Posted on by Jón Frímann

Inflation starts in Katla volcano

There seems to be inflation going on in Katla volcano, at least according to the automatic GPS data that IMO has on it’s web site. But this might be error in the data. But for the moment I do not think it is. But please do look at this data with the view that it might be wrong.


The inflation appears clearly on Lágu Hvolar GPS station. Copyright of this image belongs to Iceland Met Office.

The inflation can be seen on Lágu Hvolar and Sólheimaheiði GPS stations. But it appears to be more clear at HVOL GPS station. But that is most likey because that GPS station is closer to the magma pocket in question that creates this movement.


The inflation also appears on Sólheimaheiði GPS station. But not as clearly as on Lágu Hvolar GPS station. Copyright of this image belongs to Iceland Met Office.

What might be the critical inflation before a eruption starts is hard to know until it happens. For Eyjafjallajökull volcano that was 60mm, for Grímsfjall volcano the inflation was actually 350mm. But that volcano had eruptions that did not change it’s inflation numbers. I do not know why Grímsfjall volcano is so different in this respect from other volcanoes in Iceland (that I know of so far).

50 Replies to “Inflation starts in Katla volcano”

  2. This is what makes Iceland the most interesting place.
    Katla has ever so slowly crept on, on her walk to eruption, patience is a virtue with Katla.

    Grímsfjall volcano is so different. Perhaps the shear size of the glacier and the very complex interaction of that with the neighboring volcanoes, some of which we probably done even know because they are under the glacier. I may well be quite wrong, its just my way of explaining why Grímsfjall did not inflate.

  3. Good morning:) I’ve been reading up on katla and she sounds micheivious. Could someone explain why she is thought of as one of the “most dangerous” of volcanos in Iceland. Thanks

    1. Katla is the only volcano in in Iceland that is in close proximity to people (town of Vík) that erupts somewhat regularly and historically has violent eruptions with heavy ashfall and massive glacial flooding in as little as 3 hours after eruption begins.

      Other volcanoes are more remote or do not erupt as violently or as in the case of Vestmannaeyjar that does not erupt regularly on the human timeframe.

      Katla even threatens Vestmannaeyjar with an tsunami when the glacial flood hits the ocean.

      And if there is an “perfect storm” of condidtions it might even disrupt flights all over Europe for even longer time than Eyjafjallajökull though that depends much on the composition of the ash. It tends to be courser from Katla than from Eyjafjallajökull and therefore it might not travel as far.

      1. Nota bene, the ash from Eyja was corser the previous eruption than the last. One should always remember that volcanos can change over time, and that the longer time between eruptions, the more change can (probably) happen.

  4. Saturday
    03.09.2011 04:55:45 63.661 -19.353 1.7 km 1.7 90.02 5.6 km WNW of Goðabunga
    Saturday
    03.09.2011 04:55:43 63.607 -19.136 1.1 km 1.4 90.04 3.5 km NNW of Hábunga

    presume one of these is a ghost??s

  5. I don’t think you’re right Jón. HVO is the only station recording one single uplift. All other stations around Katla (GOD, AUST, SOHO, ENTA) are all recording subsidence. If there is inflation, all stations will notice some inflation, yes some more than others, but none will record subsidence.

    1. Inflation is not necessarily only uplift. It can also manifest itself as a movement away from or even towards to the volcano, or even subsidence. This is possible at certain distances from the volcano, the distance depending on the depth and shape of the magma chamber and some other factors.

      2. HVO lies about 15 km away from center of Katla caldera. The magma chamber for Katla lies approx. 2-3 km below the caldera floor, if I remember correctly (actually it does not even matter how deep it is, as long as it is below 10 km). So, when this chamber expands, HVO station moves SE (magma chamber is “near the surface” if looked from HVO). This is because the magma chamber is not below HVO but it is essentially beside HVO. HVO may also move UP, if there are sills under Katla, that extend at least 10-15 km away from the center of the volcano (not likely). The shape of the chamber does not matter, as long as it is less than 10-15 km wide.

        So, what has HVO done during the time span that can be seen on the graph (approx. 2-3 months)? It has moved 5 mm east, 5 mm south, and maybe a few mm up (statistics for direction up is not enough to say anything for sure).

        That means, HVO has behaved just as expected, if Katla’s magma chamber was inflating. I has moved SE, like I explained.

        If you want to dig deeper on the possibility of local subsidence when a volcano is inflating, I’ll direct you to the numerous papers that attempt to discover the depth of the magma chamber for those volcanoes by modeling.

