Category Archives: N Pacific

Measuring Fukushima Contamination in Fish Caught in Hawaii

Yellowfin tuna, Thunnus albacares leaping from the water

By Jay T. Cullen

The purpose of this post is to summarize a recently published, peer reviewed, scientific study that investigated levels of Fukushima derived contamination in fish caught in the North Pacific and sold at market in Hawai’i.  This post is part of an ongoing series dedicated to bringing quality scientifically derived information to readers so that they can form an evidence based opinion regarding the environmental impact of the Fukushima Daiichi Nuclear Power Plant meltdowns. The paper by Azouz and Dulai (both at the University of Hawai’i at Manoa) summarizes levels of human made 134-Cesium (134Cs half life ~2 years) and 137-Cesium (137Cs half life ~30 years) and naturally occurring 40-Potassium (40K half life 1.25 billion years) in 13 different fish purchased in Hawai’i in 2015.  The findings of the study were that:

  1. 3 of the 13 fish had detectable levels (above the 95% confidence interval) of 134Cs which can be linked to the Fukushima disaster
  2. Highest levels of radiocesium were found in ‘ahi tuna with 134Cs and 137Cs of 0.10 ± 0.04 Bq kg-1 and 0.62 ± 0.05 Bq kg-1 respectively
  3. Most of the fish carried no fingerprint of the Fukushima disaster
  4. Levels of radiocesium were well below intervention levels of 1,200 Bq kg-1 set by the United States Food and Drug Administration
  5. Doses to fish consumers from human made radioisotopes were 30-1,000 fold lower than the dose experienced because of naturally occurring 40K in the fish
  6. Neither the effective dose from the natural nor the human made radioisotopes represent a significant health risk to consumers of the fish given scientifically established dose-response relationships

These results agree with the results of the Integrated Fukushima Ocean Radionuclide Monitoring Project (InFORM) I head up at the University of Victoria which has been making similar measurements on North Pacific fish returning to rivers in North America.

The Azouz and Dulai paper was published recently in the journal Pacific Science and can be found here.  The authors obtained 13 different species (Ahi, Albacore Tuna, King Salmon, Cod, Dover Sole, Halibut, Mahi Mahi, Monchong, Onaga, Opah, Opakapaka, Swordfish and Yellowfin Tuna) of fish that were caught in the North Pacific (>20oN) and commonly consumed in Hawai’i at local markets.  Information about the range and size of the fish are given in Table 1:


Levels of Radiocesium in Fish From Hawai’i

Samples of the fish tissue were freeze dried and homogenized before gamma emitting radioisotopes were measured using a gamma spectrometer by counting samples for a period of 7 days. Levels of 134Cs, because of its short half life, serve as a fingerprint of Fukushima in samples as previous sources of this human made isotope (e.g. 20th century nuclear weapons testing and the Chernobyl disaster) are sufficiently far in the past that all of the isotope has decayed away and is no longer present in the environment.  Results of the analyses are summarized in the following figure:

Fig. 1 Cesium activities in fish collected in the North Pacific in 2015 and available for consumption in Hawai’i

In 3 fish statistically significant (>95% confidence interval) but trace levels of 134Cs was detected.  Given that 137Cs/134Cs ratio in vast majority of the release from the Fukushima site was ~1 the authors were able to determine the fraction of radiocesium present in these fish owing to Fukushima versus legacy sources like atmospheric weapons testing.  Maximum radiocesium levels in the fish approached 0.7-0.8 Bq kg-1 which is more than 1,500 fold lower than conservative levels thought be a health risk set by the FDA (1,200 Bq kg-1).  Most fish had radiocesium attributable to weapons testing fallout. Fukushima radiocesium accounted for ~60% of the radiocesium detected in an Ahi measured by the authors.

Levels of Naturally Occurring 40-Potassium in Fish

Naturally occurring 40K decays with a half life of 1.25 billion years and in taken up into the tissue of marine fish.  The levels of 40K in the fish measured by the authors are summarized in the table below:

Levels of artificial radiocesium and naturally occurring 40-K in fish from Hawai’i

Activity of 40K (Bq kg-1) tended be ~100 fold higher in the fish tissue than radiocesium activities.

Effective Dose of Ionizing Radiation and Health Impact to Fish Consumers

The authors determined the impact of fish consumption on the ionizing radiation dose experienced by individuals consuming an average amount of fish per year (24.1 kg per year or 53.1 pounds per year).  The table below compares the dose in nanoSieverts per year (10-9 Sv yr-1) owing to historic and Fukushima sourced radiocesium and naturally occurring 40K in seafood.

