After rumbling ominously for days, Bárðarbunga is finally erupting. Maybe. A small subglacial eruption started on Saturday. Scientists rapidly deployed additional sensors in the region, monitoring the ongoing earthquake swarm, surface deformation, and subglacial eruption.

The Icelandic volcano started erupting on Saturday afternoon, local time. At least, that's what the best-guess is based on seismic signals, as the small-scale eruption is thus far exclusively under the volcano. Similar signals can also be produced by magma-ice interactions without an eruption event, so it's be nice to have confirmation via some of the other monitoring techniques. If this is an eruption, only minor air closures are expected, although the alert level has been bumped to red as a precaution. Monitoring is ongoing while scientists await the eruption either intensifying, dying out, or at least unequivocally existing!

View from Köldukvíslarjökull (forground) to the south across Bárðarbunga, with Grímsvötn to the left. Image credit: Halldór Björnsson.



The whole fuss over Bárðarbunga started when the subglacial volcano began shaking with an entire swarm of earthquakes. Earthquakes are not inherently unusual to Iceland — the whole country is balanced along a divergent boundary of tectonic plates pulling apart at 2 centimeters per year. But this was a massive uptick in the normal seismic activity.

Earthquakes between 1991 and 2014. Earthquakes above magnitude 1.5 (red line); all earthquakes (blue line). Credit IMO.

Seismometer sensitivity has improved over time, so it's important to differentiate between detecting more earthquakes because the equipment is better, and detecting more earthquakes because more earthquakes are happening. To do this, IMO tracks all earthquakes that could be detected by older and newer seismometers (everything magnitude 1.5 or higher, red line). Instead of a few hundred earthquakes in a year, Bárðarbunga is experiencing hundreds of earthquakes in a night, thousands inside a week.


Most of the earthquakes are small, many even less than magnitude 1.5. Only an occasional earthquake surpasses magnitude 3, an earthquake equivalent to the shaking produced by a large truck passing nearby. The earthquakes themselves aren't inherently damaging, especially in such a remote and uninhabited region. What makes the earthquakes threatening is the story they tell of how magma is shifting deep in the rock, far below the ice.

One of the challenging aspects of monitoring this volcano is that it's located under hundreds of meters of ice. Luckily, we have tricks for spying on not just subglacial activity, but subterranean activity. Geophysics is the science of inversion problems, using signals detected at the surface to determine subsurface structure. All these earthquakes are signals, producing seismic waves that propagate through the rock and ice to be detected by the seismometer network. By comparing the arrival times of the different types of seismic waves and how quickly those waves propagate through different materials, geophysicists can build a three-dimensional picture of what is happening, even though they can't see it directly.

Location, depth, and age of earthquakes from August 15th to 20th. 5x vertical exaggeration. Oldest earthquakes are red, mote recent are green. Model credit: IMO/ Bogi B. Björnsson


The earthquake swarm at Bárðarbunga is allowing geophysicists to map the subsurface, watching the main magma chamber flex under changing pressure, and a new intrusive dyke dart east under Dyngjujökull. The dyke is deep, between 5 and 10 kilometers where most of the earthquakes are occurring, although shallower earthquakes on Friday morning (local time) indicate the tip may be branching or rising towards the surface. It's a massive intrusion, stretching over 25 kilometers long already, and continuing to grow both longer and wider. So far, the best estimate is that it contains 0.2 to 0.3 cubic kilometers of magma.

As the dyke grows, it's pulling magma from the main magma chamber. This change in pressure is producing a cluster of earthquakes 2 to 6 kilometers below the caladra rim, which scientists are interpreting as decompression adjustments. So far, none of the earthquakes give any indication of magma rising within the chamber in the lead-up to an eruption, yet the ongoing seismic activity also precludes ruling out the possibility of an eruption in the near future.


Taking the motto that it's better to be safe than sorry, Iceland's officials responded by increasing their monitoring and preparation efforts. Science teams with the Icelandic Meteorological Office (IMO), Coast Guard, department of Civil Protection, and Institute of Earth Sciences are all contributing to monitoring after the uptick in activity at Bárðarbunga and the Vatnajökull glacier.

