Evolving Geographical Disparities of Community Disaster Resilience in the United States

As hailed by the World Disasters Report, the proportion of disasters caused by climatic and weather events has risen from 76% in the 2000s to 83% in the 2010s. Evaluating community resilience and identifying key factors that play a role in mitigating disaster impacts is critical for locating vulnerable social groups and helping decision-makers formulate strategies to enhance resilience. In most frameworks for evaluating disaster resilience, a one-solution approach has been taken to inform policy-making. However, multiple studies have shown that urban and rural populations respond to disasters differently. There is a need to investigate if disparities in community disaster resilience exist between rural and urban communities. To address this knowledge gap, we hypothesize that rural communities have lower resilience scores and different socio-economic and environmental factors that could increase their resilience compared to urban areas. The objectives are three-fold: - (1) Developing a self-validated Customized Resilience Inference Measurement (CRIM) model which can compute the spatial and temporal patterns of resilience indexes; (2) Analyzing the difference of resilience indices between rural and urban counties spatially and temporally from 1960-2020; (3) Using Bayesian Network to identify two sets of driving factors and therefore risk reduction strategies among the rural and urban communities. First, hazard threats, damages, and recoveries every ten years were calculated utilizing SHELDUS (Spatial Hazard Events and Losses Database for the United States) and the Census data to delineate communities' evolving performance and resilience during historical disasters. The threats were calculated by a weighted sum of hazard frequencies. The damage was estimated by per capita economic losses, and the recovery was based on population stability. Second, the rural-urban continuum codes were used to obtain the difference in trends for these communities. Lastly, optimized Bayesian Networks developed separately for the two types of communities probed into the underlying relations that the indices have with socio-economic variables. This study has exhaustively shown the disparities between rural to urban communities while handling disasters, their dependencies, and the path forward in improving them.