        Any other questions?

      3. I have no doubt that Katla has been inflating for the last few months. (Which could partially be just due to seasonal changes)
        Yes, HVO shows the signs we could see during an inflation period. But all the other stations:
        http://hraun.vedur.is/ja/Katla/gps_solheimaheidi.html
        http://hraun.vedur.is/ja/Katla/gps_solheimaheidi.html
        http://strokkur.raunvis.hi.is/~sigrun/AUST_11.png
        http://strokkur.raunvis.hi.is/~sigrun/ENTC_11.png
        http://strokkur.raunvis.hi.is/~sigrun/SOHO_11.png
        http://strokkur.raunvis.hi.is/~sigrun/RFEL_11.png
        http://strokkur.raunvis.hi.is/~sigrun/OFEL_11.png

        They do all not show any signs inflation, infact they all show subsidence characteristics.
        Also, HVO has a natural movement towards S-E due to plate movement.
        And finally; the latest data from HVO shows that the short uplift trend has stopped.

      5. I’d not talk yet on subsidence. First, given the statistics, it has not yet significantly lowered from the “peak level”. Secondly, three point in a line does not yet mean that the direction has changed permanently. Those figures do also include periods ranging from days to weeks, when there has been subsidence for a while, but after that the trend turned up again. One could even say, Katla “breathes occasionally in, to continue the big blow again”…

      6. I agree, this is by far too little to make a decent prediction of the future trend. But it certainly is not a clue for inflation on the short term.

    2. From what I’ve read for the past years that Katla has been preparing for an eruption, you need not get an uplift at every station, you might get a reading of lower status for north in a station located on the south of the glacier and at the same time you get an uplift to east on a station to the east. From what I figure, you have an “relative movement”. Why else would scientists measure Austmannsbunga and Enta at a 8 hour interval?

      1. Yes you do. Inflation is the result of magma chamber expansion because of an injection of new magma. Of course you have relative movements, I’m familiar with that. But for the exception of Lagu Hvolar there is no station recording any signs of inflation. When a magma chamber the size of Katla inflates that has to been seen on stations located almost on top of them.

      2. Pieter, and this goes for everyone else.
        What is or is not happening today is basically totally uninteresting untill we have 10 or more datapoints to ruminate over that follow a trending line.
        Why? Satellite deviation…

        What is interesting is the statistically proven movement.
        Since July the inflation for Austmannsbunga and Enta has been 40 mm. Ie, 20mm per month.

        Those are the ones that are basically ontop of the inflation point and that is Why Sigrún monitors them so closely.
        I don’t care about the above mentioned stations because they are to far away to be the decisive factor. Aust and Enta are decisive.
        Compare the various ones on this link to Aust and Entra.
        http://notendur.hi.is/runa/eyja_gps.html

  7. What I think it is interesting is the shift southwards of many of these stations close to Katla. Since they are south of Katla, this could be related to magma pushing upwards inside and making the stations getting away from the center of the volcano. But it is peculiar since this is also observed near Hekla stations in recent days. Besides that, there are no changes in the west-east trends, and no significant changes in up/down too. I cannot find myself convinced that this is some signification inflation happening in Katla. Why do you so say so, Jon?

    1. I see wider up/down variations for example in Skrokkalda station (Hamarinn) in recent weeks and even a small trend upwards in Hekla in recent days. I think those two volcanoes have been showing signs of magma movements as good as Katla if not more.

    2. I was seeing now the different GPS stations in Iceland, and apparently all of them show that southwards movement in recent days, so I guess it is either a glitch, Reykjavík was moving northwards, or some general small south movement in the whole Iceland. But I don’t know any station westwards from Reykjavik to compare with these movements.

    3. I was a aware of this. But I was not sure why that might be. But it might be due to ash in the air or something else.

      But for HVO I found it less likely to be real then a error in the GPS data. But only time is going to tell us if that is the case or not.

      1. I have seen other times when a GPS movement happens in nearly all stations. I might be not a glitch, and even non geological. It might be other unknown environmental factors or changes in the reference station of Reykjavik.

      2. @ALL on GPS
        Not ash, as a former specialist in singal-systems I can say definitly that ash is not involved.
        I would say that is movement in the Reykjavik station towards the north.

        Interestingly enough AUST & ENTA is going north too, probably more so than is showing.
        But… And this is a large BUT(!) They are not as far as I know trended against REYK, they are self-referent stations as far as I know and have understood things. So for those who wish you can use them as “references” for GPS-detranding other than against REYK (which is a rather stupid choice to beginn with).

