Evaluation of decision-support tools for coastal flood and erosion control: A multicriteria perspective

Coastal areas face significant challenges due to natural and anthropogenic changes, such as sea level rise, extreme events and coastal erosion. The coastal management requires the consideration of socioeconomic and environmental factors to address these variables. The selection of an appropriate Decision Support Tool (DST) based on decision matrix method plays a crucial role in implementing coastal management strategies to tackle climate change-related issues. This has posed considerable challenges for decision-makers, aligning with the Sustainable Development Goals (SDG). This review provides an overview of the practical experience in the application of DSTs for coastal erosion and flood risk, emphasizing the use of Multi-Criteria Decision Analysis (MCDA). DST choice depends on the coastal archetype, including its geographical features and sociocultural context. The purpose is to clarify how the integration of DSTs maximizes flexibility and supports the implementation of future Decision Support System (DSS) tailored to the needs of coastal cities with development pathways (DP). This review assesses different MCDA methods, highlighting their applicability, utility, and integration in coastal management, while evaluating each method's strengths, weaknesses, and specific applications, with a focus on sustainability and resilience. The review highlights the necessity of expert knowledge in accurately defining criteria and weighting factors to ensure that the chosen MCDA method reflects the complexities of the coastal environment. Depending on the scenario, methods like PROMETHEE and ELECTRE are recommended for their flexibility and robustness in handling complex decision-making processes, especially in data-rich and well-structured environments. In contrast, TOPSIS and AHP are suitable for scenarios with limited information or requiring minimal interaction with decision-makers. For more challenging contexts, where computational resources and expertise are constrained, methods like MAUT, VIKOR, and TODAIM emerge as viable alternatives due to their adaptability and reduced reliance on extensive datasets.