The Great Mantle Plume Debate has been simmering aggressively – in a similar fashion to soup on an unattended stove- throughout all facets of Earth Science since the 1960’s. Seismologists, geophysicists, experimental petrologists, and analytical geochemists alike have invested serious time and money trying to solve the mystery that is the mantle plume. Since starting my PhD on the geochemistry of Hawaiian volcanism, I have been thrust- like a spoon in a tub of icecream- to the complex topic that is mantle plume theory, and have been surprised to learn just how much is still in dispute, and how much is still to be discovered. Sounds dramatic, but -much like a good Agatha Christie novel- a lot of headway has been made by some unassuming and cool-headed detectives (scientists).

The aggressively simmering pot of soup that is the mantle plume debate.
The aggressively simmering pot of soup that is the mantle plume debate.

First of all, what are mantle plumes?

‘Mantle plume’ is the name given to buoyant material that rises from depth in the Earth through the mantle in a narrow conduit. This material melts, and produces volcanoes at the Earth’s surface which take on a linear geometry if a tectonic plate is moving overhead- like a linear row of oozing pimples…(?)

‘I don’t care about plumes!’ I hear you say.

Plume theory, now widely accepted (although there are many influential scientists who do not subscribe to the theory in its current form), is used to explain the curious types of volcanoes that form independent of tectonic forces, unlike the majority of volcanism on Earth. For example the volcanoes which make Earth’s continental crust are produced by the collision of two tectonic plates, and most of the Earth’s oceanic crust is produced by volcanism at two diverging plates. Some volcanoes however, occur far away from plate boundaries, and the mechanism for their formation is still shrouded in mystery! Much like a man in a trench-coat would be, if he were lurking on a street corner at night.

Mantle plumes if they were a person.
Mantle plumes if they were a person.

These mysterious volcanoes are known by many different names, depending on who you talk to, including:

  • Hotspots
  • Wetspots
  • Ocean Island Basalts (OIB)
  • Or vaguely: ‘melting anomalies’

The reason for all the names is that there are still many different theories circulating in the field of Earth sciences as to how mantle plumes form, what they are exactly, and if they even exist! Yes, many Earth Scientists maintain that plumes do not exist, and that hotspot volcanism is passive, caused by surficial processes such as shallow cracks and fissures in the crust- controversial! This group prefer to use the term ‘melting anomaly’. I will continue the rest of the article assuming that plumes do exist, however plume denialists raise a good point that the current one-size-fits-all plume theory struggles to account for the many unique features of each hotspot.

At this stage it needs to be pointed out that there are over 60 hotspots/wetspots/OIBs documented globally, and every single one on Earth is unique in some way.

Many appear to start at different mantle depths, have different isotopic trace element signatures, and even highly variable major element signatures. They all have different volume fluxes, uplift and subsidence rates, some form tracks while others don’t; some have flood basalts whereas others don’t. In general though, they all tend to be located around Africa or the Pacific Ocean- kinda weird!? Some well-known hotspots include:

  • Hawaii
  • Iceland
  • Reunion
  • Yellowstone
  • Samoa
  • Kerguelen
  • Ascension
  • Cape Verde
  • Canary Islands
  • Easter Island

The two big questions that are the crux of the Great Mantle Plume Debate are:

1) How deep in the Earth do plumes originate?

The best place to start with this question, is seismic evidence. The speed of seismic waves through the mantle enables us to see – as one would see a lady’s intimate apparel by using X-ray vision goggles- material of contrasting density, and thus to see plumes. Attenuation tomography allows us to see slices through the mantle and reveals differences in porosity, permeability, and viscosity of the material through which it is slicing. Studies have shown that, yes! we can see plumes below hotspots, and interestingly most originate from highly variable depths in the mantle: some are restricted to the upper mantle, some to the lower mantle, and a few have even been tracked to the core-mantle boundary (2800 km deep in the Earth… that’s as deep as an LSD-fuelled philosophical epiphany!). This method is good, but it seems the resolution is not good enough yet to have persuaded everybody in the Earth Sciences.

