Earthquake detection and winter storms

In the last two days there have been a strong winter storm going over Iceland. The wind has been coming from the north with snow and cold weather. But this has also been followed by a strong wind all over Iceland and this strong wind has been preventing the SIL network from detecting micro-earthquakes in Iceland. Because of that it appears that Iceland has been unusually quiet on Iceland Met Office maps. This might not necessary be the case in reality. But due to a lot of wind noise the SIL network is unable to detect the smallest earthquakes that happen in Iceland. Some stronger earthquakes might appear on the SIL network if they take place.

So the quiet time in Iceland is down to weather, not something else. As the weather improves we should start to see micro-earthquakes again on Icelandic Met Office maps. It is always a good idea to check the weather forecast in Iceland if there is a sudden drop in earthquakes in Iceland.

Note 1: There are sometimes times however in Iceland when there is nothing wrong with the weather and yet still no earthquakes. What is going on there is simple, there are no earthquakes taking place in Iceland. This happens sometimes. But this is a rare, but it does sometime happen.

28 Replies to “Earthquake detection and winter storms”

  1. Well… that sort of leads me a to a slightly technical question.

    Does the SIL network incorporate an AGC (automatic gain control) based on the wind speed or wind noise? Or, does the detection algorithm simply go for the minimum discernible signal… i.e. whatever it can pull out of the noise floor?

    1. That would not work, as the wind makes the ground vibrate. This vibration is high in strong winds.

      I see this also on my geophone. What I am recording there is the vibration from the wind, not the wind it self.

    2. I would hazard and say that it does contain an automatic gain. Otherwise we would have had the oposite happening, millions of microquakes.

      There is one thing you can do, but then you need to go into an old patent of mine. That would be to set up 4 measuring devices a hundred metres out from the SIL in each direction, and then do a sum over of all the “sound” occuring in between them and the SIL. What that does effectively is to dampen out noises taking place close to the measuring station. Such as trucks, cars, wind, footsteps and soforth.
      It would then be the same system as on a sub since you have to cancel out the eigensounds from the sub itself. You can ontop of that have sound banks of known errors that also is automatically cancelled out.
      This requires more computational power, more measuring equipment, but it increases sensitivity enormously.

      For those not into subs, this is the same thing that you do on a minute scale on expensive headphones to dampen out background noise. You record that background noise, phase invert it and play it back in the earpiece, the noise played back cancells out the background noise so it appears to be quite in a noisy environment. I love that little patent.

      1. Not at you or anyone in particular… but I’m getting the distinct impression that I’m not getting my thought across. Every one keeps falling into that idea of “everything is vibrating so it won’t work.”

        First, where you get the wind speed data/information/signal doesn’t matter. If you were really adept and possibly perverse, you could derive the wind-speed from the angle that a caged goat leans as he freezes out there on the icecap. All you have to do is to detect the lean, buffer the signal and apply it to your input sensor gain control. Hell, even that “everything is vibrating” phenomena can be used to derive the AGC signal. You could even use the spectral response of that to do some really fine tune gain adjustments. This would be similar to what Carl le Strange had to add.

        And, as he noted, with the large swings in quake counts… and the odd percentages, it does appear that the integration of the new gear is having an effect on what we are seeing.

        There might have been a micro-swarm and the new gear jumped all over it. (sort of it’s job). Now they are likely pouring over the data to correlate it to what is real and what is not.

        OT @Carl.

        One aspect of doing this on a sub. If you dampen out or remove all local noise… you also reduce your ability to know when you are not quiet.

        Non-sub, but similar: on surface radar detection gear, one method is to turn your receivers off when your radar transmits, then turn them back on when the pulse from that radar gets a hundred or so yards from the ship. It works like a champ but you have to adjust/tweek the circuit to deal with radar maintenance.

        From what I’ve seen, though dealing with a threat with electronic means is effective, most commanders/skippers prefer to have the satisfaction of a physical kill. This eliminates the threat since it no longer shows up as an inbound track. With electronic means the threat is still there, still tracking towards you, and you are waiting for it to turn or miss.

        Mentally, it’s like playing chicken with a tractor trailer rig.

        I imagine that silencing local noise by removing it from the gear would also have a similar effect. Now you can see (hear) for miles upon miles with a blissfully quiet noise floor… and the cooks back in the galley slinging pots and pans and beating on dinner have no effect on how well you can hear.

        There would have to be a method for detecting that.

        With the “turn the receiver off” method that I described for radar, you flip a switch and the receivers go wide open and you can instantly tell when a local radar operator isn’t following tactical guidelines or dictates. This method works since the signal sources are known, and can be dealt with through equipment design and operational procedures.

        But noise, sound, vibrations… are more pervasive. Anything can make a sound if you hit it right. About the only way to deal with it is through brute force dampening/silencing, operational procedures, or a technical method such as you described. Not that there is anything wrong with a technical method, but you have to also deal with the cook and the meat cleaver.

