direct reflection of the underlying mantle convective structure, as it is at least three times wider than the Iceland plume, as imaged by seismic studies. Other tracks caused by ridge-hotspot interaction (e.g. segments of the Walvis and Chagos ridges) tend to be triangular with widths of about 300 km. It has been recently suggested that ductile flow of the lower crust is an important process in the explanation of off- axis crustal thickness changes in Iceland. Here we suggest that the dominant effect which produces the vastly different structure seen at the GIFR is the relatively slow spreading rate compared to those at the other ridges for which observations of plume-ridge interactions have been made. As a result of this slower spreading rate there are larger crustal thickness gradients, and this means that off-axis crustal mass distribution is an important factor. Two-dimensional modelling has previously been used to show that the crustal flow process can cause a morphology similar to what is observed, and that the flow must be primarily parallel to the axis. The present study considers the three-dimensional crustal flow problem, comparing and contrasting results using two different approaches to modelling very viscous flow: lubrication theory (`thin layer') and flow at low Reynolds number ('slow flow'). Initial work suggests that lubrication theory can only provide an accurate solution for the lower part of the channel in which the flow occurs, and that the 'slow flow' approach is more suitable elsewhere in the model. Further work will involve extending the model to incorporate more geologically reasonable boundary conditions and an analysis of the importance of hydrothermal cooling.
V31F-02
Icelandic Rift Relocations: Melting Models and Geochemical Observations
Rift relocations are distinctive features associated with plume-ridge interaction. The underlying processes controlling these relocations are not well understood. Observations of crustal thickness and basalt composition have been commonly used to investigate plume-ridge interaction, but both of these observations are affected by the rift relocation cycle. We therefore need to understand the rift relocation process better in order to isolate its effects from those caused by other processes, such as variations in plume flux, temperature or distance from the ridge. We present the first quantitative model of variations in melt production rates and composition during a rift relocation cycle including the growth and death of spreading centres. The two dimensional time-dependent model of passive upwelling beneath a spreading ridge is based on simple corner flow. The thermal structure is calculated using a control volume approach. The resulting melting structure is then calculated and used to predict the variation in crustal thickness and composition through time. The model shows an initial phase of small melt fractions as the ridge grows, moving into the steady-state phase showing much larger melt fractions and thicker crust. Finally, as the ridge dies and spreading rate decreases, conductive cooling results in a reduction in the melt fraction. The results of the model are compared to geochemical and geophysical observations of the current rift relocation in southern Iceland and a palaeo-rift relocation in northern Iceland. A comparison is also made with the abandoned Phoenix Ridge, Drake Passage, Antartica. The geochemical observations are consistent with smaller degrees of melting at both the birth and death of rift zones. Combination of these observations and the model results will be used to constrain the rate of transfer of spreading between rift zones.
V31F-03
Deformation of Torfajokull Caldera, Iceland
The Mid-Atlantic Ridge, the plate boundary between the North American and Eurasian tectonic plates, is expressed in Iceland as neovolcanic zones consisting of central volcanoes and fissure swarms. The Mid-Atlantic Ridge becomes subaerial here due to its position above the North Atlantic mantle plume. In south Iceland plate motion is accommodated across the western and eastern volcanic zones. The eastern volcanic zone (EVZ) is thought to have formed approximately 2-3 Ma and is propagating to the southwest into crust formed during the Tertiary in the western volcanic zone. This has resulted in the formation of more evolved volcanic systems of the eastern volcanic flank zone (EVFZ). Torfajokull caldera is located at the intersection of the EVZ and EVFZ, has been active since late Quaternary time, with the most recent eruption in 1477 AD, and is the largest rhyolitic complex in Iceland. The caldera hosts the largest geothermal field in Iceland and therefore has been the focus of ongoing geophysical and geologic investigation. This study investigates the complex interplay between plate spreading and volcano deformation at this caldera using space geodetic and modeling techniques. We present horizontal and vertical velocity fields based on episodic GPS measurements collected between 2000- 2006. In a stable Eurasian reference frame the horizontal signal is dominated by plate spreading through the caldera, in agreement with known Holocene fissuring northeast and southwest of the caldera and the results of LaFemina et al. (2005). The horizontal velcoity field also indicates the interaction of Hekla volcano, 20 km west- northwest, which has been uplifting since erupting in 2000. Vertical rates on average exceed 15 mm/yr and may indicate glacio-isostatic adjustment. Nevertheless, colocated GPS and dry tilt in the center of the caldera indicate subsidence at up to 10 mm/yr.
