Read our COVID-19 research and news.

From 1966 to 1974, France conducted 41 above-ground test of nuclear weapons on the Moruroa atoll, including this one of a 0.9-megaton giant in the Licorne test on 3 July 1970.

CTBTO/SCIENCE SOURCE

France grossly underestimated radioactive fallout from atom bomb tests, study finds

From 1966 to 1974, France blew up 41 nuclear weapons in above-ground tests in French Polynesia, the collection of 118 islands and atolls that is part of France. The French government has long contended that the testing was done safely. But a new analysis of hundreds of documents declassified in 2013 suggests the tests exposed 90% of the 125,000 people living in French Polynesia to radioactive fallout—roughly 10 times as many people as the French government has estimated.

“This is going to make a big splash in France,” predicts Frank von Hippel, a physicist specializing in public and international affairs at Princeton University, who was not involved in the work. Most French Polynesians were exposed to a relatively small amount of radiation, von Hippel notes, and the central issue is who is eligible for compensation under French law.

The findings come from a 2-year collaboration, dubbed the Moruroa Files, between Disclose, a French nonprofit that supports investigative journalism; Interprt, a collective of researchers, architects, and spatial designers affiliated with the Norwegian University of Science and Technology who focus on environmental issues; and the Science & Global Security program at Princeton. The findings were presented on 9 March on the project’s website, in a book, and in a technical paper posted to the arXiv preprint server.

The issue of compensation for radiation exposure from the tests has been a thorny one for decades, says Sébastien Philippe, an applied physicist at Princeton, co-author of the book, and lead author on the paper. Under a 2010 law, anyone exposed to fallout in Polynesia or Algeria, where France also conducted nuclear tests, could be compensated if they developed any of 23 cancers associated with exposure to radiation, such as thyroid cancer. Although the law did not acknowledge causing harm, it established a “presumption of causality” between the test and the cancers, explains Sonya Schoenberger, a co-author on the paper who is working on her law degree at Yale University and her doctorate in history at Stanford University.

There was a catch, however. If a standing Committee for the Compensation of Victims of French Nuclear Testing found that a person’s radiation contributed a “negligible” risk of causing their cancer—relative to factors such as smoking—the individual wouldn’t qualify for compensation, Schoenberger says. On those grounds, the committee rejected 97%—1008 of 1039—of claims made between 2010 and 2017.

The French legislature then threw out the negligible exposure exception and, in 2018, amended the law so that only people exposed to more than 1 millisievert (mSv) of radiation—about the amount in 10 chest x-rays—would qualify for compensation. The standard for determining exposure then became a set of estimates made for each test by France’s Atomic Energy Commission (CEA) in 2005 and 2006, and France began to approve about half of the claims.

But the declassified documents suggest actual exposures were between two and 20 times higher than CEA estimates, Philippe says. Reasons for the discrepancies vary from test to test, he says. For example, CEA acknowledged that the first test, dubbed Aldébaran, exposed residents of the Gambier Islands to relatively high levels of fallout. But actual exposures were likely higher still, Philippe says. Although CEA noted that contaminated rainwater fell on the island, he says, it failed to consider that many residents likely drank the contaminated water, collected in household cisterns, for days.

Most important, the documents suggest a single test in 1974, called Centaure, exposed the entire population of Tahiti—87,500 people at the time—to fallout. French authorities set off a relatively tiny atom bomb with an explosive yield equal to 4 kilotons of TNT, and weather forecasts predicted that winds should carry fallout to the north. Instead, the wind blew to the west, carrying the plume directly over Tahiti. A new simulation based on data in the documents shows how the plume of radiation wafted over the island (see video, below). CEA estimated that people on the island received a dose of about 0.6 mSv.

However, Phillipe and colleagues argue that CEA underestimated the total amount of radiation that accumulated on the ground over several days, didn’t account for radiation lingering in vegetables consumed later, and—in contrast to its analysis of other tests—included no uncertainty in its figures. Accounting for these factors and including a 25% uncertainty on the ground deposition suggests everybody in Tahiti may have received a dose greater than 1 mSv, roughly twice as much as the CEA estimate.

That’s also the de facto global standard for determining whether somebody has been exposed to radiation, says Emlyn Hughes, a physicist at Columbia University, who was not involved in the analysis. It’s roughly what a Polynesian might receive in a year from natural sources such as cosmic rays, he says, but fallout can be much more harmful if it is ingested or inhaled. “What [dose of] radiation is getting into your body for the rest of your life?” he says. “To me that’s the danger.”

The new analysis moves the vast majority of French Polynesians past the exposure threshold to qualify for compensation. Philippe and Schoenberger would like to see France do away with the exposure standard and compensate anyone who lived through the tests and developed a qualifying cancer. “Our hope is to demonstrate that this kind of threshold can be prejudicial to claimants just because of the difficulties of proving exposure,” Schoenberger says.

Philippe estimates that, assuming a cancer rate of 0.2% per year, roughly 10,000 cancer patients or their families would qualify retroactively and that compensating them would cost about €700 million. Future cancers would cost about €24 million per year, he estimates. However, Hughes says it remains to be seen whether the French government will even acknowledge the analysis. “My fear is that they will simply ignore it,” Hughes says.

The declassified documents also show the French government routinely failed to warn Polynesians about the radiation risks, Philippe says. In the Centaure test, authorities could have warned Tahitians about the approaching fallout 2 days in advance, but did not. Ironically, Philippe notes, the United States, the Soviet Union, and other countries were monitoring the tests remotely. “Everybody knew what was going on,” he says, “except the Polynesians.”