Climate Change Vulnerability Assessment by Outranking Methods: Heat Stress in Sydney (original) (raw)

2012

Abstract

ABSTRACT Climate Change Vulnerability Assessment (CCVA) allows policy makers to incorporate climate futures in planning. Indicator-based vulnerability rankings are widely-used and sometimes contested forms of risk assessment. One of the challenges of indicator-based vulnerability metrics is to combine multiple indicators from both the biophysical and socio-economic domains into indices of vulnerability for a given climatic stress, to generate a ranking of vulnerabilities. The predominant aggregation approach in the literature is based on multiple-attribute utility theory (MAUT). Scholars have critiqued this approach as theoretically flawed because a) it requires the conversion of incommensurable indicators into comparable scales; b) does not account for different types of uncertainty often present in the analysis and c) produces only a linear, threshold-free scaling of the effects of an indicator on vulnerability. In this paper, we develop an analogy between multi-criteria decision-analysis (MCDA) and Indicator-based Vulnerability Assessment (IBVA) showing that the two problems are structurally similar and share the features of incommensurability and fuzzy data. We argue that a set of techniques called Outranking Methods, based on a Condorcet approach and developed in MCDA, offer a more theoretically sound approach to aggregation. Hence, we redefine IBVA as a fuzzy problem and introduce concepts of thresholds of vulnerability difference, mirrored on MCDA, to quantify fuzziness. We use an outranking method, ELECTRE III (developed by Roy and his colleagues), to assess the vulnerability to heat stress of 15 local government areas in metropolitan Sydney, where heat-related mortality may be a significant public health threat. We apply different thresholds in the model, compare the resulting rankings and find that the use of dominance thresholds generates significantly different rankings compared to MAUT. Finally, we suggest further developments of the proposed methodology to integrate non-linearities in the relationship between vulnerability and indicator, and apply the outranking framework to a multi stakeholder setting of vulnerability assessment.

Fahim Tonmoy hasn't uploaded this paper.

Let Fahim know you want this paper to be uploaded.

Ask for this paper to be uploaded.