Edge Driven Convection: BoB’s Back-Story And Malcolm in The Middle.

Cumbre Vieja, La Palma, looking south, Teneguia is hidden beyond San Antonio (huge crater, top centre), 70 kms beyond that is the North coast of El Hierro:


(Sorry, I don’t have credit for this picture… found it via Google though…looks like an old postcard)

Whether you are new to volcanology (like me) or an old hand; if you have been following the BoB happening, you will hopefully be familiar with the work of Juan Carlos Carracedo. A brief biography may be found here: http://www.cienciasmc.es/web/biografias/juancarlos_carracedo.html

He is, pretty much, the foremost authority on the volcanology/geology of the Canaries, and he has (apparently) had a proper row with Nemisio Perez about BoB!!!

What follows is essentially a digest of some of Carracedo’s work; anything in bold is wild amateur schpeculation on my part.

Before we go into the Canaries themselves, a few points about the Hawaiian Islands, which may be considered the archetypal intra-plate, hotschpot driven, volcanoscheanic island chain (iphsdvoic)…They are located approx 2000 miles from the coast of North America and are resting on thin, fast moving Pacific Ocean crust. The Big Island is ongoingly active; the last confirmed eruption on another island was Halekala on Maui (approx 50kms from Big Island) which occurred sometime between 1480-1600. Kauai (580kms from the Big Island) is long since extinct and eroded/ subsided to a nub. The Hawaiian archipelago shows an orderly progression from youngest and most active volcano to oldest and most extinct.

Bathymetric map of Canaries and Madeira. Thick dashed lines indicate possible hotspot tracks:

(Hoernle and Carracedo, 2008)

The Canaries are approx 100kms (at their closest) from the African continental crust and are resting on thick but slow moving Atlantic Ocean crust. (Hotspot volcanism does not occur if the crust is too thick and/or moves too fast.) Forming a near perfect curve, they emerged in sequence starting with Fuerteventura which went sub-aerial approx 25 million years ago, then Lanzarote, Gran Canaria, La Gomera and Tenerife (approx 7.5 million years ago). La Palma and El Hierro are the youngest of the islands; they surfaced between 1.2 and 2 million years ago. The hotspot is somewhere between La Palma and El Hierro and the African plate is moving northeast and rotating “anticlockwise”. Incidentally the Canaries show little or no subsidence; the lower (older) islands are that way due to erosion alone. So far, so normal for an iphsdvoic…

Delving deeper and comparing with the Hawaiian iphsdvoic shows that the Canaries actually have an entirely different pattern of active volcanism:

Look again at the sequence in which the Canaries were formed; Fuerteventura before Lanzarote, La Gomera before Tenerife…

Also note in the map above; Salvagens seamount is way off the postulated hotspot track as is Lars, as indeed are Gran Canaria, Fuerteventura and Lanzarote…

El Hierro and La Palma formed (and are forming) “simultaneously”; despite being as far apart as Tenerife and Gran Canaria…

La Palma has had lots of historic activity (as may be expected), but so has Lanzarote (400 kms from the hotspot); there was a devastating eruption (though with few, if any fatalities) from 1730 to 1736 and a minor one as recently as 1824. …

El Hierro, apart from BoB has had no confirmed historic eruptions; there was a seismic crisis in 1793 but it is not known if there was an associated (presumably undersea) eruption. The most recent (proven) eruption was in approx 500BC.

Tenerife has experienced several eruptions since the beginning of the conquest in 1402 and a seismic crisis in 2004…

There is an undersea volcano 500m high at a depth > 2500m between Tenerife and Gran Canaria. Volcan Enmedio (thanks Judith! x) is in the vicinity of the strongest EQ (5.2mag, lat27.9 long-16.2, depth 36km, intensity V, 9/5/89) detected in the Canaries. This area also has regular EQs; the last one was on 10/5/12, depth 21km, 2.5 mag… The research I have seen (Carracedo) suggests that there wasn’t an eruption as such…

However, that 5.2mag as well as approx 50 EQs > 2.5mag in and around that location between 3/89 – 6/89 is perhaps quite suggestive… Imagine how many “tiddlers” were not detected!!!

