Plumes, Motion Between Hot Spots, and True Polar Wander II
Presiding: R G Gordon, Rice University; R J O'Connell, Harvard University; V Courtillot, Institut de Physique du Globe de Paris
GP22A-01 10:30h
The (Unusual?) Longevity and Spatial Stability of Hotspots and Deep Mantle Plumes in the Earth
A physical link has been proposed between hotspots, regions with particularly persistent, localized, and high rates of volcanism, and underlying low viscosity deep mantle plumes constructed of large spherical heads and long-lived narrow trailing conduits. This plume model has provided a way to interpret observable phenomena including the volcanological, petrological and geochemical evolution of large igneous provinces, ocean island volcanoes, the relative motion of plates, continental breakup, global heat flow and the Earth's magnetic field within the broader framework of the thermal history of our planet. Despite the plume model's utility, the underlying dynamics giving rise to hotspots as long-lived stable features have remained elusive. An essential test of the mantle plume hypothesis is, thus, to understand the longevity of hotspot volcanism and spatial stability of hotspots themselves in terms of the mechanics governing the formation of mantle plumes at the core-mantle boundary. Using a combination of laboratory experiments, numerical simulations and scaling analyses we show that: 1) The high temperatures (and low viscosities) inferred for mantle plumes are likely a result of strong cooling of the mantle by large-scale stirring driven by plate tectonics; 2) The head-tail structure of such plumes is a necessary but insufficient condition for their longevity; 3) The longevity and spatial stability of mantle plumes are a consequence of interactions between plate tectonics, core cooling and a dense, low viscosity layer within D", which is plausibly composed of a mixture of silicate partial melt and outer core material. Under certain conditions, analysis of entrainment from this dense layer leads to a further prediction that the variation in 3He/4He (or any tracer of the silicate component of the lower mantle plume source) will be proportional to plume buoyancy flux, which is broadly consistent with observations.
GP22A-02 10:45h
Eruption Sites of LIPs of the Past 250 My With Respect to the slow/fast D" Boundary in Four Improved Reference Frames
We have shown (Burke and Torsvik 2004) that nearly all of the Large Igneous Provinces (LIPs) erupted on the Earth's surface during the past 250 My rose, at the time of their eruption, from the base of the mantle close to the 1 percent slow S-wave delta V contour of the D" zone. Here we report on the use of (1) Improved plate-circuitry (among PAC-ANT-AUS-MADAGAS-IND-SOAM-AFR) and (2) upgraded paleomagnetic data-sets to confirm and refine our results. In an attempt to establish whether rotation of LIPs to their sites at eruption can be used to learn something about mantle and core behavior over the past 250 My we have performed rotations on assumptions of: (1) a persistent geocentric axial magnetic dipole field, (2) a time-dependent non-dipole (octopole) magnetic field contribution (Torsvik and Van der Voo (2002), (3) a mantle wind contribution to mantle flow (Steinberger et al. 2004) and (4) a revised (new timescale) fixed hot-spot (no mantle wind) reference frame. Our test consisted of estimating how closely radials dropped from rotated LIP sets based on these four assumptions plotted to the 1 percent slow contour derived from seismic tomography. We compared S-wave tomographic models at the CMB and opted to use the SMEAN model (Becker and Boschi 2002) recognizing that other models are similar and that all have similar limitations. The mean deviation of the projected rotated LIPs from the 1 percent slow contour is 6 degrees at the CMB. The four different assumptions yield results between 6.5 and 5.5 degrees which seems unlikely to be a significant difference. Our conclusion is that it is not yet possible to discriminate among the four models of mantle and core behavior using our procedure. Becker, T.W and Boschi, L. 2002: A comparison of tomographic and geodynamic mantle models. Geochemistry, Geophysics, Geosystems 3, 2001GC000168. Burke, K. and Torsvik, T.H. 2004: Derivation of large igneous provinces of the past 200 million years from long-term heterogeneities in the deep mantle. Earth Planet Sci. Lett. 227, 531-538. Steinberger, B., Sutherland, R. and O'Connell, R.J. 2004: Prediction of Emperor-Hawaii seamount locations from a revised model of plate motion and mantle flow. Nature 430, 167-173. Torsvik, T.H. and Van der Voo, R. 2002: Refining Gondwana and Pangea Palaeogeography: Estimates of Phanerozoic (octupole) non-dipole fields. Geophysical Journal International 151, 771-794.
