Toward a Permafrost Vulnerability Index for Critical Infrastructure, Community Resilience and National Security

Abstract

There has been a growth in the number of composite indicator tools used to assess community risk, vulnerability, and resilience, to assist study and policy planning. However, existing research shows that these composite indicators vary extensively in method, selected variables, aggregation methods, and sample size. The result is a plethora of qualitative and quantitative composite indices to choose from. Despite each providing valuable location-based information about specific communities and their qualities, the results of studies, each using disparate methods, cannot easily be integrated for use in decision making, given the different index attributes and study locations. Like many regions in the world, the Arctic is experiencing increased variability in temperatures as a direct consequence of a changing planetary climate. Cascading effects of changes in permafrost are poorly characterized, thus limiting response at multiple scales. We offer that by considering the spatial interaction between the effects of permafrost, infrastructure, and diverse patterns of community characteristics, existing research using different composite indices and frameworks can be augmented. We used a system-science and place-based knowledge approach that accounts for sub-system and cascade impacts through a proximity model of spatial interaction. An estimated ‘permafrost vulnerability surface’ was calculated across Alaska using two existing indices: relevant infrastructure and permafrost extent. The value of this surface in 186 communities and 30 military facilities was extracted and ordered to match the numerical rankings of the Denali Commission in their assessment of permafrost threat, allowing accurate comparison between the permafrost threat ranks and the PVI rankings. The methods behind the PVI provide a tool that can incorporate multiple risk, resilience, and vulnerability indices to aid adaptation planning, especially where large-scale studies with good geographic sample distribution using the same criteria and methods do not exist.