Ed Flashcards

1
Q

(1) Two most common tectonic settings for volcanos:

A
  1. Mid-ocean ridges.
    • These divergent plate margins represent the dominant form of volcanism on Earth. The mid-ocean ridge system extends ~65,000 km and produces around 80% of the Earth’s magma output (Crisp, 1984).
    • The magma erupted – Mid Ocean Ridge Basalt (MORB) – is remarkably uniform in composition over time and space. This is because the mantle conditions under which the melting occurs are similar everywhere, and because the magma only has to travel through a thin, compositionally uniform crust to erupt, giving it little opportunity to evolve through differentiation or assimilation.
    • MORB has a fairly low viscosity on eruption and is relatively poor in volatiles.
    • Eruptions also typically take place at a few kilometres’ depth in the oceans.
    • Together, these factors mean that mid-ocean ridge volcanism is largely characterized by gentle effusion, with pillow lavas the primary eruptive deposit.
  2. Convergent plate margins.
    • The bulk of sub-aerial volcanism occurs at convergent plate margins, above subduction zones.
    • Subduction zone volcanism is more varied than mid-ocean ridge volcanism.
    • The magma often has to ascend through tens of kilometres of compositionally diverse crust, allowing ample opportunity for it to evolve through differentiation and assimilation; consequently the composition of the erupted magma, the styles of eruption, and the nature of the landforms produced, are all very diverse.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

(1) The third tectonic setting

A

a) Rifting.
• This is cognate with mid-ocean ridge spreading and is, perhaps, not properly identified as truly ‘intra-plate’.
• It may be related to the extension of a mid-ocean ridge onto continental crust, or it may be related to the development of a new spreading centre.
• Examples include the East African Rift, and the Rhine Rift in western Germany.

b) Hot-spot volcanism.
• Most true intra-plate volcanoes are related to a ‘hot spot’ – a temporally persistent mantle melting anomaly which may last for tens of millions of years.
• It is generally, but not universally, accepted that these hot spots occur where hot, upwelling mantle plumes impinge on the base of the lithosphere.
• Hot spots can produce spectacularly voluminous eruptions, creating Large Igneous Provinces (LIPs). These may be basaltic (e.g. the flood basalts of the Deccan Plateau in India) or silicic (e.g. Yellowstone) depending on the maturity of the plume and the lithosphere on which it impinges.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

(1) Global trends of volcanic settings

A
  • Largest volcanic system – MOR system – main system of volcanic output
  • Only small fraction of magma is extruded, mostly is added to crust to produce new crust
  • Continental rift zones have different compositions as they mix with crustal material – MOR have no crust to react with
  • Much more production at hotspots due to higher initial temp and decompression melting – Hawaii
  • Subduction – dehydration melting – over water = island arc – over land = volcanic arc – larger more complex at volcanic arc due to interactions with crust
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

(1) Plutonic vs Volcanic

A
  • Plutonic = what is added to crust (intruded)
  • Volcanic (erupted) is extruded onto surface
  • Much more magma is intruded and is extruded
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

(1) Volcanos and human populations

A
  • Part of the reason for this association is that volcanoes can induce orogenic precipitation, and their eruptive products weather to produce nutrient rich soils, hence volcanic regions often provide good conditions for agriculture.
  • There are also a number of volcanoes near low-lying coastal areas, which are often sites of large urban centres.
  • As a result of these factors, nearly 10% of humans live within 100 km of a historically active volcano; nearly 12.5% within 100 km of a volcano that has been active in the Holocene (Small and Naumann, 2001).
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

(1) Plinian example and its height

A

Pinatubo, Philippines
June 15th 1991 (composite)
• The eruption was violently explosive, and most of the magma was finely fragmented to produce ash. The ash and gas rose in a lofting plume called a Plinian column, which reached to a height of around 35 km.
• This is well above the tropopause, such that much of the ash and gas was injected into the stratosphere.
• This is very important for atmospheric distribution of the eruptive products for two reasons:
o 1) most of what we consider as weather occurs in the troposphere and there is no rain in the stratosphere to wash out eruption products;
o 2) strong stratospheric winds can disperse eruptive products over a very wide area - jet streams in stratosphere – can carry around world
• Consequently, satellites tracked the ash cloud from the eruption circling the globe several times. The SO2 injected into the stratosphere caused global temperatures to drop by around 0.5°C for the next two years.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

(1) Strombolian eruption

A

Produces mainly gas – almost no lava

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

(1) Surtseyan eruptions

A

Surtseyan eruptions are submarine, hydrovolcanic and highly fragmented

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

(1) Issues of eruption descriptors?

