Helpful Hints If you are an experienced petrologist

Helpful Hints If you are an experienced petrologist you might want to miss out section 1. I’ve tried to pop in some stimulating questions and summarise the information in a way especially relevant to experimental petrology so it might be worth looking at anyway. 2. is background to the important considerations for experimentalists and a little insight into the limits of the usefulness of experiments in the shallow parts of subduction zones. It contains a few of my own thoughts, further reading etc. Hopefully these sections will help you to get more out of the two papers which I have summarised in Section 3. I’ll be interested to hear your thoughts on some of the questions I have posed (but of course I am not trying to wriggle out of answering questions myself!) Enjoy. • Welcome! • I’ve added three sections (labelled (usually )in bottom left-hand corner 1, 2 and 3) • 1. Background to rocks in relation to experiments • 2. Experimental approaches • 3. Key findings from the papers NB : some of the slides are ‘animated’

What is a rock? I’d like you to start by taking 5 – 10 minutes to discuss your experiences of rocks and then to describe what features of (igneous) rocks you think are useful for understanding magmatic processes! 1 Andesite (opx, amph, plag, Fe-Ti oxides and accessory pyrrhotite). Rodin’s ‘The Thinker’

Many subduction-related systems are open, with almost all of these processes happening to individual rocks 1 So when might analysing bulk composition be useful/important ? What would the bulk actually represent?

Important properties of the solid components (crystals) I Pressure Possible Magma Ascent path Crystallisation starts here (v. few crystals) Some phases (e. g. Plagioclase) have strong temperature dependence of the liquidus at low PH 2 O Liquidus I More crystals here (more undercooled from the liquidus) 1 This is for a system containing water, volatiles are important here Temperature

Consequences in the rock Microphenocrysts – may reflect storage OR ascent conditions Microlites – smallest crystals, may reflect Degassing or ascent related process. Larger crystals (phenocrysts)Grow in the storage region, Reflect ‘magma chamber’ processes 1

Important properties of the solid components (crystals) II I Some phases (e. g. cpx) have a less marked T dependency, thus differing phenocryst components can reflect differing conditions of formation. us id Pressure qu Li Finally, hydrous minerals (e. g. amphibole, biotite) have a positive slope at low pressures. This reverses at higher pressures. Some of you may be able to explain this in terms of thermodynamic properties of the phase and melt! II II I +II Liquidus III 1 This is for a system containing water, volatiles are important here I Temperature Now think about typical storage, ascent paths and how this might be reflected in the mineral assemblage

Important properties of the solid components (crystals) III Most phenocryst phases in subduction-related rocks exhibit considerable solid solution. This is often a strong function of one or a few of of P, T, volatile content and f. O 2. Differing phases can be used for different purposes. This is an example of the influence of P(H 2 O) and T on plagioclase composition 1 (Soufriere Hills groundmass composition experiments from Couch e

plag This is a BSE-SEM image of (predominantly plagioclase, I’ve labelled a few) phenocrysts) from Soufriere Hills (Montserrat). Brightness is proportionate to mass so the differing colours reflect differing compositions. Take a minute to look at this image and think about what these phenocrysts could be reflecting Keep thinking about that bulk composition question! Scale bar is 500 mm 1

Rocks are about more than just composition I 1 Most melts are 20% less dense than their parent rock

Rocks are about more than just composition II m (Pa s) 1014 107 Based on Lejeune & Richet (1995) 40 60 Crystal content Crystallisation, cooling and changing composition all have implications for the response of the magma to stress (affecting mobility and movement of magmas). Some magmas behave as brittle solids. 1

Thinking pause Individual crystals tell us a great deal about magmatic history But when we are thinking about volcanic behaviour it is the macroscopic changes to the system that are important Never forget to look at the rocks in the field and think about this!!! You can probably come up with some examples.

