Geoscientist Online

Gill Foulger replies to their defenders by taking the Mikado out of plumes - 14 August 2003

(Apologies to William Schwenk Gilbert)

Prof. Saunders' article Mantle plumes: an alternative to the alternative is heartening in that it engages the question of whether plumes exist, a question that he writes he would have dismissed a year ago. This is a sign of progress. Further reading, however, makes it clear that we still have a long way to go. Much of the current debate still misses the point or is rhetoric designed to discredit the questioning by attributing it to emotion. It does not help to compare plume theory with other theories considered successful. It does not follow that this Emperor is clothed simply because some other Emperor is. Furthermore, I do not agree that it is reasonable that a theory may be infinitely adjusted, even to the extent of abandoning all its original fundamental features, and still be feted as "proved". I do not share Prof. Saunders? apparent contentment to accept plume theory without, it seems, aspiring to understand it fully. Intolerance of incomplete understanding is the basic driving force of science.

If the plume hypothesis cannot be adapted to fit the observations, then the observations are commonly adapted to fit the hypothesis. "Supporting observations" are quoted that have not been observed at all but are assumed to have been because they are predicted by the plume hypothesis. The Hawaiian-Emperor chain is often quoted as the best plume example on the basis of "facts" that are unsafe or wrong and omission of embarrassing conflicting data. The average basalt production rate is not a whopping 0.13 km3/year as claimed, but only a modest one tenth of this, i.e., 0.013 km3/year1§. This is approximately the same as single mid-ocean ridge spreading centres a few tens of kilometres long, which are considered to be supplied from the shallow mantle. Similarly, enhanced topography does not require a plume ? the south Pacific Superswell clearly is not caused by a plume2. Seismology shows that seismic wave speed beneath Hawaii is relatively high, not low, compared with the rest of the Pacific3. Time-progressive volcanism occurs elsewhere where plumes have never been proposed, e.g., beneath the High Lava Plains, Oregon4.

It is a little known fact that many primary facets of the Hawaiian-Emperor volcanic chain do not agree with the predictions of the plume hypothesis5:

• The great "bend" does not result from a change in direction of motion of the Pacific plate since no such change occurred.
• The "hotspot" migrated about 1000km south while the Emperors were formed, but abruptly ceased this motion at the time of the bend6.
• There is no reasonable candidate for a Hawaiian "plume head".
• The volume flux along the Hawaiian chain has increased by an order of magnitude during the last ~ 5Ma7. Indeed, if this were not the case it is Plume III, held in Hawaii, that would have been under water, not Plume IV.
• There is no significant heatflow anomaly around the bathymetric swell8.

These are first-order observations from the Emperor-Hawaii system, and are not explained by plume theory. Furthermore, it is incorrect to assert that propagating crack models cannot explain the chain, because no-one has ever made a serious attempt to quantify such a model and as long as scientists continue to talk about "the" Hawaiian plume, no-one may ever bother to do so. The claim that the evidence "proves" the existence of a plume beneath Hawaii is unscientific. What clearer statement could there be that questioning the plume hypothesis is not allowed?

Turning to Iceland, no data require "dynamic support"9. A fraction of a percent of partial melt in a more-fusible mantle beneath Iceland can explain the seismic anomaly there better than high temperature, for which there is no evidence. It is not necessary to appeal to high temperatures to explain Icelandic picrites because they are all cumulates ? picritic glass is not observed. Furthermore, the evidence for a lower-mantle seismic anomaly beneath Iceland10 has been discredited many times11.

Large Igneous Provinces (LIPs) are often claimed to require a plume model because of their huge volumes and eruption rates and associated volcanic chains. However, what is not often pointed out is:

• In many cases LIPs, e.g., the Siberian Traps12, do not have associated volcanic chains nor the precursory uplift required by the plume hypothesis
• Quantitative modeling of plumes has been unable to reproduce the large volumes and short eruption times specifically claimed to require a plume model13.
• LIPs can be explained equally well by other models14.

I presume Prof. Saunders does not marshal the geochemical observations in support of the plume hypothesis because there is none that is not equally well explained by shallow models.

Many alternatives have been suggested15, the EDGE model16 being only one of many. However, there is frustratingly little awareness or consideration of them by a community that assumes incorrectly that the plume model explains all the first-order observations well and sees no reason to look for other explanations. Perhaps the reason for this is revealed in Prof. Saunders?s extraordinary statement that it is "curious" anyone should care! Judging from the tremendous response to the Geological Society?s discussion webpage on this subject, a lot of people care.

I conclude with an answer to the challenge to present a viable alternative. I counter-challenge the plume lobby to read the considerable literature that has been published laying out alternatives (see www.mantleplumes.org and references therein) and to consider them in a scientific way, rather than to assume the plume hypothesis is correct and from that to conclude that the observations may be assumed to support it and not require further scrutiny.

§ This error in his original text was later corrected by Prof. Saunders.

• Back to the debate page

References cited

1. D.A. Clague & G.B. Dalrymple, U.S.G.S, Prof. Paper 1350, Table 1.6, 1987.
    
2. http://www.mantleplumes.org/Superswell.html
3. Julian, B.R. & G.R. Foulger, abstract in The Hotspot Handbook, Proceedings of Penrose Conference
   Plume IV: Beyond the Plume Hypothesis, Hveragerdi, Iceland, August 2003. http://www.mantleplumes.org/Penrose/PenroseAbstracts.html
4. http://www.mantleplumes.org/HighLavaPlains.html
5. http://www.mantleplumes.org/Hawaii.html
6. http://www.geolsoc.org.uk/template.cfm/template.cfm?name=MovingStory
    
7. Bargar, K.E., and E.D. Jackson, U.S. Geological Survey Journal of Research, 2, 545-550, 1974.
    
8. http://www.mantleplumes.org/Heatflow.html
9. http://www.mantleplumes.org/TopPages/IcelandTop.html
10. Bijwaard, H., and W. Spakman, Earth Planet. Sci. Lett., 166, 121-126, 1999.
     
11. e.g., Ritsema, J., H.J. van Heijst, and J.H. Woodhouse, Science, 286, 1925-1928, 1999; Foulger, G.R., et al., Geophys. J. Int., 146, 504-530, 2001; Montelli, R. et al., abstract in The Hotspot Handbook, Proceedings of Penrose Conference Plume IV: Beyond the Plume Hypothesis, Hveragerdi, Iceland, August 2003. http://www.mantleplumes.org/Penrose/PenroseAbstracts.html
12. http://www.mantleplumes.org/Siberia.html
13. Cordery, M.J., G.F. Davies, and I.H. Campbell, J. Geophys. Res., 102, 20,179-20,197, 1997.
     
14. e.g., http://www.mantleplumes.org/Deccan.html; Tanton, L.T.E., and B.H. Hager, Geophys. Res. Lett., 27, 3937-3940, 2000.
     
15. http://www.mantleplumes.org/
16. King, S.D., and D.L. Anderson, Earth Planet. Sci. Lett., 160, 289-296, 1998.