Minor harmonic tremor in Katla volcano

There seems to have been minor harmonic tremor in Katla volcano this evening. This harmonic tremor seems to have started after a ML0.7 earthquake with the depth of ~3.7 km did happen inside the Katla volcano caldera. Last harmonic tremor spike did take place on the 24th of August according to the tremor plot on Icelandic Met Office web page.

Harmonic tremor mixed in with the earthquakes from Katla volcano on this tremor plot. Copyright of this image belongs to Icelandic Met Office.

Here is the harmonic tremor that I see on the tremor plot. It is the second spike. But the first spike is actually a earthquakes that did happen earlier today. Copyright of this image belongs to Icelandic Met Office.

This harmonic tremor can also be seen on this SIL station clearly. Copyright of this image belongs to Icelandic Met Office.

On other SIL stations around Katla volcano the signal is not so clear. But this is really a small event and unlikely a pre-eruption event. But it is hard to know it for sure, as Katla volcano has done this before and there was no eruption following this type of events. All that can be done now is to wait and see what happens next. It is also important to notice that Icelandic Met Office has not confirmed this harmonic tremor spike yet.

67 Replies to “Minor harmonic tremor in Katla volcano”

      1. Week 34 Confirmation of a harmonic tremor spike on the 28th between 11-13 o’clock, 27 quakes at more or less 5 km depth, biggest on a 2,1R at 11:41
        Most quakes within the caldera that week were at craters no 5 and 16.

        Á sunnudaginn 28. ágúst milli kl. 11 og 13 mældist jarðskjálftahrina með upptök á litlum bletti suðaustan við fyrrgreinda skjálfta og vestan við Kötluöskjuna. Alls mældust 27 jarðskjálftar í hrinunni og voru flestir á um 5 km dýpi. Stærsti skjálftinn í þessari hrinu var 2,1 að stærð kl. 11:41.

  1. @Carl le Strange (from the previous thread) says:
    August 31, 2011 at 22:05

    “And while I am at it, I finally found proof that Magma from Bardarbunga/Veidivötn, Grimsfjöll/Laki and Katla/Elgja are not related magmatically. Page 107” (and a a link to the Volcanic Degassing paper(s).

    I dug around a bit and found the Eyj ash analysis from last year. Trying to reconstruct the plot, I grew frustrated and then threw up my hands at the uncertainty of the Mg# calculation. I can’t tell if they are using MgO, or actual Mg in the formula. Rather than waste my time on what could be a misleading plot, I’ll just give you the data that I have. Remember, the Mg# number may not be accurate.

    The rest of the numbers are as they appear in the associated report. The previous report from last year is likely still on their server, but it is not quickly available. If you need it, I still have it here.

    Anyway… for your amusement (and to refer to the document you linked, specifically pg 107).


    1. Okay guys… are you up for the surprise of an eruptive lifetime?
      The magmas at Fimmvörduhals and Eyjafjallajökull are not from the same volcano. They are not in any way related it seems.

      Just to step it backwards.
      The lava from Fimmvörduhals is complex and heavy due to being metallrich. Silicity is lower than in the Eyjafjallajökull lava.
      But what we should concentrate on here are 4 metals. Magnesium does not say that much actually since it is a lightweight metal. I will though return to magnesium and silicity…
      The metals I am looking at are in order Iron, Niobium, Vanandium and Cuprium.
      All percentages below is how (ruffly) much more the FIMM lava contains compared to EYJA lava.
      25% more iron
      100% more cuprium
      240% more vanandium
      400% more niobium

      The clue here is the heavier elements. Due to gravity, heavier elements is believed to have sunken downwards and will thusly be progressively more abundant the closer to the core one comes, untill you in the end have a molten heavy ball mainly constructed of iron and other heavier metals. This is as you know the standard modell. Nothing fancy here.
      This leads to volcanos with deeper feeder tubes having heavier magmas. Since the magma at FIMM is heavier must it then really be from a different and deeper source?
      Not really. It is supposed (Klemetti among others) that the heavier elements will, if magma is left standing in a mature magma-reservoir, start to deposit in lower areas and in nooks and crannies as the magma start to become semi-solid (rhyolitic mush). In that way it is suposed to be able to change it’s chemical composition over time due to gravitational fall-out and chrystalization during cooling.
      So the answer is that it can still be the same feeder, just that the EYJA lava is older and has undergone the above mentioned process. But is this true in this case? Probably not at all…
      Let us look at the magnesium silicic proportionality.
      SiO2 and MgO has about the same suspended weight, so they should not change in proportion that much. But we still see that the magnesium is 4 times as common at FIMM compared to EYJA while the silicic value is higher for EYJA. Why does the magnesium/silicic proportionality not change? Hm, well because it forms a rather nasty surprise as the magma cools into the rhyolitic mush. SiO2/MgO forms into Olivine, a rather stable rock as long as it is not hammered with water. Ie, it does not change untill erupted. Well, that leaves us with two things.
      Wrong SiO2/MgO proportionality, and some really nasty and heavy metals in “large”abundance at FIMM.

