Posted on by Jón Frímann

Earthquake swarm in Katla volcano

Today (3. September 2011) at 16:45 UTC a earthquake swarm did start in Katla volcano. So far this earthquake swarm has been small, with the largest earthquake just with the size ML2.7 at the depth of 1.1 km (automatic data). But the quality of this earthquake was good, so location is accurate from what I can tell.

I am not sure at this moment if there is a harmonic tremor spike following this earthquake swarm or not. As the earthquakes are long period due the fact that they are created by magma movement in Katla volcano. This might be dike intrusion. But for the moment it is hard to know that for sure.


The area of earthquake activity today in Katla volcano. Copyright of this picture belongs to Iceland Met Office.


The ML2.7 earthquake (automatic size) in Katla volcano caldera. This picture is released under Creative Commons Licence, see Licence web page for more details.


Harmonic tremor plot of the current activity. Latest activity is at the end of the tremor plot. Copyright of this picture belongs to Iceland Met Office.


Harmonic tremor plot of the current activity. Latest activity is at the end of the tremor plot. Copyright of this picture belongs to Iceland Met Office.


Harmonic tremor plot of the current activity. Latest activity is at the end of the tremor plot. Copyright of this picture belongs to Iceland Met Office.

I cannot tell at the moment what is going on in Katla volcano. But the earthquakes tell me that the reason why they are happening is because of magma is moving inside Katla volcano. But latest data (not shown here) do however show that no harmonic tremor is currently ongoing, if it ever started during this earthquake swarm. I am expecting the earthquake swarm to continue and last for the next few hours. But I must point out that that is what might happen, as sometimes earthquake swarms just stop when you think that they might be about to start for real.

89 Replies to “Earthquake swarm in Katla volcano”

  1. Thanks for the update, Jón.
    Every day Katla is up for a tiny variation. Wonder how long it wil take until she decides to go.
    Hope not violently.

  3. The location in Krýsuvik is rather exciting though, it is under the lake this time around. I hope it does not start a new episode of water disapearing, all that water going away makes me nervous when it happens in a volcano.

  4. Transient time again.
    As some of you have noticed Hekla is having a transient again that is quite large.
    It is Búrfell having the transient, but there is nothing saying that an eruption is incoming at this moment.
    The transient is about one tenth of necessary, but the essential thing is that the counter-transient is to small for it being an opening transient of an eruption.

    What is new this time is that the large transient is hiding another transient pattern.
    At the same time as the large transient there are smaller transients going both in positive and negative creating a sawblade-like pattern on the corrected long-term plot.
    This is very interesting since it does not follow any known pattern.
    I think this might be related to the current activity in Katla somehow, and the magma-mevements there creating some kind of strain-changes, but I am not shure at all.
    I will mull more over this if it continous, or, if it is a new phenomenon that is repeated.

    1. Carl, could you kindly explain me how do you see the transients in a picture/link, and what to expect before an eruption (how much was this transient, and the 10x more threshol you say that it is necessary for an eruption). I haven’t learnt to read the strain/borehole graphs yet.

  7. Here you have the instant graph where you see the volymetric analog strain-meter of Búrfell (seen as a green line). That is showing volume changes in the borehole, the other red one shows the tension, or pull/push if you wish. That green line is here shown in potentials of mikro-strain counts.
    http://hraun.vedur.is/ja/hekla/borholu_thensla.html

    Here you have the long term strain. This is in the same scale as the next plot from the Heklugos of 2000. This plot is corrected for pressure changes in the atmosphere. The large jumps up and down are the main transients I am talking about. The smaller (they are actually not that small) that looks like a long series of twitches up and down are the new feature. All at Búrfell.
    http://hraun.vedur.is/ja/strain/str_corr/index.html

    Hekla eruption 2000.
    Look how Búrfell goes down and all others go up. This is the eruptions fissure ripping open starting 5,8 and then going all the way up to the top.
    http://hraun.vedur.is/ja/hekla2000/heklumynd1.html

    As you see we have a transient series of about 120, and the eruption in 2000 was 300, it is not a linear scale.

