Predicting Volcanic Earthquakes & Eruptions?

Thousands of volcanic experts & enthusiastic amateurs follow volcanos, one of the fundamental questions in modern volcanology is the manner in which a volcanic eruption is triggered; via the intrusion of fresh magma into a reservoir, the amount by which previously pooled reservoir magma interacts with a newly intruded magma will determine the nature and rate of eruption, as well as the chemistry of erupted and ground deformation to changes in volume of a magma reservoir.

Quake frequency evolves as a volcano develops from a state of dormancy through unrest into a state of explosive activity. Seismic activity originates mainly from mass movements inside the structure, such as magma intrusion and gas release, which produces stress variations and ground deformations. There is thus a relationship between seismic energy release, deformation, and volume change (McGarr, 1976; Yokoyama, 1988).

Analysis of the Seismic Activity  Associated with the 2010 Eruption of Merapi Volcano, Java  by A Budi-Santoso – ‎2013

The accelerating seismic activity is also related to the accelerating deformation of part of the summit and both phenomena can be considered as precursory signs of a large explosive eruption.

The main types of signal are volcano/tectonic (VT), multiphase (MP), low-frequency (LF), rockfall (RF), and tremor.


So from the above informative Earthquake event chart of the  2010 Eruption of Merapi Volcano, Java ………

DeepThought’s   ” Volcanic Earthquake & Eruption Predictor* “     * Predictions may become reality
Download the xlsx worksheet or Download the xls worksheet      to fully understand this interactive VolcanoCafe article.
Be able make your own predictions for this unique event at Bárðarbunga, plus other volcanic events.

Using IMO Earthquake data since 18 August 2014,


Go to the Energy sheet & paste special option “values” in cell Q4, todays EQ (delete any excess entry’s from yesterdays)

First up we convert the daily Earthquake work done to  KKI TNT equivalent,
look at TNT equivalent scale to the right, and transpose correct TNT value into column P for each EQ

Kelly Kiloton Index of Earthquake Moment Magnitudes
The Kelly Kiloton Index (KKI), aims at giving a “realistic” picture of earthquake energy. It uses the kiloton (= 1000 metric tons = 2,200,000 lbs) as the basic unit. Based on the seismic “moment of force,” translated to the equivalent energy released by an explosion of TNT, every positional increase, goes up by a factor 32. Thus, R 2.0 corresponds to the detonation of 1 ton of TNT, R 3.0 = 32t, R 4.0 = 1000t, R 5.0 = 32,000t, 6.0 = 1,000,000t,

Now for some eye candy

50 tons of TNT

Operation Blowdown was a joint UK-US-Australian operation at Iron Range in northern Queensland, Australia 18/07/1963. The test was conducted to simulate the effects of a small nuclear detonation over a rain forest. It involved the detonation of 50 tons of TNT atop a 43 meter tower. The explosion turned the surrounding forest into a twisted knot making troop movements almost impossible.

500 tons of TNT

Operation Sailorhat was a series of three 500-ton conventional explosive shots in near Hawaii 06/02/1965 designed to test the blast resistance of Navy ships in the advent of a tactical nuclear strike. Each “Sailor Hat” test consisted of a dome stacked 500-ton (450 t) charge of TNT high explosive detonated on the shore of Kahoʻolawe close to the ships under test and each test saw the USS Atlanta move closer to the explosion.

So far at Bárðarbunga 4,118,105 tons of TNT equivalent energy has been released underground !!!

5,000,000 tons of TNT underground thermonuclear test

Cannikin was detonated at a depth of 1,800 m underground on November 6, Amchitka island, Alaska, 1971. The announced yield was 5 megatons (21 PJ) — the largest underground nuclear test in US history. The ground lifted 20 feet (6 m), caused by an explosive force almost 385 times the power of the Hiroshima bomb. Subsidence and faulting at the site created a new lake, over a mile wide. The explosion caused a seismic shock of 7.0 on the Richter scale, causing rockfalls and turf slides of a total of 35,000 square feet (3,300 m2). 38 hours after the explosion a 1 1/2  mile wide by 60 foot deep subsidence crater formed  as the volume of material above the cavity  (the chimney) collapsed into the void

OK……… back to the the Energy xls sheet a   M7.0 = 32,000,000 of TNT on the KKI scale for EQ work done

Transfer cell P52  value to column C at todays date (5/10/2014)

