Magma levels
reach an all-time high
beneath Iceland

The ground beneath the Svartsengi power station on Iceland's Reykjanes Peninsula is holding more magma than at any point since this eruptive cycle began. Scientists at the Icelandic Meteorological Office have confirmed that model calculations based on ground deformation measurements now show over 24 million cubic metres stored at approximately 4km depth — surpassing the previous record of 22.6 million m³ set earlier in this same cycle.

A record high with a slow pulse

The headline figure conceals a striking paradox. While the total volume is unprecedented, the rate at which new magma is arriving is the slowest ever measured in this series. The current inflow is estimated at around 1 cubic metre per second — a trickle compared to earlier phases of the cycle. The result is a system under enormous pressure, filling up slowly, with no clear sign of when it will release.

Geoscientists describe this combination — maximum volume, minimum inflow rate — as one that dramatically increases forecasting uncertainty. The timing of a future magma intrusion or eruption cannot be constrained to better than several months.

The chart tells its own story. The curve that began in July 2025 has climbed steadily — with some variability — past the lower trigger threshold in late September, past the previous inter-eruption record in mid-March, and is now in entirely uncharted territory for this cycle.

How this cycle compares to all the others

Since the Sundhnúkur crater row first erupted in December 2023, the peninsula has experienced repeated cycles of magma accumulation followed by intrusion and eruption. Each cycle has accumulated a different volume before triggering. The chart below places the current cycle in context.

The current accumulation stands clear above all previous cycles. The earlier pattern showed a rough range of 17–23 million m³ required to trigger an event since March 2024, but the system has now far exceeded even the upper end of that range without erupting — which itself is new information about how the plumbing system beneath Svartsengi is behaving.

The Krafla lesson: slow does not mean safe

The IMO's March 31 update introduces a significant historical comparison that had not previously featured in these bulletins. Scientists point to the Krafla volcanic series of 1975–1984 in northeast Iceland as a cautionary parallel.

Over nine years, Krafla experienced 12 magma flow events, nine of which broke the surface as eruptions. The average inflow rate across the series was around 0.75 cubic metres per second. From late 1981 onward, that rate dropped sharply — and remained low for nearly three years. Yet when the final eruption eventually arrived in August 1984, it was the largest of the entire series.

The IMO is explicit that Krafla and the Reykjanes Peninsula are geologically different systems. But the comparison carries a clear message: a prolonged period of slow magma inflow is not evidence of a volcanic system winding down. It may simply be the prelude to a larger event.

What would an eruption look like?

The most likely scenario, unchanged across many months of updates, is a magma intrusion from the Svartsengi reservoir toward the Sundhnúkur crater row — the same corridor where all previous eruptions in this cycle have occurred. The probable fissure zone lies between Stóra-Skógfell and Sýlingarfell, though the IMO notes the risk zone could extend from Grindavík in the south to an area northwest of Keili mountain.

Because of the record volume now stored, the IMO specifically flags that the next eruption could be more extensive than those seen previously, if much or all of that magma reaches the surface.

When will this cycle end?

The IMO addresses this question directly for the first time in this update, laying out the criteria that would need to be met before the volcanic alert level could be stepped down to zero — the point at which scientists could formally declare the eruption series over.

Those criteria require sustained, unambiguous signals across multiple measurement streams over an extended period. A pause in accumulation of even several months would not be sufficient. And on a geological timescale, it is worth noting that eruption episodes on the Reykjanes Peninsula have historically lasted for decades or centuries, with pauses between individual events that could easily be misread as an ending.

For now, residents and visitors near Grindavík should remain prepared for short-notice evacuation. The IMO's 24-hour monitoring continues, and the hazard assessment has been extended to run until 30 June 2026.