Discussions on social media during major disasters are robust and often have multiple frames of reference. Temporal perspectives, however, are still lacking in current understandings of social-mediated discussions during disasters and crises, but incorporating temporal perspectives can significantly enhance environmental scanning efforts as prescribed in the issues management framework. The purpose of the current research is twofold: to apply and validate the SIR (Susceptible-Infectious-Recovered) model to examine topics’ growth over time on social media and to understand how future orientation of social media users (an indicator of temporal distance) affects their construal of a disaster through supervised machine learning. We based our analysis on Twitter discussions during the Texas winter storm in 2021. Results of the study show great fit of the SIR model for topic growth, and that temporal distance affects users’ construal of the event in line with core predictions of construal level theory. Theoretical, methodological, and practical implications on social-mediated discussions related to climate change-induced and -intensified disasters and issues management are discussed.

中文翻译：

---

灾难期间在线讨论的时间方法：应用 SIR 传染病模型预测主题增长并检查时间距离的影响

重大灾难期间社交媒体上的讨论很活跃，并且通常有多个参考框架。然而，目前对灾害和危机期间社会中介讨论的理解仍然缺乏时间视角，但纳入时间视角可以显着增强问题管理框架中规定的环境扫描工作。当前研究的目的有两个：应用和验证 SIR（易感-感染-恢复）模型来检查社交媒体上话题随时间的增长情况，并了解社交媒体用户的未来取向（时间距离指标）通过监督机器学习影响他们对灾难的看法。我们的分析基于 2021 年德克萨斯州冬季风暴期间的 Twitter 讨论。研究结果表明，SIR 模型非常适合主题增长，并且时间距离影响用户对事件的解释，这与解释水平理论的核心预测一致。讨论了与气候变化引发和加剧的灾害和问题管理相关的社会媒介讨论的理论、方法和实践意义。