Part 1 of a Greip Series
This week, under the rumblings of Torfajökull, we return to one of our favourite areas within Vatnajökull that is yet to show it’s true hand. Or has it already and is there evidence to support that? Due to some recent activity, Carl has been getting increasingly excited in the back channels and put his secret volcano lair on hold. He wanted to write another article on the subject of Greip with some interesting interpretations and speculations about future possibilities. It’s been a couple of years since I wrote the original Greip article and she’s been rather busy during that time. Since then I had wanted to revisit the deadzone as well as some other article ideas but a rather long period of anxiety and depression decided to make an appearance and five months off work flew by without any constructive input into anything, let alone the VC blog. I had lost all interest in my interests, hobbies and passions. Other than my family and friends that supported me through it, the one passion that did help was music. That gave me the drive to get up and do something constructive as well as appear in front of a crowd. Well, as much of a crowd as you get with doom metal! So, this is my cathartic piece now I’m back and able to contribute.
Carl and I decided to collaborate on this one as we wanted to do an in-depth series and make a claim on Greip and what it could become. Carl will be discussing GPS movements and musings and myself covering the earthquake activity and graphical speculations. So, without further ado…
Wind of Change
So, what’s changed since July 2017? Well, let’s take a broader look at the deadzone and Vatnajökull first.
Things seem to have picked up in recent months with a few volcanoes showing signs of unrest. Öræfajökull has been discussed at length and excitement was growing last year that we might see some red hot volcanic action, but now it’s gone relatively quiet, save for a few scattered quakes. We’ve also been excited to see a few quakes in the deadzone, some on the same line as the 939 AD Eldgjá fissure and some on the Bárðarbunga and Grímsvötn fissure swarms. We know about the other usually noisy volcanic culprits so I won’t detail them, but what stands out here is the activity along the Grímsvötn fissure swarm; from the eastern and southern edges of the Grímsvötn caldera system, through Haabunga and onto Thordarhyrna.
We’ve discussed Grímsvötn’s climb up the cumulative plot in other posts and an eruption here could occur soon, with the right nudge and preceded by a decent earthquake swarm. But, I’m not going to speculate on that here as Albert and Carl have already done that. What I’d rather speculate on is the nature of a future eruption, which I will do so as we work our way through Andrej’s map porn.
So, is there anything interesting going on? The middle deadzone is still a relatively quiet place and any quakes are in the top 10 km of crust. Whilst we’ve seen a few quakes on old fissure lines it certainly does not mean they’re about to wake up and have ‘Jesperitus’ (lots of runny lava). It could, however, point to increased tectonic strain in the area resulting in brittle fractures of upper crustal rock. As we near the peak of the rifting cycle, this strain will only increase in intensity until relieved at some point along the rift. We’ve also seen that Torfajökull has been noisy recently with a few swarms occurring near the Bárðarbunga fissures that enter it on its western margin. The recent activity along the southern tip of 1477 Veidivötn fissure occurred after this plot was made, so for those details please refer to the plots in this article. Some deep quakes can also be seen under Vatnajökull, under the Thordarhyrna and Hamarinn systems.
Grímsvötn Fissure Swarm
Now we’ll move on to Grímsvötn and friends to see what’s been going on there.
Whilst we’ve waited an age for Grímsvötn’s cumulative plot to get out of the blocks, it has recently started to pick up speed, finally. Activity around the caldera is commonly found around the southern caldera rim where the most recent eruptions have occurred. Lately, more shallow earthquakes have occurred towards the eastern caldera rim as well as some mid-deep quakes. As is Grímsvötn’s nature, deep quakes are rarer than potassium woven swimming trunks. We can postulate that the lack of deep >20 km quakes under Grímsvötn is indicative of an open conduit system or a wide area of hot ductile rock underlying the colder intrusive bodies of the upper caldera system. Melt influx from depth into the upper magma chamber(s) should be mostly silent, mostly. Expansion of the upper magma chamber(s), and the increase in earthquakes generated may be our only seismic clue to the pre-eruptive state of the volcano, we don’t get the earlier warning of deep quakes. There have been a few quakes stretching out to the NE towards Greip, but not enough to draw any serious beard stroking of a connection between the two, but more on that later.