Picture/ bathymetric map of Enmedio:


(Credit as in picture)

So why don’t the Canaries show an orderly progression like the Hawaiian Islands? Carracedo asserts that edge driven convection is the reason.

Edge driven convection diagram:


(Juan Carlos Carracedo, 1998)

The hotspot warms and “circulates” the nearby athenosphere. This warming and circulating doesn’t amount to much (it gradually cools and sinks) except where the circulation goes east; and runs into the African continental crust (approx 500kms from the hotspot). When it hits the cold crust it cools and sinks; this in turn pushes more of the athenosphere up and past the plume “driving” the convection.

Hopefully we can see how edge driven convection could explain some of the “anomalies” in the mantle plume model of the formation of the Canary Islands. Specifically; the historic eruptions at sites far removed from the hotspot.

So what could account for the geologically simultaneous formation of El Hierro and La Palma and the seemingly “off track” formation of Salvagens, Lars, Gran Canaria, Fuerteventura and Lanzarote? Carracedo says that this may simply be a function of the thick, slow moving plate. He has a further theory as to why La Palma and El Hierro are evolving simultaneously, albeit with apparently alternating eruptive periods. He suggests that changes in the tectonic stress regimes caused by large gravitational landslides are the explanation. I haven’t found a paper which goes into this, but a short discussion can be found at the end of the Giant Quaternary paper and in the book referenced below.

Carracedo also discusses the idea that edge driven convection is more “chaotic” than the diagram implies i.e. it’s not just a simple “loop”; slow moving as it is, there may be some turbulence in the system.

This got me thinking about possible causes of turbulence:

Eruptions have diverted energy from the system, as have the conduits and chambers under the islands.

There are an (increasing?) number of islands resting on the oceanic crust; which distorts it.

The African continental crust is not a perfect rectangular slab.

The hotspot is not a perfect “tube”.

Some (but not all) of the currents will interact with the continental crust: Those that do will interact with an uneven crust at varying distances from the hotspot, having passed a variety of “obstacles”.

The convective currents themselves (given the above) presumably interact.

The chaos in the system could therefore be “perturbing” the mantle plume itself… Eruptive action has moved from the southern tip of La Palma (possibly via Malcolm Enmedio) to just south of El Hierro (which is about 100kms) in only 40 years. This wild theory of mine may also go some way towards explaining the “anomalies” in the hotspot model of the Canary Islands discussed above.


Conclusion and discussion:


The Canary Islands and their volcanoes are complex and enigmatic, they’re not in Iceland’s league, but for an iphsdvoic they have a certain, as yet unresolved mystery about them…


I’m hoping that this post will raise as many questions as it answers and provoke as many arguments as it settles and that those questions and arguments will get answered, debated and settled (and so on and so forth!)…Thanks for reading.



P.S. Big thanks to Lizzie for all her help x.

References and further reading:


Hot spot volcanism close to a passive continental margin: the Canary Islands:

Los Volcanes de las Islas Canarias IV; La Palma, La Gomera, El Hierro. Juan Carlos Carracedo, Editorial Rueda S.A, ISBN 978-84-7207-190-2

Growth, structure, instability and collapse of Canarian volcanoes and comparisons with Hawaiian volcanoes:


Giant Quaternary landslides in the evolution of La Palma and El Hierro, Canary Islands:


Youngest lava flows on East Maui probably older than A.D. 1790:


http://www.ign.es/ign/layoutIn/volcaFormularioCatalogo.do is the source of EQ data.


104 thoughts on “Edge Driven Convection: BoB’s Back-Story And Malcolm in The Middle.

  1. Very nice introduction, Schteve, well done! Since a “hot-spot” is not a pin-point source but rather a wide and diffuse plume, isn’t it possible that KarenZ video shows the width of it and where/when certain areas of it became active?

