Understanding and Simulation of Human Behaviors in areas affected by disasters : from analysis to anticipation – Com2SiCa

While the past forty years have witnessed a significant increase in the number of disasters, they also have shown the intricacy of the events where many causal relations intermingle (physical, biological, technological and human related causes). Those trends should not reverse over the coming years, for numerous risk factors remain: climate change, geopolitical strains, risks associated with technological development and the needs of human societies, population growth and poverty, environment degradation and urban pressure, etc. (URD, 2010). Modern societies, whatever their stages of development, are still inadequately prepared to cope with the complexity and suddenness of disaster events. Populations often do not know how they should take action or react to protect themselves against a threat or danger (CEPRI, 2013). If some behaviors prove to be appropriate, some other ones, unfortunately more numerous (Boyd 1981; ISI 2012), turn out to be inappropriate (stunning, escape towards the danger zone, etc.) or clearly improper (curiosity, goods protection, etc.), as compared to the behaviors expected by the operational stakeholders (Quarantelli 2008) and that are recommended in prevention tools. This partial misconception is not confined solely to populations and policy-makers; it also relates to the difficulties encountered by the research community in identifying the range of behavior patterns actually triggered in the face of a disaster (Crocq 1994), their sequence, dynamics, and interdependence (Provitolo et al. 2015).

One of the major challenges today in the field of the safety and security of populations is therefore to better understand and anticipate both individual and collective human behaviors in the face of threats or disasters, whatever their causes and complexities. Actually, the way we predict or anticipate human response determines how we manage emergency situations (Drury et Reicher, 2011); it helps enhance the resilience of human societies, keep events under control as well as the management of their spatial and temporal extents.

The Com2SiCa project aims at meeting this challenge, by removing conventional barriers and addressing 3 crucial research issues : acquisition (observation/survey) and analysis of real data that could be used to identify, describe and classify human behavior types, whatever the root causes of the disaster; mathematical modeling of the spatio-temporal dynamics of behaviors, which reliability is tested based on actual empirical and statistical relationships; development of a web map interface for simulating human responses, and transfer of the technology to the scientific community, professional bodies in charge of peoples’ safety and security, and the general public for research, training and risk culture purposes. Considering the observation/survey part, analyses will be conducted at 2 different scales, based on: data from global disaster events; and data from two types of local events including a tsunami hitting the Côte d’Azur coastline already affected by the terrorist attack of July 2016, and a technological disaster affecting the city of Le Havre.

In order to achieve these objectives the consortium has gathered expertise on all the themes raised and the related investigation methods, participant observation, and conceptual, mathematical and computational modeling. It includes researchers, i.e. geographers, psychologists, mathematicians, computer scientists, geophysicists, and institutional and operational partners in charge of the safety and security of populations and territories. Their joint preliminary works, supported by the COMUE University Côte d’Azur (IdexJedi) and the French National Center for Scientific Research (CNRS, Peps HuMaIn 2013 and 2014), and which results were published (Provitolo et al. 2015; Cantin et al. 2016; Verdière et al. 2014), promote the success of the program.