Agent-Based Modeling of Behavioral Housing Recovery Following Disasters

In the context of postdisaster recovery, housing reestablishment has been known to be a critical factor as it has ripple effects on the overall timing of recovery. This reestablishment is based on the dynamic behavior of homeowners in the aftermath of disasters. Therefore, to have an effective disaster planning it is essential that policy makers recognize this dynamic aspect of disaster recovery to effectively enhance the recovery process. This entails a thorough understanding of how homeowners react to recovery signals. One of the main sources of these signals comes from neighbors' activities including reconstruction and relocation. The goal of this research is to develop a preliminary temporospatial agent-based model that can account for homeowners' dynamic interactions with their neighbors. These interactions are based on homeowners' equilibrium strategy derived from a game-theoretical approach, which is modeled in a multiagent system framework. The results highlight the significant impact of discount factor and the accuracy of the signals on homeowners' reconstruction decisions as well as formation clusters by reconstructed properties.