Greip, Gjálp’s Mysterious Sister

Mythology and Name

In Norse mythology, there were two tempestuous giantesses, Gjálp and her sister Greip. They were the daughters of Geirröd, a giant who really didn’t like Thor… maybe it was hammer envy? Surprise surprise, Loki was involved as well, up to his usual tricks of winding people up and leaving joke-shop items lying around no doubt. So, Geirröd wanted Thor dead and so hatched a plan to lure him to his demise. On Thor’s way to Geirröd’s castle, Gjálp tried to help her father out by flooding the river Virmur as he crossed. This didn’t go well for Gjálp who ended up with a large boulder tossed at her from a damp and moody Thor. Once at Geirröd’s castle, an incident with a booby-trapped chair backfires, literally, on Geirröd and Thor ends up falling through the floor. The mass of Thor’s armour-clad beefcake body combined with gravity induced velocity results in the breaking of both Gjálp’s and Greip’s backs. This then escalated into a bit of a Kill Bill moment, complete with Norse ‘alarm’ music and a heavy dose of artistic license from myself, and Thor ends up killing them all. Probably brutally, with a red hot iron bolt involved… They should probably make a film out of that.

On to the name. The IMO had already assigned the name Gjálp to the 1996 eruption between Grímsvötn and Bárdarbunga, for sound reasons. The eruption melted 3km3 of ice and the resulting meltwater flowed via a narrow channel into Grímsvötn’s subglacial lake which then later escaped as an impressive jökulhlaup. Whether Thor was crossing the river Skeidará at the time has yet to be confirmed, but it certainly took out the bridge there.

We now arrive, logically, at the name Greip as put forward by Andrej during an evening of back-channel discussions. But what/where is Greip? I hear you cry! Well, it’s a volcanic feature that has been mentioned on the blog over the last year or so as another name. That name is ‘Unknownabunga’, however, the -bunga suffix is misleading in this case as there is no ‘bulge’, but it was useful as a stop-gap name whilst we consulted with the Norse gods.


Some of you may already know the location of this feature and have been drawn to the earthquakes that have happened in this area over recent years. This area was probably best revealed during the August 2014 eruption of Bárdarbunga. During the eruption, the dike that travelled out of Bárdarbunga along a radial fault in an SE direction suddenly turned NE creating a distinctive elbow shape in the dike path. It has been proposed that magmatic pressure within the Greip area was higher than the magma in the migrating dike and so deflected the dike along a lateral fault at the SE edge of Bárdarbunga’s fissure swarm.

Location of Greip on the VC map with nearby features labelled for reference. It is implied that Greip is an extension of the Grímsvötn fissure swarm in this image. Image: Beardy Gaz, Data: IMO, FutureVolc, et al. Base Map: Google Maps.

There could be alternative reasons for this deflection though, one of which could be that this was the optimal line of least resistance after the dike travelled along the radial fault. This lateral line of weakness could have been a pre-existing fault or graben that was hidden from view under many hundreds of meters of ice. This deflection may have happened regardless of the magmatic pressure in both systems. But if we base our hypothesis of the deflected dike path on higher magmatic pressure from Greip, what could the potential different outcome have been with lower magnetic pressure in this system? Maybe the dike would have entered Greip instead of the lateral fault and we could have seen a rather more different eruption spectacle. We’ll never fully know the list the variables and potential outcomes during these volcanic events, but aren’t chaotic systems fascinating!

Another assumption I have made is the parental system. In this case an extension of the Grímsvötn fissure swarm. I’ve made this assumption purely on the orientation of the fissure swarms in this area and the fact that the ‘official’ fissure swarm, as implied by FutureVolc et al, sits right next door to Greip, to the SW. Another theory is that this is an extension of the Bárdarbunga fissure swarm along a similar trend to the line of the 2014 eruption radial fault. I will return to this later when we look at the topography of the area.There is also potential that Greip is a completely separate entity and may not belong to any identified fissure swarm or central volcano with its own unique magmatic signature. With no known eruptions and no rock samples to perform a petrological analysis on the origins and family history of this feature remain a mystery. I will return to petrology later in the article.


