90Sr and 89sr in seawater off Japan as a consequence of the Fukushima Dai-ichi nuclear accident

Biogeosciences, 10, 3649–3659, 2013 www.biogeosciences.net/10/3649/2013/ doi:10.5194/bg-10-3649-2013 © Author(s) 2013. CC Attribution 3.0 License. EGU Journal Logos (RGB) Advances in Geosciences Open Access Natural Hazards and Earth System Sciences Open Access Annales Geophysicae Open Access Nonlinear Processes in Geophysics Open Access Atmospheric Chemistry and Physics Open Access Atmospheric Chemistry and Physics Open Access Discussions Atmospheric Measurement Techniques Open Access Atmospheric Measurement Techniques Open Access Discussions Biogeosciences Open Access Open Access Biogeosciences Discussions Climate of the Past Open Access Open Access Climate of the Past Discussions Earth System Dynamics Open Access Open Access Earth System Dynamics Discussions Geoscientific Instrumentation Methods and Data Systems Open Access Geoscientific Instrumentation Methods and Data Systems Open Access Discussions Geoscientific Model Development Open Access Open Access Geoscientific Model Development Discussions Hydrology and Earth System Sciences Open Access Hydrology and Earth System Sciences Open Access Discussions Ocean Science Open Access Open Access Ocean Science Discussions Solid Earth Open Access Open Access Solid Earth Discussions The Cryosphere Open Access Open Access The Cryosphere Discussions Natural Hazards and Earth System Sciences Open Access Discussions 90Sr and 89Sr in seawater off Japan as a consequence of the Fukushima Dai-ichi nuclear accident N. Casacuberta1,2, P. Masqu´ e2, J. Garcia-Orellana2, R. Garcia-Tenorio3, and K.O. Buesseler4 1Laboratory of Ion Beam Physics, ETH-Zurich, Schafmattstrasse 20, 8093 Zurich, Switzerland 2Institut de Ci` encia i Tecnologia Ambientals & Departament de F´ ısica, Universitat Aut` onoma de Barcelona, 08193 Cerdanyola del Vall` es, Barcelona, Spain 3Departamento de F´ ısica Aplicada II, Universidad de Sevilla, Avda. Reina Mercedes, 2, 41012 Seville, Spain 4Woods Hole Oceanographic Institution, Dept. of Marine Chemistry and Geochemistry Clark 447, MS 25, Woods Hole, MA 02543, USA Correspondence to: N. Casacuberta ([email protected]) Received: 27 December 2012 – Published in Biogeosciences Discuss.: 5 February 2013 Revised: 25 April 2013 – Accepted: 2 May 2013 – Published: 3 June 2013 Abstract. The impact of the earthquake and tsunami on the east coast of Japan on 11 March 2011 caused a loss of power at the Fukushima Dai-ichi nuclear power plant (NPP) that resulted in one of the most important releases of artificial radioactivity into the environment. Although several works were devoted to evaluating the atmospheric dispersion of radionuclides, the impact of the discharges to the ocean has been less investigated. Here we evaluate the distribution of Fukushima-derived 90Sr (n=57) and 89Sr (n=19) throughout waters 30–600km offshore in June 2011. Concentrations of 90Sr and 89Sr in both surface waters and shallow profiles ranged from 0.8±0.2 to 85±3Bqm−3and from 19±6 to 265±74Bqm−3, respectively. Because of its short half-life, all measured 89Sr was due to the accident, while the 90Sr concentrations can be compared to the background levels in the Pacific Ocean of about 1.2Bqm−3. Fukushima-derived radiostrontium was mainly detected north of Kuroshio Current, as this was acting as a southern boundary for transport. The highest activities were associated with near-shore eddies, and larger inventories were found in the closest stations to Fukushima NPP. The data evidence a major influence of direct liquid discharges of radiostrontium compared to the atmospheric deposition. Existing 137Cs data reported from the same samples allowed us to establish a 90Sr/137Cs ratio of 0.0256±0.0006 in seawater off Fukushima, being significantly different than that of the global atmospheric fallout (i.e., 0.63) and may be used in future studies to track waters coming from the east coast of Japan. Liquid discharges of 90Sr to the ocean were estimated, resulting in an inventory of 53±1TBq of 90Sr in the inshore study area in June 2011 and total releases of 90Sr ranging from 90 to 900TBq, depending upon the reported estimates of 137Cs releases that are considered. 