This article on Hekla is my personal birthday gift to Mike Ross on his big day. Congratulations Mike!
About once a year I write a piece about which five Icelandic volcanoes would be the most likely to erupt given the signals they have fired off during the last year. This list normally contains roughly the same volcanoes, but the order of them tends to change due to activity levels. This year I thought I would write five individual pieces so I could go a bit more into depth at how and what I look for and what to look for when an eruption would start.
A couple of months ago I would have put Grimsvötn as number one on being the most likely to be next to erupt. But with the activity lately at Hekla it is time to take yet another look at things there. As Boris Behncke noted, this is a volcano with a following in here, undoubtedly it is in part my fault since I can’t put a sock in my pie-hole.
Spring of Hekla
From a historical standpoint a majority of Heklas eruptions has started in the spring, somewhere between January and May. So it should not be any great surprise that she is starting to show activity in March would it?
I believe that the statistics is a bit off due to a forgotten reason. Hekla used to have a glacier on top of her. Not a big one like Vatnajökull or the other great Icelandic Glaciers, but enough to probably give a seasonal stress load imbalance on the volcano. As the glacier melted in spring the load diminished and that set it off.
The eruption of 1947 put an effective end to that Glacier. Before that eruption the glacier had sufficient time to grow back between eruptions, and indeed the glacier started to show signs of returning before the 1970 eruption, but it never formed fully.
Later on the rapid series of eruptions kept the mountain to warm for a glacier to form, and to top that off, climate changes has by now most likely put Hekla below the height needed for a glacier to form anyhow.
I do not expect the “Hekla spring” pattern to hold in the future. This said, it still seems to be more unrest in spring to early summer, so there might be something else affecting the volcano, perhaps it is available water causing hydrothermal effects. Here more research is needed.
Hekla is not a volcano
No, I have not gone crazy. What I mean is that Hekla is not a proper central volcano. Hekla is an over-grown fissure volcano on the road to form a central volcano. This gives her a few rather odd characteristics for a volcano.
Hekla most likely started as the most eastern of the Icelandic Sprungur on the Southern Icelandic Fissure Zone. As such it is situated over the rift between two micro plates that rotate apart from each other. This explains two very odd things, the first being the rapid growth of the volcano. One should remember that Hekla is just 8 000 years old.
It also explains that unlike other fissures in Iceland it erupts at the same place. Veidivötn, Laki and the rest does not do that, they erupt in a series of parallel fissures. The reason for Hekla continuing to erupt at the same place is due to the fissure being continuously kept open at dept by the slowly rotating adjacent plate.
As the plates move apart there is a continuous lowering of the pressure at depth causing melt to form. That melt goes up through the fissure silently since it is moving through already existing melt-material and this causes no new earthquakes. One would now think that this would be a noisy process, but since there is a large sheet of melt material covering the fissure from the MOHO up until just a couple of kilometers below the surface this process is by now quiet. The only noise is at the ends of the fissure, about 15 kilometers SSW and 7km NNE of the outer edges of Heklas edifice.
The rest of the noise is happening in other places as the micro plate rotates, and in the case of SIFZ, is being ripped into small ribbons of ground with large fractures in between. So, in a way the Hekla earthquakes are the same as the Sprungur earthquakes, and even other earthquakes much further away. It is after all a dynamic interdependent system. But it does very well explain why Hekla herself is so quiet. It also explains why almost every Hekla-earthquake we see is at the outer ends of the fissure.
Erik Sturkell in his brilliant 2013 paper favored a model with a more classic approach, he used a tubular conduit going down to a magma chamber at 4km depth and a second conduit to a deep magma storage around 15km that intriguingly could be open at depth. His paper was based on the best data available and is as highly accurate. All I ever did to get to what I wrote above was to over-lie an obloid-model to account for Hekla being a fissure.
In the end the obloid-model yielded the endpoints of the fissure, and gave endpoints that I could check against earthquake data. The modeling gave the same endpoints where we see the earthquakes. Otherwise it yielded exactly the same data as Sturkells point modeling. In the end I really love when modeling yields the same data as reality. I should here point out that Sturkells modeling yielded data that exactly fit reality to, but he had asked the model a set of slightly different questions. No shade on Sturkell here!
