The monthslong eruption of Hawaii’s Kilauea volcano in the spring and summer of 2018 was the most destructive of its recorded history. As the summit crater periodically collapsed, 320,000 Olympic-sized swimming pools’ worth of lava was squeezed out of fissures on its eastern flank. Flaming rivers of molten rock destroyed 700 homes while building new volcanic shorelines, vaporizing lakes and conjuring bizarre weather.
What kick-started such a dramatic sequence? A study published Wednesday in Nature suggests an unusual suspect: rainfall.
In the months before the eruption, Hawaii was inundated by above-average precipitation. This rainfall would have found its way into the pores present within Kilauea’s volcanic rocks. If too much water tries to cram into these pores, the rock fragments from the inside-out.
This new model suggests that, just before the outburst of lava in May 2018, the rocks surrounding the volcano’s cache of magma were experiencing a 47-year peak in pore pressure — enough to break down the walls holding its magma in place until the liquefied rock made a break for it.
The study’s authors, Jamie Farquharson and Falk Amelung, both earth scientists at the University of Miami, note that 60 percent of the volcano’s eruptions since 1790 have occurred during the area’s shorter rainy season rather than its dry season. That makes them suspect that rainfall has played a key role in triggering not just the 2018 eruption, but also many throughout its history.
It’s a bold claim. “Approaching any new theory with a healthy dose of skepticism is important,” said Dr. Farquharson.
But many other scientists aren’t buying it.
There are known cases where environmental changes around a volcano affect its eruptive activity. Evidence suggests that big earthquakes, for example, can trigger eruptions but perhaps only if the volcano is primed to erupt. There is also tentative evidence that rainfall could trigger small earthquakes along some faults.
And the notion that water accumulating in rock cavities can destructively forge new corridors for magma isn’t unreasonable either, said Michael Manga, a geoscientist at the University of California, Berkeley, who wasn’t involved with the study but wrote a perspective piece accompanying the paper. That’s the basic principle of hydraulic fracturing, where pressurized fluids are injected underground to break rocks open and liberate their fossil fuel contents.
But this process simply isn’t needed to explain why Kilauea went haywire in 2018, said Michael Poland, the scientist-in-charge at the U.S. Geological Survey’s Yellowstone Volcano Observatory.
In the weeks before the appearance of those lava-effusing fissures on its eastern flanks, the volcano had been inflating. The lava lake at the summit crater had also overflowed. Both indicated that the magma cache was accumulating pressure. It was fairly clear why: the lava flow out of Kilauea’s Pu‘u ‘O‘o crater had diminished, while the flow of magma into the entire volcano continued.
“It’s sort of like a kink in the hose,” Dr. Poland said. “Eventually, the plumbing system just backed up.” And throughout, those monitoring the volcano saw no changes explained by rainfall.
It did rain a lot before the volcano’s outburst in May. But Kilauea has been erupting for 35 years and experienced plenty of heavy, prolonged rain storms during that time, said Janine Krippner, a volcanologist at the Smithsonian Institution’s Global Volcanism Program. But nothing like the 2018 inferno — the final, furious flourish of that prolonged period of eruptive activity — took place during other wet occasions.
And although this rainfall would have certainly percolated underground, the study’s model is too simplistic to assume it made its way down to the magma’s depth.
The subsurface of Kilauea is a byzantine collection of rocky features, including many impermeable walls of frozen magma. Much of the water there doesn’t move downward, but sideways. “The hydrology of Kilauea is incredibly complex,” Dr. Poland said. “It’s not something that can be well approximated by simple models.”
If the rainwater did fracture layers of rock, starting at the surface and reaching rocks near the magma, this should have produced a downwardly propagating wave of seismicity. This pattern was not seen, said Dr. Manga.
Even if the water did find its way to the depth of the magma cache, the peak stresses in the study’s model are too small to make a difference, even to a volcano ready to erupt. Such stresses are weaker than those caused by the gravitational pull of the moon. “If rainfall is going to trigger a change in behavior, why don’t the tides?” Dr. Manga said.
The study also contradicts a historical record of hundreds of volcanoes around the world that have been saturated by rain throughout recorded history. “We haven’t seen any convincing evidence for rainfall being a trigger for an eruption,” said Dr. Krippner.
The study’s authors are aware that their model simplifies the deep architecture of Kilauea, as is the case for any model trying to virtually reconstruct a complex volcanic system. Dr. Farquharson said they were nevertheless “confident that it is able to reproduce the subsurface movement of water in the rift zone reasonably well.” But as sensors were not tracking fluid pressures at depth during the 2018 eruption, this model’s results are difficult to validate.
This eruption — from its dramatic paroxysm in May 2018 to its abrupt decline that August — “highlights how difficult eruption forecasting is,” Dr. Manga said. Volcanoes are not stand-alone sentinels but components of a complex Earth system, he says, and investigating the forces that could be influencing them is a worthwhile endeavor.