A recent and exciting paper within the VIPS community explores the compositional make-up of ocean island volcanoes, and their tipping mechanisms. Lead author Teresa Ubide has written a summary of the work for our VIPS readers. Teresa also met with the blog team for a behind-the-scenes interview on the journey leading up to the publication. We hope you enjoy it as much as we did!
Ocean-island basalts (OIBs) are considered messengers from the deep mantle, however new research shows that their compositions are filtered to a ‘tipping point’ that dominates OIB systems in the Atlantic, Pacific and Indian oceans, and may represent a set of properties that make the magma erupt.
In a recent paper published in Geology, a team of researchers, and friends, spread across three continents show that erupted OIB melts are not as primitive as traditionally considered. When petrography is taken into consideration, whole rock chemistries that appear primitive (>10 wt% MgO) turn out to be cumulates with abundant mafic minerals such as olivine and pyroxene. Most crystal-free samples and melt inclusions have relatively evolved basaltic compositions of only around 5 wt% MgO. These are most carrier melts that erupt.
So, what propels these 5 wt% MgO melts to the surface? Fractional crystallisation of parental OIB magmas in the upper mantle reduces the melt density and increases its volatile content, reaching a ‘tipping point’ of volatile saturation in CO2-bearing magmas like OIBs. The separation of bubbles decreases the density and viscosity of magma, increasing its buoyancy and potentially propelling it to the surface, like the cork popping from a bottle of champagne. This is the eruptible ‘sweet spot’.
How did the journey begin that led to your new Geology publication?
We had been working on samples from El Hierro – the westernmost of the Canary Islands – for a petrological study with Laura Becerril, who did her PhD on the island’s volcanology and is a co-author of the study. We realised that there was significant geochemical variation within the sample set, closely linked with textural variations. Samples that contained large proportions of phenocrysts (mostly olivine and pyroxene) were much more mafic in composition than samples without them. This was particularly evident in cases where we had samples from lava flows and the dykes feeding them. If one of the samples was porphyritic with lots of crystals, and the other one was aphyric without crystals, their compositions were super different, even if they were sampled in close proximity from a single system. Only flow-dyke pairs with similar textures had similar compositions.
The key discovery was to realise that the aphyric samples, free of large crystals that can modify bulk compositions, always had ~5 wt% MgO. That’s not a lot considering the Canary Islands are hotspot-related ocean island basalts (OIB), commonly considered primitive, mantle-derived melts. Then we found that melt inclusion data from the literature also clustered at 5 wt% MgO, even if these melt inclusions were trapped in mafic phases at mantle depths! We thought “Wait a minute, these melts are not that mafic… They’re basaltic, but they’re not very primitive!” Something was filtering OIB magmas to an eruptible tipping point at 5 wt% MgO. That was a big realisation, and we were really excited.
How long has this research been in the works?
My co-authors and I are all PhD friends from Spain, but now we are spread across three continents. We still work together and collaborate, and we’re friends anyway so we are in contact. It’s hard for us to work together because we live in different time-zones, but for Christmas we would always go home (before Covid at least!). In 2018 we said “Okay, let’s work together after Christmas for a few days and get to the bottom of this.”. So we spent a few days together in my family home. We had a great time and worked together exclusively on this paper. We gave it a good push, and it was really fun. Then, each of us traveled back and things slowed down, but that first push was very important. That makes you say “Okay we’ll set up a meeting, we’ll catch up and keep going.”.
Did you have samples and laptops around the dinner table?
Yes, that was great! We all sat around the table in the dining room. We had enough space for the four laptops, notes, papers, and the samples. We didn’t bring a microscope, but we brought thin section scans and geochemical data with us, so we could work with all the information. It was amazing; it wasn’t a long time, but having four people working on something for a few days can be very productive!
What did you enjoy most about the process?
We get on very well, so it’s just wonderful to do things together. It’s easy and fun – getting into the discoveries and discussions… We all enjoy research, and doing research with people you’ve met during your PhD is very special.
Did you face any challenges along the way?
Of course! The reviewer’s comments were really supportive and insightful, but they made us think a lot more about the research and what is actually happening. They made us think about how mantle-derived melts fractionate as they pass through upper mantle plumbing systems, and how that changes the physical and chemical properties of magma. The review process was wonderful because we had a total of four reviewers in two rounds, and the feedback and suggestions helped us a lot. Working on the reviewer’s comments is always a challenge, but this is a great example of a peer review process where we learned a lot.
What is the impact of this research, and can it be applied to eruptions that are happening now?
