V33B-1380
Body-wave tomography at Hawaii from the first PLUME deployment of ocean-bottom seismometers
The Hawaiian PLUME (Plume-Lithosphere Undersea Melt Experiment) project is a multi-disciplinary program to study the deep mantle structure of the Hawaiian hotspot and address the debate over whether one end of the island chain is underlain by a classical plume from the deep mantle and how mantle upwelling interacts with the overlying lithosphere beneath the Hawaiian Swell. PLUME involved two consecutive ocean-bottom seismometer (OBS) deployments and a concurrent deployment of 10 land seismometers along the islands. The first deployment of 35 broadband OBSs in 2005-2006 was centered on the island of Hawaii with stations spaced about 75 km apart. A second OBS deployment with larger aperture and larger station spacing was carried out in 2006-2007. Here, we present analyses of body-wave arrival time data from the first PLUME OBS deployment. The OBS horizontal components were azimuthally oriented from particle motions of teleseismic P waves. High quality relative arrival times of P and S waves were measured with the multi-channel cross-correlation method of VanDecar and Crosson [1990]. In the frequency band 0.05-0.1 Hz, P and S wave analyses each yielded more than 1000 arrival times from about 40 earthquakes moderately well distributed in azimuth. Because of the high noise levels on the seafloor at Hawaii, most measurements are from Mw ≥ 6.0 earthquakes. We also obtained about 300 P-wave arrival times from 10 earthquakes in the frequency band 0.5-1 Hz. P-wave arrival times were used in checking clock drift corrections to the OBS data. The separate arrival time datasets have been inverted for P- and S-wave velocity models beneath Hawaii using both ray theoretical and finite-frequency methods; results from both approaches display general similarity. Data from the second OBS deployment, once analyzed, should be useful in broadening the region of consideration and extending downward the depth resolution of the imaging.
V33B-1381
Crustal and Upper Mantle Structure Beneath the Canary Islands From Teleseismic Receiver Functions.
The Canary Islands are situated in the North Atlantic Ocean, <200km west of Morocco, Africa. The islands are volcanic ocean islands, associated with the classic hot spot characteristic combination of bathymetric, gravity and geoid anomalies. However, unlike the classic hot spot location of Hawaii, the archipelago is located on a slow moving plate, showing more similarities to the Cape Verde Islands, but unlike both Hawaii and Cape Verde, the Canary Islands are close to the continental shelf. The aims of this work are to provide seismic constraints on the structure beneath the Canary Islands to determine whether this structure indicates a clear age progression across the archipelago as observed at Cape Verde and to determine whether deeper structure may illuminate the source of the hot spot features. To take a transect through the Canary Islands using receiver function analysis, we re-analysed broadband data from the MIDSEA project station (available through IRIS), CDLV on Lanzarote (1999-2001), but apply the multiple- taper spectral correlation estimate for receiver function calculation. We also analysed broadband data from the IRIS Network station of TBT from La Palma (1993-1996). Additionally we also use data from a short period seismic network consisting of 150 short period stations installed for 2 weeks as part of the TOM-TEIDEVS project on the island of Tenerife. Only 1 teleseismic event suitable for receiver function analysis was recorded during this period. Initially an average of all events was to be used for modelling, but significant differences in receiver function shape between different areas of the island, suggested separate stacks for different regions was more appropriate. Initial forward modelling for the average azimuthal stack for CDLV, supports earlier receiver function work that indicates a crust thickened up to ~20km depth, but no evidence of a continental like structure. The average azimuthal stack for TBT shows few details, but when events are grouped by back azimuth and stacked, significant differences in shape are observed. Comparison of the tangential components for different back azimuths, suggests the existence of dipping and/or anisotropic layers. Forward modelling for a stack of data from the caldera in Tenerife, indicates no crustal thickening. This suggests a relationship between crustal thickening and age across the archipelago. Further analysis and application of grid search methods will reveal the structure beneath each of the islands in more detail and indicate whether a thermal or a compositional origin is more appropriate for the islands.
