Complex dynamics of single-file pedestrian flow under varying flood water depths

Under the growing threat of flood disasters exacerbated by climate change, understanding human movement in water-inundated environments has become crucial for enhancing evacuation safety and urban resilience. Pedestrian crowds exhibit complex, self-organized behaviors that arise from simple local interactions; however, how these behaviors adapt to water-related constraints remains insufficiently understood. This study presents the first systematic experimental investigation into single-file pedestrian flow under varying water depths (0.35 m, 0.60 m, and 0.90 m), with a no-water scenario serving as the baseline. High-resolution trajectory data are utilized to explore how water depth influences the emergence of collective patterns and adaptive behavioral responses. These findings reveal that increasing water depth alters macroscopic flow regimes (velocity–density-flow relationships), including a shift in optimal density and the emergence of distinct interaction phases governed by hydrodynamic resistance, water–human contact, and interpersonal spacing. At the microscopic level, step length, step frequency, and walking velocity exhibit a two-phase organization relative to headway, indicating consistent coordination patterns across all water conditions. To investigate the underlying adaptation mechanism, this study introduces the concept of anticipated headway, which captures how pedestrians self-regulate through spatial anticipation and temporal delay. This regulatory behavior remains consistent (∼1.41 s) across all water depths, suggesting a robust self-organizing mechanism embedded in human locomotion, even under environmental disturbances. These findings offer empirical evidence on how complex collective dynamics emerge and adapt in perturbed environments, providing new insights into self-organization, behavioral coordination, and model calibration for crowd systems under flood scenarios.