Louisville Hotspot Research Articles

Four distinct pulses of volcanism built the Melanesian Border Plateau: Implications for oceanic mid-plate superstructure formation

The ocean basins contain numerous volcanic ridges, seamounts and large igneous provinces (LIPs). Numerous studies have focused on the origin of seamount chains and LIPs but much less focus has been applied to understanding the genesis of large volcanic structures formed from a combination or series of volcanic drivers. Here we propose the term Oceanic Mid-plate Superstructures (OMS) to describe independent bathymetric swells or volcanic structures that are constructed through superimposing pulses of volcanism, over long time periods and from multiple sources. These sources can represent periods when the lithosphere drifted over different mantle plumes and/or experienced pulses of volcanism associated with shallow tectonic drivers (e.g. plate flexure; lithospheric extension). Here we focus on the Melanesian Border Plateau (MBP), one example of an OMS that has a complex and enigmatic origin.The MBP is a region of shallow Pacific lithosphere consisting of high volumes of volcanic guyots, ridges and seamounts that resides on the northern edge of the Vitiaz Lineament. Here we reconcile recently published constraints to build a comprehensive volcanic history of the MBP. The MBP was built through four distinct episodes: (1) Volcanism associated with the Louisville hotspot likely generating Robbie Ridge and some Cretaceous seamounts near the MBP. (2) Construction of oceanic islands and seamounts during the Eocene when the lithosphere passed over the Rurutu-Arago hotspot. (3) Reactivation of previous oceanic islands/seamounts and construction of new volcanos in the Miocene when the lithosphere passed over the Samoa hotspot. (4) Miocene to modern volcanism driven by lithospheric deformation and/or westward entrainment of enriched plume mantle due to toroidal mantle flow driven by the rollback of the Pacific plate beneath the Tonga trench. The combination of these processes is responsible for ∼222,000 km2 of intraplate volcanism in the MBP and indicates that this OMS was constructed from multiple volcanic drivers.

Temporal Evolution of the Swell and Magmatic Fluxes Along the Louisville Hotspot Chain

AbstractHawaii and Louisville are the two long‐lived hotspot chains in the Pacific basin. Characterizing the swell flux, Qs, a measure of the plume buoyancy, and the flux of the magmatic material accumulated along the chain, Qm, brings important constraints on the plumes' temporal evolution and on their interaction with the lithosphere. Numerous studies focus on the temporal evolution of the Hawaiian chain, but no study investigates the temporal evolution along the Louisville chain. In the present study we constrain the temporal evolution of the swell and magmatism fluxes along the Louisville chain, from volcano ages and bathymetry data. The overall Louisville magmatism flux appears nearly constant, that is, Louisville plume activity has exhibited remarkably little variation over the past 30 m.y., in agreement with geochemical data. These results are also in agreement with recent studies along Hawaii, which show that the Qm increase is due to plume migration toward a more fertile mantle reservoir. This means that, in general, long term variations in Qm relate to variations in the deep source of the plume. Along Louisville, we find short‐term variations occurring with a periodicity of ∼5–10 m.y. in both Qm and Qs, similar to the short‐term variations observed along Hawaii by former studies. As these variations occur along hotspot chains associated with either a heterogenous or homogeneous chemical signature, they must be created by shallow phenomena. The tilt of the plume conduit by large‐scale mantle advection is our preferred model, as it can account for both the geophysical and geochemical observations.

Quantification of Pacific Plate Hotspot Tracks Since 80 Ma

AbstractThe motion of the Pacific plate relative to Pacific hotspots produces age‐progressive chains of volcanoes. Methods of analysis of volcano locations and age dates using a small number of adjustable parameters (10 per chain) are presented. Simple fits to age progressions along Pacific hotspot chains indicate 1σ dispersions of age dates of ≈±1.0–±3.0 Ma. Motion between the Hawaii and Louisville hotspots differs insignificantly from zero with rates of 2 ± 4 mm/a (=±2σ) for 0–48 Ma and 26 ± 34 mm/a (=±2σ) for 48–80 Ma. Relative to a mean Pacific hotspot reference frame, motions of the Hawaii, Louisville, and Rurutu hotspots are also insignificant. Therefore plumes underlying these Pacific hotspots may be more stable in a convecting mantle than previously inferred. We find no significant difference in age between the Eocene bends of the Pacific hotspot chains. The best‐fitting assumed‐coeval age for the bends is 47.4 ± 1.0 Ma (=±2σ), coincident with the initiation of the doubling of the spreading rate of the Pacific plate relative to the Farallon and Vancouver plates. The initiation of the Eocene collision of India with Eurasia preceded the formation of the bends and was completed after their formation. Initiation of subduction of the Pacific plate in the west and southwest Pacific Ocean Basin likely preceded the formation of the bends, consistent with subduction initiation changing the torque on the Pacific plate such that it started moving in a more westward direction thus creating the Hawaiian‐Emperor Bend.

