Rolli
Research

Investigating Science Misinformation and Coordinated Campaigns on Social Media

A deep investigation into coordinated campaigns and science misinformation related to Japan's Fukushima wastewater release — tracing how geo-political narratives shaped public sentiment across platforms using Information Tracer data.

Yuki Cui, Zhouhan Chen
8 min readLast updated: March 12, 2025

The Fukushima Daiichi nuclear power plant in Japan was damaged in the 2011 earthquake and tsunami. Japan used water to cool the reactors, causing the water to become contaminated. After 12 years, Japan began releasing wastewater into the sea on August 24, 2023, as part of a continuous process that will take at least 30 years. According to the International Atomic Energy Agency (IAEA): 'Based on its comprehensive assessment, the IAEA has concluded that the approach to the discharge of ALPS treated water into the sea, and the associated activities by TEPCO, NRA, and the Government of Japan, are consistent with relevant international safety standards.' However, the release of wastewater was met with varied responses across different countries, where social media played a critical role in amplifying these voices, shaping public sentiment through information, or potentially misleading them through coordinated campaigns. In this report, we investigate and demonstrate the presence of coordinated campaigns and science misinformation related to Fukushima wastewater release on social media. All data and analysis are collected and generated by Information Tracer, our proprietary software.

After manually reviewing posts, we identified two key competing narratives: (1) 'nuclear wastewater is safe' — associated with pro-Japan western countries relying on IAEA reports and scientific research; and (2) 'nuclear wastewater is not safe' — associated primarily with China, largely using rhetorical arguments such as 'if it's safe, put it in Japan.' The focus — water safety — is a scientific issue, but the narrative divide aligned well with geo-political alliances. By closely examining the data, we observed many Chinese accounts repeatedly posting the same phrase, 'If it's safe, put it in Japan.' The slogan was first posted by the Fiji Exposed Forum on Facebook, a local news channel. When Chinese state-affiliated media account 'Trending in China' mentioned this protest, engagement spiked, tripling its audience reach within a single day. Major Chinese media outlets including People's Daily and CGTN Frontline joined in spreading the message. On August 28, Chinese Counsellor Zhang Heqing tweeted the same phrase, followed by another Chinese Counsellor Zhang Meifang the next day — clear evidence of coordinated amplification.

During our investigation, an image called 'The Beast of Sin' stood out — created by Wuheqilin, a Chinese cartoonist known as the 'Wolf Warrior Artist' for his politically charged nationalist illustrations. On August 24, this image was shared on Twitter by Chinese diplomats, Chinese media, and other accounts, stirring strong emotions regarding the nuclear wastewater release. Weighted sentiment analysis revealed China and Pakistan exhibited particularly strong opposing sentiments. A contrasting episode involved the 'Simpsons Fukushima' meme that emerged after Japanese Prime Minister Fumio Kishida publicly ate fish from Fukushima. Key influencers amplifying the meme included Hu Xijin (former Global Times chief editor), New York Post, Insider Paper, and Wall Street Apes — a genuinely cross-national spread that Information Tracer assessed as organic rather than coordinated. The contrast between the organic meme spread and the coordinated state-media messaging illustrates how different actors shape the same information environment.

A deep investigation into coordinated campaigns and science misinformation related to Japan's Fukushima wastewater relea…

We hope this report makes you think: when the amount of information we receive exceeds our ability to process, what should we do? Who should we trust? There is no simple answer. Taking a bit of time to search and verify what we see on the Internet would be a good first step. We are grateful to the Pulitzer Center who provided funding for this investigation, and to Professor Haohan Chen from The University of Hong Kong and Katrine Krogh Pedersen from The University of Copenhagen for developing workshops with us. This research is a repost from the Information Tracer Blog. If you have ideas, want to learn more, or would like to collaborate, we welcome your contact.

Related to this topic: Rolli IQ · Pricing

Yuki Cui, Zhouhan Chen

Intelligence Analyst · Rolli Intelligence Desk

Covering narrative manipulation and authenticity intelligence for the Rolli Intelligence Desk.

Share this article:Share on XShare on LinkedIn
EducationMarch 5, 2025

AI's Impact on Search Traffic – and How Experts Can Stay Visible

AI-powered search is reshaping how people find information online — with 58.5% of U.S. searches n…

Read →
ResearchFebruary 27, 2025

1 meme coin, 72 hours, 422,071 tweets, and a $10 billion market cap

At 2am UTC on January 18, 2025, Trump posted about $TRUMP on Truth Social. Within 72 hours, 422,0…

Read →
Rolli IQ™August 5, 2025

Rolli IQ Explores the Labubu Phenomenon: Organic or Manufactured?

A $10 keychain sparked a global craze after Lisa from BLACKPINK was spotted with one. Rolli IQ's …

Read →

Social is full of fake signals. Start acting on the real ones.

Rolli IQ detects coordinated inauthentic behavior across 8 platforms in real time — so you respond only to what's actually real. Free trial, $99/mo, no credit card required.

Joined this week by 47 communications, security, and research teams

Start Free Trial — No CC RequiredSee a 20-Min Live Demo →
Join 400+ organizations protecting their reputation. Average setup: 8 minutes.

Monitoring live in under 2 minutes  ·  No credit card  ·  Cancel anytime  ·  SOC 2–aligned