    As you notive we only have a strain at Burfell and that the other stations are not having a transient moment. This is what I call aborted eruptions.

    I hope that you see the transient in time on the first link. I could not post an image.

    1. Damn, the large transient disapeared. It was into the 5e-region though.
      But now the sawblade-transient are clearly visible. Spikeing up and down in the region of 4 thousand straincounts per minute is something new.

      1. So, 4000 is a lot you say. What are the normal spikes in strain counts as background noise? Does the red line ever changes?

        And what is the normal variation that the long term strain-meter (second link) can have? About up to 20 or down to -20? Or is it usually zero?

        Sorry, for asking this lots of questions. Then, in future we can reference other users asking this, to this thread.

      2. 4000 is not a lot, but it is a lot for the instant graph for being that spiky wave-pattern. The wave pattern are not transients, it is the same as the normal pattern, just about a hundred times bigger swings…

        That varies between strainmetres, but between 20 and 80 plus and minus. Things is that the normal up and down is those soft curves. Transients are very rapid in it’s change.

        Ask away! 🙂

    2. Thank you Carl.

      So, if you follow the instant graph in first link, it will show the transient of the other long-term graphs. It is already late to see the transient at 17.00. But I saw it before. Does it has also -120 in the first link, or its a different scale or measure?

      Could the transient be related to the Katla quake?

      Third question: you call it aborted eruptions. So, is it expected to see always Burfell going down to circa 300 and the other going up to circa 100 or 200. I see that the other stations only went up to 20, which is very little. Did this behavior also happened in former eruptions, or it is only known for the 2000 eruption.

      Why there are no strain meters for other Icelandic volcanoes?

      And last question: what does it mean the strain going up or going down, what is the difference in this up or down? Thanks!

      1. It is in a different scale, here they count it differently in the instant graph. 5eish translates roughly into 120, but it is also factored against the duration of the transient so it can vary. It is irritating that they do not give it as an energy equivalent, or even state the equation.

        I would say that there is a time causality to the Katla quake. It is though the first time I have seen that connection. Normally a transient is following a quake in the Sout Icelandic Fracture Zone (what I before called the Sprungur area, I found the scientific term for it in a paper of Sturkleton the Confuser (Sturkell). I think that somehow the difference this time might be that the quakes now was very magmatic. But I have seen as week as 0,8 in the SIFZ cause transients, from Katla it takes 2,6 or more.
        The reason for it taking more is that any small quake in SIFZ are on the third order harmonic bands on the Lurking plots, and Katla is NOT on a third order harmonic band, so the resonance takes more energy.

        As long as the eruption happens in the same place it will be Burfell going hard with Geldingaá going hardest the other way.
        It is known to have happened exactly the same way in the 1991 eruption.

        I wish there where on other volcanos, I guess this has to do with two things. It is together with Grimsvötn the most eruptive volcano. It is also a volcano that is close to people, and causes a lot of harm. Grimsvötn doesn’t disturb in the same way. So I guess it is more important to have them around Hekla then any other volcano. But in the end I guess it has to do with money (they are not for free) and that you cannot place them under a glacier. First of all the glacier is physically in the way, and secondly the glacier shifts would make the strain unreadable. A prime candidate for getting strainmeters are Askja, and as far as I know it has a couple. But I do not know if they are online anywhere, I have tried to look for them, but they seem to be hidden behind a passworded page on IMO.
        Krafla also has some I think, but they are either not IMO (belonging to the power-plant, or are behind that password page…)

  8. Carl le Strange says in the last thread:
    September 3, 2011 at 20:18

    “Can you do the same plot, but without the time-separation? Ie, treat all the data as one day? And do them east/west and one north/south.”

    It wasn’t until after I started rendering it that what you essentially asked for was a time filtered version of my original plot set.

    Had an issue with whitespace (margins) from one view to the next since they were each done as a separate plot rather than a rotated view.

    View East Jul/Aug/Sep quakes
    http://i54.tinypic.com/e13lle.png

    View North Jul/Aug/Sep quakes
    http://i51.tinypic.com/33w2hhy.png

    1. It told me something interesting. The Godabunga quakes are not well localized, but the quakes of Katla (more proper) is localized with something that looks like a tubeformation on the east/west plot.
      The formation is at ca. 63,61 ish… 63,6 + 2/5 out of 63.63 more like.

