Enhancing Flood Early Warning Dissemination And Evacuation Response Through Agent-based Modelling

Flood events remain among the most significant natural hazards, often leading to extensive socio-economic damage and loss of life. To lessen and mitigate the risk of such calamities, Flood Early Warning Systems (FEWS) have been developed and adopted to monitor, evaluate, and predict the likelihood of floods (Golding,2022). However, as highlighted by (Fernández-Nóvoa et al., 2024), most research has concentrated on the first two components of FEWS—detection and monitoring, and analysis and forecasting—while less attention has been given to the communication of flood warnings and the corresponding response to risks (Cools et al.,2016). With an aim to improve the general efficacy of FEWS, this work simulates the social and behaviour impacts of different agents or stakeholders during flood response and evacuation using agent-based modelling. The proposed approach integrates key variables inundation rate, warning efficiency (EW), population awareness (A), mobility and response time (Rt). Simulation results underscore that agent-based modelling offers actionable insights for improving risk communication and evacuation strategies, presenting a scalable model for urban flood management. Key recommendations include integrating real-time data analytics, localised risk communication strategies, and adaptive evacuation planning tailored to urban demographics. These findings contribute to the development of resilient flood management frameworks, aligning with global best practices for mitigating the impacts of extreme hydrological events.