Ruminarian VII – Mass wasting…

Okay… what is a ruminarian?   Well, figuratively, we all are.   “Ruminate” is a slang term for standing around, thinking about something.   I adopted/co-opted the word to describe our activities, in part because it’s what we do, and also due to the sheep nature of the blog, following a hapless sheep incident that someone observed on a web cam where one demonstrated it’s lack of aerodynamics during gale force winds.

Before I continue, an apology.  I had intended this post to be a starting point for Carl to elaborate on, but since he is indisposed, I’m fleshing it out for him.

“A cascading failure is a failure in a system of interconnected parts in which the failure of a part can trigger the failure of successive parts. Such a failure may happen in many types of systems, including power transmission, computer networking, finance, human bodily systems, and bridges.”

You can see this quite often at construction sites where they have piled up and stored fill dirt for use in doing the building.   As loads of dirt are removed from the edge, the top of the pile develops cracks where the unsupported mass starts to calve away from the rest of the pile, eventually falling into the vacant area, or gaining support from the material that remains as it leans over and makes contact with it, shifting the mass load over to the new contact area. This is a gravity driven process. In this scenario, the “system” is a pile of dirt. Each cohesive section/block relys upon the next block over to resist the force of gravity and provide support. As each section is removed (or fails), the next section over has to adjust to the increased strain and find equilibrium.

Hawaii is prone to flank collapses of its islands.   Hilina Pali is probably the most recent example. It is the scarp-wall of the Hilina slump, a 19,800 km3 chunk of the south slope of Kīlauea volcano.   A section of the underwater shelf has formed a “bench” and is acting to anchor the foot of the slide and slow it’s progress to about 10 cm/yr. In 1975, a Mag 7.2 quake showed a 3.7 meter drop of the block as it suddenly slid 7.9 meters towards the ocean.

In recent comments, Ian F and DeepThought have both posted graphics that seem to illustrate calving along the rim of Bardabunga. Both lines of evidence are from the plots of quakes there. I mentioned to Ian F that the focal mechanism of the quakes should match the geometry of the failure plane if it were to support the idea. Ian F did a check of the mechanisms from EMSC and they do seem to match, with the strike angles fitting a tangent to the positions along the ring-fault. Since these quakes would be normal mode faults if they signify the two sides moving apart, each Mag 5.0 quake would indicate a motion of about 3 to 5 cm separation along the fault face. (0.5 to 1.1 cm for Mag 4.0s).


Ideally, the dip angles should match the geometry, but as of right now, I don’t have the time to dig through that information to check on it.  Scratch that, I found time.

The really cool part about this is that BOTH IanF and Deepthought independently saw fault structures in their peeking at the data.  I took a look at both plots and did some note taking.  Of the six faults that BOTH of them noted, IanF’s dip angle is 32.52° to 49.40° (95% conf), and DeepThought’s angles are 32.55° to 47.17° (95% conf).  Looking at the actual average dip angles from EMSC’s listing of GFZ provided quake solutions in the vicinity of Bardabunga, the value is 46.3°.

I think they may have hit upon something.  If the values are correct, it appears that each slab is well in excess of 13262417.92 metric tonnes for each meter of radius around the rim.  Think about that whenever you see a large magnitude quake pop up on the listing.  As the inner sections shift, the outer sections loose a bit more support and are able to slowly shift down in the slow collapse of the caldera wall.  Where it will get dicey is if one of those fault planes allow water to seep down to the magma.  Then it’s anybody’s guess as to what will happen.

Now, to give you an idea of just how freeking remarkable it is that both Deepthought and IanF saw the fault structure in the quake data:  Inflection points are usually a really good indicator that some controlling phenomena is at work on a physical process and that at that point, something causes a state change.  Plotting out the dip angle verses depth derived from the two independent plots, shows something changes at about 7 to 7.5 km depth.  The depth value is the halfway point on each perceived fault face.

