Probabilistic Modeling of Hurricane-Induced Debris Impacts for Coastal Community Resilience Analysis

Abstract

Climate disasters such as hurricanes significantly impact coastal communities, posing critical challenges to their resilience. Besides direct economic and social losses, coastal communities suffer from indirect cascading consequences of these extreme events. In particular, debris-related impacts pose significant economic burdens, while also resulting in cascading consequences. These consequences include, for example, structural damage due to debris impact, functionality impairment to transportation networks affecting access to emergency facilities, and delayed recovery of other systems. As a result, there is a need to better understand and model debris and its uncertain impacts on coastal communities in the face of storm events. This paper puts forward a probabilistic framework to evaluate hurricane-induced debris and its impacts at the community scale, which is essential in conducting a comprehensive resilience analysis of coastal communities. This framework poses interdependent probabilistic models spanning from the spatial estimation of debris presence and volume for hurricane events, to debris-induced physical damages and network-level performance impacts (considering transportation infrastructure as an illustration). Moreover, this study uses Monte Carlo approach to conduct simulations, which is accelerated by utilizing a deep neural network surrogate model in transportation network connectivity analysis. Select features of the proposed framework are illustrated using testbed community data and existing or approximated input models relevant to the Galveston region in Texas, USA. The results indicate the importance of capturing debris impacts when considering community-scale resilience metrics in coastal regions, without which the consequences of these events and equity of access to emergency facilities in the aftermath of them can be underestimated.