V42B-01 INVITED
Melt Flux Around Iceland: The Kolbeinsey Ridge Seismic Experiment.
Seafloor spreading within the Iceland region has been complex since the opening of the North Atlantic in late Paleocene-early Eocene. Whereas symmetric magnetic anomalies can be traced parallel to the Reykjanes Ridge and Mohns Ridge back to chrons 23-24, anomalies within the Iceland Plateau and Aegir Ridge in the Norwegian Sea, as well as along the Greenland-Iceland-Faeroe Ridge reflect plate-boundary irregularities associated with multiple-branched crustal accretion zones, rift jumps and plate boundary segmentation (volcanic systems). We observe large variations in crustal structure along two refraction/reflection/gravity profiles, a 700 km EW-profile straddling 66.5°N between the Aegir and Kolbeinsey Ridges and a 225 km NS-profile along the southern Kolbeinsey Ridge. These profiles enable us to quantify how melt flux at the N-Atlantic spreading center has been influenced by the Iceland hotspot from the initiation of spreading to present time. The westernmost 300 km of the EW profile lies across the Iceland shelf, considered to have formed by rifting at the Kolbeinsey Ridge whereas the easternmost 400 km lie across the Iceland Plateau and Norway Basin, a region formed by rifting at the Aegir Ridge and possibly containing slivers of older crust rifted off the east Greenland margin along with the Jan Mayen Ridge. Crustal thickness varies from 4-5 km across the Aegir Ridge, 12 km just east of the Iceland shelf, and 24-28 km beneath the outer shelf, to 12-13 km near the southern tip of the Kolbeinsey Ridge and 9-10 km further north along the ridge axis. Pronounced undulations in lower crustal structure across the Iceland Plateau are most likely associated with extinct spreading centers indicating that branched crustal accretion zones existed west of the Aegir Ridge prior to the westward ridge jump forming the KR at 26 Ma. Crustal thickness at the Kolbeinsey and Reykjanes Ridges support the asymmetry in plume-ridge interaction north and south of Iceland that has been inferred from differences in axial depth (400-750 m) and geochemistry between these two ridges. Gravity inversions indicate a sub-crustal residual gravity anomaly that decreases by about 20 mGal toward Iceland which implies a density reduction of about 10 kg/m3 in the upper 100 km of the mantle along the Kolbeinsey Ridge profile. If solely due to thermal expansion this suggests a temperature increase of about 100°C toward Iceland.
V42B-02
Propagating Rift Explanation for the V-Shaped Ridges South of Iceland
During June-July 2007 we surveyed the Reykjanes Ridge system south of Iceland to investigate the origin of the V-shaped ridges seen in both bathymetry and gravity, 1st identified by Vogt, and usually interpreted as evidence for pulsing Iceland plume material flowing down an axial conduit. However, there are 2 processes that can produce V-shaped wakes along ridges: flow down an existing axis, and rift propagation. Flow down the pipe models, whether of magma pulses (Vogt, 1971) or magma deficits (Hardarson et al., 1997), predict V-wakes symmetric about the axis, whereas the propagating rift model predicts asymmetries caused by the lithospheric transfer. Contrary to previous interpretations from south of 62 N which suggested simple and continuous ridge evolution, and symmetric V wakes consistent with flow down the pipe models, our survey between 62 N and Iceland found evidence for multiple ridge reorganizations caused by new rifts propagating south from Iceland, and asymmetric V wakes. The asymmetry is highest near Iceland but moderate even south of 62 N. Thus at least some of the V-shaped ridges are the pseudofault/failed rift wakes of propagating rifts which are breaking through the Eurasia plate and transferring lithosphere to the North America plate, rather than resulting from plume pulses flowing down the existing axis. A plume pulse model could still be valid, but only with the modification that the plume pulses drive propagating rifts ahead of them that provide new conduits for the extra plume magma. Alternatively, it might be possible for a steady-state plume interacting with a non-steady state ridge axis, rather than vice versa, to produce the observed pattern. Propagating rifts that periodically open more favorable conduits for the plume flow could conceivably produce V-shaped ridges even from a non-pulsing plume. http://www.soest.hawaii.edu/HIGP/Faculty/hey/rr2007/index.html
V42B-03
Structure of the Mantle Transition Zone Beneath Kenya, East Africa
The relationship between the surface expression of magmatism and extension with the deeper mantle processes remains enigmatic. While Cenozoic volcanism, plateau uplift, and rifting in the Kenya Rift of eastern Africa must be associated with a mantle thermal anomaly, the exact nature of the anomaly is highly speculative. Previous studies of the mantle structure beneath Tanzania suggest that the region is underlain by a plume head, but that the plume stem may be located further to the north in Kenya. To further define the thermal structure of the mantle beneath the Kenya Rift we analyze teleseismic P-waves and PP-waves from a temporary PASSCAL Kenya Broadband Seismic Experiment, permanent stations in Kenya and Tanzania , and re-processed teleseismic P-waves and PP-waves from the Tanzania Broadband Seismic Experiment. The resulting coverage is densest in south-central Kenya and northern Tanzania. Results from stacking the receiver functions with a 1D velocity model reveals depressions in of the 410 km discontinuity, suggesting thermally perturbed mantle structure. However, the depth to the 660 km discontinuity is slightly elevated throughout much of the study area, inconsistent with the presence of a mantle tail. The calculated thickness of the mantle transition zone (MTZ) in southern Kenya shows no systematic trend. In Tanzania, the MTZ is somewhat thinned in the north-central area and thickened in the west-central area, consistent with previous receiver function analysis in Tanzania.