      So, either one of two things has happened.
      1. It is the same magma, during the eruptions before the last one a sufficient amount of magma has stayed and undergone complex to very complex chemical transitions, while at the same time un-altered magma has stayed somehow under the feeder (below MOHO) as some sort of hanging blob, without the heavier elements stratigrafying out due to gravity. The blob then got pushed up through a very open conduit that was fresh and free from any of the remnant magma. To be honest, not likely for so many reasons.
      Second alternative. Fimmvörduhals erupted with new magma from a deeper feeder than the original magma in Eyja. What happened next is anyones guess. Either that commotion knocked Eyja open, or it was as Erik Klemetti stipulated, that it was FIMM magma that pushed old mushy rhyolitic lava up through the craters of EYJA. If so logically it follows that then the amount of heavy elements should have increased over the time the eruption continued, this did not happen. The heavier elements even dropped in quantity over time.
      Assumption here is that the FIMM magma is either from a volcano feeder of it’s own, or is from the Godabunga magma-reservoir. If the later (seems likely due to GPS and other activity in the area of Godabunga) FIMM was the first eruption known of Godabunga Volcano. And then Eyja would have been triggered into eruption by the Godabunga radial fissure eruption.

      Damn, now it got complicated… Read and weep.

      Weirdly enough, the difference in lavas are bigger between FIMM and EYJA, than between Eldgja, Laki and Veidivötn. Just so everyone would get the proportionality of the difference. Ie, the magma from Bardarbunga is more related to the Magma of Katla, then FIMM and EYJA is.
      Gah, now I need whisky… Lots of it…

      1. After rereading this I am still not believing my own eyes.
        I think I will go and get drunk now while the rest of you try to rip this into pieces. Or I just continue packing before going home for a week.

      2. This none smoking period comes easier you know, as time goes by (quit January 😉 ) recommend gum (nicotine in between, not all the time since you want to get rid of an addiction, not to keep it at an even level) and all the fruit/weggies you can eat.

        Still know today, that I can’t take a smoke (ergo no drinking either, need all my wits around lol) , I’d fall from my heaven ; )

      3. Very interesting line of reasoning Carl! First of all, the differences in abundance of Potassium (K), Iron and Magnesium is consistent with the Fimvörduhalsi magma coming straight from the mantle whereas once that was blocked (M3.5 eq that stopped the eruption) and the eruption from the crater being remobilised 1821-3, 1612 and even possibly 920 AD eruptive phases. Since Eyjafjallajökull does not possess a magma chamber but rather a series of sills within basaltic rock (think Devil’s Postpile with magma forcing itself in between the posts, explains the myriads of detected eqs since 1990 as well), the fractionation process must be very different from those observed in volcanoes with regular magma chambers located in dacitic to rhyolitic rocks.

        Second, when magma cools it is subject to the process of differentiation, as the temperture in the melt drops below the solidifying/melting point of a mineral, it crystallises out of the melt. The first minerals to be deposited by this process are the sulfides. That accounts for a lot of metals such as Iron, Nickel & Cobalt. Incidentally, repeat the Lady E eruptive cycles for half a million years and you’ll have a grand mine!

        Third, the mineral Olivine is the first to fractionate as it forms at a relatively speaking high temperature. Furthermore, Olivine is a really a mixture of two intermiscible minerals, Forsterite [Mg(SiO4)] and Fayalite [Fe2(SiO4)], so with those two metals being abundant, I’m not surprised by their dearth in the E magma in relation to the F deep-sourced magma.

        I’m not saying that you’re wrong Carl, just throwing a spanner or two into your reasoning.

        PS. Cuprium® – registered trademark (alloys), Cuprum – Copper metal proper. 😉

      4. So that could also mean that FIMM is likely to erupt again? and was not only a fissure of Eyja..? Since when it has its own magma source it is most likely to become a “new volcano” (I know it is not always the case that a volcano erupts more then once)

      5. They are from the same volcano. Because it is the same volcano.

        Fimmvörðuháls is just a local name for part of Eyjafjallajökull volcano.