1 Introduction Radiostrontium, and in particular 90Sr (T1/2=28.9a), is an artificial radionuclide that has been liberated to the environment in the last decades mainly due to the atmospheric nuclear weapon testing performed in the 1950s and 1960s and the direct releases into the sea from nuclear reprocessing facilities. Given that oceans cover more than 70% of the Earth’s surface, much of these releases of 90Sr now remain dissolved into seawater (IAEA, 2005). Concentrations of 90Sr in oceans are summarized in the World Marine Radioactivity Studies (WOMARS), and a global 137Cs/90Sr ratio of 1.5 has been established. As its chemical behavior is similar to that of calcium, strontium is deposited in biological systems, especially in those tissues with high calcium content such as bone. The highly energetic 0.546MeV beta particles, and the additional exposure to the 2.2MeV beta particles released during decay of its short-lived 90Y daughter isotope, make 90Sr a radionuclide of health concern (ICRP, 1993). However, not much literature exists on the presence of 89Sr (T1/2=50.52d) in the oceans. This radionuclide is Published by Copernicus Publications on behalf of the European Geosciences Union. 3650 N. Casacuberta et al.: 90Sr and 89Sr in seawater off Japan also a primary fission product, but with a much shorter halflife than 90Sr. As a result of the earthquake on 11 March 2011, and subsequent tsunami, radionuclide contaminants were accidently released into the environment from the Fukushima Dai-ichi nuclear power plant (NPP). The emissions of artificial radionuclides both to the atmosphere (Chino et al., 2011) and the marine environment (i.e., Povinec et al., 2012; Buesseler et al., 2011) have been broadly reported. In particular, 134Cs and 137Cs concentrations at the nuclear power plant ocean discharge channels peaked in early April at levels that reached several millions times higher compared with the preexisting levels in this area (Buesseler et al., 2011), thus representing the most important accidental artificial radioactive release into the sea (Bailly du Bois et al., 2011). Simulations of regional ocean models using 137Cs concentrations in the ocean off Fukushima allowed estimating the total amount of 137Cs directly discharged to the ocean, ranging from 1 to more than 20PBq (Kawamura et al., 2011; Tsumune et al., 2012; Bailly du Bois et al., 2011). Although temporal changes of 131I, 134Cs and 137Cs concentrations in surface water adjacent to Fukushima have been well documented, releases of radiostrontium into the sea are poorly investigated (Povinec et al., 2012). The main reason for this lack of data resides in the difficulties of preconcentrating and analyzing 90Sr in seawater. The recent study by Povinec et al. (2012) was the first report assessing the impact of radiostrontium released from the Fukushima Dai-ichi nuclear power plant into the marine environment. Their data on 90Sr in coastal seawater (near the discharging channel in Fukushima Dai-ichi NPP) were obtained from the Ministry of Education, Culture, Sports, Science and Technology (MEXT). Concentrations of 90Sr reached up to four orders of magnitude greater than the pre-Fukushima levels (1.2Bqm−3). Between March and November 2011, maxima activities of 90Sr and 89Sr were of 69 and 9.3kBqm−3, respectively, but even higher values were reported after a December 2011 accidental discharge (400 and 140kBqm−3for 90Sr and 89Sr, respectively) (Povinec et al., 2012). This study is part of an international project, which aimed at understanding the amount, type and fate of radioactive materials released into the northwest Pacific Ocean and to see whether the Kuroshio Current was acting as a southern boundary for radionuclide transportation in the marine environment. This is the first survey of 90Sr and 89Sr offshore Japan after the Fukushima NPP accident in March 2011. In particular, here we evaluate the distributions of 90Sr and 89Sr released from Fukushima into the water column in an area encompassing 30–600km off Japan in May–June 2011. 