Scientist installing an additional seismometer on Dyngjujökull on 17 August 2014. Image credit: IMO webcam

A Coast Guard helicopter was quickly tasked with ferry scientists around to set up additional monitoring stations. Within a day of earthquake swarm starting, they had ventured out to install additional seismometers to monitor the earthquake swarm.

Bergur H. Bergsson installing equipment on Dyngjujökull. Image credit: IMO/Þorbjörg Ágústsdóttir.



Another clue to subglacial activity is monitoring the rivers for for any change in the geothermal activity. Thus, out went the hydrologists, setting up stations to have consistent sampling locations and solid preliminary data. So far, no increase in conductivity has been detected, but the monitoring continues.

New GPS installed on Kverkfjöll, with Dyngjujökull in the background. Image credit: IMO/Benedikt G. Ófeigsson

As the magma is moving around, the volcano itself is moving and flexing in response to the changes in pressure. This produces seismic waves, but it also results in surface deformation that can be tracked with GPS stations. The background rate of deformation in Iceland is slow spreading, with the east and west coasts pulling 2 centimeters further apart every year. The surface around Bárðarbunga and Dyngjuháls has done that and far more in the past week, with 14 centimeters displacement relative to places 15 to 20 kilometers away.

TF SIF, the Coast Guard plane tasked with volcano monitoring. Image credit: LHG


Iceland officials know that everyone is wary of their volcanoes after the mess Eyjafjallajökull made of air traffic in 2010, so it didn't take long for them to recall their specially-modified Coast Guard patrol aircraft dedicated to volcano monitoring. The craft is used for flyovers to visually observe the unsettled region, and for survey flights.

On Wednesday, the plane flew its first radar and thermal survey while also checking for any travellers who might be within the evacuation area. Cloudy skies hampered their efforts, but they were able to use the synthetic aperture radar to image both the glacier and the river. This will not only help them monitor the situation as it develops, but will also provide essential pre-event data if the volcano does erupt and melt ice to create a jökulhlaup, or outburst flood.

Radar images of the caldera lake (left), and infrared image of Bárðarbunga (right). Images credit: LHG



At 2:10 pm local time on August 23rd, seismic signals were received that were interpreted as a small eruption starting under approximately 150 to 400 meters of ice. No change has been observed at the surface, but keep an eye on the webcams in case that changes. (But be cautious in identifying the eruption: a dust storm fooled people earlier!)

So far, it's producing no detectable plume or leading to an increase in ash or volcanic gasses above normal for the region. The alert level has been bumped up to red, but widespread airspace closures are unlikely.

Saturday airspace closures. Check with Almannavarnir for updates. Image credit: Almannavarnir


If this is an eruption, lava should be melting ice. However, no abnormal increase in meltwater flow rates or hydrogeological composition has been detected, making the seismic analysis of an eruption more uncertain. The alternative is some form of ice-magma interaction without lava breaching the surface in an eruption.

Assuming this is an eruption, it could take anywhere from zero to twenty hours for the lava to melt through the ice to the surface, or it may remain subglacial for the entire duration of the eruption. With nothing to trap the water, meltwater will flow unobstructed out of the glacier and into the drainage river, avoiding a major outburst flood. If the eruption rate stays small, the meltwater may be produced slowly enough to not overwhelm the river system. While the lava chemistry will determine how explosive an eruption is, if water starts melting and pooling in the vent, the water vapour can mix with the magma and make the eruption more explosive.

Amusingly, this level of uncertainty is pretty much what Bárðarbunga does. Its eruption history is full of unconfirmed subglacial eruptions, where the seismic activity looked like an eruption, but hydrology didn't. This is part of the frustration of geophysics: although it's amazing for letting is peek into processes otherwise hidden from view, interpretation is complicated and requires confirmation with other techniques.



With current information, Bárðarbunga continues to be in a wait-and-see mode, with the alert level bumped up to red as the situation develops. Ongoing seismic activity prevents anyone relaxing, and a small eruption appears to be underway under the ice. For now, scientists have moved ash-tracking radar into the region, and will keep monitoring every scrap of data they can collect until either Bárðarbunga settles down and goes back to sleep, or erupts.

For the latest updates, see the Iceland Meteorological Office official updates. For further reading, check out our previous update on Bárðarbunga, this article by vulcanologist John Stevenson, or this Q&A on troublesome Icelandic volcanoes with Alexandra Witze.