        What happened was magma mixing when the top crater did start to erupt. As a left over from the eruption of the year 1823 in Eyjafjallajökull was still melted inside Eyjafjallajökull volcano. When new and fresh basaltic magma came into contact with that almost rhyolite magma the result was explosive and andesitic magmas where formed (SiO2 ~ 60%) (http://en.wikipedia.org/wiki/Magma).

        Now. Please keep the fact and not fantasy.

      6. lol, Jon stuck with the role of Fantasy-Buster again.

        It is all good speculations though, and I like such thinking and searching for possible mechanisms … creative thinking might generate valuable new insights … but , yes, it is good to get authoritative facts in order to stay ground in reality.

        Anyway, great info on magma types and magama processes, even if the hypothesis is not correct. So thanks for proposing it and thanks for the elaborations on it! I like when the blog becomes a classroom. 🙂

      7. Jon, you’re learning! ;o) You gave justification for your statement. That’s good!

      8. First of all: Both eruptions took place in the volcano: Eyjafjallajökull. Fimmvörduháls is the name of the area, where the flank eruption took place. It is not a volcano of its own.
        The eruption of Fimmvörduháls consisted of basaltic lava, the percentage of silicate was around 48% if I remember correct. This lava was pretty thin and very hot, so it distributed very easy.
        The second eruption (at the main crater of Eyjafjallajökull) had low silica levels in the beginning but this changed over the length of the eruption. In the end this was around 62% – which is rather explosive. It is thought that this happened through the mixing of fresh basaltic lava with evolved rhyolitic lava from the 1823 eruption which got remobilized then.
        The basaltic lava inside of Eyjafjallajökull got very fast to the surface, if I remember an earlier blog post of Jón correct, the rate was 1,6km per day.

  2. What do you mean Carl? That the lavas of Eldgjá, Laki and Veidivotn are all different between themselves, or you mean that the lava of those fissure eruptions are different from the central volcano associated (for example, different between Grimsvotn and Laki)?

    1. I meant that the lavas of Eldgja, Laki and Veidivötn are different from each other. Not so surprising since they come from three different volcanos.
      I did not check between central volcano and fissure eruption, ie Grimsvötn and Laki.

  3. I found really interesting that Torfajokull and Veidivotn nearly always had eruptions at the same time. But you were telling us that those magmas are different. Well, the magma of Torfajokull in 1477 ~(Laugarhraun) and the other magmas in other places of the caldera (to the west) are also different. This is quite peculiar. So to where does it eruption belongs: Laugarhraun 1477 and Veidivotn 1477? Both to Torfajokull, Both to Bardarbunga? One to Torfajokull and another to Bardarbunga? Or are Veidivotn magma a unique kind?

    What about Grimsvotn and Laki? This should be easy to test, because material from Grimsvotn is everywhere and Laki lavas too. The same goes for Eldgjá and Katla.

    1. In the eruption in the year 1477 it was Bárðarbunga volcano that started a volcano eruption in Torfajökull volcano (according to what I have read anyway).

      The reason is simple. Fissure swarm of Bárðarbunga volcano cut into Torfajökull volcano and when magma travels it in a fissure eruption, it can break into magma chamber of Torfajökull volcano and start a eruption. The resulting eruption is explosive due to different type of magma in both volcanoes.

      Torfajökull volcano fissure swarms also cut into Tindafjallajökull volcano. So it is a chance that eruption in Torfajökull volcano can start a eruption in Tindafjallajökull volcano.

      Hekla volcano fissure swarms cut into Torfajökull volcano. So it can also start volcano eruption in Torfajökull volcano, if magma goes that way in a large eruption. Torfajökull volcano can do the same for Hekla volcano if magma travels in that direction. But only to a limited extents.

      So far this did not happen in 20th century. As it was rather quiet period in terms of eruptions in Iceland.

      Iceland volcanoes overlap in few areas. Both fissure swarms and the main volcano them self. The result from that can I interesting in my estimate when a really large eruption happens. But that is rather rare event I think.

      1. I think this should all be confirmed by checking chemical composition of the lavas… As we just noticed it can throw us odd surprises…
        I am never again going to believe two volcanos are the same again, or even different parts of the same apparant eruption is really from the same volcano…
        Jón, you where prophetic last week when you said “no volcanos share magmatic links in Iceland”…
        More chemistry less guessing I think for us all.