V33B-1382
Surface Wave Data Collected During the First Hawaiian PLUME OBS Deployment
During the seismic component of the Hawaiian PLUME (Plume-Lithosphere Undersea Melt Experiment) project from January 2005 through June 2007, we collected continuous seismic data at nearly 80 seismic stations. Ten broadband land stations were equipped with Wielandt--Streckeisen STS-2 seismometers, and about 70 ocean bottom sites were occupied with Güralp CMG-3T, Nanometrics Trillium 40 or Trillium 240 seismometers and a Cox--Webb differential pressure sensor. Data collected with such sensors provide the ideal basis to analyze surface waves across a broad period band. Our initial assessment concentrates on long-period teleseismic Rayleigh waves collected during the first phase of the two-stage OBS deployment from January 2005 through January 2006. In this one-year deployment, 35 sites were occupied in an elongated array centered on the island of Hawaii, with a station spacing and average aperture of roughly 75~km and 500~km, respectively. The OBSs at 32 of these sites were recovered, with 25 providing useful vertical component or pressure records. We collected records from upward of 95 suitable large, shallow earthquakes with surface wave magnitudes 5.6 or larger and source depths less than 200~km. Records from numerous smaller events are also available and initial data inspection suggests that many of these can also be used. In the first step of our analysis, the frequency--dependent phase is measured with respect to a reference station. This primary phase database contains about 1500 unique high-quality source--receiver OBS measurements and the land stations are projected to add another 500 data. We utilize this phase database in subsequent tomographic procedures, for example when we model two--station path--averaged phase velocity curves. We constructed dispersion curves along numerous crossing two--station paths that are each well constrained by several earthquakes. The dispersion curves are usually internally consistent between 15 and 50~s, allowing us to image the lithosphere and upper asthenosphere. Some larger events provide constraints beyond 100~s, thereby illuminating the lower asthenosphere. We find significant heterogeneity across the array. Higher phase velocities to the southeast of the island of Hawaii resemble those of 52-100 Myr old Pacific lithosphere as found by Nishimura and Forsyth (1989) though velocities do not reach the much higher off--swell velocities found to the southwest during our 1997/98 SWELL pilot experiment. The lowest phase velocities are found to the west and northwest of the island of Hawaii, suggesting a strong asymmetry of cross--swell structure. http://mahi.ucsd.edu/Gabi/plume.html
V33B-1383
Characterizing Offshore Earthquakes at Hawaii Recorded by the First PLUME Temporary Ocean-Bottom Seismometer Network
In 2005-2006 and again in 2006-2007, the Plume-Lithosphere Undersea Melt Experiment (PLUME) deployed successive networks of ocean-bottom seismometers (OBSs) around the Hawaiian Islands. The experiment consisted of a 2-year deployment of broadband land seismometers and two year-long deployments of broadband OBSs, the first with a station spacing of about 75 km centered on the island of Hawaii and the second with larger spacing of about 200 km. PLUME's major objective was to determine the mantle structure beneath the Hawaiian hotspot and swell; however, these unique data are also potentially valuable to the study of small offshore earthquakes. The Hawaiian Islands are marked by significant and continuous seismic activity. In addition to the thousands of microearthquakes that are detected and located by the USGS Hawaiian Volcano Observatory (HVO) seismic network each year, Hawaii also experiences occasional large, damaging earthquakes. Several of these large events occurred in Hawaii's offshore region (e.g., the 1871 Lanai earthquake, the 1938 Maui earthquake, and the 2006 Kiholo Bay earthquake), and such events pose a significant seismic hazard for the state. We assess whether data from the first PLUME OBS deployment and land data can improve the detection and location of offshore microearthquakes around Hawaii. We are particularly interested in whether the PLUME data set may reveal offshore fault zones not detected to date by the HVO seismic network. Initial tests indicate that many offshore earthquakes already in the HVO catalog produce detectable P and S waves on the PLUME three- component seismometers, and earthquake detection rates are improved when seismograms are high-pass filtered above about 5 Hz to reduce the seismic noise from wind-generated waves. Differential pressure gauge data yield far fewer detectable events (with the exception of a swarm of Loihi earthquakes in December 2005) and appear less promising for improving our knowledge of offshore seismicity patterns.
