Being considered the world leading authority on something is probably more of a curse than a blessing. All of a sudden you have a reputation to defend and everyone will be annoyed with you when you do not agree with their interpretation about what is happening. To top it off there will be a not so small amount of professional envy directed at you.
I am the first to admit that accurately predicting the 2014 eruption at Holuhráun was a lot of fun. I will let the other parts leave be, but I will here address a certain problem that has been around for the last year or so.
During the last year I have been stating that the volcano was not ready to erupt and that it was still recuperating after the large eruption. I know this has annoyed quite a few of my readers, and it has also annoyed certain volcanologists since I have pointed out that there were no signs at all of what they were claiming to be happening in the volcano.
I have tried to explain why I held that firm opinion, but to put it in a short format. There has up until recently not been a single detectable signal indicating an increased risk for an eruption at the Bárðarbunga Volcanic System, instead all signs pointed towards a recuperation occurring after the eruption.
In the end science is reading the scientific evidence that is at hand, misconstruing evidence or stating what is not is not science, that is wishful thinking. Now that I got this off my chest we can get to the fun parts.
The 2014 eruption
To understand what is happening now we must go back in time to a year before the start of the large Holuhráun eruption.
First we have to remember that for reasons not entirely understood Bárðarbunga is the world’s noisiest volcano, even at the best of times it was repeatedly suffering from massive earthquakes and earthquake swarms. So, just because this volcano is causing a lot of seismic noise it is not a sign in and of itself that is close to an eruption.
The trick is to understand that not all earthquakes are equal. Most earthquakes at Bárðarbunga are tectonic in nature and most often are not indicative of volcanic activity. To understand what is happening you need to understand the importance of location, earthquake type differences, earthquake progression and so on and so forth. Otherwise you will inevitably over-interpret things at an alarming rate.
The last large eruption prior to 2014 was in 1910 and as far as we know there was no eruption between that one and a 30-minute-long intra-caldera eruption that happened in 1996. This eruption was probably caused by the same M5.2 non-double-couple non-volumetric-change earthquake that caused the Gjálp eruption at Grimsvötn.
In retrospect we can state that 1996 is the year when Bárdarbunga was fully charged and primed to set off in an eruption. Most likely this eruption was delayed by distal stress release from the 1.5 cubic kilometer eruption at Gjálp. Be that as it may, for some unknown reason the fuse was not lit on Bárðarbunga until 2014.
It is well known that the mantleplume under Iceland is cyclic in nature with a periodicity of 100 to 150 years between cycle peaks. There is as of now no explanation behind this cycle in magma production from the mantleplume.
The effect of this cyclicity is that we get far more eruptions during the peak and that those eruptions are larger than those during the waning part of the cycle. We knew already that we had entered a peak of that cycle, both from the time that had gone since the last peak and from the increased activity of Grimsvötn that first suffered the massive Gjálp eruption in 1996 and two medium-sized eruptions in 1998 and 2004. And as the massive 2011 eruption happened we saw that the magma composition had changed in such a manner that we knew that fresh magma had entered the system indicating the effects of a mantleplume activity peak.
The difference between Grimsvötn and all other volcanoes in Iceland is that Grimsvötns high rate of eruptions has kept all conduits open, this means that there will be little or no seismic evidence of magma rising.
We therefore had no clear image of what the start of the peak of mantleplume magma production would look like on the Icelandic seismometers. In 2012 I hypothesized that the effect would be like a pillow hitting a madras from below and that we would see a marked increase in deep earthquakes of low magnitude at depths larger than 20 kilometres and that they would be dispersed over many places directly affected by the head of the mantleplume.
In 2013 we started to see exactly that happening at Askja, Bárðarbunga, Kistufell, Urðarhals, Vónarskárð and at a spot that is most likely an unnamed volcanic field situated very close to where the dyke turned northwards toward Holuhráun. One of the least affected areas was Bárðarbunga, but there was a few 25km deep earthquakes 8km West-Northwest of the caldera towards Vónarskárð.
At that point my theory was that all we had to do was to sit back and wait for the fireworks to begin and that we in a while would see seismic activity pick up at one or more of those spots and that the pertinent place would be where we got a clear upwards moving stack of earthquakes indicating the formation of a magma conduit.
In early 2014 we started to see that there was increased seismic activity at (in order of amount of unrest, least to most) Kistufell, Bárðarbunga, The unnamed volcanic field and Urðarhals. At Bárðarbunga we saw earthquakes all over the place between 7 and 15 km depth around the magma reservoir indicating an increase in pressure. At the Unknown volcano a conduit formed from depth up to around 12km depth. Urðarhals never really got going beyond dispersed earthquakes. Instead it was Kistufell that turned really rambunctious.