A
  • Firstly, each descriptor can cover a wide range of scales of eruption. The difference between a ‘small’ Plinian eruption, and a ‘large’ Plinian eruption can be very important, particularly for the hazard posed to surrounding communities.
  • Secondly, what features are used to classify the eruption? The presence of a lofting, buoyant column of ash and gas could be diagnostic of a Plinian eruption – but how can this be applied to eruptions that were not observed?
  • Thirdly, the descriptors may be too vague for some purposes.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

(1) Semi-quantitive classification:

A

qualitative descriptors on a more quantitative footing.
• For example, plotting the degree of fragmentation (percentage of material of ash size) against the area covered by ash. T
• his scheme has the advantage that it can be readily applied to un-observed eruptions from the geological record, as long as there is adequate preservation of deposits. This is essential, because (as we shall see shortly) very large eruptions are rare, and so we must rely on the geological record to study them.
• However, this scheme is far from perfect. For example, it doesn’t capture eruptions that do not produce fragmental material – where would a Pelean eruption plot? Or a unit within a flood basalt province?

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

(1) Quantitative classification

A
  • The most widely used quantitative classification for eruptions is the Volcanic Explosivity Index (VEI).
  • This is a discrete scale that uses volume of tephra produced and eruption column height to assign a numerical value between 0 and 8 to an eruption.
  • The scheme has been very widely adopted, but it, too, has limitations.
  • The chief amongst these is that it only captures explosive eruptions – i.e. those that produce tephra and/or an eruption column. A further limitation is that it assumes a link between eruptions that have a very high flux (hence a tall column) and those that produce abundant tephra.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

(1) Magnitude vs Intensity

A

Magnitude = log10(mass erupted in kg) − 7

Intensity = log10(mass eruption rate in kg/s) + 3

  • Both scales are logarithmic, such that the mass of eruptive products from a magnitude 6 eruption is 10x greater than from a magnitude 5 eruption. The magnitude is closely related to VEI – you can approximate VEI by rounding the magnitude to the nearest integer.
  • Calculating both the magnitude and the intensity of an eruption allows greater discrimination among different types of eruption. the simple relationship between intensity and magnitude implicit in the VEI does not hold universally. Low magnitude eruptions may still have high intensity if they are very short (e.g. the Halemaumau explosions). On the other hand, low intensity eruptions may still achieve relatively large magnitude if they are very long-lived (e.g. some lava-dome forming eruptions).
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

(1) Problems with the record of eruptions over time

A
  • Loss of small eruptions in the record through subsequent resurfacing, or erosion.
  • These same processes will introduce greater uncertainties on older eruption deposits.
  • The uncertainty on eruption age is generally greater for older eruptions.
  • The very largest eruptions are very rare, making those data less statistically meaningful.
  • It can be hard to distinguish lots of small eruptions from one large eruption, particularly far back in the geological record.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

(2) Modelling of surface and atmospheric transport processes is a more mature discipline than modelling of subsurface processes. There are at least three reasons for this:

A
  1. 1) eruptive products are the primary agents of hazard;
  2. 2) transport process are important in many disciplines and have been intensively studied;
  3. 3) the physical processes that mediate dispersal can be observed directly.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

(2) Of these three, the first provides the motivation, the second means that there is much that can be borrowed from other disciplines, and the third allows for effective validation of the accuracy of the models. Developing models of subsurface processes is more challenging for two main reasons:

A
  1. 1) the processes cannot be observed directly;

2. 2) magma has extremely complex behaviour that has not yet been fully characterized.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

(2) Models of subsurface processes fall into two categories.

