Using structural topic modeling to explore the climate change discourse about the Paris Agreement on social media

Abstract

The Paris Agreement emphasizes the inclusion of multiple actors and the importance of public information, representing notable changes in Global Environmental Governance (GEG). The Twitter (now X) public sphere presents a unique media environment that transforms political communication and influential actors. Under these circumstances, we explored how influential multi-level GEG actors discussed climate change on Twitter and whether and how there have been any significant changes in topics during the five years since the adoption of the Paris Agreement. We classified actors in climate discourse on Twitter and applied Structural Topic Modeling using Twitter data consisting of tweets and account information. We found that the topics discussed in the Twitter public sphere after five years were more inclusive of diverse regions and specifically discussed certain topics, such as climate action and climate change denial. The topics discussed by intergovernmental organizations, governments, and non-profits showed unification among actors but diverged from the public. The results showed a unified understanding of climate change among actors and increased public interest, implying a positive outlook for the success of the Paris Agreement.

Keywords

• Global environmental governance
• Multi-level actors
• Paris Agreement
• Structural topic modeling
• Twitter public sphere