Forget the flowers: Sometimes April showers could bring May lava flows. A new study suggests that the spectacular eruption of Hawaii’s Kilauea starting in May 2018 might have been triggered by excessive rainfall during the preceding months.

Two years ago, the May-August eruption on the island of Hawaii unleashed rivers of lava that poured into the Pacific and inundated neighborhoods, destroying hundreds of homes. Before that blow-up, however, came a torrent: greater than average rainfall and some extreme precipitation events. The highest 24-hour rainfall total in U.S. history occurred April 14-15 on Kauai, which lies to the northwest of Kilauea. Meteorologists measured an astounding 49.69 inches of rain.

The authors of research published in Nature this week suggest that the extra rainfall permeated the volcano’s subsurface to depths of nearly two miles. Inside the volcano, they contend, that seepage raised the pressure of groundwater held within the porous rock to its highest levels in a half-century. According to the theory, those increased pressures weakened the structure of rock channels inside the volcano.

“Rainfall-induced pressure changes like this could be a trigger, especially in cases where the system is already critically stressed or primed,” says co-author Jamie Farquharson , a volcanologist at the University of Miami (Fla.) “Then even a very small stress change could be enough to initiate new fracturing, creating a new pathway for magma to get to the surface.”

The inherent difficulty of seeing what’s happening deep inside a volcano makes them especially mysterious. In this case, the authors used weather and rainfall records, as well as the ample volcanic monitoring data sets that have been created at Kilauea, which is one of the world’s most-studied volcanoes. They used all this information to model the pressures created by groundwater inside permeable volcanic rock over time.

Farquharson also took into account the timing of past eruptions. Nearly 60 percent of the eruptions since 1790 have occurred during the rainy season (March-August) even though it’s shorter than the dry season, he notes. Observations of subsurface magma migrations also seem to correlate, to some degree, with periods of elevated pore pressure.

“Although no one particular line of inquiry would produce a smoking gun,” Farquharson says, “we’re confident that tying together all these lines of evidence gives us a pretty good picture of how rainfall can be important at Kilauea.”

Michael Manga, a geoscientist at the University of California, Berkeley who isn’t affiliated with the new research, says it’s been well-documented that surface processes can influence volcanoes.

“We’ve known that as glaciers come and go , volcano e s also change their behavior ,” he says. “There are places where heavy rainfall definitely makes earthquakes ; we can see that as the fluids are moving down through the crust the earthquakes are moving down. So the starting point for this hypothesis is firmly rooted in solid science.”

Previous research examining how rain might influence volcanoes suggested impacts in the shallow subsurface layers. One group correlated violent storms with explosion-like seismic events at Mount St. Helens while others investigated different mechanisms for rain-influenced volcanism at Soufrière Hills Volcano at Montserrat and Piton de la Fournaise, a volcano on Réunion Island in the Indian Ocean .

“These questions are really important to try and investigate the connection between what happens at Earth’s surface and deeper underground,” Manga says. But, he adds, he doesn’t agree that rainfall played a significant role in this Kilauea eruption. He argues the authors’ calculations of how much pressure changed as a result of rainfall aren’t large enough to induce an eruption.

“It’s an interesting idea, but there are bigger sources of stress changes,” he says. “Tides made by our moon in a single tidal cycle produce bigger changes of stress than what they calculate. So if tides are not responsible for the eruption, it’s not clear why smaller pressure changes from rainfall would matter.”

Michael Poland , a geophysicist with the U.S. Geological Survey, agreed that the idea that rainfall would increase pore pressure in a volcano and weaken rock to the point of fracture is at least plausible. Poland’s colleagues published a study just last year linking earthquake swarms at California’s Long Valley C aldera to snowmelt over the past 33 years . As for the 2018 Kilauea eruption, however, Poland said, “the rainfall story we don’t think was important in this case.”

Instead, the evidence points to massive pressurization of magma, he said. Before the 2018 event at the Pu'u ‘Ō'ō eruption site, where lava had been seen coming out of the ground for some 35 years, the location saw a decrease in lava volume “like someone kinked a hose,” Poland says. That signaled a pressure buildup back through the system. By April, evidence of the increased pressure on the volcano’s internal plumbing could be seen at the summit, where a lava lake’s rising and falling surface had served as a sort of natural pressure gauge for a decade.

“It rose to the highest level ever, and overflowed out of the vent onto the floor of the crater that it was in,” Poland said. The pressurization was noticeable enough for the Hawaiian Volcano Observatory (HVO) to issue warnings of imminent activity on April 17 and April 24.

If a super-pressured magma system indeed caused the eruption, it would likely be independent of the influence of water-induced pressure changes that weakened the rock. Poland compares the processes to two different methods to pop a balloon. “One way would be to blow it up to the point where the pressure is too much so the balloon explodes,” he says. “The other would be to somehow weaken the balloon itself so it would still pop at lower pressure.” He says all the data that the HVO collected shows that the system pressure most likely caused the 2018 eruption, and that rainfall would have played a small role at most.

Even if rainfall didn’t play a significant role in triggering the Kilauea eruption, that doesn’t preclude the possibility it could impact different mountains, in different climates and geographic locales, through similar processes.

“It would be interesting to find out if this is something that we could see at multiple volcanic systems around the globe, and if we can, what would be the underlying similarities or differences between these various volcanoes,” Farquharson says. “It’s a fascinating area of research but it requires a lot of detailed further study before we can say anything conclusive.’

In other words, it may pay to keep your eyes on the skies but don’t equate eruption forecasting with the weather report quite yet.