    • I wondered about that. Don’t know whether the plot showed the exact / entire width but it shows a larger area of EQ activity which implies a larger plume. A large plume might go some way to explaining the harmonic tremor on other islands, when CHIE appears to be relatively quiet.

      Rifting has played a part in volcanic activity in the Canaries which also has to be built into the model.

  2. If you take a “full pull” of all the quakes available from IGN, you will find that there is a trend of quakes to be deeper off to the North West.. tangential to the proposed hotspot track.

    Another competing meme to this, is the progression of the hitspot leading to dual active areas similar to Hawaii, with parallel eruptive centers. One article that I read proposed that the Canary hotspot had recently (geologically) taken up that behavior and that La Palma and El Hierro were the two sides to this new track. But… there is a seamount further to the South West that goes against that scenario since it would like up akin to La Gomera, right in the middle of the track.

    But as far as what goes on there, Carracedo is probably the one most likely to get it right. I’ve read some of his work and it’s very insightful.

      • we can not say that la Gomera will be extinct forever, many other canary islands have had also inactive periods in the past, many of them also lasting several million years, so could la Gomera someday (geologically speakig) wake up again? nobody knows…

      • Also note… “Jackson Volcano,” eroded flat and overlain by sediments over several millions years (more than 60 mya) would by all definitions be considered “extinct.”

        But in that sediment, you can see evidence where it has uplifted at some time after the sediment was laid down. That means that even after it was dead and eroded flat, there was still some life left in it.

        “Extinct” is a relative term.

        • yes you’re right, what I wanted to say is that all canary islands experience this so called erosive phase, after their main shield stage, after the erosion normally one last big edifice starts to grow, could one day a Teide or a Roque Nublo start to grow on la Gomera? hahaha certain is that I wont be there when it starts.

      • The Canaries are different to the Hawaiian islands. Apart from Gormera, all volcanoes could be considered active. With some erupting fairly recently (Hierro, 2011-2012; La Palma, 1973), and others with cinder cones and lava flows so young-looking (e.g. Gran Canaria) that they must haave erupted just before the spanish conquest of the islands.

          • Just think about it for a while…
            Do you really want to know what the Conquistadors did to the dogs? Considering Perfidio and Nemesio…

  3. Schteve ! What a thought provoking post. Certainly you have given me plenty to read. My first reaction is that at last here is a possible explanation for as you put it, the”Chaotic” behaviours of this hotspot. Also it makes me want to look more carefully below the sea in a few other areas. The canaries are so close to the African plate it makes sense that it affects the crust at the edges.
    Job well done there!

  4. I have had this for a while because it was saved as a draft, but i want to express my thanks here too Schteve and congrats. Very good first post.

  5. Hi everyone. Is there anything known about the rising speed of hot spots in general? Or how it compaires to mid ocean spreading speeds (20-150mm/year) ? In all drawings i always see them coming up perfectly vertical. If you look at the canary islands that seems rather unlikely considering the north east movement of the oceanic crust. You’d expect the eastward movement of the asthenosphere to deflect the plum. This defection might be enough to explain the residual vulcanism. Also if the plum hits the crust at a more shallow angle it seems much more likely that it can surface on a wider area, like now la palma en el hierro. Especially considering a bit of turbulance and an uneven crust or moho thickness causing more deflections/turbulance.

    Just a wild idea from some one that’s interrested in the matter.

  6. You guys are pyschic, really.

    Just a few days ago I was thinking about a model of the Icelandic hotspot and it was just like this model. The plume coming upwards under Vatnajokull and deflecting more horizontally under as we progress from Katla/Hekla, Reykjanes/Hengill and Snaefellsnes.

    Likewise for Hawaii or Canary Islands.

    My question is why is the Madeira hotspot not active now?

    Second question is having a second mid rift hotspot, Azores, after Iceland. Can these plumes be driving the rifting itself?