Committed effective dose to fish consumers from artificial (human made) and naturally occurring 40-K

Converting isotope activities in the fish to dose demonstrates that 40K is responsible for ~100 times higher dose than 134Cs + 137Cs. Doses to humans from consuming the fish owing to radiocesium were 0.02–0.2 µ Sv yr-1, while doses of 6–20 µ Sv yr-1 were contributed by the natural 40K present in the same fish. These levels of radioisotopes and the calculated doses to consumers are similar to those reported by the InFORM project who have looked at Pacific salmon returning to rivers and streams in North America over the last 3-4 years. It is important to note that the bulk of ionizing radiation dose to fish consumers normally results from 210-Polonium (210Po half life 138 days) naturally present in the fish but this isotope was not measured in the Azouz and Dulai study.


Fukushima derived radioisotopes 134Cs and 137Cs were detected (at 95% confidence interval) in 3 of 13 fish caught in the North Pacific and commonly consumed by people living in the Hawaiian islands.  The radiocesium in most fish reflected contamination largely present in the North Pacific Ocean owing to atmospheric weapons testing during the last century.  The levels of radiocesium in the fish were a small fraction of the levels of naturally occurring radioisotopes like 40K.  The committed effective dose of ionizing radiation to fish consumers is dominated by the naturally occurring isotopes and do not remotely approach levels known to represent a significant or measurable health risk to human beings.  The results of this study agree with previously published research and results of the InFORM project which focuses on the impact of the Fukushima disaster on the marine ecosystem and public health in North America.

Open Access Review of Fukushima Radionuclide Source Term, Fate and Impact in Pacific

Schematic of Fukushima Daiichi sources of 137-Cs from Buesseler et al. (in press). Atmospheric fallout (1) and direct ocean discharges (2) represent total petabecquerels (PBq = 10^15 Bq) released in the first month of the meltdowns. Groundwater fluxes (3) and river runoff (4) are approximate ranges for the first year in terabecquerels (TBq = 10^12 Bq), a unit 1,000 times smaller than the PBq used for fallout and direct discharge. Details on source term estimates can be found in the paper ( (Buesseler et al. 2017)

by Jay T Cullen

The purpose of this post is to bring to the attention of readers here a review of the available measurements and science based investigations of the Fukushima Daiichi Nuclear Power Plant (FDNPP) and its impact on the Pacific Ocean ecosystem and public health. This post is part of an ongoing effort to summarize scientifically rigorous information about the disaster for interested readers. The new paper is a product of a working group on radioactivity in the ocean convened by the Scientific Committee on Ocean Research (SCOR) an international non-governmental non-profit organization. I highly recommend this paper for anyone who wishes to better understand what the international scientific community has found about the marine release, fate and impact of FDNPP-derived radionuclides in the marine environment.  The working group was made up of 10 experts from 9 different countries, including Japan, and published the open access paper in Annual Reviews.  The main findings of the review were as follows:

  • The amount of 137Cs released from the plant was ~50-fold less than the fall out from nuclear weapons testing in the 20th century and ~5-fold lower than that released from Chernobyl in 1986. Total releases from Fukushima are similar to the discharges of 137Cs from the nuclear fuel reprocessing plant Sellafield in the UK
  • Initial releases in the weeks to months after the disaster which began on March 11, 2011 dwarf those from aggregated ongoing releases from the plant site
  • The majority of radionuclide releases ended up in the Pacific Ocean with most deposition and input occurring close to the FDNPP
  • Current range of estimates of the total 137Cs ocean source term are 15-25 PBq (PBq = 1015 Becquerel where a Bq is one nuclear decay event per second). While many other radionuclides were released from FDNPP, the most likely isotopes to represent a health risk to the marine ecosystem and public are those of Cs given their longer half-lives for radioactive decay (134Cs = ~2 yrs; 137Cs = ~30 yrs) and higher relative abundance compared to other isotopes of concern in the FDNPP source term
  • Because Cs is very soluble it rapidly dispersed in the ocean after the disaster given mixing, transport and dilution by ocean currents.  Peak levels of 137Cs occurred close to the plant in 2011 where activity concentrations near FDNPP was tens of millions of times higher than before the accident. By 2014 137Cs concentrations in the central North Pacific was about six times the remaining weapons testing fallout and about 2-3 times higher than prior fallout levels in the northeast Pacific near to North America. Most of the fallout remains concentrated in the top few hundred meters of the ocean. Measurements being made by the Fukushima InFORM project indicate that maximum 137Cs levels off the North American coast are likely to occur this year before declining to levels associated with background nuclear weapon testing before the accident by about the end of this decade
  • There are unlikely to be measurable effects on marine life with the exception of coastal areas very close to FDNPP immediately after the disaster. Monitoring of fish species in Fukushima Prefecture show that about 50% of samples in coastal waters had radiocesium levels above the Japanese 100 Bq kg-1 limit, but that by 2015 this had dropped to less than 1% measuring over the limit. High levels continue to be found in fish around and in the FDNPP port
  • Given levels in seawater and marine organisms measurable impacts to human health through contact with the ocean and the consumption of seafoods are very unlikely

