Roadmap | The review process | Home

Ultimate Test Of The Plume Hypothesis

Don L. Anderson
Friday, May 2, 2014

Morgan (1971, 1972) proposed that “hotspots” (midplate or intra-plate volcanoes) are manifestations of convection in the lower mantle, and provide the motive force for continental drift. In his model about twenty deep mantle plumes,“pipes to the deep mantle” , bring heat and primordial material to the asthenosphere. Horizontal currents flow radially away from each of these plumes. The points of upwelling have unique petrological and kinematic properties but there are no corresponding unique points of downwelling, the return flow being uniformly distributed throughout the mantle. Wilson (1963) had earlier pointed out that Hawaii, the largest of these “hotspots”, could be sourced as shallowly as 200 km, just below the “jet-stream” of the rapidly moving surface layer.

Morgan proposed that “highly unstable mantle”, just as the atmosphere, would yield a thunderhead pattern of flow rather than the convection cell patterns calculated from linear viscous equations. The currents in the asthenosphere spreading radially away from each upwelling produce stresses which, together with plate-to-plate interactions, determine the direction in which each plate moves. Hotspots come and go but they are motionless with respect to each other, as Wilson proposed in his shallow source model for Hawaii.

This model was deemed to be compatible with the observation that there is a difference between oceanic island and oceanic ridge basalts. In Morgan’s mantle plume model “relatively primordial material” from deep in the mantle rises adiabatically up to asthenosphere depths. Ridge basalts come entirely from the asthenosphere, passively rising to fill the void created as plates are pulled apart by the stresses acting on them. The 2 billion year “holding age” needed to explain lead isotope data reflects how long the “relatively primordial” material was stored and isolated in the lower mantle without change.

Morgan (1972) argues that the hotspots provide the driving force for plate motions because almost all of the hotspots are near rise crests (agreeing with the notion that asthenosphere currents are pushing the plates away); hotspots become active before continents split apart. Morgan argued that neither rises nor trenches are capable of driving the plates; ridges are passive. He argued that dikes cannot push the plates apart and that sinking lithospheric plates do not provide the main motive force. The alternative notion is that gravitational forces in the cooling plate and sinking slabs pull the plates away from ridges and in general control the directions and velocities of plate motions. Morgan specifically excludes this possibility. It is now generally agreed, from detailed modeling, that plates and slabs are the main drivers of plate tectonics and, perhaps, mantle convection as well.

The kind of convection advocated by Morgan occurs in the atmosphere and laboratory fluids when strongly heated from below. The thunderhead analogy, if valid, suggests a geometric and dynamic scaling relation that allows the mantle to develop structures similar to those seen in the atmosphere and the laboratory, in spite of the large differences in scale, viscosity, thermal conductivity and mode of heating. In the mantle, cooling from above adds a concentrated downwelling aspect to mantle convection, and radioactive heating causes diffuse upwelling, not allowed for in Morgan’s scheme.

Morgan suggested that plumes are about 150 km in diameter, rise at 2 m/year and bring up a volume of 500 km3/yr, which is greatly in excess of the amount of crust formed each year by seafloor spreading. Furthermore, the plume heat flux is predicted to be half the global heat flux. He suggested that plumes are 300°C hotter than normal mantle; even higher temperatures are required for narrower plumes. In Morgan’s model these very large volumes and heat flows are necessary in order to keep ridges from closing. Specifically, he argued that plume fluxes as low as 10 km3/yr would contradict the model.

Current estimates of plume heat and volume fluxes are less than 20% of those required by Morgan to make the plume hypothesis viable. The magma and heat flow budgets of the mantle are dominated by plate tectonics, not by plumes, and plates and slabs apparently drive themselves, and mantle convection. The isolation time required by isotopes is comparable to the revisitation times of migrating ridges and long term storage in the deep mantle is not required. Petrological estimates of differences in magma temperatures between ridges and hotspots do not exceed 200°C and are generally much less (Presnall & Gudfinnsson, 2008; Green & Falloon, 2005). It is not clear that they represent differences in the deep mantle adiabat which may, on average, start near 300 km depth (Kaula, 1970). Large radial temperature gradients occur in the near-surface thermal and chemical conduction boundary layers and magmas extracted adiabatically from various depths can be expected to have 100°C higher temperatures for each 10 km increase in depth. Lateral temperature changes of at least 200°C occur in the shallow mantle as a result of plate tectonic processes. High-temperature komatiites apparently have higher temperatures but this may reflect cooling of the mantle since the Archean.

Ironically, high-resolution global tomography has confirmed that there are about 20 features in the upper mantle that might be interpreted as upwellings. However, their dimensions and inferred ascent speeds show that they are passive features, similar to those imaged under ridges.

The thermal and chemical conditions assumed by Morgan for the deep mantle actually exist in the surface boundary layer (Anderson, 2011, 2013). The 20 or so upwellings have lateral dimensions of 1000-2000 km (Ritsema & Allen, 2003) confirming earlier studies by Anderson, Tanimoto and Zhang. Using Morgan’s relations the inferred ascent speeds are now 1-2 cm/yr, not meters per year. These considerations not only falsify the Morgan plume hypothesis but make it unnecessary.


  • Anderson, D. L. (2011), Hawaii, boundary layers and ambient mantle–geophysical constraints, J. Pet., 10.1093/petrology/egq068.
  • Anderson, D. L. (2013), The persistent mantle plume myth, Australian Journal of Earth Sciences, 60, 657-673.
  • Green, D. H., and T. J. Falloon (2005), Primary magmas at mid-ocean ridges, “hot spots” and other intraplate settings; constraints on mantle potential temperature, in Plates, Plumes & Paradigms, edited by G. R. Foulger, J.H. Natland, D.C. Presnall and D.L. Anderson, pp. 217-248, Geological Society of America.
  • Kaula, W. M., Earth’s Gravity Field; Relation to Global Tectonics, Science, 169, 982 (1970).
  • Morgan, W.J. (1971), Convection plumes in the lower mantle, Nature, 230, 42–43.
  • Morgan, W. J. (1972), Plate motions and deep mantle convection, Geol. Soc. Am. Bull., 132, 7-22.
  • Presnall, D. C., and G. H. Gudfinnsson (2008), Origin of the oceanic lithosphere, J. Pet., 72, 615-632.
  • Ritsema, J., and R. M. Allen (2003), The elusive mantle plume, Earth Planet. Sci. Lett., 207, 1-12.
  • Wilson, J. T. (1963), A possible origin of the Hawaiian Islands, Can. J. Phys., 41, 863-870.


last updated 2nd May, 2014