Evaluating resilience of urban lifelines against flooding in China using social media data

Urban lifelines are the backbone of fundamental services that require stability to enable all other aspects of society to function during natural hazards. However, few studies have focused on measuring lifeline performance and resilience to hazards, especially from the public perspective. In this study, taking flood as the research object, we developed an enhanced urban lifelines classification scheme that integrates the Federal Emergency Management Agency (FEMA)'s lifeline framework with the co-occurrence relationships of keywords derived from 430 million Weibo posts. Leveraging this framework, we formulated two key indicators: Public Concern Ratio and Public Emotion Ratio, to evaluate the resilience of urban lifelines during the 2017 and 2020 flood in China. The results demonstrate the robust resilience of urban lifelines against flooding, with notable improvements over time. The study also identifies certain vulnerable lifelines, notably in the ability of early warning and control of rumor spreading, which often lead to an increase in social media posts expressing negative emotions during flooding. These areas are pinpointed for necessary enhancements. Employing a data-driven methodology, the study provides a novel and insightful approach to assessing urban lifeline resilience against flooding.