There are many informative graphics and moderately technical summaries of available studies found in the new paper.  The authors highlight the difficulty of monitoring radionuclides in the ocean  given the dynamic nature of the sea and logistical challenges presented by the temporal and spatial scales and low levels of FDNPP derived contamination going forward.  In addition to providing ongoing assessments of risk to the environment from the disaster it is likely that useful information about ocean circulation will be obtained through continued monitoring efforts.

Authors Lower Fukushima Cesium in North Pacific By Order of Magnitude

By Jay T. Cullen


One of the goals of the InFORM project is to make measurements of radionuclides in the North Pacific Ocean to determine maximum activities that will determine impacts on the marine ecosystem and residents of the west coast. The purpose of this post is to bring to the attention of readers a recently published correction to a prominent model that predicts the activity of Fukushima derived Cesium-137 (137-Cs, half life ~30 years) in seawater of the North Pacific. The diary is part of an ongoing series aimed at discussing research addressing the impact of the Fukushima nuclear disaster on the health of the North Pacific Ocean and inhabitants of North America’s west coast. Predictions of a model by Rossi and colleagues published in Deep-Sea Research in 2013 of the evolution of the plume of seawater contaminated by the Fukushima triple meltdowns are an order of magnitude too high. Rather than a range of ~1-30 Bq/m^3 reported previously maximum activities off the west coast of North America are likely to be ~3 Bq/m^3 or about more than 25 times lower than maximum activities measured in the Pacific in the mid-20th century resulting from atmospheric weapons tests. These activities are not likely to represent significant radiological health risks to the North Pacific ecosystem or residents of the North American west coast.

A paper by Rossi et al. (2013) used a Lagrangian model to predict the temporal and spatial evolution of the seawater plume contaminated by the Fukushima nuclear disaster beginning in March 2011. The model predicted a range of 10-30 Bq/m^3 137-Cs in waters off the coast of North America at 49 degrees North latitude as demonstrated in the figure shown below:

Activities of 137-Cs predicted by the Rossi et al. (2013) model on the continental shelves of North America at two latitudes and off Hawaii over time.

This model predicted higher maximum 137-Cs activities in seawater in the North Pacific compared with a similar model published by Behrens et al. (2012) that had maximum activities off of North America reaching only ~1-2 Bq/m^3.

Recently, after comments from Professor Michio Aoyama of Japan, Rossi and colleagues recognized an error in their model and have published a correction to their 2013 study here. The error resulted in a factor of 10 overestimation of maximum activities of 137-Cs in the Pacific such that maximum 137-Cs off N. America will likely be between 1 and 3 Bq/m^3. The corrections to the model do not affect the conclusions of the study and results from the 2013 study are easy scaled to the more accurate values given the Langrangian approach used by the authors in the original work.

The figure below shows the time evolution of the plume at various latitudes along the international date line and compares the model output with measurements made by Aoyama et al. (2013) along the international dateline at about 40 degrees N in 2012.

Activities of 137-Cs predicted by the Rossi et al. model along the international dateline in the N. Pacific over time at various latitudes.

The factor of 10 lower activity correction better agrees with the Behrens et al. (2012) modeling study and measurements of 137-Cs in seawater made by Japanese and North American scientists.

Maximum activities of ~1-3 Bq/m^3 as the heart of the contaminated plume reaches the North American coast in the coming 2 year period are roughly 25-fold lower than 137-Cs activities in the North Pacific circa 1960 resulting from atmospheric weapons testing. Therefore, it is unlikely that 137-Cs activities of 3 Bq/m^3 or associated radionuclides released at lower total activities from Fukushima will represent significant health risks to the North Pacific ecosystem.

