Skip to content

Operational ashfall forecasting in New Zealand: Current status and future perspectives

Paul A. Jarvis 1, Victoria Miller2, Josh Hayes1, Tony Hurst1, Rosa Trancoso3, Oliver Lamb4, Aleksandr Spesivtsev1, Anna Perttu5, Andrew T. Prata6, Valentin Freret-Lorgeril7

  • Affiliations: 1GNS Science | Te Pū Ao, Lower Hutt, New Zealand; 2GNS Science | Te Pū Ao, Auckland, New Zealand; 3MetService, Wellington, New Zealand; 4GNS Science | Te Pū Ao, Taupō, New Zealand; 5Volcanic Risk Solutions, Massey University, Palmerston North, New Zealand; 6CSIRO Environment, Clayton, Australia; 7Laboratoire Magmas et Volcans, Université Clermont Auvergne, Clermont-Ferrand, France

  • Presentation type: Talk [Invited]

  • Presentation time: Thursday 09:00 - 09:15, Room R280

  • Programme No: 6.4.3

  • Theme 6 > Session 4

Abstract

All of Aotearoa-New Zealand's (NZ's) active volcanoes can erupt explosively, potentially dispersing ash across much of the country. Ash can be highly disruptive and damaging to the agricultural industry; critical infrastructure, including transport, power and water supplies; and human health. During eruptions, rapid and reliable ashfall forecasts are necessary to inform impact assessments and enable infrastructure and emergency managers to make informed decisions. GNS Science, through the GeoNet programme, has responsibility for monitoring NZ's volcanoes, as well as providing associated advice products, including ashfall forecasts. Through collaboration with the NZ MetService, ashfall simulations are performed every 6-hours for various eruption scenarios at 10 different eruptive locations, using the most up-to-date high-resolution meteorological forecast. From the model outputs, ashfall forecast maps are automatically generated and made available to duty personnel through a dashboard. In the event of an eruption, duty personnel can select the most appropriate scenario and insert the relevant map into an ashfall forecast template document, which can then be disseminated to emergency and infrastructure managers, as well as the public and media. Limitations of the current system include inefficiency in updating forecasts with observed data, an inability to produce probabilistic forecasts in response, and a lack of uncertainty information in the disseminated forecast. Current and future work to try and address this includes the development of new tools to rapidly characterise eruption source parameters in near-real-time, as well as continued engagement with next- and end-users to ensure that disseminated forecasts meet user-needs in regard to accessibility and usability.