Transfer Total # Quakes to column E at todays date (5/10/2014)  Note *current chart lists the # of all EQ >M 0.2 at BB

Transfer cell L105  value to column K at todays date (5/10/2014)

We’re all done   ….scroll up back to row 1

This chart shows the daily TNT Equivalent Earthquake energy release at Bárðarbunga (5/10/2014 is your data inputs),
Discussion:  note the # 5% or 200,000 Tons days, so this could be taken as a average “maximum daily release”,
the “middle daily release” seems to be merging with the “minimum daily release”, also the expended energy level has been rising for the last 5 days.
We keep an eye on all this & use some analytical grey matter, question, is released energy rising or falling?,

& magically  out spits a number of answers / conclusions with out even having to interpret & comprehend some weird formula.   Winking smile

This next chart shows the daily TNT Equivalent Earthquake “forecast & remaining/unleashed pent up energy”  at Bárðarbunga (5/10/2014 & 6/10/2014 is your data inputs),

The ” Volcanic Earthquake & Eruption Predictor” models concept is to have a # Quakes & achieve a “zero value” on the brown-yellow bar i.e. “an equilibrium state”.

The blue–red bar is the next days forecast EQ energy to expend, which is derived from yesterdays forecast (blue–red bar),  less EQ that happened yesterday to 0.00hr – 24.00hr (green bars in the first chart),  plus todays anticipated energy release. (the white line)

So while both bars are above “zero value” there is “X energy” still to be released…. i.e. more or BIG or Lots of quakes to come.

[A] Think of a bucket on the end of long stick, hose water goes in & you have to bail it out with a cup, bail it out too slow an the stick bends(tension), bail faster to reduce tension (small quakes), or fall behind & the extra waters weight eventually breaks the stick (a BIG quake) &”an equilibrium state” now exists ……… go back to [A]

So……….WTF…… yet another uncomprehendable even weirder formula, some mathematician probably spent days on & got quite tumescent over. Smile


?????………..  OK…..lets just grab a drink of your favourite poison & mull this over & use some more analytical grey matter,  magically  out spits quite a number of answers / conclusions which we auto arrange into improbable and likely scenarios Winking smile

So we’ll re put this chart again conveniently here……. instead at the “end of the paper” and refer endlessly to it (cause were so smart, we don’t need picy’s).

The models concept is to have a # Quakes & achieve a “zero value” on the brown-yellow bar i.e. “an equilibrium state”. at the end of each day.

So we adjust the white line forward “expected EQ energy release” in the Energy sheet column M to have almost no negative values ….OK  we underestimated in late August as  Bárðarbunga had a “hole lot’a a shak’in go’in on”. Smile
Discussion:  note the 250,000+ Tons days, so this could be taken as a average “maximum daily forecast release”, before  BIG or Lots of quakes happen en mass and not the end of day brown-yellow bar falls close to a near or zero “an equilibrium state”,  on the 15,16,21,22,29,30 so this infers/means the white line forward “expected EQ energy release” is set about right for now.
If the projected “unreleased energy” reaches  close to 375,000+ TNT and EQ are still happening but at a reduced rate we would go back to say the 1/10/2014 and adjust the white line value down for each day, to have on the next biggest EQ day  achieve a close to “zero value” on the brown-yellow bar i.e. “an equilibrium state”.  Theres a few more self explanatory graphs within the ” Volcanic Earthquake & Eruption Predictor”

Use all info & data you obtain from everywhere, use some more analytical grey matter, it matters not if your out, nobody gonna die, or be evacuated,
but if you get it right ………… you’ll be   “the one”     of the few who has predicted Earthquakes………….
Enjoy,   DeepThought

630 thoughts on “Predicting Volcanic Earthquakes & Eruptions?

    • Which Game? JayBee.

      Their own description says the lava measurements are from the 4th, and the modis image from this morning. When there was full overcast across the site.

      • The heat image shows that there is little movement on the long soits. as soon as they solidify there is no turning back. The lava movement along those paths is no longer possible until after other “lands have been concured”. The activety is heightening the field and pushing south. We would need to see double the of lava in order to fence in the river.

