Abstract Details
(2020) Novel Constraints on Magmatic Mushes from Kīlauean Olivines
Wieser P, Edmonds M, Maclennan J, Jenner F, Kunz B & Wheeler J
https://doi.org/10.46427/gold2020.2860
05h: Room 2, Saturday 27th June 05:30 - 05:33
Penny Wieser
View abstracts at 4 conferences in series
Marie Edmonds View all 3 abstracts at Goldschmidt2020 View abstracts at 10 conferences in series
John Maclennan View all 3 abstracts at Goldschmidt2020 View abstracts at 17 conferences in series
Frances E. Jenner View all 5 abstracts at Goldschmidt2020
Barbara E. Kunz View all 2 abstracts at Goldschmidt2020 View abstracts at 3 conferences in series
John Wheeler View abstracts at 5 conferences in series
Marie Edmonds View all 3 abstracts at Goldschmidt2020 View abstracts at 10 conferences in series
John Maclennan View all 3 abstracts at Goldschmidt2020 View abstracts at 17 conferences in series
Frances E. Jenner View all 5 abstracts at Goldschmidt2020
Barbara E. Kunz View all 2 abstracts at Goldschmidt2020 View abstracts at 3 conferences in series
John Wheeler View abstracts at 5 conferences in series
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 Aaron Pietruszka on Monday 22nd June 21:28
Hi Penny, thanks for presenting this exciting talk! Do you think that compaction, deformation, and/or mobilization of an olivine crystal mush beneath the summit of Kilauea could have played a significant role in "making room" for the collapse of the caldera in 2018? If so, how big of a role might it have been compared to withdrawal of magma from the summit reservoir? Best regards, Aaron Pietruszka
Hi Aaron, just to follow up, I think Dennis' point about erosion of these mushes is a really good one. While I don't think the 5-10vol% olivine in the lavas had a big role to play in the total volume collapse, I don't think its a coincidence that we get more and more of these high Fo grains as the eruption proceeds- it seems likely the draining of the summit reservoir disturbed the pile - I really enjoyed the paper Dennis talks about in the Galapagos about the "melt flush - then mush flush" - it seems very analogous to 2018, with the initial flushing of LERZ storage reservoirs, followed by erosion of the summit olivine mush.
Hi Penny, thanks for presenting this exciting talk! Do you think that compaction, deformation, and/or mobilization of an olivine crystal mush beneath the summit of Kilauea could have played a significant role in "making room" for the collapse of the caldera in 2018? If so, how big of a role might it have been compared to withdrawal of magma from the summit reservoir? Best regards, Aaron Pietruszka
Hi Aaron, just to follow up, I think Dennis' point about erosion of these mushes is a really good one. While I don't think the 5-10vol% olivine in the lavas had a big role to play in the total volume collapse, I don't think its a coincidence that we get more and more of these high Fo grains as the eruption proceeds- it seems likely the draining of the summit reservoir disturbed the pile - I really enjoyed the paper Dennis talks about in the Galapagos about the "melt flush - then mush flush" - it seems very analogous to 2018, with the initial flushing of LERZ storage reservoirs, followed by erosion of the summit olivine mush.
Submitted by Kendra Lynn on Tuesday 23rd June 17:54
Hi Penny - thanks for sharing your work in this talk. Your final/main conclusion cautions us when trying to link MI trace elements to eruption characteristics (e.g., fountain height). Wasn't that the main point of the Sides et al. (2014) paper - that more energetic eruptions had more fertile signatures in MIs? So I'm wondering if your results negate the previous study? If not, why not? I'm also curious to know If you compared MI trace element ratios of deformed vs. non deformed olivine? In the 300 yrs of the Keanakakoi Tephra eruptions I found primarily euhedral non-deformed phenocrysts with >Fo87, and didn't see much of the deformed population (confirmed with EBSD). It would seem to me that not all eruptions sample the long-lived mush pile inferred from your detailed work - what do you think? Thanks!