A
  • The first includes models that link monitoring signals – seismicity, gas emission, ground deformation, etc. – to physical processes in the subsurface. These models can be extremely useful in supporting the interpretation of monitoring data.
  • The second category of models have the grander aim of modelling eruptions themselves, usually focussing on processes that operate in the volcanic conduit (the subvolcanic ‘plumbing system’) that links the shallow reservoir to the vent.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

(2) Consider what eruption products can tell us about the eruption that produced them

A
  • Vesical size – bubble size and gas content
  • Hydrous minerals = how wet of a magma = how explosive it was – melt inclusions
  • Texture – cooling history + vesicles
  • Composition = type of eruption = tectonic setting + viscosity
  • Type of product + main product
  • Mineral diffusion to determine residence time in magma chamber – know how long until an eruption from a recharge event
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

(2) Coke and mentos

A
  • Surface energy from scratches on mentos – nucleation site – gives activation energy for bubbles to form – coke is above its saturation point before hand and gas is forced out after activation energy reached
  • Bubbles have to rise through liquid to burst – when bubbles form there is a volume change – a race for bubbles to rise before time runs out and liquid stops rising
  • Relate to explosive eruptions–
  • Bubbles don’t have enough time to rise through liquid and burst – bubbles are small. Liquid is rising very fast and liquid is viscous
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

(2) A simple conceptual physical model of volcanic eruptions

A
  1. Silicate melt, containing dissolved volatile species (primarily H2O and CO2) is stored in a crustal magma reservoir.
  2. The magma becomes supersaturated and bubbles nucleate and grow.
  3. The growth of bubbles causes the density of the magma to drop, driving ascent of the magma.
  4. Ascent promotes further growth of bubbles through exsolution and decompression – this positive feedback drives and sustains the eruption.
  5. Under certain conditions, the growth of the bubbles may become so energetic that it leads to fragmentation of the magma.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

(2) Liquid saturation at depth

A
  • At depth liquid is undersaturated
  • Something will cause magma to become supersaturated – one cause is rising magma – bubbles will nucleate – crystals forming cause a nucleation site – bubbles decrease density – decompression occurs – bubbles rise faster – run away effect
  • Viscosity = magma is more viscous than water – lowest is 10Pa/s – magma can break apart in a brittle way unlike water – if it is stretched too fast (fast rising bubble) it begins to act as a solid and cracks
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

(2) Volcanoes as syphons

A
  • The pressure in the crustal magma reservoir is largely controlled by the weight of the overlying country rock – we can say that the reservoir is lithostatically pressurized.
  • A column of magma will rise up a conduit leading from the reservoir until it reaches a height at which the ‘magmastatic pressure’ at the base of the conduit is in equilibrium with the reservoir pressure.
  • The magmastatic pressure is controlled by the density of the magma and the height of the magma column: 𝑃 = 𝜌 𝑔h (neglecting atmospheric pressure). The 𝑚 magma reservoir pressure is controlled by the density of the country rock and the depth of the reservoir:
  • Reservoir is lithostatically pressurized – from overburden of rocks
  • Height of magma in conduit is balanced by pressure in reservoir
  • Density of the magma is greater than the column of rock
  • Lavas are full of vesicles – these fill with water so the density is lower
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

(2) Lava lake eruption in terms of density

A

– density of magma is the same, on average, as the density of the rock so the height of the magma is the same as the depth of the reservoir - surface

23
Q

(2) Density of rock is greater than magma

A
  • Height of magma in conduit is greater than depth of reservoir so it goes higher than surface
  • Lithostatic pressure allows the reservoir to constantly be pressurized and ‘squashed’ so it can empty in an eruption
  • Gas bubbles form provides buoyancy and lowers density, so flow is upwards due to difference in pressure
24
Q

(2) Solubility and the formation of gas bubbles

A
  • Gas bubbles form in magma when it becomes supersaturated in volatile species.
  • This occurs when the concentration of the volatile species dissolved in the melt exceeds the saturation value – this value is called the solubility 𝑆 of the species.
  • Water is usually the most abundant volatile species in magma, hence it plays the most important role in determining the gas volume fraction.
25
Q

(3) Magma

A
  • In order to understand the phenomenological complexity of volcanic eruptions, therefore, it is essential to understand the physical complexity of magma.
  • The exact composition of the primitive melt depends on the conditions under which the mantle is melted and varies a little with tectonic setting.
  • More dramatic variations are introduced if the magma assimilates crustal rock that it melts on its ascent, or if it stalls and crystallizes in the crust, causing certain oxides to partition preferentially into the crystal phase.
  • It is rare for truly primitive melts to be erupted
  • At depth – single phase Newtonian fluid
  • As it cools and decompresses – bubbles form as pressure drops – as it degasses crystals form
  • Magma is a complicated soup of elements
  • Obisidian – quenched silicate melt – rhyolotic – rhyolite is a stable crystal composition = doesn’t like to crystalise – forms under fast cooling – quenched immorphous solid
  • Visiculates under heating becomes its got enough residual water to foam under heating
26
Q

