COMPARING MACROSCOPIC AND MICROSCOPIC APPROACHES FOR OPTIMIZING FLOOD EVACUATIONS IN HISTORIC URBAN BUILT ENVIRONMENTS

Worldwide, millions of people are forced to evacuate their homes and areas due to floods. During such dangerous procedures, pedestrians experience difficulties in choosing or applying the right evacuation strategy due to difficult interactions with outdoor environments deeply modified by the floodwater. Flood-prone Historic Urban Built Environments (HUBEs) are of particular interest due to their complex and compact layout, general poor implementation of risk-mitigation strategies, and attractiveness for tourists with a low level of familiarity with spaces. Evacuation planning and management are fundamental to guarantee pedestrians’ safety, and can be pursued through macroscopic or microscopic approaches. The formers are less detailed (as they do not consider single pedestrian features) and can be represented as optimization problems. The latter consider each pedestrian singularly in spite of higher computational costs. This work compares the capabilities of a macroscopic and a microscopic assessment approach for flood evacuation strategies optimization. The macroscopic approach consists of an Integer Linear Program (ILP), which computes optimal evacuation paths depending on pedestrians’ rules in path choice. The microscopic approach simulates the evacuation paths through a commercial, generic software set up on purpose to reproduce pedestrians' flood-related behaviors and crowd dynamics ignored by the ILP. A Risk Index to jointly consider the urban layout, the human factor, and the event intensity, is then developed and applied to evaluate assessment differences among the approaches, thanks to the application to a typological riverine HUBE in the Italian context, characterized by a narrow streets grid. Hydrodynamic conditions refer to a real-world event with a return time of 100 years, to assess floodwater impacts on pedestrian motion and safety. Results demonstrate a general correspondence between the approaches, although the microscopic one seems to be more conservative in terms of pedestrians who can be hindered by floodwaters during the evacuation. The two tested approaches could be also combined in a single tool for technicians of local authorities, including lowtrained ones, to preliminary assess evacuation risks in HUBEs and to propose risk-mitigation strategies, (i.e., architectural integrated solutions, implementation of wayfinding and alert systems).