      • The modis images may well be useful to them. Even though the ground level webcams show us nothing, the satellite views can be illuminating. I’ve only bothered with posting final, heavily-processed Modis images (ones that fit into a whole Arctic mosaic). Today’s show small clear patches amongst the heavy cloud cover. Nothing useful for our purposes. But there are actually many different raw images that are sorted in NASA’s algorithms, and those raw images are available. Look at enough raw images with bits of clear area, and you can, indeed, draw conclusions about what’s happening on the ground.

    • i wouldnt trust your statements about the tremor plots given how you presented them to the viewers here before, the lines in the middle are just the time the station was offline and the data is missing, i see nothing of interest there now in the end either.

      Lugn, inget att spräcka ett blodkärl över…

      • I think it was just generally a humorous comment on the fact that IMO kicked the SIL-station into action at the same time the earthquake took place.

        • haha, sry my carl and rat, my joke detecting skills needs some polishing :p i just saw desert rat posting tremor plots again and was too fast to make a reply hehe 😛

      • Hi burre01, I take it DR meant it literally… the nuts and bolts of the station, or the needles and pins in its inner workings got shaken loose 🙂

    • also this station is located near Reykjavík and not really of any interest regarding the state of bardarbunga right now.

      Lugn, inget att spräcka ett blodkärl över…

  1. Hmmm … spooky … very, very few EQs since the 5.5 this morning … are we supposed to be holding our breath or letting out a gentle sigh of relief I wonder 🙂

        • Regardless of any of the volcanology, if a mass 10km by 7km by possibly 4-8 km deep attains any speed, the inertia would incredible. We are talking an asteroid sized ‘plug’ here and from all I’ve read it is moving as one, not in pieces.

          If someone did the maths it would be possible to calculate its rough weight, and the maximum speed it has thus far achieved in one of its ‘drops’, then you would get an idea of its kinetic energy. One thing for certain is whatever lies below the plug is taking one mother of a pounding!

          • And here is the approximate answer you are looking for…
            1 455 300 000 000 ton. At the surface of the ice the M5.5 earthquake caused a 40cm drop.
            This then gives the Kinetic Energy by using KE = 0.5 x m x v2 and that would be cirka 1 455 300 000 000 Joules. That would be the equivalent of 350 kiloton or equivalent of a 6.9 earthquake. The last part is a bit baffling. But, we should remember that the functions for calculating the shear from an earthquake are not really made to take into account planetoid sized cylinders dropping… But, how exactly that happens I think I will leave to Albert… after all, planetoids moving ponderously in slow motion inside terra firma is not my subject…

            • 🙂 I wasn’t going to attempt it since I don’t know the amount of ice/water in the ‘cauldron’, nor the exact depth of rock in the plug or its density. I thought someone might know 😉

              That’s quite an object to try and stop if it gets going, I suspect if it does then the frictional and stress factors will fracture it, then what is the next question?

            • Moving ponderously on a leash gravitic
              the plug-sized planetoid
              as measured on a scale explitic
              1.5 teratons through the void
              25 kay-emm-per-ess relativistic
              455 ZetaJoules as it Earth destroyed

            • If we make that 400 cubic kilometre, it would weigh a million billion kilogram. Which is better written as 10^15kg or 10^12 ton (Carl’s number). Total energy if you drop the whole thing by 40cm is m*g*h = 4 10^15 Joule (not the same as Carl’s number). That is 360 times more than the energy in the M5.5 quake this morning and is about equal to a M7.2 (again close to Carl’s number). But if the energy is converted into heat, the rock would heat up by a milli-degree centigrade. There is no problem storing the excess energy.

              The speed it achieves in the drops may also disappoint you. The fastest you would expect is free fall which after 40cm is about 3 m/s. In reality it will be slower. Asteroids come in 10 thousand times faster.

              The volume you lose, assuming 40cm per day, comes to 0.03 cubic kilometre per day. Which over the 30 day eruption is – let’s see – 0.85 cubic kilometre. Fairly close to the total estimated volume of the fissure eruption.

            • My numbers was slightly approximated and I used the duration time of the earthquake as the time for the drop. So it was really slow.

              Approximation from me being lazy and doing it in my head.

            • I liked the idea of storing the excess energy as heat. I wonder if that would not cause friction melt and that it could by now work as lube for the plug.

            • I don’t get that. OK, m=1.5 x 10^15 kg, that’s an energy of E=mgh approx 6 x 10^15 J. That’s a moment magnitude of 2/3*log(E) – 6, approximately 4.5. There’s far less energy in the falling stone than in the EQ. This means that some stress is released as well?

