Online community discourse during the Deepwater Horizon oil spill: an analysis of Twitter interactions

Following the Deepwater Horizon oil spill on April 20, 2010, information was widely broadcast through social media platforms such as Twitter. This study aimed to gain insights into the content and flow of the tweets that had shaped the conversation related to the oil spill within the first 4 months of the rig explosion and to understand how these online interactions may have contributed to disaster response and resilience. A set of tweets (n = 876,298) was identified and processed to yield 736,324 clean tweets. The study used a mixed-methods approach, including human coding, machine learning and social network analysis to examine the online discourse. Visualizations were used to present the patterns and connections within the dataset into a more readable and interpretable format. Our analysis revealed that popular themes such as environmental and economic concerns, clean-up and volunteering, health impacts and frustration towards BP had contributed to the shaping of the discourse. The Twitter users included seekers and providers of information and resources. Another group of users, local volunteers and digital activists, functioned as boundary spanners by facilitating information flow between the seekers and the providers. Additionally, Twitter had been used for risk communication and the dissemination of factual health information. The online communication and coordination efforts may have contributed to the mitigation of the impacts of the oil spill. Our findings suggest that social media empowered community-based users, affording access to power brokers, reliable information, the formation of online networks and social capital and potentially contributed to their resilience.