Iceland: Plumbing the Plume

Image borrowed from the University of Iceland.

In the last decade, science has been under increasing attacks and have lost in status among the general population. Where famous scientists a century ago could rake in some serious dosh by going on lecture tours explaining their findings, today’s scientists are increasingly unknown and scorned by the general population.

Instead, we see the rise of pseudo-sciences ridiculing evidence and holding “feelings” in high esteem while screaming the words “fake news” at the top of their lungs.

A few decades this was just a movement on the fringes of society, but about a decade ago this started to happen inside science too. Some scientists started to take advantage of the peer-review system to push papers based on false data, or that are just nonsensical.

A few days ago, two of these “scientists” died due to a lack of vaccination against Covid-19. I am obviously talking about the twins, Igor and Grichka Bogdanov. Igor got himself a Ph.D. in Physics and Grichka one in applied mathematics.

Based on this and a few published papers they got their professorial robes. The only problem was that it turned out that their entire list of publications including what they did for their Ph.Ds., was entirely without a shred of substance or meaning. It was all a huge joke on science.

In this case, it was probably an art project of sorts, we are after all talking about a couple of real characters, complete with record-breaking amounts of facial silicone and potentially being the partial fathers of Bitcoin and Ethereum (the Ethereum part is canon, the Bitcoin part is at least very likely).

So, what on Earth does this have to do with geosciences? Sadly, the same thing is going on in the geosciences with bogus papers proliferating to bolster careers that should be non-existent. But where the Bogdanov’s lied about the moment right after Big Bang, the geoscientists are congregating around something that should not even be contentious, mantleplumes.


The Definition of a Mantleplume

“A mantleplume is caused by upwelling of hot ductile material from deep within the mantle conveying this hot material to, or near, the surface of Earths crust.”

This is the modern definition that is atomic in so much that it covers the essentials of all known mantleplumes.


The Origin of the Plume

Kistufell has the distinction of being Iceland’s least photogenic volcano, and also being Iceland’s least photographed volcano. This is as far as I know the only image.

At first, the mantleplume was a hypothetical solution to explain the existence of volcanism in Hawaii. This Hawaiian origin greatly influenced the theory building at the beginning, somewhat to the detriment of our understanding of mantleplumes.

The original definition, and the list of prerequisite evidence for being a mantleplume, is something that has rightfully been attacked by scientists in the field. This has in turn forced into place refinements of the theory.

If we look at the original list of required evidence, we find the following:

  1. There will be a linear chain of progressively older volcanic features as the crust moves above the mantleplume.
  2. The amount of primordial He3 should be anomalously high compared to He4 in collected volcanic gases.

The idea behind 2 is that He3 can’t form on Earth and is of stellar origin, whereas radioisotope decay can form He4. All helium is leached out of the atmosphere, so an anomalously high He3 count can only come from ancient deep trapped Helium from the time of the formation of Earth.

There was much rejoicing in the geosciences, finally, we could say something about the innards of our planet. Much lip-banjo was played by volcanologists.

The problem was that Nature was about to wreak havoc on this original definition of what a mantleplume was.


The anomalous Mantleplumes

Nature enjoys throwing spanners into science at the most inopportune times it seems. After finding the Hawaiian plume, and a couple of others (Albert did a fine job on one of those in the previous article), more and more very odd plumelike things started to crop up.

So many of those odd plumelike things cropped up that it started to look like it was the Hawaiian type of plume that was the anomaly, and not the other way around.

The likes of the African plume(s), The Azore plume, The Canary plume, The Icelandic plume, and so on seemed to be nature peeing mightily on the original plume theory. None of those fulfilled the first prerequisite of having a nice volcanic track, and to make things even worse, volcanoes that are decidedly not of plume origin showed up with anomalous He3/He4 content.

It was time to go back to the drawing board. And this time the entire barrage of the theory of science was deployed.


A Plume of good science

Here we must pause and debate what is good science for a moment. First, there must be a hypothesis.

