Grindavik dropping into the sea

/ 11/11/2023

Mount Thorbjorn

Foreword

Like minds and all, both me and Albert set out to write an update article unbeknownst to each other. I guess that Albert has not yet fully come to grips with me returning back to “life”. But, this is a good thing for you as a reader, you get twice the fun from two different minds as a nice weekend surprise.

Albert is the top part of the article, and I will be down far into the bottom.

/Carl

Thorbjörn Update by Albert

Credit: Astrograph. Click here to see a high resolution 360 degree panorama in Google Maps of Grindavik, taken on 4th July 2021, with the first Fagra eruption in the background

O what a night.. At the last count, there were 252 earthquake of magnitude 3 or more, the largest at M5.2. That is 252 stars on the map, scattered all over the place as the automatic locator cannot cope with this amount of shaking. It is Christmas come early, with 252 stars of Grindavik. The evacuation of Grindavik adds another semblance of that first Christmas story. But so far, it is a lot of noise but no action. The magma has (so far) remained underground.

To recall, this all started with inflation over an area centred roughly on Thorbjorn, ‘child of Thor’. Thor, according to wikipedia, is a “hammer-wielding god associated with lightning, thunder, storms, sacred groves and trees, strength, the protection of humankind, hallowing, and fertility”. We get our name for Thursday from it. There is certainly a lot of thunder and hallowing going on at the moment – but the protection of humankind has been deferred to the lesser gods of the Icelandic authorities. They are rather good at it and overnight they evacuated both Grindavik and the power plant. Work to protect both against future lava flows has been started, by building embankments of several meters high. It worked for the flows two years ago, although in one case only because the lava did not actually reach the wall. The sign ‘lava – no access’ appeared to be sufficient. The risk, of course, is that a wall may deflect lava onto some other property, raising questions of liability. But not knowing where lava will surface makes planning more difficult.

The following map was uncovered by one of our commenters. It shows how far lava needs to travel to get to Grindavik from various locations. At the moment there is no ‘most likely’ location along the rift. The further north, the better, obviously.

It started mid October when inflation took off. A sill formed with an inflow of 5 m3/s. Over three weeks, that amounts to 10 million m3, or 0.01 km3. It added to previous intrusions which had been happening since 2020.

Why a sill? A sill is a horizontal magma layer, which pushes up the layers above it. This is what is causing the earthquakes and the inflation. The earthquakes happened at 5-6 km depth, suggesting the sill was just below this. How does it form? Magma in rock has to find a weakness, a crack where it can insert itself. This can happen at a place where two different layers connect, for instance where the lava pile which has build Reykjanes lies on top of the oceanic crust below. The two are already a bit separate and magma can prise them apart. This happened over an area of up to a few square kilometer: the sill would have been 10-20 meters thick, in my estimation. There is minor magma around than is in the sill: it has a feeder system below although we don’t seem to know where this feeder is.

Yesterday the sill found another weakness. To move up, it has top break rock and that is not easy in such an old, cold lava pile. So it looks for another connection which goes up. This can be a fault, but in this case it found something else: an old fissure system. The fissure sits on top of an ancient dike, and this dike means there is a vertical discontinuity between the rock and the dike. The first earthquakes along the fissure were several days ago, but yesterday it took off. The magma began to build a new dike along the side of the old one.

The magma moved up by 1-2 km but mostly moved sideways. The dikes here run along the direction of the rift, which in Iceland is SSW-NNE. Because of the spreading, it is easier for magma to insert itself in this direction. And now it went fast. As of this morning, the dike extended over 18 km, from a few kilometer off short to far in-land. That does not mean the magma has traveled all this distance. It means that the old fissure dike is breaking away from the surrounding rock.

So where is the magma? The GPS shows an indications. As this morning, Thorbjorn has sunk by 40 cm. This happened because the dike is forcing the rock apart. That the sinking is so much indicates the magma has shallowed. How shallow? My guess is 2 km but it is only a guess. Both Grindavik and Thorbjorn are moving northwest, but that is harder to interpret. Where would an eruption be? Impossible to tell. It can be anywhere along the line. The highest probability is in the original location which is east of the Blue Lagoon. There is a bit of a gap in the earthquakes and that can be a sign – or not. The second probability is indicated by the deflation – that would put the eruption in the harbour of Grindavik!

RUV put out the following (again taken from a comment)

“The signs of magma movement include significant subsidence in the Sundhnúkur craters, indicating that the magma might be shallow below the surface.

The southern end of the crater row is about 1 km from the nearest buildings in Grindavík and approximately 1500 meters from the Svartsengi power plant.

The Icelandic Meteorological Office notes a substantial change in seismic activity, moving south towards the town of Grindavík. According to the Icelandic Meteorological Office, there is a likelihood that magma movement has extended beneath the town of Grindavík.

The volume of magma involved is considerably larger than seen in the major magma intrusions related to the eruptions at Fagradalsfjall.”

Here is the onset of activity, with M2+ magnitudes twice per minute at one point.

An interferogram (thanks Gaz!)

And here is a list of cameras mentioned by our contributors

RUV

RUV 1

RUV 2

RUV 3

Other

View of power plant from Thorbjorn

A Thorbjorn view

A multiview

Live from Iceland has views all over iceland, including:

Fagradalsfjall and
Svartsengi

Albert, 11 Nov 2023

 

Grindavik update by Carl

Grindavik, photograph taken by VC FB regular Roman Zacharij and used under Wikimedia Commons.

I promised everyone an update today, so here is an update about why I and every other volcanologist should go and drink some tea and ponder why we underestimated a volcanic risk. I will get back to this below.

