In this article I will pick up with the questions regarding age, GPS-tracks and the general setting. Knowing me it is quite likely that I will meander out into wild tangents as I go. Time to Greíp my pen.
The age of Greíp
For a volcano that has not erupted (we will get back to that in the next chapter) Greíp is a wild ride. As I noticed it back in 2012 my initial reaction was that it was a run of the mill deep dyke injection of magma. Interesting as such, but not that unusual.
It did though grow on me as I noticed that it was unusually persistent and long-lived, and that it ran up to a certain point and then dead-ended. This normally means that there has been activity there for some while.
Checking backwards we could see that there had indeed been activity prior to this, and that it formed a very well-defined stack of earthquakes reaching upwards. I then surmised that it was a formative sill-dyke system indicating the possible birth of a future central volcano on the edge of the Grimsvötn fissure swarm.
Then came Holuhraun II and stole quite a bit of time from us. So, I handed over the ball to Gaz who ran with it together with Andrej Flis. Up until that article we had called it Unknownabunga, but he named it with Greíp, and since it is a smashing forn-nordic mythological name I hope that the Icelandic naming committee will make it official in some form, perhaps Greípur or Greípagigur.
As I looked at his article, I felt a strong urge to write about it too, but since I am lazy it ended up with me talking endlessly about it with Tommy, Andrej and Gaz. In the end we felt that we had refined our ideas enough to write a more developed series of articles.
As Andrej started to churn out earthquake plots something struck me with the force of a super tanker hitting a jetty. And that was that there is no simple sill and dyke system to be found, instead there is a well-defined magma chamber with all the bells and whistles.
Problem with that is that takes a bit of time to form a magma chamber, far longer than we could find instrumental data for. There was also an intriguing anomaly within the first propagating dyke leg that led from Bárðarbunga towards Greíp during Holuhraun II.
During the winter of 1796/1787 a lava field mysteriously came into existence between Vatnajökull and Askja. It was found by a couple of Icelanders travelling through the volcanic hinterlands of Iceland.
They noted that the lava was horrible to walk across (illuhraun), and that it had not been there the previous summer when they last ventured out there. I would dearly want to know why they meandered through that neck of the woods two summers in a row, but that is lost in the mist of time.
They then eye-balled the nearest volcano, Askja, and blamed it for the nuisance to their trek. This is the only “evidence” of Askja being the culprit there is, but steadfastly some sources, including Global Volcanism Program, has this eruption as a Askja-eruption against a veritable wall of evidence placing it at the feet of Bárðarbunga.
There is also no compelling evidence or reason to believe that Holuhraun I was not sourced and propagated in the same manner as Holuhraun II, quite the opposite.
This brings us back to the anomaly of the dyke between Bárðarbunga and the vicinity of Greíp. As you analyse the earthquake data for the dyke it rapidly becomes evident that the wall of the magma chamber at Bárðarbunga where the dyke ventured out was closed off, then the earthquake track rapidly diminishes before going silent, and then growing in strength again as it comes closer to the stress-field of Greíp.
The propagation speed here was also quite fast compared to the second leg of the dyke between Greíp and the Baugur vent.
Taking all of this into account leads to the solution that exactly the same dyke (LD1) was in play during Holuhraun I, and that it was left open with quite a substantial amount of residual magma.
And if it bounced at the same knee in 1796/1797, then there either was a stress-field in place there, or Greíp was born out of a large amount of residual magma that through tectonic stress formed a dyke down into the magma reservoir under it.
This means that we can safely date Greíp as being a minimum of 220 years old. If it was already a central volcano with a large enough magma reservoir to create a local stress field capable of diverting a propagating dyke it is older than that, if it formed out of residual magma, then we have a surprisingly exact date for the formation.
The three-body problem
This is a famous problem in mathematics regarding orbital mechanics. We can mathematically predict the location of two balls orbiting each other pretty much ad infinitam, but if we chuck in a third ball, we can’t deduce any stable orbit, nor can we predict where the balls would be in the future. Inevitably at least one of the balls will be spun out of the group orbit.