      1. Are the low depth quakes at 63.54 and -19.35 the same? Are they the quakes in the south rim of the caldera outside the glacier, the interesting spot as we called?

        If so, look! There is a kind of sub-chamber that you can see in the longitude graph. You can also see it as a dyke/possible fissure in the latitude graph.

  9. Oh my Gawd… Giggle celebrates a new triumph! 🙂

    Here is the article with Jón Frimann Grant! (Google translated Jónsson in a interesting way…)

    “Apparently jarðskjálftahrina began in boilers of today. So far, the largest earthquakes in the 3 on the Richter scale of 1.1 kilometers. Jon Frímann Grant, who keeps a lot of ground movements in Iceland, however, that earthquakes are good quality. Most earthquakes originate in Hábungu. Earthquake was also Katla yesterday and he was over three.

    On its website, said Jon Frímann not sure what the cause quake. He wrote earlier today that Katla was beginning to swell, which usually means that the magma flow into the volcano. When dialed Employees at supper was an earthquake scientist beginning of shift and not considered likely that any further burst of earthquakes were coming. “

    1. I have the latest craziest theory regarding Icelandic volcanoes.

      And it’s much crazier than you expect.

      No, it’s nothing to do about the moon alignments, or whatever. It is the following…eheh:

      Icelandic volcanoes work during the week, and so they remain calm. But in Saturdays, they decide to awake, and come to life. Yes, it’s true. See for yourself. Grimsvotn erupted in a Saturday. Katla had a flood in a Saturday. Two weeks ago, on a Saturday, Katla had a 3.2 earthquake, probably the biggest earthquake this year. One week ago, on the Saturday-Sunday night, Hekla had a 2.0 earthquake. And now, Krisuvik and Katla are having large swarms and Hekla a big transient.

      Why do Icelandic volcanoes seem to seek eruptions on Saturdays?
      Is it to maximime their killing potencial?

      1. As someone said, and being Icelanders themselves, they drink heavily on Fridays nights, and so they tend to vomit on Saturdays.

        The big Volcanic Hangover!

        Remember Grimsvotn eruption? Katla flood! Yes, they were only Saturday Hangovers!

      2. We all know that the Volcanos are really bad drunkards that can’t keap down their beer. It is that simple 😉
        Worst of them all is Hekla, she only need one small zip and then she hurls all over.

    1. Yepp, I love the cumulative seismic moment, I think that in the end that will show itself to be the most important number for many volcanos when predicting eruptions. The sad thing is that we do not know the figure for the last eruption though… 🙂

      1. I hate the cumulative seismic moment.

        I can’t reliable replicate the calculation. I can make traces that are similar, but until I can reproduce the official ones, I am not willing to post them.

      3. Okay, here goes…

        IMO gives the number for the cumulative seismic moment. But the question is what the hell is that really?

        Let is beginn what you are calculating (if I have understood that). You are doing a sum over the released energy in the earthquakes. Yes?

        Cumulative Seismic Strain (CSS) = The energy built up that is the potential of possible quakeing. Measured by summing over history of STR (Seismic Tremoring Release)
        Cumulative Seismic Release (CSR) = The energy released in quakes. (Your plot I think)
        Cumulative Seismic Moment (CSM) will then be logically formulated as ((CSS-STR)+CSR)=CSM. CSS-STR then gives a residual value of strain not being able to tremor our that gives a surplus potential for the future (as in an eruption or a major quake).

        I think this is what makes your plot not give the same numbers as IMOs. It is not a question about apples and pears, it is a fruitbowl having a swingers-party in the seventies. Your plot says the same thing, it is just giving other values.

      4. CSS-STR then gives a residual value of strain not being able to tremor our that gives a surplus potential for the future (as in an eruption or a major quake).

        Should be:
        CSS-STR then gives a residual value of strain not being able to tremor OUT (as in away) that gives a surplus potential for the future (as in an eruption or a major quake).

