Bárdarbunga – The Elevator to Hell

Icelandic BBQ. Photograph copyright by Eggert Norðdahl.

Due to me having had a couple of hectic weeks at my day job and catching this year’s influenza I have not gotten around to writing as much as I have wished. What I had wanted to do by now would have been to explain more in detail what is happening at the Icelandic eruption, but perhaps mostly why it is happening as it does.

The advantage is that we have seen volcanic history unfold and a type of eruption never witnessed before in the age of instrumentation. Early on Icelandic scientists tried to use modeling from the Krafla Fires eruption sequence and to be honest that led them quite astray. The Krafla Fires came out of a shallow dyke, whereas this one is much larger and deeper, skirting or breaching down to the mantle. Also, the size of the eruption is quite different.

As we look at the eruption via webcams, or look at pictures it is easy to think that this is a small eruption. But nothing could be further from the truth; it is just that the sheer scale of the Icelandic landscape is fooling us all.

In reality this is a major eruption, not on the brutal scale of let us say the Lakí eruption, but it is still massive. By now the lava flood covers 50 square kilometers, the edges are reported to be between 6 and 10 meters high, but that is not the average depth of the lava, that is more likely to be 30 meters and that would put the total output at around 1.2 cubic kilometers if we allow for the lower edges. Now, let us start comparing with other large eruptions.

Tolbachik eruption in december 2012.

Eyjafjallajökull is an eruption that has etched itself into our brains. That eruption coughed up as little as 0.15 km³ of lava equivalent. The lesser known Grimsvötn 2011 was the largest eruption in more than a hundred years in Iceland. It erupted about 0.4 to 0.5 km³ in lava equivalent. No, we must go further back and further out.

In 1975 the Great Tolbachik eruption took place, it erupted 1.2 km³ and it used to be the standard that every later effusive eruption was measured against. It was even called “The Great Tolbachik eruption”. But now Holuhraun is busily surpassing that figure.

Instead we have to go for the VEI-6 1991 eruption of Pinatubo to get figures that are larger. In lava equivalent that would be 4.5 km³. So far there is nothing saying that the Holuhraun will reach that large a figure, but on the other hand there is not much actually saying that it will not.

Pinatubo’s eruption in 1991, the second largest eruption of the twentieth century.

If we instead look at Iceland we will have to go back to the 1874-1875 cataclysmic Askja eruption to find a bigger eruption. That would place us at 1.8 km³ of lava equivalent. To get a larger Icelandic effusive eruption we have to go back to the 14 cubic kilometer Lakí eruption. I find it rather improbable that we will see something on the scale of Lakí.

At the current stable eruptive rate of 350 m³/s Holuhraun erupts 0.9 km³ per month and at that rate it will become a pretty large eruption soon, and remember that it is the world’s largest eruption in the last 23 years and Iceland’s largest in the last 139 years. Next time you wonder when the “real eruption” will begin, think about it again.

The caldera plug

I have used the word plug for the huge lump of rock that sits inside the circle of earthquakes that form the ring fault around the caldera floor. But plug is not a good description really, so perhaps we should come up with a better name. On the other hand it kind of describes it quite well.

This plug has been dropping at an even rate since the onset of the eruption, but there have been a couple of misconceptions about it that I wanted to explain better. I did so in a comment earlier today, but I thought it deserved to be expanded into a proper article.

The dynamics of the plug

Image by the Icelandic Met Office. The image shows the earthquakes outlining the ring fault and the dyke leading all the way to Holuhraun.

There is a huge lapsus in thought running around concerning this eruption. I will try to correct that now.

We know that magma is leaving the system at a rate of 350 m³/s lava equivalent and that it ends up at Holuhraun. We also know from GPS evidence that this is creating a system less pressurized. I am not using the word under pressured since this would implicate something else. I will now try to tie the knot on the sack.

Now, if a pressure of 1 is needed to keep the caldera lid(s) in place (the plug) we now have less than 1, it would probably be possible to calculate how far below 1 we are, but that would take quite a bit of calculations and contain a great bit of uncertainty at best. For now <1 is enough to know.