Let us reformulate the definition into a hypothesis. Are there cases of upwellings of hot ductile material from deep within the mantle conveying this hot material to, or near, the surface of Earths crust?

To move this from a hypothesis we need to follow a few rigorous steps to get it elevated into a proper scientific theory.

The first step is, can you derive predictions from the hypothesis? In this case, there were 3 predictions accepted (later there was more). He4/He3-anomaly (not conclusive, but pointing towards the hypothesis having merit).

The second one was that there should be zones in the mantle related to the proposed plume that evidenced areas where the speed of sound changed due to increased heat, if so this should be possible to measure with instruments.

And the third was that there should be inclusions of materials that can only form in higher temperatures and at higher pressures than what is possible at normal magma-formation depths. This is if you think about it the deal-breaker. No such inclusions and it is not a plume.

The second step is, can the hypothesis be easily falsified? The requirement is even more stringent, as you formulate your hypothesis you must stipulate ways that the theory can be falsified (sticking your neck out). It is also obviously free for any scientist to ad things that are falsifiable within the theory, and then try to falsify it. We find that our formative plume-theory is eminently possible to falsify at this point.

And the third step, all experiments and data collection should be possible to repeat by other scientists in the field. It turned out that this was possible to do at numerous instances and plumes.


Plume Problems

Even though we by now have a well-established theory of plumes it turns out that we have something of a headache on our hands. Our proposed solutions caused far more questions than they answered.

How does a plume form? Sounds simple enough, but we do not have the foggiest idea, even though some interesting hypotheses have been brought forth. In fact, we only know parts of how a single plume formed, and that particular plume is the definition of anomalous.

Are the number of plumes increasing, or decreasing, over geological timespans?

Are plumes nicely shaped conveyor like things, or are they messy vortex-like spinning constructs?

To a layperson, this sounds like a mess best avoided, but in science, this is akin to a Gold-rush from all of those majestic questions that are open to solving. This is a moment where strapping Ph.D.-students across the globe crack their knuckles to create lifelong careers borne aloft on the rise of The Plume.

Now, before I start to play lip-banjo out of scientific joy we need to tackle to elephant in the room. The most anomalous plume of them all must be conquered. And it is honker of a problem that could easily derail Plumology.


The Icelandic Plume

Some days I feel that the entire reason for the existence of Iceland is nature’s way of taking a dump on all that we know about volcanology and geology. Whereas most of the rest of the planet is neatly ordered and understandable, and mainly only lacking enough instrumentation, Iceland is a chaotic maelstrom.

Iceland is quite like a psychotic kid running around giggling maniacally as it kicks you in the gonads for the sheer heck of it.

The problem is that no single theory can explain what is going on in Iceland. You must fully understand ALL of the scientific theories that are relevant for the place and then use ALL of them at the same time to model an understanding of what is happening there at a given point of location and time.

Sadly, not even the most picture-perfect and tested theory of mantleplumes will explain more than parts of Iceland’s very existence.

And to pile insult upon gonadal pain, the Icelandic Plume is setting records in being the most anomalous, making even the African Superplume look like a normal garden variety rose. If there is one plume that could topple the scientific plume-carriage, it is the adorable lump named Iceland.

First of all, it is the youngest of all known mantleplumes, being between 14.4 and 17 million years old. We know this from geochemical data from the oldest volcanic rocks in Iceland that is evidencing plume-origin factors like deep-inclusions and He3/He4 anomalies.

It is the most location stable mantleplume on Earth, neither the crust nor the plumehead itself is moving about to any great extent.

We also know that it is growing in intensity and depth over time. We know this since we can track both eruptive volumes related to the Mantleplume, and also the depth that the deep-inclusions form at.

This means that we can say that the Icelandic Mantleplume was born under what would become Iceland and that is not in any way form or shape came wandering from someplace else.

In other words, the Icelandic Plume is in fact a rather psychotic gonad kicking child of a plume, we also know that it originated from the top and is overtime burrowing downwards.