Hi Kendra, just to follow up with a more detailed answer after the session! The vast majority of the "energetic" eruptions do have these high Fo olivines - of the 17 eruptions I show on the Nb/Y figures for Fo>84, 4 are my work, 2 are Robin Tuohys, and the others are Izzy's data directly. Both Izzy and Robin did suggest there was something slightly suspect about the trace element correspondence between melt inclusions and glasses, but alone didn't have enough data to be sure. It was really the LAICPMS work on the ~50 inclusions from each Mauna Ulu period eruption (far more than for any other individual eruption) that gave us enough data to be sure that all these high Fo inclusion populations are identical, regardless of the carrier melt composition, and therefore their crystal cargoes are antecrystic. Also, our work was the first to subdivide these melt inclusion populations by forsterite content (and therefore, the extent of olivine-melt disequilibrium). Because the low Fo olivines do follow a 1:1 trend in Nb/Y glass/melt inclusion space, if you view all eruptions together on a single plot, the trace element disequilibrium is much less clear, because you're seeing both phenocrysts and antecrysts. It seems that instead of the melts hosted within the inclusion marking more "fertile sources", these more energetic eruptions are the ones that can disturb the mush and pick up these olivines. It just so happens that the mean melt inclusion populations sampled in the high Fo eruptions have quite high mean Nb/Y value, while those in the less energetic ones have lower Nb/Y in melt inclusions, as that was the composition of the carrier melt at the time of most of izzy's intracaldera summit eruptions. So if we look at the matrix glass composition, which is actually doing the driving of the eruption, does still correspond to Izzy's story, the more enriched carrier melts near the distinct maxima in Nb/Y corresponding to Kilauea Iki etc- so more enriched carrier melts, sufficient disturbance to plumbing system to scavenge these grains. Less enriched, low Nb/Y intracaldera summit eruptions, less energetic eruptions, mostly phenocrystic olivines grow in the shallow reservoir. But the melt inclusion record is far more obscured for trace elements than the glasses. We did compare deformed and trace elements-about 50% of the "antecrysts" were deformed - some were not, but that is possible with the mush model, in that some olivines in the deforming mush may be held as secondary or tertiary members within the shifting force chains, so not end up visibly strained. With the EBSD data, did you process it with a small color scale and small subgrain segments with the MTEX calc grains? There is some code in the supplement of my EBSD nat comms paper to do this in MTEX- when I've reprocessed Icelandic EBSD maps collected for diffusion studies in this way, we did see distortions that were missed by previous data processing. Happy to chat more about this!
Hi Penny - thanks for sharing your work in this talk. Your final/main conclusion cautions us when trying to link MI trace elements to eruption characteristics (e.g., fountain height). Wasn't that the main point of the Sides et al. (2014) paper - that more energetic eruptions had more fertile signatures in MIs? So I'm wondering if your results negate the previous study? If not, why not? I'm also curious to know If you compared MI trace element ratios of deformed vs. non deformed olivine? In the 300 yrs of the Keanakakoi Tephra eruptions I found primarily euhedral non-deformed phenocrysts with >Fo87, and didn't see much of the deformed population (confirmed with EBSD). It would seem to me that not all eruptions sample the long-lived mush pile inferred from your detailed work - what do you think? Thanks!