    • Sort of depends on your definition of Hotspot. Ultimately, yes. By pushing the crust along. Once it gets into a situation where other forces can act upon it, the “Slab Pull” scenario seems to generate more driving (pulling) force than Slab Push.

      I made some gravy the other day, and sat in amazement as I watched the surface of it start to congeal, drift to the center and collide with other congealed pockets on the surface… all driven by the upwelling from the heat, but eerily similar to the crust movement on the earth. Take a look at some video of lava lakes (specifically 7:00 in this one) and you can get an idea of the dynamics in play (in miniature). Whether upwelling magma or sinking crust is the driving force is a matter of scale and heat flux.

      Now.. as for hotspots, it depends on your definition. In the paper “Antipodal hotspots and bipolar catastrophes: Were oceanic large-body impacts the cause?” by Hagstrum there is a compiled listing of about 122 “hotspots”.. some primary, some secondary, depending on how well accepted that particular hotspot is. His list is derived from several sources, his “primary” list coming from the intersection of about seven different lists put out by various researchers.

      Click to access Antip_hot.pdf

      And here is how his list plots out.

      So… pick your hotspot.

      • I am really into sea mounts and ocean floor rifting at the moment. I am looking forward to see if there will be any exploratory expeditions that will have live feeds this year. I so enjoyed those on the Axial sea mount and the Nautilus voyages last year.
        Nautilus will be back at sea next month.
        With both ships, the ROV’s took samples and explored various volcanic features. I thoroughly recommend an hour or so beneath the oceans. It gives not only a whole new perspective of our world but also I found the experiences exciting but I am sure my blood pressure dropped a few points whilst sitting watching the ocean floor pass before me.
        There are some amazingly complex fracture zones in the far South Atlantic which have only recently been discovered.. Here is an introduction to undersea Volcanic research, this time in the South West Pacific. It is not difficult to associate what you see in this video to what is happening off the canary islands.

  7. Morning, A guy named Frank posted a Youtube video of a flight over the south of Iceland on Jon Frimanns Blog. The names of the volcanoes are added.

    Is Etna at 2.01 just to test if we pay attention?????

    • Entajökull, the NW outlet from Katla. I believe it was mentioned a couple of posts back, so yes, it could be seen as a test of attention. 😉

    • from that vid I screen captured this image..

      (i have that as my desktop now. lol)

      and edited it to bring out a pretty cool image of what looks like a round crater, with what look like cauldrons inside it..

      this is a slow gif of the two of them..


  8. Creamy soups work well too 🙂 Just let it simmer for a while without the lit on and you can see continental crust being formed with a central seabed subducting underneath it.
    It doesn’t look very appetizing though.

      • Jam during the making also works well if the white frothy scum is left on! That’s Jelly to all you in the USA.

  9. There is as lot of discussion on Avcan FB regarding the latest earthquake in the North of Italy and all the aftershocks that followed just noticed this comment from Javier Hidalgo.

    ,, What really scares is the depth of the singulars of some only 5 km away, so they are very shallow. Why wonder when we had to wait or to know if the process is tectonic or data on the contrary are facing a new volcanic system that is developing in this area of Northern Italy. Also a few days ago Mount Vesuvius presented signs of activity

    • I visit the tremor graphs around the Bay of Naples at least once a week. This morning in the early hours the signals from Pozzuoli station started to look rather alarming. I checked back to other early hours and yesterday I know it was Sunday so no working. Apart from a major volcanic tremor which I do not think this i,s all I can think of is some sort of constant drilling. Has anyone any information please? The Vesuvius always show heavy traffic and railway activity during the week but Pozzuoli is usually much more quiet.