Also, it gave an intriguing lead to why Hekla is so explosive. In a wedge shaped obloid the heat convection would be higher, so the cooking off of lighter high cilica minerals would be much faster, also the internal transportation of light materials would be significantly faster. In effect it would become a colon-refinery, working in the same way as a petroleum-refinery. It would also yield the famous bimodal lava of Hekla where you first have very explosive lava that changes within a minute or so into effusive lava. I am though not by far a good enough petrologist or petrochemist to do something with this. It just fit neatly into what we can observe during an eruption and general chemical processes.
GPS at Hekla
Fissures operate different in comparison to real central volcanoes like Eyjafjallajökull. Before Eyjafjallajökulls 2010 eruption all of Iceland started to move rapidly as the mountain inflated in a so called bradyseism.
We will never see that at Hekla since the magma is produced as the plates are pulled apart, the net increase would be very small and produce just a steady minute uplift. Also, the E/W and N/S motions would be fairly stable and be more related to earthquake swarms at other places as the plates rub against each other.
Minute changes might be seen as magma is squeezed from different places in the fissure, but there will most likely never be any big changes. It will most likely not be on the GPS we see a coming eruption. Yes, there will be a significant and very brief motion on the two closest stations in the hour leading up to the eruption, but that will most likely have reset at the next visible data point. In other words, you would need minute by minute updates to get anything useful out of the GPS to predict Hekla. It does though give data on the plate motions causing Hekla to exist to begin with.
Strainmeters at Hekla
The activity at the strainmeters at Hekla have been very low so far this spring, no large sudden drops have been recorded at Búrfell strainmeter as it has been during other episodes of high activity. One can see these sudden drops as failed eruption attempts by the mountain as it tries to rip apart.
The lack of these drops actually speaks against Hekla being on the way to erupt now. But, in reality all it takes is one big drop and we are on the go. If you see the Búrfell strainmeter drop in a few minutes below -1e+05 and it continuous to drop, well then you can start holding on to your hat. Brutally fast, and brutally sharp, that is the Hekla way. If you see that starting set the timer, 30 minutes or so later the explosion will occur. This is the premier warning sign.
Tremor at Hekla
Any tremor at Hekla before an eruption would be minute and we will probably miss it due to noise caused by wind and waves. There is really no big reason for Hekla to be tremoring before an eruption. The noise will not come until the “cork” is pulled from the champagne bottle and the entire top of the magma column transform into gassy foam and it all squirts out.
So watching tremor at Hekla is a rather futile thing to do.
Earthquakes at Hekla
Hekla is famous for being aseismic. New equipment at Hekla have though proven this to be either slightly wrong, or Hekla has been unusually active in the last 3 years.
I think it is a combination, we are seeing more of the small earthquakes due to all the new SIL-stations around Hekla. But, we have also seen a couple of earthquake swarms at Hekla that is out of sorts. Especially one year ago we saw a significant swarm that looked a lot like what we would see before an eruption. It is at this point unknown why that swarm did not lead to an eruption. So, seeing a swarm at Hekla should be seen in the light that she once upon a blue moon can have a swarm without an eruption happening.
Up until the 2013 March swarm Hekla had suffered 53 earthquakes in 13 years since the last eruption. For being such a large and active volcano that is a totally unheard of silence. Just to put it into perspective, in the last two weeks we have seen 15 earthquakes at Hekla, 4 of these at Hekla proper. And the March 2013 swarm contained more earthquakes than had happened in the previous 13 years.
So, a single earthquake is not much to jump up and down from. If you see 3 or more in rapid succession below Hekla itself and they are ranging between 15 km and 2 km, then it is time to make popcorn and wait for the strain drop to happen. If an eruption will occur from the small swarm you have seen, it will most likely happen somewhere between 60 minutes to 120 minutes later. Anything much larger than a M2 earthquake is highly unlikely. These few earthquakes would be all the notice you get before the strainmeter at Búrfell drops.
What you will never see at Hekla
You will never see Hekla steaming before erupting; also you will not see any other normal signs of an upcoming eruption like spine-extrusions and dome building. Hekla holds her cards very close to her rocky chest indeed. Also, again… Do not expect to spot something on the GPS.
2 significant Papers on Hekla
http://www.lsndocuments.info/Documents/wp0062.pdf (Big thanks to Boris Behncke for finding this)
http://www.sciencedirect.com/science/article/pii/S0377027313000462 (Erik Sturkells 2013 master-piece. Main pay-walled paper for below)
http://www.unavco.org/science/snapshots/solid-earth/2013/sturkell.html (Short version of above)
http://kortinu.info/test/pages/Hekla.php (On the March 2013 unrest)