Late last week, on Iceland’s Reykjanes Peninsula, a concerning sequence of earthquakes suddenly turned into a full-blown volcanic crisis. A burst of intense and frequent seismic shaking, accompanied by a convulsing crust, suggested that a huge volume of magma was rapidly burrowing its way toward Svartsengi, the site of a major geothermal power plant and, close by, the coastal town of Grindavík, home to 3,500 people.

The region now nervously sits atop a vast sheet of magma simmering just half a mile belowground. At some point, likely within the coming days, it will probably erupt somewhere along a 10-mile-long line stretching from northeast of the town to a little ways out to sea. The two big questions, exactly where the eruption will get started and how severe it will be, are impossible to answer. But scientists watching the area closely have other questions, too: How did these eruptions go from safe spectacle to a potentially town-smothering danger? And why, after so much recent and violent movement, is the magma now just sitting there?

In some ways, this is exactly what magma here is expected to do. “This activity is very much in line with much Icelandic volcanism,” says Mike Burton, a volcanologist at the University of Manchester. “But the specifics of each crisis are always unique.” And much of the problem this time around has to do with the location this batch of magma has decided to camp in—and why it’s never obvious where an eruption on the peninsula may occur.

When people usually think of eruptions, they picture a mountain-shaped edifice with lava exploding out of a central vent at the summit or bleeding out of its flanks. Iceland does have those sorts of volcanoes, but the Reykjanes Peninsula also specializes in fissure-style eruptions: cracks in the ground that open, often with little warning, when magma below forces its way to the surface.

Magma cracking through the crust creates specific types of earthquakes, and along with the changing shape of the ground you can broadly track where this magma is going and how much magma is involved. But when molten rock reaches the uppermost section of the crust, it can very easily push those rocks aside—and the seismic activity often drops off just prior to an eruption commencing somewhere in the area.

“In the preceding Fagradalsfjall eruptions, we had a pause in the earthquakes some days before the eruption started. Just based on that, I would expect the same, but of course that’s not a guarantee,” says Evgenia Ilyinskaya, a volcanologist at the University of Leeds.

That makes it extremely difficult to know in advance precisely where the next fissure will appear. Fortunately, the seismic storm that has rocked the peninsula in recent days has indicated that it is probably going to emerge within or somewhere very close to Grindavík, a vital clue that ultimately allowed authorities to get people out of harm’s way before any lava saw the sky.

Curiously, the peninsula’s past three eruptions (in 2021, 2022, and this summer) all emerged from closely spaced fissures near to the isolated Fagradalsfjall mountain. These outpourings filled up uninhabited valleys with crimson and tangerine rivers of molten rock and were often watched by curious onlookers from the surrounding hills, poked at by scientists, and celebrated by Icelanders as a showcase of their geologically dynamic country’s natural splendor.

But earlier this month, scientists tracked what appeared to be a huge volume of magma gathering below the Svartsengi area. It quickly rose toward the town of Grindavík last Friday night, stopping just shy of the surface and consequently prompting the town’s swift evacuation.

Even knowing that the next eruption could happen in one of several places on the peninsula, including somewhere a little closer to urban infrastructure, this development still shocked scientists. “The extension of the seismic activity under the town of Grindavík and the shallow waters south of the town did come as a surprise, simply because previous volcanic fissures have not extended so far southwest,” says Þorvaldur Þórðarson, a volcanologist at the University of Iceland.

Why the sudden shift? Scientists suspect that the 2021 eruption kicked off a decades-long period of fissure eruptions across the peninsula; something similar happened 800 years prior. This possible fourth eruption is certainly part of that new era. But it isn’t clear how the magmatism at Fagradalsfjall and Svartsengi are connected. These are not clearly segregated volcanoes, but rather volcanic networks with poorly defined boundaries.

“Some have thoughts that the systems are linked at depth,” says Edward Marshall, a geochemist at the University of Iceland—either directly, with magma flowing between the two subterranean mazes, or indirectly, where they trade pressure. But any geologic connection between Fagradalsfjall and Svartsengi is tenuous at best, making understanding why magma ascends at the former several times, then switches to the latter, a tall order.

This investigative effort is further complicated by the current crisis’s additional idiosyncrasies. Over the past few years, Thorbjörn—a volcanic mound close to the Svartsengi geothermal power station and Grindavík—has occasionally inflated, perhaps due to the movement of magma somewhere below, but this has always ended without incident. The events of the past week “certainly mark a break in that pattern,” says Tom Winder, a volcano seismologist at the University of Cambridge.

Initial estimates hint that the amount of magma involved is more substantial than the peninsula’s past three eruptions, and it also flowed into the Svartsengi area at an astonishing speed. “Why the magma inflow rate appears to be so much higher this time, and indeed where it was sourced from, remains an important open question,” says Winder. Considering the seemingly hefty volume of magma, the potential for a long-lived eruption, or an otherwise very prolific eruption of lava, is high—but paradoxically, as with many eruptions, it could be that only a fraction of that molten rock sees daylight.

That the magma hurriedly rose toward Grindavík late last week, then paused just beneath its now-empty streets, has engendered both curiosity and anxiety. The reasons for this interlude are not quite clear. During the 2021 eruption, there was a three-week gap between the magmatic curtain invading the shallow subsurface and the emergence of the eruption itself. The same may transpire this time. Or it may erupt after you finish reading this article—there is no surefire way to know.

That there will even be an eruption isn’t certain. Presently, based on the proximity of the magma to the surface and the constant seismic rumbling, Iceland’s Meteorological Office suspects that there is a very high likelihood of an eruption, somewhere along that 10-mile-long line of deformed and quaking ground, in the coming days. But there is nevertheless a small chance that the magma cannot find an escape route and remains belowground for the foreseeable future.

Forecasting the nature, timing, and—in this case—location of upcoming volcanic eruptions are exercises in reducing uncertainty. Volcanology, as a research field, has made huge scientific and technological leaps in recent decades, giving researchers an unprecedented level of understanding of the nature of Earth’s magmatic depths.

But just think about weather forecasts. Weather is something scientists can directly sample, observe, and study, and forecasts a few days into the future can be very accurate. But the weather in several weeks’ time cannot be accurately predicted. Volcanologists have to deal with something that, until it erupts, is out of sight—so, for now, forecasting the style, onset, and duration of the next Icelandic eruption is exceedingly difficult.

The peninsula, though, isn’t helpless. The two things Iceland’s scientists and emergency responders needed to do—monitor the magma around the clock while using that data to make sure harm to life and property is minimized—are being efficiently acted upon. The residents of Grindavík are being kept away from the volcanic risk, and a protective wall is being constructed around the Svartsengi geothermal plant to redirect any incoming lava.

Whenever and wherever the outburst in this region begins—if it happens at all—the events of the past week have “brought home how fortunate we have been over the past three years,” says Winder. Sadly, it seems it was only a matter of time before this new eruptive era turned from a delight into a plight.