Texture-Informed Approach for Hurricane Loss Estimation: How Discounting Neighborhood Texture Leads to Under-Valuing Wind Mitigation

Abstract

The focus of emergency management is shifting from response and recovery to predisaster mitigation. And, a grand challenge in championing for this shift is effectively communicating natural hazard risks and the value of mitigating structures (to reduce those risks). Present tools for loss estimation overlook building-level variations in wind loading induced by the configuration of surrounding buildings, called neighborhood texture. By doing so, such tools under-estimate expected wind-related losses and under-value wind mitigation – significantly in densely built-up areas susceptible to adverse texture effects. In this thesis, those texture effects are incorporated into a widely recognized loss estimation framework. The impacts of local texture are approximated on the recurrence of wind loads on structures. And, in the case study, the benefits of mitigating are re-evaluated for the residential building stock of the hurricaneprone state of Florida – with a focus on five densely populated counties representing a range of exposure to wind-related hazards. Each home is individually assessed with its prevailing local texture evaluated and its occupancy and building characteristics probabilistically assigned. Mitigation measures considered include shutters, straps, and tie downs. For these mitigation measures, the model results yield annualized benefits of $8.1 billion statewide (80% higher than conventional estimates) ranging from $2.0 billion in Miami-Dade County to $56 million in Duval County (respectively, 90% and 100% higher than conventional estimates).