V33B-1384
Influences of Source Heterogeneity and Potential Temperature on the Generation of Basalts from the North Atlantic Large Igneous Province
Source heterogeneity and mantle potential temperature are important factors affecting the mantle melting process and must be adequately accounted for to interpret the compositions of mantle-derived melts derived from a variety of tectonomagmatic settings. Several previous studies have attempted to simplify the problem by modeling the melting of a homogeneous source, or by creating 1-D melting models of a heterogeneous source (i.e. solid-solid source mixing). The results of these studies are limited as the latter approach places no constraints on pressures and extents of melting while the former approach does not account for the role of non- peridotite lithologies in the source. We have begun investigating 2-D models using a modified version of the REEBOX polybaric melting model of [1] in an attempt to gain insight into melt compositions related to the effects of the melting process (related to potential temperature) while simultaneously identifying the chemical effects of source heterogeneity. This new model incorporates experimentally - determined melt productivity functions, melting reactions and solidus pressures for eclogite and peridotite source lithologies. At a given potential temperature, model outputs produce grids in trace element and isotope space that reflect variable mixing of melts derived from eclogite and peridotite source lithologies at different segregation pressures from the melting column. Consequently, we are able to qualitatively assess relative pressures and extents of melting within the melting column for comparison with basalt compositions. To better understand the influences of source heterogeneity and potential temperature within the North Atlantic Large Igneous Province, we apply this model to volcanic products from the Paleocene rifted margin of Greenland and the modern ridge system. In the case of the east Greenland flood basalts, variations in contributions from a source involving highly depleted peridotite and eclogite and changes in potential temperatures can be distinguished. These insights are useful in constraining the nature of the chemical and thermal anomaly associated with the ancestral Iceland plume. (1) Fram and Lesher, 1993, Nature, v 363, p 712-714
V33B-1385
P- and S-wave Tomography of the Cameroon Volcanic Line Used to Discriminate Between Formation Models
The Cameroon Volcanic Line (CVL) is a 1600km feature traversing both continental Cameroon in west Africa and the offshore islands of Bioko (part of Equatorial Guinea), Sao Tome and Principe, and Annobon (also part of Equatorial Guinea). The CVL is a fairly linear feature, suggestive of the movement of the African plate over a stationary hotspot, but the volcanic rock ages of the CVL range from 42Ma to the present (with present volcanism occurring in the center of the line at Mt. Cameroon), contrary to what would be expected from a stationary hot spot. Several hypotheses have been proposed for the formation of the CVL. One possible explanation has been that there may be a plume rising to the surface along of previously weakened linear zone. In this scenario, the magma may rise sporadically to the surface at different points along the line thereby explaining the apparent lack of age progression. Other suggested models include edge-flow convection from the nearby Congo craton, the existence of multiple plumes, and a leaky transform. The Cameroon Seismic Experiment was deployed in Cameroon from January 2005 to January 2007, with 8 stations active the first year and an additional 24 stations installed in January 2006. The data from the 32 broadband seismometers is currently being used for a body-wave tomography study to study the upper mantle structure beneath Cameroon. Preliminary results from a P- and S-wave travel time tomography show a linear negative velocity anomaly directly beneath the CVL that extends from shallow mantle depths to at least 350km. This finding suggests that the anomaly is not caused by regional flow patterns associated with the Congo craton.
V33B-1386
Rayleigh Wave Phase Velocity Structure of the Cameroon Volcanic Line Region
The Cameroon Volcanic Line (CVL) is an 1800 km lineament of Cenozoic volcanism that extends from the Gulf of Guinea into Central Africa. With most of the line having experienced recent holocene volcanism, the CVL is one of the more prominent volcanic tracks exhibiting little to no discernible age progression. Previous interpretations of the CVL have been diverse and included volcanic activity associated with laterally transported material from a plume, multiple plumes, rift-flank processes, plate-driven Richter rolls, plume-fed small-scale convection cells, or propagating lithosphere cracks in order to reconcile the lack of age progression with the strong linearity and position of the CVL. Discriminating between these models can be significantly aided by the determination of the extent and depth of the source of the volcanism. In early 2005, a dense broadband seismometer array located over the continental section of the CVL was deployed to image the underlying mantle structure and resolve the source of the volcanism. This two-year PASSCAL deployment consisted of 8 trial stations for the first year and the following year was upgraded to 32 stations with a station spacing of 50 to 200 km. We present preliminary results of surface wave phase velocity maps from this new dataset that successfully image a low-velocity anomaly beneath the CVL as well as high velocities associated with the lithosphere of the Congo Craton. Phase velocity maps were determined by an inversion of teleseismic Rayleigh waves with Forsyth and Lee's two-plane wave method to reduce bias from multipathing. The low-velocity anomaly is observed over a period range of 20 to 140 s suggesting that the CVL anomaly extends across a substantial depth range of the lithosphere and asthenosphere. Furthermore, the lowest velocities of the anomaly are centered at the junction of the 'Y' in the volcanic track, 400 km northeast (inland) of the CVL's largest active volcano Mt. Cameroon.