V42B-04
Two Plumes Beneath the East African Rift System: a Geochemical Investigation into Possible Interactions in Ethiopia
East African Rift System magmatism began over 40 my ago and has continued through the present. Numerical models have determined two plumes are necessary to create the spatial and temporal distribution of volcanism. Geochemical data support the presence of two chemically distinct plumes initially located beneath the Afar Depression (NE Ethiopia) and the Turkana Depression (SW Ethiopia/N Kenya). The timing and eruptive of the Afar and Kenya plumes are also distinct. While there is growing evidence to support the existence of two dynamically and chemically distinct plumes beneath the East African Rift System, the interactions between them remain unclear. Our study focuses on the geochemistry of mafic shield lavas from three locations on the eastern flank of the Ethiopian plateau. These lavas are spatially located between the surface manifestation of the Afar and Kenya plumes. The majority of the lava is alkaline and has experienced varying degrees of olivine and pyroxene fractionation. The northernmost lavas (9°10'N) are transitional and display the most fractionation. Primitive mantle melts were generated at depths near the fertile mantle garnet-spinel transition zone and deeper (80-100km) and are free of metasomatic influence. Minor HREE depletions also support derivation of melts from a garnet-bearing source. Lavas with lithospheric influence are generated from shallower depths and show minor amphibole influence. Overall, geochemical data show the lavas in this study closely resemble those from various episodes of Kenya plume magmatism with modifications attributed to lithospheric contamination. This interpretation is consistent with current numerical models suggesting episodic northward movement of Kenya plume magmas along the lithosphere-asthenosphere boundary. The data imply that the Kenya plume has a much larger spatial influence and therefore a larger geodynamic influence in the EARS than previously recognized.
V42B-05 INVITED
Temperature Versus Buoyant Mantle Heterogeneities, Evaluating the Origin of OIB Using the Galapagos Archipelago
The existence of hot mantle plumes has recently been questioned; instead tectonics and shallow low-melting heterogeneities, rather than excess temperatures, have been proposed to explain the upwelling of mantle beneath oceanic islands basalts (OIB) [1]. Geophysical, petrological and geochemical arguments have been used to suggest that enriched heterogeneities located shallow within the upper mantle are buoyant due to either melt retention or water lowering its solidus and consequently triggering early melting of the fertile blobs. Thus, the opponents of the hotspot model have suggested that shallow fertile and buoyant blobs could reproduce the effects of temperature, including isotope gradients, as the cause for the generation of OIB [1]. Uranium-series disequilibria measured in oceanic lavas may provide insight into the origin of OIB that is independent of previous arguments. Earlier works have shown that U-series isotopes in basalts are sensitive to variations in the upwelling rates of their mantle source, and those can be used to invoke upward movements that may be associated with mantle plumes [2–4]. Those work also demonstrated that there exist a significant relationship between the extent of U-series disequilibria and the buoyancy flux beneath OIB; where lower buoyancy fluxes are associated with lower rates of melting and hence higher (230Th/238U) values [2]. However, most of those works have assumed that the effects of source heterogeneity can be neglected relative to the effects of other key parameters. If the increase in melting rates is truly due to the presence of fertile blobs, then there should be a correlation between clear indices of enrichment such as radiogenic isotopes and extent of U-series disequilibrium in OIB. Furthermore, the fertility of the mantle heterogeneities may or may not be associated with the enrichment of water, which has an important effect on the melting rate and consequently in the U-series isotope of the basalts. To evaluate what factors (temperature versus mantle heterogeneity) are controlling the upwelling mantle beneath oceanic islands we used our data (including volatile and U-series isotopes) for lavas from across the Galapagos archipelago combined with previously published results from the adjacent Galapagos Spreading Ridge lavas [5]. These results suggest that the most important factor controlling the variation in mantle upwelling velocities beneath the Galapagos archipelago and adjacent ridge is the difference in mantle temperature. References 1. Plates, Plumes, and Paradigms (eds Foulger, G.R., Natland, J.H., Presnall, D.C. & Anderson, D.L.) (GSA Special Paper 388, Geological Society of America, Boulder, 2005). 2. Bourdon, B. et al. Nature 444, 713-717, 10.1038 (2006). 3. Bourdon, B. & Sims, K. W. W. in U-series Geochemistry (eds Bourdon, B., Lundstrom, C., Henderson, G. & Turner, S. P.) 215–253 (Mineralogical Society of America, 2003).