2 Materials and methods 2.1 Seawater sampling A total of 57 20L seawater samples, both from surface and vertical profiles, were collected from the R/V Ka’imikai-o- Kanaloa (KOK) between 4 and 18 June 2011 (Fig. 1). Sample codes, collection date, locations, depths, salinity and temperature of each sample are provided in Table 1. Surface waters were collected using a deck-mounted pump and hose, which was placed about 2m below the sea surface. A pre-filter cartridge was used to remove the suspended matter. Deeper samples (10–200m) were collected with Niskin bottles mounted on a CTD/Rosette at stations #3, #19, #21, #23, #25, #27 and #29 (see Table 1). Samples were spiked with 200mg of Sr2+and stored for its subsequent analysis at the Universitat Aut` onoma de Barcelona and Universidad de Sevilla. 2.2 90Sr purification and quantification through 90Y counting by proportional counter Once in the laboratory, samples were weighted and acidified with concentrated HCl. The purification method for the analysis of 90Sr was based on the protocol described in Waples and Orlandini (2010), and adapted to seawater samples by Casacuberta et al. (2013). The method consists in the measurement of the daughter of 90Sr, 90Y (T1/2=64h), which is in equilibrium with 90Sr, by beta counting. Samples were spiked with 10mg of stable Y, and 30mg of Fe3+as ferrous sulfate was added to the samples. After letting samples equilibrate overnight, pH was raised up to 8 to coprecipitate yttrium with the iron hydroxides. The precipitate was then collected onto a nitrocellulose filter (0.22µm, 14cm, Whatman) and subsequently dissolved in 20mL HNO38M. Isolation of yttrium from other beta emitters (mainly Th) was performed by passing the solution through an anion ionexchange column (Bio-Rad AG1-X8, 100-200 mesh). The eluate, 40mL of HNO38M, was then evaporated to reconstitute the sample in 20mL 1MHCl. A further separation on a cation exchange column (Bio-Rad AG50W-X8, 100– 200mesh) was required for the final purification of yttrium from lead isotopes. Sample was transferredto the column and 80mL of 1MHCl were added to wash the resin. Finally, yttrium was eluted by adding 30mL 6MHCl recovered by precipitation using ferric hydroxides and filtration onto a nitrocellulose filter (0.22µm, 47mm, Whatman), which was dried and placed onto a beta holder for the measurement of the 90Y beta emissions. Samples were measured on a 5-position beta counter manufactured by RISØ National Laboratories (Roskilde, Denmark) during 60min intervals over approximately 60–64h. Filters were re-measured after 1week and 3weeks to verify that detected counts corresponded to the decay of 90Y and no other beta emitters, including 91Y (halflife: 58.5d), interfered. Yttrium chemical recoveries (that Biogeosciences, 10, 3649–3659, 2013 www.biogeosciences.net/10/3649/2013/ N. Casacuberta et al.: 90Sr and 89Sr in seawater off Japan 3651 Fig. 1. Sampling stations, with their corresponding station numbers (see Table 1 for more details), superimposed on surface velocities (from Buesseler et al., 2012). Grey contour on the left represents the Japan coast, whereas the red star stands for the Fukushima Dai-ichi NPP. ranged from 10 to 85%) were assessed by determining the stable Y concentrations by ICP-MS. 2.3 89Sr purification and quantification by proportional counter The supernatants from the first separation of iron hydroxides described in Sect. 2.2 were stored and sent to Universidad de Sevilla to proceed with an alternative direct determination of 90Sr in the same samples. The initial objective was to perform these measurements as a quality control exercise to validate the results obtained by the method described previously in Sect. 2.2. However, the method used for the validation of 90Sr allowed for the determination of both 89Sr and 90Sr (the first one only measured in some samples due to the long time delay between sampling and measuring). The conventional radiochemical procedure (Harvey et al., 