The first thing is that the make-up of OIB volcanoes is much less primitive than we considered, and it seems to be the same for many OIB systems worldwide… which was unexpected. The other thing is why? Why do many OIB melts have 5 wt% MgO? Why not more?
When olivine and pyroxene crystals are removed via fractional crystallisation, the density of the residual melt decreases, increasing buoyancy relative to the surrounding rock. Unlike in other systems such as mid-ocean ridges where plagioclase fractionation can lead to an increase in the density of residual melt, the density of fractionating OIB melts decreases continuously. As the melt ascends and fractionates, the water content increases. In OIB magmas there is also CO2, and the volatile content can reach saturation at mantle depths of around 10-15 kilometres. This happens after ~50% crystallisation of the original parental magma, which brings down the magnesium content of the residual liquid to 5 wt % MgO, which is what we observed! Bubbles are generated when the tipping point of volatile saturation is reached, which decrease the density of the magma and can propel it to the surface.
If we can detect magma accumulation at the depth where these OIB melts reach the tipping point, or eruptible sweet spot, then we might have an upcoming eruption. That’s what happened in the recent eruption on La Palma, located just north of El Hierro. Deep seismicity started a few years ago, indicating deep intrusion of magma. A week before the eruption started, seismicity was detected below the MOHO, where we expect OIB melts to reach the eruptible tipping point. Our study supports the notion that seismicity coming from that depth may be particularly worrying.
The La Palma eruption destroyed homes, infrastructure and farmland – affecting thousands of people. This was horrible and highlighted the key role of volcano monitoring and hazard mitigation. The Instituto Geográfico Nacional and Insituto Volcanológico de Canarias did an amazing job with all the monitoring and informing evacuation efforts.
Could your findings be used to help mathematical and numerical modelers develop their models?
For models where you need to input the rheological properties of magma and country rocks, our findings show that OIB volcanoes might be less primitive than we thought. The depths at which ascending magmas reach volatile saturation and may tip the system to erupt are also important. It is interesting to explore this depth in other systems. In OIB magmas there is a lot of CO2, which has low solubility compared to water. The system might behave quite differently to say, an arc, where you have more water and less CO2, and magma stagnates at shallower depths in the crust. Interestingly, a recent paper published in Science (Rasmussen et al. 2022) finds a correlation between depth of magma stagnation and magmatic water contents in arc volcanoes! The authors suggest that higher magmatic water contents lead to higher depths of degassing, which results in crystallisation, increased magma viscosity, and magma storage.
Can you tell us a little bit about your current research?
I really like investigating magmatic systems from a high-resolution perspective. I like to investigate what individual crystal populations tell us about the inner workings of volcanoes and the triggers of volcanism, and also what individual melt batches tell us about pre-eruptive processes. With high-resolution geochemistry, we can resolve subtle variations in magmatic histories that would remain unresolved with a bulk rock approach. For example, my former PhD student Ruadhán Magee worked on melt variations throughout individual eruptions at Mount Etna, Sicily, one of the most active volcanoes on Earth. We were able to resolve individual batches of melt mixing on timescales of days during the 2002-03 eruption, and link that to monitoring signs and crystal records of eruptive activity. I love using detailed work to answer big questions, because the big questions need to be broken down and the fundamental processes understood. It’s like a jigsaw puzzle. Of course, every time you try to answer a question you find ten more. It’s always fun.
Do you have a favourite mineral?
I like this question. I love all minerals, but I find pyroxene especially interesting. It is a super mineral because it records lots of processes. It is stable in a range of conditions in mafic to intermediate systems, and sensitive to changes in the magmatic environment. In addition, many elements have a slow diffusion in its structure, so pyroxene is able to preserve processes which may be lost in minerals that have faster diffusivities. So, pyroxene can ‘see’ and ‘preserve’ lots of processes in its growth texture and chemical zoning. It’s great!
Teresa Ubide is a Senior Lecturer in Igneous Petrology/Volcanology at the University of Queensland, Australia, and will be giving an invited talk in the upcoming IAVCEI 2023 Conference Session – Architecture and dynamics of volcanic plumbing systems. To keep up to date with Teresa’s research you can follow her on twitter.
Would you like to be featured on our blog?
We are on the hunt for fellow VIPS enthusiasts to share their stories, whether you’re a student, an early career researcher (ECR) or more experienced researcher – we want to hear from you!
Do you have a new paper to tell the world about? Or have you been involved in some exciting fieldwork recently? Get in touch at firstname.lastname@example.org to tell us what’s going on in your field of VIPS.