            • Is that right?


              10km diam (I assume was given) x 6km depth.
              These are order of magnitude figs so that’s 500 km^3 or 5×10^11m^3

              For simplicity (some is ice anyway) lets give density of 2000kg/m^3
              that’ll be 10^15kg, dropping 0.04m (using mgh) thats about 400×10^12J.

              That’s quite a lot….
              Really quite a lot!

              I’m not sure where you used E=1/2mv^2 but a body dropping 0.04m would reach a speed of

              Hmmm (thinks back 50 years) v^2 = u^2 +2aS so here v^2 = 20 x 0.04 m/s = 0.8

              v= 0.9 call it 1 m/s.

              Hmm, check, 1/2mv^2 = 0.5×10^15 (see above) = 500×10^12J pretty close!

              Unless I have (as usual) slipped a few magnitudes…

            • What if the plug is smaller than that? Say, it’s more conically-shaped and only 2.5km deep…?

      • There’s almost 60 cm of drop locked in today (Tuesday), not counting the past 3 hours. Wonder if this might be a record?

    • Oh no, really? Look, I’ve had words with them. The earthquakes in the rift zone can propagate there all they like as long as they keep off the ice! If I’ve told them once, I’ve told them a dozen times. :<

  2. As a non expert reading an expert:
    thru Giggle translate.
    I understand that another reason for the difference in volume between BB plug drop and volume at the dike is the pressure at the vent. As the lava has lower pressure when it surfaces, more gas is formed and so the volume is expanded, but the mass is the same.

    • I wrote 350m3/s lava equivalent, I had allready translated the values so that they are the same at both ends to take into account such effects.

      • Couldn’t the lowering of BB due to a venturi effect as the lower level lava flows past a constriction? This would have the effect of sucking the lava out of BB at the same time as having an increased flow in the dike.

  3. some teachers resources for Iceland

    Volcanologist: The Coolest Jobs on the Planet

    Hugh Tuffen @HTuffen · 21h 21 hours ago … Great to rediscover a classic paper about crystallisation of rhyolites from Swanson. Relevant to many ongoing studies

    Spica might already have this in the library
    Relation of nucleation and crystal-growth rate to the development of granitic texture

    anyone have #elenskya twitter account?
    She is installing some new hydro stations near Katla or Hekla i think…

  4. The deep M 2½ quakes noted under Bardarbunga a couple of days ago (~18km) have been followed up by a few at ~12 km depth including the M 3.4 at 11.2 km 7.0 km ESE of Bárðarbunga and the M 4.2 at 12.4 km 4.5 km NNE of Bárðarbunga today. The question is if any of these are connect to or if all are independent of those earlier, deep quakes. Since we have had an unusually large drop today, maybe there was a connection.

  5. And a comment on Carl’s long comment which he claims was a response to my comment (but I think he wanted to write that comment anyway). The chamber is indeed not overpressured. If it was, the overlying rocks would not respond to a draining of the chamber and we wold not see the deflation. The whole thing is in equilibrium, with the magma holding up the heavier rock above. You would not expect the magma to be driven upward in such a situation especially when the reservoir is being drained.

    But in reality the situation is a bit more complex. Magma is lighter than the rock above. It would happily move up, letting the rock drop. This does not happen (I assume) because magma is not that liquid: it is a mix where only a small fraction of the rock is molten. Also, the rock above is rock solid. But the melt fraction may change, for instance if temperature goes up (new magma convecting up) or pressure drops. So I would not exclude that some magma does manage to rise in some locations from 8km to 6km in response to the earthquakes. Earthquakes can reduce pressure and tension.

    But note that we see this earthquake gap at 5.5-6km only along the northern rim. On the southeast side there are some earthquakes at this depth. If this magma layer exists it could be a local thing only.

    In the end, data will decide.

    • I must plead guilty on that one.

      Actually your comment about energy potentially being stored as heat gave me an idea. The bulk of the heat would for obvious reasons stay along the edges of the “plug” (we do need a more scientific sounding name for it) potentially causing melt to occur. That could form a conduit were buoyant magma could come upwards.

      • @Carl –I am going to stick my neck out and suggest that your long comment does really deserve to be a post. I have waited a while to see a convincing explanation of what is going on, it was gratifying to see your comment. I felt lucky to have caught it. It would be a shame to leave it buried in comments. It is begging for an illustration however. I would volunteer but I am a notoriously poor artist. Maybe add some EQ postions?

