Review of Research on Human Behavior in Large Outdoor Fires

Large outdoor fires such as wildfires, wildland urban interface (WUI) fires, urban fires, and informal settlement fires have received increased attention in recent years. In order to develop effective emergency plans to protect people from threats associated with these events, it is important to understand human behavior in large outdoor fires. This paper provides a broad review of existing empirical research and modeling efforts on human behavior in large outdoor fires to provide researchers and practitioners with reference material. In this paper, first, the various methodologies on how information on human behavior is collected from large outdoor fire events is reviewed. Next, available empirical literature that synthesized this information from various survey techniques to paint a picture of the human behavior from large outdoor fire disasters is summarized. The review then moves into how evacuation models have been developed to potentially predict human behavior during various large outdoor fire disasters, and how these models are dependent on the data collection methodologies above. This review concludes with emerging issues and challenges. The review found that interviews and surveys taken from pre-and post-disasters have greatly advanced the understanding of human behavior in large outdoor fires. Yet, past modelling efforts were mainly limited to traffic modelling approaches. The current research direction has shifted to the use of Artificial intelligence (AI) and Geographic Information System (GIS) methods to model evacuation behavior coupled with fire spread. Most of the reviewed research was concerned with wildland-urban interface (WUI) fires, due in part to the greater number of disasters across the globe. As there were less informal settlement fire and urban fire disaster studies in the literature, much more work on those two fire types is needed in the future. There also needs to be broader comparison to all large outdoor fire types to potentially harness similarities amongst all of these types of disasters.