Cross-Platform Analysis of Climate Discourse on Meta and Bluesky

Abstract

Climate discourse on social media varies across paid advertising and organic public discussion, but most cross-platform analyses focus on themes or stance rather than argumentative structure. We analyze 16,157 Meta advertisements and 18,781 Bluesky posts to compare how climate claims and supporting grounds are constructed across platforms. Using a Toulmin Argument-inspired LLM pipeline, we extract and validate claims, then examine claim topics, named entities, and LIWC-based linguistic style. Our results show that Meta advertisements more often emphasize solution-oriented, locally targeted, and campaign-style claims, while Bluesky posts are more political, event-driven, and issue-centered. Linguistically, Meta claims show more analytic and authoritative language, whereas Bluesky claims are more personal and expressive. These findings show that claim-level analysis reveals cross-platform differences missed by topic- or stance-level approaches.