• Program

    The Goldschmidt2020 program

  • Venue

    All about the venue and city

  • Registration

    Details of how to register to attend the conference and/or submit an abstract

  • Exhibition

    Information for and about exhibitors and sponsors

  • My Goldschmidt

    My program, purchases, connections, etc.




  • Sponsor Seminars

  • Training Events

    These are scheduled events to help delegates test out the tools and platform we will be using for the Q&A and other events at the conference.

  • Workshops

    Our workshop program provides training and teaching in topics across geochemistry and related fields. We are currently liaising with the workshop organisers to ascertain if any workshops can become virtual. Any updates will be added to this page.


  • Conference Locations

    Location of the convention center and social events

  • Hotels

    The hotels are no longer available to book at the special rates agreed by the conference. Should you wish to cancel or change your hotel reservation please see the details below.


Present your work




My Goldschmidt

Role functions

Abstract Details

Days to Weeks of Syn-Eruptive Magma Interaction: High-Resolution Geochemistry of the 2002-03 Branched Eruption at Mount Etna

Magee R, Ubide T & Caulfield J

Magee R, Ubide T & Caulfield J (2020) Goldschmidt Abstracts, 2020 1698

Sorry, the PDF cannot be displayed on your browser.

Download abstract

The author has not provided any additional details.

05d: Room 2, View in program

Listed below are questions that have been submitted by the community that the author will try and cover in their presentation. To submit a question, ensure you are signed in to the website. Authors or session conveners approve questions before they are displayed here.

Submitted by Georg F. Zellmer on
Hi Ruadhan, Thanks for this interesting presentation on how to link petrography / mineral chemistry to volcano monitoring. What strikes me as the most fundamental aspect of your work is that you considered the groundmass composition, as opposed to the whole rock composition, and how this changed chemically and isotopically through time. This really makes sense. My question is of technical nature. How do you separate the groundmass? By hand picking? I presume given microcrystallinity, the handpicked particles are non-transparent, so how are you sure to exclude particles that have phenocrysts inside but are entirely mantled by groundmass? Or did you pick from grain sizes smaller than complexly zoned phenocrysts? Do you think it would be possible to do these analyses on the groundmass using microbeam techniques (EPMA combined with LA-ICPMS for majors and traces, and potentially Sr-isotopes using a quadrupole ICPMS in MS/MS mode - see talk by Thomas Zack on thermochronology)? And a related question to this is obviously the origin of the microphenocrystic (rather than microlitic crystal cargo? How are you sure that the microphenocrysts are not also of antecrystic nature? Thanks! Georg
Hi Georg, Thanks for the question! You are correct, the groundmass was handpicked with a tweezers under a microscope. You are also correct that the samples were crushed to < 2mm so that the grains picked were smaller than the complexly zoned phenocrysts. I was very aware that phenocrysts could be hiding in larger grains so I made sure to avoid them. Regarding microbeam techniques, yes I think that is a possibility. Something to explore in the near future is whether the same tends that we observed with solution analysis are reproducible with LA-ICPMS and EPMA analysis of the same samples. This would certainly preferable as groundmass picking is a tedious task! As for the origin of microphenocrysts, in a hybrid melt anything other than crystal rim (formed after mixing) could be considered antecrystic. The assumption we make here is that the majority of micropheocrysts formed after the onset of mixing (or upon ascent) as opposed to being disproportionately inherited from either endmember magma. Thankfully in the S-rift the groundmass microphenocrysts are not very abundant, so the potential for the composition to be skewed by microphenocryst populations is relatively low. However, I agree that in samples with high porphyricity (such as the NE-rift) it is more difficult to pick a crystal-free groundmass fraction and therefore you can be less confident of the result. That being said, most of the microphenocrysts in the NE-rift are plagioclase (likely to have formed at a late stage, perhaps upon ascent and degassing) which, if removed, may leave you with a residual melt composition that does not reflect composition prior to eruption. I think a good indication that the trends are largely reliable in the S-rift is the narrow range of compositions observed for all elements in the groundmass compared with the wide range in whole rock. Groundmass composition may of course vary within the same sample but not to the same extent as whole rock with variable phenocryst abundance. Lastly, I certainly think it important to verify the the trends with reference to the petrography and monitoring, which in this case agree with mixing. I hope that goes some way to answering your question, please let me know if want to discuss it further! Cheers, Ruadhan

