In recent years, the uncontroversial and widely accepted scientific facts about climate change have become a polarizing and politicized issue, dividing society into hostile camps. This has been exacerbated by growing hostility to science and the multiple political crises of the 2020s, which have led to widespread uncertainty and increased susceptibility to misinformation, disinformation, and conspiracy theories among the general population.
Against this backdrop, the final report of the research project NOTORIOUS, jointly conducted by the Leibniz Institute for Media Research | Hans-Bredow-Institut (HBI), ISD Germany and the Hamburg University of Applied Sciences (HAW Hamburg), examines the cross-platform dissemination of climate-related misinformation and disinformation. Philipp Keßling, Dr. Felix Victor Münch, Dr. Gregor Wiedemann and Mattes Ruckdeschel from HBI contributed to the report.
The research report can be downloaded here .
Key Findings
- Semantic similarity to misinformation or disinformation narratives and the presence of far-right actors in posting clusters are effective indicators for identifying misinformation and disinformation in large datasets.
- Twitter/X and Telegram are important platforms for climate-related misinformation and disinformation, while Instagram is rather underrepresented.
Individuals, pseudonymous accounts, accounts that primarily share content from other accounts, and partisan accounts are responsible for the majority of posts classified as misinformation or disinformation. - Both mainstream and alternative media are used as sources: In addition to right-wing alternative media such as “Tichys Einblick” [Tichy’s Insight] and “Junge Freiheit” [Young Freedom], links to other posts on social media platforms are also used, as well as to liberal-conservative daily newspapers such as “Die Welt” [The World].
- Personal attacks and delay narratives dominate the narratives and are used more frequently than climate change denial and climate skepticism.
Discussions on climate change and climate protection serve as a starting point for more far-reaching critiques of the system by the extreme right.
Methodologically, this research report has developed an approach that takes into account the connections and interactions between different platforms in order to identify and analyze climate-related disinformation and misinformation narratives. To this end, 3.3 million posts from Telegram, Instagram, Facebook, and X (formerly Twitter) between 2019 and 2023 were collected based on climate-related keywords and further processed using natural language processing methods, language models, similarity networks created from them, and hierarchical clusters.
Based on four technical indicators – semantic similarity to known misinformation and disinformation narratives, the presence of far-right actors within the posting clusters, the proportion of Telegram posts within the posting clusters, and a random sample – 38 clusters with 1,484 posts were selected for further quantitative analysis. Nineteen posting clusters were also analyzed qualitatively in terms of the narratives used.
The report provides valuable insights into the mechanisms of “destructive discourse” in climate-related online communication. For example, it analyzes prominent actors, sources used, and narratives that fuel the spread of climate-related misinformation and disinformation.