Visual digital intermediaries and global climate communication: Is climate change still a distant problem on YouTube?

PLOS ONE Visual digital intermediaries and global climate communication examination reveals that language groups articulate the scope and character of the issue with notably distinct profiles. The aggregate analysis focusing on the top-20 results suggests that YouTube supports ‘vis- ibility winners’ [66] on climate change. A concentrated set of channel types dominate top results, and their videos are consistently more likely to top search results across different regions. As Table 1 shows, media organisations are the most prevalent, with a higher frac- tion of videos in the top results. Their videos are highly amplified — the API returns the same video in several regions’ rankings — but not the most so. Actors fresh to conventional climate communication are also highly amplified, and audi- ence engagement with these channels is higher than for professional media channels. User-led channels are relatively prevalent at the top, highly amplified, and (often) highly engaged with. Videos posted by digitally native popular science groups and online education actors are the most amplified, although with limited prevalence in top results. For example, the same Ger- man popular science channel appears across several global regions. These popular science and education channel types are also the most engaged with. By contrast, other traditional climate communicators have little visibility in the aggregate results. Authoritative channels such as (inter-)governmental and science organisations, and potential stakeholders, such as political

actors, civil society, and religious organisations, are neither prevalent, amplified, nor engaged with. Nevertheless, the iconic themes of the distancing frame continue to be part of the climate narrative. One fifth of the search results show classic arctic themes (and recall that we are using a conservative coding for these themes). While fresh actors from beyond the conven- tional climate communication circle do appear in the top ranked content, they do not notice- ably reduce the prevalence of the arctic theme. If anything, fresh climate communicators sup- port its persistence. Table 1 shows the extent to which videos in the respective channel cat- egories include the themes. Professional media maintain arctic content, and online educa- tion actors match this. Yet, the other newly prominent actors, user-led channels and popu- lar science groups, surpass these levels. By contrast, traditional communicators such as sci- ence organisations, civil society, and (inter-)governmental organisations show little of these themes, but as just noted, are also not particularly visible themselves. Beyond the aggregate, there is significant variation across contexts. Language is the main differentiating factor, as opposed to geographical location on the globe or categorisations such Table 1. Composition and content of the combined search results from all country-language pairs, considering the top-20 results (videos) for each pair. #res: number of search results. %res: fraction of search results. #cha: number of distinct channels. #vid: number

of distinct videos. amp: fraction of search results divided by fraction of videos minus 1. #vie: median number of views per video. #com: median number of comments per video. #lik: median number of likes per video. %PB: fraction of videos showing polar bears. %Ice: fraction of videos showing icebergs or ice cliffs in water. %PBI: fraction of videos showing polar bears, icebergs or ice cliffs in water. Only categories with at least 2% search results. Category #res %res #cha #vid amp #vie #com #lik %PB %Ice %PBI All Categories 5060 1.00 1067 1369 0.00 2785 34 2 0.08 0.18 0.20 Prof. media org. 1932 0.38 310 436 0.20 3402 39 3 0.08 0.20 0.22 User-led channel 965 0.19 183 199 0.31 13778 192 32 0.13 0.27 0.29 Pop. sci. groups 424 0.08 43 55 1.09 28369 542 63 0.15 0.31 0.31 No description 397 0.08 199 222 -0.52 706 10 0 0.07 0.13 0.14 (Inter)gov. org. 317 0.06 80 131 -0.35 1253 10 0 0.02 0.13 0.14 Civil society 245 0.05 74 93 -0.29 841 10 0 0.06 0.11 0.13 Online education 235 0.05 41 48 0.32 8516 117 8 0.17 0.12 0.21 Business 119 0.02 32 35 -0.08 3595 23 3 0.09 0.14 0.17 Science org. 110 0.02 44 52 -0.43 1240 10 0 0.06 0.13 0.15 https://doi.org/10.1371/journal.pone.0318338.t001 PLOS ONE https://doi.org/10.1371/journal.pone.0318338 April 8, 2025 11/ 23