MODELING TSUNAMI EVACUATION OR SHELTER-IN-PLACE ALTERNATIVES FOR WAIKIKI, HAWAII

Using geospatial analytical techniques we estimate surface transportation demands associated with a plausible worst case disaster scenario in Waikiki, Hawaii. We begin with background on Waikiki, illustrating development patterns, population densities, and socio-economic risks. Four groups are considered: 1) tourists; 2) workers; 3) residents of Waikiki; and 4) others living on the island of Oahu who travel to Waikiki. This is the largest tourist area in the state with the highest concentration of employment and economic activity. Approximately 19,000 people live in Waikiki. Many thousand residents of Oahu visit Waikiki because of activity generators (entertainment, restaurants, shopping, beaches, festivals, etc.). We estimate the population size and location based on land use and parcel level building data. We demonstrate how GIS and transportation planning tools are used to identify hazard exposure, risks, and vulnerabilities. We then focus on a disaster created by a tsunami, although the conditions and choices apply to other catastrophic flooding events brought on by storms and coastal flooding. The needs of visitors, workers, and residents and their decisions to evacuate as well as well preferences for certain travel modes are considered as well as likely outcomes in terms of injuries and fatalities. Using survey data, decision tree analysis, and GIS (geographic information systems) estimated evacuation by car, bus, pedestrian, and bicycle modes are estimated. Impacts on local roadways are estimated, including requirements for maintaining LOS B in order to determine shelter-in-place requirements. Among the challenges and risks identified are those arising out of the geography and physical conditions, the density and nature of coastal development, the perceptions and attitudes expressed by those in the area as to what they would do in the event of a catastrophic event, and lack of analysis of the viability of different options focusing on either evacuation or shelter-in-place strategies. We conclude with three contributions: 1) discussion and summary of data requirements and modeling efforts; 2) estimation of both evacuation and shelter-in-place requirements; 3) an agenda for future research and policy analysis necessary to reduce injuries and loss of life associated with catastrophic events in Waikiki.