This will mainly just be an update that I will edit as things unfurl, and I will throw in a little prognosis of what will happen, and where it will happen, based upon the situation at 11.00 Icelandic Time on Saturday.

 

Short Recap

Roughly two weeks ago Thorbjörn started to show signs of an eruption being possible as a secondary intrusion started, and a sill formed where magma accumulated.

On Sunday evening I felt confident that the prerequisites for an eruption was there based on GPS-data, spectrography over seismic data, earthquake locationing, etcetera.

And based on that data I pointed at a possible location for the eruption to break out, if activity had continued like that I would probably not have been far off, nor would IMO either, we are after all talking about as little as 800 meters between assumed centroids from our individual models.

IMO weighted GPS-data higher than I did, I tend to weight Earthquake-locationing higher. Both methodologies are solidly based on science, and give similar results most often, well similar enough to give an 800-meter difference. Something that in the greater scheme of things is not a lot.

Earthquake map by the Icelandic Met Office. Note that there are loads of “ghost-earthquakes” that never happened, this is caused by over-saturation of the system by signal reflections.

Yesterday at around 12.00 Icelandic time, Iceland decided to kick those models in the proverbial nuts and a dyke started to form that was NNE/SSW trending, putting the most likely location of breakout at Sundhnukagigar, East of Bláa Lonid (Blue Lagoon) and the Svartsengi Powerplant.

This later led to the powerplant evacuating the workers and going into remote controlled operations after some sort of minor accident that was probably caused by strong earthquake activity.

That part of the formative dyke rapidly stalled out, and the bulk of the activity switched down the other dyke leg towards Grindavik.

This forced the evacuation of Grindavik Town itself, a decision that was proven to be a very good idea, but perhaps not for the reason it was based on.

Right now, the dyke is hammering away South of Grindavik in under the sea.

 

What We Forgot

Let us leave aside the risk for an eruption, and the earthquakes, and even the dyke, and instead look at another side-effect of the earthquakes and the propagating and dilating dyke.

Iceland is truly spectacularly good at moving about, it is extending itself in all sorts of interesting tectonic ways. Faults spread and drift apart, micro-plates dance about and rotate… it is a mess of simultaneous movement.

One of those movements have a scientific name, and that is Graben-formation. It is when due to spreading-movement a lineament of land is dropping down into the void created by the spreading land.

Some Graben like the Eldgjá-Graben form sharp sides, and if you have your house on the exact spot where that side form your house will start to lean, or even fall on its side. But if your house is in the middle of the forming graben you might end up with your house remaining perfectly straight if you are lucky.

This happens in large caldera volcanoes; I saw a house topple over in Amatítlan in Guatemala in a single night from this phenomenon. So, obviously it is well known thing among volcanologists, and something we take into our risk-calculations.

This time though it had an effect that I am pretty certain that nobody had considered, but that we will do in the future when applicable.

Let us phrase it as a question. What happens if a Graben form under a town adjacent to the ocean, and the Graben continues extensionally into said ocean?

Let us begin with numbers as of 11.00 Icelandic time, the Graben has dilated 120cm and the Graben floor has dropped a whopping 96cm (I will edit this as often as I can).

This means that conversely the ocean level has risen permanently 96cm in Grindavik. Obviously, it is not the water level that has risen, it is the town that has dropped.

Grindavik is built so that 96cm is not drowning houses, there is a margin to account for storm surges etcetera, but that margin has shrunk with said 96cm, meaning that storms are more likely to cause flooding along the shoreline and in the port.

Obviously, this is now. If the Graben continue to widen and drop it will get worse.

My favourite Graben in Iceland is Eldgjá, it is 8.5km long, 600 meters wide and 150 meters deep. Obviously the Grindavik-Graben will not become nearly that wide and deep, but a meter or two more depth is obviously not out of the question.

So, there is an idea to look at those Volcanic Hazard Maps and see if another town is at risk of Grabening itself into The Big Fishtank in the Ocean.

 

Prognosis

Station GRIC 4-hour solution. Image from Sigún Hreinsdottir’s page.

Remember that this is based on data at 11.00 on Saturday, so take into account that conditions could change with time, it is a snapshot of “now”, and based on future trends that are realistic and congruent with previous eruptions of similar nature in Iceland.

I am writing the above paragraph so that even Charles Gregory can understand, hopefully, that a prognosis changes over time and with shifting data. I do though not have great hope of that.

Eruptions are statistical games based on data, and the length of a dyke is indicative of where it will erupt. Thorbjörns dyke has two dyke-legs based on the breakout point (feeder), the shorter leg is the one running East of Bláa Lonid, and the longer leg is running out into the ocean south of Grindavik.

On a purely statistical standpoint it is more likely that a longer dyke-leg will encounter a weakness that the magma can use to pop up to the surface.

What we know so far is that the Northern leg up to now remains dry (uneruptive), but there are faults that way that may in turn end turn up to be the weakest spot along the dyke.

There are also faults and weaknesses on the longer Grindavik leg of the dilating dyke, but they are a bit more spaced out. There is though one thing that is probably going to affect things more than everything else.

It is that the ground level is constantly dropping as the Grindavik dyke-leg extends and dip its Grabening foot into the water. This means that the distance a breakout needs to travel to the surface is decreasing since the overburden become increasingly thin.

Let us assume that the top of the sheet-dyke is 2km deep at the breakout point from the sill near Bláa Lonid, as the dyke goes south in Grindavik the height of the overburden drops around the point it hits the water and it slowly get deeper until it hits the continental shelf, and it drops below the top of the dyke. But the continental shelf margin is quite some ways out.

On top of that the Graben itself is fertile ground for an eruption to pop up, and on Graben margins you often get a formation of a crater row, see Eldgjá and Lakí just as a couple of examples.