V31F-04
Tristan da Cunha Hotspot Tracks and the Seafloor Spreading History of the South Atlantic
Careful mapping of more than 50 distinctive seafloor spreading magnetic anomalies between magnetochrons C5 and M4 has been used to develop a detailed spreading history for the South Atlantic between 15°S and 45°S and to investigate spreading axis interactions with the Tristan hotspot. Spreading appears to be roughly symmetrical with 3 separate phases identified: (1) steady spreading of ~30 mm/yr from 10 to ~45 Ma, (2) slower (15-18 mm/yr) and more variable spreading from 45 to ~70 Ma, and (3) an earlier faster spreading (45 mm/yr) from 70 to 84 Ma. C34-M0 distances provide an average value for pre-C34 spreading of ~27mm/yr. The noticeably larger C34-M0 distance on the S. American side near 32°S, 40°W is attributed to one or more eastward ridge jumps that occurred between 84 and 120 Ma. Residualized free air satellite gravity data have been used to delineate fracture zones (FZs) associated with the Early Cretaceous through Tertiary opening. More than 20 flow lines determined from these FZs intersect the magnetic lineations mapped between C5 and M4. The FZs and isochron data have been used to compute stage poles from 130 Ma to the present for both the South American and African sides, and the corresponding total reconstruction poles. Features attributed to hotspot activity on each plate have been rotated about these poles to determine when the hotspot was beneath the spreading axis. Coincident on-axis locations and the age of the underlying oceanic crust have then been used to document the history of the interaction of the Tristan hotspot with the spreading axis. In a fixed Africa reference frame the hotspot appears to move generally west-southwest. From 120 Ma to ~70 Ma, spreading was sufficiently rapid that the hotspot was maintained on or close to the ridge axis by eastward ridge jumps. From 70 to ~45 Ma the slower spreading allowed the hotspot to remain on the ridge without ridge jumps. At 45 Ma the hotspot crossed beneath the Meteor transform fault, was displaced from the ridge axis and became isolated beneath the African plate. The subsequent increase in spreading rate resulted in the hotspot remaining beneath the African plate where it has produced that part of the Walvis Ridge track between the Meteor FZ and Tristan from 45 Ma and the present. Comparison of paleolatitudes determined for Walvis Ridge features north of the Meteor FZ with the present location of Tristan suggests no significant latitudinal motion of the Tristan hotspot since ~ 80Ma. Between 130 Ma and 80Ma, however, the paleolatitude of hotspot features are roughly 10° more northerly that present day Tristan.
V31F-05
Does the Reunion Hotspot Volcano Emplace on a Fossil Ridge or a Fracture Zone?
It is now clear that the structure and the mechanical properties of the lithosphere have to be taken into account to understand how mantle plumes are expressed by surface volcanism. Reunion, a large volcanic system in the Indian Ocean, is widely considered as the most recent expression of a mantle plume. Previous studies have suggested that it might be located above pre-existing lithospheric structures such as a fossil spreading ridge or a fracture zone. Cruise FOREVER (FORmation and Evolution of the Volcanic Edifice of Reunion) of R/V L'Atalante has recently surveyed the oceanic plate around Reunion Island in order to investigate possible relationships between the structures of the plate and the emplacement of surface volcanism. The cruise collected multibeam bathymetry and back-scatter data, as well as magnetic, gravity and 24-channel seismic reflection profiles. The coverage extends up to 250 km around the island. The new data confirm that the formation and evolution of the oceanic plate in the Reunion area is more complex than in the adjacent compartments. Oceanic magnetic lineations display various directions (discussed in Dyment et al., this meeting). In addition to the Reunion large volcanic edifice, the high resolution bathymetry coverage reveals numerous volcanic structures on the surrounding oceanic plate. Their morphology varies from elongated ridges to isolated seamounts. The seismic data, complemented by older multichannel seismic data (cruise REUSIS, 1993), allow investigations of the oceanic plate topography below the sedimentary cover. Preliminary results indicate that a fossil spreading axis is unlikely to underlie Reunion but reveal a N35° E topographic structure, possibly a short fracture zone. The seismic data also document the evolution of lithospheric flexure related to Mauritius and Reunion loading. FOREVER cruise scientific party : IPGP : Georges Boudon ; CNRS/IPGP : Christine Deplus, Jerome Dyment, Anne Le Friant ; Univ. Reunion/IPGP : Laurent Michon, Francky Saint-Ange ; Univ. Pau: Beatrice de Voogd; OPGC : Jean-François Lenat ; ENSG Nancy : Marine Perus ; IRD/Géosciences Azur : Bernard Pontoise
V31F-06
Do the Isolated Volcanic Islands of Mauritius and Reunion reflect a Thicker Underlying Lithosphere or a Weakening Reunion Hotspot for the last 10 Ma?
