Fukushima Radiation Risks from Eating Fish

Japan’s damaged Fukushima nuclear power plant is releasing radioactively contaminated water into the Pacific Ocean. The Japanese government has placed restrictions on fishing in waters near the reactors. Are radioactive substances from Fukushima being found in fish caught and sold in the United States? The answer is “Yes” for tuna, as measured by researchers at Stanford University, and Fukushima radiation is expected to be detectable for other species of fish which swim long distances. Does this measured radiation level in U.S.-caught fish present a significant health impact to individuals? The answer is “No.” Could Fukushima radiation in U.S.-caught fish increase over time to levels that would present a significant health impact to individuals? In our judgment, the answer is “No” due to dilution of the radioactivity released from Fukushima in the vast waters of the Pacific Ocean.

But bear in mind that the likelihood of Japan’s success in cleaning up the Fukushima accident is uncertain, and the United States imports over 80 percent of its seafood, with a large proportion from across the Pacific Ocean. The full impact of the Fukushima nuclear accident on marine life and ocean ecosystems, and on the commercial seafood industry, is still unfolding and deeply troubling.

An excellent resource on this topic is a webpage of Woods Hole Oceanographic Institution scientist Ken Buesseler, who in his career has also studied the impact of the Chernobyl nuclear accident on the Black Sea. NRDC’s physicist Tom Cochran also recently looked at this issue of Fukushima and seafood caught off the U.S. West Coast and Hawaii, and his analysis is presented below. The Stanford measurements were interpreted as the tuna having been exposed to radiation in waters close to Japan and then swimming to U.S. coastal waters where they were caught. Dr. Cochran also looks at the impact of Fukushima radiation carried directly to U.S. waters in ocean currents, and the effect of that radiation concentrated in the bodies of fish through bio-accumulation.

Assessing the Radiation Risks Associated With the Leaks at Fukushima
Thomas B. Cochran, Ph.D., September 2013


In the aftermath of the nuclear accident at Fukushima Daiichi, radioactive contamination at the plants has been leaking out of control into the sea. It is widely believed that these leaks into the sea are primarily the result of groundwater that becomes contaminated as it flows under the damaged reactors and adjacent buildings at the plant site and then into the sea. The three reactors whose cores have partially melted require continuous water cooling. Some of the contaminated cooling water is collected and stored in numerous temporary tanks at the site, and some of it continues to leak into the basements of buildings and from there it leaks into the ground contaminating the aquifer flowing beneath the plant. In addition, a few of the storage tanks have begun to leak, further contaminating the groundwater. To date, TEPCO, the plant operator, and the government of Japan have been unable prevent the continuous flow of radioactive contamination into the sea. The leaks are not under control.  This has caused widespread concern regarding the potential radiation risks to the Japanese people and even to some U.S. citizens.

There is a dearth of publicly available data, or at least little data known to us, regarding the levels of radioactive contamination in the sea and in seafood as a consequence of these chronic leaks of radioactivity into the sea from Fukushima.  Nevertheless, we can gain some insight into the risks from the limited data that we have seen.

In assessing health risks from radiation exposure the conventional approach is to examine the various pathways be which people are exposed to radiation, calculate the radiation dose to exposed individuals from the various pathways, and then estimate the risk by multiplying the dose by risk factors derived largely from past epidemiological studies.  As in this case the uncertainties in this chain of calculations often lead to very large uncertainties in the calculated risks. Alternatively, we can bound the estimates of risk by making conservative assumptions. Exposure pathways include direct external exposure, inhalation, and ingestion. In the case of the leaks it is obvious that the risk to the public is dominated by the ingestion pathway, so we can ignore the other pathways. This would not be the case regarding workers at the site. Since the leaks are into the sea and people do not drink seawater, assuming they are not swimming in the ocean, we can limit our analysis to exposure from eating contaminated seafood.

It is well known that fish concentrate certain radioactive elements in their flesh and bone—particularly, strontium and cesium. We also know, in part from historical experience, that cesium-137 and -134 and strontium-90 are the radioisotopes that will dominate the exposure dose in this situation.  Evidence suggests that levels in fish of strontium-90 remain much lower than that of cesium-137. With this background we proceed with our calculations to assess the potential risks to Japanese people that hypothetically might eat a lot of fish taken from the sea near Fukushima, and then the risk to U.S. citizens that eat seafood. Before doing so, let’s say a few words about units. Radioactivity is measured in Becquerels (abbreviated "Bq"), where one Bq equals one disintegration per second. The old unit was the Curie (abbreviated “Ci”), where one Ci equals 37 billion disintegrations per second. This older unit was derived from the fact that this was the rate of disintegration associated with one gram of radium.


Case 1: Releases at the time of the Accident. Vincent Rossi, et al.,have modeled transport of radioactive cesium-137 from Fukushima across the Pacific Ocean.  This analysis assumed the release of 22 Peta-Becquerels (PBq: 1 PBq = 1x1015 Bq = 1 million billion Bq = ~0.6 Mega-Curies) released from Fukushima over a period of one month starting in mid-March 2011. The following excerpts from the Rossi, et al., paper provide modeling results. In their model:

The plume is quickly advected away zonally from Japan due to the vigorous Kuroshio Current and  Kuroshio Extension as it passes over the Izu-Ogasawara Ridge (~142°E). Within this pathway there is a rapid dilution of Cs-135 with time. Concentrations fall below 10,000 Bq/m3 everywhere by the end of July 2011, from a maximum concentration of about 150,000 Bq/m3 near the Japanese coast at the end of March 2011, in good agreement with coastal studies (Buesseler, et al., 2011; Tsumune et al., 2012).