      2. Ah, but the beautiful thing with the system I described is that you have a recording of the cook banging, and on a sub you can even say which pan he was banging with which of his knives and where he was standing within an inch when banging.
        You truly do not have any privacy when you are on a sub. Believe me, the sonar guys will now when you have been thinking about the wife/girlfriend in the bunk…

        Second good thing with sound, you can actually with an active system use active components to remove internal noise by active dampening. You just take you banging pot sound, phase invert it, and transmit that one too. Ie, with sound active stealth is not only possible, it is fairly easy.
        But is it practical? Well you need to make large parts of the hull into a combined speaker/microphone…

  2. Well, that’s what I’m getting at.

    Is there a method in place that backs off on the system/sensor gain as the wind speed increases?

    The other method would just rely on the signal to noise ratio of the raw input. Automatic detection and tracking systems do this and toss any “hits” into a data bin for that azimuth and range, each having a time decay value. If the number of “hits” in a bin crosses a certain threshold, a new “target” is generated for the computer to munch on.

    I was just wondering if SIL incorporated something like that (or the AGC method) in the automated seismic systems.

    1. I suppose that would not work as expected. As wind puts things above ground vibrating (e.g. trees, water, etc.), those things transfer part of the vibration to ground. Hence, the ground is vibrating creating higher back-ground noise, thus preventing the detection of smallest quakes.

  3. IMO may have some work to do in the integration of the new seismometers… One recent quake near Katla was initially nearly M2, quality~30%. After review it is only M0.5 but quality is 99%. So no frost quake.

    1. I saw that one, it proved my theory of the new eqiupment not being gain-scaled to match to program-values set to the old gain-values. Love it when I am right, even though I have done this for 20 years… 🙂

      Sound-engineering and building measuring equipment that is, not volcanoes. I’ve only done them for 3/4 of a year…

      1. Since Eyja’s eruption? Me, too. Basically you’re suffering the same syndrome with me: Knowledge.

      2. Yes, it was Lady Eyja that seduced me 🙂

        Yepp, it is a terrible syndrome… Wild thoughts at night, over-consumption of coffee, eyes staring out into the great blue yonder, and a strange demented laughter…

        But it is fun when you notice that you can apply your own specialisation onto something so wastly different as a volcano measuring system. I had never even thought about how they operated before, and now I am thinking about building an active system just for fun to see if I can do it better.

      3. The best part of it is finding totally new meaningful connections between seemingly totally unconnected things…

  4. Jon,

    What Eyjaf cameras are still working? Most have been pulled due to lack of interest

  5. Míla has a number of new cams up – and it must be cold because there are people walking on the ice on the city lake. Gullfoss has quite a few spectators, and there are people enjoying the Blue Lagoon.

    1. These cams already existed actually, Mila just took these over from InspiredByIceland I believe, because the Jokulsarlon, Gulfoss, Tjörnin and Austurvollur cams were located on that site before they transfered to Mila. 🙂

    2. Are the waterfalls at Gulfoss partially frozen by the way? Or is it just the camera having malfunctions.

      1. These cameras have always been operated by Míla, although they where only accessible via the campaign.
        And there is always a lot of ice around Gullfoss in the winter since the spray of this waterfall freezes on rocks etc. The last days have been really cold.

    1. I suspect this is ice on the northwest side of the antenna (during the last storm wind was from northwest). If it was ladies B or H, I’d expect movement to northwest. But movement is northeast, so the source must be in the southwest, if it was due to inflation of magma to somewhere.

      1. I meant to say ice on the northeast side of the antenna, i.e. the direction of the apparent movement.

      2. I would say that Bardarbunga and Hamarinn (Hammer) is definitly male volcanos. 🙂
        Other male ones would be Theistareykjarbunga, Hengill and Kverkfjöll.
        Then we have the sexually ambigious volcanos like Öraefajökull and Krisuvik.

      1. It is odd how one starts to fancy things… I have some odd thing going with Theistareykjarbunga (but that has more to do with the dream of hearing all american newscasters trying to pronounce it).
        But among the SIL-stations Kaldarsel is a favourite, it suits my silly scandinavian humor.

  6. @Pieter, definitely frozen (mostly) at Gullfoss – some frozen people too, earlier, at least a lot of jumping-to-stay-warm…the winds are why many houses in Iceland have (old style) corrugated metal siding over wood or (new style) are made of concrete. High winds not unusual ( and would be considered REALLY high in other places..) and when accompanied by driving ice pellets… can see that the very early turf houses (cut turf laid up like big bricks over more scarce wood) were probably reasonably snug and definitely necessary for survival. [Icelanders correct me if I’m wrong here..]

Comments are closed.