Image: Eruption in south Iceland, 21 March. Photo by DTM visiting investigator and former postdoc (1993-95) Ingi Th. Bjarnason, now at the University of Iceland, who led an expedition to the site in the early stages of eruption.

As has been widely reported, a volcanic eruption occurred in southern Iceland last week, near the Eyjafjallajokull glacier, beginning shortly before midnight on Saturday, March 21. Rescue teams evacuated 500 people on Sunday from the rural area surrounding the eruption; no injuries or damage to property were reported. Researchers are now concerned that this event could cause a large eruption at an even more powerful volcano, Katla, partially covered by a large icecap. An eruption there could cause major floods and potentially affect people who live nearby.

Alan Linde, Selwyn Sacks, visiting investigator Erik Sturkell, and colleagues have been studying seismic and volcanic activity in Iceland for years, though their instruments are not located near the site of last week’s eruption. Currently, the group operates a strainmeter network on Hekla, one of Iceland’s most active volcanoes. Hekla has erupted approximately every ten years since 1970, prior to which the repose interval was about sixty years. For the last eruption in 2000, the group studied the geometry of surface magma in the area with a wide variety of deformation measurement types including InSAR interferograms, tilt data, and borehole strain data. They determined that the dike that is responsible for surface fissuring extended no more than ~0.5 km in depth, and the magma reservoir depth is ~10 km. Both of these volumes are connected by a conduit. Because this conduit remains fluid during the short interval between eruptions, only a small pressure increase can rupture the crustal seal. Such information is consistent with precursory seismicity at very shallow depths and may apply to other volcanoes that undergo abrupt changes in eruption interval.

Prior to Hekla’s 2000 eruption the group was able to issue a short-term warning. Hekla is aseismic except just before an and during an eruption. A colleague of Linde, Sacks, and Sturkell, Páll Einarsson of the University of Iceland, came into the office to change the paper on a drum recorder for an analog seismic station close to Hekla on the day of the eruption and immediately noticed that there were small earthquakes occurring near the volcano. He alerted Ragnar Stefansson, a collaborator at the Icelandic Meteorological Office, Vedurstofa, who confirmed the earthquakes and alerted civil defense and aviation authorities. Stefansson checked the wind directions where North American-Europe flights would be crossing Iceland, determining that Hekla could create an ash column at those heights within a matter of minutes from eruption. Stefansson then began to watch the strain record from the closest strainmeter, because the group knew from Hekla’s 1991 eruption that there would be slow strain changes, indicating magma movement, about thirty minutes before surface breakout. As soon as he could see those changes occurring, Stefansson made the official call that Hekla would erupt in twenty minutes. The prediction was broadcast on a local news program and planes were diverted. The eruption took place within two minutes of Stafansson’s prediction.

To view more images, click here. For more information on the current volcanic activity in Iceland, click here.