    2. At this rate we will get an eruption when? If it has the same threshold as Eyja (not likely) it will reach 6e+09 within two months. So we would have an eruption by late October.

      1. Well,, we do not have a clue what it takes. But I would guess a few months at the minimum. Cumulative Seismic Strain only works well when we know the numbers needed.

  11. Jon Frímann Grant, who keeps a lot of ground movements in Iceland…

    Heheh, stop keeping all the ground movements, Jon, it can’t be good for you!

    Congrats on the article, though! 🙂

    1. And we also have an age on Jón now… 🙂
      “and he was over three”… 😉

      Starting to be quite a celebrity. But the good thing with the Icelandic pappers is that they actually get the picture.
      But Giggle Transfuse seems to have a personal grudge with Jón. I have never seen anything that warbled coming out of it. Even I would have done a better jobb of it, even though The Other Lurker cries every time I translate Icelandic… Minn íslensk gerir Öðrum Felur gráta í hryllingi. 😉

  12. @ Jón Frímann at 21:43

    This has not been updated for over a month – beware of old data

    1. When one speaks of the… 🙂

      Gögn eru alltaf gömul og í augnablikinu er liðinn.

      Sorry, I will now shut up forever… at least in Icelandic and this week 🙂

      2. Now, I actually did that on my own… But I guess it was horrible… 🙂
        I am trying to learn Icelandic due to contemplating where the heck I am going to move…

      3. Data is always old = Gögn eru alltaf gömul
        but the other part, the words are right but don’t quite make sense
        for the moment over = í augnablikinu er liðinn.

        Were you perhaps trying to say

        Data is always old as soon as it has been written ? Gögn eru alltaf orðin gömul um leið og þau eru skráð 😉

        The moment has passed = Augnablikið er liðið

      4. I tried to write:
        “Data is always old, as the moment has passed.”

        It is so irritating, I should be getting it better than I do, after all it is very close to my native language.
        Carpe cerevizi…

      5. Gögn eru alltaf gömul þar sem augnablikið er liðið

        You’re not picking the easiest sentence to translate, even if you think it similar to Swedish (believe me it’s not, I speak Norwegian and understand Danish) , so you set yourself up for failure. Cry, no lol. Kudos for trying

      6. Actually, in some way you are closer to swedish than to norwegian.
        Take my favourite missunderstanding. Kaldársel. I missunderstod it to be Cold Arse, since my map didn’t have the á in it. (Swedish Kallarsel), but Kaldársel, has the exakt same meaning is Swedish (correct this time) if slightly changed in spelling to Kallarsel. As you notice the cold butt and the Kaldársel will be spelt the same way in swedish, though we pronounce it differently. KAllarsEll and KallArsel. We kind of sloppyfied things on the road through time.
        That is why we have Mensträsk, that depending on how you pronounce it is either Menstruation Swamp, or “The swamp that will leave you damaged for the rest of your life” Men=handicapp (in this case).
        And then we haven’t gotten around to Kräkångersnoret (Vomit regret snot) that could equally well be Bovine Small Field River-bank. With my humor I prefer the first, but the second is the correct one.
        So what trips me is mostly that I cannot get my hand around the differentiators you guys are using determining the way to interpret some words like Kaldársel.
        And then the grammar, but that is just to hammer in…

      7. Why and how on earth should you know the origin of the words, that’s what one uses and how you know where to split it, you want to accomplish something not all natives can do 🙂

      8. Now you are finally starting to understand how esoteric I can get in my lust for knowledge.
        I sometime have a tendency to not see the bark on the tree that is in the forrest because I am counting atoms in a single cell of the tree…
        Sometimes I get to new result this way, but normally I get lost in the forrest (metaphorically).

      9. Hi Carl,

        You have a very interesting interpretation of Germanic placename etymology. I would suggest you read up a bit and you will find the names even more intriguing or perhaps less.

        This is a volcanic blog, I know, sorry for the off topic.

        I am certain Kaldarsel has nothing to do with your arse being cold. Sel is calm water, kaldar is most likely an adjectival form of cold (I am not absolutely sure since I don’t have a dictionary at hand and don’t know Icelandic).