In other words, based on current data a breakout is most likely from about a kilometre north of Grindavik down to a couple of kilometres out into the ocean.

As magma rushed into the dyke, overall systemic pressure has dropped, so for the moment the pressure is probably somewhat to low for continued lengthwise progress, and to form an upwards directed conduit, but the last is depending on how firm the overburden is and how much resistance it can put up.

Eldgjá Graben, photograph by Andreas Tille, Wikimedia Commons.

The diminishing systemic pressure is why we are seeing a drop in the amount and strength of the earthquakes (11.00), it is though still quite impressive.

As magma continue to enter the system from the bottom feeder conduit from the mantle pressure will though go up, and we will get increased activity again, and probably the expected breakout. This can happen anytime from a few hours, up to a few days.

But, at this point it is pretty much a sure thing that something will pop up.

Here I would like to make a point, normally the initial breakout is 10-20 percent from the furthest point of dyke propagation, it is not clearly understood why really, it might have to do with some sort of fluid-dynamic hammer-effect, but why and how is not well understood as I said.

If that will turn out to be true here, we will see the eruption start outside of the port. It would be a tad ashy as it breaks the surface and “rooster-tails” form, but as soon as it is above the surface ashfall will rapidly decrease.

In other words, do not expect the end of the world, well perhaps the fishermen of Grindavik have a different opinion, for them it might end up as the end of the world.

I have seen people “biggying up” the upcoming eruption to proportions it just can’t take. No, it will not be even remotely near the size of Holuhraun. That being said, judging from magma-influx and previous eruptions a likely figure would be a peak average discharge rate doubling that of Fagradal eruptions.

If memory serves that was 40 cubic meters per second during the fountaining phase of Fagradal II eruption, so perhaps as high as 80 cubic meters per second during the peak hour, and then slowly falling to 10 to 20 cubic meters per second.

If we assume the length of the eruptive episode to be somewhat equal of Fagradal eruptions, we end up with an eruption at around 0.1km3, and that is not a bad figure as sizes go. In other words, it will most likely be a fairly impressive tourist eruption, but nothing more.

If the influx is steady and continue, we could end up with a longer eruption and the amount of magma would slowly climb upwards. At the extreme end of probability, we get a small shield formation that lasts for up to a century, but that is a very low probability indeed.

As a final thing I would mention that I seriously hope that Grindavik will be affected as little as possible, and that the lives of the inhabitants soon can return to normal.

CARL REHNBERG

P.S. Charles Gregory, “Told you so”. It is now the third time you have been rude and wrong and gotten “I told you so’d”. I suggest pulling in your attitude and stop trying to bully people. I am definitely unbullyable if nothing else.

Related posts

1,132 thoughts on “Grindavik dropping into the sea”

  1. Thanks for the update!

    RUV cam links are broken by the way.

  2. Thank you very much for summarizing the events since yesterday. The pure numbers alone are very impressive… not only the number of earthquakes, but especially how enormous the GPS displacements are during the rather short time period. Thor doesn’t seem to have a hammer, there is a full sized Caterpillar equipment at work. 😉

  3. Can someone please explain the subsidence taking place? It seems counter intuitive?
    Posted in the previous article, just as the new one was going up. So reposted it here.
    Thanks.

  5. The quakes are a lot smaller and shallower now. The majority of the hard work has been done.

  6. Thanks for the update!
    When I looked at the IMO GPS charts, last evening there must even been a quite sudden movement.

    Fagradalsfjell moving south east and Svartseni moving mainly west resulted in a ‘gap’ of about 120 cm!

    • Festarfjell (not Fagradal, sorry typo) did have an uplift after snapping.

      • Can you supply a link to the above graphs. I seem to have lost my link.

  7. As Tomas mentioned the spot of maximum deformation tends to be where the dike is most likely to erupt, this is the maximum dike opening. I overlayed the fringes from the interferogram shared by Beardy Gaz. Green are fringes away from satellite, orange towards. Blue circles are earthquakes from the dike intrusion (mostly along the base of the magma intrusion). The red lines represent the maximum deformation and the Sundhnukur fissure row. More or less where the two cut is the most likely place for an eruption. 63°52’21.79″N, 22°23’56.22″W. If so, it can reach the Svartsengi Power Plant quite easily, and if big enough could also reach the east side of Grindavik:

    • The maximum deformation of the butterfly’s wings are probably a bit off due to the look angle of the satellite, but the place where the two red lines intersect I think should be accurate

      • There’s also a very obvious offset across the Reykjanes fault that can be seen all the way out to the tip of the peninsula. I assume that would also offset the wings. I’d say within a 200m radius of your mark will be the spot to watch.

    • very interesting, could well be the spot and you will win the sweepstake!
      Interesting if it actually happens offshore, presume not so much info about inflation / deflation available there

    • Hector, I am concerned about the possible gas content of this magma. I think it is probably going to be more than we anticipate. If the fissure eruption comes up where you have indicated, it might steam roller Grindavik. It was very interesting that last June 5, 2022 Just Icelandic made a video on earthquake clusters happening in this very exact area, indicating this as an area to be watched.

  8. RUV English live reporting here https://www.ruv.is/english/2023-11-10-magma-tunnel-lies-under-grindavik-396249

    Some photos on an empty Grindavik posted and one very interesting one showing a huge crack across a road and house. So this confirms the problem isn’t just due to eruption risks, but the damage caused by the quakes and land movement will be a problem too, especially if the town has no power or water.

    Reporting also confirms the power has gone out in the town. And with crack that size, that is not a surprise!

    • Didn’t they just update it to 800 meters underground now? Maybe we’ll see a phreatic eruption in the Grindavik harbor soon.