Submitted by Maurizio Petrelli on
Hi Ruadhan, thanks for sharing your results. As Georg, I also have a question about the picking procedure. How do you can you ensure avoiding chemical biases related to the picking procedure?
Hi Maurizo, thanks for the question, I hope that in my answer to Georg I have somewhat answered it. There are certainly potential biases in the approach that are more difficult to avoid the more crystal-rich the sample. Generally speaking, however, I think picking does quite a good job of removing the larger crystals. As Georg mentioned, variability in microphenocrysts abundance/composition could potentially effect the results (e.g clinopyroxene accummulation effecting Cr composition). I have tried to check whether this is an issue in two ways. Firstly by looking at the composition of clinopyroxene in the X-ray fluorescence maps (slide 4 of my presentation). What I observed is that the majority of clinopyroxene microphenocrysts have the Cr-enrichment (observed at the rims of larger crystals) at their core, suggesting that they represent post-mixing crystallisation. Secondly, I point counted each sample checking if peaks in Cr were related to the accumulation of clinopyroxene microphenocrysts in different samples but could find no such relationship. There was no trick to the picking procedure itself other than zoom in and find the smallest, dullest looking pieces :) Thanks again and let me know if you would like to discuss anything further! Cheers, Ruadhan

Submitted by Maurizio Petrelli on
Hi Ruadhan, thanks for sharing your results. As Georg, I also have a question about the picking procedure. How can you ensure avoiding chemical biases related to the picking procedure?

Submitted by Lucy McGee on
Hi Ruadan - I really enjoyed your talk! The combination of chemistry and monitoring is really powerful. The story of mixing between those undegassed and degassed magmas would be perfect for short lived U-series isotopes (the 210Pb/226Ra pair is strongly affected by volatile degassing and accumulation). I was interested in this suggestion of an input of more evolved/radiogenic magma - does that suggest a complex storage system? Can the crystal textures and zoning patterns reveal more detail about how extensive such a plumbing system could be? Look forward to chatting more on Friday.
Hi Lucy, thanks for the question, i'm delighted you enjoyed the talk! As far as I know, some U-series isotopes have been measured for this eruption in a paper by Clocchiatti et al (2004). They concluded that ratios of close to 1, found in the most primitive and radiogenic samples, represented the arrival of undegassed magma. They did find one sample which had a ratio of less than 1 (0.95 ish) and attributed it to magma that had migrated to the shallow plumbing system a couple of years earlier (similar to the magma erupted on the NE-rift). Unfortunately all of these samples were from phases 1 and 2 of the S-rift eruption (last slide of my presentation) during which the most radiogenic, undegassed magmas were erupted. Samples from phase 3 may well show lower ratios and agree with progressive mixing with the more degassed magma! I know very little about U-series, so hopefully I have understood the results correctly! Regarding the plumbing system, yes it is definitely complex and open-system and that is reflected in the complex mineral textures. Generally speaking there are thought to be two primary levels at which ponding and crystallisation occurs, at shallower levels (< 6 km-ish) and and at > 10 km's, constrained by both geophysics and mineral barometry. The peculiarity of this eruption is, as you said, that it involved magmas with different isotopic signatures. The more radiogenic of the two (erupted on the S-rift) was thought to be an endmember that has been gradually infiltrating the plumbing system over many years and leading to an increase in eruption frequency and explosivity. To get to your question about what the mineral zoning can tell us about this complexity, one of the main observations in this study is that the 'endmember' magma is not in isotopic equilibrium with its crystal cargo. Rather, a less radiogenic cargo seems to have been entrained in a more radiogenic melt that generated a reverse, hybrid rim. So it seems mineral zoning (of clinopyroxene at least) records the history of the intruded resident magma but tells us very little about the more radiogenic magma itself, other than it was was probably crystal-poor. Some normally-zoned primitive olivine (as well as the deep seismicity) do suggest that it ascended from the deep plumbing system. Sorry for the long-winded response.. I got a bit carried away. Looking forward to discussing more on Friday as well :)

Sign in to ask a question.