V33B-1387
Mantle support of the East African Rift System
The African Superplume is a region of slow seismic wave velocities in the lower mantle under southern Africa. The uplift, volcanism and rifting that defines the much of eastern and southern Africa suggest a dynamic link between lower mantle dynamics and near-surface processes affecting the African plate. The dynamic link between the lower mantle and the surface, and the structure and dynamics of the upper mantle below the East African Rift System (EARS) remain unclear. As part of a comprehensive geochemical and numerical investigation of basaltic magmatism in the EARS we have modeled the interaction between putative upper mantle plumes and the rifting continental lithosphere. The modeling provides dynamically tested scenarios that explain the observed episodes of Cenozoic volcanism. Results from recent models that provided an explanation for the present day distribution of volcanism (Lin et al., EPSL, 237, 2005) suggest two plumes below Afar and Tanzania whose uplift is influenced by lithospheric topography. In new 3D modeling we provide improved quantification of the mantle involvement in generating EARS volcanism as constrained by the timing of uplift and regional volcanism. The time scales of episodicity of the volcanism observed at Turkana (related to the Tanzania-Kenya plume) since 45 Ma can be explained by deep- seated time-dependent plume activity. We suggest that this time-dependence is due to thermochemical interactions of dense recycled oceanic crust in the thermally hot regions in the African superplume region (Lin and Van Keken, Nature, 436, 2005).
V33B-1388
Water in Arctic Mid-Ocean Ridge Basalts: Evidence for a Wet Recycled Crustal Component in the Jan Mayen Plume
We present H2O concentrations in fresh mid-ocean ridge basalt (MORB) glasses dredged from 3 contiguous Arctic MAR ridge segments (Kolbeinsey, Mohns, and Knipovich Ridges) to determine the volatile content of the Jan Mayen plume end-member within this low 3He/4He region. Water concentrations (0.15-1.30 wt%) correlate with other indicators of mantle enrichment. H2O/Ce ratios vary from 208 to 428, consistent with high H2O/Ce of North Atlantic basalts (regional means of 240-280±50; Michael, EPSL, 131, 1995), but extend to values higher than previously reported. Previous Pb-Sr-Nd-Hf-H isotope studies of basalts from the Mohns Ridge confirm a strong binary mixing relationship along this ridge segment, grading from the Jan Mayen platform (enriched) eastward to the Greenland FZ where the Mohns Ridge meets the more depleted Knipovich Ridge. We model trace element concentrations in these basalts as binary mixing between depleted mantle and an enriched Jan Mayen component, followed by 7% batch melting of the mixed source using variable mineralogies. Depleted compositions are consistent with melting of a spinel lherzolite mantle, whereas the most enriched Jan Mayen lavas are best modeled by deeper melting of a mantle assemblage richer in clinopyroxene (>50%) and garnet suggesting involvement of a recycled crustal (eclogite) component. Moderately enriched lavas require intermediate mineralogies. Involvement of a recycled crustal component is consistent with the positive Nb and Ta and negative Pb anomalies in enriched samples. Water concentration in the Jan Mayen mantle component is estimated to be ~3000 ppm, significantly higher than previous estimates of water in other mantle plumes. The H2O/Ce of melts derived from melting of this component is ~300. A possible origin for the enriched component is subducted oceanic lithosphere retained in the shallow mantle during formation of the Caledonian suture at ~410-410 Ma as proposed for Iceland by Foulger and others (JVGR, 141, 2005). While high 3He/4He values in Icelandic basalts would appear to preclude such a model for Iceland, the Arctic ridges are characterized by low 3He/4He values, and thus could have a foundered slab source with little or no lower mantle involvement. A foundered slab may have a less dehydrated composition (~88% dehydrated) than those that subduct into deeper mantle (Dixon et al., Nature, 2002). Variable contributions of this "wet" crustal component into North Atlantic mantle may account for the high water concentrations in this region.