GP22A-03 11:00h
A consistent model of large-scale mantle flow, hotspot motion, true polar wander and plate motions
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. Our model of large-scale flow is based on mantle density anomalies inferred from seismic tomography and mineral physics, and a viscosity structure that is also consistent with mineral physics. Flow and density field are also in accord with the geoid and global heat flux. Using the same viscosity model and new experimental constraints we compute depth-dependent plume radius and buoyant rising speed. We expect thermal plume radii of about 100 km in the upper mantle, increasing to about 300 km in the lower part of the mantle, in agreement with recent results from tomography. This yields somewhat larger conduit rising speeds than previously assumed, and predicted hotspot surface motion is often similar to the horizontal flow component in the lower part of the mantle. Our flow model also considers advection of density anomalies, hence changes of the geoid with time and true polar wander can be computed. Thus we are also able to consistently compute hotspot paleolatitudes - an important observational constraint on hotspot motion. Combination of hotspot motion and plate motion allows computation of hotspot tracks; if there are alternative plate motion models we are able to determine which one yields the best fit. Our results are of particular importance for the relative motion between Pacific and African plates between 47 and 83 Ma.
GP22A-04 11:15h
Reconciling hotspot and TPW reference frames within the uncertainties by evolutionary computation of hotspot motion
A significant discrepancy exists between TPW and fixed hotspot reference frames in the time period between 90-120Ma. A number of Indo-Atlantic hotspots, including Tristan da Cunha, Great Meteor and Kerguelen, display a significant paleolatitude drift at this time, which cannot be explained by polar wander. However, the degree of motion implied is of a similar magnitude to the uncertainties inherent in the hotspot reconstructions. We present a method for contraining the uncertainties in hotspot reconstructions, and present a model for the motion of the Indo-Atlantic hotspot group based on an interactive evolutionary approach to the inverse mantle convection problem. The hotspot motion is consistent with available paleolatitudinal constraints, and calculated plume conduit tilts are similar to tomographically imaged conduits. We show that fixed and moving hotspot reference frames are not significantly different until 80Ma, and the discrepancy between TPW and hotspot reference can be greatly alleviated for moving hotspots.
GP22A-05 INVITED 11:30h
Upwelling Plumes, Superswells and True Polar Wander
The geological evolution of the Earth's rotational axis is most likely controlled by internal mass redistribution within the mantle. Paleomagnetic observations suggest that it is episodic in nature, with periods of quasi-standstill alternating with periods of faster wander. Here, we investigate two models for the influence of mantle plumes which vary at different spatial wavelengths on the time variations of the rotational axis (True Polar Wander). In the first model, we represent an upwelling plume as a sphere whose radius varies as a function of the flux of material in the conduit, and which traverses the mantle at the Stokes velocity. Such a plume produces very little wander of the rotational axis. We then study the effects of two superswells which mimic the ones observed with seismic tomography, and conclude that a doming regime within the mantle involves significant polar wander. Some of the features of this TPW which are directly linked to the periodicity of doming are reminiscent of observed phases of slow and fast true polar wander, with similar peak velocities.
GP22A-06 11:50h
New Tests of the Fixed Hotspot Approximation
We present new methods for estimating uncertainties in plate reconstructions relative to the hotspots and new tests of the fixed hotspot approximation. We find no significant motion between Pacific hotspots, on the one hand, and Indo-Atlantic hotspots, on the other, for the past ~ 50 Myr, but large and significant apparent motion before 50 Ma. Whether this motion is truly due to motion between hotspots or alternatively due to flaws in the global plate motion circuit can be tested with paleomagnetic data. These tests give results consistent with the fixed hotspot approximation and indicate significant misfits when a relative plate motion circuit through Antarctica is employed for times before 50 Ma. If all of the misfit to the global plate motion circuit is due to motion between East and West Antarctica, then that motion is 800 ± 500 km near the Ross Sea Embayment and progressively less along the Trans-Antarctic Mountains toward the Weddell Sea. Further paleomagnetic tests of the fixed hotspot approximation can be made. Cenozoic and Cretaceous paleomagnetic data from the Pacific plate, along with reconstructions of the Pacific plate relative to the hotspots, can be used to estimate an apparent polar wander (APW) path of Pacific hotspots. An APW path of Indo-Atlantic hotspots can be similarly estimated (e.g. Besse & Courtillot 2002). If both paths diverge in similar ways from the north pole of the hotspot reference frame, it would indicate that the hotspots have moved in unison relative to the spin axis, which may be attributed to true polar wander. If the two paths diverge from one another, motion between Pacific hotspots and Indo-Atlantic hotspots would be indicated. The general agreement of the two paths shows that the former is more important than the latter. The data require little or no motion between groups of hotspots, but up to ~10 mm/yr of motion is allowed within uncertainties. The results disagree, in particular, with the recent extreme interpretation of Tarduno et al. [2003], who assume (1) that motion of the Indo-Atlantic hotspots relative to the spin axis can be ignored during the past 85 Myr, and (2) that the Hawaiian hotspot has been fixed relative to the spin axis since the age of the Hawaiian-Emperor bend. Our results indicate that both assumptions are false.