This obviously means that we have one class of plumes that is confirmed to have a top-down origin and that we perhaps have a bottoms-up category to be proven to exist. In other words, we can make the prediction that if Hawaii is of the top-down origin category we will find a curve of depth-created inclusions leaning shallower the further away we sample down the track from current Hawaii.

Easily falsifiable obviously, if they turn out to be ultra-deep from the starting point, we have a case for a second bottoms up variety of plume. Just a thought as I am writing.

So, let us look for evidence of an Icelandic plume. One after all has to try to find evidence for a scientific theory.


He3/He4 Anomaly Problems

He3/He4 anomaly data is problematic for Iceland. The first problem is that since the plume is not as deep as let us say Hawaii, we have less pronounced He3/He4 anomalies. But the really irritating gonadal giggle-kick is that Iceland is eminently producing non-plume related He3/He4 anomalies. Let me explain with a recent example.

Fagradalsfjall had a deep origin not related to any known plume activity. Instead, it was caused by crustal rifting sucking up mantle material, and as the eruption went on the erupted material came from ever deeper origin showing constantly increasing amounts of primordial He3.

First of all, this makes He3/He4 anomalies fairly useless as an investigative tool in regards to the Icelandic plume.

Secondly, it is perhaps giving an inkling towards how the Icelandic Plume started. Perhaps it was a Fagradalsfjall like “origin-eruption” that was larger, and that started to burrow down deeper and deeper until we got a downwards self-propagating mantleplume. This is just me thinking loudly as I write, not even a hypothesis at this moment in time.

Anyway, let us chalk up the He3/He4 anomaly studies as being muddy, but that the requirement is fulfilled. Since it is not giving the clarity we need on the subject, let us just state that there is such an anomaly and move on.


Seismic Tomography of the mantle under Iceland

We find that the Icelandic Mantleplume is evident at depths below 670 kilometres, and maybe even deeper. We also find scant evidence in the tomography for the Hawaiian mantleplume, at least a deep origin is not evidenced.

By tracking the wavefronts caused by large earthquakes we can make detailed (well, sort of detailed) maps of the interior of the planet. What we can track is how what is down there alters the speed of sound of the wavefronts of the P- and S-Waves. This alteration is predominantly caused by variations in temperature.

Warmer temperatures will cause the speed of the wavefront to slow down, and cold areas (sunken plates for instance) will cause increases in speed.

And as we can see there is a warm temperature anomaly under Iceland consistent with a mantleplume.

Now, there is a problem here that most people do not understand. A tomographic image in and of itself is only showing that there is a tubelike form that is hotter than the surrounding mantle, but it does not de facto prove that the heat is caused by hot ductile material being conducted upwards.

Let us pause here. This last problem is real, and it is something that I am surprised that plume-denialists have not dug into. After all, it could derail plumes completely. On the other hand, they never discuss seismic tomography, to begin with for some reason.

The reason is that then the onus would be upon them to come up with a workable hypothesis that is falsifiable, producing testable predictions that are also falsifiable and possible to repeat, to explain how the heat-tubes in the mantle are formed.

Let us just state for now that we have indeed found the predicted “heat-tube” and move onwards to a more inclusive ground.


Deep Magma Inclusions

Beautiful Chrome Spinel. Depending on the crystal shape it has many interesting uses.

In February of 2002, the angels of science sang as Kresten Breddam published the seminal paper Kistufell: Primitive Melt from the Iceland Mantle Plume, in Journal of Petrology. It is not hyperbole to state that it is one of the 3 most important papers published in regards to Icelandic volcanology and the single most important one in regards to mantleplumes anywhere.

My only criticism of the paper is that it is insanely dense and requires you to have above average knowledge on a scientific degree level in geochemistry, to be able to understand it. Words like highly technical do not even begin to cover it.

So, most of you will have to take my word on it being very rigorous and a showcase of good science. I will now try to explain what is so special.