We know the activity of the dike during the 2014 Bárdarbunga eruption highlighted Greip on the map, literally, but what evidence is there of previous activity? Prior to 2004, there was little detectable earthquake data available, I found only one reference to a Greip earthquake in the period 1975-1985 and a handful of ones between 1995-2004. We know that the sensitivity of the seismic network has increased 10-100-1000 fold over the recent years and this area is no exception. Positioned in the middle of the Vatnajökull ice cap, and with no protruding bedrock, detection of earthquakes in this area relies on the sensitivity of the surrounding seismometers at Grímsvötn, Bárdarbunga and Kverkfjöll, as an example of three, to triangulate the location, magnitude and depth.

An aerial view graphic map showing 2004-2017 earthquakes over the northern part of Vatnajökull. The main central volcanoes are outlined. Greip can be seen to the SE of the elbow in 2014 dike. Map: Andrej Flis, Data: IMO.

From this relatively short period of time, we do have a good picture of what’s hiding under the ice other than ‘The Thing’ and a crashed spaceship. The map is saturated with shallow earthquakes, especially from the dropping of Bárdarbunga’s caldera and the dike path during 2014-2015. Hydrothermal activity shows up in the top few kilometres of crust and this can be seen along the Hamarinn-Loki-Fögrufjöll system(s) and Kverkfjöll. Activity around Grímsvötn is from the eruptions in 2004 and 2011 and the build up to the next eruption. The activity at Greip is predominantly between 15-25km deep indicating that magma is the culprit here but has little surface manifestation.

If we look at the depth profile of Greip from the south with all earthquakes plotted we see a column of earthquakes from ~25km deep all the way to the surface. However, this is not a clear picture of the system and it highlights the importance of having alternative views, models and different measurements to build up an accurate picture of what has and could be happening.

Depth plot of all 2000 – 2017 earthquakes in the Greip area looking from the South. Plot: Andrew Flis, Data: IMO.

So if this isn’t a single column then what is it? It’s actually two separate events that look like one entity on this view. The bottom half, below 15km, is predominantly Greip whilst the top half is predominately the 2014+ dike activity. This becomes clear if we take a south-westerly view of the area.

Depth plot of all 2000 – 2017 earthquakes in the Greip area looking from the SW. Plot: Andrew Flis, Data: IMO.

We now see the that the single column as seen from the south is detached around 15km deep with dike activity sitting to the north of Greip and this correlates with the aerial presentation of the data. Greip proper consists of, dare I say it again, a wedge of earthquake activity between 25-15km depth. Andrej refers to this as a dike swarm, pictured as a branching lightning bolt discharge, and his description has merit when it comes to visualising what’s happening deep in the crust. In this case, the branching matrix is made up of multiple dikes throughout the crust. There’s no evidence of any magma chamber or conduit from the seismic data and the best way to describe this, from my point of view, is a proto-chamber/reservoir of magma, almost like a honeycomb/sponge filled with water, but not as densely intruded with magma like the mature chambers we’re familiar with at the main central volcanoes. This could be a very young feature we are starting to see and something that may have happened at the other volcanoes during their gestation. There appears to be a ceiling around 15km which contains the majority of the dike swarm below this level. The overlying rock strata may be harder in this area and therefore less susceptible to failure {cue imaginary PHD research montage}.

Static graphics are extremely useful for displaying data, but we are still limited in the clarity of the presentation in two dimensions. If we bring time into the equation then we get a fantastic visualisation tool of cumulative activity year by year and how activity in one area affects other areas and even the cascade of earthquakes along multiple lines of weakness.

Cumulative depth plot of all 2003 – 2017 earthquakes in the Greip area looking from the East. Plot: Andrew Flis, Data: IMO.

At the start of the cumulative time sequence, we do actually build a picture up of a single column with earthquakes at all depths. Although the main Greip activity is below 15km, from magma influx and therefore increasing pressure, it will have corresponding upper crustal earthquakes associated with stress changes from doming of the area as well as seasonal earthquakes related to ice loading from changes in mass of the overlying glacier.