  4. Question: Couldn’t Magma composition vary according to the time at which the chamber fills i.e why does magma from the crust have to be consistent and uniform in composition with time? There is much to suggest that the inner workings of the core and mantle are quite turbulent so perhaps this is more a maker of when the chambers fill rather than an indication that it fills from different volcanic systems?

  5. M 5.4 – NORTHERN MID-ATLANTIC RIDGE – 2011-08-31 12:17 UTC
    could it be the spark of this harmonic??

    Does somebody have ever try to follow up the possible relation between earthquakes on the mid-atlantic ridge and increasing or apearance of magmatic harmonics on Iceland hot-lava spot end-ups (volcanoes)?

    1. YES! … sort of … I did that and got a surprising correlation which led me to predict a couple of months in advance of a 4+ EQ in Iceland somewhere along and mid Atlantic ridge, and most like to have a volcanic activity at the same time, most likely in SE Iceland … and thereafter, about 4 weeks later than my best estimate for it’s occurance (wide sigma on the time estimate), a 4.1 EQ hit ( Saturday 21.05.2011 18:42:19 65.069 -15.133 1.1 km 4.1 39.26 36.2 km NE of Snæfell ) and Grimsvotn erupted May 21, 2011.

      I was going to predict that here on Jon’s blog, knowing it would get incredulous laughs for the most part. But it was very long and detailed with explanation and data, so never bothered to paste the whole thing here (I have it as a file, of course) … and prefered to wait to see if any additional data could back up the idea, to reduce how naive and unsupportd it would appear to be.

      So, I’m looking to acquire more data on such a hypothetical process before posting about it. Since it is long and detailed, I figure I should create a WordPress blog on Volcanic/EarthQuake stuff first and post it there and make a link-reference to it here for possible comment … or amusement.

      I think it is interesting and it makes a lot of logical sense, to me; and some data can be seen to support the idea … but need more confirmative data.

      1. The northern mid Atlantic ridge is massive, the chance that an earthquake coincides with an earthquake in Iceland is large.

    2. There is none. Iceland is on a large rift zone. But that rift zone is in parts so to speak. So activity at one place does not mean activity on other place.

      Do not let random chance fool you in this matter.

  6. When looking at the tremor plot of Slysaalda, it see double as many blue and green spikes than some days ago, and they are larger too. Also all tremor bands are rising at the moment a little, but that might be by chance, due to weather or so.

  7. Looks like the harmonic tremor coincided with a marked southwards movement of Godabunga:


    I find this interesting as the harmonics was seen mostly on the Slysaalda station north of Katla according to Jon. Perhaps something is pushing up there, which is weird given that most of the quakes are seen south of or in the middle of the Katla Caldera. But it correlates well with the fact that AUST, ENTA and OFEL are the most recently inflated GPS stations at Katla and they are all north of the Caldera.


    Combined with the increasing tremor at slysaalda as noted by Amandus I think this is starting to look very interesting!

    1. I’ve been looking at http://notendur.hi.is/runa/eyja_gps.html daily for weeks, and I agree that recent movements of ENT, AUST and even OFEL look interesting, and could be construed to be consistant with inflation … but a big problem with that picture is why is GOD just sitting there not moving up and not moving west, as I’d expect with inflation of Katla? Can Katla be inflating if GOD is not moving?

      Additionally, can there be magmatic-inflation of Katla without SOHO moving south? I’d expect SOHO to be heading south with any meaningful inflation.

      But the recent weeks of REFL’s upward and eastward movement would also be consistant with Katla inflation.

      So it appears to be a mixed picture. But, as you point out … hmmmm, looking very interesting.

      I think that page is the best one to scrutinize to get an early view on Katla inflation and its subsequent eruption. Along with th quake data, of course.

      1. It depends on the shape and depth of the inflating body “down there” what kind of movements can be seen on the surface.

        Also, it depends on if Katla and Godabunga are inflating together and on what are the relative inflation speeds. Or, it may even be Torfajökull or Tindfjallajökull.

  8. Aaaaaaaaah, it’s really starting to irritate the s**t out of me that I can’t press any “like” button here 😛 Everytime I read something interesting here, I wanna press “like” 😛

    Can u fix a button like that, Jon?

    Katla: Is there any link between what happens now and next and what happened in July?

