Terrae Nova: the Eye of Africa

Richat's structure: the Eye of Africa. http://www.okwave.com/arigato/en/posts/29583

Richat’s structure: the Eye of Africa. http://www.okwave.com/arigato/en/posts/29583

The previous post on Venus described a peculiar type of volcanic construct called a corona, unique to Venus. This blog is read by knowledgeable people with a somewhat critical attitude to authority, while at the same time having a strong respect for experts [For UK readers: see footnote], and this statement was immediately questioned. A structure was uncovered on Earth with showed a remarkable similarity to Venusian coronae: Richat’s structure, also known as the Eye of Africa. Could this be an interplanetary corona? A Venusian refugee hiding in the Sahara?

To deal with the most important point first: the word corona comes from Latin, and means crown, or garland for those crowned with fewer financial resources. I couldn’t take Latin at school, as that was on offer only to the brighter buttons with linguistic ability (my strength was more on the mathematical side). But if you can’t trust a non-expert, who can you trust? As a Latin word, the plural of corona is coronae, unless the word is used in the accusative, when the plural becomes coronas. Thus, you can say coronae are beautiful (nominative) or This post discusses coronas (accusative). In the dative, the plural is coronis: Plumes give coronis their rings. Luckily the English language has lost these distinctions (an echo still remains in the word ‘them’). The grammatical complexity may be why that show is called Games of Thrones rather than Crowd of Crowns. It may be best if readers of this blog don’t go looking for further coronas on Earth: while there is only one (suspected) corona, coronal grammar is a lot simpler. I will use ‘coronae’ for multiple coronas.

(Don’t be fooled by some English words. ‘Conundrum’ is a made-up word, so its plural is ‘conundrums’, not ‘conundrae’. ‘Octopus’ is from Greek, and its proper plural is octopodes (not octopi – it isn’t from Latin) but this is so awkward in English that it makes even ‘octopusses’ sound acceptable.) (And no, I do not know how English lost the plural of sheep.)

Coronae – the story of Venus

Venusian coronae were first seen in 1977, in radar images taken from Earth. Valeri Barsukov gave them their name and first identified them as volcano-tectonic features. Coronae are circular or oval, 75-2000km diameter, with concentric and radial fractures around a central depression with an inner plateau. Their interior can be above the surrounding the plain or below it. Artemis provides a spectacular example: 2100 km across, and criss-crossed with fractures. The numerous fractures may be magma dikes. Several ‘small’ volcanoes occur inside the corona, but don’t be fooled, the largest of these is only a little smaller than Mauna Loa. In between the double ring fracture is a deep trench, reaching 4 km below the surrounding plain. It is a funny feature, almost like a circular continental rift.

Artemis Corona, Venus. Source: NASA.   Click to enlarge to full glory (beware of data allowance): note how much more difficult is to see the corona at full resolution

Artemis Corona, Venus. Source: NASA. Click to enlarge to full glory (beware of data allowance): note how much more difficult is to see the corona at full resolution

3d structure of Artemis

3d structure of Artemis

Plume models

The rift-like structure and the large size suggest that Artemis sits above a mantle plume. A model for this was presented by Teras Gerya (Zurich). It assumes that a mantle plume begins to melt into a thick crust and sets up convective cell within the melted crust. This convective cell has an up flow in the centre, and a down flow around the edges. The cell is as wide as the plume head, which can be 1000 km across or more. The upflow pushes up the surface over a very large central area. The down welling gives a depression or trench around the edges. The stress causes extension, and fractures develop, filling with the crustal melt. The main melt stops below the surface, but some shield volcanoes do develop.

Once the plume cools, the central area drops down again. The model does not address the question why more of the magma does not reach the surface: when do you get a corona with some volcanoes inside, and when do you get flood basalt covering half the planet?


A model showing a plume causing crustal melt and a corona. Taken  From Tara Gerya:  http://www.ises.su/2013/pdf/03092014_Gerya.pdf

A model showing a plume causing crustal melt and a corona. Taken From Tara Gerya: http://www.ises.su/2013/pdf/03092014_Gerya.pdf

Nova model

Nova model

The model also can create a structure called a nova. A nova is a star-like pattern on the surface of Venus, with numerous radial cracks extending from a small central rise or depression. In the model, these are the early phases of the formation of a corona. I’ll come back to this later.