        Swedish Mensträsk which you often talk about does most likely not imply menstruation swamp. “Men” or “Mens” at the beginning of the word, can mean man, while, harm, change. I do not think (and I might be wrong here, since I have not checked) that the Swedish word “mens” for “menstruation” was used in Västerbotten at the time the placename Mensträsk came into use. After having had a look on Wikipedia, they say “mens” is a Sami word meaning calf of reindeer and this is most likely the correct interpretation of the placename.

        But I have to agree Icelandic is intriguing. Do you know any dialects in Västerbottten? I am sure that knowledge would help you a great deal since these dialects are very old and have many things in common with Icelandic both etymology wise and also when it comes to grammar.

        Carl wrote:
        So what trips me is mostly that I cannot get my hand around the differentiators you guys are using determining the way to interpret some words like Kaldársel.
        And then the grammar, but that is just to hammer in…

      10. Hello AK!

        I do not know if you understand Swedish, but I guess that you do.
        “Men” in this case is the plural form av “injured for life”. As in the swedish sentence “Men för livet”. It is not harder than that.
        Kald ár sel Cold area with slackwater. But, I prefer the Cold Arse since it is so hillarious.
        I think you missed that I wrote about it in a style that was intended to make fun of myself.

        I speak 3 different dialects, out of one that nobody understands if I do it properly, but that is not a Västerbotten dialect.

        I am surprised that you succeded in missing that I wrote that Mensträsk did not bear the meaning of Menstruation Swamp, but that you with a little twisted sense of humor could read it out as that.
        I know that irony is the hardest form of humor to translate and understand, but a simple verbal juxtaposition shouldn’t be that hard.

  13. Hi Jon,

    What is going on on the Reykjanes Ridge? It seems to have a lot of activity all of a sudden.

    Cheers,

    Karen

  14. You are all brilliant with your banter and information, thank you, I have learnt so much over the last year and am completely hooked.
    It seems that these girls (or are volcanos boys?) are warming up, especailly as it’s a saturday and obviously “things” seem to happen in iceland on a saturday!!! all getting very interesting specially for a learner like me. I hope the wonderful icelandic people will get away quickly if what JReykdal and Carl tell us, (many thanks for your informative response in the last thread explaining why Katla is considered the most dangerous volcano in iceland…..) Another Question (I hope they are not too stupid 🙂 but has it ever happened or is it ever likely to happen that two volcanoes will erupt at the same time, are they linked? It just seems that there is so much activity going on all over the place such as Katla, Hekla, Krysuvik etc that I instictively think that there must be a connection underground somewhere linking all these volcanoes together.

    1. The only way they are connected is that they are all subject of the forces of the tectonic movement in the northern part of the Mid Atlantic Rift, and the influece of the Icelandic Hotspot. But there is no direct connection between most of the volcanos, although there are a couple of suspected cases of nearby volcanos having som sort of connection, but this is unproven.

      I do not think Hekla is Icelands most dangerous volcano. Yes, it can cause harm to Vík. But Hekla can cause harm to other areas, and is the volcano that has caused most ashfall over europe. I would say that is the most disruptive volcano on Iceland.
      The largest potential of destruction for the Icelanders is probably Hengill and Krisuvik since they can affect the economic and populative center of Reykjavik, either directly, or economically due to destruction of the power and electrical infrastructure in Southwestern Iceland.

    2. “It seems that these girls (or are volcanos boys?) are warming up”

