An island of glass - the explosive history of Pantelleria, Italy

By @volcanologist

If you’ve heard of Pantelleria, it may be because you’ve seen A Bigger Splash, or read about it in a travel magazine extolling its yet undiscovered virtues (unless you are Madonna or Georgio Armani who have holiday homes there). If you’re a geologist, you may know the name from Pantellerite, the name given to peralkaline rhyolites, which are so abundant on the island. Or perhaps, the enigmatic mineral Aenigmatite, also known as Cossyrite after Cossyra, the ancient name of Pantelleria. What you may not know is that Pantelleria is an island of glass and has a history of catastrophic, caldera forming eruptions. Our recent work has pieced together the island’s explosive past in a new, comprehensive volcanic history.

The last eruption to occur at Pantelleria was a submarine basaltic eruption that occurred 4 km NW of Pantelleria in 1891. The eruption lasted for 9 days and produced floating scoria bombs which eventually exploded and sank. But this gentle, effusive basaltic activity is not typical of the island’s history, instead the island more commonly erupts trachyte and rhyolite, regularly covering the island in volcanic glass of one form or another. For the last 46,000 years, eruptions have mostly been small (strombolian in scale) with local pumice fall deposits and lavas. Around 46,000 years ago, a large eruption occurred generating a hot,sticky pyroclastic density current which covered the island in a welded, rheomorphic ignimbrite known as the Green Tuff. The Green Tuff blanketed the island, and covered the older deposits. The older rock record was known to include eruptions of at least the same size as the Green Tuff, but as their deposits are best exposed in dramatic sea cliffs it has been difficult to piece together this story of explosive eruptions until now.

Panoramic view of a section of the sea cliffs on Pantelleria showing the complicated and largely inaccessible geology: a succession of laterally discontinuous lavas and pumice deposits can be seen, draped by ignimbrites at the top. White and cream units are either non-welded pumice fall or PDC deposits.

Our study brings together field volcanologists, palaeomagnetists and experts in radiometric dating to put together the complete story of the pre-Green Tuff eruptions. The field volcanologists carried out detailed studies of the rocks left behind by these older eruptions, interpreting the rocks in order to understand the processes that formed them. The palaeomagnetists used palaeomagnetic data as a correlation tool, to help match up some of the deposits where these couldn’t be easily traced in the field. Finally, we used Ar/Ar radiometric dating so that we know when the different eruptions occured.

General vertical stratigraphy of ignimbrite-producing eruptions on Pantelleria

We find that the island’s history is dominated by large ignimbrite-forming eruptions. Ignimbrites are the deposits of pyroclastic density currents; dramatic hot flows of gas, ash and rocks which can travel at speeds up to 450 mph (to learn more about ignimbrites and how we use them to reconstruct PDCs, read this blog). Some of the ignimbrites are related to eruptions which resulted in caldera collapse. This is where the magma reservoir underneath the volcano is evacuated so dramatically during an eruption that the roof of the reservoir collapses - the volcanic edifice disappears into the space created by erupting the magma. Caldera collapse eruptions are thought to be some of the biggest, most violent eruptions that a volcano can produce. We found that Pantelleria had experienced at least five of these catastrophic eruptions. The Green Tuff eruption is commonly thought to have ended with a caldera collapse event, but recent work suggests that this isn’t the case. These large eruptions occur every few thousand years up to a gap of around ~40 kyr - we found that there didn’t appear to be any cyclicity or pattern to the timing of the eruptions.

Sea cliffs at Scauri. The bump on the right is a small local centre, draped by ignimbrites.

We also found that in between these large eruptions the island was far from quiet. Small scale eruptions were producing eruption columns that covered local areas with pumice, or perhaps generated small lava flows. These ancient small eruptions are very similar to the activity that has occured on the island since the Green Tuff. So then, is the island currently in a phase of activity so very different to the island’s volcanic past? Probably not. Studies on the last 46,000 years have suggested a decrease in eruptive activity, particularly in the last 15,000 years. But we cannot assume that no large, catastrophic eruption will occur in the future. Importantly though, there is no current evidence that an eruption is expected imminently.

