Hot or not? Developing a spectrum of indicator-based assessments in approaching vulnerability to climate change

ABSTRACT

Vulnerability assessments to climate change are instruments to support the design of mitigation and adaptation strategies. They are relevant to cities or regions where the impacts may be significant and politicians are keen to avoid economic losses due to maladaptation or inefficient policy courses. Despite the relative simplicity of indicator-based assessments (IbAs), their reliability has been questioned due to their non-robustness and inconsistent outcomes due to changes in modelling assumptions. Nonetheless, politicians still require evidence-based tools to make decisions to signal adaptation and policy approaches. This article develops a range of IbAs through the Ordered WeightedAverage (OWA) approach to construct a decision space for policy-makers. The OWA incorporates the possibility of non-robustness and inconsistency, and improves our understanding about vulnerability. We take Auckland, New Zealand, as a case study and find that if policymakers are risk averse, policy focus is on minimising vulnerability to coastal inundation due to sea level rise. As policymakers ease risk aversion, focus switches to enhancing natural capital and ecosystem services. The OWA reveals the trade-offs prevalent in complex socio-ecological systems and coupled human-infrastructure systems. Therefore, it consolidates a knowledge base for decision-making, which could be adapted internationally and create knowledge spillover and exchange of expertise.

KEYWORDS:

• Robustness
• spatial dispersion
• consistency
• ordered weighted average

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 In this article all indicators are assumed equally important (importance weight equal to 0.05), we focus on the order weights and the effects of changing the assumed ORness values on the decision space for policy-makers.

2 Indicators are normalised as $y=(x-\mathrm{M}\mathrm{i}\mathrm{n}(x\left)\right)/\left(\mathrm{M}\mathrm{a}\mathrm{x}\right(x)-\mathrm{M}\mathrm{i}\mathrm{n}(x\left)\right)$ for those that are positively associated to vulnerability, or as $y=\left(\mathrm{M}\mathrm{a}\mathrm{x}\right(x)-x)/\left(\mathrm{M}\mathrm{a}\mathrm{x}\right(x)-\mathrm{M}\mathrm{i}\mathrm{n}(x\left)\right)$ otherwise.

3 The NZDI is a score for each meshblock (the smallest area for which statistical information is collected) in New Zealand. This index is scaled to have mean 1000 index points and standard deviation 100 index points (Atkinson, Salmond, and Crampton 2014).

4 Table S1 in the Supplementary Material shows the order weights estimated through the OWA programme (Equations 1–3). For an ORness value equal to 0, all the order weight is applied to the indicator located in the 20th position for each CAU. As the ORness value increases, other indicators are assigned higher weights and then become contributors to the formation of the composite vulnerability indicator. In turn, for an ORness value equal to 1, all the order weight is applied to the indicator located in the 1st position for each CAU.