Social Media and Sever Weather: Risk Communication Before, During, and After Tornados

In recent research scholars have found that real-time data on social media platforms like Twitter enhance our ability to detect and monitor natural disasters (e.g., Sakaki, Okazaki, and Matsuo 2010) and epidemics (e.g., Achrekar, Gandhe, Lazarus, Yu, Liu 2011; Garcia-Herranz, Egido, Cebrian, Christakis, Fowler 2012). Building upon this research, this paper uses an original archive of approximately 3 million public posts (“Tweets”) about tornados collected between May and January 2012 to advance our understanding of how affected publics perceive and communicate about severe weather. More specifically, we use these data to characterize and compare patterns of communication across three phases of a given incident, including (a) the “bow wave phase” consisting of posts prior to the event in which discussion of forecasts, watches, warnings, preparations, and emotive responses to the prospective storm are expected to predominate; (b) the “storm phase” in which postings reflect the arrival of the severe weather, its effects on structures, the behavior of those in the path, and effects (injuries, stress) on people; and (c) the “post storm phase” in which posts are more likely to have content reporting what happened, provide data on damage and location, and discuss personal and official responses to the storm and evaluate them. Following this characterization, the paper concludes with a general discussion of how data of this sort can be used to improve severe weather forecasting, communication, and post-event assessments.