(3) Physical processes in magma

A
Flow
Crystalisation
Bubble nucleation
Bubble growth
Fragmentattion
Outgassing
  • Crystals can form in crustal storage – if stored for too long without eruption = plutonic batholith
  • If bubbles grow fast enough = magma fragmentation – flows too quickly and breaks apart
  • Outgassing = large bubbles can move through magma without dragging magma with them
27
Q

(3) Properties of magma

A
  • Viscosity – dictates flow
  • Density – drives flow
  • Surface tension – controls bubble nucleation and deformation
  • Diffusion – dictates bubble growth speed and crystallisation speed
  • Strength of magma – if we fragment magma
28
Q

(3) Flow

A

This is the most fundamental process, playing a role in every part of the magmatic system.
a. Flow is driven by pressure differences and opposed by viscosity – recall the Hagen-Poiseuille equation (Eq 2.3). In order to understand and model flow of magma, we therefore need to know its viscosity. As we found in exercise 2.5, pressure differences in the subvolcanic plumbing system are often created by density contrasts, so we also need to know its density.

29
Q

(3) Crystallization

A

Crystals form in response to cooling and degassing (loss of dissolved water from the melt). Their growth changes the composition of the residual melt phase, usually making it more silica-rich, and richer in volatile species. Consequently, crystallization changes the viscosity and density of the melt, and the growth of the new phase changes the magma’s bulk rheology.

30
Q

(3) Bubble nucleation

A

Bubbles are hugely important in the volcanic system. They form through a process called ‘nucleation’, which requires that the melt is supersaturated, and therefore depends on solubility of the volatile phase. It also depends on the viscosity of the melt and the surface tension between the melt and the gas. As with crystals, the presence of bubbles changes the magma’s bulk rheology.

31
Q

(3) Bubble growth

A

As we saw in the last session, the growth of bubbles is the primary driver of volcanic eruptions, mainly through the resulting change in magma density. Bubbles grow though decompression, so we need to know the equation of state of the gas. They also grow though exsolution (i.e. degassing) of dissolved volatiles, which is limited by the diffusivity of the volatile species through the melt. Growth is opposed by the viscosity of the melt, and by surface tension.

32
Q

(3) Fragmentation.

A

This occurs when the bubble growth rate is so high that the melt surrounding the bubbles breaks in a brittle fashion. Fragmentation therefore depends on the strength of the melt, in addition to the properties on which bubble growth depends.

33
Q

(3) Outgassing.

A

This is the loss of the gas phase from its parent packet of magma, and is distinct from degassing, which is the loss of volatiles from the melt by exsolution. Outgassing takes place via two main mechanisms:

  1. Permeating flow. This occurs when gas flows through connected networks of bubbles and/or cracks and depends on the permeability of the magma.
  2. Bubble rise. This occurs when bubbles are able to rise appreciably through the melt on the timescale of eruption; it is sometimes referred to as ‘separated flow’ or ‘slip velocity’. This depends on bubble size, viscosity and density of the melt, and surface tension.
34
Q

(3) Density

A
  • The density of silicate melt depends on its composition, temperature, pressure, and dissolved water content.
  • These relationships are encapsulated in an equation-of-state (EOS), which is a function that links volume, pressure and temperature of materials – the ideal gas law that we used in exercise 2.4 is another example.
  • In general, density decreases with increasing silica content, increasing water content, decreasing pressure, and increasing temperature (in approximately that order of importance). Silicic magmas tend to be richer in dissolved water than mafic magmas, hence have lower density. This is offset slightly by the fact that silicic magmas tend to have a lower eruption temperature than mafic magmas.
35
Q

