An integrated model of crowd evacuation considering the coupling effects of temperature, density, and fall risk

Crowd evacuation is a critical aspect of emergency management, with ambient temperature playing an important role in shaping evacuation dynamics. Although previous studies have focused on temperature, limited attention has been paid to its interaction with crowd density and fall risk. To address this gap, an integrated evacuation model that incorporates pedestrian movement, pushing detection, fall adjudication, and emotion updating is proposed in this paper. This model introduces a safety factor into pedestrian movement and explicitly examines the coupling among temperature, density, and fall risk. Furthermore, the model is also applied to explore how initial population size, spatial distribution, exit configuration, and obstacles affect falling incidents under varying temperature conditions, and is validated using a real-world stampede case. Results indicate that temperature exerts little influence on the number of pushed individuals but significantly affects the number of falls, with larger crowds more likely to escalate pushing into falling. Compared with concentrated distribution, random distribution reduces falling risk. Under identical exit capacities, the most falls occur at low temperatures and the fewest at high temperatures. With fixed total capacity, central exits reduce fall risk compared with side exits, whereas adding more exits unexpectedly decreases safety due to insufficient flow dispersal. In high-density conditions, obstacles near exits further prolong evacuation time and increase the number of falls. Simulation results closely align with observed data, supporting the validity of the proposed model. Overall, this paper highlights the interplay among temperature, density, and fall risk, offering theoretical insights and practical implications for safer emergency evacuation management.