Our best-known example, the Hatepe or Taupo eruption around 1800 years ago, would have created effects visible in China and Rome and fired a devastating 1.5km-high pyroclastic flow that covered the landscape with ash and pumice for a distance of 80km.
To create a new probability model, Professor Mark Bebbington of Massey University will poll experts to form a consensus view.
And he will also build a dataset of historical records from similar volcanoes worldwide.
"We've got quite a number of volcanoes in New Zealand, some of which haven't erupted for quite a while, like Mt Taranaki," he said.
"But there are analogues - volcanoes that look very similar and erupt similar products in similar settings, which have been more active recently."
One example was Mt Merapi in Indonesia. Before it erupted in 2010, killing 353 people, it had a long quiet past, just like Mt Taranaki.
"So the idea is if we can work out how the two records can map on to one another, then we can extend it to other ones worldwide."
Another Massey University scientist, Dr Gert Lube, who specialises in devastating pyroclastic surges, is starting another study with colleagues here and overseas which will use a large-scale eruption simulator based at the Manawatu campus to understand what happens to infrastructure when it is hit by a volcanic current.
Research published last month found a catastrophic volcanic eruption in Auckland's industrial heart could have an economic impact of up to $10 billion in the first year and knock out a large chunk of the city's GDP.