1989) applied for isolation and purification of the 89Sr and 90Sr in the supplied supernatants was based in an initial oxalate precipitation, followed by a transformation of the oxalate precipitate into carbonate by calcination to 550◦C. Subsequently, calcium was mostly eliminated by fuming acid precipitation, followed by the dissolution of the precipitate in water, its evaporation to dryness and its posterior redissolution in nitric acid. Barium was afterwards separated from strontium by formation of barium chromate precipitation, and yttrium was removed by two successive hydroxide precipitations. Strontium was finally purified by its precipitation with a saturated solution of sodium carbonate, subsequent conversion of the precipitate to strontium nitrate by adding concentrated HNO3, the dissolution of the precipitate with hot water, and the final precipitation of the Sr as SrCO3, by adding again a saturated solution of sodium carbonate. The final SrCO3 precipitate was homogeneously distributed over the counting planchet and dried until constant weight in a desiccator. The radiochemical yields were determined gravimetrically and double-checked by ICP-MS in a selected number of samples. The presence of natural Sr in samples was considered after its determination in aliquots by ICP-MS. The quantification of 89Sr and 90Sr was performed by using a 10-channel low-level proportional counter (LB790, Berthold Technologies). Two measurements at different time intervals were performed to determine the concentrations of 89Sr and 90Sr in the sample under study at the time of sampling. The first one was performed 36–72h just after the 90Y separation and the second one at least 18days after the first measurement. In all cases, each measurement was performed during at least two cycles of 600min. From the net counting rates obtained in the two measurements, the activities of 89Sr and 90Sr were determined, following the methodology described in Herranz et al. (2011), which implies the previous calibration of the counter for determination of the counting efficiencies of 90Sr, 90Y and 89Sr. 3 Results Concentrations of 90Sr (n=57) and 89Sr (n=19) in seawater samples collected at the different stations and depths ranged from 0.8±0.2 to 85±3Bqm−3and from 19±6 to 265±74Bqm−3, respectively (see Table 1). All activities refer to sampling date. 3.1 90Sr in surface seawaters Activities of 90Sr in surface seawater samples ranged from 1.1±0.2 to 85±3Bqm−3. The highest concentrations were found along the northern coast of Japan, being particularly enhanced about 100–200kmSE from Fukushima, at stations 29 and 30 (Fig. 2a). The concentration of 90Sr in surface samples from near-shore stations ranged from 10 to 85Bqm−3. The lowest activities were located in the southeastern areas of the studied area (i.e., stations 1, 2, 3 and 10) and northern from Fukushima NPP (stations 20 and 24). These concentrations were close to the background value of 1.2Bqm−3 reported for this area before the Fukushima NPP accident (IAEA, 2005). The distribution pattern of 90Sr concurs with the surface 137Cs and 134Cs dispersal reported in Buesseler et al. (2012), from the same cruise (Fig. 2b and c). 3.2 90Sr in shallow profiles Specific activities of 90Sr in shallow profiles (down to 200m) taken at stations 19, 21, 23, 25, 27 and 29 are detailed in Table 1 and summarized in Fig. 3. In all cases, the majority of 90Sr was found in the first 50m depth horizon. In general, 90Sr concentrations measured at the deepest waters of each profile correspond to the expected for background levels (1.2Bqm−3) of 90Sr in this area. The only exception was www.biogeosciences.net/10/3649/2013/ Biogeosciences, 10, 3649–3659, 2013 3652 N. Casacuberta et al.: 90Sr and 89Sr in seawater off Japan Table 1. Concentrations of 90Sr and 89Sr in seawater samples collected off Fukushima in June 2011. Information on collection date location, depth, salinity and temperature is also provided. All activities refer to sampling date. 90Sr activities reported here correspond exclusively to the samples measured following