How We Analyze Samples and Report Data
Samples collected by scientists and citizens for Our Radioactive Ocean are analyzed in our labs at WHOI using a method that is capable of detecting extremely low levels of radioactivity produced by cesium isotopes in seawater. We report our data in units of Bequerels per cubic meter of seawater (Bq/m3), where one Bequerel is equal to one decay event per second.
- Once we receive a sample in the lab, we first weigh it and measure the salinity of the water. Then we add a known quantity of stable cesium (Cs) to the sample (see step 4 below) and slowly pass (1 ml/min) the sample through a 5 ml column of potassium-nickel-hexacyanoferrate composite ion resin beads. This resin is specifically designed to selectively separate cesium (stable or radioactive) from the sample and has been optimized here for use in seawater samples.1
- We dry the resin and transferred it to a plastic “counting jar,” which we then place in an expensive ($75,000), high-purity germanium well detector made by Canberra Industries2 for between 24 and 72 hours (see 3:05 here). Every time a cesium atom decays, that event is registered in the instrument’s multichannel analyzer, which has the ability to discern energy given off by decay of two critical isotopes of cesium: 134Cs and 137Cs.3 By counting the decay events associated with each isotope, we calculate the total counts per second (cps) for a given sample.
- By periodically analyzing standards samples with known levels of cesium, we can calculate the detector efficiency. With this information, the sample weight, the cps, and the number of cesium gamma events per decay (the so-called “branching ratio”), we can calculate the total activity of 134Cs and 137Cs in each sample.
- By adding a known quantity of stable cesium to each sample at step 1, and then measuring how much passes through the resin column, we can determine a chemical yield for the extraction procedure. This stable cesium “yield monitor” is determined using an inductively coupled mass spectrometer. Extraction yields for the resin columns are typically 96 to 99 percent.
The result for each sample includes the total activity (Bq/m3) and an associated counting error, which is a measure of how precisely we can analyze for Cs isotopes: for example, 2.0 ± 0.2 Bq/m3. This number reflects a combination of uncertainties related to the number of decay events in a sample and are larger for a smaller cps), as well as uncertainties in the processing and calibration steps. To minimize these uncertainties, we regularly participate in proficiency tests with the International Atomic Energy Agency (IAEA) to ensure that our results are not just precise, but extremely accurate when compared to international seawater standards.
Our current detection limit using this method is about 0.1 Bq/m3 for 137Cs and 0.2 Bq/m3 for 134Cs. Values below this are reported as “below detection,” but this detection limit will vary with the sample size, the methods and detector used, and the total time each sample spends on the gamma detector. In general, larger sample sizes (we process a relatively large 20 liter sample), longer counting times (we typically leave a sample on for 48 hours or more), and more efficient detectors (we use some of the world’s most sensitive gamma detectors) lead to the lowest possible detection limits.
We expect samples from the surface waters of the western Pacific that have not been contaminated by the Fukushima source to have 137Cs activity of between 1 and 2 Bq/m3 and for 134Cs to be “below detection.” This is because the only significant source of cesium in the Pacific prior to Fukushima was nuclear weapons testing during the 1950s and 1960s, and with its shorter 2-year half-life, all of the 134Cs from this source would have decayed by now, but because 137Cs has a 30-year half-life, we still see about 25 percent of the amount that was released (50 percent lost in first 30 years, half again of the remaining 50 percent lost in the following 30 years).
By January 2014, about 40 percent of the original Fukushima 134Cs remains in the environment compared to March/April 2011 when the disaster occurred, so we correct our data to account for decay of both cesium isotopes from the time of peak release directly to the ocean from the reactor complex in Fukushima: April 6, 2011. We do this to look for changes in the levels of cesium that result from ocean mixing and dilution, rather than just radioactive decay. For human health concerns, the activity at sampling may be of greater interest, and will be lower than the decay-corrected value we report.
1Kamenik, J., Dulaiova, H., Sebesta, F., Stastna, K., "Fast concentration of dissolved forms of cesium radioisotopes from large seawater samples," Journal of Radioanalytical and Nuclear Chemistry 296(2012): 841-846.
3We look at following energies: 661 keV for 137Cs; 604 and 795 keV for 134Cs.
4 cf., M. Aoyama, M., et al., "Cross equator transport of 137CS from the North Pacific Ocean to South Pacific Ocean (BEAGLE2003 cruises)," Progress in Oceanography 89(2011): 7-16.
January 28, 2014
The first results from seawater samples come from La Jolla and Point Reyes, Calif., and Grayland and Squium, Wash. Four samples from these three locations show no detectable Fukushima cesium. We know this because Fukushima released equal amounts of two isotopes of cesium: the shorter-lived cesium-134 isotope (half-life of 2 years) and the longer-lived cesium-137 (half-life of 30 years). Cesium-137 was found at levels of 1.5 Bq per cubic meter (Bq/m3), but this was already detectable prior to releases at Fukushima and came primarily from nuclear weapons testing in the Pacific during the 1950s and 1960s.[ MORE ]
This so-called "negative" result has two immediate implications. First there should be no health concerns associated with swimming in the ocean as a result of Fukushima contaminants by themselves or as a result of any additional, low-level radioactive dose received from existing human and natural sources of radiation in the ocean (existing levels of cesium-137 are hundreds of times less than the dose provided by naturally occurring potassium-40 in seawater).
Secondly, and just as important from a scientific perspective, the results provide a key baseline from the West Coast prior to the arrival of the Fukushima plume. Models of ocean currents and cesium transport predict that the plume will arrive along the northern sections of the North American Pacific Coast (Alaska and northern British Columbia) sometime in the spring of 2014 and will arrive along the Washington, Oregon, and California coastline over the coming one to two years. The timing and pattern of dispersal underscores the need for samples further to the north, and for additional samples to be collected every few months at sites up and down the coast.
For this reason, we are also pleased to report that funds are already in hand to continue sampling at both the La Jolla and Pt. Reyes locations thanks to the foresight and generous donations of the groups who volunteered to adopt these sites. We expect levels of cesium-134 to become detectable in coming months, but the behavior of coastal currents will likely produce complex results (changing levels over time, arrival in some areas but not others) that cannot be accurately predicted by models. That is why ongoing support for long-term monitoring is so critical, now and in the future.[ LESS ]