Wild AJ, Lindsay JM, Bebbington MS, Clive MA, Wilson TM. 2020. Suitability of quantitative volcanic hazard and risk assessment methods and tools for crisis management in Auckland, New Zealand. Lower Hutt (NZ): GNS Science. 34 p. (GNS Science report; 2020/16.). doi:10.21420/NGM3-5R75.
In the lead up to the next Auckland Volcanic Field (AVF) eruption, it is likely decision-makers will need to determine when and where to call an evacuation. Over the past two decades there has been a growing body of research aimed at exploring the use of quantitative hazard and risk models to support decision-makers. There is a wide range of existing quantitative approaches for assessing risk and its components: hazard, exposure and vulnerability. A detailed understanding of the differences between these approaches is needed in order to select the most appropriate one for a specific decision-making context. In this report quantitative volcanic hazard models for assessing eruption probability and vent location distribution are compared, along with relevant exposure, vulnerability, impact and cost-benefit analysis approaches. A selection framework for each of these components is presented for a range of assessment requirements. The selection framework presented in this report aims to support any risk scientist in selecting an approach to use when developing a quantitative volcanic hazard and risk model. This selection framework is then applied in the context of an AVF eruption evacuation decision-support framework. The criteria for selection of approaches used for the AVF example are informed by the outcomes of a stakeholder engagement workshop that was carried out to understand requirements for AVF decision-support. For the AVF hazard assessment, two pre-developed event tree tools were identified as appropriate, BET_EF and ST_HASSET. The models were both run based on published examples and tested against simulated event data from a 2008 tabletop AVF eruption exercise, Exercise Ruaumoko. This highlighted some functionality differences between BET_EF and ST_HASSET, with BET_EF using predefined thresholds and inputs for monitoring observations, and ST_HASSET relying on subjective selection for any changes in monitoring observations. Due to these differences and the wider experience in New Zealand for both the AVF and other New Zealand volcanoes, BET_EF is selected as a more appropriate quantitative volcanic hazard assessment tool for the purposes of an AVF eruption context. The availability of detailed population and road spatial data for Auckland and suitable vulnerability functions allow for the development of an evacuation decision-support tool for an appropriate scale. The findings from this report will support ongoing research developing a quantitative hazard and risk model to support decision-makers in the event of an evacuation during an AVF event. (auth)