An eclogitic component in the Pitcairn mantle plume: Evidence from olivine compositions and Fe isotopes of basalts

Subducted oceanic crust can transform into eclogite in the upper mantle, and generate chemical heterogeneity of mantle plumes, as recorded by elemental and radiogenic isotopic variations in oceanic island basalts (OIBs). The secondary pyroxenite produced by the reaction between eclogite-derived melt and peridotite is increasingly regarded as a major source component of OIBs, as well as peridotite. However, it remains unclear whether eclogite can be a direct source component of OIBs. To test this possibility, we present high-precision whole-rock Fe isotopes and the chemical compositions of olivine phenocrysts from well-characterized EM1-type basalts from Pitcairn Island. The Pitcairn basalts are characterized by moderate 87Sr/86Sr, low 143Nd/144Nd and 206Pb/204Pb isotopic ratios, and lowest δ26Mg values among OIBs, suggesting a contribution from recycled ancient crustal components (oceanic crust plus sediment). For comparison, we also report the Fe isotope compositions of FOZO-type basalts (with low 87Sr/86Sr, moderately high 143Nd/144Nd and moderate 206Pb/204Pb ratios) from the Louisville hot spot track, which were suggested to be a typical peridotite-derived OIB. The Louisville basalts have MORB-like δ57Fe values (0.06‰–0.15‰), whereas the Pitcairn lavas have substantially heavier Fe isotopic compositions (δ57Fe = 0.17‰–0.31‰) than MORBs. Quantitative evaluations suggest that magmatic differentiation, partial melting, and elevated oxygen fugacity in the source are insufficient to generate the heavy Fe isotopic compositions of the Pitcairn basalts.A good correlation between δ57Fe and εNd(i) (or δ26Mg) values in the Pitcairn basalts indicates binary mixing, and the melts derived from the EM1 endmember have unusually heavy Fe and light Mg isotopic compositions. In order to explain the origin of the Fe and Mg isotopic compositions, we calculated the Fe, Mg, and Nd isotopic compositions of partial melts of eclogite, secondary pyroxenite and peridotite, respectively. The results indicate that eclogite is the only suitable candidate to generate melts with both heavy Fe and light Mg isotopes. This inference is strengthened by the major and minor elemental compositions of olivine phenocrysts from the Pitcairn basalts, which show low Fo (73.2–82.5) and Ni contents (620–1949 ppm), and low Mn/Fe (0.011–0.015) and Ni/(Mg/Fe) (524–1041) ratios. These characteristics are markedly different from those of olivines from MORBs and the Koolau lavas from Hawaii, which represent phenocrysts that equilibrated with peridotite-derived and secondary pyroxenite-derived melts, respectively. We therefore argue that eclogite is the source lithology of the EM1 endmember of the Pitcairn basalts. Binary mixing between our modelled eclogite- and peridotite-derived melts produced magmas with relatively low Mg# value, Mg/Fe ratios and moderate Ni contents characteristics, which are preserved in the low-Fo Pitcairn olivines. Our results highlight that, in addition to peridotite and secondary pyroxenite, eclogite may survive in mantle plumes at shallow depths and make a substantial contribution to the source of OIBs.

