Shelter location and transportation planning under hurricane conditions

► A stochastic bilevel optimization model for shelter locations is proposed. ► Dynamic user equilibrium is used to describe route choice behavior. ► Large-scale real world example is provided for the state of North Carolina. This paper develops a scenario-based bilevel programming model to optimize the selection of shelter locations with explicit consideration of a range of possible hurricane events and the evacuation needs under each of those events. A realistic case study for the state of North Carolina is presented. Through the case study, we demonstrate (i) the criticality of considering multiple hurricane scenarios in the location of shelters, and; (ii) the importance of considering the transportation demands of all evacuees when selecting locations for public shelters.