Reunion Island is the most recent expression of a hotspot which formed the Deccan Trap flood basalt, the Chagos-Laccadives Ridge (CLR), the southern part of the Mascarene Plateau (SMP), Mauritius and Reunion Islands. Whereas CLR and SMP are continuous structures, both Mauritius and Reunion Islands are isolated structures. Such an observation may reflect a thicker oceanic lithosphere under the Mascarene Islands, a weakening Reunion hotspot, or both. A related question for the Mascarene Islands is whether the rising hotspot material has used pre-existing structures of the oceanic lithosphere to reach the surface and create the volcanic edifices, or if the isolated islands reflect pulses in the hotspot activity. Recent bathymetric, magnetic and seismic data collected by R/V L'Atalante in 2006 as part of cruise FOREVER (FORmation and Evolution of the Volcanic Edifice of Reunion) help to address these questions. The full bathymetric coverage and the dense seismic survey allow the construction of a basement map that reveals a structure, possibly a short fracture zone, beneath the volcano (Deplus et al., this meeting). The magnetic anomaly map built from FOREVER and older reprocessed data displays coherent magnetic anomalies. Interestingly, the seafloor spreading anomalies can be deciphered under most of the edifice (radius ~ 100 km) with only an inner zone of radius ~ 50 km showing shorter wavelength anomalies related to the volcanic structures of the island. Anomaly 32, 31 and 30 are tentatively identified west of Mauritius, north and east of Reunion Island, and south of Reunion Island, respectively. The seafloor spreading lineations show two orientations, N120° E- N140° E and N90° E-N110° E in the central and eastern part of the compartment, respectively. These different orientations may be related to the presence of a triple junction trace which could have been the one connecting the Mascarene fossil spreading centre to the Southeast Indian Ridge.
V31F-07
Melt Production and Morphology Variations Along the Southeast Indian Ridge
The Southeast Indian Ridge (SEIR) is an intermediate spreading rate ridge with an along-axis transition in axial ridge morphology from an axial high to an axial valley, which occurs independently of geochemistry and spreading rate. We investigate possible mechanisms for this transition by considering two segments separated by 600 km, one characterized by an axial high and the other by a shallow axial valley. The crustal structure of these two segments is well constrained through seismic reflection and refraction studies. We find that crustal structure differences account for only 70 m difference in off-axis (isostatic) depth, compared with an observed 237 m depth difference. We explore two models to account for the variation in morphology, crustal structure and depth by varying mantle properties. First, we consider a model that assumes a decrease in mantle temperature approaching the Australian-Antarctic Discordance. Mantle melt production models imply a 12°- 15°C mantle temperature difference between the two ridge segments based on the difference in crustal thickness. This result is compatible with the observed difference in mantle Bouguer anomalies between the two segments. Since the transition from an axial high to an axial valley occurs within a single segment located between the two segments for which we have seismic data, the actual temperature difference over which the change in the mode of crustal accretion occurs must be significantly less than 15°C. In the second model, we consider a constant temperature asthenosphere that thins toward the AAD. If the base of the asthenosphere lies above or cuts across the base of the potential melting zone, melt production would decrease as the AAD is approached. If a 100°C difference in temperature between the asthenosphere and mesosphere is assumed, then an ~3° slope (33 km depth difference over 600 km) on the base of the asthenosphere is required to match the gravity data. Both models predict that axial morphology and shallow crustal structure are highly sensitive to small changes in melt production. This is consistent with a threshold type relationship between melt production and the mode of crustal generation that can result in an abrupt along-axis transition in crustal structure and axial morphology driven by small-scale changes in mantle properties.
V31F-08
Triple-Junction Magmatism: Arc-Affinity Lavas on the Woodlark Spreading Center
The Woodlark spreading center exhibits unusual, arc-affinity lavas near its triple-junction intersection with New Britain trench, at the New Georgia Group of the Solomon Islands. These lavas have major element, trace element, and isotopic compositions that are generally intermediate between the N-MORB erupted on the mid- ocean ridge further from the trench and enriched arc laves from the New Georgia Group. These unusual lavas may be the result of mixing between MORB and arc magmas that have been channeled across the trench boundary within the conduit of a slab window. Near-trench volcanism at Kavachi Seamount in the forearc region about 30 km from the trench is postulated to be a result of melts produced by the ‘blowtorch effect' above the slab window. Similar forearc volcanism and arc-affinity lavas are found at the triple junction formed by the Chile Rise and trench, suggesting that these magmatic effects may be common consequences of ridge subduction.