High‐Resolution 40Ar/39Ar Geochronology of the Louisville Seamounts IODP Expedition 330 Drill Sites: Implications for the Duration of Hot Spot‐related Volcanism and Age Progressions

AbstractThe Louisville Seamounts display age progressive volcanism and are thought to have formed as the Pacific Plate moved over a long‐lived primary hot spot. Here we present 70 new 40Ar/39Ar age results from the Integrated Ocean Drilling Program (IODP) Sites U1372, U1375, U1376, and U1377 drilled and cored during Expedition 330 to the northern, older end of the Louisville Seamounts. The five seamounts drilled are flat‐topped guyots with ages ranging from ∼74 Ma (Canopus Guyot) to ∼51 Ma (Hadar Guyot) recovering up to ∼510 m of basaltic material beneath thin sediment interfaces. Our 40Ar/39Ar measurements reveal that throughout each drill hole, most dates are constant within ∼500‐900 kyr at the 2σ confidence interval. In this study we use the new Louisville age information to compare against the Hawaiian‐Emperor trail on the Pacific Plate. The Louisville hot spot trail is low volume, dominantly alkali basalt, and seamounts have a relatively short lifespan up to ∼4 Myr, whereas the Hawaiian hot spot trail is high volume, has a tholeiitic shield‐building stage capped by an alkalic post‐shield stage, and has a lifespan up to ∼6‐7 Myr. Here we show a new approach to estimating a seamount's inception age based on the known cumulative age distributions for seamounts in the Louisville Seamount and Hawaii‐Emperor Seamount trails. Based on our new 40Ar/39Ar analyses, we conclude that existing absolute plate motion models misrepresent the age progression of Louisville Seamounts and that the timing of the Hawaiian‐Emperor Bend and the 169°W bend in the Louisville Seamounts are asynchronous by ~3.7 Myr.

Flexure and gravity anomalies of the oceanic lithosphere beneath the Louisville seamount

We have calculated the elastic thickness (Te), flexural deflection, and gravity anomaly of the oceanic crust beneath the Louisville seamount (LSC-03), near the Kermadec trench. A regional-residual separation of the bathymetry was performed to remove the effect of other geologic features (e.g., the trench). We used the uniform density and dense core models to approximate the total mass of the seamount, which was defined as the surface load required for flexural deformation. From the flexure modeling results, we found that more flexural depression was predicted by the uniform density model than by the dense core model. However, the uniform density model predicted a significantly smaller gravity anomaly than observed, whereas the dense core model minimized the prediction misfits reasonably. The best flexure model was found with a Te of 16km for the uniform density model and 6km for the dense core model. The flexure computed with the dense core model was consistent with the seismically detected Moho. The flexure modeling for LSC-03, thus, indicates that the dense core model better approximates the inner structure of the LSC-03. Based on the crustal age and geochronology of the given seamount, the age of the oceanic crust at the time of seamount formation (Δt) is 20Ma. If this is the case, however, the Te estimates from both flexure models require some degree of lithospheric reheating by Louisville hotspot activity. Alternatively, considering the tectonic plate motion of the Osbourn Trough, Δt becomes approximately 4Ma. This younger lithosphere model is more consistent with the observed flexural deformation and the Te estimate from the dense core model. Therefore, the time that the seamount-induced lithospheric deformation occurred may be far earlier than the age-dated volcanism.

A rapid burst in hotspot motion through the interaction of tectonics and deep mantle flow.

Volcanic hotspot tracks featuring linear progressions in the age of volcanism are typical surface expressions of plate tectonic movement on top of narrow plumes of hot material within Earth's mantle. Seismic imaging reveals that these plumes can be of deep origin--probably rooted on thermochemical structures in the lower mantle. Although palaeomagnetic and radiometric age data suggest that mantle flow can advect plume conduits laterally, the flow dynamics underlying the formation of the sharp bend occurring only in the Hawaiian-Emperor hotspot track in the Pacific Ocean remains enigmatic. Here we present palaeogeographically constrained numerical models of thermochemical convection and demonstrate that flow in the deep lower mantle under the north Pacific was anomalously vigorous between 100 million years ago and 50 million years ago as a consequence of long-lasting subduction systems, unlike those in the south Pacific. These models show a sharp bend in the Hawaiian-Emperor hotspot track arising from the interplay of plume tilt and the lateral advection of plume sources. The different trajectories of the Hawaiian and Louisville hotspot tracks arise from asymmetric deformation of thermochemical structures under the Pacific between 100 million years ago and 50 million years ago. This asymmetric deformation waned just before the Hawaiian-Emperor bend developed, owing to flow in the deepest lower mantle associated with slab descent in the north and south Pacific.