Other models have been proposed. Early on, it was thought that coronae were impact craters. Their sizes are similar to the maria on the Moon, for which an impact origin is well determined. But that was very quickly abandoned when better images showed a raised inner plateau, complex tectonics and internal volcanic features. Impact craters also do show at most two concentric rings. An indirect relation to impact craters has also been suggested, where an old large scar has weakened the crust allowing plumes to push through. But impact craters would be randomly distributed on the surface. Coronae are not.

It has been suggested that the outer trench is an arc-like subduction zone, making coronae similar to island arcs on Earth. A Venusian Carribean. But radial fractures often run through the coronal trenches, and that would not happen if they were subduction zones.

Ring-dike intrusions have some similarity to coronae and it has been suggested that coronae are oversized ring dikes. But ring dikes do not show the outer depressions that many coronae show.

Drifting plates and the battle against mantle plumes

This leaves one question unanswered: why coronae are so common on Venus (513 are known) whilst absent on Earth? Mantle plumes happen on both planets, after all, and Earth has some thick crust in the old cratons. Kimberlite plugs form when mantle melt punches directly through the crust, and these are not uncommon on Earth. Are Kimberlite plugs micro-coronae? Or a mini-nova?

An interesting suggestion is that on Earth, continental drift stops coronae from forming. The idea is that a corona needs a long time to form, perhaps 100 million years. In that time, Earth’s continents move and the same point does not stay above the plume. Only Africa is currently near stationary, and that has only been since the last 30 million years. Instead, on Earth a mantle plume develops into a hot spot trail. Put the plume underneath the same bit of continental crust for 100 million years, and perhaps you would get a corona. If only these continents could stay put long enough.


Pluto's moon Charon, as seen by New Horizons. Click to get full resolution - be aware, it is huge.

Pluto’s moon Charon, as seen by New Horizons. Click to get full resolution – be aware, it is huge.

Are coronae really unique to Venus? The New Horizon images of Pluto have shown an amazing scenery, but the images of its moon Charon have not received as much attention. Right at the rim of the main image of Charon, a deep long chasm is seen, perhaps 10 km deep (it is hard to tell) and 700 km long. It may well be round: it curves behind the horizon. Could this be a corona? It has the right size, and the visible part has the right structure. Too much is hidden to be sure, though.

The Eye of Africa

Richatt’s structure, 40 km diameter, is located in the sand deserts of Mauritania, in the western Sahara. It was discovered from space in the 1960’s. The ring is too large to be easily visible from the ground: you can stand inside it and not see it. It is hard to recognize that a low, 10 meter wide ridge is a circle if you can only see a small part, and it stretches beyond the horizon. The largest impact crater in Germany was also only discovered from above, although there is a town in it.

Richat's structure

Richat’s structure

The 100-million year old structure consists of three rings, dipping outward. The main rocks here are sedimentary (the blueish tint is salt). Two of the rings are mafic ring dikes, which seem to have formed from the same magma chamber. The inner mafic ring has a gap, caused by a maar explosion which ripped through it. The outer ring has a slight offset in the north due to a fault. There are volcanic rocks, but the main structure is not volcanic and the volcanic activity was only near the centre. The remnants of two craters are here, each about 2 km diameter, on either side of the centre. The craters are largely buried under the sand. Both are considered maars; they formed after the main structure. The rocks in the centre show that they have been extensively affected by hydrothermal activity. To the north, just outside the outer gabbro ring, is a kimberlite sill. It has the same age as the main structure (give or take a few million year) but there is no clear relation. The central area has a large number of fairly short, radial dikes, containing carbonitite.



The structure is considered an uplifted but eroded dome. It formed 100 million year ago, around the time the Atlantic Ocean began to open. The area has multiple sedimentary layers, mostly marine from the continental shelf but some formed above water. The layers are vertically offset near the centre, indicating a collapse has occurred there.