V33B-1389
Geochemical variations at a ridge-centered plume caused by variable melting of a veined mantle
Geochemical observations at ridge centered hotspots offer clues into the nature of mantle melting and composition feeding volcanism. For example, observations at Iceland show that lava composition varies with distance along the ridge axis, gradually shifting from relatively incompatible element enriched compositions near the hotspot to relatively depleted compositions far from the hotspot. This observation has been interpreted as indicating mantle compositional heterogeneity below the Mid-Atlantic Ridge, where an "enriched" component is preferentially sampled by hotspot volcanism. Further, an observed gradient in variation along the ridge indicates two end-member scenarios (1) the plume is compositionally distinct from the ambient asthenosphere and the gradients reflect progressive dilution of the plume along the ridge axis or (2) gradients reflect variation in melting of the same heterogeneous source. We model the dynamics and melting of a ridge-centered mantle plume and predict the geochemical composition of magma at the surface. Our model couples flow, heat transfer, and melting of a heterogeneous (veined) mantle that is the same in the plume and ambient mantle. We assume that the heterogeneous mantle comprises two lithologies with different solidi, and that the less refractory lithology has an isotope signature of long-term enrichment in incompatible trace-elements. Our previous work on intraplate plumes reveals that the dynamics of shallow mantle flow and melting of the two lithologies generate magmas with isotope compositions that are very different from those of the initial mantle and lead appreciable geographic patterns in lava composition. Preliminary calculations of a ridge-centered plume predict that the same basic processes can give rise to along-ridge-axis variations in lava compositions with increasing contribution from the enriched lithology towards the center of the plume. This effect is likely augmented by an increase in viscosity due do melting dehydration. Thus variations in melting can induce significant compositional variations at the surface in a plume-ridge setting independent of any geochemical distinctions between plume and ambient asthenosphere.
V33B-1390
What is the threshold plume flux for a persistent plume-fed asthenosphere?
Chao Shi and Jason Phipps Morgan, EAS Dept., Snee Hall, Cornell Univ., Ithaca, NY 14853 In past years we have presented observation evidence that in Earth's mantle convection there exists a buoyant asthenosphere layer fed by upwelling in mantle plumes, and consumed by accretion and transformation into overlying lithosphere by ridge upwelling and melt-extraction (which creates a ~60km-thick layer of compositional lithosphere at mid-ocean ridges), by plate cooling (which accretes a further ~40km of asthenosphere after 100 Ma of near-surface cooling), and by dragdown by subducting slabs (which drags a further ~20km sheet of buoyant asthenosphere on either side of the subducting slab). This scenario has been recently reviewed in Yamamoto et al (GSA Vol. 431). We believe that the reason this mode of mantle convection has not yet been seen in numerical models of mantle convection is due to the inability of current models to model the correct magnitudes of upwelling in focused lower- viscosity plumes (which they currently underpredict) and to correctly model the magnitude of downdragging of a more buoyant but lower viscosity asthenosphere layer (which they currently overpredict, cf. Phipps Morgan et al., Terra Nova, 2007). Here we present results from a suite of 2-D and 3-D calculations that include the effects of ridge accretion, plate cooling and well-resolved asthenosphere dragdown by subducting slabs. In these experiments we do not let mantle plumes spontaneously form at the hot base of the mantle. Instead we extract mantle at a prescribed rate from a single region near the bottom of the mantle (the ‘base of the plume stem') and inject this hot material into the uppermost mantle. The point is to bypass the correct treatment of plume upwelling, in order to explore what upwelling flux is needed to form a persistent plume-fed asthenosphere. We will present results on the plume-flux needed to create a global sub-oceanic plume-fed asthenosphere as a function of: (1) the rate of lithosphere creation and subduction; (2) the buoyancy contrast between asthenosphere and underlying mantle; and (3) the asthenosphere viscosity.
V33B-1391
The European Cenozoic Volcanic Province: The Type Example of an Implausible Plume (IMP)?
The links between plate tectonic processes and magmatism are less obvious in the interiors of plates than at MORs and subduction zones. This has led to simplistic suggestions that intraplate volcanism results from mantle plumes, and extreme pushing of the envelope of plausible variations of the plume model. As the limitations of the mantle plume model have been highlighted over the last few years, considerable interest has arisen in testing the model, and even questioning if mantle plumes exist at all (see http://www.mantleplumes.org/). The plume model is fundamentally a scientific idea, and thus it should make predictions for real cases. These predictions should be borne out at presently proposed plume locations, and should successfully predict what is observed at new locations. In 1972 the plume model was first applied to the cluster of small volcanic areas that form the European Cenozoic Volcanic Province (ECVP) via the so-called Eifel plume. However, radiogenic age data do not support the predicted age progression of the volcanic fields of central Europe. In addition, the geochemical and seismic tomography data, the history of vertical movements and the tectonic evolution of the lithosphere during ECVP formation are all inconsistent with the predictions of the plume model. This proposed mantle plume is one of the most implausible in the world. The