Posted on by Jón Frímann

The harmonic tremors of a volcanic eruption

Many people want to know how to identify a volcanic tremor when they appear. Here are few examples of volcanic tremors that I have seen over the past few years.


The volcanic tremor following the eruption in Grímsfjall in 2004.


Harmonic tremor spike in 2005, following a glacier flood in Skaptárkötlum in Vatnajökull glacier (they are in Hamarinn volcano).

I hope that this two examples show people what to look for when something is going on in Icelandic volcanoes. All images are from Icelandic Met Office web page.

33 Replies to “The harmonic tremors of a volcanic eruption”

  1. Jón: Thanks for elucidating this particular aspect of tremors. Now I hope we’ll be able to tell the difference (and what a difference) between magma and water/ice.

  3. @renee, Besides the strange changes in the harmonic tremors that started yesterday everything looks normal and nothing strange is going on.

  4. Thx Jón! In the light of this, Askja shows some very interesting patterns over the past three days as does several more stations as Renee points out.

  5. Jón, I’d be interested to see the Grimsfjall chart annotated to show when in this tremor sequence the actual eruption began and ended?

  6. For those like me who know even less, what are the key parts of the graph that say this is a harmonic tremor, is it the periodicity or those sudden sharp spikes?

    Thanks.

  7. @M, At the top it is the rise just before the eruption that is a clear give away on what is going on. In the second on it is the oblivious spike that can be seen on it. The smaller spikes there are earthquakes, not harmonic tremors.

  8. Uhm, annotated even in the crude sense as I have done here ? [blush]

    Enjoy reading your observations and looking at your accompanying examples but I don’t know where and how you wish for us to direct our gaze.

  9. interesting picture on the Helka cam is that a bright light or star in the background

  10. @Raving: Good question. One that can only be answered like this: the first “here” is meaningless if you don’t have the second “here”, and by then, you have a big boom. Jón is quite right when he says “it could be, yet it could be not” leading to an eruption. He, and us all, need the second “here” to be sure. This is the way I see it, and can pretty much understand why it is so hard to make forecasts. When you’re 100% sure, probably it will be to late.

  11. I’ll place here a part of the article Erik mentioned at his blog. Maybe this could shed some light to owr discussion: “VOLCANIC eruptions are sometimes accompanied by a characteristic type of seismicity known as harmonic tremor, in which the signal is dominated by discrete vibration frequencies1–4. This harmonic structure could reflect resonance behaviour in the excitation source4–6 or filtering of the seismic waves as they propagate through the surrounding rocks7–10; but complexity and variability in the properties of volcanic systems make it difficult to discriminate between such mechanisms. To address this question, we have analysed the source and propagation characteristics of seismicity at Old Faithful geyser (…) We find that sharp pressure pulses inside the water column trigger distinct seismic events that give rise to a harmonic ground response whose frequency varies spatially but not temporally. A superposition of these seismic events creates the appearance of continuous harmonic tremor. The absence of resonance within the water column suggests that the harmonic motion must arise from the interaction of the seismic waves with heterogeneities in the surrounding elastic medium—most probably a near-surface soft layer.”
    Now my question is: do we need to have a tube or conduit to get a ressonance from magma or it could also happen in a fissure?

    1. We are talking two different phenomena here: Resonance waves and transient pressure waves. I’ll use water pipes as an analogy, but magma flowing inside rock behaves essentially the same way – the only difference is characteristics: volumes, pressures, speeds, viscosities, etc. being different.

      Loosely speaking, resonance occurs when energy leaves a given place, and comes back with the same phase as it left (e.g. hammer noise in a water pipe). It involves pressure waves, but these are not transient, they are either stationary or moving. And, you usually need a chamber, a hole to a chamber or a “straight” pipe for a resonance to occur. Transient pressure waves can occur anywhere, when an external force sends an one-time kick or pressure front to a medium (e.g. an over-pressure valve closes in a water pipe). They can put resonance modes “ringing”, but it is a decaying case (otherwise it woke up a resonance).

      As magma pushes its way up to the surface, it creates these “kicks” all the time, as surrounding rock bends, cracks, etc. If magma were to create a resonant wave, it must be in a essentially “straight” pipe, or in a chamber. So, magma can create resonances, but I guess tremor (high number of transient waves created for a long time) is more common.

    2. It can be confusing because the item that is ‘resonating’ is misleading. For volcanic tremors it is not the magma, fluid or gas which resonates. Rather it is the enclosing rock. A famous example of a resonating solid driven by the passage of fluid which induces sympathetic vibrations: Tacoma Narrows Bridge Collapse “Gallopin’ Gertie”

      Think of a big organ pipe. In that situation the air resonates in the cavity.

      Think of a bow being drawn across taunt cello strings. The bow acts as a fluid which resonates the the taunt cello strings which transfer the vibrations to the hollow cello body cavity which resonates as an organ pipe (complex hollow cavity) and whose resonance is colored by secondary sympathetic resonances which are established in the wooden cello cabinet.

      1. @Raving, Are you sure?

        For Tacoma bridge, the wind would have not been strong enough to irritate the bridge’s own structural resonances, if the bridge was properly designed in terms of insensitivity to torsional vibrations. And, even after the design failure, it was not the wind directly as its density is too low. It was the drag generated by the off-planar wind on the deck of the bridge. Wind was still of low density, but it can generate appreciable force, if the area is large. And, that drove the torsional forces leading to torsional vibration (and finally to collapse).

        In organ pipes, yes, it is the air that resonates inside the pipes. And the sound originates from the holes in the pipes, not from the pipe structure.

