Climate change vulnerability assessment is a complex form of risk assessment which accounts for both geophysical and socio-economic components of risk. In indicator-based vulnerability assessment (IBVA), indicators are used to rank the vulnerabilities of socio-ecological systems (SESs). The predominant aggregation approach in the literature, sometimes based on multi-attribute utility theory (MAUT), typically builds a global-scale, utility function based on weighted summation, to generate rankings. However, the corresponding requirement for additive independence and complete knowledge of system interactions by analyst are rarely if ever satisfied in IBVA.

We build an analogy between the structures of Multi-Criteria Decision Analysis (MCDA) and IBVA problems and show that a set of techniques called Outranking Methods, developed in MCDA to deal with criteria incommensurability, data uncertainty and preference imprecision, offer IBVA a sound alternative to additive or multiplicative aggregation. We reformulate IBVA problems within an outranking framework, define thresholds of difference and use an outranking method, ELECTRE III, to assess the relative vulnerability to heat stress of 15 local government areas in metropolitan Sydney. We find that the ranking outcomes are robust and argue that an outranking approach is better suited for assessments characterized by a mix of qualitative, semi-quantitative and quantitative indicators, threshold effects and uncertainties about the exact relationships between indicators and vulnerability.