Data-Driven Community Flood Resilience Prediction

Climate change and the development of urban centers within flood-prone areas have significantly increased flood-related disasters worldwide. However, most flood risk categorization prediction efforts been focused on hydrologic features hazards, often not considering subsequent long-term losses recovery trajectories (i.e., community’s resilience). In this study, a two-stage Machine Learning (ML)-based framework is developed to accurately categorize predict communities’ resilience their response future hazards. This step towards developing comprehensive, proactive disaster management planning further ensure functioning mitigate catastrophic events. framework, indices are synthesized goals robustness rapidity) using unsupervised ML, coupled with climate information, develop supervised ML algorithm. To showcase utility it was applied historical records collected by US National Weather Services. These were subsequently used indices, which then associated data, resulting in high-accuracy predictions and, thus, studies. demonstrate utilization spatial analysis quantify vulnerability across selected domain. The presented study employable studies patio-temporal identification. Such can also empower decision makers effective data-driven strategies.