IanF’s data:

and Deepthought’s data

Whats it mean?  Beats me.  That’s what we sit here and think about.  If you wish to get into the more technical side of it, this may help.  Tension and Compression in Bars

Now… to make a bit of use of the data.   The ring faults are composed of six fault planes separating individual contiguous blocks.  Fault 1 separates the first slab from the surface.  The bottom of this slab is fault 2.  that chunk, if you count just the northeastern quadrant, has a mass of about 52,548,131 metric tonnes.  I used an inner and outter boundary of 2.88 km and 6.54 km, the extents apparent from IanF’s plot.  That first slab is about 2.62 km thick.  Slab 2, underlying Slab 1, is about 1.8 km thick and clocks in at about 36,162,125 tonnes mass. Slab 3 – 1.20 km thick and 24,013,900 tonnes. Slab 4 – 1.23 km and 24,578,944 tonnes. Slab 5 – 1.67 km and 33,336,995 tonnes.  Slab 5 is also the closest to the magma chamber, and is holding up the rest of the stack of blocks. (as far as I can tell).  It should be the one directly feeling the effects of a draining chamber.  From the geometry, is appears as if they fit together like a set of books that have been pushed over on their side, with each book resting on the ones underneath it.  Again, how it plays out or works together, that’s for you guys to discuss.  All I can do is point out what was seen in the quake data by IanF and Deepthought.

To get an idea of the energy accumulation of the quakes in the area, KarenZ provides this plot.

Meanwhile, out at the fissure eruption, avalonlightphotoart provides us with a match of Landsat image of 24 Oct and IMO map 7 Oct.

As some of you know, I have this thing for volcanic gases.  Though Holuhraun is far smaller than the 1783 Lakiskagar event, volcanic gases are still something that should be paid attention to.  User scotsfjohn found a nice couple of info graphics about the gas hazards.

861 thoughts on “Ruminarian VII – Mass wasting…

  1. Ok now that DeepThought has let the cat out of the bag, you can see the enigmatic SIL station named ‘12939’ I mentioned earlier:

    Judging by the strength of the signal on that station, wherever it is, it’s *very* close to the action!

    • ok 12939 is the purple which does look to have been close to those quakes – rather than the pale yellow which is VON which seems to be showing more of the Harmonic tremor looking stuff – how much of that is wind noise ??
      if you want to roughly locate 12939 thenlooking for the epicentre for those particular quakes and knowing that Von Urh grf and Vsh also sometimes show the same quake – as smaller/bigger should let you triangulate a bit – in that image at 05.01 von looks closest then 12939 then grf then the others are too far away

      but at 06.15 von and 12939 look to be equidistant, with urh and then vsh being further away.

      I guess you need a map, put a hypothetical dot for 12939 on the map at the epicentre of the 0606 quake and see if that works for the comparing the distances for the other quakes – if not move the dot a bit one way or the other and so on until you narrrow it down.

  2. Hi @ all!
    Short question:
    The amplitude of strain measurements from Hekla and Búrfell were growing bigger since the day before yesterday evening. There has not been any wind/storm at all yet. All was calm that evening with a colorful sunset on the Mila’s. So what could that mean?
    Or are those instruments also influenced by weak wind?

    • I fail to see how the strain meters could be affected by any wing. These are borehole strain meters which literally means that they bore a deep hole, lower the meters into them and cover them over. The only part above ground is the aerial that transmits the information and that sends, not receives. So you see, they are well insulated from wind or rain.

    • Yes, they are considerably affected by the weather – they always get very fuzzy during a storm. IMO had been warning of a storm for two days now. Nothing strange happening there at all.