    1. If you mean the glacial flood that took place earlier this year then yes there is a link. Actually it is the same activity. The volcano might have had a subglacial eruption (very small) which was not seen on the surface. This would have caused the glacial flood. But that is just one of many signs of unrest as magma is pushing up through myrdalsjökull there are alot of earthquakes, tremors, inflation and in that case a glacial flood.

      So basically all events are linked to Katla. Its her way of saying she will wake up soon. And by soon I mean within a 10 year period..Maybe 1 year, maybe 10 years.

    1. It has been explained here on the blog before. It is not a “negative” earthquake per se.

      Magnitude calculations are based on a logarithmic scale. A tenfold drop in amplitude decreases the magnitude by 1. Example below…

      an amplitude of 20 millimetres corresponds to a magnitude 2 earthquake.
      10 times less (2 millimetres) corresponds to a magnitude of 1;
      100 times less (0.2 millimetres) corresponds to magnitude 0;
      1000 times less (0.02 millimetres) corresponds to magnitude -1.

  9. Good morning 🙂
    Ah Ha! At last people are looking at Slysaalda!
    I posted a plot from Lurking made last year, in a previous post yesterday. here it is again.

    I also looked at the all the quakes under Myrdalsjokull in the last 6 months. I didn’t post the whole of the last 12 months due to the quakes from the Eyja… eruption earlier last year, just for clarity. The quakes to the North & West towards Tindfjallajokull have not been reviewed. Maybe because at the time they seemed unimportant, I don’t know. However there seems to be a definite trend toward the North west facing side of the icecap. Here also is a graph showing numbers of quakes for each area.

    and Numbers of earthquakes for each area

    and finally Priest Bulge for Goodabunga updated recently.

    The Cumulative strain

    I cannot find a map showing quakes at and near Torfajökull but they certainly have been on the increase in the last year.

    Reading what has been said by Jon, Carl ,Irpsit and others I am thinking……
    “Curiouser and Curiouser ” said Alice. (Alice in Wonderland by Lewis Carroll) but eventually something has got to give…… but what will trigger it and where?

    1. When I look at the cumulative strain for the period 2009- 2010, then it shows the same diagram for Eyjafjalla (blue) as it does in the last diagram
      (2011 – 2012) for Myrdallsjökull: a very steep climb!
      If Katla does the same as Eyjafjalla, it could erupt in 1 to 2 years?

  10. Apologies to Jon but this is Off Topic but may be of interest to Iceland Volcano fans also 🙂

    Just to remind people that there is an underwater survey of Colombo Volcano starting this afternoon. There will be professional Volcanologists aboard and it is interactive. You can ask questions and get almost immediate replies. It’s brilliant!


    GRRRRR! I have been waiting for this and now the weather is so good here in the North West of England that Family matters have put paid to an afternoon glued to my PC!.
    It is not often I would have prefered a wet afternoon!!

  11. Jon, In the notated tremor plot shown below, you’ve indicated “G” is “harmonic” tremor.


    But how does that differ from A, B, C, D, E, F & H? Do they also possibly show harmonic tremor? If not,what is different about each of them in a rule-based criteria context?

    Also is the icelandic usage of “harmonic” tremor the same as the english usage of the expression … that is, as it is generally defined?

    In identifying “harmonic” tremor, wouldn’t it be especially important to see the difference in amplitudes in the three different frequency channels? With “harmonic” tremor likely to show up predominantly in the lowest frequency channel? But that comparison can’t clearly be seen on the standard tremor plots?

    1. This is simple, harmonic tremor spike is different the a earthquake.

      A, B, C, D, E, F, H are earthquake, both close spaced and large spaced ones. The one you marked as G, is both earthquake and harmonic tremor. But the spike closer to F is a earthquake. But the spike closer to H is a harmonic tremor spike or a swarm of really small earthquakes. But it is a question of H is a second harmonic tremor spike or not.

  12. Thursday
    01.09.2011 18:44:30 63.764 -19.539 7.4 km 2.2 76.52 10.1 km NNW of Básar
    01.09.2011 18:44:27 63.702 -19.303 1.1 km 2.5 90.01 7.4 km NNW of Goðabunga
    01.09.2011 18:44:27 63.660 -19.194 2.9 km 3.2 90.03 3.5 km NE of Goðabunga

  13. What is happening at Leirhofn? The tremor graph shows a large spike. Also Katla seems to be having a busy day. Is Katla going to errupt 18 months, after Eyjafjallajokull as was recently predicted?

Comments are closed.