>From The 'eye of Africa': An isolated Cretaceousalkaline-hydrothermal complex Guillaume Matton, Michel Jebrak, 2014,  Journal of African Earth Sciences

From The ‘eye of Africa’: An isolated Cretaceousalkaline-hydrothermal complex
Guillaume Matton, Michel Jebrak, 2014, Journal of African Earth Sciences

So what happened? About 100 million year ago, heat from below causing doming, causing the various sedimentary layers to bulge up. Two ring dikes formed, filled with mafic mantle magma, supplemented with some crustal melt; the ring dikes stayed underground and these did not break the surface. The radial carbonitite dikes probably formed next, from melted sediment mixing in with the magma, but these also remained below ground. Water percolated through and a phase of hydrothermal activity began. The heated water dissolved much of the sediment, especially the lime stone, and caused collapse of the central area. The two maars also dated from this time, and this may have been the first volcanic activity on the surface. Later, the remnant of the dome underwent deep erosion. The harder layers survived better and now stand above the softer sedimentary layers. Sand filled in the craters, but left some (but not all) of the rings visible. It is almost perfectly circular because the dome was beautifully symmetric, and the sedimentary layers were undisturbed and horizontal, a beneficiary of Africa’ solidness even while the Atlantic was opening. The role of the kimberlite sill is a big unknown in the process: was this related, or was its appearance at roughly the same time a coincidence?

Terrae nova

The evidence for doming in the Eye of Africa seems to set it apart from the coronae. But it is akin to another type of volcanic feature on Venus, closely related to coronae. Novae are large domes with numerous radial fractures, and a central depression. About 60 are known on Venus. In contrast to coronae, they lack concentric rings. Typical sizes are 50-100 km but some are larger.

A nova on Venus. Note the star pattern and absence of concentric rings.

A nova on Venus. Note the star pattern and absence of concentric rings.

Novae are considered an early form of a corona, something that may continue to develop into a full-blown corona, or it may form when the plume is smaller or the crust thicker, preventing the formation of a full corona. Novae don’t look like Richat’s structure because of the lack of rings. But take a nova and apply erosion, something that does not happen on Venus, and the result could be revealing. The erosion could remove some layers, leave others, causing the tilted layers to form concentric rings. This is what a nova on Venus may have looked like had the planet kept its water. Thus, Richat’s stucture may be a nova on Earth: Terrae Nova.

Earth’s corona?

Is Richat’s structure a corona? No. It is too small, and it formed mostly from sediments, ring dikes, and erosion. There are similarities and both did form from volcanic heat from below which failed to reach the surface, but Richat’s structure never developed into a corona. Continental drift may be to blame. The structure is more akin to a Venusian nova, a dome which failed to become a corona.It is a beautiful feature, the biggest eye on Earth. But it ain’t Venus.



[1] Footnote for UK readers: exactly the opposite to Michael Gove, who favours authority but told the electorate they shouldn’t believe experts. In his defence, he was minister for education.

More reading

  • Geology
  • Coronae on Venus
  • Mantle plumes and coronae
  • Video report (in French but worth watching even with the sound turned off)
  • Description of Richat’s structure (not easy to read)
  • Novae
  • 119 thoughts on “Terrae Nova: the Eye of Africa

      • Scratch that. Microliths are real. (Tool flaking debris)

        Millilithic for the sake of humor.

        • Hm. Even megaliths are much smaller than this. Perhaps this should be a Gigalith?

        • Heh… with regards to overall size, yep.

          But I was mainly referring to the size of the component parts.

    1. Reading the article, I looked for a good overview of Oregon Maars and came across this very good paper on central Oregon volcanics.
      : ntheplaygroundofgiants.com/geology-of-central-oregon/geology-of-the-oregon-cascades/
      The opinion of the writer on our local galloping hotspot (now Yellowstone)
      is interesting.
      I had no idea the “Eye of Africa” was so large..

    2. Has anyone noticed this yesterday? A deep one at Oraefajokull.

      03.07.2016 01:16:11 27.7 km M1.9 99.0 3.8 km WSW of Hvannadalshnjúkur

      • I read somewhere that under Oraefajokull subduction could be occuring of a leftover of continental microplate. The crust is quite deep there.