The first waters characterized by Cs-137 levels above 1 Bq/m3 are projected to reach the US west coast at around 45°N in mid-2013,while rapidly increasing up to 5 Bq/m3 from early 2014 (Fig.2a  and b). Maximum concentrations between 20 and 30 Bq/m3 occur from mid-2014 (3 years after the initial release) to mid-2018 off Oregon/British Columbia (45–55°N), in good agreement with the simulations by Behrens et al. (2012). Cs-137 concentrations above 10 Bq/m3 in surface shelf waters near Vancouver (49°N) are modeled from 2014 to mid-2021, i.e. over ~6 years.


Slightly contaminated waters start to reach the area [the Hawaiian Archipelago] by 2013, but Cs-137 concentrations above 5 Bq/m3 only arrive in early 2015 and remain above this concentration until early 2025. Maximum concentrations of about 8 Bq/m3 occur from mid-2017 to mid-2021.The archipelago is not situated along the main pathway of the plume and so is only affected by water that recirculates or mixes southward...

Let’s begin by considering the maximum concentration of Cs-137 in seawater reported in the quote above, about 150,000 Bq/m3, near the Japanese coast.  Incidentally, Ken Buesseler has told The New York Times that he has received samples of seawater taken in July [2013] from near the plant that contained 10,000 Bq/m3. The corresponding level last year, only months before the disaster, was just 1.5 becquerels. Since there are 1,000 liters of water in a cubic meter and a liter of water has a mass of one kilogram (kg),150,000 Bq/m3 corresponds to 150 Bq/liter (or Bq/kg) of seawater. We further assume that the cesium bio-accumulates in fish flesh such that the concentration of cesium in the fish meat is 100 times the average concentration of cesium in the water there the fish is swimming. This is at the upper end of the range of measured ratios. Consequently, fish meat from fish taken in waters contaminated at 150 Bq/liter will be expected to have cesium-137 concentrations below 15,000 Bq/kg.

Let’s assume we are very risk averse and we want to maintain the risk of getting cancer from eating contaminated fish to less than 1x10-5, that is, the probability of getting a cancer from eating the contaminated fish will be less than one in 100,000. Keep in mind, for an average individual in the United States, the probability of getting cancer is about 40 percent and the probability of dying of cancer is about 20 percent. The U.S. Environmental Protection Agency (USEPA) estimates that for dietary intake, the risk of getting cancer (risk of cancer morbidity) from ingesting food containing cesium-137 is 1x10-9 cancers/Bq of ingested cesium-137 (USEPA, “Health Risks form Low-Level Environmental Exposure to Radionuclides,” Federal Guidance Report No. 13, - Part 1, Interim Version, EPA 402-R-97-014, January 1998, p. 35). That is a risk of one in a billion per Bq of ingested cesium-137.  

To keep the risk below 1x10-5 the consumer must limit his/her dietary intake to less than 10,000 Bq of cesium-137.  Therefore this risk limit would be reached after eating about 0.7 kg of fish meat. While this is a conservative estimate of what is required to achieve a low risk, one could make a good case for quarantining fishing off the Japanese coast near Fukushima, which of course is what the Japanese government has done.

Near the west coast of the United States the maximum projected concentration is about 30 Bq/m3 some three years after the initial release. This is 5,000 times less than the 150,000 Bq/m3 concentration we have assumed near the Japanese coast.  Therefore, to keep the risk below 1x10-5 the consumer must limit his/her dietary intake to less than about 3,000 kg (3 tonnes) of fish. In other words, do not worry about eating fish taken from US coastal waters. Since the concentration projected for waters near the Hawaiian Archipelago are even less than that projected for the West Coast, the same admonition applies to Hawaii.

Case 2: Chronic Leakage into the Sea from Fukushima.  From samples of seawater, Jota Kanda of Tokyo University of Marine Science and Technology estimated last year that about 0.3 TBq of radioactive material are leaking into the sea each month. And in this article in New Scientist, Ken Buesseler says the Kanda estimate is probably the best he is aware of, and closely matches figures released on 21 August by TEPCO, of 0.1 to 0.6 TBq per month for cesium-137 and 0.1 to 0.3 for strontium.  At an average leak rate of 0.3 TBq/month, it would take more than 6,000 years to equal the 22 PBq release assumed under case one above. Consequently, the current chronic leaks do not increase the risks associated with consuming fish caught in waters off the west coast of the United States or Hawaii.

Case 3: Recent leaking tanks. It has been reported in the press that a storage tank containing 24 terra-Becquerels (TBq = 1x1012 Bq = 1,000 billion Bq) of radioactivity in in 300 cubic meters of water was leaking.

Let’s be exceedingly conservative by assuming all the radioactivity is cesium-137 and all of it leaks into the groundwater and from there into the sea over a very short period of time. The source term, 24 TBq, is about 1,000 times smaller (and a couple of years later) than the source term used in modeling by Rossi, et al. Clearly, this does not change the conclusions regarding eating fish caught near the West Coast or Hawaii.


For the foreseeable future, one should avoid eating fish caught near Fukushima. Buesseler says that during his own sampling survey in waters 30 to 600 kilometres from Fukushima in June 2011, three months after the meltdown, the highest levels he found were 3 Bq/liter of cesium-137. This suggests that the consumption of fish caught in these waters would not represent a significant risk to individuals. There is not a significant radiological risk to individuals associated with consuming fish caught near the West coast of the United States and Hawaii.

 Ken Buesseler notes that the north Pacific contains an estimated 100 PBq of cesium-137 from H-bomb testing in the 1960s, so the fallout from Fukushima is adding only a fraction of that. Total discharges from the Sellafield nuclear plant in the UK released 39 PBq over 40 years of operation, according to Buessler.

About the Authors

Matthew McKinzie

Senior Scientist, Land & Wildlife program, and Director, Nuclear program

Join Us