I have always found this one to be highly interesting, and the reason for that is that nature itself seems to be quite happy with creating stable 3-, 4, or even 5-body orbital groups. And we know of such a surprising number of stable multi-body orbital groups that it defies logic, so this is not caused by a random lucky system that somehow against all odds found the one viable option.
Obviously, nature is not very good at mathematics, or our mathematics is lacking something obvious in regards of multi-orbitals. (I can now hear Albert groan over the extreme simplification of a very complex subject.) Finding glaring holes in our understanding of mathematics and physics is one of my great joys in life, I am just too lazy to be getting around to plugging them.
Now why on earth did I amble off into mathematics and the stars orbiting each other? Because when analysing GPS-data in Vatnajökull you rapidly get stuck in a massive multi-body problem of your own. And to confound things even more, you must contend with that depending on which half of the world you view the problem from, you get two different sets of multi-body problems at the same time from the same pesky data.
If this is giving you a headache, I advise you to go back to the prequel and lament over problems with reading instruments and ways that errors can creep in.
The GPS-multi body problem
If you are dealing with one volcano only it is very easy to calculate the trajectory of it, and if you have two of them, all you have to do is subtract the influence of A to get the true trajectory of B, and vice versa. Add a third and you are out in the great blue yonder.
In the map above you can see a lot of culprits jostling for space with each other. And to confound the problem a bit more, the GPS-trajectories for the volcanoes change constantly over time. I think we need a brief exposé here over what the individual volcanoes and features are doing right now (as far as is possible to understand.
I am going from North to south and will for now omit Greíp:
Herðubreið: The Holocene volcano of Herðubreið last had a minor eruption at the beginning of the current-interglacial. It was previously believed to be either deeply dormant or gone with the dodo. It has slowly and steadily been inflating since 2007 due to root-infilling with associated magma movement upwards, indicating a possible future eruption.
Askja: Askja inflated slowly between 1961 and 1975 and deflated between 1975 and 2012. After that a very gentle inflation period started indicating that a new eruptive cycle might be coming.
Trölladyngja/Urðarhals: These polygenetic shield volcanoes started to inflate in 2013 and has continued to inflate rapidly well into 2018. Then the inflations stopped quite abruptly but they are not deflating.
Kistufell: This volcano inflated rapidly in 2013 and 2014, deflated during Holuhraun, and then continued to inflate up unto 2018. After that it started to slowly deflate.
Kverkfjöll: This sad has been of a volcano has slowly deflated since at least 1968. This former major player has now permanently moved into a volcanic retirement centre where it now and then whacks a nurse with a cane. Expect it to continue to slowly deflate into oblivion as it is moved off from the volcanic centre of Iceland.
Bárðarbunga: This volcano had slowly been inflating for decades up until the eruption of 2014. During the eruption rapid deflation took place, and as the eruption ended rapid inflation indicated refilling of the volcano. During 2017 the inflation rate started to slow down, and during 2019 the volcano has had a strong deflation signal indicating that the volume of magma is not enough to cover the EW-spread of the fissure swarm.
Hamarínn: No other volcano in Iceland has had such a long and strong inflation period as Hamarínn. Problem is that nobody knows why it is inflating, and why it is not giving off signs of an imminent eruption. It is though the strongest shoving force in this part of Iceland. Basically, this fella needs some serious study down the line (but not here and now).
Grimsvötn: Due to the location of the GFUM GPS-station this volcano is its own three-body problem, or even a four-body problem. The problem is that the volcano has 3 distinct magma chambers that are connected into a deep common magma reservoir.
This means that the magma will always enter into the magma chamber with the lowest pressure, and as magma enters that chamber becomes top dawg and either erupt or end up last in the fuel line. And since magma is almost constantly entering into the system this means that pretty much all the time something different is inflating squeezing the other two.
In the end this means that the poor GPS will move in every conceivable way and change trajectory almost on a monthly basis, with odd spurts that are almost impossible to divinate the origin of.
And every now and then one of the magma chambers will erupt, resetting the system pressure and the refuelling game starts anew.