        Still this deep quake could be related to magma movements.

    3. “Corona Special” is my favorite and needs to be served cool. These coronae and novae are also served in a cool way! Thanks Albert for another interesting post!!

    4. What about Cordillera del Paine in Chile? (also a Corona Especial fan.)

      (from Google Maps)

      • Just glacial erosion, or less likely, an ancient caldera basin with a resurgent edifice.

        Coronae would greatly dwarf this.

    5. OT, sort of. Mainly about sheep.


      A measure of distance equal to about 7⁄8 of a mile (1.4 km), defined as the closest distance at which sheep remain picturesque.

      • This looks kinda like a Gjalp thing or I would dare to say even Hamarinn which is inflating relatively fast lately and might be causing stress on the outher layers of its deeper magmatic system. And it could be Grimsvotn too. Its kind of in a triple point zone really. We might have to wait for the official word if there will be one at all.

      • You can see why English lost most of this, and why it is so popular as a second language! Of course you lose the grammar at the cost of adopting a lot of irregular verbs. It took me a while to realize that Earth’s star would not be ‘terra nova’ but terrae nova – assuming I got this right

    6. I think sheep is the plural…and there is no singlar term for both the male and female sex.

      One would not say “This sheep has footrot” They would say “This ewe has footrot” or “This ram has footrot”

      • Another interesting find, Ian. Thanks for posting the link.

        I find it interesting that the drop during the BB eruption looks a bit larger than that during the 2000 eruption of Hekla itself, and the fact that it did drop bears out your thesis about there being some connection between the two volcanoes – if you still hold that opinion.

        Also interesting that this “academic” piece was published at about the same time as Pall Einarsson made his well-publicised (on here) comments about Hekla being ready to erupt. He makes a comment in the piece you linked to that bears out our view that he was issuing a warning partly because it was the beginning of the tourist season:

        “This short warning time is of concern because of the increasing tourism around the volcano. Groups of hikers on the flanks of the volcano cannot count on getting warnings before it is too late.”

        • I think it’s one of the other regulars here that pushed the connection between these two, but when you look at this information it is difficult to ignore, it really does support this.

          I recently read another paper that put forth this speculation, I will have to see if I can dig it up

          The amount of change is really interesting along with the recovery!

          • Thanks for the correction, and apologies for mistaking you for someone else – iirc, their thesis was thought to be rather implausible, but with that graph published there does seem to be cause for further thought.

      • These are fairly small tilt changes which may not always be due to Hekla. The station is 11 km west of Hekla, which puts it close to the main road, on the flood plain (I think). Deeper magma pushes up a larger area. Therefore, as magma moves up, the outer regions lose the inflating effect and start to deflate. That gives the strongest change in tilt near the outer regions which may be what this is measuring. But note that from one station you can’t tell where the inflation/deflation is centred, unless of course it coincides with an eruption as it did in 2000.

        A 10 microradian tilt over a distance of 11 km corresponds to an inflation or deflation at the top of Hekla of 10 cm. (This is a linear extrapolation which is unlikely to be correct but it should give a rough number).

        The decline during the Bardarbunga eruption is a bit deceptive. To me it looks more like a positive jump just before the eruption. The jump upward was 5 microradians. The distance to Bardarbunga is about 150 km, so this would corresponds to Bardarbunga inflating by roughly a meter before the eruption. That is not implausible, so this station may just have been measuring Bardrabunga directly rather than its effects on Hekla.

        In the months after the eruption of Bardarbunga, GPS measurements over most of central Iceland showed a small deflation, probably because of the loss of magma from underneath Bardarbunga. The jump down on this Hekla tilt measurement may just be due to that.

        If you subtract these two movements, what you get is a fairly steady increase of the tilt since 2000. That may well be Hekla. The most recent point is a bit higher and may be what caused the warning to be given out, but it is always dangerous to stress one single data point.

        The bottom line is, be careful with interpreting a single-station tilt measurement. To get a pattern and a centre of action you need more.