V42B-06
Geochemical Evolution of the Louisville Seamount Chain
The Louisville seamount chain is a 4300 km long chain of submarine volcanoes in the southwestern Pacific that is commonly thought to represent a hotspot track. It spans an ~80 Myr age range, comparable to that of the Hawaiian-Emperor chain (Koppers et al., G-cubed, 5 (6), 2004). The few previously dredged igneous samples are dominantly basaltic and alkalic, and have been inferred to represent post-shield volcanism (Hawkins et al., AGU Monograph, 43, 235, 1987). Their isotope and trace element signatures suggest an unusually homogenous mantle source (Cheng et al., AGU Monograph, 43, 283, 1987). Dredging in 2006, during the AMAT02RR cruise of the R.V. Revelle, was carried out in the hope of recovering both shield and post-shield samples and of exploring the geochemical evolution of the chain. Igneous rocks were recovered from 33 stations on 23 seamounts covering some 47 Myr of the chain's history. Our study, focusing on the major and trace element and Sr, Nd and Pb isotopic characteristics of these samples, shows that all are alkalic basalts, basanites and tephrites containing normative nepheline. Variations in major and trace elements appear to be controlled predominantly by variable extents of melting and fractional crystallization, with little influence from mantle source heterogeneity. Indeed, age-corrected isotopic values define only a narrow range, in agreement with long-term source homogeneity relative to the scale of melting; e.g., εNd varies from +4.1 to +5.7, 206Pb/204Pb from 19.048 to 19.281, and 87Sr/86Sr from 0.70362 to 0.70398. These values broadly fall within the fields of the proposed "C" or "FOZO" mantle end-members. However, small variations are present, with less radiogenic Nd and Pb isotope ratios at the older, western end of the chain, defining a trend toward a broadly EM2-like composition. Although some workers have postulated that the Louisville hotspot was the source of the ~120 Myr Ontong Java Plateau, our samples are isotopically distinct from any known Ontong Java compositions.
V42B-07
Constraining the Early Isotopic and Trace Element Signature of the Yellowstone Mantle Plume: Evidence from Imnaha Basalts
Characterizing the geochemical signatures of plumes is critical for evaluating the petrogenetic evolution of plume- related volcanic rocks. The main phase of Columbia River flood basalt activity (16.6 – 15.5 Ma), considered to represent the first clear manifestation of the Yellowstone hotspot on the North American continent, includes the Steens Mountain, Imnaha, Grande Ronde and Picture Gorge basalts. Isotopic and trace element covariations defined by Grande Ronde, Steens Mountains, and Picture Gorge basalts diverge radially from the field of Imnaha basalts, which retain signatures that most closely reflect the ‘undiluted' geochemical characteristics of the Yellowstone mantle plume. Sr, Nd, Pb isotope ratios and incompatible trace element abundances and ratios of Imnaha basalts closely resemble those of some Pacific EM II OIB groups. Nonetheless, the compositions of some Imnaha lavas reflect mixing of the plume with different mantle types. Others have clearly been affected by interaction of plume-derived basalt with continental crust, although the latter process is much more significant in the genesis of the succeeding Grande Ronde basalts. We will review the geochemical characteristics of the Yellowstone plume as recorded in the Imnaha basalt in the context of later-erupted volcanic products of the Columbia - Snake – Yellowstone system, and Pacific mantle plumes more generally.