Twitter During Sandy: Data Analysis and Insights For Future Events

Superstorm Sandy was the second-most damaging tropical cyclone in modern U.S. history. The unusual track and characteristics of Sandy presented serious challenges for both forecasters and the general public. How did confusion over the risks presented by Sandy propagate through the public? We have analyzed over 180,000 Twitter posts during Sandy that included terms relevant to the cyclone using C# code and a MySQL database. The quantitative results reveal scant attention to official statements but a strong focus on anthropomorphized social media creations that personalized the event for Twitter users. Fake reports and imagery also received far more attention than National Weather Service or local/state governmental statements. We will present lessons learned from this research regarding attention-generating social media posts, as well as some disquieting patterns regarding reported personal behavior during the event.