On to the troubled sister. She’s been busy, busy enough to repeatedly catch many an eye on the blog and stir up continued interest. What stands out more than anything in the figure above is just how noisy Greip has been in just two years. That’s a lot of data points and I’d hate to imagine the global warming input from Andrej’s PC from running and analysing all this data.
As you can see from the graph above, the frequency of earthquakes has been much higher in the last couple of years. If we look back further and discounting the noise from the August 2014 dyke intrusion, we see that there is a burst of activity following the end of the eruption in 2015. It appears that something changed at this point in the Greip system. The deep area between 15 – 25 km depth has been very active since Holuhraun finished erupting and is the main focus of Greip’s activity. Magnitude also appears to be on the rise but we may need another few years of data to establish any definite upward trends for this.
We’re not sure why the 2004 – 2010 time period has such high magnitude data in comparison to the rest of the data and it skews the plot somewhat. Maybe mislocation of noisy quakes from bardy and Grímsvötn were attributed to Greip due to the poor seismic network at the time. It does seem strange that all lower magnitude quakes are missing even though they were detected before this time period. Moving to more recent times, the events of 2014 can be seen as a prominent vertical line. But there’s that other ‘stack’ that appears in 2015 as bardy finally stops dropping and reinflation starts anew.
Here we see the same trends as the other plots in terms of activity bursts. Another noticeable feature is the separation of surface quakes from deep quakes. Maybe a possible lid on surface progress due to a more resistant crustal layer or the existence of an already present melt pathway laterally at approximately 12 km into other volcanoes in the area. Something I’ll discuss later.
Starting in the lair of Cthulhu (unofficially >20 km depth), we see many, small magnitude quakes indicating a continuous rate of influx of melt from the MOHO. How do we know they are due to magma movements? Well, thankfully lots of other scientists have studied these quakes and their low-frequency signals. Being a bass player in B standard tuning, it’s not what you hear, it’s what you physically feel as your internal organs start to liquify (think just off brown note territory). Deep, magmatic earthquakes have low-frequency profiles. There’s only hot, ductile crust down there so no high-frequency brittle fractures like you get at the surface. Some fracturing of the crust does occur during dyke and sill formation, but any high frequencies are tempered somewhat by the Queen-sized pressure. Here is a snippet from a paper on Askja discussing deep, magmatic quakes:
“There is a possibility that the lower-crustal earthquakes have a tectonic, not magmatic origin. However, the striking mid-crustal gap in seismicity from 8–12 km depth, and non-overlapping epicentral distribution between the upper-crustal (tectonic) seismicity and the lower-crustal events is evidence against this explanation.” H. Soosalu, et al. (2009)
This has been a common pattern since the end of Holuhraun, see Figure 6. above. Are we seeing increased melt influx during recent years or are we seeing a particularly noisy episode of sill storage expansion via CO2 exsolution or crystallization in situ? We’ll have to wait on that one until more evidence comes to light or we see more data patterns over time. The earthquake scattering pattern splays out towards depth as we would expect from it sitting right on the active extensional rift zone, with each side being pulled apart in opposite directions. Remember the upside down boat hull that’s often discussed? More on that later…
Moving up, the 20 – 10 km midsection is represented by a vertical column that is fairly horizontally confined. We’ve seen some greater magnitude earthquakes in this region recently, above that of the general tick. If we infer that all quakes in this area are due to sill expansion then we could imply that the growth of the plumbing system has accelerated in recent years due to higher melt influx. With the additional data from the next 2-3 years a much clearer pattern may be apparent and we could start speculating on cumulative plots like the Grímsvötn one. Estimates of the volume are possible using the denser earthquake clusters, but they are likely to have a large margin of error. I’ll leave that for others to speculate on.