proposal that the ECVP is the manifestation of a mantle plume is mainly based on the OIB-like (or "intraplate") geochemistry of the magmas. However, such a geochemical character can tell us nothing regarding depth of origin of the source. This is mainly because geochemical tools cannot differentiate between an in-situ lithospheric source and/or lithospheric material recycling in the sublithosphere. Plumes have often been invoked simply because the erupted basalts have an OIB signature, irrespective of whether the volumes of magma, rates of eruption, or anything else are consistent with that model. Other evidence cited to justify the proposal that one or more plumes underlie the ECVP is from mantle tomography. Low seismic-wave-speed mantle anomalies are detected beneath some ECVP sub-areas, but not beneath another with similar geochemistry. Where the anomalies are seen they do not extend down into the lower mantle. Temperature is not the only physical property that influences seismic velocities. Composition and the presence of partial melt can also lower wave speeds, and these factors are more likely to explain the tomographic structures beneath the ECVP. The whole ECVP is far to small to be classified as a large igneous province that might correspond to a plume head. No primary high-Mg magmas (picrites) consistent with excess temperatures are reported. The He isotopic ratios are much lower that expected for a plume. Lastly, the uplift/volcanism history of the area is not as predicted for mantle plume lithospheric interactions. The ECVP contradicts essentially all the foundation predictions of the mantle plume model. ECVP volcanism is more likely linked to Alpine subduction processes that affect stress, deformation and flow in the European continental crust and underlying shallow mantle, coupled with the influence of local lithospheric conditions. RM is funded by BFR 05/133 http://www.mantleplumes.org/Europe.html
V33B-1392
Volcanism of Nanpu Sag in the Bohai Bay Basin: Geochemistry, Origin and Tectonic Implications
The recently discovered large Nanpu Oilfield is located in the NW part of the Bohai Bay Basin, eastern China. Extensive drilling and geophysical research have revealed widespread basaltic volcanic rocks of Cenozoic age. On the basis of stratigraphic constraints six cycles of vocanism have been recognised. Petrological and geochemical studies of drill core samples indicate that the basalts belong to the alkaline series. Some basaltic rocks experienced fractional crystallization, dominated by clinopyroxene. All samples examimed are enriched in high field strength elements and have trace elements patterns similar to oceanic island basalts (OIB). High ratios of La/Yb suggest that significant fractionation of rare earth element has occurred. The Sr and Nd isotopic system shows that 87Sr/86Sri range from 0.7036 to 0.7082 and ¦ÅNd(t) vary from -2.336 to 3.564. Measured Pb isotopes range from 17.434 to 18.354 for 206Pb/204Pb, from 15.43 to 15.609 for 207Pb/204Pb and from 37.352 to 38.454 for 208Pb/204Pb, respectively. These geochemical and isotopic characteristics indicate that they are similar to oceanic island basalts and were generated from a within-plate setting. These basalts were formed by low degrees of partial melting from a hybrid mantle of EMI and DMM types, with no significant crustal assimilation. Considering that there is widespread and coeval basaltic volcanism in the Bohai Bay Basin, with comparable geochemistry, we suggest a mantle plume model for the origin of these basalts. We argue that a mantle plume was activated since the Cretaceous, which resulted in uplift and rifting in eastern China. Continuing ascending movement of the mantle plume induced multi-stage mantle upwelling and melting. This process caused crustal uplifting and thinning, resulting in the inception of the Bohai Bay basin, in which a succession of interbedded volcanic and sedimentary rocks accumulated
V33B-1393
Supertoxic Flood Basalts: The CAMP – Siberian Trap Connection
Several diverse magma types are represented throughout the CAMP and Siberian Trap LIPs, however, the main extrusive phase of each province is highly unusual among continental flood basalts. The most widespread extrusions were intermediate titanium (ITi-type) CAMP basalt and the lower portion of the Upper Sequence of Siberian Trap. New and recently published data indicate that the geochemistry and petrology of these basalt suites closely resemble each other and infer similar origins. The basalts are characterized by strong negative Nb- Ta anomalies and unusual island arc-like depletion in high field strength elements, particularly Ti, plotted on spider diagrams. The geochemical data is consistent with significant contributions from subducted slabs into the magma source regions. If contaminated, volatile enriched mantle wedges were trapped beneath thick continental plates during the assembly of Pangea, fertile magma sources would have remained dormant until decompression melting was triggered during failed rift, then early rift stages of continental plate disassembly. The combination of volatile enriched sources and highly extensional tectonism would create rare perfect storms of toxicity. Calculated low viscosities assuming negligible carbon dioxide are consistent with rapid crustal penetration. Resulting aphyric melts extruded at enormous effusive rates as thick sub-parallel flows across wide subareal terrains through fissures extending several hundred km in length. High fountain heights would afford ample opportunity for efficient degassing, perhaps into the stratosphere. When the supply of volatile flux was exhausted magmatism ceased. The mass extinctions that coincide with CAMP and Siberian volcanism contrast with some large plume and superplume events that correlate with expansions of biodiversity. This may be due in part to contrasting magma access to sources of toxic volatiles, particularly sulfur concentrations in anoxic subducted sediments.