the method described in Sect. 2.2. Station Sample Collection Depth Longitude Latitude Salinity Temperature 90Sr 89Sr # code date (m) (psu) (◦C) (Bqm−3)(Bqm−3) 1 Sr-1 06.06.2011 2 146.5924 33.5967 34.72 20.46 1.3±0.2 2 Sr-2 07.06.2011 2 147.0002 34.2989 34.69 28.81 1.4±0.9 3 Sr-3 07.06.2011 200 147.0004 34.9980 34.65 16.23 1.5±0.6 Sr-4 07.06.2011 100 147.0004 34.9980 34.76 18.37 2.3±1.4 Sr-5 07.06.2011 50 147.0004 34.9980 34.72 19.12 4.8±5.7 Sr-6 07.06.2011 20 147.0004 34.9980 34.72 20.66 1.5±0.2 4 Sr-7 08.06.2011 2 147.0026 35.2986 34.33 17.29 10.1±0.4 24±4 5 Sr-8 08.06.2011 2 147.0227 36.0090 34.71 18.23 2.7±1.2 6 Sr-9 08.06.2011 2 147.0068 36.2998 34.57 15.95 3.4±0.2 7 Sr-10 09.06.2011 2 147.0000 37.5947 34.18 15.53 6.3±0.3 8 Sr-11 09.06.2011 2 146.5999 37.3010 34.34 15.21 5.6±0.4 9 Sr-12 09.06.2011 2 147.0124 37.5901 33.59 10.85 3.2±0.3 10 Sr-13 10.06.2011 2 143.5978 37.5986 33.92 12.27 9.3±2.5 27±5 11 Sr-14 10.06.2011 2 144.0018 37.3066 34.24 13.77 7.8±2.0 26±6 12 Sr-15 10.06.2011 2 143.5953 36.5623 33.86 11.71 3.5±0.3 14 Sr-16 11.06.2011 2 144.0033 35.5976 34.73 21.15 1.1±0.2 17 Sr-17 12.06.2011 2 143.0061 36.3021 34.53 20.96 1.1±0.2 18 Sr-18 12.06.2011 2 142.5998 36.5994 34.27 15.62 6.1±0.3 19 Sr-19 12.06.2011 2 142.5972 37.3029 34.20 14.96 8.8±0.4 27±6 Sr-20 12.06.2011 200 142.9980 37.4940 33.81 5.90 0.8±0.2 Sr-21 12.06.2011 100 142.9980 37.4940 34.15 9.42 1.7±0.2 Sr-22 12.06.2011 50 142.9980 37.4940 33.99 10.03 4.7±0.4 Sr-23 12.06.2011 20 142.9980 37.4940 34.19 14.85 8.1±0.4 25±4 20 Sr-24 13.06.2011 2 143.0016 37.6000 34.53 14.25 1.7±0.2 21 Sr-25 13.06.2011 2 142.2947 37.3008 34.18 14.88 7.4±0,4 22±9 Sr-26 13.06.2011 200 142.5020 37.5010 33.71 4.66 1.0±0.3 Sr-27 13.06.2011 100 142.5020 37.5010 33.69 6.46 1.2±0.3 Sr-28 13.06.2011 50 142.5020 37.5010 33.98 9.39 2.6±0.2 Sr-29 13.06.2011 20 142.5020 37.5010 34.19 14.87 7.3±0.4 22 Sr-30 13.06.2011 2 142.3005 38.0001 34.34 18.31 6.9±0.4 19±6 23 Sr-31 14.06.2011 2 141.6000 37.2974 34.79 17.14 26±2 57±14 Sr-32 14.06.2011 100 142.0050 37.4810 33.73 7.32 4.8±0.7 Sr-33 14.06.2011 50 142.0050 37.4810 34.32 12.83 34±2 137±43 Sr-34 14.06.2011 20 142.0050 37.4810 34.67 16.06 37±3 132±34 Sr-35 14.06.2011 10 142.0050 37.4810 34.79 17.17 28±1 65±21 24 Sr-36 14.06.2011 2 141.5984 38.0004 34.48 19.21 2.5±0.2 25 Sr-37 15.06.2011 2 141.2673 37.3187 33.98 15.08 17±1 29±13 Sr-38 15.06.2011 90 141.4410 37.5170 33.89 8.98 3.6±0.9 Sr-39 15.06.2011 50 141.4410 37.5170 33.89 9.30 12±4 43±16 Sr-40 15.06.2011 20 141.4410 37.5170 33.88 9.60 12±2 Sr-41 15.06.2011 10 141.4410 37.5170 34.20 14.11 11±1 Sr-42 15.06.2011 120 141.4410 37.5170 33.89 8.98 1.9±0.2 Sr-43 15.06.2011 100 141.4410 37.5170 33.90 9.74 3.3±0.3 27 Sr-45 15.06.2011 100 141.4000 36.5010 34.20 10.13 2.2±0.4 Sr-46 15.06.2011 50 141.4000 36.5010 34.13 11.59 16±6 Sr-47 15.06.2011 20 141.4000 36.5010 34.14 13.74 12±1 35±14 Sr-48 15.06.2011 10 141.4000 36.5010 34.11 14.80 18±1 45±18 Sr-50 15.06.2011 2 141.2390 36.2981 34.11 14.78 33±8 67±27 28 Sr-49 16.06.2011 2 141.2391 36.0033 34.38 20.63 1.1±0.2 Biogeosciences, 10, 3649–3659, 2013 www.biogeosciences.net/10/3649/2013/ N. Casacuberta et al.: 90Sr and 89Sr in seawater off Japan 3653 Table 1. Continued. Station Sample Collection Depth Longitude Latitude Salinity Temperature 90Sr 89Sr # code date (m) (psu) (◦C) (Bqm−3)(Bqm−3) 29 Sr-51 16.06.2011 2 142.0058 36.2986 34.03 17.06 85±3 265±74 Sr-52 16.06.2011 100 142.0030 36.4990 34.20 10.37 1.2±0.2 Sr-53 16.06.2011 50 142.0030 36.4990 34.38 12.50 4.4±0.3 Sr-54 16.06.2011 20 142.0030 36.4990 34.28 14.52 12±1 32±7 Sr-55 16.06.2011 10 142.0030 36.4990 34.03 17.06 15±1 30 Sr-56 17.06.2011 2 142.2972 36.2975 33.94 17.49 28±3 31 Sr-57 17.06.2011 2 142.3001 36.5992 33.91 16.55 15±2 32 Sr-58 17.06.2011 2 142.0030 36.5989 34.19 17.66 11±1 41±14 Fig. 2. (a) Activities of 90Sr in surface seawater samples (expressed in Bqm−3). Grey contour on the left represents the Japanese coast. The red star stands for the Fukushima NPP. The right-hand plots show the specific activity of 137Cs (b) and 134Cs (c) in surface waters from Buesseler et al. (2012). station 23, where