  10. IMO Update: 12:00

    Analysis of the earthquakes from today and yesterday is ongoing. The goal of this work is to better understand the evolution of the magma intrusion. Presently, the data indicates that the magma intrusion extends from Stóra-Skógsfell in the north to Grindavík in the south, where it extends beneath the sea. In accordance with the latest preliminary models, using the most recent satellite data acquired last night, the shallowest depth of the top of the magma intrusion north of Grindavík is 1.5 km. Joint interpretation of the ground and satellite measurements indicate that the size of the magma intrusion and the rate at which it is moving are several times larger than have been measured previously on the Reykjanes Peninsula. Our assessment is that an eruption, if it were to occur, will originate from the northern side of the magma intrusion. This means that there is a greater likelihood of an eruption beginning close to Sundhnjúkagígur.

    Scientists are meeting regularly to interpret the data and update the latest models and hazard assessments. A meeting for journalists will be held at 12:00 at the Department of Civil Protection and Emergency Management. The current conditions and future scenarios will be discussed.

    The likelihood of a volcanic eruption occurring in the near future is deemed considerable.

  12. My (uneducated) guess is that the fissure will open up somewhere between Þorbjörn and Lagafell. Or to the east of Þorbjörn. Assuming that one does at all.

  13. Obviously the coy magma will wait until the visibility range is down to a minimum.

    • Probably during a deep fog bank and heavy rain…
      And it will a storm so that the webcams shake until we suffer from a collective seisure. 🙂

      • Would be percieved as bad luck, but we know better, haha (<- crazy laugh).

        • We know Iceland!
          The only place where you are sitting divining what is happening on a violently shaking webcam and a sheep flies past and drop over a cliff, to later be found and eaten by Bear Grylls, who then continue with vomiting profusely.

          Poor sheep is the reason we have a sheep as a patron saint to this day…

    • Ah, it was all those foggy days trying to observe Fragradalsfall that did you in? 😉

  14. https://www.mbl.is/frettir/innlent/2023/11/11/naer_staerdargradu_holuhrauns_heldur_en_geldingadal/?utm_medium=Social&utm_campaign=mbl.is&utm_source=Facebook&fbclid=IwAR2kKc6ALaFSQTcHialai2lQLkZx0sKm7RGMLNzHuKFi8TuoRAIbm3yDtgc_aem_AaiDra_ylubuXAVovlB4kXx3zguXvL2ObjhQrH3hPtgHwZfDhtG8f_hgmUYIA6lBek4#Echobox=1699703044

    They think it will be a large eruption there is lots of magma avaible, this really shows how monsterious Icelandic volcanoes can be when they are activated. Iceland on avarge tends to do the largest one go eruptions togther with Hawaii, for comparsion on etna this woud be a large flank eruption If it erupted.. but here in Iceland its avarge with 0,1 km3 per eruption.

    they think it coud be as large as Holuhraun intrusion or something thats a major part of its size, that woud be very bad near Grindavik, but eruptions from Svartsengi/ Reykjanes tends to be much smaller in volume and even faster, so I dont think he is correct, but they say lots of magma is avaible as its a large dyke, MAR dykes can be immense and are long as its deep tectonic fissures that haves to be filled, the magma dyke is now 1,5 km under the surface the geology experts say.

    • The terminology that scientist is using is that “the eruptive volume will be closer to that of Holuhraun than Fagradalsfjall”.
      So no, not Holuhraun sized. I blame the way the article was written, a little too excitable for what he said

  15. I updated the article by adding an entire article to the end of it that I had written for today.
    Twice the Saturday Fun, Two articles to the price of one!

    • I’ve trouble picturing the sill and the dyke (two legged) in 3D. So there is a “source dyke” coming up, feeding a huge flat sill, a plaice in fish terms, and from there a dyke runs upwards, forks and ends up NE and SW of Grindavik?

      • There is a bottom feeder conduit going from the mantle that formed during the previous intrusion, it then formed a sill-layer more towards Thorbjörn, the new intrusion formed a second sill near Bláa Lonid.
        From that sill an earthquake stack formed that slowly moved upwards indicating a formative conduit to the surface.
        Then the new sill broke on the eastern side of it and NNE-SSW sheet-dyke formed, that in turn killed off the previous formative conduit at least for now.

        The question now is where this dyke-sheet will break out.

        I hope I was a bit clearer now?
        If not ask more questions and I will try to clear things up further.
        This is a moment where I wish I was good at drawing things, but alas, I draw like a 4-year old with MS-Paint… 🙁

        • Didn’t get it. What about a hand drawing + photo of it? No MS-pain(t) involved…

          • Seriously, that would be even worse…
            The other admins have voted and banned me from doodling.
            But, if one of the more drawsome admins see this they may be able to help out with a little image.

          • @Carl: I found “Storage and Transport of Magma in the Layered Crust — Formation of Sills and Related Flat-Lying Intrusions” (and it’s also for free! Bargin day.)

        • Carl, one of the Iceland volcanologists said that he’s not sure of the feeder to this sill/dike, it indicated it still might be Fragradalsfall ? Do you think his idea is rather dubious?

  16. Thanks Albert and Carl!
    Besides the intense earthquake swarm since yesterday, there also seems to be an intense comment swarm here 😀
    Keep up the much appreciated work you and many others do. I never posted until a few days ago but I’m reading the blog for many years.

  17. Am I thinking about graben formation correctly? I’m picturing the lava as a long skinny balloon that is expanding in length with lava (a dyke?). As the balloon expands, the crust above it stretches. At some point the stretching becomes too great and strips of land (parallel to the dyke below) collapse thus forming the graben?