V33B-1394
A Revised Geochemical Grouping of Gondwana LIP: Distinctive Sources and Processes at the Weddell and Limpopo Triple Junctions
The magma types belonging to the Gondwana large igneous province can be divided into two categories based on, respectively, their primitive mantle-like and fractionated Sm/Yb and Sr/Zr values, different Sr and Nd isotopic trends, and geographic affinity to the Weddell and Limpopo triple junctions. In the new grouping, the Ferrar magmas, the Karoo Central Area magmas, and the Kirwanveggen-Sembberget magmas from Dronning Maud Land are viewed as three major magmatic lineages generated at the Weddell triple junction. These Weddell group magmas were produced by voluminous low-pressure partial melting of possibly subduction-modified upper mantle and tended to be laterally transported over long distances. The Limpopo group magmas include the Karoo high-Ti magma types and various low-Ti types from Lebombo (Karoo), Vestfjella, and Heimefrontfjella (Dronning Maud Land). They are likely to represent magmas that were produced at high pressure from an eclogite-bearing mantle source below the Kaapvaal craton. The distribution of the Limpopo group magmas was mainly confined within the rift valleys of the Limpopo triple junction. The chemically distinctive magma types of the Gondwana LIP probably reflect heterogeneities in the Weddell and Limpopo magma sources and lithospheric level differentiation within Archean (Limpopo group) and younger lithospheric terranes.
V33B-1395
Geochemistry of Deccan Traps Dikes: Insights Into the Evolution of a Flood Basalt Feeder System
Three large dike swarms are exposed in the 500,000 km2 Deccan Traps of India: the dominantly N-S trending West Coast swarm, the ENE-WSW trending Narmada-Tapi swarm in the northern Deccan, and the Nasik-Pune swarm in the central western Deccan. Dikes of the Nasik-Pune swarm show no strongly preferred trend. This swarm is commonly postulated (e.g., Hooper, Nature, 349, 246, 1990) to be the principal locus of feeders for the lava pile, and the lack of a preferred trend taken as evidence that the flood volcanism was not accompanied by significant rifting-related lithospheric extension. Our combined major and trace element and Pb-Nd-Sr isotope data reveal that dikes with signatures matching those of the three major lava formations in the upper part of the lava pile (Poladpur, Ambenali, and Mahabaleshwar formations) are abundant in the coastal and Nasik-Pune swarms. As a group, these dikes have no preferred trend. Dikes with similarities to formations lower in the lava stratigraphy (i.e., the Igatpuri, Jawhar, and Bushe formations) are present but rare in these two swarms. However, many dikes with strong affinities to the lower and middle lava formations (e.g., Igatpuri, Jawhar, Bushe, and Thakurvadi formations) are present in the Narmada-Tapi swarm. These dikes have geometries indicative of N-S extension. We infer that rifting did not drive emplacement of the upper lava formations, but was occurring in the Narmada-Tapi region during the earlier phases of volcanism. Thus, N-S rifting cannot be ruled out as a trigger mechanism for the massive melting event. In contrast, our data provide no evidence that E-W extension along the coast triggered the event.