concentrations at 100m depth were still enhanced (i.e., 4.8±0.7Bqm−3). In general, specific activities of 90Sr decrease with depth, yet some differences are noticed in the profile trends: whereas concentrations of 90Sr in stations 19, 21, 27 and 29 decrease exponentially with depth, the two closest stations of Fukushima Dai-ichi NPP (23 and 25) showed high 90Sr concentrations between 20 and 50m depth. The particular behavior of 90Sr at stations 23 and 25 corresponds to an area affected by near-shore eddies. This is observed in the section represented in Fig. 4a, which comprises stations 25 (left), 23, 21, 19 and 11 (right). 90Sr concentrations at station 23 are significantly higher (from 5 to 37Bqm−3) than the concentrations in the two profiles closer to shore (stations 25 and 21), due to the presence of this nearshore eddy (see Fig. 4b and c). Distribution of 90Sr according to water masses is further discussed in Sect. 4.2. Inventories of 90Sr in the upper 100m were calculated for the six profiles located closer to the coast (Fig. 5). These inventories were in all instances higher than the expected inventory of 90Sr for this area prior to the Fukushima accident, calculated assuming a background 90Sr concentration of1.2Bqm−3in thewater column (i.e., 120Bqm−2). Station 23 holds the greatest inventory of 90Sr (2511±102Bqm−2), followed by stations 27 and 29, these last ones corresponding to the stations that presented the highest 90Sr concentrations at surface. 3.3 89Sr results in surface and shallow profiles 89Sr was quantified in 19 samples (Table 1), and concentrations ranged from 19±6 to 265±74Bqm−3in surface waters, reaching a maximum subsurface value of 137±43Bqm−3at 50m depth at station 23. Due to the short half-life of 89Sr (50.53d) and the long time delay between sampling and counting (up to 5–6months), uncertainties were larger than 40% in some of the measured samples. www.biogeosciences.net/10/3649/2013/ Biogeosciences, 10, 3649–3659, 2013 3654 N. Casacuberta et al.: 90Sr and 89Sr in seawater off Japan Fig. 3. Specific activities of 90Sr in shallow profiles at the 6 closest stations of Fukushima Dai-ichi NPP. Note the different scales on the x-axis. 4 Discussion 4.1 Impact of the Fukushima Dai-ichi nuclear accident in seawaters off Japan Contrary to other radionuclides such as 137Cs, 134Cs and 131I (Hisamatsu, 2012), the amounts of radiostrontium in the form of liquid radioactive wastes released into the sea after the Fukushima Dai-ichi nuclear accident are less known, mostly due to the more complex and time-consuming analytical procedures used for its determination. However, as the cooling water directly interacted with ruptured nuclear fuel rods, large amounts of radiostrontium may have been released directly into the ocean as well (Povinec et al., 2012). Indeed, our results of 90Sr and 89Sr in both surface seawater and depth profiles showed a clear impact of the Fukushima nuclear accident in the waters off Japan. Before April 2011, the main 90Sr source in the Japanese waters was due to global fallout from atmospheric weapons testing, including the influence of 11 Chinese atmospheric nuclear weapons tests in the 1970s (Ikeuchi, 2003). Just after these events, the average 90Sr concentrations on the eastern coast off Japan were between 4.3 and 5.7Bqm−3. Average concentrations measured during the 1980s and 1990s ranged between 2.5 and 2.8Bqm−3, and the last 90Sr concentrations reported ranged from 1.4 to 1.8Bqm−3in surface waters (Ikeuchi et al., 1999 and Povinec et al., 2010) and between 0.21 and 0.58Bqm−3 in deep waters (>2000m) (Ikeuchi et al., 1999). Decaycorrected concentrations of 90Sr in the West Pacific Ocean in 2011 would therefore have decreased to a range of 1.1 to 1.4Bqm−3(IAEA, 2005; Povinec et al., 2012), agreeing with the lowest value found in this study for surface waters (i.e., 1.1±0.2Bqm−3). The Fukushima Dai-ichi nuclear accident produced an increase of the 90Sr concentrations in the waters off Japan, such that three months after the accident they were 1 to 2 orders of magnitude higher than the background