The premise is quite simple. What inclusions can we find in the lava that has erupted out of Kistufell, a volcano that is suspected to be located above the centre of the plumehead itself? What depths are those inclusions formed at? And what inclusions are missing from greater depth than those found? The last one gives depth constraints for the plume.

The first part to note is that the paper finds evidence of subducted oceanic gabbro. This is interesting as such, but it is not the same thing as a sunken continent under Iceland, instead, it is the original oceanic Icelandic crust in and of itself that has been pushed down and morphed into sub-crustal gabbro. This formed as oodles of magma pushed down the oceanic crust it erupted on top of.

I think that we should invent a new term for this process since it is not classical subduction in the common sense of the word. Interestingly enough, this is also the case over in Hawaii.

The paper goes through in detail the gabbro part and the underplating material for those who are interested in those things (I am for one), but it also notes how rare those xenocrysts are from the deep crustal parts.

Let us now discuss the melt depth of the magma, it occurred just below the crust at circa 45 kilometres depth. This high depth/melt figure is most likely explained by the high temperature of 1270 at the point of melting.

This is interesting in a theoretical sense; it means that if the crust became even thicker it would effectively hinder any melt from occurring under Kistufell. Interesting, but it has nothing to do with our discussion as such. This is evidenced by the chrome-spinel formation depths.

Lavas from Kistufell shows that the plume derived magmas are like this: “The isotopic heterogeneity within the Iceland mantle plume may thus be viewed as a result of mixing between plume material rising from a layer of subducted slabs (which have partly maintained their geochemical integrity and heterogeneity) and lower-mantle material (FOZO) entrained in the initial stages of plume formation.” (Kresten Breddam, 2002, linked below)

The sentence above might be the most explosive sentence in contemporary volcanology. It is like someone had chucked the Tsar Bomba into the classical mantleplume model for Iceland. I will try to explain it by quoting myself… Massively.

“Kistufell is situated straight on top of the Icelandic mantleplume core. The petrochemical analysis gives at hand that a large part of the magma comes from the 670-kilometre discontinuity… …and is consistent with a formative mantleplume in the lower mantle.

Now, what on earth is the 670-kilometre discontinuity? Well, the material above that has the spinel crystal structure and below you have perovskite structure. In short, if your basic magma has spinels in it you have magma from above the discontinuity. If you have a marked lack of spinels the magma formed deeper than the discontinuity.

And the Kistufell magma is poor in chromium spinels, and the few that are seems to have come from xenoliths from the magma conduits rather than from the basic basalt (ol-tholeiite). Also, the high amount of Sr points towards a deep source.

Now over to garnets, they form at about 35 to 45 kilometres depth, and the Kistufell lava is very poor in garnets, so it is safe to assume that the magma has formed below that. This differentiates the Kistufell (and other mantleplume volcanoes) from other Icelandic volcanoes far away from the plume core.”

Is there any evidence that the mantleplume is indeed formative? Yes, there is. The amount of spinels increases with the age of the lavas tested in Iceland. Or, in other words, the older magmas came from an increasingly shallow depth as we progress backwards in time.

As such the mantleplume is not more than 14.4 million years old, at least in a way that we define as a mantleplume. That puts quite a spanner in the Alpha Ridge Theory, or any other theory stating that the Icelandic Plume has meandered over from somewhere else.



The mantleplume theory itself is not problematic, what is problematic is the lack of understanding of mantleplume formations. It is also surprisingly lacking in the ordering of types of mantleplumes because there seems to be more than one subtype that follows the basic definition of being a mantleplume.

We have up above defined the Icelandic Mantleplume as the most anomalous plume, and we have put it to the test against the definition and the prerequisites of being defined as a plume.

We find that it fulfils the prerequisite of having a high He3/He4-anomaly. We found that this is a must, but that it is not defining in and of itself. A lack thereof would though be constituting falsification of the mantleplume theory.