        You should also ask the question why their other stations did not show the same movement. The say this one was the clearest. Why?

        • Thanks for the detailed and very informative explanation Albert.

          I found the paper I was referring to, in was written before the 2014 eruption and on page 65 it make reference to the theory of a pressure connection between the systems.


          I would imagine that rifting alone would have a system wide impact on the pressures within individual systems. I see the rift down the middle of Kverkfjoll and Kitsufell has either relieved pressure or maybe increased stress and shut down the systems? These areas went from seismically crazy for years to dead silent after Aug 2104

            • Interesting looking at the comparisons. A few things I noticed.

              1. There is a high concentration of quakes to the west of Öræfajökull close to where the Vatnajokull glacier runs towards the lowlands, draining into the Atlantic. I don’t know of any volcanic systems in this area, this seems a bit odd to me.

              2. It can be somewhat more confirmed that Öræfajökull has started to reactivate in the last year or two, at least tectonically speaking. This does not seem to be entirely related to Bardarbunga in the way that Tungnafellsjokull’s quake activity was. There was very little quake activity in Öræfajökull, and the quake activity didn’t start to pick up until after the Holuhraun eruption ended. With that said, the quake levels are still pretty mild, so I don’t expect anything to happen here soon, but it’s interesting to see a rising trend at the very least.

              3. Tungnafellsjokull is significantly more active, which seems to be a direct result of what happened at Bardarbunga. Once again, this is not anything particularly new, but is interesting to note. Both Tungnafellsjokull and Öræfajökull are infrequent eruptors that can potentially erupt quite explosively.

              4. What is the odd cluster in the 1995-2011 period to the east of Askja? Whatever it was, it has completely died down in recent years.

            • I think Albert is spot on. That inflation was probably BB on top of the background of small Hekla inflation.

              Nevertheless the fissure swarm of BB seems to run southwest towards Hekla, and the Hreppar microplate edge continues along both volcanic systems.

              I think during Holuhraun the Hreppar microplate moved somewhat, to adjust to all other largescale tectonic movements.

              This means that everything is now different across Iceland. Previous tensive areas are now without tension, and viceversa. New volcanoes have been awakened. Tungnafellsjokull and Oraefajokull being two examples.

            • Very informative! Thanks. Curious about the region with many black only spots stretching southwest of the reykjanes ridge in the atlantic. Wonder if these quakes came in a few swarms or if the were spread more evenly over time?

            • I turned on the click for information feature so now you can click on a quake and see the date, depth and magnitude

              if the info is not showing, refresh the page to update the map in your browser.

        • Thank you, Albert, for a very instructive and helpful post. Your concluding question had been lurking in my mind as I read your views, so I am glad you raised it… Intriguing.

          • One possibility which had occurred to me is that they had placed their tilt meters at various position surrounding Hekla. In that case this one may have been the only one which happened to be oriented towards Bardarbunga (i.e. Hekla and Barda were lined up from its position).If the inflation jump came from Bardarbunga, the other station may have missed it. While it was for Hekla, all stations would have picked it up. Just my suspicious mind.

            • Heres the thing, it may mean nothing and the rise and fall may not be joined with anything BB related?

              The deflation ends when the eruption at BB starts in Aug 2014

              But it still begs the question “what caused the deflation” ? It was significant when compared to the previous eruption but I understand your train of thought, it could have easily been from a different source and the magnitude then cannot be compared to the previous eruptions.

    7. Thanks Albert for further elucidation on the eye of Africa and the coronae in Venus. I made a half baked attempt to correlate a few ideas into a theory on the previous thread. I will have another go here. Recently I watched a documentary on the holes in Siberia and Mongolia that appear in the night with a rumble and rattle and slowly fill with water. They are found in areas of tundra lakes and mounds. It now seems that the mounds explode leaving deep cavities hundreds of feet deep and fully circular and have always happened but are now happening more often with global warming. Then I considered my preference that the Richat structure was pushed upwards leaving the geological ages of the concentric rings revealing their history. Next I looked at the geology of the region. Gabon, further south, has natural formations which are spent nuclear reactors as measured in the radioactive decay whereas Mali has uranium which has not been spent underground. The article on arc electrical coesite fracturing would seem to be debunked and the Harvard link didn’t show the research paper on electrical storms and coesite. Anyway the idea I had then linked back to Venus as perhaps having more propensity to subterranean explosions rather than magma movement. I will stop here as this is just a speculative linking of diverse phenomena. Best wishes all.

