Behavioral and Contextual Factors for the Perceived Vulnerability of Hurricanes

Keywords: Risk perception, hurricane, hazard, machine learning, Florida
Session Type: Virtual Paper Abstract
Day: Friday
Session Start / End Time: 2/25/2022 05:20 PM (Eastern Time (US & Canada)) - 2/25/2022 06:40 PM (Eastern Time (US & Canada))
Room: Virtual 67

Abstract

Hurricane risk perception refers to people’s expected threats or harms from a hurricane. Factors influencing people’s perceived hurricane risks are not fully understood. In this study, we examined residents’ perceived vulnerability towards hurricanes based on behavioral data collected through household surveys as well as contextual setting variables extracted from GIS data sources. Factors of interest included risk area, hurricane knowledge and experience, hazard proximity, housing and neighborhood characteristics, and socioeconomics. Bayesian network models were applied to discover relationships between the perceived hurricane vulnerability and those potential influencing factors. Results of our study contribute to the prediction of people’s hurricane risk perception, which in turn helps the future intervention strategies for evacuation decisions.