      Anything this unpredictable and full of wrath must be a woman…hehe

      1. Icelandic volcanoes are usually regarded as females. Katla and Hekla are actual names for girls.

    1. This is the Katla swarm from today (cluster within a few km area): “qu” means it has been checked for quality, “auto” means automatic (SIL system). This list is from IMO “Skjálftavefsjá” tab, then zoomed in, choose “greina” on menu bar on right. You see only the last one seems shallow (others are still “auto” and not checked) but the big first was deep. I think it will not vomit yet, but he** I am not shure.
      date – time – position – size – dept – remarks
      2011-09-03 22:02:05 -19.1 63.642 1.9 1.2 auto
      2011-09-03 17:49:05 -19.106 63.644 1.4 1.1 auto
      2011-09-03 17:49:05 -19.087 63.643 2.2 1.1 auto
      2011-09-03 17:13:48 -19.078 63.64 0.5 1.1 auto
      2011-09-03 16:51:38 -19.072 63.628 2.3 1.1 auto
      2011-09-03 16:51:37 -19.074 63.64 2.2 0.1 qu
      2011-09-03 16:50:01 -19.058 63.645 2.4 0.0 qu
      2011-09-03 16:48:16 -19.065 63.637 1.6 13.3 qu
      2011-09-03 05:10:13 -19.079 63.646 0.4 0.1 qu
      2011-09-02 05:17:17 -19.057 63.638 1.5 4.4 qu

      1. I think we all should use the He** not shure, phrase 🙂

        Linkur please 🙂

        There is always the problem with those set as 1.1 depth. I saw a coment during the Eyja eruption about the 1.1 depth marker over on Klemettis blog from the dude who had written the auto-locating software, and it was a dummy depth for all non-depth located quakes. So a 1.1 is almost always a bug.

        I though think that Katla now has officialy started the run-up for an eruption the will happen sometime between 3 and 123 months… Or not…
        He** I am not Shure! 🙂

      3. Changed my mind: I think it will vomit yet, but he** I am not shure. Definately whould like see “live” TV broadcast from GOD looking to south-east..

      4. Why to the south east? Wouldn’t an eruption just south of Austmannsbunga be due east of GOD?
        And on the humoros side… I never knew God was an Icelandic volcano that is to lazy to erupt…

      5. finger troble, was on phone …, meant view east-south-east from GODabunga (god) meter, but have feeling IF it goes, it will come just south of this cluster (a fissure extending from due east to the south-west, possibly) but there seems low (red line) frequency increase on god tremor plot. No real wind that I know of in area….

    1. That one we know, problem is in how IMO has formulated the CUMULATIVE seismic moment. Because seismic moment is only a part of the CSM.

      Normally one might think that a cumulative seismic moment should be just a summing over the shear modulus, but then the figures does not compute correctly. Both me and Lurking have tried it ad nauseam, and we would have gotten there because the math is simple; Shear module/Shear stress as a tensor.

      So, the thing has been to retro-enginer how IMO computes the not-defined term cumulative seismic moment (it is not defined anywhere).
      So the reasoning behind how to set up a formulation above is the last attempt to retro-engineer it. Because to get anywhere we have to find out the clues in how they are doing it, then we can construct a mathematical formulation, Lurking puts in the numbers and POOF, the numbers add up. Sofar no add up.

      I hope this clarified what we are talking about.

      1. The formula for that is J (erg1/2) and J=10(5+M), where M is the size of the earthquake.

        It is just the size of the earthquake collected over time. So 40 earthquakes equals one earthquake in given size.

        Iceland Met Office has been using this for a long time now.

      2. Yes and no, they do it for the Cumulative Seismic Release, not for the Cumulative Seismic Moment. There is a difference in those numbers.
        Yes, one could argue that we will get the same shape of graph, and that it says the same thing, but it is actually not the correct figures.
        We are nit-picking here Jón.
        Just trust us on this, when we use the J (erg1/2) and J=10(5+M) we get an end result that does not come up with the numbers from IMO. Ergo, IMO is doing something fishy with the CSM, and we are trying to find out what it is.
        The CSR value you are giving is not the problem.

      3. Well… I generally attribute it to some arcane knowledge about the shear modulus of the area that we are not party to. That’s the only thing that really explains it.

        If it weren’t for that I should be able to find an adjustment factor and scale my plot by applying that correction.

        Every time I tried that one end of the range or the other would take off to a different and unexpected value.

        I finally gave up.

      4. @Lurking:
        Tried something on the line of the logic formulation I gave above? I think it is something close to that they are doing.
        Otherwise it will be the monster solution of renormalisation I wrote about below…

    1. Actuallly it is very human. Our brains are connectivity machines that is constantly trying to make order out of seemingly chaotic behaviour. Normally it is a very good feature, but sometimes we over-interpret things and start seeing ghosts.