The view from the east shows the cooling dyke as a separate column to the right of Greip. Can we imply a connection from the few quakes at ~15 km that seem to connect the two columns? Not from this data set, but the evidence for a connection was much stronger during the 2014 eruption and from an increased dataset and subsequent papers on the subject.
The upper 10 – 5 km section is mostly void of earthquakes, especially if we discount the dyke. As discussed in H. Soosalu, et al. a gap of seismicity between lower and upper crustal earthquakes lends weight to the deep quakes at Greip being of magmatic origin.
The upper 5 km of the crust has a reasonable scattering of earthquakes and we can infer, through the logic above, that these are of tectonic origin. We are, after all, sat on the middle of the rift. Greip also has some big neighbours, all of which offer their bulk and stress to the surrounding area. I’m not convinced of any hydrothermal activity here, I’m certainly not aware of any evidence for hydrothermal activity at this point. The cauldrons seen in the ice 2014-2015 were much further north above the dyke path and into the graben. Do correct me if I’ve missed something in Icelandic news.
The shape profile of each depth layer lends itself to a rather interesting situation. We see the deep earthquakes aligned with the Grímsvötn fissure swarm in the SW-NE bearing and we see the yellow midsection reaching out to Bárðarbunga and the dyke path. That’s not to say that it’s definitely on Grímsvötn’s fissure swarm; this orientation is common due to the rift angle at this point in Iceland.
Volcanoes in Iceland have a bit of a habit of stepping on each other’s toes, especially where fissure swarms overlap. Take a look at the recent unrest at Torfajökull along the historic Veidivotn fissure lineament and, after layering historic fissure eruptions on top, you can see what Bárðarbunga has done to poor Torfajökull over the years. Various intrusions have ripped this volcano apart and left it a steaming pile of beautiful rusty rhyolite hills. Holuhraun I was initially attributed to Askja, but after an investigation during Holuhraun II, this was revealed as a first jab by Bárðarbunga. Redrawing the fissure swarm areas of these volcanoes with the increased data from petrology led to a fair overlap in the lava plains beyond Vatnajökull. There’s no shortage of takers in Iceland when the rift opens and they certainly don’t abide by our lines and boundaries drawn on maps.
Activity to the NE, Kistufell, is often suggested on the blog to be the originating point of the 2014 dyke intrusion due to its place on top of the plume head and the presence of precursor earthquakes prior to the main unrest at Bárðarbunga’s caldera. Now, after a bit of digging and nodding off whilst reading scientific papers in the early hours of the morning after being woken up by the kids, I came across a paper by Green (2015). In it, he discusses the stress changes induced by the dyke’s propagation. He discounts the precursor activity at Kistufell as the source of the intrusion and attributes it to a building stress field from magma accumulation under the caldera. During the dykes propagation to the SE and then to the NE, secondary tectonic earthquakes were generated and then dampened by the changes in the positive and negative stress field, respectively.
This figure above from Hudson, et al. (2017) implies a connection with Bárðarbunga using the data from the 2014 eruption. It also gives a nice hypothesis on why there was an area of aseismicity at the knee/elbow of the propagating dyke; an area of shallow melt accumulation, and therefore ductile rock, at the base of the brittle upper crust. The seismic analysis of some of the quakes within the Greip column has been interpreted as non-double-couple brittle rock fracturing within the sill complex as melt enters the system once rock fracturing has taken place. This appears to be a continuous process as earthquakes appear weekly, if not daily at times at Greip.
“There appears to be an approximately constant average rate of deep seismicity over the time period, with no obvious change associated with the eruption. An apparent lack of deep seismicity during the dyke intrusion and eruptive periods shown in Figure 2 could indicate a genuine lack of any melt movement at depth during that period, but it is more likely that the seismometer network is less sensitive due to the many earthquakes generated during the dyke intrusion and eruption”
Looking at the activity from the last two years between Bárðarbunga and Greip, we see no clear signal of seismicity connecting the two. We see residual quakes from the 2014 dyke cooling and contracting. If there is an aseismic zone connecting the two we are best looking at alternative sensors to help us see what’s happening. There is evidence that melt accumulation started again following the end of the eruption in 2015 (Jónsdóttir 2017) but GPS readings don’t tie up – Carl will discuss this in his part.
