WHERE THERE’S FIRE

The eruptions on the Reykanes peninsula in 2021 and 2022 once again propelled Iceland’s volcanoes into the global consciousness. The last time this happened was in 2010, when Eyjafjallajökull’s ash cloud grounded an estimated 10 million

air travellers. By contrast, the latest social media-friendly eruptions, a half-hour drive from the capital, attracted even more tourists to the island’s already strained post-COVID hospitality industry.

Although at any given time there’s a 50/50 chance a volcanic eruption is ongoing in Iceland, the last time the Reykjanes peninsula witnessed an active volcano was in the 12th century. The eruptions were part of a sustained period of volcanic activity that spanned more than 200 years, known to Icelanders as the “Reykjavík Fires.” Could these latest eruptions herald a new era of volcanic activity on the peninsula, near the homes of over two-thirds of Iceland’s population?

In search of answers, I meet with scientists from the University of Iceland’s Earth Sciences Institute and visit Veðurstofa Íslands, Iceland’s Meteorological Office (IMO) home to the volcanic hazards department ten minutes outside Reykjavík.

Dr. Sara Barsotti is the co-ordinator of Volcanic Hazards and operational geophysical monitoring at the IMO. “Volcanic hazards are manifold. The most common, of course, are lava flows, tephra [ash] fallout, and gas emissions, but In Iceland we also have pyroclastic flows and jökulhlaup [glacial floods caused when ice encounters volcanic fluids] in Iceland. These are the most common volcanic hazards. We might have lightning; we might have landslides. There is a whole range of events that might occur before, during, and after an eruption.”

Sara arrived in Iceland in 2013 from the Italian Institute of Geophysics and Volcanology. The focus of her PhD research

was forecasting volcanic ash dispersal in the atmosphere. The department she heads up oversees the Danger of Eruption project (Gosvá). Set up following the disruption to aviation caused by Eyjafjallajökull in 2010, Gosvá is a collaborative research project that brings together the IMO’s Volcanic Hazards Department, the Institute of Earth Sciences at the University of Iceland, and civil protection authorities. Projects focus on specific hazards and form the basis for efforts to mitigate projected dangers. One project, for example, gathered data from soil sections to quantify the volume and distribution of tephra fallout from the eruption at Öræfajökull, Iceland’s highest peak, in 1326. The paper projected forward to calculate the impact this event would have on infrastructure if it happened again today.

The IMO’s monitoring room is the nerve centre for a network of equipment spread across the country. Banks of monitors are watched 24/7 by rotating shifts of scientists on the lookout for anything unusual or slightly above the background noise. There are more than 30 active volcanoes in Iceland. Each is unique and behaves differently. Most eruptions are preceded by earthquake swarms, but that’s not always the case. When Hekla erupted in 2000, it gave just 80 minutes’ warning. “You need a hundred eyes. Since I’ve been here, we have been jumping from Eyjafjallajökull to Askja to Grímsvötn to Reykjanes to Hekla to Katla. It is just continuous.” (Since speaking, we can add Grímsey island off the north coast, which at the time of writing is experiencing an earthquake swarm.)

Iceland’s location on the boundary of the diverging North American and Eurasian continental plates creates a continuous background level of seismic activity. This background itself fluctuates. “There are seasonal trends. For example, an increase in seismicity in Katla is something we see in the summertime. As the ice melts, the geothermal system responds to this availability of water.” As the water turns to steam, it expands, demanding more space, and the increasing pressure triggers seismicity. But, of course, an explosive eruption might also occur during the annual seasonal uptick in activity. To tease out what is significant from the noise, scientists rely on accurate data sets of seismic, deformation, and monitoring data. These records are researchers’ key tool in predicting if a volcano is about to erupt.

Professor Emeritus Páll Einarsson at the University of Iceland has played a key role in producing seismic measurements and data sets. “There were four rather primitive seismographs

in the country when I started out in the 1970s. My first job was to expand that network.” Together with colleagues at the university, Páll built and installed a network of about 50 seismographs. These large and heavy instruments recorded seismic activity onto paper rolls, which were collected by local non-scientists and posted back to the university every week.

Over five decades, Professor Páll has participated in several revolutions that have advanced our understanding of the Earth’s processes. As an undergraduate, he assisted an American scientist who brought the first geodimeter to Iceland. The optical device, now obsolete, replaced the centuries-old method of measurement using tape measures and angles, enabling the accurate measurement of the Earth’s crustal movements and determining that Iceland’s plates diverge by about 19mm annually. The seismicity arising from this spreading is now considered background noise. “All these measurements started to fit together into a general theory. And that came about in the early seventies. Most of my fellow students at Columbia were working on the seismicity of different patches of the Earth, and my part was Iceland. I was trying to fit Iceland into the puzzle.”