      • As in maybe there is a correlation with methane making the holes in the tundra and maybe uranium pushing up and making the Richat structure in Mali and maybe a similar process pushing up and crumbling the circular structures on Venus and Richat? Lateral thinking…..

      • The two maars inside the Eye show that there were indeed subterranean explosions. Perfectly circular depressions are normally due to underground excavations, often explosive (even for impact craters, for there too the actual explosion happens underground). But these don’t give the regular titled layers you find in the Eye. That is best explained with doming, following by saving off the top by erosion. Magma is an obvious heat source for the doming but other sources of heat may play a role. But you need more than just heat: to move up a mountain you need to fill the void below. Only magma can easily do that. Radioactivity or methane does not provide new material and so can’t really push up things much. Thermal expansion helps a little bit but rocks don’t expand much when heated. Heated water can move around and push ground by many meters (this happens in many calderas) but it doesn’t make circular domes. In this case, magma is the most likely cause.

        • I defer to your expertise, Albert, though because of the sedimentary nature of the rings, and the fractured coesite which allegedly requires water and an electrical charge, I remain suspicious of something forcing the central rocks upwards, shattering the dome, allowing it to be eroded more easily. I am not aware that any of the rocks are magmatic in origin and the event appears to be unique. Shattered coesite is the ‘smoking gun’ perhaps. My favourite mystery however remains the launching of India off the east coast of Africa. Best wishes and thank you for this brilliant article.

          • The surface rocks indeed are not magmatic: apart from the two ring dikes, the magma stayed further down. There was shattering, attributed to water dissolving the lime stone, which cause collapse of the central area. Is this Eye unique? I am not aware of anything on earth quite like it but I am not the expert on this: perhaps other people on this blog know more!

            • Water dissolving limestone?
              The Mark-1 mod A Sinkhole. But on a fairly huge scale… 😀

    8. A new star 2 km northwest of Krysuvik. Wonder if the 24 hours of increased tremor at Krysuvik station has something to do with this or if it was a technical issue wit the station.

      • Wednesday
        06.07.2016 03:06:23 63.897 -22.109 5.3 km 2.8 99.0 2.4 km WNW of Krýsuvík

    9. Speaking of Latin words in English: I cringe at “Venusian”. The root is vener-. The s is just the nominative marker and has no place in the adjective. I realise that “venereal” would have distracting connotations, so “venerean” has been suggested.

      • You are quite correct, of course, and the proper form is in use in medical circles. The unwanted connotations are understandable given the origin of the word! But in astronomy, which tends to present itself as disease-free, I have only seen ‘venusian’ used. Venerian might be better

        • Astronomers, as a human sub-species (!) tend to read a lot of SF (some of ’em write it too), and SF authors have always preferred ‘Venusian’ -or at least their editors did

          • It is commonly assumed that english words with latin and greek roots should be conjugated in the fashion of their original languages. For some time now it has been the norm to consider these words english and thus follow english norms. This is good because english already has more than enough oddities for foreign speakers. So a noun venus should transfer to venusian, and plurals should add an -s, no matter how unerudite that sounds.

    10. Under optimal conditions, warning time for an eruption could be less than an hour (last line of the article). I wonder what sub-optimal conditions are!

    11. Audio warning in the Reykjavik ops room; just a thought, wouldn’t it be a good idea to have a relay of the audio warning -something like a WWII air-raid siren- on the slopes of Hekla herself? Just in case there were any foolhardy tourists in the vicinity. A cheap bit of kit (since it would be destroyed by the eruption) but just possibly a life-saver

      • The main problem with that would be false alarms. Eventually people will just blow it off and not pay attention. With as many times as I heard the Tornado warning sirens in Jackson MS while growing up, its surprising that town still exists. (If course, it’s not much of a town anymore, more of a war zone from the competing street interests. And, in my opinion, that town got what it deserved.)