In 2014, and subsequent years, you can see the dike earthquakes propagate in the upper right of the plot which quickly fills this area. At the same time, there is an increase in the number of earthquakes at Greip proper. The effect of the dike passing by may have weakened the surrounding crust allowing for more rock fracturing and magma influx. However, during this period many smaller earthquakes were recorded in both areas and this increase in quakes at Greip proper may be related to increased sensitivity and the resulting higher resolution of the network during the 2014 eruption. As we are dealing with such a ‘new’ process, from a human observed timescale and viewpoint, we can’t be sure of all the cause and effect processes and the nature of the beast. We would need many years of data from this area to build up a clearer image of what Greip is up to.

Topography and Anomalies

So we finally get to the topography of the area and here’s where the assumptions and possibilities of the area really start flying around like deranged monkeys on an evil mission. Although this original map was created in 1988 and the resolution is not as high as it would be if conducted today, it does help in probing the depths and trying to work out what’s going on at Greip.

Topography map from radio-echo soundings with an overlay showing the location of the main volcanic features in relation to the underlying topography of northern Vatnajökull. Overlay: Beardy Gaz. Base map: Björnsson 1988.

We’ll start with the Grímsvötn area first and work our way round. Although the base map was produced before the Gjálp eruption in 1996 a small ridge can be seen extending to the north of Grímsvötn around the overlay marker for Gjálp. It appears this area has been active before in the past and this ridge may well have been formed from multiple radial eruptions from Grímsvötn creating a hyaloclastite ridge. To the NW we have the Loki-Fögrufjöll ridge and its connection to the Hamarinn system. A link between Bárdarbunga and Hamarinn could be inferred purely from topography, but the existence of Hamarinn as a separate central volcano, rather than part of Bárdarbunga’s fissure swarm, is still hotly debated. To the South, we see the connection to Háabunga, again another area of debate that I’ll leave for another day. To the East, we see a complicated series of ridges and domes that appear to intertwine with the ridges from the Kverkfjöll system. Maybe a connection existed here many millennia ago when Kverkfjöll had its day sitting on the plume head barking ash and lava at the sky. Finally, we come to the NE. A ridge extends out from NE Grímsvötn, but turns to the NNW and appears to intermingle with a ridge coming from the SW of Bárdarbunga. As Gjálp showed us, the large magnitude 5 non-double-couple earthquake in Bárdarbunga caused all manner of magma shenanigans under the area which has been implicated in kicking off the Gjálp eruption from Grímsvötn. Multiple past and future interactions between these two monsters cannot be ruled out and hopefully, we’ll see more in the near future which will further our understanding of large volcanic system interaction.

So what about Greip? Well, this area sits to the NW of this Grímsvötn-Bárdarbunga ‘handshake’ and doesn’t seem to have a pre-existing ridge associated with it from Grímsvötn nor does it seem to have any existing eruptive surface deposit visible with this data. Greip seems to have more in common with the ridges extending from the SW of Bárdarbunga than Grímsvötn. Another thing that caught my eye was a series of ridges extending to the NW from Greip that run between, and in parallel to, the 2014 dike path and Kverkfjöll. This might have been the site of previous lateral dike eruptions from Bárdarbunga and may link up to the graben that exists out in Holuhraun. From this evidence, the parental system could be implied as Bárdarbunga and a re-drawing of the fissure swarm outlines may need to be done if more evidence comes to light.

Looking at the anomaly data, the older established intrusive bodies can be clearly seen around the big central volcanoes. There is some nice correlation between these dense intrusive bodies and the bedrock topography. Again, some limited interconnection between the central volcanoes could be implied by this data at some point during their birth and subsequent evolution. But poor young Greip is sat all alone in the big blue…

Map showing the magnitude of the horizontal gradient of the Bouguer anomaly with the VC map overlay. Overlay: Beardy Gaz, Map: Gudmundsson 2007

There doesn’t appear to be any intrusive bodies underlying Greip so past intrusions must either have been minor or the intrusions are still hot and therefore less dense. Negative anomalies can indicate areas of fresh hot magma, this can be seen within the caldera complex of Grímsvötn and Bárdarbunga. Vonarskard and Kverkfjöll still have negative anomalies and may indicate that magma is still is present even in these post-caldera and less active systems. Kverkfjöll still has hydrothermal activity and the occasional phreatic explosion like that in 2013 (GVP) and this fits nicely with the implied underlying magma body. Greip sits within a large negative anomaly, but this is not to say that there is a large magma body ready to destroy us all like the media likes to portray at Yellowpebble. It could simply indicate an area of lower density rocks, like hyaloclastite, which have a lower density than gabbro and other dense intrusive rocks. The negative anomaly may also indicate this area is hotter, and therefore less dense, than the average background crust of the greater area. This area does sit above the proposed mantle plume head and this additional heat may lower the density of the rock over a large area.