Dynamics and excess temperature of a plume throughout its life cycle

Measurements of the velocity field associated with plumes rising through a viscous fluid are performed using stereoscopic Particle-Image Velocimetry in the Rayleigh number range 4.4 × 105 − 6.4 × 105. The experimental model is analogous to a mantle plume rising from the core-mantle boundary to the base of the lithosphere. The behaviour of the plume is studied throughout its life cycle, which is broken up into four stages; the Formation Stage, when the plume forms; the Rising Stage, when the plume rises through the fluid; the Spreading Stage, when the plume reaches the surface and spreads; and finally the Declining Stage, when the heat source has been removed and the plume weakens. The latter three stages are examined in terms of the Finite-Time Lyapunov Exponent fields and the advection of passive tracers throughout the flow. The temperature at the heater and near the fluid surface are measured using thermocouples to infer how the presence of a mantle plume would produce excess temperature near the lithosphere throughout the various stages of its life cycle. In all experiments a time lag is observed between the removal of the heat source and the decline in the excess temperature near the surface, which is proportional to the rise time. A simple analytical model is presented, which suggests that under mantle conditions (i.e. negligible thermal diffusion), the relationship between the time lag and the rise time is robust and independent of the Rayleigh number; however, the constant of proportionality is closer to unity in the absence of diffusion. Once the heat source is removed, the excess temperature near the surface declines exponentially at a rate that is inversely proportional to the rise time. The implications of this result are discussed in terms of the decline in volcanism in the Louisville hotspot chain over the past 20 Ma. The rise velocity of material in the plume is examined; the rise velocity is found to vary significantly with the plume height in a manner that is inconsistent with many of the common semi-analytical models of thermal plumes in the literature. It is also argued that this height-dependency will cause estimates of the rise velocity based on the decay series of Uranium isotopes to significantly underestimate the true value.

Louisville Seamount Chain: Petrogenetic processes and geochemical evolution of the mantle source

AbstractThe Louisville Seamount Chain is a ∼4300 km long chain of submarine volcanoes in the southwestern Pacific that spans an age range comparable to that of the Hawaiian‐Emperor chain and is commonly thought to represent a hot spot track. Dredging in 2006 recovered igneous rocks from 33 stations on 22 seamounts covering some 49 Myr of the chain's history. All samples are alkalic, similar to previous dredge and drill samples, providing no evidence for a Hawaiian‐type tholeiitic shield‐volcano stage. Major and trace element variations appear to be predominantly controlled by small but variable extents of fractional crystallization and by partial melting. Isotopic values define only a narrow range, in agreement with a surprising long‐term source homogeneity—relative to the length scale of melting—and overlap with proposed fields for the “C” and “FOZO” mantle end‐members. Trace element and isotope geochemistry is uncorrelated with either seamount age or lithospheric thickness at the time of volcanism, except for a small number of lavas from the westernmost Louisville Seamounts built on young (<20 Ma old) oceanic crust. The Louisville hot spot has been postulated to be the source of the ∼120 Ma Ontong Java Plateau, but the Louisville isotopic signature cannot have evolved from a source with isotopic ratios like those measured for Ontong Java Plateau basalts. On the other hand, this signature can be correlated with that of samples dredged from the Danger Islands Troughs of the Manihiki Plateau, which has been interpreted as a rifted fragment of the “Greater” Ontong Java Plateau.

Deep plume origin of the Louisville hotspot: Noble gas evidence

AbstractNoble gas compositions have been reported for basaltic core samples from Louisville seamounts recovered during IODP Expedition 330. The in‐vacuum crushing techniques were employed to extract noble gases from fresh olivine phenocrysts and submarine glasses with ages between 50 and 74 Ma. Stepwise crushing tests confirmed the extraction of magmatic noble gases from the olivine samples with minimal release of posteruption radiogenic nuclides; however, this was not always the case for the glass samples. The 3He/4He ratios of the studied samples range from a value similar to those of mid‐ocean ridge basalts (MORB) to slightly elevated ratios up to 10.6 Ra. These ratios are not as high as those observed in other ocean island basalts, suggesting that the Louisville mantle plume was weak or the samples represent late‐stage magmatic activity of the seamounts. However, two Louisville seamount basalts exhibit a primordial Ne isotopic signature that can be clearly discriminated from MORB Ne ratios. The He and Ne isotopic compositions of the Louisville seamount basalts can be explained by the mixing of less degassed mantle and depleted upper mantle with different He/Ne ratios. The presence of the less degassed mantle component in the source of the Louisville seamounts documents a deep origin of their mantle plume.

