The Role of Social Media in Facilitating Collective Action for Risk Response during Disasters: A Case Study of the 2021 "Henan Floods" in China

With the rise of an interconnected society, risk communication during public emergencies has increasingly shifted to social media platforms. Individuals in high-risk environments adopt risk communication to reduce risk perception and guide decision-making (Prybutok & Ryan, 2015). Social media provides an open platform for information exchange, shaping risk response (Yoo, 2019). Unlike stable communities, emergency groups focus on sharing information and disband once the crisis ends, forming 'temporary publics' (Bruns & Burgess, 2011). How does social media play a role in this process?

Our study explores this phenomenon using the example of the 2021 Henan Floods, during which people used Weibo for risk communication. Weibo, China's leading social media platform, uses hashtags to connect communities and disseminate information, forming strong connections between dispersed users (Bennett & Segerberg, 2013). Hashtags enhance searchability and visibility, bringing users with common interests together (Saxton et al., 2015). Emotions play a key role in risk communication, as personalized content triggers emotional mobilization, which can influence public risk perception (Gu, Guo, & Zhuang, 2021). We proposes a framework where weak connections emerge through risk communication topics, and strong connections form through hashtags, creating stable issue spaces during public emergencies on Weibo.

We analyzed 51,442 Weibo posts about the 2021 Henan floods using Python for data collection and STM in R for topic modeling, identifying eight topics and constructing a hashtag co-occurrence network. We reveals four findings: (1) Weibo users form weak connections through shared topics, driven by specific information needs; (2) hashtags create strong, locally centralized yet distributed topic networks; (3) weak topic connections and strong hashtag connections shape a locally centered, distributed issue network; and (4) issue network structures influence emotional communication, with overall emotional contagion but varied emotions within topics. These four findings illustrate the role of social media in shaping risk communication dynamics.