Seismic waves (X-ray goggles) travelling through the Earth's mantle (lady’s peticoat).
Seismic waves (X-ray goggles) travelling through the Earth’s mantle (lady’s peticoat).

A good way to show you how seismic imagery may not be high resolution enough is to show you what mantle plumes are supposed to look like, based on fluid dynamic experiments, and then compare it to what we actually see:

Left: what plumes are supposed to look like. Right: actual tomographic images of the plume beneath Hawaii.
What plumes are supposed to look like.
plume hawaii
Actual tomographic images of the plume beneath Hawaii.









As you can see, some colourful imagination is required to convert between the two. What we can say for sure though, is that large areas of mantle below Africa and the Pacific contrast seismically to surrounding ambient mantle. As I said before, this also happens to be where most of the hotspots are located. Coincidence…? I think not!

In addition, geochemists –represent- know that the isotopic signature of hotspot lavas is very different to that of mid-ocean ridge lavas, indicating they must have come from much deeper in the mantle than the strongly depleted mid-ocean ridge lavas. Further geochemical evidence for a core mantle boundary source for some hotspots comes from the helium isotopic signature, although this isotope as a tracer has come under scrutiny because other processes may affect the ratio.

Some disagree with a deep source however, and suggest that the source depth may be restricted to the upper mantle (shallower than mid-ocean ridges in some cases), or at least to the 660km discontinuity, and that upwelling occurs as a result of mineral phase change there.

2) What causes the material to upwell?

This question appears to divide people the most. There are broadly two possibilities: it is a thermal anomaly at some boundary layer that causes material to become warm and buoyant, or it is a geochemical compositional anomaly that causes buoyancy.

Another possibility which does not get discussed nearly enough in my opinion is that it could be some combination of the two. Both work on the same principal that warmer material or material with a different chemical composition would become buoyant and upwell to create a plume- in the same way you can melt ice by either warming it up or by adding salt to it.

A few methods have been used to investigate this; one is using calculation of buoyancy flux to see what temperatures would be required to produce large amounts of melting- many people believe that for any large amount of melting to occur an increase in temperature must exist. This is where the name ‘hotspot’ originates from. Similarly, topography can be used to infer the amount of heat entering the plume, and subsequently the depth the plume originates.

Another method is the use of geothermometry to determine mantle temperatures. One that is commonly used is the olivine geothermometer, however the use of this method has produced excess temperatures of over 200̊C in some studies, but zero excess in others. Either way the thermal argument would require long-lived thermal anomalies at the core-mantle boundary, and heat conducting from the core heterogeneously.

Geochemists and petrologists are more in favour of a chemical argument for plume formationfunnily enough.

The general idea is that subducted crust may sink down to a boundary layer where it sits, warms up and becomes buoyant, and melt is produced because the reaction between this crust and mantle forms material with a lower melting point than ambient mantle. There is growing evidence that the source rock for OIB lavas may be pyroxenitic, and not peridotitic, based on experimental petrology and the ratios of major elements present in the lava.

Others speculate that the amount of water and carbon in this subducted material may be the cause for melting. ‘This is why ‘hotspots’ have started to be called ‘wetspots’ by some scientists.

Other studies have rejected the need for a deep source for this crustal material, suggesting that it may simply be delaminated lithosphere that has metasomatized the asthenosphere to cause melting.

Seismic studies have found ‘blebs’ of material sitting at the core-mantle boundary (called large low shear velocity provinces), but it cannot distinguish the difference between thermal or compositional ‘blebs’, so unfortunately this method is less useful for answering this question.

In the end, we have many different unique volcanoes that vary in so many different ways, and we don’t have a mantle plume model that explains all of them. We also don’t know the exact mechanism that causes plumes, and we don’t know for sure at what depth in the Earth they form. This is good news for us geologists as it gives us something to do with our attention deficit brains.