Háabunga: In a way it is debatable if this is a fourth magma chamber of Grimsvötn or if it is a separate central volcano. There is small amount of earthquake data placing it as a separate entity. The problem is that this volcano is just south of the poor GFUM-station.
Háabunga seems to be inflating in distinct spurts violently pushing GFUM sharply north and up, before the general push south resumes from the 3 chambers of Grimsvötn proper.
Thorðarhyrna: This volcano seems to be a neutrum, neither inflating, nor deflating.
Öraefajökull: Up until 2017 this volcano did not amount to much as GPS-trajectories go. But after that it suffered from a marked episode of inflation that ran up until a few months ago when it calmed down.
Leg Dyke 1: This is the leg between Bárðarbunga and Greíp. Normally we do not talk about separate dykes and their effect on a larger scale, but in this case, it is warranted. During the Holuhraun II eruption it inflated massively, and after the eruption it remained in a neutral dilated state.
Leg Dyke 2: Obviously this part of the dyke inflated widely during the eruption, even more so than LD1. But, after the eruption this part of the dyke started to deflate rapidly as magma started to cool down.
Now that we know the culprits, we need to know two facts. A long-term deflating volcano will gently suck things towards it like a mild-mannered black hole, and a long-term strongly inflating volcano will gently push things away.
Or in other words, everything should be moving away from Hamarínn and towards Kverkfjöll (obviously in different trajectories since there are quite a few volcanic obstacles on the way jostling each other). Think of it as a pool-table that is slightly tilted towards the east corner-pocket, with Hamarínn being the pool cue. It should all generally be ambling towards Kverkfjöll.
Up until around 2011 this was indeed true, but after that things start to become ever more out of whack. It is like an X-factor is creating a mound on our otherwise flat and obstacly tilted pool-table.
The volcano that should be most clearly be pushed away from Hamarínn and towards Kverkfjöll is Grimsvötn. It should be pushed ESE by Hamarínn, and sucked NNE by Kverkfjöll, giving it a general trend NE.
Problem is once more the placement of GFUM, the signal of Grimsvötn is happily masking that motion and showing general SE movement. The east is congruent with Hamarínn, but the massive inflation is reverting any N-signal into south movement instead. Pesky volcanoes interfering with hypothetical trajectories showing up.
Now, if we take the longer trend instead and would sum over several eruptions, we would indeed get the NE signal, well up until 2011. Then it changes with a more ENE trend if we try to remove the inflation cycle effects.
The same very weak anomalies show up on most volcanoes affected by these two long-term effects. It is like there is a slight mound diverting them just a little bit. An X-factor that should not be out there.
Problem is that there are several strong signals out there muddling the waters as we try to hear a whisper of things to come. And to boot this has been the most raucous decade in Vatnajökull in instrumentally recorded times.
Still, in the faint signals we get the possibility to deduce that there is something more than LD1 and LD2 affecting things, and that highest lump on the road seems to be at that infamous knee where LD1 turned sharply into LD2. There is definitely a stress-field there slowly picking up strength as evidenced by the flatter slightly flatter east trajectory and a clear change in the NS-trajectory. Hopefully it will soon become clearer as the noise of 2014 diminishes, and more magma arrives into the reservoir.
Before Albert comes around janking my hypothetical chain of circumstantial evidence, I will freely admit that this is quite a bit of theoretical ho-hum, and that I would not even write about if Andrej and Gaz had not come up with quite a bit of a more substantial case in the previous chapter.
As it is, I have one firm leg to stand on, while the other leg at best proves that it is hard, but not impossible, to say anything with a high degree of certainty from the GPS-system.
As global warming does its thing and more nunataqs crop up (and receives GPS-stations), things will become much easier. The vagueness in what I have written about above is mainly due to lack of near-field data for a lot of volcanoes.
Perhaps Albert in a later chapter will invoke his awesome powers of satellite derived InSAR to bring clarity. I can just imagine what a headache all the ice-movement and ice-fluctuations will bring.
Now, what would happen if we skipped the long-term trend changes and instead had a hypothesis that the increasing number of earthquakes on the plots of Andrej Flis that Gaz mentioned should be moving the nearby GPS-stations around. Will that be fruitful?