in some areas. Fukushima-derived radiostrontium was present not only in surface waters but also at depth, reaching maximum values of 37±3Bqm−3at 20m. The presence of 90Sr in the waters off Japan derived from the Fukushima nuclear accident was accompanied by the presence of 89Sr. Given its short half-life (50.57d), the occurrence of this radionuclide is a faultless signal of the Fukushima-derived release. Moreover, the 89Sr/90Sr ratio pointed to be Fukushima-derived radiostrontium, as the decay-corrected ratio measured in this study (see Sect. 4.3) is in excellent agreement with the ratio reported for the reactor core inventory on 11 March 2011 (Nishikara et al., 2012). 4.2 Distribution of 90Sr on the coast off Japan Three assumptions were made for discussing the 90Sr distribution in the waters off Japan in May–June 2011: (i) the Sr peak in direct oceanic release into the Pacific Ocean took place on 6 April, as it did for Cs isotopes (Buesseler et al., 2011); (ii) both Cs and Sr behave geochemically similarly (Fig. 2) in the waters off Japan, as both are dissolved in seawater and have a weak interaction to sediment (IAEA, 2005; Povinec et al., 2003); and (iii) most of the 90Sr left in the KOK sampling area at the time of the cruise came from the direct oceanic discharges, as it was the case for Cs (Rypina et al., 2013, this issue). After the radionuclide peak release, 90Sr in seawaters would be dispersed and distributed through the Japanese coast governed by advection processes occurring in this specific area of the Pacific Ocean. In particular, this is a region where three main currents converge: the Oyashio, the Kuroshio and the Tsugaru currents (Talley et al., 1995). The Oyashio front is defined as the southern limit of subpolar waters with very low salinity, which are driven from the northern to the southern coast of Japan (Fig. 6). The Kuroshio Current originates in the subtropical gyre, consisting of low density, warm, high salinity surface waters, coming from the southwest of the Japanese coast (Fig. 6). Finally, the Tsugaru Current (not present in May–June) is a source of warm and saline water that enters the Sea of Japan through the Tsushima and Tsugaru straits (Talley et al., 1995). Using data from surface drifters released during the KOK cruise, Rypina et al. (2013, this issue) performed a numerical model to investigate the dominant mechanisms governing the short-term spread of radionuclides in this area. From their model, and similar to 137Cs results, the following can be stated: (i) the absence of the 90Sr at the southernmost stations (stations 1, 2, 3, 14 and 28) was attributed to the Kuroshio Current, which prevents a southward progression Biogeosciences, 10, 3649–3659, 2013 www.biogeosciences.net/10/3649/2013/ N. Casacuberta et al.: 90Sr and 89Sr in seawater off Japan 3655 Fig. 4. (a) Specific activities of 90Sr (expressed in Bqm−3)and (b) temperature and (c) salinity contour plots in a section at Fukushima NPP latitude. It comprises stations 25 (left, closest to Fukushima), 23, 21, 19 and 11 (right). of Fukushima contaminants such as 90Sr; (ii) the largest concentration of 90Sr was shown to be associated with the semipermanent near-shore eddy and stirring them for a long time around the eddy perimeter (stations 29, 30 and 31); and (iii) the intermediate 90Sr concentrations at the westernmost row of stations (stations 25 and 27) were explained by the coolant water that continued to leak from the reactor after the peak release. In addition to the dispersion of radiostrontium in surface waters off Japan, 90Sr may also have penetrated into deep waters, either by diffusion or advection processes, as shown by our measurements of Fukushima-derived radiostrontium down to 100–200m depth. 