We also found that it had a seismic tomographic heat-tube structure. As discussed above this is a must, but not a definite for being a mantleplume. We do find though that the tomography suggests a much deeper origin (2500km) than what is evidenced by the inclusions, this is though trivial since Breddams work is not as such excluding a deeper origin, it just states a minimum depth origin.

We also find problems in the seismic tomography data in regards to Hawaii being a true mantleplume. Interestingly enough, no plume denialist has ever denied that Hawaii is a mantleplume, and still, it is the most likely spot to not be a real mantleplume. I mean, at least one of them should be able to read a seismic tomography plot, no?

Do note, I am not stating that Hawaii is not a mantleplume. I am though stating that it seems to lack the second component needed to be counted as such. Anyhoos, back to Iceland.

When looking at inclusions in the Kistufell samples we find evidence of an origin deeper than 670 kilometres. We find that this means that there is indeed material moving up from depth that is constrained into a tubelike conduit or vortex.

Incidentally, even though it was outside of the scope of this discussion, we find that the mantleplume under Iceland is young (14.4 million years) and that it formed at the surface and is burrowing downwards. We do though acknowledge that there are seismic tomography indications that it may be older and may have deep origins. More studies are indeed needed.

Carl Rehnberg


Sources (among many, but I am limiting myself to the three most relevant)

Kistufell: Primitive Melt from the Iceland Mantle Plume | Journal of Petrology | Oxford Academic (

Mapping out the conduit of the Iceland mantle plume with helium isotopes – ScienceDirect

Mantle plume tomography – ScienceDirect





660 thoughts on “Iceland: Plumbing the Plume

  1. Back to Iwo-Jima, your NDVP #1 candidate, if a similar event were to happen there is it possible that the tsunami wouldn’t be as destructive as thought? Granted Iwo-Jima has a bigger magma chamber, and the coastal shelf is fairly flat.

    • Iwo Jima coud be almost Sakursunarvatn sized If it descides to go really big

    • Hunga Byebye had a reservoir of a few cubic km, Iwo-Jima is a couple of orders of magnitude larger.
      Bigger chamber, bigger boom.

  2. Map I made of the volcano, showing where most recent eruptions were.

    The yellow octagon is about the size of a crater with a radius of 1 km. Albert before said the resulting crater would need to be at least 1 km wide to be a VEI 5, though a 6 would need to be a 3 km wide crater. So this would reflect a sizable 5, be about 2 km3 of material blasted out. It looks like it would take something way bigger to completely destroy all of the island, would need to be a crater of at least 6 km diameter to encompass all of the above sea level parts, which would take over 50 km3 of material to be ejected and make it a VEI 7.
    I think this is simplifying too much because eruptions are not really only explosions for the most part but still, point is the majority of the island is still going to be there. However, it is likely most of the tuff cone that was made over the last month has been destroyed along with half of the 2015 cone and maybe a bit of the northeast island.

    • Agreed. Seems to be corresponding to the films made yesterday.

  3. I think this is very interesting although I do not understand how they got to insinuating a caldera. This is a long row of volcanic cones, mostly submarine, stretching from that structure in the Australian Plate across Samoa resembling a shoulder blade all the way down to New Zealand (North Island), a typical volcanic row parallel to a subduction zone.
    White Island is possibly the last oceanic one, and then this continues into the Taupo Zone.
    It is also logical that there must have been volcanism in the whole stretch after collision and then subduction had begun, albeit not always in the same location.
    So this is a typical subduction-caused volcano, one of many, mostly submarine.

    Whakaari had an earthquake in the NE two weeks prior to the eruption, and this time there was an earthquake of 4.9 on Jan,11th, NE of Tonga.

    So, this could be altogether similar to the eruption of Whakaari 2019, albeit bigger than that one, and very typical for the whole area and – it seems – very sudden which also continues into North Island:

    Better though being uninhabited plus no tourism. Plus no pink terraces 😉

Comments are closed.