A Kiwi volcanologist is in Tonga to reconstruct what he says was likely the planet's largest eruption in recent times – and one that could have easily been far more deadly.
The spectacular eruption of Hunga Tonga–Hunga Ha'apai on January 15 sent shock waves, tsunamis and ash hundredsof kilometres across the Pacific – killing at least three people, injuring others, and damaging dozens of homes and buildings.
Yet University of Auckland volcanologist Professor Shane Cronin, who's just arrived in the country, said the people of Tonga had still been "incredibly lucky" to avoid what could have happened.
"Had this event occurred in the middle of the night – the casualties would have been enormous."
Instead, the giant ash cloud it produced could be seen by all on what was a clear and sunny afternoon.
As well, Cronin said the "spectacular, cannon-like" blasts the volcano fired out during its first 30 minutes of activity warned people to get away from the ocean – potentially saving many hundreds of lives.
Two smaller tsunami waves, measuring just a few metres high, preceded a main tsunami wave which had a run-up of about 14m – providing another alarm for people too near the water to evacuate.
In another fortunate twist of fate, the ash largely fell into the ocean, rather than all onto the populated islands of Tongatapu and Ha'apai, which were both at the edges of the plume.
Finally, Cronin said, the eruption happened to strike in the wet season.
"With rain falling every day and many heavy rainstorms, any chemical impacts of the ashfall have been rapidly cleansed from the land."
In the two months since, Cronin and other scientists have been able to learn much about how the eruption occurred.
Ash samples that had been sent to New Zealand for analysis had indicated magma within the volcano, which lies mostly submerged about 65km north of Tongatapu, was ripped apart "very violently" - and not in the usual way undersea eruptions unfolded.
"This is more than just magma being erupted into sea-water and reacting – that would have led only to a smaller eruption," Cronin said.
"To make this eruption happen there must have been an extremely rapid squeezing of the magma out of the deep Earth while it was still hot and very runny - while also fracturing and breaking up the upper volcano to allow the water pressure of the overlying ocean deep into the edifice."
The chemistry of the magma was also much richer in iron and magnesium, and more mixed, than what the volcano had previously belched out.
In the eruption this denser, "mafic" magma combined with silica-rich magma above it at an incredibly quick rate.
"This shows us that whatever caused the eruption really mixed the full depth of the magma system and squeezed a huge amount of the whole system out in one go."
One possibility was that a flank collapsed, which could have taken off the top of the volcano and led to fresh magma and parts of its edifice being violently ejected at supersonic speed.
Another was that the "lid" of the volcano – a caldera spanning four to five kilometres in diameter – dropped, causing the magma below to be rapidly squeezed out of the ground on its edges.
A third scenario was a combination of both of those effects - either of which could have triggered the tsunami.
In any case, Cronin said the event – the climax of an episode that kicked off in late December - could prove the largest ever measured in the modern era of volcanology.
"The combination of the violent squeezing of the mixed magma out of the ground while it was still hot - we estimate 1100C - and the intense interaction of the magma with seawater in the fracturing edifice led to the most violent eruption ever recorded," he said.
"During the first 1.5 hours of the eruption, the rise rate of the plume and especially the expansion rate of the ash column allows us to calculate an eruption rate of around one billion kilograms per second - which is about double that of the peak phase of the 1991 Pinatubo eruption."
The environment around the volcano had been dramatically transformed – with major changes observed to Hunga Tonga and Hunga Ha'apai, the two uninhabited islands it sits between.
Cronin said Hunga Tonga has been reduced to just 10 per cent of its original size, as most of it collapsed into the caldera and below the sea, while Hunga Ha'apai had shrunk by about three quarters.
"It has also been stripped completely by violent waves that washed over the island – stripping off between 30 to 50m of weakly consolidated volcanic deposits, down to the oldest and hardest rocks on the island," he said.
"I have never seen anything like it – the island is barely recognisable."
Alongside Tonga's geosciences agency, Cronin will travel from village to village to measure and collect ash and pumice samples, gather recollections from local residents, and identify tsunami current markers.
"This will allow other collaborators to recalibrate tsunami models for the event and also feed into tsunami warning systems."
The trip will also enable him to compare samples with those taken from his earlier field work in Tonga.
"By comparing the 2022 ash with those from past events, we will gain an excellent calibration of the ancient volcanic record," he said.
"I have already identified at least 25 large eruptions have affected the inhabited islands over the last 6500 years."
As for the future, Cronin couldn't rule out more blows.
"This volcano is a very active system. It has had at least five historic eruptions we know about and probably several others."
A reanalysis of ancient, eruptive sequences has suggested the volcano might produce large explosions as frequently as every 500 years.
"This tells us that the volcano will build to eruption again."
The eruption might have also completely removed a hydrothermal system within the volcano's upper edifice that once trapped magmatic gases, while also sealing off salt water.
"Its likely fractured state may continue to be unstable, and we may see further smaller-scale collapses from the edifice, in association with regional earthquakes, or possibly the rise of new magma."
Yet, because the event had effectively emptied out the system's magma, future eruptions might not be as spectacular.
"New, hot rising magma that penetrates to the surface will now encounter little resistance for rise and gas escape - hence future eruptions are more likely to be gentle and effusive," he said.
"There is also the possibility for less-violent forms of magma-water interaction and small explosions to form – similar to the size and scale of events in 2015 and 2009."