Kiwis are all too acquainted with the catastrophic potential of quake-making subduction zones that lie within close striking distance of many of our coastal towns and cities.
Now scientists are investigating a tsunami threat less well understood: volcanic eruptions.
In a new study, led by Niwa and GNS Science, a team of international scientists will model the different ways that volcanoes could send huge waves surging toward land.
The potential scenarios - an explosive eruption or a super-heated pyroclastic flow hitting the ocean, the collapse of a caldera volcano, or even a shockwave in the air striking the sea surface - were as numerous as they were dramatic.
These could play out at offshore volcanoes like White Island in the Bay of Plenty, or in Auckland's own volcanic environment.
Niwa hydrodynamics scientist Dr Emily Lane, who is leading the project alongside GNS geophysicist Dr William Power, said Lake Taupo had also experienced episodes other than the monstrous 180AD eruption that proved New Zealand's biggest in the last 20,000 years.
"If one of these smaller eruptions occurred now, the eruption mightn't have much direct impact on people, but it could cause a tsunami that could badly affect Taupo and other townships by the lake edge."
Further away, there were underwater volcanoes along the Tonga and Kermadec Ridges - one of which may have been responsible for a small tsunami recorded northeast of New Zealand this month.
"So there are many potential volcanic tsunami sources out there," Lane said.
"We really want to know how much of an issue they could be in terms of tsunami hazard."
MODELLING A TSUNAMI
While there had been some research on how volcanoes caused tsunamis, most of it had been focused on finding a trigger that produced a plausible tsunami.
Scientists have long studied underwater explosions - some even considering tsunamis as weapons that could be activated through large detonations - but there was much still to learn.
"We want to delve a little deeper and try to better understand what's going on in there as the tsunami is created," Lane said.
"We want to be able to answer the big questions like 'how does this eruption generate a tsunami?', 'how big could the tsunami be?' and 'do different eruptive processes generate different sorts of tsunamis?'."
Fresh insights into the way water, gases and other elements interacted could even lead to new discoveries in fluid dynamics.
Eruption-charged tsunamis were naturally challenging to study, not least because they happened infrequently and without much warning.
"And given their nature it would be difficult and dangerous to get there and measure many things directly."
That meant the study, supported with an $858,000 Marsden Fund grant, would be mainly based in labs, and drawing upon numerical models rather than real-life events.
The team will build scale models representing features of the volcanoes they want to study, such as underwater air and steam jets to simulate eruptive explosions and column collapses, and heated particles flowing into water to mimic pyroclastic flows.
"While these only capture some aspects of the volcanic eruptions, they're still plenty complicated enough as they are and understanding exactly what is going on in these situations will advance the field a lot."
The numerical modelling will be used to combine information about the depth and properties of water with equations about different eruptions, to calculate how waves might propagate away from the volcano.
"The computer solves those equations step by step through time on a grid that represents the area around the volcano," Lane said.
"From this, we can model how the water responds to the eruption and the sort of tsunami it generates.
"We will use all the information from the laboratory experiments to make sure that the equations we are using to describe what is happening capture the event properly."
Some of the data was being drawn from areas elsewhere that had experienced tsunamis, such as Russia's Lake Karymskoye and the Caribbean island of Montserrat.
"There are lots of different mechanisms involved which makes it very complicated to understand the individual parts.
"In the lab we need to repeat the experiments and know we are getting the same results so that we can measure what is going on.
"For the numerical modelling, we need to understand the physics well enough so that our equations describe the process well."
Lane said the researchers would start with simple scenarios to study and make them more complex as they worked out the mechanisms involved.
Ultimately, she said, it was crucial to understand what threats the events posed.
Lane pointed to the 1883 eruption on the Indonesian volcanic island of Krakatoa - one of the deadliest in history - which killed more than 36,000 people in resulting tsunamis.
"They talk about tsunamis 'expanding the radius of disaster' beyond the distance the eruption can reach directly, and in New Zealand we are a very geologically active area, with lots of volcanoes.
"If we understand the mechanisms that cause these tsunamis better, we will be able to identify and plan for eventualities better.
"We'll know which types of volcanoes are most dangerous and how big an impact they can have.
"We're not the only place in the world where there is the threat of tsunamis generated by volcanoes either, so the research we do with this Marsden study will have benefits far beyond our shores."