抄録
This paper aims to develop alternative forms of emergency trip destination (emergency shelter) and investigate the evacuee behaviors affecting the shelter choices for future tsunamis and other disasters. The shelter allocation forms are developed using the logistic regression model and neural network model in estimating the probability of evacuee choice of selecting evacuation destinations between public and private shelters. Not only the socio-economic factors that affect the selection decisions are presented, but this paper also intends to analyze the shelter attributes based on evacuee’s perceptions. These perceptions are expected to influence evacuee’s preferences among different alternatives. Data are collected in one of the most recent tsunami affected areas, Phuket, Thailand. Results of this study revealed that the emergency trip destination demand estimation does not provide only information for shelter requirement and shelter space but it can also be used to plan for emergency materials needed.
