Integrating Multi-Criteria Decision Analysis and Uncertainty Quantification for Climate Adaptation

Climate risks are increasing globally due to climate change and socio-economic development. Societies must implement adaptation measures today despite deep uncertainty regarding future climate trajectories, socio-economic pathways, and intervention effectiveness. Because no single strategy performs equally well across all impacts, for instance, protecting infrastructure versus saving lives, decisions depend on which outcomes are prioritized.

Most assessments focus on a single criterion, most often the cost to benefit ratio of measures, overlooking other trade-offs and risking maladaptation. Multi-criteria decision analysis (MCDA) addresses this by explicitly evaluating and weighting multiple objectives. When coupled with probabilistic risk modeling and uncertainty quantification, MCDA can identify strategies that are robust across various futures and stakeholder priorities.

In this project, we develop and test an integrated framework by coupling the new MCDA module of the open-source platform CLIMADA with its uncertainty and sensitivity quantification engine. Using a stylized case study from the Economics of Climate Adaptation (ECA), we explore how methodological and normative choices shape adaptation outcomes through three primary research questions:

• How do different impact units influence the prioritization of adaptation measures? We systematically compare rankings derived from multiple types of impact (e.g., population affected, economic losses, infrastructure exposure) to identify measures that perform consistently well across criteria versus those that are context-specific.

• How does the choice of risk metric affect the evaluation of adaptation measures? We quantify how rankings vary when using expected annual impact versus tail-risk metrics (high-impact, low-likelihood events), clarifying the normative implications of how "risk" is formulated.

• How sensitive and robust are MCDA-derived rankings to the weighting of decision criteria? We explore how results shift when assigning equal weights versus emphasizing specific priorities, making explicit how the assignment of preferences affects evaluations.

Across these questions, we perform an uncertainty and sensitivity analysis that propagates uncertainty through all model components. This allows for a quantitative assessment of decision robustness and identifies the assumptions to which results are most sensitive.

The key contributions of this work include the integration of MCDA with uncertainty analysis in a global modeling platform (CLIMADA); a systematic exploration of how normative modeling choices affect adaptation prioritizations; and a transparent, reproducible workflow for more integrated and value-aware climate-adaptation assessments.