    • Yes, that is what happens, except what drives the graben formation is the extension from the tectonic plates pulling apart from each other which the dike intrusion then concentrates into a narrow band as it literally cuts the plate boundary open. Here is an experiment that simulates a dike intrusion and what it happens to the ground above:

      https://www.volcanocafe.org/wp-content/uploads/2020/01/word-image.jpeg

      From:

      https://www.usgs.gov/observatories/hawaiian-volcano-observatory/ground-fractures-and-subsidence-hazards-island-hawaii

      • Thanks, that is helpful, as I was trying to reconcile the tectonic vs. magmatic causes. The tectonic forces pulls the crust apart, allowing lava to rise in that area (forming a dyke) that further pulls the crust apart in a more localized area?

        • A dike is like a sharp knife that cuts open the plate boundary and fills it up with magma. The space opened comes from the tectonic plates spreading. However, what happens to the rock above and below the dike where the knife doesn’t reach? It gets stretched out. Above the dike, close to the surface, the ground sinks to fill the space that is being created, hence the graben. Below the dike there’s rock all around so this cannot happen, what you get instead are a lot of earthquakes, usually strike-slip earthquakes, small rearrangements of the rock to cope with the spreading, and that is why the base of dikes have very intense seismic activity.

  18. Thank you for the double-bill of articles. Very helpful. I am wishing the inhabitants of Grindavik well, and that they have a town to return to when this is over. The graben pictures are the ones that grab me. Wow!

  19. Latest data being urgently reviewed
    Scientists are working hard to review the latest data, to shed more light on what is happening in and around Grindavík. Later today, the results of studies of new [land] deformation data are expected to be reported.

    This morning it was stated that the deformation happened much faster than previously seen in Iceland. The size of the magma tunnel and the speed of the magma flow seemed to be many times what has been observed during the activity of recent years.

    • Hmm, in hindsight, I guess watch the Torbjorn inflation a few years should have keyed us that something big is one the way, but it’s easy to forget once nothing is happening for several months.

  20. Thank you for the insightful articles.

    A new article on Visir based on comments at the public safety meeting mentioned that peak magma inflow was 1000 cubic metres per second. (Can’t find the specific article link as my Icelandic is a tad rusty.) I can’t find out how long that was for, maybe someone can do a bit of detective work.

    Quite a difference between that and the official figures of around 5 cubic metres per second just a few days ago. It’s easy to see why they’re saying it’s going to be several multiples larger than Fagradalsfjall versions one, two and three. As mentioned in the article above, Holuhraun 2.0 does seem unrealistic unless those crazy magma inflow rates were sustained over several days.

    • I do not believe that number of 1000cm/s one bit, I have seen several insane numbers without any foundation in reality lately.

      But if true, I think that the high flow rate number would be for just as the dyke opened, and the pressure drop caused instand foaming and nucleation, and I imagine that would have dropped down to something far more normal in a matter of minutes.

      But, I have not seen that high a number anywhere, and I would want to see the data behind it before believing in it.

      • Any idea where that number could come from? As pure inflow, 1000 m3/s it would imply an inflation rate over 1 square kilometer of 1 meter per hour or so. Seems a tad high. Perhaps they meant something else than inflow?

        • If the dyke is 12km long, 2km high, and 1m wide and the bulk of it formed in 6-7h, that makes an average of around 1000m3/s. We don’t really know the true dimensions, and it’s probably not box shaped like that, but ballpark numbers…

          Do we have any estimates of the actual dimensions?

          • The growth rate of a dike can be very high as it quickly draws magma from a reservoir, if lava erupts early in the formation of a dike, effusion rates of 1000 m3/s can be easily reached (some Mauna Loa initial outbreaks are around that much). Sundhnukur clearly has a large reservoir backing it up to mobilize this much magma this quickly, I have already talked about this in some of the Reykjanes articles here, fissures in the area have long curtains of fire, and lavas bear a flood-like morphology, there is some reservoir of considerable size. Luckily the magma has not broken out yet, so when it does if it does it will probably be more subdued than usual. The last Krafla Fires dikes, when the rift was almost filled up, made it to the surface in just an hour. When that happens you get a Mauna Loa-like fissure eruption.

          • Yes, they may have meant the dike formation rather than the sill. The number was compared to the 5m3/s sill formation. That would make the dike more voluminous than the original sill, though

          • Looking at the GPS information, it does look like the entire sill may have emptied into the dike. Grindavik has stopped going down this afternoon, by the way

        • I don’t think they are referring to the sill, rather the dike. To form the dike at the rate it did yesterday would require something like 1000 m^3/s with how fast it expanded. Like this dike at this point is about 2X longer than Fagradalsfjall’s was and quite possibly wider judging from the amount of stretch/subsistence. Though likely less vertical height. That is a lot of magma, maybe not in *absolute* terms, but certainly in *rate* of feed terms.

          • Like Fagradalsfjall dike is estimated to have been about 0.03km^3 in volume, formed over 3 weeks. This is probably at least as large in volume and less than 24 hours old.

        • I guess once the dike broke through the whole sill got essentially squeezed out like dropping a cinder block on a tube of toothpaste?

  21. The bulk of small earthquakes is gradually decreasing in magnitude and is now below M2 (with occasional larger quakes still happening). When that has happened during previous intrusions it has meant one of two things will happen next: Either the direction of expansion will switch from south to north (or vice versa), or the eruption is not far away. Well a third option is possible and that is that it stalls completely without an eruption, but that does not seem likely at this point.