Ongoing monitoring of radionuclide activities in the North Pacific is required to ground-truth models of Pacific Ocean circulation and plume evolution and to provide the best information to determine likely impact to residents of North America.

Comparing 20th Century Strontium and Cesium Isotopes From Atmospheric Weapons Testing in the Pacific to Fukushima Sources

by Jay T. Cullen

The purpose of this post is to compare the concentrations of Sr-90 and Cs-137 in the North Pacific Ocean over the last 50 years to the concentrations predicted to arrive on the west coast associated with waters affected by release of radionuclides from the Fukushima-Daiichi Nuclear Power Plant. Given present levels that are being measured in the eastern Pacific and barring release rates that significantly exceed past rates in March-April 2011, when release rates were 10,000-100,000 times greater than ongoing releases at the plant, the impact on marine organisms and the marine environment is likely to be less significant than impacts owing to radioactivity in the 20th century. What follows is a comparison of the concentrations measured and predicted over much of the Pacific owing to Fukushima to the concentrations that were present in the mid-1960s from the fallout of atmospheric weapons testing that is free from any discussion of safe doses or models of radiation exposure to organisms.

Continue reading Comparing 20th Century Strontium and Cesium Isotopes From Atmospheric Weapons Testing in the Pacific to Fukushima Sources

Misunderstanding Ocean Transport Models of the Fukushima Radionuclide Plume in the Pacific

by Jay T. Cullen


This post is part of an ongoing series that endeavors to provide useful and accurate information about: 1) the fate of Fukushima derived radionuclides in the Pacific Ocean, and, 2) the impact of these radionuclides on the marine ecosystem and the west coast of North America. The purpose of this diary is to draw attention to a number of poorly researched posts about a recently published study (unfortunately this study is behind a publisher pay-wall) in a Chinese journal that predicts a concentrated plume of radioactive elements from Fukushima arriving on the west coast. It is an unfortunate but common example of how news aggregation sites can misinterpret the results of a scientific study and misinform the public.

What models can and cannot say about the Fukushima plume

The study in question by Fu and co-workers published in the Journal of Ocean University of China in 2014 (behind pay-wall unfortunately) is wholly incapable of describing the behavior of dissolved radionuclides in the plume that is now arriving on the west coast of North America.

From the paper the authors themselves state in the methods that:

“In the study, the radioactive pollutant in the ocean is treated as a mixture of multiple Lagrangian particulates, and each particulate represents a radioactive element. The particulates can move in both horizontal and vertical directions, but cannot diffuse and mix with surrounding seawater.”

What this means is that rather than being allowed to mix and diffuse (or decay or sink after becoming associated with particles) the radionuclides are treated as neutrally buoyant drifters. The model, therefore, greatly overestimates the concentrations of radionuclides reaching the west coast of North America in the plume.

For those interested in models using accurate physics that will allow for an accurate prediction of radionuclide concentrations consult the following studies:

Behrens et al. (2012) and Rossi et al. (2013) (behind pay-wall)

Snapshot of the high-resolution (0.1°) model field, taken at the end of the tracer injection period (end of April, model year 0): shading indicates the thickness of the surface mixed layer (in m); contouring illustrates the surface velocity field indicated by local stream lines.
Snapshot of the high-resolution (0.1°) model field, from Behrens et al. taken at the end of the tracer injection period (end of April, model year 0): shading indicates the thickness of the surface mixed layer (in m); contouring illustrates the surface velocity field indicated by local stream lines and clearly identifies the high velocity Kuroshio and Kuroshio extension.

The Behrens et al. study is open-access while the Rossi et al. study is not. Measurements taken in the North Pacific by Canada’s Department of Fisheries and Oceans and InFORM team member Dr. John Smith indicate that the Rossi et al. study predicts the arrival time of the plume on the west coast but overestimates the activity of the Fukushima derived radionuclide 137-Cs. Behrens et al. predict a too late time of arrival but with lower activities that appear to more realistic. It important to note that these models carry the own simplifications and assumptions (e.g. see section 3.4 Caveats of the Behrens et al. (2012) study) and that recent measurements suggest that some of the Fukushima plume is being dispersed to the south rather than to the east in the Pacific (e.g. Kumamoto et al. (2014) open-access; more on this study in a forthcoming post).

Articles that confuse the conclusions of the Chinese study are a good example of poor reporting on an important subject. The example here was originally spawned by Energy News who have a history of inaccurate reporting on Fukushima and then propagated through the web by uncritical followers of the site.