Robustness, Uncertainty and Sensitivity Analyses of the TOPSIS Method for Quantitative Climate Change Vulnerability: a Case Study of Flood Damage

Abstract

Multi-criteria decision making (MCDM) techniques have been used to evaluate and rank the spatial flood vulnerability to climate change. However, various sources of uncertainty, such as the determination of evaluation criteria, the assignment of criteria weights and performance values, exist in the application of MCDM methods. In this study, three existing methods were combined to quantify the risk and uncertainties inherent to the process of climate change vulnerability assessment, which is called the TOPSIS-based Robustness-Uncertainty-Sensitivity (RUS) approach. The A1B scenario was used to assess the vulnerability of seven metropolitan cities in South Korea to climate change. Twenty indicators that are closely related to the cause of and deterioration from the flood risk and the resulting damages were selected by two surveys of experts, and the weights of these factors were determined by using the Delphi technique, which can derive the subjective weights. Based on the derived weights, the vulnerability ranking was calculated using the TOPSIS method, one of the most popular MCDM methods. This TOPSIS-based RUS approach was used to analyze the robustness of the vulnerability rankings for the assessed cities, to derive the minimum changed weights of the single and multiple criteria that determine the rank equivalence (or reversal) between any two cities and to check the sensitivities of the performance values to the vulnerability rankings. This study showed the effectiveness of the RUS approach for assessing the vulnerability to climate change, demonstrating the application of flood vulnerability.