Constraints on past plate and mantle motion from new ages for the Hawaiian‐Emperor Seamount Chain

Estimates of the relative motion between the Hawaiian and Louisville hot spots have consequences for understanding the role and character of deep Pacific‐mantle return flow. The relative motion between these primary hot spots can be inferred by comparing the age records for their seamount trails. We report 40Ar/39Ar ages for 18 lavas from 10 seamounts along the Hawaiian‐Emperor Seamount Chain (HESC), showing that volcanism started in the sharp portion of the Hawaiian‐Emperor Bend (HEB) at ≥47.5 Ma and continued for ≥5 Myr. The slope of the along‐track distance from the currently active Hawaiian hot spot plotted versus age is constant (57 ± 2 km/Myr) between ∼57 and 25 Ma in the central ∼1900 km of the seamount chain, including the HEB. This model predicts an age for the oldest Emperor Seamounts that matches published ages, implying that a linear age‐distance relationship might extend back to at least 82 Ma. In contrast, Hawaiian age progression was much faster since at least ∼15 Ma and possibly as early as ∼27 Ma. Linear age‐distance relations for the Hawaii‐Emperor and Louisville seamount chains predict ∼300 km overall hot spot relative motion between 80 and 47.5 Ma, in broad agreement with numerical models of plumes in a convecting mantle, and paleomagnetic data. We show that a change in hot spot relative motion may also have occurred between ∼55 Ma and ∼50 Ma. We interpret this change in hot spot motion as evidence that the HEB reflects a combination of hot spot and plate motion changes driven by the same plate/mantle reorganization.

Scientific Drilling

Deep-Earth convection can be understood by studying hotspot volcanoes that form where mantle plumes rise up and intersect the lithosphere, the Earth's rigid outer layer. Hotspots characteristically leave age-progressive trails of volcanoes and seamounts on top of oceanic lithosphere, which in turn allow us to decipher the motion of these plates relative to deep-mantle plumes, and their (isotope) geochemistry provides insights into the long-term evolution of mantle source regions. However, it is strongly suggested that the Hawaiian mantle plume moved approximately 15 degrees south between 80 and 50 million years ago. This raises a fundamental question about other hotspot systems in the Pacific, whether or not their mantle plumes experienced a similar amount and direction of motion. Integrated Ocean Drilling Program (IODP) Expedition 330 to the Louisville Seamounts showed that the Louisville hotspot in the South Pacific behaved in a different manner, as its mantle plume remained more or less fixed around 48 degrees S latitude during that same time period. Our findings demonstrate that the Pacific hotspots move independently and that their trajectories may be controlled by differences in subduction zone geometry. Additionally, shipboard geochemistry data shows that, in contrast to Hawaiian volcanoes, the construction of the Louisville Seamounts doesn't involve a shield-building phase dominated by tholeiitic lavas, and trace elements confirm the rather homogenous nature of the Louisville mantle source. Both observations set Louisville apart from the Hawaiian-Emperor seamount trail, whereby the latter has been erupting abundant tholeiites (characteristically up to 95% in volume) and which exhibit a large variability in (isotope) geochemistry and their mantle source

IODP Expedition 330: Drilling the Louisville Seamount Trail in the SW Pacific

Deep-Earth convection can be understood by studying hotspot volcanoes that form where mantle plumes rise up and intersect the lithosphere, the Earth's rigid outer layer. Hotspots characteristically leave age-progressive trails of volcanoes and seamounts on top of oceanic lithosphere, which in turn allow us to decipher the motion of these plates relative to "fixed" deep-mantle plumes, and their (isotope) geochemistry provides insights into the long-term evolution of mantle source regions. However, it is strongly suggested that the Hawaiian mantle plume moved ~15° south between 80 and 50 million years ago. This raises a fundamental question about other hotspot systems in the Pacific, whether or not their mantle plumes experienced a similar amount and direction of motion. Integrated Ocean Drilling Program (IODP) Expedition 330 to the Louisville Seamounts showed that the Louisville hotspot in the South Pacific behaved in a different manner, as its mantle plume remained more or less fixed around 48°S latitude during that same time period. Our findings demonstrate that the Pacific hotspots move independently and that their trajectories may be controlled by differences in subduction zone geometry. Additionally, shipboard geochemistry data shows that, in contrast to Hawaiian volcanoes, the construction of the Louisville Seamounts doesn’t involve a shield-building phase dominated by tholeiitic lavas, and trace elements confirm the rather homogenous nature of the Louisville mantle source. Both observations set Louisville apart from the Hawaiian-Emperor seamount trail, whereby the latter has been erupting abundant tholeiites (characteristically up to 95% in volume) and which exhibit a large variability in (isotope) geochemistry and their mantle source components. <br><br> doi:<a href="http://dx.doi.org/10.2204/iodp.sd.15.02.2013" target="_blank">10.2204/iodp.sd.15.02.2013</a>