      • OK I know I am thick, but reading Scotsfjohn and Carl’s earlier comments above on Bardar’s kettle lid slip-sliding down slowly generating earthquakes reminded me that there is science detailing the lubrication of rock at fault line interfaces. Where heat and pressure stress building at the fault interfaces reaches a critical point that causes the contact surfaces momentarily to liquidise and allow short bursts of movement before cooling and re-frictioning. This mechanism suggests Bardar’s plug is unlikely to go into “freefall” and we’ll continue to see earthquakes until as such a time magma finds its way through (i.e. due to lid fracture) or the process ‘grinds’ to a halt. John Spray’s article “Evidence for melt lubrication during large earthquakes” kind of is in the right direction, but I’ve seen a much more recent paper that follows up on this and damned if I can find it now.

        I know my comment is a bit half-witted but I thought it might contribute to the discussion. 🙂

        • not half-witted at all. This whole question of coupling has not been properly addressed here IMO. In a normal fault regime we would normally assume, given the sizes of the quakes and their number, that a pretty huge surface has decoupled and slipped. Given that the energy required to decouple a fault plane depends on a huge range of factors, including the amount of water, heat, the chemical bonding involved, existing fault structures, etc. it is kind of hard to make any meaningful comment. But the fact that this is an existing caldera that has obviously seen serious subsidence in the past makes me a bit surprised to see such a huge amount of seismic energy getting released to break these bonds, i.e. this makes me lean more in favor of Gudmundsson’s idea that the caldera is getting squeezed by wider regional forces induced by the dike and is not just an expression of a falling plug into an evacuating chamber.

        • It seems like sometimes we talk as if chamber walls are vertical and smooth ie cylinder-like. More likely they are not smooth and have somewhat smooth protuberances and also have waviness, or snake-like configuration and variable diameter thus creating more or less friction resistance to downward movement of the chapeau or plug at different depths. However, unlike ice, the solid plug should not give compressional melting since the solid form of the magma occupies less volume than the liquid form. This is opposite of the case with the water/ice relationship. If the diameter of the chamber becomes too large as the plug moves downward, liquid magma could move around the plug and move upward and the plug downward but probably at a slow and rather controlled rate due to viscosity consideration and time it takes for liquid to move around the plug.

          So the question in my mind is what is the mechanism of overcoming the enormous friction of plug chamber wall contact. It is likely that the chamber walls are slightly mushy due to contact with hot liquid or partially liquid magma. The presence of some liquid between mush grains of the wall can allow the grains to move and let the plug continue downward in jerky movements like we are seeing on the seismograph readings.

          Of course in the case where the magma level lowers and creates a significant void between plug and liquid magma the movement may be anything but smooth and steady.

          Just thinking out loud. Probably this has been said above and I am just repeating common knowledge or I am just wrong.

    • The presence of eruptible melt is kind of conspicuous by its absence, isn’t it? I mean, I can’t think of a comparable instance where we have had such a huge volume of major seismic activity that ranges all the way from the assumed depth of a proven large magma chamber up to the surface in a volcanic environment and at the same time NO surface manifestations whatsoever (disregarding the cauldrons under the radial dike early on). It’s like putting a casserole on the fire and coming back an hour later and lifting the lid only to discover it hasn’t even started simmering yet someone has come along and eaten it already. This makes me think the only comparisons worthy of the name are Pu’u ‘O at Kilauea and Piton de Reunion where a caldera has its guts drained out of it before our eyes:

      or Dolomieu before:



      Question: why is Bardarbunga so noisy when it falls when these examples are pretty silent jobs?

        • Fernandina is an interesting comparison in that it features both radial fissures draining the main chamber (like Holuhraun) as well as explosive volcanism from the main vent. GVP says the caldera in its current form was created by an explosive eruption so in that sense it is not the same as Bardarbunga, which is being formed by passive draining, but Bardarbunga might still surprise us yet. There might be quite a few parallels between the two but I don’t know enough about Fernandina to say… hmm.. where’s Boris when you need him? He’d know.

