Modelling the conditions that may lead to phreatic eruptions, with comparison to Whakaari volcano, New Zealand

Phreatic or steam-driven eruptions are the most common type of volcanic eruption in New Zealand. They are notoriously difficult to forecast and have caused fatalities globally. There are two main conceptual models that have been proposed to lead to phreatic eruptions: a) cracking of a hydrothermal seal that is trapping pressurised fluid, and/or b) an injection of hot magmatic gas interacting with cold groundwater that then flashes to steam resulting in rapid fluid overpressure and overburden failure. Testing these hypotheses can help to determine the subsurface processes and timescales leading to an eruption, a possible key to unlocking phreatic eruption forecasting.

We have created generalised 2D numerical models of heat and fluid flow based on conceptual models of Whakaari/White Island volcano in New Zealand. Using CSMP++, we have explored the conditions that allow a build-up of fluids with sufficient overpressure to cause rock failure and potentially initiate an eruption. Our models simulate a sealed hydrothermal system above a magma reservoir. In some models, there is a high permeability zone linking the two which represents a highly fractured zone inferred from fumarole and degassing locations at Whakaari. A subset of these models also includes low permeability zones on either side of the highly fractured zone, which correspond to inferred regions of hydrothermal alteration.

Model results suggest that the permeability of the rock affects how quickly fluids move, but not the long-term pressure distribution. Our models show pressure increases of more than 5 MPa beneath the hydrothermal seal purely due to magmatic heating of groundwater. However, the timescales are on the order of decades. If low-permeability alteration zones are included which limit lateral flow of fluids, pressure increases by 5 MPa from hydrostatic in less than two years. We are now exploring the effects of adding magmatic fluids as well as heat, and will compare results of these models with field and experimental observations from Whakaari volcano.