The theory, now universally accepted, is the basis for understanding volcanoes and earthquakes. It proposes that the earth’s surface is composed of continental plates floating on the mantle. These plates have been pulling apart and colliding for billions of years. Their motion is driven by the slow convection of the upper mantle. The constant movement means that the plates’ boundaries are dynamic. In subduction zones such as on the west coast of the US, where plates collide, one is forced beneath the other. Here we commonly find earthquake zones. Conversely, at spreading centres, the plates are pulled apart to create a gap in the crust. When enough magma squeezes through this gap to the surface, we get a volcano. This spreading generally happens on the seafloor. But Iceland, sitting smack bang on the divergent boundary of the North American and Eurasian Plates, is an exception.

Another feature marks Iceland out as a special case. For most of its length, the mid-Atlantic plate boundary takes a sort of zigzagging, staircase-like path. As a rule, the horizontal, east-west sections (the flat parts of the staircase) are associated with earthquakes and the vertical, north-south sections with volcanoes. The southern half of Iceland’s plate boundary, including the Reykjanes peninsula, is, however, somewhere in between, at an awkward diagonal angle. Classified as an oblique transform fault, the boundary here is subject to greater, hyperbolic forces capable of creating exceptionally deep faults in the crust. Most volcanoes are fed from magma chambers at shallower depths in the crust, but chemical analysis of the lava from the recent eruptions on Reykjanes has shown that these vents were tapping into magma 20 km [12.4 mi] beneath the surface, where the mantle and crust meet. This affects the composition of lava and gases emitted by the eruption.

Analysis by Dr. Olgeir Sigmarsson and his team at the University of Iceland detected a large volume of halogen gases like chlorine and fluoride degassing from the lava following the eruption. These elements are released when salt is heated and suggest that the lava erupting on the peninsula may be ancient, recycled seabed subducted beneath the crust, perhaps billions of years ago. “These gases bring all the nasty stuff to the surface, chemicals like arsenic, cadmium, thorium, and other rare-earth elements. This is most likely part of what happened during the scatter fires at Laki.” Olgeir is referring to the Laki eruption of 1783, an enormous volcanic event that caused a famine on the island and darkened the skies of Europe, North Africa, and parts of Asia, disrupting the global weather system.

Speaking specifically of Reykjanes, the 5.7 magnitude earthquake that occurred there on February 24, 2021 was a turning point. “We looked at each other and knew there was going to be an eruption,” Sara tells me. “Based on the previous activity, we considered the most likely scenario was an effusive eruption (i.e., lava flowing from a vent), as indeed happened. Though the Reykjanes Svartsengi system, it is worth saying, reaches the tip of the peninsula, where the fissure could open under the sea.” Eruptions in shallow water, known as Surtseyan eruptions, last occurred in the Reykjanes Svartsengi system in 1216. In the days following the earthquake, this represented the worst-case scenario. “Our international airport is nearby, it would be just a matter of minutes for the ash cloud to get there, and Keflavík and Reykjavík would be downwind of an eruption at the tip of the peninsula.”

When lava, at 1,500°C, meets sea water, the reaction is explosive. With a sufficient supply of lava, the steam, ash, and gases produced can form a column of ash and gas to altitudes where it disrupts aviation and can cause tephra fallout. A similar reaction between water and lava was responsible for the disruption to air travel caused by Eyjafjallajökull in 2010. In that case, the water source was a glacier, melting like a

waterfall into the erupting vent below. In the Reykjanes event, by the beginning of March, deformations in the hills and valleys around Fagradalsfjall captured by InSAR analysis confirmed the imminent eruption would be on land.

The most recent eruptions on the peninsula were part of the Krýsuvík volcanic system, one of six systems on the peninsula. Four of these have shown signs of unrest in recent years. The last eruptions on the peninsula, recorded as the “Reykjavík Fires,” occurred in the 12th century and belonged to a period of effusive volcanic activity that lasted over 200 years. The Krýsuvík system stretches to the outskirts of Reykjavík, raising the billion-dollar question: Are we entering a prolonged period of activity and could the Reykjavík area itself witness renewed volcanic activity?

When I put the question to Olgeir he tells me I’ll need to visit the oracle at Delphi. “This is not engineering, where you know the physical laws you are dealing with. Even forecasting weather is easy in comparison. We can’t see the mantle. The only information we get is when the magma comes up. Of course, we would like to be able to make predictions. The only thing we can do is base our thinking on what we know from history. In that sense, our prediction is based on what’s happened before. Is that the best way to predict the future? Maybe, maybe not. You wouldn’t predict someone’s behaviour in their forties based on how they acted at twenty years old. With some kinds of volcanoes, we can measure parameters that are indicative, and these we can predict quite well. For most, we cannot.” I’m learning that every opinion that could in any way be construed as a prediction comes with a caveat. Volcanoes are unpredictable, predictably so.