Up above we definitely found something at Grimsvötn (GFUM), but that could be because of local changes instead of my fevered hypothesis. We need more evidence, so let us hunt among the other stations in the vicinity.
Gengissigid (GSIG), is the closest station to Kverkfjöll and should be fairly stationary with just the tectonic movement east and be deflating. In April of 2016 something drastic happened and it started to push north and show inflation. Either a volcano known for deflation and no NS-trajectory has decided to inflate like mad, or something near it is inflating quite a bit. Let us check further and if it is local shall we.
Station Kverkfjöll (KVEC), is showing the same pattern, just a bit less of inflation since it is further away. Problem is just that it would be showing the same regardless, we need to hunt for another GPS-station less probable to be affected by Kverkfjöll.
GIGO is fairly far away and is impacted hard by the inflation at Urðarhals. Same goes for the intermittent station URHC.
DYNC is out of commission since a year ago, and it is also a bad one to do long trendlines on since it is so heavily affected by Holuhraun II. RJUC, Rjúpnabrekkukvísl, is mainly just showing things related to Bárðarbungas deflation, same goes for Kista (KISA).
The signals are just to big from Bárðarbunga and Urðarhals for anything to be visible, also this is where the main spread of Iceland is occurring, and that produces a lot of noise.
So, in the end we find that there might be a weak signal caused by blockage to the Hamarínn toward Kverkfjöll. And that Grimsvötn seems to react to something close by, and that either Kverkfjöll is inflating, or there is a point of inflation behind in under the ice.
What to do now, well, we can remove Kverkfjöll at least as the point of inflation, there has been no such signals detected in the form of earthquakes, and we do know that there has been signals indicating influx of magma into Greíp. And looked at things in that light, we now can reverse our viewpoint and state that we seem to have a bit of corroborating evidence in the form of GPS-data for Greíp.
The hunt for GPS-signals did not produce a huge smoking gun, but it confirmed the earthquake data debated by Andrej and Gaz in the previous chapter. Order is restored in the Universe.
Eruptive characteristics of Greíp
Even though there are a couple of unknowns of the volatile content of the magma in Greíp we can safely assume that it will be congruent with either Grimsvötn mainline type basalt derivatives or be a Gjálp-type intermediary basalt in between Grimsvötn and Bárðarbunga.
In other words, a nasty smelly volatile gas-filled basalt. The water content in the basalt itself will though be small since it is plume derived.
We do not know how long the magma would have resided in situ, but it is probably not long enough to have evolved anything more interesting than andesite, most likely it is a slightly evolved basalt.
Here our friends of all things Hawaiian concluded that any eruption would be in the form of large strombolian eruptions or lava-fountains. Not a bad guess, except for two things.
The first is that the emergent volcano would come up in a valley below the ice. The second is that the heat would rapidly melt the ice, giving the recipe for ash. Combine those two and you get a pooling of liquid melt water surrounding the emergent volcano.
The likelihood that Greíp in a single eruption would be able to build an edifice high enough to avoid the water and the ice is slim to none. Yes, Gjálp built an elongated cone sticking up above the ice and water, but it started its life on a pre-existing ridgeline.
So, expect an ashy Grimsvötn style eruption ranging in about VEI-3 to a small VEI-4. Anything smaller would just immediately quench out and be noticed just as another Icelandic tremor episode, with perhaps a bit of stinky water emerging from under the glacier.
When dealing with something new that is not mainline in science it is best to take baby steps. The only thing I have been able to come up with above is basically a few things to look for in the future in regards of GPS-trajectory changes, and the possibility to perhaps shed more light via long-term plotting of InSAR.
I could though say more about a future possible eruption, and that saved this chapter a bit. Science is equally much about failing to find evidence, as finding evidence. But I will try other ways to find smoking guns.
I also completely left out the part of the possible intermingling of magmas at Greíp, and its effects on Bárðarbunga via LD1. I will return to that in the next chapter, because in that regard I found something more fruitful to say.
So, I will be back with the astounding tale of the quad-junction everyone missed.