Back in early 2011 I was sitting down for a late light evening read of “Lower-crustal earthquakes caused by magma movement beneath Askja volcano on the north Icelandic rift” by Soosalu et al. As I read angels started to sing a glorious hymn as I had an epiphany and my view on how volcanoes operate changed forever.
For me this is the single most important paper in how to understand volcanoes, and I have been meaning to write about it for almost 8 years. So, a bit overdue, and with some mighty rambling detours I will describe the paper, the implications, and the current tidings of Askja.
The paper is straight-forward and easy to understand, it describes the findings of two seismic studies done in 2006 and 2007 of Askja volcano. It was done with 20 seismometers working in conjunction with the national Icelandic network.
In 2006 this highly augmented network detected circa 100 deep earthquakes between 14- and 26-kilometres depth. In 2007 the same network detected more than twice that number, some as deep as 34 kilometres deep.
This was interpreted as short magma dyke intrusions caused by increase in magma pressure, presumably as magma moved upwards.
The earthquakes were not of the same type as the upper crustal tectonic earthquakes, instead they had a lower frequency with a perceived coda produced by reverberations presumably from adjacent magma bodies.
In Soosalu et al they perceived this as short dykes being a few metres long, or as being brittle quakes in hot ductile material caused by increasing magmatic pressure.
After this paper the model was proven to be correct as Askja was seen as partially inflating, a change from the previous state of long-term deflation. This was also found by Hazel Rymer in a study of gravimetric changes.
Epiphany and Vivaldi
I remember that I put on Thomas Wilbrandt’s reinterpretation of Vivaldi’s Four Seasons on my Bowers & Wilkins Nautilus speakers and sat down for a think.
As I was thinking about deep stuff I decided to start very deep indeed and work my way up. I started with the Icelandic plume, pondering that it was cyclical in nature, and how that might show itself as and when a new cycle emerged.
Since we are talking about something akin to treacle moving at the speed of an inch per year, I thought that it would be a gradually emerging event moving at glacial speed. Sort of like if a pillow was hitting a foam mattress in slow motion from below, slowly pushing itself up into the cracks and nooks of the Icelandic volcanoes.
Then I started to ponder how that would be evidenced in light of the paper I had just read. And I thought that this would give a very spread out and diffuse amount of deep earthquakes centring around Kistufell, since this is the volcano sitting directly on top of the plume head.
At this point I connected this to another thing floating around in my head. All the volcanoes around the plume head had been erupting magma from the previous plume pulse roughly 250 years before.
Grimsvötn was a stunning example of this, it had been coughing up ever staler dregs after the eruptions it had back around the Skaftár Fires in 1783.
Grimsvötn had then violently changed in 2011 when all of a sudden fresh hot magma filled with large amounts of volatiles pushed out in an eruption bordering near the VEI-5 range. At this moment it all clicked in my head and I felt pretty sure that the magma pulse had started not long before the 2006 survey in the paper I had just read.
It then dawned on me that clumsy me had stumbled on a theory that could be a way to prognosticate something never seen on instruments previously, and that was eruptions caused by a plume pulse, and that if I was correct these deep earthquakes would coalesce on the volcano that was the easiest to break through at. Ie, at the point of least resistance.
This lead me down a two-and-a-half-year long hunt for deep earthquakes in Iceland, much of this was possible thanks to Lurking as he introduced the world to do-it-at-home volcanic plotting as a prognosticative tool.
As chance would have it, I got to test it against a more unexpected plume pulse over in the sunny Canary Islands as El Hierro decided to become festive.
There we could see initial deep quakes turning into sill-formations as magma pushed itself into the dormant volcano of Tanganasoga, before it all went down into a fissure swarm and erupted harmlessly out in the waters to the south of the island.
I must admit that I at that point did not see this being copied so exactly on a far grander scale two years later in Iceland in such exacting detail.
Prior to Icelandic festivities
As the eruption of El Hierro ended I turned back to my hunt for deep earthquakes under Iceland with new vigour, feeling that the model I had envisaged would work. Soon enough the earthquakes started to coalesce around a fairly unknown volcano named Kistufell, known to be situated directly on top of the plume head.
In 2013 the activity had increased quite a bit and I was convinced that either Kistufell would erupt, or we would have a shield formation north of it, or that Bardarbunga would erupt, possibly as a rift eruption.