          • Reason I mentioned it: although the 1968 collapse was ‘explosive’ in that there was an enormous dust/ash cloud, the subsidence of the caldera floor -several hundred metres- was so, well, gentle, that a cinder cone on the down-dropped floor was essentially undamaged: GVP quote:

            The central cone besides having fallen some 250 m in altitude appears to be further tilted towards the SE. Its height however (approximately 115 m above the surface of the lake) has not altered significantly. It is covered to a varying extent with a mantle of gray-black ash.

            But yes, we DO need Boris!

            • oh cool, I missed that. That must have been a pretty gentle explosive eruption then! But yeah, looks exactly like the same kind of phenomenon.

      • The associated rifting caused caldera deformation, maybe? Round plug dropping in what is now a less than round hole.

        Can we really consider the “plug” a consolidated relatively rigid mass?

        • I agree. Probably a combination of both. Deformation and the fact that the plug is soooo big it is kind of coupled with the caldera wall. The other thing that I think might be happening is that the 10 km chamber is in the process of moving northwards (which explains the skewed plug we were talking about) and the current activity is in response to the wall trying to straighten itself.

          • I really am fascinated with the amazing picture of the drained caldera.

            I wonder what the structure surrounding the BB caldera looks like?

            Anyone care to speculate, it must be starting to look like gravel

            • I agree, the plug must be starting to get pretty fractured by now, which is another explanation of the lack of quakes there.

              This is also why I am a bit wary of equating the release of seismic energy with the kinetic energy of the falling block (plug). Seismic energy represents a release of pent up strain which would imply the plug must still be coupled with the surrounding wall and I am not so sure that is the case. The longer term chart looks more like the plug is just sitting on a wad of magma and is falling in steady state.

            • My view (could be wrong) is that the quakes are from rifting. The caldera subsidence may be caused directly or indirectly by this: directly – rifting gave magma from beneath the caldera a route out to the north; or, indirectly, (actually this might just be a different directly), rifting weakened support for the caldera so it is sinking. The scenarios are not mutually exclusive – we could be seeing a combination.

    • interesting. 6 to 11 km depth. Is this a delayed response to the “near-earth asteroid” hitting the main chamber and shunting a pressure wave through the system or signs of further rifting unrelated to magma?

    • There’s still the option of the rift opening up further back along that route, under the ice, isn’t there?

      Also there’s a small cluster of quakes at the very north of the TFZ, just to show us that other Icelandic options are available. 🙂

  6. I’ve noticed recently that some quakes are coming in twos or threes, pretty much at the same time and quite close to each other – is there any particular significance to this?

    From IMO’s table:
    Tuesday 07.10.2014 18:46:11 64.676 -17.450 8.0 km 4.2 99.0 5.5 km NE of Bárðarbunga
    Tuesday 07.10.2014 18:46:11 64.669 -17.390 1.1 km 1.8 90.13 7.3 km ENE of Bárðarbunga
    Tuesday 07.10.2014 18:45:55 64.680 -17.460 2.5 km 3.5 99.0 5.4 km NE of Bárðarbunga
    Tuesday 07.10.2014 18:45:03 64.676 -17.489 5.1 km 3.7 99.0 4.4 km NNE of Bárðarbunga

    • Maybe an analogue to the phenomenon of ‘aftershocks’ from tectonic quakes – a quake jolts the area, just enough to trigger movement from another nearby location where strain had been building up

  7. It’s amazing how the gps graphics are developing. The last M5 seem to have had a serious fall as result. The M5 earlier seem to heave had a steady (nicely curved shape!) rise as result. There was discussion earlier about how to read (or not read) these graphics. But following the moves of the blue line is a nice thing to do. 🙂 And leaves you with many questions 😀 .

    • aagghhh.. and as so often.. here we have almost direct evidence that everything I just said above must be wrong. The correlation between subsidence and big quakes is just too glaring:

      stuck, stuck stuck, bang, slide, slide slide, sticking sticking, bang slide (with a couple of odd bangs thrown in for good measure) 😉

      ok, time for bed. sweet dreams of sheep everyone.

  8. Download DeepThought’s ” Volcanic Earthquake & Eruption Predictor V2”
    I’ve improved & added new features and corrected a addition error in Col J
    (make sure you copy the whole url less the xxx)
    Download the xlsx worksheet xxx
    Download the xls worksheet xxx

  9. Pingback: Predicting Volcanic Earthquakes & Eruptions? | Explosive Earth

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 )

Facebook photo

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

Connecting to %s