Magma properties and Lava Fragmentation – in simple words


Scanning Electron Microsopic (SEM) image of highly fragmented ash from the explosive Kelud 2014 eruption. Image by Spica, AEC.


One of the most evident volcanic manifestation are eruptions. There are different types, which depend mainly on the type of magma produced by the volcano. Over the times one same volcano can produce widely different magma/lava types. We will try to give some insight on magma formation and type of lavas. We will try to explain also the mechanisms of magma fragmentation which leads to the production of lavas and tephras.

To write this article the authors have widely used the materials of a recent Coursera MOOC, “Volcanic eruptions a material science” which have been provided by Prof Dingwell who is a very respected member of the academic community.

To define things more precisely, magma is a mixture of molten rock, dissolved gases and sometimes crystals, held at depth, before the eruption. Once the material is out, it is lava.

Magma is produced through different mechanisms. The most widely known are tectonic plate subduction and hotspots. As the geological settings are widely different, the magmas produced are usually very different from a chemical point of view.


Viscosity value of some substances

One of the main classification of eruptive material refers to the silica content versus the total.

Silica (Si02), is one of the main constituent of sand and can be found under different chemical forms crystalline or not (in that case it is a glass, which is a solid having a non ordered structure contrary to a crystal which shows always the same pattern in the position of the different atoms).

One of the physical characteristics of lavas is viscosity which is the resistance to flowing shown by a material. Water at normal temperature has a low viscosity (ie it flows easily). On the contrary toothpaste does not flow so easily so its viscosity value is higher).

To see the effects of viscosity, you can check this little animation from university fo Rhode island

Lava viscosity is mainly a function of temperature (usually a higher temperature means a lower viscosity) and chemical composition.

On this picture you can see the differences in composition, temperature and viscosity depending of the lava type.
Note that there is a relationship between composition and eruption temperature.


Silica content plays a major role as lavas poor in silica like basalt are fluids lavas and on the contrary lavas high in silica are much more viscous like andesite. This has a tremendous influence on the behaviour of the volcanoes during eruption.

As said before lavas come from magma which are deep under the volcano (usually several kilometers) in a magma chamber. In this chamber the magma is submitted to very high temperature (about 1000 °C or more) and enormous pressure too (typically several hundred bar, one bar being the atmosphere pressure at sea level).

From a chemical point of view, this has many implications.

In the chemical world, reactions happen all the time. A chemical reaction means that several molecules (a group of atoms) or single atoms are going to “do something” together to produce another chemical. It is, for example what happens in your car when you start the motor : hydrocarbons (made of carbon and hydrogen) react with oxygen from air (with a spark, under pressure) to produce carbon dioxyde, water vapor….and motion and heat.

Depending on the pressure and temperature conditions, the reactions vary and can produce different chemicals. One good example is pure carbon, which can be in the form of graphite, a very soft mineral (used in pencils, yes, the one you use), or diamond, the hardest natural one on earth (and the girls’ best friend).

So, a chemical reaction is a little like a vacation trip. You can go from Paris to Marseilles in different ways (train, plane, foot, car, going through Lyons or going through Clermont Ferrand). The big difference with a chemical reaction is that the Marseilles you’ll get to will not be necessarily the same depending on the followed path. So, with the same chemical mix, depending on the reaction conditions, you can obtain different results, or products, or a mixture of different products.

This being said, let’s apply it to magma. It was resting nicely in the magma chamber and then its surrounding conditions (temp and pressure) changed (why is the subject of another post). So as it begans to move up, the chemical equilibrium is no more and the chemical reactions will begin. Its composition will change as the depart from the equilibrium will produce new mineral species (solid crystals), and gas bubbles.

The multiple path explained before takes its sense now. Depending on the conditions (pressure gradient, temperature gradient and reaction time), the results can be very different in terms of produced lava type, even with a similar initial composition. Chemical reactions do take some time (it’s called kinetics – things always sound better with a greek root) from some part of a millisecond to several centuries. As the path is different, the result will also be.


And this leads us to the post’s title second part: fragmentation.

First a definition: Fragmentation is where the flow of magma undergoes a transition from a laminar flow of viscous bubbly liquid to a turbulent flow of gas carrying liquid fragments.

Laminar flow, when used for a fluid means that it is moving quite slowly and that there is no overall mixing. You can compare that to several layers of the same fluid, moving along in the same pipe, but not mixing.

Turbulent flow implies high velocity and a lot of mixing.

There is of course an major influence of the fluid viscosity, and also of its density – both values depend on temperature and pressure. For more information look up “Reynolds number”.