        A long time ago in a place not that far away, I was talking to a fellow fire fighter when he leaned up against the wall. Shortly after that, we heard the old siren on tower of the station go off. He had leaned on the switch. The siren was left over from the early days of the station when that was the only way to summon the volunteers. Even though it was pretty much in the middle of a neighborhood and never used…. no one called to ask what was going on. My guess is they just thought it was a new engine with a larger Federal on front. (A Federal is a brand of impeller style siren)

      • Interesting comment. There are several boards in the vicinity of Katla (eg at Emstrur on the Laugarvegur, and then at Basar on the extension down to Thorsmork) that describe in quite a bit of detail both the warnings that will be given (iirc, flares and siren) and what to do (eg head to high ground). However, when our bus stopped at the point on the Landmannaleid (= F225) where one of the routes up Hekla starts, there was no warning sign – though it is possible that it was out of sight further up the route.

        But, as Geo says, there’s a far greater likelihood of false alarms when it comes to Hekla than Katla.

        • Point taken about false alarms; although I still feel that there should be some sort of warning in place on Hekla’s flanks in the event that IMO detect a worrying anomaly under her Ladyship. (if only to prevent irate relatives of the deceased suing IMO for negligence in not giving a warning) Katla is a different katla-fish (sorry!) since the principal hazard there is the jokulhlaup, surely? Which will develop rapidly and can pose a serious danger to visitors at a considerable distance, not just to Darwin Award candidates

          (Yes, Hekla can be dangerous at a distance, too, but a 15km high eruption column is a fairly obvious warning to stay clear)

          • The rocks travel a lot faster than the ash. Didn’t a rock from Hekla kill someone 40 km away?

            • I believe so; that is what I had in mind. Hekla’s eruption column is a danger signal that can’t be ignored; jokulhlaups are sneaky by comparison

            • Well, an optimal solution would be for Iceland to lease the top of Hekla to the US and for us to level it and put a congressional dormitory up there. That way if anything happens, they will definitely start yacking about it. Those buggers never shut up.

          • It is possible that there is a warning sign at the end of the 4WD track from which the walking path up Hekla starts. If Icelanders put warning signs everywhere there are potentially life-threatening hazards (eg waterfalls, solfataras, glaciers, deserts, powerful waves, avalanches, river crossings, quicksands, etc. Oh, and volcanoes), the country would be plastered in them.

            The Katla sign states the hazards (tephra, gas, lightning, jokulhlaups) and warning signals that will be given in the event of an eruption, it shows escape routes, and gives advice on what to do. For walkers and vehicle drivers, jokulhlaups are relatively less significant hazards, and also more predictable.

            • Well, I would point out in my defence that Hekla -because of her short warning times combined with being a popular tourist destination- is a hazard of a different order to waterfalls or solfatara fields. I will close my comments there, lest I suffer the fate of Mr Cryptodome

        • See the discussion initiated by Ian F at 19.13 on 05/07 above, and in particular the analysis given by Albert. However, the possibility of a ‘failed eruption’ was not discussed there, so we need to add that to our considerations.

          • Ahhh yes the Hekla end tilt response to the inflation/deflation at Bárðarbunga… intriguing thought… and perhaps a better explanation than ‘failed eruption’ – although there was a point fairly recently where IMO were concerned enough about Hekla they issued some kind of warning as I recall?

            • My memory isn’t clear, but are you thinking about the reports in the Icelandic and then foreign media (Daily Mail!) in the last month or so? If so, then these were based on interviews with Pall Einarsson in which he was saying pretty much the same thing as in the IMO report, but with less scientific content.

            • There was a warning around start of april in 2013 warning people to stay away from the summit.

              released from the holding pen where it waited for approval

      • The next eruption is most likely larger, than all the previous ones since 1947, because Hekla eruption intensity correlates quite well with dormancy time since previous eruption.