V33B-1396
Isotope-heterogeneity of the monogenetic Thjorsa lava eruption, Iceland – source and crustal influence
The Thjorsa lava in the Eastern Rift Zone (ERZ) in Southern Iceland erupted 8600 years ago. It is the largest Holocene fissure lava in Iceland with a total volume of 25 km3. It is one of several highly plagioclase-porphyritic tholeiitic fissure eruptions from the early Holocene. These porphyritic lavas belong to the Bardarbunga volcanic system which forms the highly productive eastern edge of the South Iceland microplate: The plagioclase content ranges from 4-30 vol% - about 7 vol% on average. The origin of the large plagioclase phenocrysts is uncertain, i.e. are they crystallized from the host magma or its precursors, or are they accidental xenocrysts. In this study, elemental and isotope (Sr, Nd and Pb) analyses were made on a suite of groundmass samples, and 87Sr/86Sr ratio in plagioclase/groundmass pairs. Samples are from various parts of the Thjorsa lava – including a number of samples from drill cores. Additional whole rock/glass samples from the Bardarbunga system lavas are also included for comparison. The host lava groundmass (i.e. the liquid phase) has a restricted variation in elemental and isotopic composition. The plagioclase separates show a range of 87Sr/86Sr isotopic values, but with hardly any overlap with the groundmass. The isotopic heterogeneity of the plagioclase phenocrysts covers almost the entire variation observed within the Bardarbunga spreading center during the Holocene. The plagioclases are therefore xenocrysts and significantly more isotopically primitive than the host groundmass, and derived from a different mantle source. Correlation between of cpx bearing elements (Cr/Y, Sc/Y) and isotopic ratios of Sr, Nd and Pb demonstrates addition of remelted cpx in the groundmass. Relics of Cr-rich cpx microliths and Cr- diopside xenoliths are assumed to be the leftovers and examples of the digested cpx. The ERZ is about 35 km wide with an array of sub-parallel fissure swarms and eruptive lineaments. The age of the ERZ is estimated at 2- 3 mill. years, and thus needs to accommodate rifting of similar width. Ascending magmas will, therefore, frequently penetrate the roots of earlier spreading swarms and possibly older crust. The large range of the isotope signature of the plagioclase xenocrysts indicates that they are formed by multiple intrusion events. These results indicate large-scale and rapid remobilization of crustal material by disintegration and partial assimilation of plutonic rocks, which significantly influenced the isotopic composition of the erupted lava. The mineralogical and geochemical character of this hybrid rock and its heterogeneous components sheds light on the origin of scattered chemical and isotopic data from Icelandic rift zone basalts.
V33B-1397
Permian Large Volume Basalt In Tarim Basin Of China And Its Geological Implication
The Tarim basin is the biggest sedimentary basin in China with an area of 560,000 km2, and is surrounded by the Tianshan, Kunlun and Altun orogenic belts. It consists mainly of Precambrian basement and Phanerozoic stratum, and has recorded a magmatic history at ages ranging from 2500 Ma to 270 Ma. Based on geological studies and oil-gas surveys, the total area covered by the Permian basic igneous rocks is estimated to be more than 200,000 km2. The Permian rocks are mainly composed of basalts, with a maximum thickness of ca. 600m, and diabases, and some occurrences of basaltic andesites and ultramafic rocks. The field survey and drilling core data show that the Tarim basin has experienced a major tectonothermal event in the Permian, with strong imprints in the central and western parts of the basin. An important unanswered question is whether the Tarim Permian magmatism is associated with the Permian mafic magmatic events from the surrounding regions, such as the 258 Ma Emeishan large igneous province (ELIP) in SW China. The goals of this study are to constrain precisely the time of formation of the Tarim Permian basalts using 40Ar-39Ar ages, and to discuss its origin using geochemical and isotopic tools. Using our new 40Ar-39Ar dating combined with other age data and the paleontological and sedimentary evidences, we estimate a time interval for the Permian basalt formation ranging from 280 to 290 Ma. In detail, distinct ages have been obtained for the different units, i.e. 290 Ma and 282 Ma for Kupukuziman (lower section) and Kaipaizilaike (upper section) basalts, respectively. Our study reveals that the different rock suites belonging to the Permian volcanic suites represent an igneous activity which spanned for about 10 Ma ranging from 280 to 290 Ma. We suggest that the timing of these rocks is intimately related to a lastly widespread and large thermo-tectonic event in the Tarim basin. The Permian basalts and diabases have compositions of tholeiite and alkali basalt series, and display high K, Rb, Ba, Th, LILE and LREE concentrations with no Nb-Ta anomalies. The basalts from the Yigan section of NW Tarim have initial 143Nd/144Nd of 0.5120-0.5122, initial 87Sr/86Sr of 0.7068-0.7077, 206Pb/204Pb of 17.90-18.05, 207Pb/204Pb of 15.52-15.53, and 208Pb/204Pb of 38.33-38.73. The trace element contents and Sr-Nd-Pb isotopic ratios suggest that the basalts are derived from partial melting under an extensional setting of an enriched mantle reservoir, likely related to a mantle plume as explained below. In addition, the trace element contents and Sr-Nd-Pb isotopic ratios are similar to those of the ELIP, suggesting a genetic linkage between them. Based on the new data from the Tarim event, a genetic link between the Tarim basalt (285 Ma) and ELIP (258 Ma) is that the existence of two short magmatic pulses at 285 and 258 Ma may indicate two discrete events in the source region, and two discrete plume/LIP events originated as discrete plumes arising from the deep mantle.