During ascent towards the surface, magma composition will change along the way. As the temperature and pressure both get lower along the magma path, its composition will be modified, and some new minerals will appear (forming a crystal mush) and some gas bubbles will appear also (the solubility of gases in liquids is a function of concentration, pressure and temperature). Depending on the type of gas (water, CO2, mainly), the overall result will be a diminishing gas solubility in the liquid phase, hence the bubbles. Some bubbles will stick together and grow. The nucleation is triggered when a sufficient amount of supersaturation is reached. It can be facilitated by crystals which will help forming bubbles by providing what is called an interface – a zone where two different states of matter meet (solid, liquid, gas). As pressure decreases during ascent, the solubility of water in the melt decreases rapidly, which increases the degree of supersaturation, more and more bubbles will nucleate and grow.

Caption: Vesiculation of a Chaiten rhyolite at 875°C

The overall effect is drastic. The water content of the magma is a key parameter to viscosity. Water is the most important volatile in magma which may solve up to 10% of it´s weight. The water vapour has left the liquid to get into the the new bubbles and the viscosity of the magma rockets up.  A magma containing 4 wt.% is likely to erupt explosively.

In the chemical world movement can come from differences in concentration. For any species, the general movement will be from the higher concentration zone to the lower one. Water (in its supercritical state?) will migrate from the liquid phase of the magma to the bubbles and so the bubbles will grow. To a point. By doing so they will first act as little undeformable spheres.

While the spheres are still small, it will cause the overall viscosity to get much higher as the fluid has to flow around the small gas spheres as if they were solid.

This notion is counter intuitive, as the viscosity of the bubble is only a very small fraction of the viscosity of the liquid, but the bubble is maintained spherical by other forces.

This viscosity rise can lead sometimes to a “viscosity quench”, the bubbles will stay as they are because the surrounding viscosity is too high to allow for more growth. This can lead to very explosive magma, as the water gas will go on diffusing in the bubbles and the overpressure in the bubbles will grow. The depth at which critical overpressure occurs, in the bubbles, if at all, is an important factor determining the transition of eruption style from effusive to explosive. Critical overpressure can e.g. be achieved by placing dry ice (solid CO2) in a solid container at room temperature as you can see in this video.

To grow, the bubbles have to push the surrounding liquid away and that can be done only if the viscosity is not too high for that.

So, mainly depending on the water content of the magma, the bubbles will in some cases continue to grow. Once they are big enough the forces exerced by the surrouding fluid will begin to deform the bubbles, stretch them. In doing so the resistance to flow due to the bubble will diminish and so will do the viscosity.


Scanning Electron Microsopic (SEM) image of ash from the explosive Grimsvötn 2011 eruption. Image by Spica, AEC.

But in the end the water content of the lava gets to a minimum and at this time we are very near the surface. So at one point, and depending on the lava viscosity, we’ll see a different comportment.

First the easy ones :

Caption: Bursting bubbles on the surface in a Hawaiian style effusive eruption. Kilauea’s lava bubbles at Waikupanaha Ocean Entry 18 Aug 08. 

If viscosity is still not too high, gas bubbles will manage to get out quite easily. This will be the case for fluid basalts for instance.

Then if the viscosity is intermediate, the gas bubble will manage to get elongated, sometime giving birth to woodworm like structures. Most of the gases will get out not too explosively.

Finally, if the lava acts nearly like a solid, then the nasty stuff will happen. The gas pressure in the bubbles will exceed the rock resistance and the rock will eventually give way. Violently. Right beneath this level that is called the fragmentation heigth, the decompression rate increases sharply by more than two orders of magnitude. This means that even more bubbles may nucleate because of oversaturation. The water depletion of the magma and corresponding increase in viscosity carries on into the depth of the conduit in a self-reinforcing process.

As long as rate of destruction of lava and of the flow incoming lava is equivalent, the eruption will continue. The interface between the decomposing lava and the upcoming one stay at about the same level (ie the vent). Once the rate of lava alimentation diminishes, the eruption will slow and eventually stop. Of course, some times explosive behaviour will empty the fluid column and it will take some time for the lava to reach the surface again. That is what we see in vulcanian explosions for instance.


Massol and Koyaguchi (2005) “The effect of magma flow on nucleation of gas bubbles in a volcanic conduit”, Journal of volcanology and geothermal research.

Chryphia and Dfm


This week´s riddles: 

Grover1. President + image Answer: Mount Cleveland, Grover is the muppet, and Grover Cleveland was a US president. Bobbi, 2 points. 

2. I’m older than the one in Spain, but they still call me “new”? Answer: the New Madrid Fault. KarenZ, 2 points. 

3.  This hot spot has cabbage, but you’re going to have to travel a long way to get it! Answer: Kerguelen hotspot, one of the most isolated islands in the world, it is home to the endemic Kerguelen cabbage. Sissel, 2 points. 

