Social Responses and Urban Resilience Building in Extreme Precipitation Events: An Analysis of Social Media Data from the July 31, 2023 Beijing Rainstorm

As climate change intensifies extreme weather and climate events, enhancing urban climate resilience has become a global priority. Social media plays an increasingly important role in disaster response, yet limited research has used large-scale data to reveal differentiated response patterns among multiple actors or the mechanisms of resilience capital generation. This study develops a three-dimensional actor–stage–response analytical framework grounded in disaster risk management, social cognition, and social resilience theories. Using the July 31, 2023 Beijing Rainstorm as a case study, we collected 280,052 Weibo records and applied topic modeling, sentiment analysis, and text classification to examine responses by official media, online opinion leaders, and individual users across pre-disaster, during-disaster, and post-disaster stages, focusing on information dissemination, emotional expression, and response behaviors. Results show that multiple actors formed a collaborative information network in which official media provided authoritative guidance, individual users generated the bulk of content, and rapid rumor refutation was achieved—though vulnerable groups were inadequately reached. Positive emotions dominated throughout, with temporary negative spikes during the disaster that subsided following official reassurance and expert-led science communication. Self-help and mutual aid were the most prevalent behaviors; official media focused on resource coordination, while online opinion leaders and individual users contributed more to emotional support and social mobilization, yet information gaps delayed responses among vulnerable groups. These findings reveal that social media facilitates the generation of crisis-responsive social capital through coordinated multi-actor participation across information, emotional, and behavioral dimensions, thereby contributing to disaster resilience. The study proposes a four-dimensional framework—knowledge translation, emotional coordination, capacity building, and intelligent decision-making—to guide multi-actor collaboration in climate-resilient city development.