Disaster evacuation of the old city of Nicosia

Historic urban centers face unique evacuation challenges due to dense morphology, limited network redundancy, and cultural heritage constraints that restrict infrastructural modification. This study presents an exploratory, decision-support analysis of evacuation dynamics in the walled Old Town of Nicosia, Cyprus, using a high-fidelity digital twin and a hybrid agent-based simulation framework integrating pedestrian and vehicular movement. Rather than modeling a specific hazard, the study evaluates how structural stressors commonly produced by disasters, including exit closures, road blockages, and shelter saturation, influence evacuation clearance, congestion formation, and overall system robustness. Controlled simulation experiments examine the sensitivity of evacuation performance to network disruptions, the effectiveness of existing and expanded shelter configurations, and the role of population familiarity and guidance in constrained historic environments. Results show that evacuation outcomes are highly sensitive to a small number of critical exits and that uncoordinated increases in shelter capacity, without adaptive guidance, may increase congestion and clearance times. Findings also reveal nonlinear congestion dynamics and long-tailed evacuation delays that persist under idealized conditions. The study demonstrates the value of high-resolution digital twins as exploratory tools for identifying structural vulnerabilities, evaluating mitigation strategies, and supporting risk-informed evacuation planning in historic districts exposed to multiple hazards. • Developed a hybrid agent-based simulation coupling pedestrian and vehicle flow models. • Builds on a digital twin of Nicosia Old Town for evacuation decision support analysis. • Evaluates evacuation scenarios to quantify clearance times and congestion hotspots. • Shows evacuation performance is driven by a few critical exits in heritage areas zones. • Finds that added shelter capacity may worsen congestion without adaptive guidance aid.