    Carl says above that it is because the systemic pressure is dropping, but how about adding the dimensions of the intrusion to the equation as well? As the dyke gets longer and more shallow (i e taller), the increase in the dimensions as a function of a constant inflow rate will decrease. If all three dimensions grow, then each dimension will grow proportional to the cube root of the total volume. Each cubic meter that flows into the intrusion will have a gradually smaller effect on the expansion.

    Taking the reasoning one step further. If we think of the magma as a hydraulic fluid, pressing on the walls of the dyke, like the piston in a hydraulic cylinder. Then as the dimensions grow larger, the same amount of fluid will cause a smaller movement, but is at the same time capable of moving a larger mass. Is this why in 2 of 3 eruptions so far, we have seen M5+ quakes (and aM4.5 in the 3:rd) near the very end of the intrusion, when everything else has seemed to slow down?

    Just a few thoughts, not saying they are correct.

    • If I can make an opposing suggestion (speculation – not knowing whether it is correct), at the end of the intrusion the magma begins to break through to above and that creates more space. The pressure on the magma comes from the sides of the dike, so magma get squeezed into the top. That makes the bottom of the dike collapse – hence the M5 earthquake. Just room for thought.

      • Yes, maybe that makes more sense. With my suggestion the M5 could relieve the pressure and take away a bit of the driving force for an eruption…

        • I thought that collapses might be an easier way to generate an M5 than expansion. Based on Holuhraun..

    • Size increase drops pressure, size stops increasing, pressure starts to build up… rinse/repeat.

    • Ok then, I’m going outside for a couple of hours to do some chores. You’re welcome 😉

    • You can watch Grindavik here. Hoping for completely uneventful viewing.

        • it took me a while to realize that Google Maps silently ignored my high resolution 360° panorama as the resolution was TOO HIGH (23414 x 11707 pixel(, I downsized it to a more digestible size, and it is now available in Google Maps to pan around and have a look: https://maps.app.goo.gl/7ws7PPA2jZan5VYw6

          The steam of the Svartsengi Powerplant is visible behind Thorbjorn and as the image was taken on July 4th 2021 the original Fagradalsfjall eruption is visible in the east.

        • Electricity bills through the roof. I don’t leave any lights on when I’m not in.

          • Iceland is all renewable electricity. Presumably they leave the lights on to drive the solar panels.

  24. Could someone explain me why earthquakes have to be manually checked? I feel like some pretty simple machine learning should be able to locate precisely earthquakes, but maybe I’m missing something?

    • Mike Ross could explain this better.. To get a location, you need to measure arrival time at several stations. To get depth, you need to get the arrival time of the reflection of the earthquake of the bottom of the crust. It is normally no problem, but with so many earthquakes it is hard to disentangle which events go together. The automated software is not able to deal with this and puts the earthquakes all over the place. A person can do better. At the moment it seems they focus on events above 2.5, and even then I am not sure that even all ’99’ events are correctly placed. They may be, but I would not be surprised if mistakes slip in when working under such stress.

      • It’s pretty simple actually. How it works depends on the software and the settings, but generally, there’s an ‘autopicker’ program running that just looks for a significant departure from background ‘noise’ on each seismic instrument. When it finds one, it assume it MAY be an earthquake signal, and picks it – marks it with a ‘P’ to signify a P-wave arrival.

        Then there’s an ‘autolocator’ program running; it scrutinises the ‘picks’ and, if it can fine ten or more (a typical setting) correlated in space and time, it uses crustal velocity models to attempt to associate them into an earthquake and determine the least-worst location for it.

        Usually this works fairly well, although some locations can be wrong due to random seismic noise being wrongly ‘picked’ as an earthquake signal. But with a lot of events close in time, as in Iceland, it can be very wrong – you can get picks made on different earthquakes being wrongly associated, picks made on the S wave as well as the P wave etc, it can be a real mess to untangle. Far beyond the capability of any machine learning at this stage; it needs a skilled seismic operator, like me! 🙂

  25. Watching this seismic event unfold as it happened was both worrying and exciting at the same time. Worried for the local residents who had to leave their homes and excited because I have never seen that many green stars on the Icelandic Met Office earthquake reports. Learning that Grindavik has dropped by almost 1 meter is a reminder to us all that Iceland is a dynamic place.

    • You need to refresh the site from time to time. There is still a constant stream of quakes.

  27. Not sure where else on here to ask but I can’t log on. The message says that there’s no such email address. Did you have a member purge that I don’t know about?

  28. it took me a while to realize that Google Maps silently ignored my high resolution 360° panorama as the resolution was TOO HIGH (23414 x 11707 pixel(, I downsized it to a more digestible size, and it is now available in Google Maps to pan around and have a look: https://maps.app.goo.gl/7ws7PPA2jZan5VYw6

    The steam of the Svartsengi Powerplant is visible behind Thorbjorn and as the image was taken on July 4th 2021 the original Fagradalsfjall eruption is visible in the east.

  29. Sorry, you can ignore that last message. I worked it out, thanks.

  30. Carl, If I read the latest GPS data correctly, Grindavik’s recent subsidence has simply returned Grindavik to a similar elevation as in May…i.e. the 96cm drop merely erased a similar amount of uplift seen in the last couple of weeks. Hence, I don’t think the town is appreciably lower nor sea level higher than historical? Please correct me if I’m interpreting something wrong, or if you have better data at your disposal. Thanks.

    “Let us begin with numbers as of 11.00 Icelandic time, the Graben has dilated 120cm and the Graben floor has dropped a whopping 96cm (I will edit this as often as I can).

    This means that conversely the ocean level has risen permanently 96cm in Grindavik. Obviously, it is not the water level that has risen, it is the town that has dropped.

    • I think you are reading the charts wrong

      Note that the year plot hasn’t updated with the huge drop since yesterday.

