A few days ago, I was sipping on a rum and coke as I was flying over Kilimanjaro and the great continental divide ripping East Africa away from the embrace of the rest of Africa. As I did that, I got to ponder the other major tectonic powerhouse, Iceland.
People tend to forget or underestimate the effects of the African superplume and the Icelandic Mantle-plume on global tectonics. Only these major plumes can jostle about entire continents, or split continents apart.
There are of course a few other tectonic forces driving plate tectonics but compared to these plumes they are mere afterthoughts.
How do we know that these plumes are so important? The reason is simple, we can temporally lock the birth of the Icelandic plume (14.4 to 15.0 million years ago), with continental drift trajectory changes in the North American continent. But that is something best left to another article.
As I flew over Lake Turkana I got to think about the elephant in the room. The Icelandic volcano that I have not written about, the direct sibling of Hekla. In a tectonic sense of it, this volcano might be the most important volcano on earth. In other words, I have only told half of the volcanic story of Iceland. Time to fix things and go to the continental drift centre-of-the-world.
Unlike it’s mechanical sibling Hekla, Vatnafjöll is a highly unassuming volcano from a visual perspective. It looks like any other volcanoclastic ridgeline in Iceland. In fact, it was so unassuming that up until recently it was just seen as some sort of dead-beat sub-set of Hekla, and since it had not erupted since 350AD, it was even seen as a dead volcano.
Before we go into that, we need to look at the tectonic setting, because that one is interesting. Vatnafjöll is first of all a triple junction volcano connecting the end of the Southern Icelandic Seismic Zone (SISZ) to the Eastern Volcanic Zone (EVZ) and the beginning of the Vestmannaeyjar Volcanic Belt (VVB).
Most often Hekla get the distinction of being at the end of the SISZ, but that is not true as we will see below.
Now most of you will be slightly oogle-eyed trying to get the tectonic setting to fit with the lack of large earthquake swarms. A place like that should be rife with large tectonic and volcano-tectonic earthquake swarms.
And it is indeed highly tectonically active, it is even giving Bárðarbunga a run for its money if you tally up the Cumulative Seismic Moment over time. The difference is just that Vatnafjöll does them far bigger, and further apart.
So, according to volcanic wisdom, as we partied ourselves into the year of 1987, Vatnafjöll was a dead volcanic subset of Hekla. As per usual, nature was about to throw us a spanner into our wheel of faulty preconceptions.
Before we move further, I must mention a few things about the paper I am referencing, it contains a glaring error. It sets the MOHO at 13km depth, instead of the 24km (and counting), that we now know to be true.
The reason for this is not to be put on the highly accomplished scientists in the field that wrote the paper, it is just that new and far superior equipment has made it possible to find those tiny deep earthquakes that we nowadays love to follow.
This leads to errors in ground plasticity in their modelling and is making the depth centroid dubious. It also has effects on the size of the intrusion that I will discuss below.
That being said, the paper is still sound at the core, and it is one of the more important volcanic papers that nobody has read.
11.31 May 27, 1987 M5.8 Earthquake
On May 27 at 09.01 in the morning one of those events that will make volcanologists slobber with saliva started at the southern end of the Vatnafjöll proper. It was an M4.0 Earthquake with a tectonic signal, so far not much to write home about. But, at 09.11 the Búrfell strainmeter started to detect an anomalous strain change signal indicating that an intrusion had started somewhere.
Initially it was believed that the earthquake might have kickstarted Hekla, after all the epicentre was inside what was then believed to be the Hekla volcanic field.
As more fore chocks came rolling in at the same spot as the initial earthquake this was ruled out, partially by utilizing changes at other strainmeters.
At 11.31 the main earthquake occurred. It was a North-South strike-slip earthquake lateral dipping slightly to the right. The depth of the fault plane was 4-13 kilometres deep and the length was 10-12 kilometres long.
In the region (SISZ) large earthquakes tend to be slightly offset from the transform fault patterns that run to the NNE. In other words, small earthquakes tend to be running from SSW to NNE, but the large ones go on a more NS trajectory.
For the next couple of days after chocks followed and the strainmeter changes indicated that magma moved into the newly formed fault. In the end a 0.8-meter-wide, 10km long and 9-kilometre-deep dyke formed containing a sum of 0.072km3 of magma according to the signals detected at the time.
If we now would remove the errata up above from this calculation, we get a minimum volume of 0.16km3 of intruded magma in 2 days flat.
Now, in most cases a dyke carrying that volume residing at 4km depth would become a problem of eruptive nature. But, it obviously did not. It didn’t even result in the usual seismic swarms as the magma would try to move upwards due to natural buoyancy.
Instead the magma ended up in what I believe to be a central magma chamber. We will obviously return to that in the next part where we plot the living daylights out of Vatnafjöll. Caveat here, I have not seen the upcoming plots by Andrej Flis, so I am writing in the dark here and nature might show me wrong.
Now we have a constraint downwards on how large a seismic event is needed for Vatnafjöll to erupt. A volcano-tectonic M5.8 is just not cutting it.
May 6, 1912 M7.0 Earthquake
This earthquake had a hypocentre 10km west of the 1987 earthquake. It was NS strike-slip earthquake that also was slightly right-lateral dipping.
This earthquake extended to the surface and is as deep and extends for 30 kilometres. As such it was a tectonic heavy hitter. We can safely assume that it too was a volcano-tectonic event, because this time it produced two eruptions.
The known width of the dyke is 8 metres, so we get an intruded volume of 4.8km3.
On the 25th of April 1913 eruptions started at Lambafit and Mundafit along the earthquake surface rifts, and they lasted until the 18th of May the same year, the total length of the erupting fissures was 10 kilometres and about 1/10th of the intruded magma extruded through the vents. As such it was not that much to write home about on an Icelandic scale of eruptions.
Especially from a volcano that at numerous times during the Holocene has erupted between 5 and 15km3 of lava. For being an unassuming hyaloclastite ridge, Vatnafjöll is a major leaguer.
Constraints and effects
Now we know that an M5.8 is not large enough to cause an eruption, and that an M7.0 is enough to cause an eruption. Somewhere in between those two figures we would find the eruption threshold for Icelandic crust of similar type to suffer from a rifting fissure eruption originating bottom up, and not out of an obvious central volcano.
A volcano-tectonic event able to produce a large eruption (5 to 15) km3 of erupted magma would probably require an even larger main chock (or a series of large earthquakes). I have calculated the needed seismic force to be equal, or larger than M7.5, and since we know that large eruptions have occurred frequently here, we should probably up the largest possible magnitude for the area to around M7.5.
If we now transpose this into a similar setting with long repose times interspersed by major tectonic activity, like the Dead Zone, we get a major spanner in the flywheel for those who believe that we can have a Lakí event without major earthquakes.
To get Lakí we would need 8748 M5s, or 324 M6s, 12 M7.0s, or a single M7.5. If we average things a bit, we would most likely have a large number of earthquakes topping out at a maximum of an M7.
The conclusion is that we need to look closer at the volcano of Vatnafjöll to see if it is indeed a central volcano, and to closer study seismic trendlines. To do that we will use all of the recorded seismic data on the volcano, while trying to remove Hekla proper from the picture.