Scientists have shed new light on how Auckland's volcanic underbelly is fed with magma and then primed for eruption.
While its more than 50 volcanoes are thought unlikely to erupt again, the Auckland Volcanic Field (AVF) is young and potentially active - with the potential to bring lava, ash and fast-moving surges of hot rock and gas from vents that could form within days.
Now, in a newly-published study, volcanologists turned to a spate of eruptions which took place 30,000 years ago to pin-point two main ways that magma is formed and pulled closer to the surface.
Lead author Professor Ian Smith, of the University of Auckland, said this cluster, involving five eruptions that occurred within less than a century of each other, had been known about for some time.
Yet it was only recently discovered that each event produced magma with distinctly different compositions.
Although the cluster itself was likely triggered by a major tectonic event beneath the North Island, the differing magma compositions showed that each lot had separated from their main source area, before being individually "ripened" prior to eruption, said co-author Professor Shane Cronin, also of the University of Auckland.
These insights allowed the study team to infer that the mantle - a layer between 30km and 2900km beneath the surface - existed in a partially molten state, or that magmas were separated and stored away from where they were created.
In the first case, Cronin explained, eruptions happened when magma was squeezed from a sponge-like mush where it was constantly forming.
In the second, the magma had already separated from its original source site in the mantle, and was being stored as a stand-alone reservoir that fuelled eruptions when triggered by wider processes tied to the tectonic plate boundary running through the country.
"So we think the AVF has two apparent modes of eruption: one where individual events occur on their own timescales, and controlled by magma pressure, and another that is controlled by the regional tectonism," Cronin said.
"Nonetheless, our work suggests that Auckland volcano must have magma ready for an eruption to occur, even if there aren't triggering conditions."
The scientists suspected there was magma present beneath the city right now, and most likely stored within several batches with varying compositions.
"However, we wouldn't expect a series of five to six eruptions, unless there was major tectonic activity in the North Island, such as a subduction zone earthquake," study co-author Dr Marco Brenna, of Otago University, added.
Smith said their findings raised another interesting notion.
That was whether large eruptions hundreds of kilometres away in the Taupō Volcanic Zone could be linked to events in Auckland, given they may be tied to the same major movements along the plate boundary.
The new research at least recognised, Smith said, that the AVF was one of four similar fields – the others being South Auckland, Ngatutura and Okete, to the south – that were all linked to a common tectonic process.
All four had been successively active in the last two million years.
Smith said the study findings would likely have little bearing on how the natural hazard – constantly monitored and researched through the Determining Volcanic Risk in Auckland (DEVORA) programme – was managed.
"However, it is part of our project aimed at understanding how magmas are generated and - most importantly - how quickly they rise to the surface."
He pointed to a recent study that suggested magma could take just days to weeks to rise from the mantle to the surface – leading scientists to estimate that officials could have only between five and 15 days to evacuate perhaps as many as 400,000 people from their homes.
Another new modelling paper found it could take less than 50 hours to safely evacuate several hundred thousand people in the event of an impending local eruption - provided the threat was well understood.
NZ scientists simulating eruptions
Meanwhile, Kiwi scientists have been using innovative experiments to explore how erupting volcanoes can also cause deadly and damaging tsunamis.
In a new project supported by the Marsden Fund, researchers from Niwa, GNS Science and Otago and Auckland universities have been creating miniature underwater eruptions to study how they can trigger giant waves.
Niwa hydrodynamics scientist Dr Emily Lane said some underwater blows occurred in such deep water that there was little noticeable effect, but those that occurred at shallower depths could come with larger impacts at the surface.
The project team was investigating a so-called "sweet spot" where the depth of the water and strength of the eruption combined to make huge and damaging waves.
Their experiments have involved using compressed air, gas and steam in water tanks to simulate waves, which were captured by a network of sensors and video cameras.
The project has also seen glass beads tipped down aerated ramps into water to mimic what happened when pyroclastic flows hurtled down the sides of erupting volcanoes.
While there wasn't strong evidence of past volcanic tsunamis in New Zealand, there were many places – including Lake Taupō, Whakaari/White Island, offshore parts of the Auckland Volcanic Field, and the geologically restless Kermadec Arc - with all the right ingredients to drive them.
"Taupō caldera, Auckland Volcanic Field and offshore Bay of Plenty are places where we're missing information about tsunami hazard from eruptions," Lane said.
"It's a possibility in those locations, so we're working to find out how much of an issue it may be."
Big underwater shakes are responsible for most of the tsunamis that reach New Zealand shores.
While our tsunami hazard planning focuses mostly on earthquake-generated tsunamis, large waves generated from volcanic activity did happen – and they could prove every bit as destructive.
Perhaps the most famous case was 1883's Krakatoa eruption, which generated multiple large tsunamis that killed more than 36,000 people.
More recently, a 1996 eruption near the northern shore of Russia's Lake Karymskoye created waves up to 30m high, and more than 400 people died in a tsunami when the flank of Indonesia's Anak Krakatoa collapsed in 2018.
"We're combining our knowledge of what has happened in different places around the world with these lab experiments and numerical models," Lane said.
"That will help us better prepare for possible future volcanic tsunamis."