      • Thanks as always Talla – but how is it affected by the weather when it’s buried? I just can’t get my poor old head around that! :-S

        • To be honest I’ve no idea – but I know that the instruments are incredibly sensitive and that I’ve seen many storms come and go and with it the fuzziness. I think it must be in the same way that the sea and tides affect the seismometers near the coast, also the cars passing on nearby roads. As we’ve all seen the faraway Bardar quakes show up hugely on the strain meters so I’m sure that rain pounding down up above and wind buffeting around would also show up. 🙂

  3. Mopshell had asked for this and I am posting it here so it is not lost way back in the comments. Cross section through the center of the caldera in the region with the dense quakes and running full north to south

    I hope the dots are not too small, if they get too big you lose perspective on how far each quake is from the next one.

    • It is indeed. However, the signatures on the drumplots (see previous page or the one before) was purely tectonic with well-defined P- and S- waves. Since this is a small swarm, and clearly defined in four dimensions (time = 4th), it would be nice to see a rotating high-resolution graph of it which ought to confirm the nature of it. 🙂

      • Pyrite, Krisuvik and the whole Reykjanes peninsula stands at 45º angle at the rifting. So think of it as a in between transform zone and pure rifting zone.

        That’s why we ocasionally see large tectonic events in there, and volcanic activity is more rarer and sparse in time. But Reykjanes seems to exhibit periods of volcanic activity lasting a few centuries and then going back to quietness (or just earthquake activity)

        Actually this swarm belongs to the Brennisteinsfjoll volcanic system. But that’s a minor detail.

  4. Hello everyone.
    Interesting addition to the daily notes of the advisory board:

    Energy of the geothermal areas in Bardarbunga is now few hundred megawatts and the melting of water is
    estimated around 2 cubic meters per. second. The water goes into Skjálfandafljót og Jökulsá
    á Fjöllum. The flow is
    too small to effect the total water flow of the rivers.

    • 80 km2 and 700m gives roughly 50km3 of ice in the caldera. This will melt in 20 years at 2m3/s.
      Suppose that IF this has been going on for the last 2 months. this could account for roughly 6 meters of drop of the top of the ice. Nowhere close to the now estimated 42 m.
      These quick calculations are based on a cylinder shape. But still. the caldera floor must going down clearly.

        • Yes pyrite, another nail on it.

          I still can’t understand why people still suggest the melting scenario rather than the caldera subsidence scenario. I mean, how could just ice melting cause so many M5 events, it’s physically impossible. Also all the drops of 40cm occur at same time as large M5 events. It’s clearly a caldera sinking event. All data points to it. Even GPS data.

          And clearly linked to the large regional tectonic/rifting event occuring. It always does that. In Iceland, rifting is often linked to grabens and subsidence. (as magma drains to erupt at the extremity of such rifting regions). Another examples: Thingvellir in southwest Iceland, Askja in 1875, Krafla in the 1970s. And most Icelandic calderas.

          • We all get bees in our bonnets sometimes, even the best of us. Need I mention Katla and the firm conviction of a certain internationally renowned vulcanologist? Myself, there’s a whole hive inside, bzzzzzzzzzz… 🙂

            • Well, we can at least forgive him.

              Katla clearly had an “event” in July 2011, possibly a VEI1 subglacial eruption.

              It is just not the type of eruption that everyone expected, the “10 times bigger than Eyjafjallajokull” VEI5 event.

              But yeah, in 2011 Katla followed Eyjafjallajokull event, again, within a year or so, just not the way thay all of us expected. So in the end we were all happy.

              His mistake was to advertise it too much. People expected a large Katla eruption and it didn’t happened. Next time, better to say, expect some restless subglacial behavior at Katla, within a year of Eyjafjallajokull eruption. Be a large VEI4+ event, or be a tiny VEI1 event, beneath the ice.

            • Again… there is no demonstrable connection between the two. Leaving that aside, statistically, there is virtually no connection at all. The way it typically goes is that Katla always goes when it’s neighbor does. The problem is, that Katla is always going… any time Eykaf goes, Katla has either just gone, or is already on it’s way to going. It’s akin to walking outside and seeing a bus pull up to the corner. Is it there because you came outside? Or is it there because it’s always coming by at some time during the day?