At the top it has always been assumed that there was a pancake shaped layer of magma, or a set of pancakes. The depth for these pancakes was assumed to be 2 km by the Icelandic scientists. Remember that this might not be entirely correct given new data. But, if we assume it to be correct this would most likely be stale magma that is unlikely to be eruptible and form a rhyolitic mush.

Below that there is now believed to be a secondary magma reservoir between 5 and 6 km depth (source: IMO), I think this forms another set of pancakes, or it might be a more solid “chamber” like structure. This most likely contains hotter eruptible material.

At about 10 km we have a larger chamber and between 16 and 20 km you have the start of a boat hull shaped reservoir stretching down to the mantle.

Image by the Icelandic Met Office. This is the station BARC residing on top of the ice on the Bárdarbuna caldera, notice the very steady drop.

And it is also believed that the entire shebang is connected via a permanently open conduit. Since we are talking about connected pressure vessels the pressure should be pretty much the same equilibrium of 1 everywhere, or in our case an equal <1.

Now, if it was higher somewhere in a single part of the system it would rapidly equal out to be the same all over. So, if we get a pressure below 1 at the deep chamber that feeds Holuhraun we get lowered pressure in all reservoirs causing a cascading chain of readjustments.

The magma at the top is most likely not going to go anywhere since it is stale rhyolitic mush, but magma from reservoir two is most likely going down to reservoir 3 to feed Holuhraun, magma should also be moving up from the “hull” into the feeding reservoir number 3. Basically, the system is at <1. This makes it pretty impossible for magma to move upwards and form new intrusions since the easiest path would be to continue out towards Holuhraun. At least until a path opens that goes straight to the top, then the resistance of the dyke might be greater than the extra height to reach the surface at the caldera floor.

Now, the general belief seems to be that the pressure is greater than 1 however impossible that would be. This would mean that more magma is entering into the system then is going out. This would at the very least leave the plug where it is. But it is more likely that it would either push the plug upwards, or that magma would move up from the “hull” into reservoir 3, onwards into reservoir two and then up into reservoir 1 were the rhyolite would be reheated, and we would have noticed that at least. But even if the rhyolite didn’t blow up we would be seeing inflation and a GPS permanently moving upwards and not down.

By now any good physics student should be raising their finger to say “but won’t the system be striving to achieve equilibrium in the pressure, i.e. 1?” But of course! And the only way to achieve that is by the lowering we are all following on the GPS at Bárdarbunga. It is equally marvelously linear as the output of at Holuhraun is stable.

Now one last thing that seems to have swept past everyone, the missing magma. Seemingly twice as much magma is either residing inside the dyke leading to Holuhraun or have been erupted at Holuhraun as is seemingly leaving Bárdarbungas magma reservoirs if we calculate it from the drop.

First, a small part of that differential is what is giving the <1 value of pressure, but it is a rather small part. Initially when I noticed this I thought it was magma from the initial intrusion that took place in the months prior to the eruption. But, with time it became apparent that this seemingly missing magma is also ridiculously linear.

350 m³/s of lava equivalent goes out, and the caldera plug drops 175 m³/s in lava equivalent. Now, the solution is to be had by the pesky pressure differentials. Not only is the plug dropping, there is also decompression melt running at the bottom of the “hull”.

Now, when I speak about “low pressure” it should be understood that the pressure is still tremendous. It is just lower then what is needed to keep the caldera plug up and lower than is necessary to keep the mantle material in a semi-solid. With decompression melt in this case it means that the pressure drop makes the semi-solid mantle transition into a more liquid state (magma).

Now some people seem to think that it is the dropping plug forcing the magma out, but this would just create another pressure level of 1 and we know that the drop only equates to half of the erupted lava, so there just simply has to be pressure below 1 at play, and it also pans out with the decompression melt model.

CARL

Sheepy Dalek Update (The bar is open) :

New riddles are up!!!

https://volcanocafe.wordpress.com/riddles/

Answer on that page only, please!