Author(s): Nasrul Hakim Ghazali; Xilin Xia; Nigel Wright
Linked Author(s):
Keywords: Agent-based modelling evacuation; Flood early warning; Climate resilience; Taman Sri Muda; 2021
Abstract: Accurate and timely warnings do not always translate into effective evacuations. This study combines semi-structured interviews with stakeholders, from the federal government and local authorities to the affected residents, with spatially detailed Agent-Based Modelling (ABM) to examine gaps in Flood Early Warning Systems (FEWS) and evacuation during the 2021 Taman Sri Muda, Shah Alam flood event. Findings highlight: (i) current gaps in the dissemination of early warning, (ii) interpretability gaps between technical warnings and community action, and (iii) heterogeneous community risk perceptions that shape compliance with evacuation advice. These themes are directly translated into model rules that represent how people receive warnings, perceive risk and decide where and when to evacuate as flooding develops. The model is then used as a policy laboratory to explore how changes in dissemination coverage, message actionability, response times, and shelter guidance may improve evacuation outcomes in dense urban environments.
Year: 2026