Limited latitudinal mantle plume motion for the Louisville hotspot

Hotspots that form above upwelling plumes of hot material from the deep mantle typically leave narrow trails of volcanic seamounts as a tectonic plate moves over their location. These seamount trails are excellent recorders of Earth’s deep processes and allow us to untangle ancient mantle plume motions. During ascent it is likely that mantle plumes are pushed away from their vertical upwelling trajectories by mantle convection forces. It has been proposed that a large-scale lateral displacement, termed the mantle wind, existed in the Pacific between about 80 and 50 million years ago, and shifted the Hawaiian mantle plume southwards by about 15° of latitude. Here we use 40Ar/39Ar age dating and palaeomagnetic inclination data from four seamounts associated with the Louisville hotspot in the South Pacific Ocean to show that this hotspot has been relatively stable in terms of its location. Specifically, the Louisville hotspot—the southern hemisphere counterpart of Hawai’i—has remained within 3–5° of its present-day latitude of about 51° S between 70 and 50 million years ago. Although we cannot exclude a more significant southward motion before that time, we suggest that the Louisville and Hawaiian hotspots are moving independently, and not as part of a large-scale mantle wind in the Pacific.

Reconstructing Ontong Java Nui: Implications for Pacific absolute plate motion, hotspot drift and true polar wander

The Taylor (2006) hypothesis suggesting a common origin for the Ontong Java, Manihiki, and Hikurangi large igneous provinces provides an opportunity for a quantitative reconstruction and reassessment of the Ontong Java–Louisville hotspot connection. Our plate tectonic reconstructions of the three plateaus into Ontong Java Nui, or greater Ontong Java, combined with models for Pacific absolute plate motion (APM), allow an analysis of this connection. A new survey of the central Ellice Basin confirms easterly fracture zones, northerly abyssal hill fabric, as well as an area of sigmoidally-southeast-trending fracture zones associated with a late-stage spreading reorientation. From the fracture zone trends we derive new rotation poles for a two-stage model of Ellice Basin opening between the Ontong Java and Manihiki Plateaus. We use these and a single stage pole for separation of the Manihiki and Hikurangi Plateaus, together with three different Pacific APMs, to reconstruct the Ontong Java Nui super plateau back to 123Ma and compare its predicted location with paleolatitude data obtained from the Ontong Java and Manihiki plateaus. Discrepancies between our Ontong Java Nui reconstructions and Ontong Java and Manihiki paleolatitudes are largest for the fixed Pacific hotspot APM. Assuming a Louisville hotspot source for Ontong Java Nui, remaining disparity between Ontong Java Nui's paleo-location at 123Ma and published paleomagnetic latitudes for Ontong Java plateau imply that 8°–19° of Louisville hotspot drift or true polar wander may have occurred since the formation of Ontong Java Nui. However, the older portions of the Pacific APMs could easily be biased by a similar amount, making a firm identification of the dominant source of misfit difficult. Prior studies required a combined 26° of hotspot drift, octupole bias effects, and true polar wander just to link the Ontong Java Plateau to Louisville. Consequently, we suggest the super plateau hypothesis and our new reconstructions have considerably strengthened the case for a Louisville plume origin for Ontong Java Nui.