(3) Solubility

A
  • In general, the solubility of a volatile species depends on pressure, temperature, and melt composition
  • Solubility and exsolution are fairly straightforward when only a single volatile phase is considered, but become rather more complex when multiple volatile species are considered.
  • From a physical volcanology perspective, it is H2O and CO2 that are the most important species (though SO2 is very important for monitoring and gas geochemistry).
  • The solubility of H2O depends on CO2 concentration and vice versa.
  • Although CO2 is usually present at much lower concentrations in silicate melt than H2O, it also has much lower solubility.
  • Consequently, CO2 is often the first volatile species to reach saturation as melt ascends, or as crystallization drives up volatile concentration in the residual melt.
  • Once CO2 bubbles have formed, the other volatiles species will partially degas into the bubbles, because their solubility at an interface depends on the partial pressure of the volatile species in the gas bubble, rather than the absolute pressure.
  • It is therefore possible to have H2O in bubbles when the absolute pressure is above the saturation pressure at which pure H2O bubbles could nucleate.
  • The importance of solubility from a physical volcanology perspective extends beyond its role in controlling the growth of bubbles – the concentration of dissolved water in a silicate melt also has a profound influence on other magma properties.
  • For example, melt viscosity depends strongly on dissolved water content, and the presence of water depresses the liquidus of the melt, such that degassing of water promotes crystallization.
36
Q

(3) Surface tension

A
  • Nucleation and growth of bubbles involves the creation of new interface between the melt and the gas phase. This interface has an energy associated with it, which is usually described as a surface tension. Surface tension acts to minimize surface area, which has some important consequences for volcanic processes, for example:
  • Surface tension delays nucleation by posing an energy barrier to the formation of new bubbles. Consequently, melts have to become supersaturated before nucleation will occur, and the degree of supersaturation required depends on the availability of nucleation sites, such as crystal phases.
37
Q

(3) Diffusion

A
  • Molecules within a melt are constantly in motion and are therefore able to move through the melt by diffusion.
  • If there is a spatial gradient in the concentration of a particular molecule species, then molecules of that species will tend to diffuse down the concentration gradient, smoothing it out.
  • One of the main reasons we care about diffusion is that it limits the rate at which volatiles can degas from a melt, into bubbles. In exercise 2.4, we assumed that degassing is immediate once a decompressing melt reaches its solubility curve.
  • This is equivalent to saying that melt around a bubble always loses (or gains) volatiles immediately in response to changing pressure and temperature conditions, in order to stay at the saturation level appropriate for the conditions – we call this equilibrium bubble growth.
  • In fact diffusion acts as a kinetic ‘brake’ on bubble growth, so that bubbles often grow in disequilibrium, allowing melt to become significantly supersaturated.
38
Q

(3) Melt strength

A

• Silicate melt typically behaves as a viscous liquid such that, when a stress is applied to it, it deforms in a ductile fashion (i.e. it flows).
• However, if a very large stress is applied, it can fail in a brittle fashion.
• This is a property that is common to all viscous liquids, but we rarely experience it because most liquids that we commonly encounter have a lower viscosity than silicate melts.
• Failure indicates a transition in material behaviour from viscous (ductile) to elastic
(brittle).
• In the viscous regime, deformation (strain) in response to an applied stress can be accommodated through structural reorganization of the molecules within the liquid.
• In the brittle regime, the material cannot reorganize rapidly enough to accommodate the strain, and so it fails.

39
Q

(3) Magma rheology

A
  • Rheology is the study of the deformation and flow of materials in response to an applied stress – the term is also used to describe the deformational behaviour of materials.
  • The most commonly used descriptor of rheology is the viscosity, which is the ratio of stress and strain rate.
  • For many common fluids (e.g. air, water, oil, syrup) the ratio of these quantities is constant, which means that the rate of deformation doubles if we double the stress applied.
  • We call these fluids ‘Newtonian’, and their rheology is completely described by a single quantity, which we call the viscosity, 𝜇:

• Many fluids do not obey this simple relationship and are called ‘non-Newtonian’.
• For these materials, there is a more complex relationship between applied stress and resultant strain rate.
• For any such fluid, an ‘apparent viscosity’ 𝜂 can be defined which is the ratio of stress and strain rate at a given strain rate – indicated by the dashed line in Fig
• Pure silicate melt is approximately Newtonian over a very wide range of strain rates; however, magma is composed of three phases – liquid silicate melt, solid crystals, and gas bubbles – and the presence of the suspended phases introduces non-Newtonian behaviour.
• The fractions of each of the three phases may vary from 0 to 1 and, for
any given packet of magma, the fractions change dramatically as the magma moves through the volcanic plumbing system.