      • Also 80 km2 and 42m of subsidence gives roughly 2km3 of magma drained out of the caldera (cylinder volume calculation). Guess how much is the total gas released + magma released at Holuhraun + magma intruded in the Dyngjojokull dike? Yes, you guess it, 2km3.

        But yeah, I am not saying that Holuhraun is not receiving magma straight from the mantle., but even magmatic analysis suggested a slight evolved basalt, indicating it originate from Bardarbunga central volcano.

        Even earthquake energy of all these M5 events is somehow similar to displacement of several dozen of meters of a cylinder of 80km2.

        3 more nails in the coffin for the ice melting hypothesis..

        • Yeah it all fits so nicely. Never understood the need for a different hypothesis. I think the coffin is properly nailed shut and it’s time for burial.

        • I do not believe it is melting either but just to keep the volumes correct, this is from IMO Powerpoint and I think it says .75km^3 of sink??

          • Good point. A cylinder 80km2 by 42 meter is 3.3 km3. A cone would be 3 times less so about 1.1 km3. And it seems to subside on a somewhat smaller area as the total caldera area.
            So 1km3 from barda and 1km3 from decompression melt ?

            • This is a really good question 🙂

              Where is all the magma?

              I have read many different takes on where it may be.

              The Rift Stress paper by Gudmunsson says it may be in the cracks and none went into the fissure, but to be fair this paper was written early on and it may not or may be still relevant

              Dike emplacement at Bardarbunga, Iceland, induces unusual stress changes, caldera deformation, and earthquakes Agust Gudmundsson & Nora Lecoeur & Nahid Mohajeri & Thorvaldur Thordarson

              “Magma flows from higher to lower potential energy (e.g.,Gudmundsson 2011), so that chamber expansion and faulting in these parts tend to drag in, or act as sinks for, magma from other parts of the chamber. The resulting magma readjustment in the chamber (assuming little or no inflow),including possible ring-dike formation, could result in some rise of the ring-fault areas, particularly in the breakout parts 1 and 2 (Fig. 4), and a corresponding subsidence of the central part of the caldera.

            • Good point Ian!
              Hard to envisage that all the “missing” magma actually made its way all the 40km and nothing stick to the rims! LIke nice cake batter, just try to take ALL of it away from the bowl… And this is some sticky viscous magma, that has to squeeze through all kinds of holes and crevices?? Like trying to squeeze it through a sponge or brush rather… 🙂

          • What I mean is, the visual observations of these ice cauldrons began on Oct 10 – there have been at least another three or four since to take measurements of the cauldrons.

  5. Thanks Ian.
    The nearly vertical sidewall of the plug/segment in the south vs. the outward dipping wall to the north is quite evident. I am also intrigued by the horizontal width of the fracture zones…about 0.8km wide in the south and about 2km wide in the north. However, not sure how real is this is since we are looking what essentially is an arc of hypocenters viewed head-on.

    Lastly, to aid us ultra-visual types, could you present the scale of the x-axis (longitude) relative in size to distance in km? From Google Earth, it looks like longitude 17.4 – 17.5 is about 8km? That would make each x-axis division 0.8 as long as each y-axis division (1km)? i.e. a 0.8 : 1 box. Thanks a lot if you could accommodate?

    • The latest plot is a way takes out the “arc effect” since it is not plotted by latitude but by distance from the GPS position and the scale is in kms, I know it is hard to see but they are 1×2 km boxes 🙂

      • Thanks for the clarification Ian! So, in summary, the data is suggesting the width of the fracture zone in the north is about 4 km and the width in the south is about 2 km?

