Fitzgerald, R.H.; Wilson, T.M.; Hayes, J.L.; Weir, A.M.; Williams, J.H.; Leonard, G.S. 2023 A stocktake of global volcanic vulnerability models to inform future volcanic risk research in Aotearoa New Zealand. Lower Hutt, N.Z.: GNS Science. GNS Science report 2023/12. 39 p.; doi: 10.21420/NQ3E-9M78
Abstract
There is increasing demand for volcanic impact and risk assessments and associated decision-support applications with recent disasters highlighting information needs, an increasing population exposed to volcanic hazards, and the adoption of the Sendai Framework for Disaster Risk Reduction. Vulnerability models are an essential component of a risk assessment, relating the hazard intensity (e.g. ashfall load) to the exposed societal elements (e.g. buildings), to discern the potential impact, or risk. As of March 2022, there are multiple research programmes in Aotearoa New Zealand that are focused on volcanic risk, with many generating vulnerability models and impact data. A unique opportunity exists to coordinate and leverage these efforts. This report summarises a stocktake of volcanic vulnerability models that are available globally as of March 2022. This stocktake allowed analysis to determine the common hazard intensity metrics (HIMs) and exposure attributes utilised in vulnerability models for different element types. It also allowed identification of gaps and limitations of vulnerability models for certain hazard or asset types. This information has been compiled into a spreadsheet (Appendix 1), contributed to by attendees of a research workshop (held in October 2019 at the University of Canterbury) and supplemented by a literature review – first undertaken in 2019 and updated in March 2022. We report on the methodology used, results and highlight knowledge gaps where future work could be focused. We identified 217 volcanic vulnerability models, covering buildings, critical building infrastructure, roads, rail, bridges, water networks, electricity, telecommunications, economic elements, environment elements, airports, vehicles, maritime elements, agriculture, humans, forestry elements, non-water pipelines (such as gas pipelines), clean-up and societal elements. Of these sectors and elements, buildings were the element with the most available volcanic vulnerability models (26%). The vulnerability models covered many volcanic hazards including tephra, volcanic ballistic projectiles (VBPs), lahar, gas, lava flow, pyroclastic density currents (PDCs), edifice formation, hydrothermal eruptions, volcanic earthquakes, volcanic hazards (non-specific), and volcanic-induced tsunamis. Tephra has the most dedicated vulnerability models with 53%, followed by PDCs with 12%. The majority of models are threshold and function-type models, though indexes and qualitative indicators are also used. We identified 29 different impact metrics in the models reviewed in this study, with damage and loss of service overwhelmingly utilised. Hazard intensity metrics also varied across the reviewed models, with many hazards having multiple metrics. Eighteen different HIMs were identified for tephra fall, five for PDC and volcanic hazards (in general), four for lahar, three for lava and volcanic ballistic projectiles, two for gas and volcanic earthquake respectively, and one each for hydrothermal eruption, edifice formation and volcanic tsunami. This stocktake also highlighted that volcanic vulnerability models do not exist for many hazards, societal elements and types of vulnerability. We provide some commentary within the report to guide future research directions and prioritisation. Vulnerability models by volcanic hazard: While vulnerability models that assess impacts from lahars, tephra, PDCs, VBPs, gas, volcanic tsunami, and volcanic earthquakes exist, there are numerous societal elements that are entirely without vulnerability models for these hazards. This is most often due to a lack of empirical data to draw upon. Additionally, there are no currently known specific volcano vulnerability models for volcanic ground deformation, sector collapse, geothermal fluids, shockwaves, fire following eruption, VBPs from littoral explosions, and volcanic acid rain/gas. However, vulnerability models and/or research from other related fields (earthquake-, geotechnical-, fire- and explosion-engineering; social vulnerability and resilience; environmental impact assessment; etc.) may be able to inform volcanic vulnerability models for some of these hazards. Limited multi-hazard and dynamic vulnerability models: Volcanic eruptions are multi-hazard and typically multi-phase events, which can have compounding impacts on a societal element. There is considerable opportunity for future vulnerability models (and frameworks) to better consider dynamic vulnerability throughout a volcanic disaster, where volcanic hazards, societal decisions (e.g. mitigation decisions) and other factors (e.g. hydrometeorological hazards) may all influence potential impact. We have not critiqued the robustness and efficacy of the volcanic vulnerability models included in the stocktake, as this was beyond the scope of this exercise. However, we note that a) many models are based on limited empirical or analytical vulnerability data; and b) vulnerability models often use very basic statistical approaches in model construction. This is particularly acute when comparing volcanic vulnerability models to comparable model suites from other related natural hazard risk sub-disciplines, such as earthquake or flood risk assessment.