40
Q

(3) Different fluids

A
  • Newtonian fluid – where stress and strain rate is proportional = water, air, oil, syrup etc – double the stress = double the strain rate
  • Elastic fluid = hookian fluid
  • Yield stress – has to hit a certain stress to flow – Newtonian fluids do not have this – typical of something with a large number of particles
  • Shear thinning rheology – paint – as we increase the stress the strain rate increases non linearly
  • Melt = Newtonian until crystals and/or bubbles form – becomes shear thinning
41
Q

(3) Melt viscosity

A
  • Natural silicate melts lack long-range order but, at the shortest range, are primarily composed of tetrahedral units of SiO4 and AlO4, which polymerize via bridging oxygens.
  • Consequently, higher abundance of Si and Al generally leads to greater polymerization, hence higher melt viscosity.
  • Other species, such as Na and Ca, act to depolymerize the melt and decrease its viscosity. The major element composition of a melt therefore exerts a strong control on its viscosity. Temperature, too, plays a crucial role – viscosity decreases dramatically as temperature increases.
42
Q

(3) Crystal-bearing magma

A
  • Magma crystallizes as it cools and as it exsolves water.
  • Within the conduit, exsolution-driven crystallization likely dominates; for erupted lavas, cooling-driven crystallization is more important.
  • The growth of crystals causes a dramatic increase in magma viscosity.
43
Q

(3) Bubbly magma

A
  • Bubbles have a profound influence on magma rheology.
  • For many years the scientific community was divided over whether the growth of bubbles increases or decreases the apparent viscosity of a melt.
  • This apparent contradiction was resolved in the early years of the millennium when careful analysis of experiments on analogue and natural bubbly magmas (see Llewellin and Manga, 2005) revealed the central role of the capillary number, which is the ratio of deforming viscous stress (equation 3.5) to restoring elastic (surface tension) stress
  • At low Ca, elastic stress dominates, and bubbles remain approximately spherical, distorting flowlines within the magma, and therefore increasing viscosity.
  • At high Ca, viscous stress dominates, and bubbles become elongate in the direction of flow.
44
Q

(4) The 3 Es

A
  • Explosive = produces clastic materials
  • Effusive = not clastic = lava flow
  • Extrusive = also not clastic but more like a lava dome than a flow
45
Q

(4) Flow or blow?

A
  • The most fundamental dichotomy in volcanic eruption styles is between effusive (or extrusive) and explosive. From a physical volcanology perspective, the discriminating event is magma fragmentation.
  • Fragmentation occurs when magma undergoes a topological inversion from a continuous liquid suspending bubbles and crystals, to a continuous gas phase containing discrete clasts of magma. As we will see, it is possible – and common – for magma to comprise both a continuous liquid and gas phase, so the defining feature of fragmentation is the disruption of the continuous liquid phase.
  • Note that fragmentation usually requires that the gas phase is over-pressured with respect to the ambient conditions, such that the magma around the gas phase (in bubbles or cracks) experiences a differential stress that is large enough to overcome its strength
46
Q

(4) This development of gas overpressure is enhanced by a number of factors:

A
  1. Rapid decompression of the magma, for example through rapid ascent, or unloading by collapse of a lava dome or part of the volcanic edifice.
  2. High initial water content providing an abundance of the gas required for bubble growth.
  3. High magma viscosity such that bubble growth is too slow to keep pace with decompression
47
Q

(4) Outgassing

A
  • Degassing = gas out of magma
  • Outgassing = gas fractionating from magma

The primary factor that mitigates the development of overpressure – along with the inverse of the three factors listed above – is the loss of gas from the magma. This is termed ‘outgassing’ (not to be confused with degassing, which is the exsolution of dissolved volatiles) and is mediated by two key processes: bubble rise and permeable flow. The influence of the latter is illustrated in Fig 4.1, which shows a sequence of event by which outgassing through permeable flow transforms an explosive eruption into an effusive, dome-forming eruption.