    • Funny that. Every focal solution I’ve seen shows reverse faulting and not one geologist has complained about an “intrusive” signature. (Yeah, it would stick out like a sore thumb) And this guy tries to tie it in to Vesuvious… (groan)

    • The northern Italian quakes are tectonic only.
      The small activity at Vesuvius is normal activity, presumably garbage trucks that illegaly dump garbage in smaller cones and craters on the flanks. There was a stunning french documentary about the environmental catastrophy that is happening around Vesuvius due to industrial waste and garbage being dumped by the Commora (Neapolitan maffia) into the volcano. It aired a week ago on Swedish Television.
      Running garbage trucks up a volcano gives a “wee” bit of tremor.

      • As usual, the only reason that the locals protested was not the uptick in cancers, nope, it was due to the wine being judged as toxic and not being allowed for sale inside the EU…

      • What a shame… …just imagine the media furore if they weren’t. 😈

        I’ll say one thing in favour of the Mafiosi – unlike politicians they don’t pretend to abide by the laws and discharge their duties towards their dependants scrupulously.

  10. The eathquake swarm at BOB may be continuing there have been 5 EQ already today and 17 in total since the 2nd June.

    1146541 04/06/2012 07:18:05 27.8262 -18.0778 15 1.9 4 NW FRONTERA.IHI [+] info

    1146551 04/06/2012 07:31:45 27.7878 -18.0796 17 1.0 4 NW FRONTERA.IHI [+] info

    1146553 04/06/2012 07:40:04 27.7561 -18.0863 18 1.2 4 W FRONTERA.IHI [+] info

    1146556 04/06/2012 07:43:30 7.8128 -18.0967 16 0.8 4 NW FRONTERA.IHI [+] info

    1146568 04/06/2012 08:40:29 27.8123 -18.0826 12 1.1 4 NW FRONTERA.IHI [+] info

    • By now the amount of earthquakes should have subsided quite a bit.
      Instead it is getting a bit more active again. Could perhaps maybe be a slight and small sign of new magma that perharps maybe is arriving from the depth.
      But to be sure one would need accurate GPS-data, and that we do not have thanks to Perfidio.

    • when bob started to make its presence felt, I had a look at last years EQ, a repeat ? but this time a closer to the surface start, but then which one of the cone’s is it this time?

  11. Six more to add to the list for BOB this morning.
    1146597 04/06/2012 09:23:20 27.7188 -18.1223 13 1.5 0 SW FRONTERA.IHI [+] info
    1146603 04/06/2012 09:28:58 27.9276 -18.0727 10 0.8 4 NW FRONTERA.IHI [+] info
    1146609 04/06/2012 09:29:23 27.7486 -18.0378 16 0.9 4 W FRONTERA.IHI [+] info
    1146611 04/06/2012 09:29:51 27.8026 -18.0504 16 1.4 4 NW FRONTERA.IHI [+] info
    1146612 04/06/2012 09:32:32 27.8060 -18.0390 15 1.4 4 NW FRONTERA.IHI [+] info
    1146617 04/06/2012 09:40:46 27.8000 -18.0678 16 1.0 4 NW FRONTERA.IHI [+] info

  12. Avcan are discussing a possible new induction of magma as reported on their Avcan FB Page.

    If we see the earthquakes in the graph of evolution or the energies liberated seismically accumulated straight, seems like it wants to start uploading again, although at the moment…. timidamente will know in a few days… Since tendra that climb the magnitude of earthquakes if we are going to have something of importance. In addition, the truth is that yesterday and today activity is more concentrated on the map, both the vertical and the horizontal and today as I start to lean by I can be having a small intrusion of magma at depth (Henry)…


  13. Good Mornin’ All,
    I’ve just checked in, I’m glad you liked the article 🙂 I didnt know it was going to be posted and as it happens, it’s a lovely birthday surprise for me!!!
    @ Spica: I didn’t tell you my birthday, did I?
    I’ve done a couple of plots and added to the reference list since I wrote the article, but I’m not at home so those will have to wait…
    Thanks for your kind words and interesting comments x

  14. I’m no expert but is this theory correct?
    I was under the impression that under the continents the magma wsa warmer than under the (thin, new) oceanic plates. Thats one reason they are high.
    So I would suggest reversing the argument. Hot magma from underneath a (non-subducted) continet is spilling out around the edges. Elsewhere its breaking out inside (eg african rift valley).
    Seems more plausible to me, anyway.