        But judging from the very good past correlation, presented below, I think it will stay around VEI3+, most likely it will be similar in intensity to Eyjafjallajokull but less long lasting.

        – Most eruptions in previous decades, apart roughly every 10 years, started explosive VEI3 and then with substantial effusive eruption in following couple of weeks. Strong explosive period but short duration.
        – Most eruptions in previous centuries, apart usually every 70-80 years, were more violent at VEI4, significantly more ashy, with larger effusive volumes of lava afterwards.
        – Sometimes they were apart only 30 years and the intensity afterwards was VEI3 (as in 1636).
        – In 1222, an eruption happened 16 years after rest, and it ranked only as VEI2.

        – In another way, the eruption in 1947, happened after 101 years of dormancy, and it ranked VEI4+ as a very big eruption, with a large volume of lava erupted. The same happened in the eruption of 1510. Both were quite violent and very ashy eruptions. The eruption of 1104, after 200-300 years of dormancy was even more violent and catastrophic, at VEI5. But even stronger were the pre-historic eruptions of Hekla at VEI5+ after many centuries of dormancy.

        Though the next one might rank only at VEI3+, it should start quite violently, with a column around 15km high, but that should be rather short duration.

        Usually Hekla eruptions last about a year of effusive activity, following a couple of days of the initial very explosive period. But the eruptions since 1947 only lasted a month of effusive activity. So the next eruption will probably last longer.

        We shouldn’t also ignore the fact that many eruptions have also occurred in the vicinity of Hekla, not being Hekla proper, these are mostly basaltic effusive eruptions in the flanks or vicinity of Hekla, and can start mildly explosive at VEI2, like in 1913 and 1878, lasting also about a month.

      • That M3.4 the other day was probably appointed to Grimsvotn. That or there is a glitch.

          • A good example of the effect one semi-decent sized quake can have on cumulative energy versus a bunch of small ones.

            I’m a little unsure whether that energy should really be attributed to the cumulative energy of the Grimsvotn system proper however – given the location.

    12. 3.1 at 3.7 km at the sw rift of Mauna Loa.

      3.1 2016/07/10 04:38:03 HST 3.7


    13. For general information: we hope to have a new post on Tuesday. Manpower is very limited at the moment and the frequency of posts is not as high as we would like it. Luckily in this blog the comments are more important, and often more informative, than the posts. Thank-you to all contributors, regular and occasional!

      • Therein lies the problem, actually. Since new comments tend to only be made on the latest post, when posts are frequent, it is easy to keep abreast of the valuable comments. But when the time between posts grows to well beyond a week, it becomes necessary to slog through 100+ comments again and again, daily, in order to not miss new ones. This is ameliorated somewhat if one ignores every leaf comment with a date stamp of at least 2 days prior, and every ancestor of such a comment, and checks daily, but it is still somewhat unwieldy. A chronological display order would fix that by allowing going up from the bottom until the first already-read comment, but at the cost of losing the reply structure that groups related comments. The ideal solution, which would require client-side cookies but doesn’t seem technically infeasible, would be some sort of read/unread icon next to each comment. Or else more frequent top-level posts. 🙂

        • On my view, I have a list of recent comments in chronological order at the top of the post, on the right. I don’t know whether everyone sees that (perhaps you need to log in?). More frequent posts would be nice! But they are take time to write. Over the holidays if any we’ll have fewer. I am writing this from holiday in Melbourne. Could have picked a warmer place to go!

    14. New Scientist has an article written by Stefan Kropelin on the Sahara’s Volcanoes, notable the Trou au Natron volcano. No link as it’s behind a paywall. But you can pick up a copy easily here in the UK.

      • Really impressive Ian and to see how concentrated the quakes are between south and west. The large quake to the west of Gjalp is still puzzling me though.

    15. Bard’s got a 3. Tho I don’t see a star on map yet.

      12.07.2016 09:07:44 64.660 -17.559 6.3 km 1.5 99.0 2.7 km NW of Bárðarbunga
      12.07.2016 08:58:45 64.675 -17.523 4.1 km 3.0 99.0 3.9 km N of Bárðarbunga

    Comments are closed.