    4. I think the key to the answer is that the movements are relative to Reykjavik. Hence, we see also the relative movements between American and Eurasian plates. However, you’ll need much longer data sets for removing this contribution reliably from the data. I guess at least one year to avoid mixing with seasonal effects, but possibly longer if there was an eruption to mask those seasonal things.

  10. Actually, after reviwing all the stations around Katla, it looks like that inflation seems quite clear in Katla now.

    HVOL, located southeast, moved southeast (because moved away from the caldera). But the trouble here is that this is also the natural plate movement.

    AUST, located north, has moved northwards (again away from the caldera) and upwards since the flood event in July (it reversed its movement southwards), and this also correlates with all the increased EQ activity at Katla. But the trouble is that this has also happened last year, probably part of seasonal changes on the glacier. But it is a movement worth following.

    Fimmvordugals located west, moved west, which could be related to moving away of the caldera due to magma entry, but this is ever since Eyja eruption, and so can be also related to the emptying of magma there.

    More interestingly, OFEL and SNAE, located further northeast, moved northeast, when the plate moves usually southeast, so can be linked by a rising chamber in Katla.

    With this alone it looks already clear, that even with the common seasonal changes, Katla is inflating at all its stations around it. But there are more recent clues.

    Another interesting recent change: In recent days, Fimmvorduhals, HVOL and Skogar, located in south/west side of Katla, show inflation (up), OFEL, RFEL, ENTA and AUST, located in east and north side of Katla, show deflation (down). So, there is a change happening now in the volcano. It could be that magma was pushing upwards under AUST but apparently shifted southwards (which also explains that movement in stations). Or, maybe magma is mainly pushing below the south side of the caldera (where we find most earthquakes) and so AUST is now deflating while HVOL starts inflating faster. I think this gives another clue that we are getting closer to an eruption; magma is closer to surface.

    With Eyjafjallajokull we saw a trend over 2-3 months with EQ swarms and rapid inflation. With Katla I expect something similar. However, we still don’t have the scale of swarms and inflation I am waiting before an eruption, so I think we are still at least 2-3 months away from its eruption.

  13. The Mag 3.2 shows up in the running list as ML 2.95. The Mag 2.7 hasn’t shown up in the running list yet, so I can’t tell you what the ML will be.

    Both of the two largest events (mentioned above) are called out in the plots.

    All quakes in and around Katla for 2011. 4D Plot boundaries: 63.5°N to 63.8°N, 19.5°W to 18.8°W.

    View North
    http://i53.tinypic.com/2u76mom.png

    View East
    http://i56.tinypic.com/34s5yev.png

    Plan View
    http://i53.tinypic.com/5y71jm.png

    1. I am guessing here Islander, but there is apparantly tremendous preassure in the system right now, first we had the magmatic swarm of quakes at Katla, and now we have a brutal quake-swarm at Krisuvik.
      Why do I use the word Brutal? Because it is under the lake itself. And that might not be good at all aand might herald a new drainage of the lake.

      I guess the transient was caused out of overall systemic pressure going in the South Icelandic Fracture Zone (I finally found out what the correct scientific term is for what I sloppily called the Sprungur area!)
      Something is definitly happening in the SIFZ as it hits Hekla on the SISZ.

      2. Yes, a 2.6 under the lake is nothing small. How big was the biggest earthquake in Krisuvik in the past 12 months? I remember an earthquake in Reykjanes this year that was felt in Reykjavik.

      3. There has been at least one at 3+…
        But I have a deep respect of a huge volume of water disapearing mysteriously into a volcanic fissure that is massively tectonically active like Krisuvik.

        Let’s recapitulate:
        Krisuvik has had magmatic root-filling and uplift during two separate periods. The lake of Kleifarvatn has has one large and at least one smaller instance of water disapearing. The Volcano trends in under the lake.
        My totall nightmare is that the volcanic fissure opens up in under the lake and that water drops down into the magma filled fissure. That would result in an instant tephric explosive eruption close to Reykjavik.
        The same thing happened at Hengill during 3 out of the last 4 eruptions, but on a fairly small scale. Here we have the majority of the activity at the shore of lake Kleifarvatn, and now a rather powerfull swarm directly under the lake… Not good methinks.

    1. Well, it can’t “belong” to the glacier, unless it’s over 4 km deep. Last I looked they didn’t indicate that it was being that deep.

  19. All the plots in the post are Relative to REYK, so all they may be showing is that REYK is subsiding … and none of the stations near Katla are “inflating”.

Comments are closed.