Auckland is the most populous city in New Zealand and contributes to over a third of the country’s gross domestic product. The city’s characteristic volcanic cones and explosion craters are evidence of the Auckland Volcanic Field (AVF) on which it is built. The Determining Volcanic Risk in Auckland (DEVORA) program has explored many of the volcanic eruption hazards associated with the AVF, however, before this research there has been no volcanic gas modelling of eruption scenarios. While there has been some initial consideration of the likely impacts of gas emissions from future eruptions, there are only preliminary estimates of gas emission rates. This research aims to model SO₂ gas dispersion from an eruption of the AVF throughout the city in a high pollution risk meteorological event to investigate ground-level concentrations impacted by mesoscale and topographically influenced weather patterns, and subsequent impacts on people, buildings and infrastructure in a worse-case scenario. The model was the Weather Research and Forecasting (WRF) mesoscale gridded weather model was used to provide the meteorological drivers for the Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT). The meteorological conditions were selected based on a pollution and wind analysis of Auckland, as well as synoptic weather regimes correlated with high pollution episodes in order to model a worse case air pollution scenario.

Trajectory, Concentration and Ensemble Concentration models were run on one of the eight DEVORA eruption scenarios (Scenario C). The trajectory and concentration model outputs illustrate the effect of the sea breeze convergence zone on pollutant distribution, and show that communities to the north and northwest of the eruption location are at high risk of volcanic air pollution in the selected meteorological scenario, with a large area exceeding hourly and daily air quality guidelines for SO₂ and aerosol particulate matter, from the National Ambient Air Quality Guidelines, the National Environmental Standards for Air Quality and the Auckland Ambient Air Quality Guidelines. To investigate limitations of the study and the model, two types of uncertainty were analysed: model uncertainty and input parameters such as emission type and rate. Within the model, the ensemble runs suggest that the largest uncertainty exists within the meteorological grid. Input parameter uncertainty also impacted results, however even with a lower emission plume there was a significant area which likely would exceed 24-hour air quality guidelines. Overall, this research suggests a high risk of severe air pollution in worst case meteorological conditions to communities surrounding and downwind of an eruption of Scenario C and develops a methodology for modelling gas dispersion for all the DEVORA scenarios. This research also develops a methodology for air pollution modelling in the AVF that can be applied to other scenarios or emergency response planning.