We now return to the petrology of the known eruptions in the greater area surrounding Greip. The Gjálp eruption was initially pinned to Bárdarbunga due to the aforementioned magnitude 5.6 non-double-couple earthquake (more here) and large scale magma movement and then pinned to Grímsvötn from local seismic data. It was in fact found to be a completely separate system, with the bulk of the ash identified as basaltic icelandite. Icelandite is an iron-rich volcanic rock between rhyodacite and tholeiitic basalt as it contains more silica than basalt. We have the great geologist Carmichael to thank for identifying icelandite from his work at Thingmuli (nothing to do with ‘The Thing’, thankfully). However, with further microprobe analysis, it was revealed to be heterogeneous, with compositions ranging from Grimsvötn-like tholeiite to icelandite, an intermediate magma between the olivine tholeiites of Bárdarbunga and Grimsvötn. Crustal melting has been implicated in this change from theolitic basalt, common from the plume/rift volcanoes, and the resulting increase in the silica content. Gjálp truly was a mix of parental magma from Grimsvötn with a little bit of its own concoction to make it an individual in its own right.

As we have no evidence from Greip of past eruptions we have nothing to base its magma composition on. All we can do is rest uneasily on raw theory and extrapolate data from the surrounding volcanoes and intermediate eruptions between them. It’s possible that if Greip were to erupt it would have a composition closely resembling theolitic basalt from Grimsvötn and Bárdarbunga, depending on who initiated or contributed to the eruption. Crustal melting may contribute to its own individual petrology, and may be higher or lower than Gjálp due to the length of time the magma has been accumulating in this proto-reservoir and how much crust it melts on its way to the surface. Of course, if it has it’s own unique deep source of magma from the mantle it will have its own individual signature regardless of crustal melting.

The Future?

After experiencing the recent ‘slaps-in-the-face-with-a-wet-fish’ results of Brexit and Trump in regards to poll predictions and the actual outcomes, I’m hesitant to predict which way this one will go. Of course, nature is far more complex than deciding between say glacial exit, glexit, or remaining in restless slumber beneath a frozen sky.

Aerial map of North Vatnajökull showing recent earthquakes in the last 18 months. Map: Andrej Flis, Data: IMO.

The graphic above shows earthquake data from the previous 18 months. The 2014 dike can clearly be seen popping and contracting away as the magma slowly cools and the surrounding crust shifts and adjusts to the rather rude intrusion. Bárdarbunga is still noisy post-eruption with the ring fault clearly seen as well as the deeper activity to the NW. The line of quakes indicating the Hamarinn-Loki-Fögrufjöll systems can be seen as it ‘points’ towards the Gjálp eruption area to the North of Grímsvötn. Greip can be clearly seen as the cluster of deep to mid-crustal depth earthquakes all alone to the SW of Bárdarbunga. Activity appears to be increasing in this area as we’ve already had numerous earthquakes in just 18 months. A word of caution though, as I’ve stated previously, this could be purely down to increases in the sensitivity of the network.

So, what if we are seeing an increase in earthquakes, where’s all this recent activity heading? Should we all be pulling up a chair and putting the popcorn on in anticipation for an eruption? Should the Daily Fail crank up their Mislead-o-MaticTM headline generator? Hopefully, not ever on the latter point. We need to rely on actual data rather than sensationalism from the media. So let’s look at where the data is heading.

Up to date cumulative energy graph for the Greip area 2005-June 2017. Graph: Andrej Flis, Data: IMO.