Hi Kendra, just to follow up with a more detailed answer after the session! The vast majority of the "energetic" eruptions do have these high Fo olivines - of the 17 eruptions I show on the Nb/Y figures for Fo>84, 4 are my work, 2 are Robin Tuohys, and the others are Izzy's data directly. Both Izzy and Robin did suggest there was something slightly suspect about the trace element correspondence between melt inclusions and glasses, but alone didn't have enough data to be sure. It was really the LAICPMS work on the ~50 inclusions from each Mauna Ulu period eruption (far more than for any other individual eruption) that gave us enough data to be sure that all these high Fo inclusion populations are identical, regardless of the carrier melt composition, and therefore their crystal cargoes are antecrystic. Also, our work was the first to subdivide these melt inclusion populations by forsterite content (and therefore, the extent of olivine-melt disequilibrium). Because the low Fo olivines do follow a 1:1 trend in Nb/Y glass/melt inclusion space, if you view all eruptions together on a single plot, the trace element disequilibrium is much less clear, because you're seeing both phenocrysts and antecrysts. It seems that instead of the melts hosted within the inclusion marking more "fertile sources", these more energetic eruptions are the ones that can disturb the mush and pick up these olivines. It just so happens that the mean melt inclusion populations sampled in the high Fo eruptions have quite high mean Nb/Y value, while those in the less energetic ones have lower Nb/Y in melt inclusions, as that was the composition of the carrier melt at the time of most of izzy's intracaldera summit eruptions. So if we look at the matrix glass composition, which is actually doing the driving of the eruption, does still correspond to Izzy's story, the more enriched carrier melts near the distinct maxima in Nb/Y corresponding to Kilauea Iki etc- so more enriched carrier melts, sufficient disturbance to plumbing system to scavenge these grains. Less enriched, low Nb/Y intracaldera summit eruptions, less energetic eruptions, mostly phenocrystic olivines grow in the shallow reservoir. But the melt inclusion record is far more obscured for trace elements than the glasses. We did compare deformed and trace elements-about 50% of the "antecrysts" were deformed - some were not, but that is possible with the mush model, in that some olivines in the deforming mush may be held as secondary or tertiary members within the shifting force chains, so not end up visibly strained. With the EBSD data, did you process it with a small color scale and small subgrain segments with the MTEX calc grains? There is some code in the supplement of my EBSD nat comms paper to do this in MTEX- when I've reprocessed Icelandic EBSD maps collected for diffusion studies in this way, we did see distortions that were missed by previous data processing. Happy to chat more about this!
Submitted by Keith Putirka on Wednesday 24th June 01:46
Hi Penny - So do MI in the highest Fo olivine grains record moslty, or only late stage interstitial melts in a mush? And I thought MI needed a substantial delta-T to allow their capture by a growing crystal. Can we be sure that the MI in high Fo grains are not captured at shallow depths where delta-T, and crystal growth rates, would be greater? Thanks for the interesting talk. -Keith
Hi Keith, just to follow up, Our model is that these high Fo olivines grow in the magma reservoir, -likely during undercooling involved in the injection of new melts in the reservoir- and then settle into the mush, rather than growing in the mush. Adrien Mourey's presentation gives really nice evidence for a similar model to this-https://goldschmidt.info/2020/abstracts/abstractView?id=2020003748 The melt inclusions are distributed throughout the core and rim of the crystal, and give similar volatile and trace element chemistry throughout. My new work on the 2018 samples reconstructing melt inclusion CO2 using both vapour bubbles and the melt shows that these crystals formed at ~3-5km depths. Similar depths are found based on fluid inclusion barometry. This is supported by the depths estimated from Robin Tuohy (2016, GCA) using rehomogenized inclusions, although its worth noting that the largest dataset by Sides underestimates storage depths because they didn't measure bubble carbon (which, in the F8 samples, accounts for up to 95% of the carbon).
Hi Penny - So do MI in the highest Fo olivine grains record moslty, or only late stage interstitial melts in a mush? And I thought MI needed a substantial delta-T to allow their capture by a growing crystal. Can we be sure that the MI in high Fo grains are not captured at shallow depths where delta-T, and crystal growth rates, would be greater? Thanks for the interesting talk. -Keith
Hi Keith, just to follow up, Our model is that these high Fo olivines grow in the magma reservoir, -likely during undercooling involved in the injection of new melts in the reservoir- and then settle into the mush, rather than growing in the mush. Adrien Mourey's presentation gives really nice evidence for a similar model to this-https://goldschmidt.info/2020/abstracts/abstractView?id=2020003748 The melt inclusions are distributed throughout the core and rim of the crystal, and give similar volatile and trace element chemistry throughout. My new work on the 2018 samples reconstructing melt inclusion CO2 using both vapour bubbles and the melt shows that these crystals formed at ~3-5km depths. Similar depths are found based on fluid inclusion barometry. This is supported by the depths estimated from Robin Tuohy (2016, GCA) using rehomogenized inclusions, although its worth noting that the largest dataset by Sides underestimates storage depths because they didn't measure bubble carbon (which, in the F8 samples, accounts for up to 95% of the carbon).
Sign in to ask a question.