Nina J. Jordan, Silvio G. Rotolo, Rebecca Williams, Fabio Speranza, William C. McIntosh, Michael J. Branney, Stéphane Scaillet, 2018. Explosive eruptive history of Pantelleria, Italy: Repeated caldera collapse and ignimbrite emplacement at a peralkaline volcano. Journal of Volcanology and Geothermal Research, 349, 47-73. https://doi.org/10.1016/j.jvolgeores.2017.09.013.

A PDF is available at http://www.sciencedirect.com/science/article/pii/S0377027317300781; from the University of Leicester or Hull’s repository; or by emailing Nina Jordan or Rebecca Williams.

Kinematic indicators in the Green Tuff Ignimbrite: can they tell us about the timing of caldera collapse?

By Dr Rebecca Williams (@volcanologist) & Jodie Dyble

In the summer of 2014 I have had a Nuffield Foundation student, Jodie, working with me towards a Gold CREST Award, which we blogged about the other week. Here, I’m going to talk a bit about the research she did.

Jodie looked at the Green Tuff Ignimbrite on the island of Pantelleria, Italy. The Green Tuff Ignimbrite is a rheomorphic ignimbrite which was emplaced during an eruption about 45 thousand years ago. An ignimbrite is the deposit from a pyroclastic density current. Rheomorphic means that the deposit was still hot when it was formed, so that the shards of ash welded together and was able to be deformed ductiley. Rheomorphic ignimbrites are common on places like Gran Canaria, in the Canary Islands (where the classic work of Schmincke & Swanson 1967 was done) and the Snake River Plain in the western US. You can get two types of rheomorphism, that which occurs during deposition of the ignimbrite (e.g. the overriding current exerts a shear on the underlying deposit) and rheomorphism which occurs after the deposit has been fully formed (e.g. the deposit starts slumping under gravity). I’m avoiding using primary vs secondary here, as actually the historical meaning of those words and their relative timings can be difficult to disentangle. For a very good, concise overview take a read of (Andrews & Branney 2005). Either way, rheomorphic structures within the deposit like lineations, folds, tension gashes and rotated crystals or clasts, can tell us about this sense of movement. Volcanologists interpret these kinematic indicators in the same way a structural geologist would interpret verging folds, or rotated porphyroclasts in a mylonite (e.g. Passchier & Simpson 1986). You can even determine the direction a pyroclastic density current flowed if you map out these kinematic indicators across the ignimbrite (e.g. Andrews & Branney, 2011).

Schematic diagram of the development of rheomorphic structures in a syndepositional shear zone during the deposition of an ignimbrite. Taken from Andrews & Branney, 2005.

The Green Tuff eruption was said to have been a caldera forming eruption, but the details of this have been debated. Two different calderas have been proposed: the Cinque Denti caldera (Mahood & Hildreth 1986) and the Monastero caldera (Cornette et al. 1983; Civetta et al. 1988). These share the same scarps to the east, west and south but while the Cinque Denti caldera has exposed scarps in the north (the Costa di Zinedi scarp, the Kattibucale scarp and the Cinque Denti scarp), the Monastero caldera has a buried northern scarp. During my PhD on the Green Tuff (Williams 2010; Williams et al. 2014) I found that the Costa di Zinedi scarps, the Kattibucale scarps and the Cinque Denti scarps were extensively draped by the Green Tuff, right down to the bottom of the exposed caldera walls.

The map shows the two different proposed calderas for the Green Tuff eruption. Panoramics and sketches show the draping Green Tuff down the three disputed scarps. Localities used in this study are highlighted. From Williams, 2010.

