Hurricane Risk Assessment for Residential Buildings in the Southeastern US Coastal Region in Changing Climate Conditions Using Artificial Neural Networks

Abstract

Hurricanes cause severe damage along US coastal regions, which may be aggravated in changing climate conditions. This paper investigates the impact of climate change on the hurricane risk for residential buildings in the southeastern US coastal region. To understand hurricane characteristics under changing climate conditions, relationships between climate variables and hurricane parameters are investigated. Climate variables considered for the investigation include regional sea surface temperatures and regional relative humidity. The impact on hurricane characteristics is studied with respect to the central pressure difference and the ratio of the number of major hurricanes to the number of total hurricanes. The future hurricane parameters for near- and long-term projections considering a particular climate scenario (RCP8.5) by the Intergovernmental Panel on Climate Change are predicted using nonlinear autoregressive neural networks. Hurricane tracks are developed for four study regions along the southeastern US coast, and associated economic losses are estimated using a natural hazard analysis tool. The result shows that the estimated losses under changing climate conditions increases dramatically in all study regions. The findings of this study indicate the urgency and necessity of considering climate change scenarios in building design codes and standards.