        • Yes that is correct for this strip of the sample which is from -17.4 to -17.5 longitude which is about 2.2 km wide strip running north to south

          Also interesting is the west side, I did a cross section running east(right side of plot) to west through the center and that cross section, it is very wide and the trend line is flat like the south

          Of course there is no reason to plot the east, the lack of quakes is shouting out to us but I cannot tell what it is saying. However, the comment in the paper you posted earlier may be a hint??? Is the fact the east is inward dipping a reason for the lack of quakes? Or, is it due to rock being ductile? Is it due to the lack of stress from rifting?


          Insights into the kinematics of a volcanic caldera drop: Probabilistic finite-source
          inversion of the 1996 Bárdarbunga, Iceland, earthquake
          Andreas Fichtner a,⁎, Hrvoje Tkalčić b
          a Department of Earth Sciences, Utrecht University, Utrecht, The Netherlands
          b Research School of Earth Sciences, The Australian National University, Canberra, Australia

          “The inclination angle of the subfault segments is less well
          constrained than the rupture times and relative magnitudes. There is
          nevertheless a clear preference for outward-dipping segments in the
          western and inward-dipping or vertical segments in the eastern part of
          the ring fault. Our results indicate that the geometry of the fault is more
          complex than a simple outward-dipping cone. Since the rupture is likely
          to have occurred on a pre-existing ring fault, its current shape reflects
          the conditions at the time of its formation. Factors contributing to the
          ring fault irregularity include an irregularly shaped magma chamber, an
          anisotropic stress field or structural heterogeneities.”

  6. Scotsjohn, you suggested a connection, that however controversial, its well worth to think about.

    Many quakes over last year in the MAR well outside Iceland.

    It would be interesting to see a connection between increase of activity in Iceland and across the MAR in spots well far outside of Iceland, as if the plates would be moving more actively and the release of tension would be seen all along the MAR.

    I think we have no data to either confirm or reject this hypothesis. But we can have a hint.

    Why? Because only in last 2-4 years has Iceland seen a significant increase in volcanic activity due to the hotspot peaking in its 130 year cycle.

    So if this increase is happening (as we know it is, because we see now a large regional rift event as we (and I) often predicted in this blog for the next two decades ahead). You just have to go back in history to see how intense can these peaks of activity develop in Iceland.

    And so we can check whether activity along the MAR, well outside Iceland, if also increases at the same time, to a degree not seen in 130 years (however our data for the MAR doesn’t go that in time).

    We still have the Azores and other volcanic islands to compare.

    This is just a wild hypothesis, very wild, but it would change the books if confirmed.

    We have seen a some interesting connections in the Pacific, like many strong quakes happening at same time in the Pacific in the 1960s and also in recent years.

    I wouldn’t be surprises if the causal link wound’t be a direct one, but rather one of chaotic behavior, like a large quake in a spot in the plate might cause a large event elsewhere in the plate a few ears down the road, but basically not possible to predict with accuracy.

    The same kind of thing seems to happen within Iceland.
    We do not know clearly where the next large event might be. We could suggest Reykjanes, Edlgjá, southeast of Langjokull, Hamarinn or Thordarhyma, even large quake in Tjornes, but sincerely its hard to predict. But we have seen many signs pointing to this event at Holuhraun before it happened. Carl was often bold in suggesting it, but always keeping it between the lines. The best data we have is the Icelandic historical record. We know our domino-style things develop in Iceland when such large regional events occur.

    Just check how many sequence of large events happened for instance in 1860-1900 period, 1710-1730, 870-940 or 1780-1790. It’s almost certain something like that will happen in 2010-2030. I often said it.

    • Let me say also that often it takes a huge effort to make everyone agree when we suggest a bold new theory. For instance, when the tectonic plate theory (continental drift) was suggested a few decades ago (yes, 99 years ago) it was made ridicule, even when you see that the continental shapes and boundaries fit all with each other like a puzzle. Or even with fossil data that couldn’t be explained otherwise. So, we should expect the same when we go bold with our theories. We can expect the ridicule from others, and it often takes a few decades for such a theory be accepted at a mainstream level. Well, the history of science is FULL of such examples.