4. At the end of a crack, my lava will flow-right!

5. Sea dogs and monkey suits?  I am the newest.

Score board
14 Sissel
8 Inannamoon667
8 Kelda
5 KarenZ
6 Shérine France
5 Dinojura44
2 Alison
2 Evan Chugg
1 Diana Barnes
1 Edward

138 thoughts on “Magma properties and Lava Fragmentation – in simple words

    • FYI for the curious. That link directly covers stuff that was written about in this post. I highly recommend it. Api is discussed using ideas that were introduced here. Almost made to order.

    • Wow, that was a veritable tour de force. Love the way petrologists unravel the puzzle. So what we have here is a close relative of tambora with evidence of recent magma injections and fractionation occurring at least at two levels, one below the crust in the Moho and the other in a crustal chamber. The lava is silica under-enriched (trachy-basalt) but high in potassium and volatiles making it fairly explosive stuff. Have I got at least that much right?

  1. Well, I thought I was about the only dragon wandering around at this hour, but Spica is already on the new post tweaking it for release.. so, I will turn in and check it out tomorrow.

    Good on ya Spica! 🙂

    • Okay… I cheated and snuck a peek. It’s a good write up!

      Food for thought: Peak SO2 to sulfate conversion will happen in about 2 to 3 months, for me, that is in the middle of hurricane season. If Api did reach as high as I’ve heard, it may cut back on the energy budget for the tropics just in time to make the latter part of the season less energetic.

    • Yeah KarenZs possible post is edited in. I am just looking for an image .. copyright issues..besides that it is done. ( Thanks for using my ash pics 😉 !!!!! ) Thought they were completely forgotten.

      • Hi Spica
        No they are not forgotten, I intend to do something directly with them, a sort of follow up. It shows the value of the treasury. This pictures were made more than one year ago, And now we can exploit them in a different way.

  2. Öraefajökull still going at it. Looks like it all began with that deep quake (M1.8 @~20km) yesterday.

    A question about the drumplots: Is there a visual difference between wind noise and harmonic tremor (Ignoring the infrasound plots), or does it take some mastering to distinguish the two on the drumplots?

    • In order to be able to distinguish the causing events one would first need a high temporal resolution of the drumplot (seconds, not minutes like here). Secondly a spectral analysis would help a lot. Some earthquake monitoring websites also show the raw drumplots together with RSAM (Real-time Seismic-Amplitude Measurement) and SSAM (Seismic Spectral-Amplitude Measurement) plots, for example GeoNet, but not IMO so far.
      This is a good educational introduction to seismic signatures from IRIS:

      • There are cheats one can use.
        First of all one can check the local weather in the region. If we check that we see that there is not enough wind to cause what is happening right now, nor was there enough wind during the night event.

        Second of all one can check for earthquakes. Currently there is one listed at Esjufjöll at magmatic depth, that is also a clue.

        There has also been quite a lot of smaller earthquakes not yet listed, but that has been visible on 3 or more seismographs that is surrounding Karasker.
        In this case we can also notice that there has been many microseisms associated with the tremor. So, with a bit of experience you can be fairly certain that something is going on down there.

  3. Thank you Chryphia and dfm. More brain food for me to digest. I am learning so much, if I had money to spare I would go for a distance learning degree course in Volcanology…if there is one. 😀 But as I can’t afford the money and I don’t need to prove to anyone my ability to learn facts then apply them to carry out field/laboratory work, I shall just sit back and enjoy the learning process.
    Once again stumped by riddles.
    Got coffee # 1 and planning work on the veggie patch as it’s not raining….yet.
    I am preparing for a long stay down in Brighton as I am needed to help with the arrival of Grand child # 6. I shall be 100% Toddlerized as the new Big Sister (aged 4) and I will be off adventuring whilst Mum & Dad cope with the nappies and feeds 😀 😀 So I may be away from the 16th June and ominously my Son has sent me an open ended return train ticket……..I may be some time!
    Today I must remember to take my camera to get pictures of my Allotment. It almost looks organised 😀 😀
    I have used lots of vermiculite in the process of raising my vegetable plants…. So any reference to my allotment is not OT. It just highlights the wonderful and world wide uses of Volcanic products.
    I need Coffee #2.

    • Hi Diana and thanks
      As stated in the artile, there is one Mooc on from which we took some part of the material.
      you can check if you like

  4. I have the large dog laying next to the bed waiting for me to give him another doggie treat. I’m a bit paranoid to doze off since my wife stepped on him the other day and did a faceplant as he fled out from under her. Stepped on dogs tend to flee in haste. I like him where he’s at, but I don’t need him freaking out when she comes into the room. When he moves fast to get out of the way, she gets nervous at which way he’s gonna go. In staying awake, I can be the traffic director.

  5. Spica, I do not trust myself yet to open the dashboard here and check (not enough coffee). Who is the copyright holder of the image in KarenZs article-image? Perhaps it is someone we allready have image copyrights from or who is included in the EU Fair Usage Act?

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 )

Google+ photo

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

Connecting to %s