Abstract:
The onset of power grid failures and outages due to severe weather happen
instantly. During a cyclonic weather event, critical infrastructure sustains damages and
can be destroyed from tropical-force winds, storm surge, flooding, and tornadoes.
Electricity is the most vulnerable infrastructure to severe weather impacts. Often,
damages to electrical equipment and their impact can be hard to locate. Also weather
conditions may not permit for a safe dispatch to locate, assess, and repair utility
equipment. Therefore, this study is tasked to examine how social media users report
blackouts and damages to electrical equipment of utility providers. Specifically, it aims to
explore the possible relationship between the volume and the spatial footprint of weather
and power-related tweets to the spatial extent of power outages reported by utility
companies. Social media platforms, such as Twitter, now play a pivotal role in crisis
management during severe weather events. However, limited research has been completed on the concept of using social media to predict future disruptions in the energy
grid. In order to determine whether or not this concept is feasible, a geosocial,
spatiotemporal, and geospatial analysis are carried out, examining the relationships
between social media usage and power outages in two case studies: Hurricane Harvey
(2017) and Superstorm Sandy (2012). The results of these case studies suggest that, at
least in some cases, social media can serve as a possible information source about the
occurrences and the spatial extent of power outages due to extreme weather events.