  20. There was a 7.0 EQ in Vanuatu at 22:55:36 UTC . I think it can appear in the graphs, as it does in Etna’s tremor graphs.

    1. A small spike in Jón’s geophone occurred at the same time as at Etna’s, 23:20 circa.

  21. Jón, I am going to try to explain this in a metaphor:
    Lurking is trying to build a cupbboard, he uses the tools that are given, but the problem is that however he tries to apply the tool he gets a cupboard that is in the wrong size, even though he follows the blue-print exactly. Ergo, something is wrong. Both I and if I remember correctly, Jack, looked at it, and got bewildered.
    In this case the wrongly sized tool is the mathematical formulation.
    When you have a tool that is the wrong size you build a new tool that is the right size. Here it starts to get tricky. I am a physicist, we build models with the tools that we have and know how to use. When we need a new tool we walk over to the math-department and either order up a new tool, or steal one.
    Problem here is that we have physicists only here at the blog, since mathematicians don’t believe in reallity and pesky real things like volcanos, so there is no one we can ask to build the new tool around in here.
    Physicists can be seen as carpenters, and mathematicians as the tool-makers. So right now, I (as a carpenter) am trying to build a tool that fits the blue-print.

    To construct a mathematical modell when you only have two sets of non-conforming, but related number-series, is called renormalisation, and contains some of the most weird math known to man.
    To give it perspective, we humans do not have the math to renormalise the theories of relativity and quantum mechanics into one unified theory. However we try we only get gibbletygook. The answer is probably very simple, but we do not have the tool. So what we are pondering has a simple answer, but untill we find the simple answer we are having one hell of a mathematical problem.

    So first we need to find out why the cupboard is to small, then we need to find the logic behind the process that gives the difference, then we can build a tool. Perhaps, and in some millenia.

    So, when you are saying we should use J (erg1/2) and J=10(5+M), you are just telling us to use a tool that we have proven is the wrong size, so from your simple math we have gone on to this math… Read the appendix on this link for a simplified understanding;
    http://en.wikipedia.org/wiki/Exact_renormalization_group_equation

    We can with enough data solve the problem with a Wetterich equation as the end result of a Effective average action ERGE, but there must by Gawd be a simpler way to do it… and that is the way I am trying to deduce somehow.
    Starting to see the depth of Lurkings headache now? And why I gave up to do a direct attack solution to the problem? We need a simplified formulation that instead of using cumbersome renormalisation does the job, ie. coming up with the formula used to beginn with, and that is not J (erg1/2) and J=10(5+M), because that we have proven mathematically a long time ago to be the wrong tool (and since it is in math we can actually prove things).

    1. I’m thinking that its going to reside in one of three (possibly more) ways of coming up with cumulative curve.

      1) Simple summing operation of the energy release over time.
      2) An integration of the energy release over time.
      3) A trapezoid function for the energy release over time (cheating at the integral).

    2. Think of that formulation like this. With that one you can calculate what the result will be of a Strain energy release event. Here we have the result, not the formulation so we do not need it.
      It does not help to give a way to in advance calculate how energetic a quake will be in a given space and time. We allready have the result given so graciously by the nature.
      The problem is really simple; One way it should be done which will give a set of number A. Problem is that we are getting a set of numbers B. Ie, they are using another formulation. Which is fine in every way to do. Problem is just that they have not told us what the formulation is (an no, it does not have anything to do with the formulation in the link you gave), they have just given us the answer. And going from the answer B to the formulation B is orders of magnitude more hard than if you had the formulation to beginn with.
      I could set up 10 working formulations that would give a set of numbers that actually described a function in the Cumulative Seismic Field. Problem is just that none of them would be the one Lurking needs since he wishes to be able to calculate EXACTLY as IMO does to be able to check data against their numbers.
      It is like we are blind people that have a screw-driver in one size, and screw another size, and we cannot see which size of screwdriver we need to buy instead.
      Believe me, Lurking has stumbled into one of the hardest fields of math,, it happens now and then. I though admire him for his tenacity in refusing “a good enough” solution. Because in a way we could just do a summing over 1 month, calculate correctional multiplyer, and then auto-aply it. But that would never be totally exact, and that inexactness grows disproportionaly the more we need an exact answer untill we reach infinity. Ie, if the need of an absolutly exact answer we now beforehand that it will be exactly wrong.