What Jodie set out to determine this summer was when that draping occurred. My work on the chemical stratigraphy of the Green Tuff already determined that those drapes represented the earliest part of the eruption. So, did caldera collapse happen after the deposition of the Green Tuff and did those drapes represent the rheomorphic slumping of the deposit down a newly formed caldera wall? Or, did the caldera wall exist before the emplacement of the Green Tuff, and those drapes represent a deposit formed by an overriding current? In the field, macro indicators (such as large scale folds) suggested that the deposit slumped down the caldera wall. We went in search of micro kinematic indicators to see if they would tell the same story.

 Some of the micro-kinematic indicators seen in the thin sections from the Green Tuff Ignimbrite, including verging folds and rotated clasts (δ and σ–objects). From Dyble & Williams, 2015.

What Jodie found was compelling evidence for upslope flow in the thin sections that she analysed. Thus, those deposits were formed by the Green Tuff pyroclastic density current flowing up the caldera scarps, depositing and shearing the underlying deposit as it went. Which means that those caldera scarps must have existed before the Green Tuff ignimbrite did, so we support the idea that those scarps had nothing to do with the Green Tuff eruption. We think that’s pretty neat and we’re presenting the work at the Volcanic and Magmatic Studies Group annual conference, which in January 2015 will be held in Norwich. Jodie has already made the poster we’ll be presenting as part of the assessment required to achieve a Gold CREST Award, so we’ve decided to publish that online before the conference. I’d like to thank Jodie for some stellar research this summer, despite only having done 1 year of Sixth Form (AS level) geology (she’s 17!), and answering some questions I’ve been pondering for about 6 years. Hopefully, this data will go into a couple of papers I’m working on too!

Andrews, G. & Branney, M., 2005. Folds, fabrics, and kinematic criteria in rheomorphic ignimbrites of the Snake River Plain, Idaho: Insights into emplacement and flow. In J. Pederson & C. . Dehler, eds. Interior Western United States: Field Guide 6. Bouldor, Colorado: Geological Society of America, pp. 311–327.

Andrews, G.D.M. & Branney, M.J., 2011. Emplacement and rheomorphic deformation of a large, lava-like rhyolitic ignimbrite: Grey’s Landing, southern Idaho. Geological Society of America Bulletin, 123(3-4), pp.725–743.

Civetta, L. et al., 1988. The eruptive history of Pantelleria (Sicily Channel) in the last 50 ka. Bulletin of Volcanology, 50, pp.47–57.

Cornette, Y. et al., 1983. Recent volcanic history of pantelleria: A new interpretation. Journal of Volcanology and Geothermal Research, 17(1-4), pp.361–373.

Dyble, J.A., Williams, R., 2015. Micro kinematic indicators in the Green Tuff Ignimbrite: can they tell us about caldera collapse? VMSG Meeting, Norwich, 5th-7th January 2015. http://dx.doi.org/10.6084/m9.figshare.1160476

Mahood, G. & Hildreth, W., 1986. Geology of the peralkaline volcano at Pantelleria, Strait of Sicily. Bulletin of Volcanology, 48, pp.143–172.

Passchier, C. & Simpson, C., 1986. Porphyroclast systems as kinematic indicators. Journal of Structural Geology, 8(8), pp.831–843.

Schmincke, H. & Swanson, D., 1967. Laminar viscous flowage structures in ash-flow tuffs from Gran Canaria, Canary Islands. The Journal of Geology, 75(6), pp.641–644.

Williams, R., 2010. Emplacement of radial pyroclastic density currents over irregular topography: The chemically-zoned, low aspect-ratio Green Tuff ignimbrite, Pantelleria, Italy. University of Leicester. http://dx.doi.org/10.6084/m9.figshare.789054
Williams, R., Branney, M.J. & Barry, T.L., 2014. Temporal and spatial evolution of a waxing then waning catastrophic density current revealed by chemical mapping. Geology, 42(2), pp.107–110.