Since I did not want to sound like a crazed loon I discussed this behind the scenes with my fellow moderators, especially Lurking, so that things would be saved for history if I was correct. I also started a rather long-term series of articles about that part of Iceland to give the readers a backdrop to the eruption I by then knew was coming.
Quite a few people starting to wonder if I was off the rockers writing incessantly about Icelandic unknown volcanoes and volcanic features that was doing absolutely nothing in the eyes of the average reader, even though I hinted quite a lot about something being in the making.
As the months progressed, I could refine the prognosis until I could accurately and publicly state that an eruption was imminent. In retrospect this was my 15 minutes of volcanic fame.
Obviously, I would not have gotten it right without having a bookshelf with volcanic papers in the loo, complete with the marvellous paper by Soosalu et al. All honours where due. By the way, am I the only one on the planet with a bookshelf in the loo?
Enter the Dragon
As things have it the Icelandic national network picked up 23 deep earthquakes at Askja last week at the same NNE corner of the volcano as Soosalu et al described in that paper indicating that once more the pressure is increasing over there.
I should here say that there were probably hundreds of them, but due to technical reasons most of those could not be detected due to the minor number of seismometers in the vicinity.
It was indeed followed by minor changes in the GPS-trajectories. And with minor I really mean minor changes, but they indicated inflation over a wider area tilting the volcano ever so slightly indicating a root-fill episode of the volcanic system.
I should here go into some detail of the state of the volcano. First of all, it is not one of the volcanoes producing high frequency eruptions, instead it goes through episodes of infill causing eruptive sequences interspersed by rather long dormancy periods. Sometimes these dormancy periods last far longer than a millennium, but as and when it has an eruptive cycle it is not uncommon that it goes caldera, something it has done 3 times since the end of deglaciation.
It last erupted in 1961 and has suffered from a long-term period of deflation caused by evacuation of magma, possibly influenced by the Krafla Fires, since the deflation started at that time.
This long-term deflation means that the volcano is fairly depleted of magma and is in a low-pressure state. The rate of deflation was deemed to be around 50 centimetres per year at the caldera centre by Hazel Rymer, a quite staggering amount.
This means that it will take quite some time before the volcano is refilled enough to erupt. It all depends on the rate of the intrusion, and how much of the available plume derived magma that will end up inside Askja in competition with other nearby volcanoes.
Bardarbunga is currently taking up quite a bit of the available magma, so is also Grimsvötn as it preps for its next eruption. Same goes for Kistufell and Trölladyngja, the latter is a bit of a mystery since it officially is a single eruption shield volcano officially not counted as a central volcano. It was though also quite active prior to the Holuhraun eruption from a seismic standpoint evidencing the possibility that it is an emergent central volcano.
To the north-east of Askja you find Herdubreid, a volcano that last had a minor eruption in the early Holocene. Herdubreid has been highly active in the last decade and is inflating as magma is entering the system.
Here comes a problem with the GPS-trajectories, it is a very noisy environment with all these volcanoes acting on the stations at the same time as tectonic swarms alter things, and at the macro level you have the continental rift running through it all.
Now combine that with the after effects of the Holuhraun eruption and you need to be the Gandalf of all things GPS to be able to divinate minute changes caused by Askja, and up until now nothing has been possible to see, and even now it is so minute that it is hard to state where and what is happening at Askja. All we can say is that something seems to have happened.
In the end Askja is a very big volcanic system, even for being in Iceland. It also tends to go for big eruptions instead of frequent eruptions. This means that it will take quite a bit of magma for it to erupt.
On the other hand, we are in the middle of a plume pulse with prodigious amounts of plume derived magma arriving into the Icelandic volcanoes, and if a sizeable portion is going up into Askja it could be cocked and ready to fire quickly.
Right now, we are looking at about a decade of inflation, with not much showing for it since the system was low pressure at the beginning. And until we start to see larger swarms, we know that the pressure is not high enough for an eruption to be possible.
Question right now is if that will happen next year, or in a millennium?
The answer is that unless the amount of magma entering the system increases rapidly, we are in for quite a bit of a waiting game.
A couple of weeks ago I was asked if the eruption of Holuhraun would have any influence on Askja. The answer to that is that it is highly unlikely. Holuhraun was a distal eruption caused by a rifting fissure.
The last time that Holuhraun erupted was in 1797 and it took 85 years until Askja erupted, so at least back then it had no effect.
Anyway, Askja is indeed doing interesting things and it is on a runup, but it is still a bit early to say when it will be interesting enough to erupt.