Kistufell had at the time suffered no postglacial eruption so it was with fascination I followed how a conduit formed from depth and with even more fascination I saw how the rising magma ended up in a living breathing magma reservoir at about 8 km depth. There the magma stalled for a while.
At the same time tectonic earthquake activity picked up at Bárðarbunga. Both Kistufell and Bárðarbunga are on the same fissure swarm with several fissures running between the volcanoes, and in the end this turned out to be too much and the magma reservoir under Kistufell breached towards Bárðarbunga and a dyke formed from Kistufell into the magma reservoir of Bárðarbunga.
As Bárðarbunga was force fed magma it in turn breached towards the East and in the end 2.5 cubic kilometers of magma left the magma chamber out of which 1.5 cubic kilometers erupted through the vents at Holuhráun.
After the eruption none of these signs have been evident due to the widespread systematic pressure loss caused by the large eruption. Instead we saw a high level of seismic unrest with signatures indicative of settlement inside the caldera and deep refilling of the lower magma reservoir under Bárðarbunga.
Due to the very large amount of magma that left the magmatic system of Bárðarbunga I stated that it would take a minimum of two years until Bárðarbunga would be able to erupt again. Even at such a productive place as this it takes time to produce a pressure equivalent of 2.5 cubic kilometers of magma in the magma chamber. Do note that pressure equivalent is not the same as the actual volume, it will take between 10 and 100 years to produce that volume of magma.
During the following two years we saw brittle tectonic earthquakes with slip direction indicating a massively dropping caldera plug and we also saw that there was a general deep inflation on the GPS stations. Or in other words, common recuperation after a large eruption in an active system on top of a large mantleplume.
So, for two years I played the part of being a spoilsport once every single week explaining that the M3+ earthquakes was boring and not a sign of impending ultra-doom.
What is new?
A couple of months ago, an uptick in those deep unassuming earthquakes at 25km depth started again. First at the unnamed volcanic field NNE of Grimsvötn that the Holuhráun dyke bounced against as it turned north, then at a spot south of Grimsvötn at a volcano named Hábunga, then at Askja and right before Christmas at a spot 8 kilometers WNW of Bárðarbunga.
If we now utilize what we learned during the runup to Holuhráun we can surmise that the next eruption will occur at one of these places. Since Grimsvötn is an open-ended system we should probably include that volcano since it will not suffer from deep pressure earthquakes since the magma would go straight up without causing noise.
From GPS-data we know that Hábunga has inflated a lot since May so that is a good spot to suspect an eruption, but since that volcano has not erupted for a long time and there has been a lack of seismic unrest there I think we must wait for anything to happen there, if anything at all happens.
So far we have not seen activity indicating rising magma at the unnamed volcanic field meaning that we are no nearer to an eruption than before. Askja could have been a good bet, but nothing seems to be out of the ordinary there as of now.
This leaves us with Bárðarbunga itself. And finally the friends of Bárðarbunga has some good news. Earlier today there was a very short lived earthquake swarm. If you just look at the location and type, there is not much that looks different compared to previous brief outbursts. Today’s event is one of the more short-lived, it was formed by rather few earthquakes and the size is to be honest not much to write home about.
So, why now should we be interested? Well, the answer is that the signatures of the earthquakes are different. Previously the signals have had a sharp clean initial break indicating tectonic type earthquakes. This time around the M3.5 earthquake started with an indistinct “muffled” part prior to a large tectonic-looking part. It is then followed by an indistinct smaller earthquake with a prolonged tremor-coda and several even more indistinct earthquakes. This points towards magma moving causing increased pressure that in turn caused a tectonic earthquake with magma rushing in filling the fractured rock with repeated smaller volcano-tectonic type earthquakes. If magma is involved we should be seeing a low-frequency component to the earthquake and lo and behold, we did.
During the Holuhráun eruption there were episodes of unrest running from Bárðarbunga towards Vónarskárð, it was at the time interpreted as a secondary dyke running towards the WNW, but there is an equal chance that this is a part of the magma reservoir of Bárðarbunga that is situated outside of the caldera, or that this is a permanent feature that is where the magma leaves Bárðarbunga when an eruption occurs at Veiðivötn.
What happened today is not the runup of an eruption, but it marks the turning point for Bárðarbunga when it could erupt. It is though good to cast one’s memory backwards and remember the incredible hubbub that preceded the 2014 eruption. But, if you see an earthquake swarm start that runs for hours with several large earthquakes then it is time to let your imagination run free. And yes, you all have my blessing to go Bárðarbonkers. Because soon the Bárðarbung will leave the Bárðarbunghole and we will have a bit of firework again. It is though still too early to say when, where and exactly how yet.