    • Note that I was using a different link that show 960mm of drop, whereas the usual that Squonk referenced below shows an uplift of 90mm…
      The one I referenced is more up to date.
      There has not been an uplif of 96cm, in total it was 15cm over both inflationary periods.

      Edit/
      I missed that Squonk showed both of them 🙂

      ?fbclid=IwAR0jyjZ86u5acDFRDX8CbD3X4BzFJHcbEiCeDOsEDcwkWLLYYkIxN8PSaWg

  31. hmm, sometimes a reply goes directly into the dungeon it seems, I tried to post it again, as it was not visible, but then I got the notice, that I already posted that reply…

    can this be caused by the reply containing a google maps address?

    • trying again: it took me a while to realize that Google Maps silently ignored my high resolution 360° panorama as the resolution was TOO HIGH (23414 x 11707 pixel(, I downsized it to a more digestible size, and it is now available in Google Maps to pan around and have a look: https://maps.app.goo.gl/7ws7PPA2jZan5VYw6

      The steam of the Svartsengi Powerplant is visible behind Thorbjorn and as the image was taken on July 4th 2021 the original Fagradalsfjall eruption is visible in the east.

    • it took me a while to realize that Google Maps silently ignored my high resolution 360° panorama as the resolution was too high (23414 x 11707 pixel(, I downsized it to a more digestible size, and it is now available in Google Maps to pan around and have a look: https://maps.app.goo.gl/7ws7PPA2jZan5VYw6

      The steam of the Svartsengi Powerplant is visible behind Thorbjorn and as the image was taken on July 4th 2021 the original Fagradalsfjall eruption is visible in the east.

    • I checked. Our spam queue (which we never look at it because it is always full of spam) is full with your comments. This is much worse than the dungeon. This is the oubliette of VC. Somehow the system took objection to either the links or the number of comments within a sort time period, or because it knows something about you which it is not telling. Pick your choice.. I can release comments if you wish, or we can just be happy that you escaped!

      • hmm, at first I thought something went wrong with the sending of the reply, then I got the duplicate message alert, then I tried to change the text, to see what was the offending part… if you could release just the first of my messages would be great, if this is not possible, the lower resolution pano needs to suffice…

        • Ok, done. But it kept the original submission time so is way back in the comment stream. You may want to resubmit and see whether it goes through this time

  32. To Carl’s part:
    “If that will turn out to be true here, we will see the eruption start outside of the port. It would be a tad ashy as it breaks the surface and “rooster-tails” form, but as soon as it is above the surface ashfall will rapidly decrease.”

    When saying outside of the port you mean the ocean, don’t you?
    So, some kind of Surtsey-like event?

    • Outside of the port is near the port from the Ocean side, typically the area within a nautical mile from port is referenced as such.

      It may that other countries or navies define it differently, I guess Lurking would be able to tell that.

  33. it took me a while to realize that Google Maps silently ignored my high resolution 360 degree panorama as the resolution was TOO HIGH (23414 x 11707 pixel(, I downsized it to a more digestible size, and it is now available in Google Maps to pan around and have a look:

    https://maps.app.goo.gl/7ws7PPA2jZan5VYw6

    The steam of the Svartsengi Powerplant is visible behind Thorbjorn and as the image was taken on July 4th 2021 the original Fagradalsfjall eruption is visible in the east.

  34. There seems to be a wide spread of M3+ quakes all over SW iceland, including a bunch near Katla/ Eyjafjallajökull.. are these incorrectly geolocated quakes, or real for their location. If real any guesses as to what is going on in geological terms over such a wide area?

    • someone mentioned that those were the automated system trying to make sense of what was happening and planting green stars everywhere, I think if they are manually checked then many would disappear – they just have other things on their hands at the mo

      • Yes, it’s not easy to see but the manually checked quakes have a black outline (star or circle) while the automated/unchecked ones have a white outline. With the amount of quakes occuring at the mo, the location of any that have a white outline should be taken with a large pinch of salt.

    • So, based on the various measurements they’ve given the dike is now 15km long, close to 3 km tall and up to 3 meters wide. Pretty big dike! Though seems that might have temporarily exhausted the available volume now that the sill is deceased.

      • We (I, at least) do not know where the feeder for the sill is/was. Perhaps that connects now directly to the dike.

    • So, based on the various measurements they’ve given the dike is now 15km long, close to 3 km tall and up to 3 meters wide. Pretty big dike! Though seems that might have temporarily exhausted the available volume now that the sill is deceased. We are just lucky it didn’t just cut straight through to the surface at 1000m^3/s while Grindavik was still in the process of evacuation.

  36. it took me a while to realize that Google Maps silently ignored my high resolution 360 degree panorama as the resolution was TOO HIGH (23414 x 11707 pixel(, I downsized it to a more digestible size, and it is now available in Google Maps to pan around and have a look:

    https://maps.app.goo.gl/7ws7PPA2jZan5VYw6

    The steam of the Svartsengi Powerplant is visible behind Thorbjorn and as the image was taken on July 4th 2021 the original Fagradalsfjall eruption is visible in the east.

  37. New IMO statement (Chat GPT translation):
    “The likelihood of an eruption has increased. It could start at any time in the coming days. Models show a 15 km long magma intrusion just northwest of Grindavík. Due to significant tension in the area, it’s not guaranteed that the eruption tremor will be detected before the eruption begins.

    As of 18:00, a meeting between scientists from the Icelandic Meteorological Office, the University of Iceland, and Civil Protection concluded the interpretation of the latest data received earlier today. Models were based on satellite imagery and GPS measurements. They indicate a significant extent of magma intrusion nearing the surface. It spans from Kálffellsheiði in the north, directly west of Grindavík and extends into the sea southwestward. The magma intrusion is approximately 15 km long, and the magma is closest to the surface at around 800 meters depth. These models rely on satellite data around 12 hours old, so the magma might have moved even closer to the surface since then. Hence, there’s a substantial probability that the magma may breach the surface. Additionally, there’s an increased chance that the magma might surface on the seabed.