48
Q

(4) Rapid acceleration

A

magma ascends, pressure drops, bubbles expand, if they can’t expand as fast as they want to they become overpressured, can go from ductile to brittle bubble films causing fragmentation

49
Q

(4) Rapid decompression

A
  • Decompression – Pressure wave brings low pressure down the layers causing fragmentation
50
Q

(4) Slug flow, pulsatory fountains, and Strombolian activity

A
  • True Strombolian activity is understood to result from the bursting of large ‘slugs’ of gas, which ascend through almost stagnant magma.
  • The slugs are large enough that they are confined by the conduit walls, adopting an elongate morphology.
  • This allows the slugs to develop a sufficient overpressure to drive the mildly explosive eruptions that are characteristic of this style of activity (Del Bello et al, 2012).
  • Mildly explosive bursting of large gas bubbles rising through stagnant magma is also observed at lava lakes, such as Halemaumau, Kīlauea. In this case, the bubbles were not confined by a conduit, though they often rise against one wall of the lake.
  • Organization of the gas phase into large bubbles is not unique to stagnant magma.
  • As we determined in exercise 2.5, lava fountains are associated with high magma ascent velocities.
  • Nonetheless, they are often pulsatory in nature – if they are particularly so, they can be referred to as ‘rapid Strombolian’ eruptions. These pulsations indicate that the ascending magma contains large pockets of over- pressured gas, demonstrating that substantial separated flow occurs in these eruptions. Analogue experiments (Fig 4.3) support this hypothesis.
51
Q

(4) Outgassing – permeating flow

A
  • When a medium contains connected gas flow pathways at the micro-scale, it is said to be permeable.
  • In magma, permeability arises from connected networks of partially-coalesced bubbles (Fig 4.4), and/or from connected networks of cracks.
  • The gas flux 𝑄𝑔 through permeable magma depends on the pressure gradient driving flow, the viscosity of the gas 𝜇𝑔, and the permeability 𝑘 of the magma.
52
Q

(4) Permeability of vesicle networks

A

• Permeability is found to be a strong function of gas volume fraction.
• For pumices with approximately spherical vesicles (the quenched equivalent of bubbles) the permeability is zero below a ‘percolation threshold’ of around 𝜙 = 0.3.
• This is because the bubbles are too far apart to create a long-range network. Above the threshold, permeability increases rapidly with gas volume fraction.
• Pumice clasts with highly elongate bubbles show anisotropic permeability – the permeability is higher in the direction of elongation.
• Dome rocks typically have highly contorted gas flow pathways as a result of their high crystal
content and the complex deformation history of the magma. As a result, they do not show a percolation threshold and have an appreciable permeability even at very low gas volume
fraction.

53
Q

(4) Outgassing through cracks

A
  • Gas may also escape from magma along cracks. Cracks may form in magma in response to significant overpressure in the gas phase, by a process analogous to hydrofracking, or by shear stress, particularly at conduit margins.
  • Ash-filled cracks are often found in blocks that have been thrown out during mildly explosive phases of dome-forming eruptions (Fig 4.6).
  • They are also found in eroded conduit margins, indicating that they not only provide a route for intra-magmatic flow of gases, but also for outgassing from the conduit into the country rock.
  • The prevalence of tuffisites in dome-forming eruptions suggests that outgassing through cracks might play a role in determining whether an eruption is explosive or effusive.
54
Q

(4) Fragmentation

A

• At the micro-scale, the process of magma fragmentation must require that packets of melt separate from one another.
• This could happen through brittle failure if the strain rate in the melt exceeds the critical Deborah number criterion. In exercise 3.3 we determined that this is most likely to occur in high viscosity melts.
• It is also intuitive that the necessary high strain rates will be reached first in the thin films between growing bubbles.
• Low viscosity melts can reach extremely high strain rates without moving into the brittle field.
• Consequently, melt films may become thin enough to fail through surface-tension driven instability, rather than brittle failure.
• Because the fragmentation process is so important in determining the eruption style, significant research effort has been invested in characterising it and encapsulating it for use in numerical models of eruption.
• Experiments show that fragmentation occurs at much lower gas overpressures for samples with high gas volume fraction.
• A range of fragmentation criteria have been adopted in numerical models over
the years. Early models simply used a critical gas volume fraction (usually in the range 0.6 < 𝜙 < 0.8). More sophisticated models have used criteria based on critical values of bubble overpressure, bulk strain rate, stress, or inertia. Cashman and Scheu (2015) give a detailed overview.