Less needful of supernatural divination, the ability to forecast volcanic ash clouds will help authorities in Iceland and abroad mitigate the consequences of largescale explosive eruptions. At the IMO, Sara shows me a live video feed of the Askja volcano. The screen is overlaid with a vertical and horizontal axis. There’s no ash plume today, but there is a lonely cloud wandering across the screen. By scrolling the horizontal line to the top of the cloud, a counter tells us the cloud is at an altitude of 2.6km [1.6mi]. Knowing an ash plume’s height enables the IMO’s simulation software to forecast the transport of the ash clouds in the atmosphere and calculate how much ash will fall in any given area.

In addition to disrupting aviation, ash can be highly disruptive to modern infrastructure. The Gosvá paper that modelled a repeat of the Öræfajökull eruption of 1326 found that up to 115km [71mi] of power lines would be at

risk from a phenomenon known as flashover, where the conductivity of the fallen ash causes power cables to short circuit. Transport would also be affected, with roads from Kirkjubæjarklaustur, Southeast Iceland to Seyðisfjörður, East Iceland having a 75-100% chance of receiving more than 3mm of ash, at which point visibility drops to almost zero and roads become impassable.

As I talk with Sara, I’m reminded of how, in every disaster movie I’ve ever seen, the volcanologist’s or scientist’s warnings are ignored. (“You’re crazy. We can’t close the beaches cause some drunk kids saw a shark!”) I ask Sara if she relates to the archetype. “No, no, no, I think most of the time we have a strong connection with stakeholders. Of course, any kind of mitigation action takes time. It’s something that needs to be planned and worked out. And that’s why we talk about long-term hazard assessment because it is, in a way, the first step to improving preparedness for events that might happen in the future.

“Of course, we also plan how to react to different events, and we try to get more and more organised.” The last caveat goes to Olgeir: “Iceland is a fishing nation, a hunting nation. We don’t predict too much. We are not an agricultural nation; you don’t plant your seeds in the spring and anticipate you will have a harvest in the fall. Here you just go hunting. And if not today, then tomorrow. You have to face up to it. You have to get your bread.” The eruptions on the peninsula led to a period of intense activity among Iceland’s small community of earth scientists, at the University of Iceland, and at the IMO. Many were pulled from their own research projects into an all-hands-on-deck effort that has led to two papers being published in the prestigious scientific journal Nature, an extraordinary achievement that reflects an exhaustive collective effort. Among the scientists I met, I felt a palpable sense of urgency as they talked about their research. It may be because, as I heard over and again, there is so much we don’t know about volcanoes and the processes that power them. It is something of a paradox that the science studying the processes that gave birth to the earth is itself such a young and exciting field, with so much left to discover. The key paper proposing the notion of plate tectonics, the foundation of the Earth sciences, was not published until 1965. Sara, Páll, Olgeir, and their colleagues at IMO and the university are adding to an incomplete understanding that contributes to the realworld need for quick answers. Hopefully, the consequences of the next eruption can be mitigated, the one thing we can be certain will happen.

InSAR (Interferometric Synthetic Aperture Radar) is a powerful tool used for tracking magma intrusions by measuring how the surface of the ground deforms during episodes of volcanic unrest. The images are created by satellites orbiting over 500 km [310mi] above Earth. An image is taken of the target area, and then at a later point in time (hours, days, or even months later), the satellite captures the same terrain from precisely the same distance and position. By calculating the phase difference between the two images, a deformation map known as an interferogram is produced. The image, a psychedelic tie-dye-like map, visualises any changes in the topography between the two images. The areas with the most coloured fringes correspond to the areas of terrain with the greatest movement. The images, accurate to a few millimetres, allow scientists to measure the inflation of potentially dangerous volcanoes and can also track the movement of dike intrusions (magma injections) by measuring deformation in the terrain above. Before an eruption, rising magma creates a telltale butterfly pattern that indicates the likely location of an impending eruption. For more information, check out icelandvolcanoes.is, where you will find the latest updates and information as well as the Gosvá reports.

This is a COSMO-SkyMed interferogram spanning from July 27, 2022 to August 4, 2022. Each coloured fringe represents 1.55cm of deformation in the satellite’s line of sight. The modelled dike location (prior to the August 2022 eruption) is displayed as the black line and the location of the eruption that commenced on August 3, 2022 as the red star. The larger lobe to the northwest of the dike shows about 25cm of deformation in the satellite’s line of sight. This butterfly pattern displayed here, showing an elongated subsidence signal and 2 lobes extending outwards perpendicular from this, is typical of a dike intrusion. The interferogram was processed by Dr. Michelle Parks (Icelandic Meteorological Office) using data acquired by COSMO-SkyMed satellites (Italian Space Agency, 2022) provided through the Iceland Volcanoes Supersite.