    Seismic activity has decreased significantly after midday today. It’s believed that the main reason for this is the significant tension in the area due to yesterday’s earthquakes and the strain caused by the magma intrusion. Due to this tension, it’s probable that the magma will swiftly make its way to the surface, making it unlikely that the eruption tremor will be detected before the eruption starts.

    Based on the interpretation of the latest data, the Meteorological Office has advised Civil Protection that the likelihood of an eruption has increased since this morning, and an eruption could start at any time in the next few days. Models also suggest that the magma might come to the surface at the southern end of the magma intrusion just outside Grindavík. The probability of an underwater volcanic eruption has increased, so there needs to be preparedness for the possibility of an explosive eruption. An exclusion zone has been defined based on the location of the magma intrusion.”

    • By the way, Kálffellsheiði is more or less the north-east tip of yesterday’s earthquake swarm.

    • Let’s just pause for a moment and appreciate the quality of that translation. ChatGPT really does a great job with Icelandic.

    • I noticed that they’re careful to not indicate where the 800 meter depth actually was along the conjectured fissure line. (unless I missed something in the article)

  40. Is it still possible for this episode to stop without an eruption? Or is magma at 800 meters past a point of no return due to gasses now expanding out of the magma etc.?

    • I think it’s past the point of no return, I’m afraid.

    • The data is 12 hours old and there’s still plenty of activity. So I think there’s no turning back and it’s even higher up now.

    • It is almost assured to be beyond the point of no return, I would even argue that it passed that point prior to even the dyke outbreak.

    • And to complicate things, I’ve been seeing a few of versions of the name:
      Sundhnjúkagígar
      Sundhnjúksgígar
      Sundhnjúkaröð
      Sundhnjúksgígaröð

      Essentially it’s Sund-Hnjúk(a/s)-gíga-röð, a rough translation would be Bay-peak(s)-crater-row

      • Hildur, I love your lessons on Iceland language and words. Google seems to be so ignorant still 😉

      • I remember way back in 2021 Valur from the Reykjavik Grapevine gave some lessons in pronunciation for some of the volcanoes, and Sundhnjúkagígaröð was one.

        To a non Icelander it literally sounds like ‘Soond-Hnyooker-Gigar-Hrroydth’ although without accents it is very easy to read what I wrote completely wrong too. It is way easier to just listen to how it is said 🙂

        Little did they, or anyone else, know how much it would matter just a few years later…

  42. Question for Carl:

    The sudden northward and westward movement at GPS station GRIC is quite significant. What could account for this? Naturally one would assume some type of inflation coming from the opposite direction of the movement? I would naturally assume that since the dike seems to be on the east side of this GPS station, that the extension of the dike walls has caused this change? Am I right? I am trying to reconcile the downward drop with the significant NW movement.

    • This can be due to movement towards the graben, or it can be movement due to the emptying of the sill, or both. It is hard to interpret.

      • I have tried to understand the horizontal movement inside the graben for a dyke source, and in all models I have found, there is very little horizontal movement inside the graben compared with outside. Now, models don’t always agree with reality, but a fair assumption is that deflation of the sill is a contributing factor in this case.

        • I agree, it is a messy signal and you need to compare all of the nearby GPS stations to grasp what is what and start to subtract interference from one factor on another factor.

          • Add to that some general slippage along the entire Reykjanes fault, and throw in a couple of large N-S striking bookshelf faults, and you have a nice deformation salad. Serve cold with a hot lava dressing. This sequence of events is really something else.

    • I have added this to the post as well! It is near the top of the post

  44. Worst case scenario distinct possibility
    Some of the worst case scenarios could be coming true, according to geophysicist Ármann Höskuldsson, talking to RÚV earlier.

    The eruption could be worse than that in Vestmannaeyjar, fifty years ago, if the magma erupts in or near Grindavík.

    • If we’re allowed to speculate, here’s my two concerns: very fluid lava gushing out, going over the top of the retaining walls near the power plant with relative ease, and very gaseous magma, leading to a massive release of S02 gas, not seen in Iceland for sometime, I guess the Holohraun SO2 release would qualify. I am particularly concerned about the SO2.

  45. Great articles as always, this place continues to be my go-to source on the internet as far as discussion and updates on volcanoes etc.

    Gauging station V263 south of Mýrdalsjökull showed a little rise in conductivity today, don’t think it is anything significant. Maybe just a coincidence..

    https://imgur.com/UttCUVM

    • I just remembered the title picture of my first volcano book (a white car, burning house and a fire spitting black mountain). Heimaey …

  46. So are we expecting something like this again or will it just appear on the floor along a line silent as an assassin?

    • It will depend on many factors; fissure length, effusion rate, gas content, magma intrusion size, and re-supply rate from MOHO (if relevant here)… We won’t know these factors until it erupts, evolves and samples are taken.

  48. As a reminder, if you comment for the first time, you will probably find that your comment does not instantly appear. The reason is that our demon may (will) put your comment in the dungeon, for approval by one of the dragons. That may take anywhere between minutes and hours (dragons also need to sleep!) Don’t despair and be assured that after approval, your comments will appear instantly. It is something everyone has to go through. Sadly is it necessary as we also receive a lot of spam and we prefer to keep the comments advert-free! So do take the plunge!

    • I’ve just released a couple of trapped souls from the spamgeon. 🙂

Comments are closed.