  24. I too am addicted to your great blog! The latest posts on cumulative strain and adding quake-energy up are hard to grasp for me. I would think some quakes add strain, but some probably release tension within a vulcano by reshuffling magma. My personal theory about predicting eruptions is that it is comparable to weather predictions in The Netherlands. some long term certainties , trends (global warming~more active fase coming)and facts (seasons, periodicity) and short therm uncertainty and facts limiting predictability to short horizons. (5 day forecast max). The levels of uncertainty are imo undefinable. As a cardiologist I unfortunately fall short to in predictions of future “eruptions”.

  25. Regarding the cumulative earthquake plots, to me, it doesn’t seem like a major issue to match the IMO numbers exactly. Out of curiosity, it would be nice to do that, but there can be many factors which could cause differences in results obtained … mainly pertaining ot what data is used ( some data changes, some quake data might be deliberately not used due to quality issues, etc) and which quakes get assigned to which of the four areas. And anyway, it’s just a cumulative energy tabulation, so regardless of whether they are the same exact curves, they are bound to tell the same story.

    Lurking, since IMO has refused to give updated seismic plots at http://hraun.vedur.is/ja/myr/myr_reg.html#img I’d be very interested to see the plots which you have. Is that simply a matter of feeding quake data into an already-existing spreadsheet?

    Of interest is: WHY did IMO stop updating their plots of Katla-related earthquake data at http://hraun.vedur.is/ja/myr/myr_reg.html#img … deliberate secrecy, since things are getting interesting now? … perhaps with concerns about getting papers published before others hijack their data to publish papers? Or government orders to keep the data secret, in order to avoid possibly adverse econmic impacts which eruption-concerns might generate (eg tourism)?

    Is it the old conflict between an ethical mandate for scientific openness and honesty versus strong special interests demands for secrecy, to protect profits and professional acclaim?

    In order to give a clearer picture of wht is going on at Myrdalsjokull, I “renormalized” (haha, noble prize work) the cumulative earthquake energy plots from http://hraun.vedur.is/ja/myr/myr_str.html#img to show a span of 2+ years rather than the limited 1 year slices at the prior link. It shows that in comparison to Eyjafjallajokull, the cumulative earthquake energy for Katla has not increased substantially … however, just as things got interesting, the agency publishing the data at the above link suddenly stopped publishing updates, in mid July, as mentioned above. Anyway, here is the 2+ year cumulative data plot …

    http://twitpic.com/6fua4w/full

    1. “Of interest is: WHY did IMO stop updating their plots of Katla-related earthquake data at http://hraun.vedur.is/ja/myr/myr_reg.html#img … deliberate secrecy, since things are getting interesting now?”

      I hope this is not the case. We see this “dumbing down” of data all the time by the YVO due to the fear that Yellowstone brings with it. I don’t expect it from Iceland, as they seem pretty open with thier information.

  26. One of the first things that absolutely has to be done before even attempting to tackle the formula problem, is to work with the exact input data that is used for the official plots.

    Importing the boundary image into DivaGIS and georeferencing it, I come up with this for the boundaries.

    http://i56.tinypic.com/2zio77n.png

    It’s not exact… but it’s as close as I can get. In the original image, there are gaps between the boundary markers. Not being sure where the actual boundary is in relation to the (thick) graphics, I went with the assumption that the border of one area runs all the way up to the boundary of the adjacent box.

    I don’t like assumptions, but unlike the Grimsvotn tensor graphs, they don’t annotate what the boundaries are in numeric form. So, this will have to do.

    That is, unless someone has a better idea.

    For the record, this is all academic on my part.

    The plots at http://hraun.vedur.is/ja/myr/myr_str.html#img are something that had wanted to take a look at for quite some time and was the reason that I had pursued the idea of constructing them myself.

    I just want to see how they are put together now.

Comments are closed.