Again, with the limited date range and seismic network sensitivity variables thrown into the mix, a true picture of trends becomes hard to formulate. We don’t have the seismic data between eruptions that we have for Hekla and Grímsvötn so we don’t know the rough failure point measured in cumulative energy release. There is an increase in quantity in the last few years, but the magnitude and therefore cumulative seismic energy isn’t increasing exponentially as would be seen during a run-up to an eruption. Again, we may have to wait a few more years before we can start to see any clear trends, but every volcano has to start somewhere…

And so this brings us almost up to date. Last month there was a triplet of earthquakes at Greip as seen from the IMO data:

19.06.2017 17:24:26 64.590 -17.180 17.0 km 0.9 99.0 17.4 km ESE of Bárðarbunga
19.06.2017 17:23:27 64.584 -17.179 15.1 km 1.1 99.0 17.7 km ESE of Bárðarbunga
19.06.2017 17:22:36 64.587 -17.170 18.2 km 2.2 99.0 18.0 km ESE of Bárðarbunga

At this depth, it’s likely this was caused by a small dike intrusion penetrating upwards. Another offspring in the large family of dike swarms at Greip. As magmatic pressure increases from depth, possibly from the plume head exerting increasing pressure as we head into a new Icelandic peak cycle, we could see increasing amounts of multiple earthquakes and maybe the odd swarm. As we’ve seen from the previous data, there could be a relatively stable and resistant crust acting as a ceiling, helping keep any intrusions small and at depth. If this is the case then this could potentially keep a cap on things for many years to come, sadly for the impatient volcaholic. However, we’ve no way of knowing what the critical failure point is in this system. A large intrusion at depth could destabilise the entire proto-chamber/reservoir and lead to the birth of a truly fascinating and wondrous thing. The birth of a new volcano in the land of giants.

Beardy Gaz


(Gudmundsson 2007)

(Björnsson 1988)

Friday Riddles

The Friday riddles have returned and can be found in the Friday posts for now, but we will look into opening the Sheepy Bar in due course. Once open, grab a beer (or drink of preference) and feel free to discuss anything you like in there as well as answer the riddles. We want to bring the community feel back to the blog and what better way to do that than having a drink and putting the world to rights!

If you’re unfamiliar with the volcanic riddles, simply deduce which volcano the clues hint at and leave your answers in the comments. Expect a ‘Ding!’ for every correct one :-):

  1. Doctors Cumbrian watery abode. – Mount Carlisle, answered by Daisaster
  2. Hot electric tunnel of love. – Gunung Kelud (Ampera Tunnel) answered by Bjarki
  3. Towering depressed volcano for all to see. – Eifel Volcanic Field, answered by Thomas A
  4. The man, the car, the movie. (Antropomorphic nickname) – Herbert Island, (The Movie Herbie) answered by Daisaster
  5. The pastoralist cowbell of: Ngorongoro (Ngoro is a massai cowbell) answered by KarenZ

In Other News…

As part of the increase in the community feel on the blog we want to put forward a few ideas. One of these is having an ‘open-conduit blog night’ in which anyone can submit a poem, a nugget of knowledge/experience or an essay on volcanoes and geology in general. I haven’t worked out the logistics yet as I’ve been too busy writing this article and moving house, etc., but if you have any suggestions for implementing this kind of functionality, or direct me towards some decent WordPress plugins, then please let me know. We also hope to integrate the Facebook group’s posts and comments on the blog and as part of this, you may notice design changes to accommodate the additional content, i.e. wide-screen content. Again, if you have any suggestions, please let me know.

And lastly… we are always on the lookout for writers for the blog. It doesn’t matter what your experience or knowledge level is if you can write an article on volcanoes, a particularly intriguing aspect of them, something within the greater geology context or even a topic within the broad scope of historical evidence of past events on Earth or beyond then please get in touch! In addition, any coding or graphical skills are always welcome and help us to further our understanding of these systems, disseminating the information to our readers and hopefully kick-starting a research venture to test the various hypotheses put forward.

133 thoughts on “Greip, Gjálp’s Mysterious Sister

  1. Catching up on my reading & was delighted to see an article about an area I’ve been curious about. Plus to see Unknownabunga has a name, Greip. Great article. Now on to the next article. 😊

Leave a Reply