Bárðarbunga is likely to have multiple feeds up from the lower reservoir, due to its mature stage in life, with some being almost aseismic. We do see deep earthquakes indicative of magma movement under the caldera with the melt constantly percolating upwards and filling the sills and magma chamber(s).
So, let’s say Greip is connected to Bárðarbunga by a noodly magmatic appendage, is there any evidence for a connection to Grímsvötn?
The greater distance between Grímsvötn and Greip is the first thing that stands out in the figure above. But, that didn’t stop Gjálp back in 1996 from getting Bárðarbunga and Grímsvötn sharing bodily fluids. Was Greip involved in priming this eruption via feeding Bárðarbunga or Grímsvötn? It’s possible, but sadly the sensor network wasn’t sensitive enough to pick up the low magnitude quakes at Greip. I suppose a by proxy connection to Grímsvötn via Bárðarbunga could be implied if a little contrived.
The main eruptive activity at Grímsvötn has recently been on the southern rim of the clustered caldera complex, quakes in this area are no surprise. What is noticeable is the deeper activity in the NE area of the complex, an area that not been active for some time. Could this eastern caldera reactivate? Are we seeing increased melt influx across the broader NE area of the Grímsvötn fissure swarm and can this be linked to Greip? That’s hard to say as it could easily be tectonic in nature with the strain being released along the rift.
If we infer that Greip is the next volcano inline within the Grímsvötn fissure swarm then we could see co-eruptive behaviour during a major rifting event. We could witness dykes propagating between them through seismic activity, but more than anything else we would be in awe of watching a volcano being born in an area where data and imagery are so rich.
One thing that stands out above all is the offset of Greip from the other neighbouring volcanoes if we are to interconnect them. We grow up seeing the classic volcano illustrations in school books of a nice straight magmatic conduit rising up and leading to a perfect layered cone belching pulverised rock to the wind. In some settings this could be true, but not in Iceland where everything has been stretched and squeezed over the millennia. The fractured nature of a rift zone creates many variable paths for magma to work its way up.
We saw this offset of deep magmatic influx at El Hierro in 2011 (Domínguez Cerdeña 2013) as well as the current and historic activity at Askja where the offset is to the NE (Soosalu 2009). Also, Katla’s main deep feed can be found on the east side of the caldera imaged by recent seismicity, but it’s not as displaced as the others mentioned above.
So, it’s completely possible for Greip to be a melt feeder, especially for Bárðarbunga, but could it be more in the future?
So what does all this evidence mean for the future? Well, for a start we desperately need some more sensors in the area, the more data we have the better supported our hypotheses are. Are we finally seeing the magmatic appendages of the Illuminati of the volcano world and its influence on the volcanic powerhouses of Vatnajökull? Is this a sign of the top of the mantle plume head nearing its peak and the main source of magma for the plume volcanoes? Or are we seeing a formative volcano in the making that reached out and had the power to slap Bárðarbunga 90° and may be about to steal its magma as well as its thunder? On to Carl for more ruminant ruminations on that subject.
Domínguez Cerdeña, et al. (2013) – Seismicity Patterns Prior to the 2011 El Hierro Eruption
Green, et. al. (2015) – Triggered earthquakes suppressed by an evolving stress shadow from a propagating dyke
Jónsdóttir, et al., (2017) – Bárðarbunga volcano – post-eruption trends following the Holuhraun eruption in 2014-2015
Hudson, et al. (2017) – Deep crustal melt plumbing of Bárðarbunga volcano, Iceland
Soosalu, et al. (2009) – Lower-crustal earthquakes caused by magma movement
beneath Askja volcano on the north Iceland rift