Exploring global sensitivity analysis on a risk-based MCDM/A model to support urban adaptation policies against floods

• Published in: International journal of disaster risk reduction Vol. 73; p. 102898
• Main Authors: da Silva, Lucas Borges Leal, Alencar, Marcelo Hazin, de Almeida, Adiel Teixeira
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 15.04.2022
• Subjects: Climate change; MAUT; MCDM/A; Monte Carlo simulation; Sensitivity analysis; Urban flood risk; MAUT; Climate change; Sensitivity analysis; MCDM/A; Urban flood risk; Monte Carlo simulation
• ISSN: 2212-4209; 2212-4209
• DOI: 10.1016/j.ijdrr.2022.102898

Strategies for reducing flood risks and adapting urban systems involve estimating parameters and conducting difficult trade-offs among human, financial, and environmental issues, which are usually conflicting with each other. This way, multicriteria models are useful as they can aid risk-based decision-making by dealing with all these aspects simultaneously, while the decision-maker (DM) exerts a great influence when establishing his/her preferences. However, this problem is usually associated with uncertainties about defining the variables required, and these certainly affect the credibility of the decision. Hence, sensitivity analysis (SA) is a powerful tool for assessing how changes in these variables lead to robust results. In this context, this paper compiles a SA protocol and this includes using a Monte Carlo Simulation in a multicriteria decision model. It aims to prioritize flood risks in urban areas under climate effects. The model quantifies the risk by using Multi-Attribute Utility Theory and aggregates five criteria: accessibility to public services, economic, human, sanitary conditions, and the need for social assistance. By undertaking a critical analysis, the SA links risk and uncertainty so as to deal with climate risks adequately. It simulates the behavior of three groups of input data: climatic variability, exposure to risk, and the DM's preference statements. Our findings explore graphical and statistical tools to provide the DM with a broad range of evidence with the potential to increase confidence in his/her own decisions. Also, innovative insights emerged from conducting this study which leads us to making suggestions for new improvements in the multicriteria model. [Display omitted] •A global Sensitivity Analysis is proposed to treat flood risk and its uncertainty.•The procedure is applied in a multidimensional model that assesses flood risks.•It aims to enhance the robustness risk ranking from statistical and tabular tools.•A case study detected a correlation between the ranking and climate projections.•12 simulation patterns evidenced how uncertainty and risk attitudes on the results.