        In water pipes it is the water, that is carrying the pressure waves (closing valve, hammering noise, etc.). The pipes only make it heard, when there are discontinuities on the water flow (bend, joint, etc.). The density difference (between the medium and surroundings) is not anymore that big, and water pressure effectively decreases it.

        Magma is much like pressurized water, only more viscous. It surely can resonate, and it surely can also transmit vibrations. Both statements are true due to the fact that sound waves are longitudinal. And fluids transmit vibration energy always longitudinally. Earthquakes can transmit energy longitudinally, too, but also transversally as transversal vibrations are permitted for solids. Generally, fluids can not transmit energy transversally.

        So, if tremor is due to magma moving, it is essentially like sound waves, i.e. longitudinal vibrations. Magma is viscous, but underground the pressure is huge, hence magma underground behaves like water in sand or air. When fluid (magma) moves, the pressure waves occurring can not generally create transversal vibrations in the surroundings (although the vibration can show as transversal movement, too, e.g. on surfaces).

        Hence, I’d say I am not that sure, is it the magma itself emitting the tremor, or the rocks responding to magma movement.

      2. You know I think that is first time that I understand the distinction between transverse and longitudinal (compression) in regard to the phase (solid, liquid, gas) of the media wherein it propagates. Thank you.

        I can see transverse waves are set up in a solid because of the strong coupling in 3 dimensions.

        Still am confused about the distinction in the nature of spatial couplings of liquid and gas. Ideally one is compressible and the other is not so. Liquid seems to exhibit some transverse (surface) waves which have some connection with liquid depth. (Knowledge long forgotten for me). Wouldn’t liquid compression waves exert a considerable transverse component if confined to a solid conduit? I expect that the gaseous state being ‘compressible’ would behave differently.

      3. The only big difference between longitudinal waves (sound) in liquids and gases is amplitude. In gas amplitude is higher due to lower density.

        Waves on the surface of water raised by wind are not transverse, see the animated gif in the end of this page. It only looks transverse, but it is not, as liquids typically do not permit transverse waves propagating. Even the tsunami generated by the Boxing Day quake on 2006 was not transverse, even though it definitely caused locally transverse motion. Instead, it was a shock wave (a form of transient sound wave).

        A longitudinal plane wave does not exhibit any transverse components. In reality, there may be a transverse component (in reference to the principal direction of propagation), but locally this minor component behaves like a longitudinal wave (to another direction).

      4. Jack @ Finland, The animated gif showed the KE+PE coupling relationship between gas and liquid on a surface and the role of surface tension in a gravitational field. Yes, I understand now 😀 . Your explanation was wonderful and clear. Thank you.

  12. I don’t actually think it’s a “resonance” per se. The only time I get a resonance in my water pipes is when you get that rapid vibrating hammer noise. Left long enough and you’ll pop a pipe. Even with out that, you can still hear the water flowing when it’s moving through the pipe. That’s non-resonant. It’s also probably closer to what you see in the tremor plots… only the tremors are much lower in frequency.

    One would think that with the right gear… and a correctly located transducer array, you could localize the source of the noise to some extent. Phase matching the signals like you do with a Rotman-Turner dielectric lens for rapid radar localization in the more advanced intercept gear. I imagine that at the really long wavelengths the gear (well, the lens) would be ungainly. Essentially, you would need correctly phased delay lines from the transducers (sensors) to several gain matched intermediate units. Then, which ever intermediate channel had the highest signal would give you the the direction that it came from. Adjusting the phasing network would in essence allow you to point your “ear” in any direction tangential to the sensor array. You would have to have a second sensor set to eliminate false direction info… at radar frequencies this isn’t needed since your antennas are already highly directional.

  13. @Lurking: Didn’t quite get, but sounds feasible. But do you need full-closed pipes or tubes (conduits) to get the “harmonic” reading, or they could also propagate in incompletely closed channels, like fissures?

      1. Sorry Jack. I haven’t read your comment above. It clearly explained my question.

  15. Eh? Well how about that. Much like the previous years, but seemingly a bit early. I never could get a ballpark measurement on the average duration of the annual uplift. I’m also starting to lean towards it being the three peak event rather than two… but I could just be wandering aimlessly out in left field.

    1. @Lurking, It is moving downward. But but it is not deflating. It is just inflating in a different direction. That direction is to the south at the moment. Why that is I am not sure. But I would not be surprised if that is because that the magma has found a weak spot somewhere in the crust.

  16. … let me explain the colloquialism I just used. “Left Field” is a region of the outfield in baseball. Due to most people being right handed, when they swing the bat, most of the action occurs in right or center field. Balls that head towards left field are usually struck late in the swing and have a high lobbing trajectory, and tend to be easier to get when compared to the high energy ball of center and right field. Part of the problem in being in left field, is that its boring, so you wind up with a player who may not not necessarily be focused on the game.

      1. @Lurking: You are amazing; but this time you were beaten by @Mots Fo: besides a poetess who could imagine you played baseball?!

      2. I don’t.

        I did. But I don’t now. Both sides of the outfield’s can be boring, but I take issue with Mots Fo’s right field opinion. Batters who don’t react as quickly will tend to swing late and nail the ball down the 1st base line, if they connect well it will go over or past the first baseman at a pretty high velocity. You need a quick right fielder to cover that.

        Anxious batters will anticipate the pitch and swing early, this usually places the bat just below the centerline of the ball, or else they catch it on the upward moving part of the swing. This additional upward vector tends to loft the ball high, and the bored [expletive deleted]’less left fielder has time to notice that something might actually be coming his way.

        {grin}

Comments are closed.