    • Here we are talking about a hotspot driven convection current.
      A hotspot is a spot anchored at depth that delivers hotter material up to a spot under a crust. The hotspot is stationary compared to the individual plates that slide over the hotspot, that is why it often leaves a tell-tale track of islands if it is out in the ocean. Examples of this is Canaries and Hawaii. Also Iceland is hotspot driven.
      When a hotspot hits a rift (like the MAR) it can cause very increased volcanism, examples of this is Iceland and the Azores.
      There is nothing pointing to there being a higher temperature under continents than under sea-crust.
      Hope that this helps a bit.

  15. Indeed, however as stated the chronological trend (or lack of it) doesn’t look like a hot spot at all.
    That would have a line of islands (or seamounts) each of which was older then the next and where eruptions went (give or take) from one to the next. Anyone who has seen Lanzarote would not class this as ‘old’. Further IIRC africa is heading westwards here so the plate movement is all wrong.

    • Hello Farmeroz!
      Nope, Africa is not heading westward here. And regardless this is on the eastward going part of the spreading atlantic crust. To further complicate it we have the northward trend of the african plate (see the alps and the apennine subductive volcanic activity), and hey presto, then you have that oddball little trail in clear logic.

      The clear trail that you are looking for is only achieved after the fact, in this case there is an ongoing process. Lanzarote had it’s latest erution 1730-36 if memory serves. So, you would have to wait untill it all has died off, and that will take a bunch of deca-millions of years. Then you will most likely have the result you are looking for.

      Personally I think that the ambivalent nature of where the eruption occur has to do with the carracedo-effect in conjunction with the duality of motion in the tectonics.

      Also, you should know that the hotspot is not travelling over remnants of both MAR volcanicity and residual volcanic activity from when Africa (moving eastward) left the westgoing south america. So, it is no mean feat always to know which seamount belong to which volcanic process.
      For instance, south to southwest of El Hierro there are two seamounts. One that belong to the Hotspot (next Island forming probably). And one that belong to a stonedead volcanic line of African separation origin. Above you have a seamount of MAR origin…

  16. Thanks for all the friendly words to me! on the other blog (article) Its good to have people around you who understand this tragedy and have sympathy with you . Wednesday is the funeral of my neice (46) and her husband(51 years). A nice day to all of you. Deanne

  17. Bob is showing some activity today 5 earthquakes today so far, 3 of them with magnitudes between 2 and 2.9.

      • yes, my mistake, Iooked at ign list of the last ten days, I forgot that they only put the ones of magnitude 1,5 or more in that list XD

  18. thats not so good charly for the people over their when you see the link of the hotel cam and you look at the sea i think that is not completely normal over their yesterday evening when it was dark their their were some lights on sea but it can be also the boats im not an professional take a look for yourself but you can see the mountain by frontera http://www.hotelidaines.com/frontera-webcam/

    • Thank you for the link. Can’t see the sea very well due to reflected sunlight at the moment but there is quite a shake on that camera at the moment.

  19. earthquake2012-06-04 14:19:53.0
    13min ago 36.94 N 28.16 E 6 mb 4.7 DODECANESE ISLANDS, GREECE 2012-06-04 14:32
    earthquake2012-06-04 14:05:55.0
    27min ago 36.93 N 28.24 E 7 ML 3.2 DODECANESE ISLANDS, GREECE 2012-06-04 14:22
    earthquake2012-06-04 14:00:52.0
    32min ago 36.94 N 28.22 E 8 ML 3.2 DODECANESE ISLANDS, GREECE

  20. I just had the rather disquieting thought that if large landslips occur at Hawaii and the Canaries as they do, then why not also in Iceland?

    Anyone know if anyone has looked for any?

    David B

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s