The fluid phase. The amount of volatiles dissolved in the magma markedly decreases as a function of pressure (or depth) 1 Moore et al. , 1998 b – ref at end

Volatiles Phase change from dissolved fluids to exsolved gases is the fundamental driving forced behind volcanic eruptions Also have profound influence on physical properties of the magma (in what way? ) Exerts strong control on crystallisation. . . Many of you will have experience of volatile emissions, perhaps good here to reflect among yourselves on the way in which volatiles drive volcanic eruptions 1

Recall ‘plagioclase’ Pressure ‘Hydrous phases’ 1 This is for a system containing water, volatiles are important here Stability and compositions of such phases could be excellent record of volatile content or degassing processes Temperature

Experiments !!! Can determine volatile concentrations (see Moore et al. , 1998 b example) Using phenocryst compositions and abundance can reproduce conditions of storage and ascent, and perturbations relevant to eruption (our papers focus on this) Dynamic experiments can reproduce degassing-induced crystallisation (see Couch et al. , 2003 as an example) BUT. . . . 2

Need to carefully consider the nature of the liquid + crystals to be reproduced What is useful about an equilibrium assemblage? Most useful to use low crystallinity magma equivalent to that under investigation (cf Moore et al. , 1998) or natural glass or experimental reversal with bulk and glass Can anything meaningful be derived from working with a hybridised magma? 2 From Pichavant et al. , 2007

Problems with equilibrium: Need to consider whether want partial or total equilibrium and what this represents in the real magmatic system. We need to think again about the complete system AND local equilibrium. Example from Pichavant et al. , 2007

Some more observations There are comparatively few studies of basaltic andesites, although these are often implicated in the recharge and triggering of eruptions (incl. solubility data relevant to excess degassing) There is something of an experimental ‘gap’ above 3 Kbars and below 10 kbars. This is a technical problem rather than a lack of scientific interest (see e. g. conclusions from Moore et al. , and Barclay and Carmichael for reasons to work in this zone). See also point above!. . . But rescue may be at hand (see Moore et al. , 2008). 2

2

The kit At the end of the experiment the magnet is pulled down and sample rapidly removed from the hot zone 2 One of UEA’s RQCS – in its shiny orange safety cage

Locations of the papers (adapted From Carmichael et al. , 2006) 3

Moore and Carmichael • i Usual Colima assemblage: plag, opx, cpx and hbl (matched as shown) No significant role for CO 2 NB Colima assemblage too crystalline to use as starting bulk. Ascent-related plagioclase growth identified Mascota spessartite much increased water content. Interesting! 3

Basaltic andesite Equilibration at water contents of 35 wt% Reproduced assemblage compositions and volumes 3

Notable findings • Crystals in both compositions result of degassing and decompression of hydrous parent magmas • Can fractionate (hbl + plag) from similar hydrous parent • Volatile phase dominated by water • Limited role for magma mixing 3

Cerro la Pilita Trachybasalt (near Jorullo Volcano) Very rare global occurrence of amphibole-bearing basalts Using a lava with amphibole allows us to consider why its absent!

Phase diagram Stability field for the Cerro la Pilita cone

Effect of mixed-volatile conditions (produces unlikely plagioclase compositions) Influence of f. O 2 (shift of amphibole)

Astonishing crystallisation! 3 Recall: the ‘mechanical’ consequence of the large increase in crystallinity in the experiments of Le. Jeune and Richet

This considers other experiments on other compositions (Sisson and Grove, Moore et al. , and Blatter et al. ) Crystallisation effect associated with ‘hornblende in’ most marked for basalts Will all hydrous basalts stall?

Is this what happens?

Taking it further…. The residual melt defines a shoshonitic trend Here, the experimental composition are compared with the Aurora Volcanic Field in California A reasonable fit to the trends of erupted material (perhaps plag removal in Aurora)– could we sometimes have re-melting of stalled material as a source of water-rich magma ? I’m using bulk compostions and experimental glass compositions here – what do you think of that?

3

Davidson et al. , (2007) Cooling rather than Decompressi on (Annen et al. , 2006) 3 • Finds geochemical evidence for fractionation from amphibole • Does this apply to Mexico? ?

Conclusions • Used well experiments are exceptionally useful at determing pre-eruptive volatile contents, storage conditions and quantifying the nature of the perturbation prior to eruption • The attainment of equilibrium is very important and the use of a ‘bulk’ that represents a meaningful magma vital. • Lots of exciting work to be done particularly at moderate pressures

References not on the website