New40Ar/39Ar age progression for the Louisville hot spot trail and implications for inter-hot spot motion

[1] In this study we present 42 new 40Ar/39Ar incremental heating age determinations that contribute to an updated age progression for the Louisville seamount trail. Louisville is the South Pacific counterpart to the Hawaiian-Emperor seamount trail, both trails representing intraplate volcanism over the same time interval (∼80 Ma to present) and being examples of primary hot spot lineaments. Our data provide evidence for an age-progressive trend from 71 to 21 Ma. Assuming fixed hot spots, this makes possible a direct comparison to the Hawaiian-Emperor age progression and the most recent absolute plate motion (APM) model (WK08G) of Wessel and Kroenke (2008). We observe that for the Louisville seamount trail the measured ages are systematically older relative to both the WK08G model predictions and Hawaiian seamount ages, with offsets ranging up to 6 Myr. Taking into account the uncertainty about the duration of eruption and magmatic succession at individual Louisville volcanoes, these age offsets should be considered minimum estimates, as our sampling probably tended to recover the youngest lava flows. These large deviations point to either a contribution of inter–hot spot motion between the Louisville and Hawaiian hot spots or to a more easterly location of the Louisville hot spot than the one inferred in the WK08G model. Both scenarios are investigated in this paper, whereby the more eastern hot spot location (52.0°S, 134.5°W versus 52.4°S, 137.2°W) reduces the average age offset, but still results in a relatively large maximum offset of 3.7 Myr. When comparing the new ages to the APM models (S04P, S04G) by Steinberger et al. (2004) that attempt to compensate for the motion of hot spots in the Pacific (Hawaii) or globally (Hawaii, Louisville, Reunion and Walvis), the measured and predicted ages are more in agreement, showing only a maximum offset of 2.3 Myr with respect to the S04G model. At face value these more advanced APM models, which consider both plate and hot spot motions, therefore provide a better fit to the new Louisville age data. The fit is particularly good for seamounts younger than 50 Ma, a period for which there is little predicted motion for the Louisville hot spot and little inter–hot spot motion with Hawaii. However, discrepancies in the Louisville age-distance record prior to 50 Ma indicate there is an extra source of inter–hot spot motion between Louisville and the other Pacific hot spots that was not corrected for in the global S04G model. Finally, based on six new 40Ar/39Ar age dates, the 169°W bend in the Louisville seamount trail seems to have formed at least 3 Myr before the formation of the Hawaiian-Emperor bend. The timing of the most acute parts of both bends thus appears to be asynchronous, which would require other processes (e.g., plume motions) than a global plate motion change between 50 and 47 Ma to explain these two observations.

Deep seismic structure of the Tonga subduction zone: Implications for mantle hydration, tectonic erosion, and arc magmatism

We present the first detailed 2D seismic tomographic image of the trench-outer rise, fore- and back-arc of the Tonga subduction zone. The study area is located approximately 100 km north of the collision between the Louisville hot spot track and the overriding Indo-Australian plate where ~80 Ma old oceanic Pacific plate subducts at the Tonga Trench. In the outer rise region, the upper oceanic plate is pervasively fractured and most likely hydrated as demonstrated by extensional bending-related faults, anomalously large horst and graben structures, and a reduction of both crustal and mantle velocities. The 2D velocity model presented shows uppermost mantle velocities of ~7.3 km/s, ~10% lower than typical for mantle peridotite (~30% mantle serpentinization). In the model, Tonga arc crust ranges between 7 and 20 km in thickness, and velocities are typical of arc-type igneous basement with uppermost and lowermost crustal velocities of ~3.5 and ~7.1 km/s, respectively. Beneath the inner trench slope, however, the presence of a low velocity zone (4.0–5.5 km/s) suggests that the outer fore-arc is probably fluid-saturated, metamorphosed and disaggregated by fracturing as a consequence of frontal and basal erosion. Tectonic erosion has, most likely, been accelerated by the subduction of the Louisville Ridge, causing crustal thinning and subsidence of the outer fore-arc. Extension in the outer fore-arc is evidenced by (1) trenchward-dipping normal faults and (2) the presence of a giant scarp (~2 km offset and several hundred kilometers long) indicating gravitational collapse of the outermost fore-arc block. In addition, the contact between the subducting slab and the overriding arc crust is only 20 km wide, and the mantle wedge is characterized by low velocities of ~7.5 km/s, suggesting upper mantle serpentinization or the presence of melts frozen in the mantle.