      Now, back to the connection Iceland-rest of the MAR.
      We know that Iceland had volcanic peaks (or large regional events) around year 900, 1240, 1480, 1620, 1720 and 1780, 1870, 2014.

      Historical data on Azores:
      Picos, at Sao Miguel island, circa 940, 1652
      Sete Cidades, at Sao Miguel island, VEI3 at circa 950, VEI4 at 1444, 1638, 1682, 1713, 1811, 1861, 1880
      Agua de Pau volcano (Sao Miguel island), VEI5 eruption in 1563 (no connection to peaks above)
      Pico volcano, 1562 and 1718
      Sao Jorge, VEI3 at 1580, 1757, 1808, 1902
      Terceira, circa 920, 1400, 1761, VEI3 at 1867, 1998
      Fayal volcano, 1672 eruption and 1957 (no connection to peaks above)
      Furnas volcano, approximate dates circa 840, 1170, 1430, 1441, a VEI5 on 1630

      Perhaps several eruptions in the Azores seem to occur at same time as Icelandic peaks (like around year 900, 1710s, 1860s, but data is scarse and uncertain, can’t make any conclusion out of it, could just be my act of cherry-picking

      However more interestingly, is that in the period circa 1563-1672, many large eruptions occured in several of the Azores islands, indicating a peak of volcanic activity. Data seems to provide a more confortable suggestion for it.

      Tried data for other Atlantic volcanoes but data is scarse or not conclusive.

  7. On the page that DeepThough linked to at 07.20ish this morning, there is an image top right of river levels and temperature. (I can’t understand the Icelandic for the middle one plot.) The water temperature seems to be rising more than recent diurnal patterns account for, and the level seems to be rising too.

    Also, the plot is for Hverfisfljót and Skaftá, I think, and these flow out of the SW side of Vatnajðkull, from Síðujökull and Tungaárjökull respectively.


  8. Does anyone know how this geothermal energy gets measured inside Bardarbunga?

    A few hundred megawatts increase on the scale of Bardarbunga doesn’t sound very spectecular yet ( or is it?), but what does it indicate?
    And maybe more important, what it implicates?
    What are possible reasons for an increase in geothermal energy inside a volcano?

    • I live around 30 miles from you in W.Sussex and i can confirm how warm it was today. I had the sense to go out in a T-shirt today, husband wore a sweatshirt. Oh was he upset he hadn’t listened to me. 😀 I think max here on the coast was 22C but still absolutely incredible. I have known some summers when it never got this warm through the whole summer. Nice way to say goodbye to October though.

        • Here it’s still October 31st. It’s currently at the high for the day 39F (3.89C). Wish it was that warm here for trick or treating. It has been in the past at times. We have chance of flurries and suppose to get down to 31F (-0.56C). Brrrrrrr Almost time for the kids to start about 2 hrs. at 6pm EDT. I’m going to my daughter’s to help out giving candy. There’s someone that rides thru the town on a horse as the headless horseman. 🙂

    • The record was first broken at my local weather station, Charlwood (near Gatwick Airport) at 12.20pm with 22.5 deg C. An hour later it was surpassed by Gravesend, Kent with 23.5.

  9. There have been some dramatic drops on the BB GPS recently followed by equally dramatic rises and then a more gentle subsidence again. I think someone else here described this as Bardy “breathing”. If the drops were caused by the earthquake activity and the sudden rising by re- inflation as magma flows into the chamber until the increased pressure forces it out along the dyke, would we expect to see this pulsing mirrored in the rate of lava outflow at the eruption?
    Is there any evidence of this?

  10. So that comment from the IMO about a few hundred MW of geothermal energy is too small to be extra geothermal energy. Simply converting the potential energy of a mass of ice that large by lowering it yields more energy output. That ice may be melting simply due to that conversion of potential energy and not by any rising heat from the volcano.

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s