I was asked to write a small weekend piece while we wait for Albert to finish the second part about Grimsvötn. My original idea was to write about people eating volcanoes, but thankfully Iceland saved us from that.
As many of you have noticed Grimsvötn has thrown some big ones since my part of the two-part article, but I know that Albert is diligently working those ones into his article, so I will leave out Grimsvötn being hard at work to meet that 31st December 23.59.59 o’clock deadline to outshine the Reykjavik fireworks.
Iceland is most likely the only place on earth that will cause people to look down towards the ground if you say, “Oh, look at the stars”.
As many of you know a “star” is an earthquake above M3 on the Icelandic Met Office earthquake page. As I write this there are 3 stars in Iceland at Bláfjallaskal, Grimsvötn and Hamarinn. And as we all know, big earthquakes in volcanoes can be signs that things are about to go down.
Problem is that Hamarinn can throw these now and then without any impending doom, and the same goes for the tectonically active part of Iceland that houses Bláfjallaskal. So, if big things are not necessarily a sign of big eruptions, where do we look?
The last 24 hours are among the most interesting and downright intriguing in volcanological history, and that is even counting eruptions like Cerro Hudson, Holuhraun and Pinatubo. The air was filled with the brimstone smell of heavy science data collection being done in the dark.
Before and during the 2000 eruption of Hekla the monitoring network was not bad for its time but compared to today it was a kayak crossing a small lake compared to a rocket hurling a Tesla playing David Bowie towards Mars.
Not only have we gotten more seismometers, GPS-stations, strain meters and a new funky multigas-measuring station. All of the old equipment has been replaced with new and far more sensitive instruments.
Basically, what will trigger the automatic system today would be hard to manually find back then. Today the automatic threshold is a staggering -1.2 on the moment magnitude scale. And it is possible to manually see earthquakes that are even smaller.
The same level of upgrades is also there for the GPS and the strain meters. And we have the gas station on top of the mountain and an infra-sound detector down at Katla.
Yes, there are more pieces of equipment on a couple of volcanoes, but nowhere on earth is the average density versus quality higher.
I am mentioning this for two reasons. One is so that you will understand that there are more earthquakes picked up now than during the last eruption. And the second reason is that we can now be almost certain that we will be able to pick up the pre-eruption signs of Hekla.
Obviously, we are not entirely sure what those signs are, since we have never seen them before, but we can make educated guesses. And, we can also compare it to the 2000 eruption signs that we did pick up and extrapolate from there.
Only problem is that apparently Hekla is throwing us a Black Swan moment. Because according to our known extrapolations and guesstimates Hekla started an eruption this morning. At least if we go with the seismometer network data.
So, where is the brimstone and fiery doom that our beloved Hekla is famous for? I will now try to answer that question.
In 2011 all seismic network upgrades except station HES was in place. From that moment the data quality does not increase a lot and we can safely say that there is no large technology skew increasing the numbers of earthquakes. Well, except in the direction of HES obviously, and that HES gave us better depth resolution.
The 2013 Hekla Seismic Crisis
In 2013 Hekla threw up it’s most numerous post eruptive earthquake swarm. At the time it was assumed that Hekla would erupt, and even the Icelandic Met Office put up a bulletin that an eruption at Hekla was imminent. I was also fooled by Hekla and thought it would go off.
In the end no eruption occurred but post the seismic crisis that has been the benchmark of what the maximum unrest Hekla could withstand without an eruption following.
I had a bit of a problem with that idea. The location of those earthquakes was not in line with what had been observed prior to Hekla’s last eruption in 2000.
In 2000 against what most people believe there was a small swarm about two weeks prior to the onset of eruption. That small swarm occurred slightly west of the southern terminus of the Heklugjá fissure swarm.
It is normally said that Heklugjá is 7 kilometres long and that it runs the length of the actual mountain. But that is the top of the fissure that normally opens during an eruption, the fissure at depth is a bit longer. Perhaps 3 km to the NNE, and some 7 to 10 kilometres further SSW. This is very clear on the plots that Andrej has made for this article.
The pre-eruption swarm in 2000 is a logical place swarming for a fissure system under intense magmatic pressure, because the pressure will attack the weakest points to make room for that magma.
The NNE end is fairly blocked off due to the neighbouring Torfajökull volcano, and the Vonarskard and Veidivötn fissure swarms taking care of the business Icelandic continental spreading.
That means that the best spot to grow is the SSW end, and this is evidenced by that part being considerably longer than the NNE part.
In 2000 the ever-increasing magma pressure tried to make more room at the southern end causing pressure induced earthquakes there. In the end that did not work well enough, and the pressure after a while attacked another weak spot, the topmost part of Heklugjá that runs the length of the edifice known as Hekla proper.
No such thing was seen in 2013, so in hindsight the obvious conclusion is that the pressure was not high enough for the earthquakes to be part of a pre-eruptive sequence.
If the relatively small earthquake crisis in 2013 can be called a seismic crisis what we are seeing now is more like seismageddon (in a piddly Hekla way). Just to put it into perspective, the total earthquake count now is 3 times larger, and the amount of earthquakes per month above M0.8 is by now twice the number.
So, just on pure numbers this has transcended 2013. But that is not the point that is important. The important thing is the location.
This time a large percentage of the seismic activity is slightly west of the southern terminus of Heklugjá indicating intense pressure. The earthquakes are though small, the same as in 2000.
Since the network now is far more sensitive than in 2000, we are seeing quite a few more small earthquakes. That is why the IMO is using the M0.8 and upwards as the divider for their master plot. A M0.8 earthquake was easy to detect in 2000, and none of those would have been missed even at the old network, so this makes it easy and accurate to compare with now.
All things taken together, we know that the GPS-system says that the pressure levels by now are higher than they where back before the 2000 eruption, we are seeing magma system pressure earthquakes at the correct location, but we are not seeing any untoward twitching on the strain meters.
So basically, we have 2 out of 3 of the prerequisites known for an upcoming Hekla eruption. And the strain twitching normally starts just a few minutes prior to onset of eruption.
I will not say that Hekla will erupt soon. Suffice it to say that there are far more signs of an eruption being on the way now compared to 2013.
It has been an interesting and exhausting 24 hours, and even if there will not be an eruption this time around, Hekla gave us all a true masterclass in volcanology, and for that we will have to be thankful.