Crustal intrusion beneath the Louisville hotspot track

We report here the first detailed 2D tomographic image of the crust and upper mantle structure of a Cretaceous seamount that formed during the interaction of the Pacific plate and the Louisville hotspot. Results show that at ∼ 1.5 km beneath the seamount summit, the core of the volcanic edifice appears to be dominantly intrusive, with velocities faster than 6.5 km/s. The edifice overlies both high lower crustal (> 7.2–7.6 km/s) and upper mantle (> 8.3 km/s) velocities, suggesting that ultramafic rocks have been intruded as sills rather than underplated beneath the crust. The results suggest that the ratio between the volume of intra-crustal magmatic intrusion and extrusive volcanism is as high as ∼ 4.5. In addition, the inversion of Moho reflections shows that the Pacific oceanic crust has been flexed downward by up to ∼ 2.5 km beneath the seamount. The flexure can be explained by an elastic plate model in which the seamount emplaced upon oceanic lithosphere that was ∼ 10 Myr at the time of loading. Intra-crustal magmatic intrusion may be a feature of hotspot volcanism at young, hot, oceanic lithosphere, whereas, magmatic underplating below a pre-existing Moho may be more likely to occur where a hotspot interacts with oceanic lithosphere that is several tens of millions of years old.

Observations of geometry and ages constrain relative motion of Hawaii and Louisville plumes

The classic view of linear island chains as volcanic expressions of interactions between changing plate tectonic motions and fixed mantle plumes has come under renewed scrutiny. In particular, observed paleolatitudes from the Emperor seamounts imply that the Hawaii hotspot was > 5–15° further north during formation of these seamounts and that rapid retardation of its southward migration was the primary agent forming the angular Hawaii–Emperor bend. Supporting this view are predictions from fluid dynamic experiments that suggest the general mantle circulation may displace narrow mantle plumes; consequently the surface locations of hotspots are not fixed and may have varied considerably in the past. However, the locations and ages of available rock samples place fundamental limits on the relative motion between the Hawaii and Louisville hotspots. Here we use such data to estimate empirical age progression curves for separate chains and calculate the continuous variations in hotspot separations through time. While the data are sparse, the inferred inter-hotspot motion for ages > 55 Myr appears significant but the observed relative motion is only about half of what is predicted by mantle dynamics models. To reconcile the observed paleolatitudes with our observed relative motion requires either a larger contemporaneous southward motion of the Louisville hotspot than previously suggested or a moderate component of true polar wander.

Conduit diameter and buoyant rising speed of mantle plumes: Implications for the motion of hot spots and shape of plume conduits

Mantle plumes are expected to be affected by large‐scale flow in the Earth's mantle related to plate motions, subducted slabs, and possibly large‐scale upwellings. Motion of plume conduits will depend on both large‐scale flow and buoyant rising speed of the conduit through the mantle. Here we present a model of depth‐dependent plume conduit temperature, viscosity, radius, and buoyant rising speed and use it to compute plume and hot spot motion. Results support a temperature anomaly of about 500 K at the plume base. In this case, sublithospheric temperature anomaly is about 150 K for the Iceland plume; transition zone anomaly is between 150 and 200 K for Iceland, about 250 K for Samoa, and about 300 K for Hawaii. Thermal plume radii are about 100 km in the upper mantle, increasing to about 200 km in the lower mantle. Beneath the lithosphere, viscosity in the vicinity of plumes is substantially reduced compared to the underlying mantle, and fast‐moving plates deflect plumes by 200 km or less, corresponding to a plume conduit buoyant rise time of about 3 Myr between 400 and 100 km depth, where most of the shear flow due to plate motions occurs. During 100 Myr, plume conduits rise buoyantly from about 1500–2000 km depth. In many cases, computed hot spot motion agrees with previous computations: south‐southeastward motion of the Hawaiian hot spot during the past 100 Myr, Louisville hot spot motion in a similar direction but at slower speed during the past 50 Myr, and westward motion of the Easter hot spot relative to Hawaii and Louisville. However, the previously computed substantial southward motion of the Kerguelen hot spot is not confirmed. If plume sources in the lowermost mantle are assumed to move with large‐scale flow, they are predicted to be displaced toward the two large‐scale upwellings beneath Africa and the Pacific relative to hot spot surface locations. For fixed sources, predicted tilts of the lower part of conduits tend to be opposite to that.