Flood shelter identification using a mechanics-based approach and triangular fuzzy analytical hierarchy process

Flood events continue to cause substantial damage to people, infrastructure and communities. Effective evacuation planning is essential, and the identification of safe shelters is a key element in protecting people and property. Shelter selection depends on a reliable flood risk assessment that considers building characteristics, environmental conditions and occupant safety. This study proposes a three-stage framework for assessing and identifying shelter risks under flood disasters. The framework considers 3 criteria, namely hazard, exposure and vulnerability, and 14 subcriteria. These criteria and sub-criteria were evaluated using a hydrodynamic model, topographical data and an innovative application of a mechanics-based method (MBM) that assesses flood hazard from modelled water depth and velocity data. The relative importance of the criteria and sub-criteria was estimated with the Analytical Hierarchy Process (AHP) and Triangular Fuzzy Number-based AHP (TFN-AHP) approaches. Furthermore, the Geographical Detector Model (GDM) was employed to investigate the driving forces behind the contributing factors. The frameworks have successfully assessed 27 candidate shelters in Pontypridd, UK. The results showed that the MBM method was scientific for the hazard assessment, and TFN-AHP was more stable than the AHP method, effectively reducing the uncertainty of the assessment. Among the 27 candidate shelters, these frameworks have successfully recommended 5 candidate shelters with lower risk in the city centre and Treforest areas and identify 6 potential shelters to have higher risks and deemed unsuitable. The proposed framework supports safer shelter planning and contributes to improved people safety, better environmental risk management and stronger urban flood resilience.