4.3 89Sr/90Sr ratio The 89Sr/90Sr net ratio (subtracting 90Sr background concentrations) can be used to estimate the time delay between sampling and the Fukushima accident (time of nuclear reactor cease). Although the date of the nuclear accident was well known, the ratio was used to corroborate it and thus prove the 89Sr source being Fukushima NPP in a sort of academic exercise. Time delay calculated for each sample (Table 2) was estimated by using the following Eq. (1): t= Ln 89Sr 90Sr obs 89Sr 90Sr reactor −λ89Sr ,(1) where (89Sr/90Sr)obs is the net ratio measured at each sample at sampling time, (89Sr/90Sr)reactor the ratio at reactor (i.e., 11.8±0.4 according to Povinec et al., 2012), and λ89Sr the decay constant of 89Sr (0.0137d−1). Resulting values range Fig. 5. Total inventories of 90Sr in the upper 100m of the water column at stations 19, 21, 23, 25, 27 and 29. The solid line represents the background value for this area (120±40Bqm−2), and the numbers above the bars indicate the Fukushima-derived inventory (subtracting the background). from 74 to 135d, which on average is 98±18d and is in good agreement with the elapsed time since the accident. 4.4 Estimate of 90Sr discharge to the sea Several works have been devoted to quantifying the amount of 137Cs released into the sea due to the Fukushima Dai-ichi nuclear accident, ranging from 3.5 to 34PBq (Tsumune et al., 2012; Buesseler et al., 2012; Bailly-du-Bois et al., 2012; Charette et al., 2013, this issue). Here we calculated the release of 90Sr by using the 90Sr/137Cs ratio data obtained from the analyzed samples (Buesseler et al., 2012), once www.biogeosciences.net/10/3649/2013/ Biogeosciences, 10, 3649–3659, 2013 3656 N. Casacuberta et al.: 90Sr and 89Sr in seawater off Japan Fig. 6. Surface salinity and temperature measured during the KOK cruise in June 2011; the Oyashio and Kuroshio currents are depicted. Grey contour on the left represents the Japanese coast. Table 2. 89Sr/90Sr ratio in seawater samples and resulting ages obtained from Eq. (1). Station Depth Sample 89Sr/90Sr Water age (m) code (days) 4 2 Sr-7 2.7±0.4 107±17 10 2 Sr-13 3.2±1.1 95±31 11 2 Sr-14 3.9±1.3 81±28 19 2 Sr-19 3.5±0.8 88±21 19 20 Sr-23 3.6±0.7 86±16 21 2 Sr-25 3.4±1.5 90±39 22 2 Sr-30 3.3±1.1 92±30 23 2 Sr-31 2.3±0.6 120±30 23 50 Sr-33 4.1±1.3 76±24 23 20 Sr-34 3.7±1.0 84±23 23 10 Sr-35 2.4±0.8 116±39 25 2 Sr-37 1.8±0.8 135±62 25 50 Sr-39 4.1±2.0 78±38 27 20 Sr-47 3.1±1.3 97±40 27 10 Sr-48 2.6±1.1 109±46 27 2 Sr-50 2.1±1.0 127±59 29 2 Sr-51 3.2±0.9 96±28 29 20 Sr-54 2.9±0.7 102±24 32 2 Sr-58 4.3±1.5 74±26 corrected for background concentrations (1.2 and 2.5Bqm−3 for 90Sr and 137Cs, respectively) (IAEA, 2005) (Fig. 7). The 90Sr/137Cs ratio for the samples collected during the KOK cruise was 0.0265±0.0006, which is higher than the ratio of 0.01 reported by Povinec et al. (2012). Their value was the average of the data obtained in April 2011 at site 1, close to one of the discharging channels. Indeed if calculating the average value at their sites 1 and 2 (close to discharging channels), 7 and 8 (15km off the Fukushima NPP) between April and June, the resulting ratio 90Sr/137Cs is 0.0282, in very good agreement with our data. Therefore, ratio 90Sr/137Cs calculated here (0.0265±0.0006) is representative of the average ratios found in Povinec et al. (2012) once surface waters off the coast of Japan were homogenized. The total amount of 90Sr in the study area in June 2011, taking the inventory of 2PBq of 137Cs estimated by Buesseler et al. (2012) and using the 90Sr/137Cs ratio of 0.0265±0.0006, would then be of 53±1TBq. This figure is constrained within the inshore box, but the actual total releases of 90Sr would have been greater. Given the range of published estimates of 137Cs releases due to direct discharge from the NPP, the total amount of 90Sr released could range from 90 to 900TBq (Table 3). It is worth noticing that in December 2011, the Tokyo ElectricPowerCompanyreported anaccidental leakagefrom the evaporative condensation apparatus at Fukushima Daiichi NPP of approximately 150 L of water containing 11GBq of 89Sr and 15GBq of 90Sr. In fact, activities of 90Sr measured at the discharge site in December 2011 exceeded 137Cs concentrations. Thus the 90Sr/137Cs ratio that we measured cannot be used to predict the total amounts of 90Sr released through direct ocean discharge after December 2011. The amount of 90Sr released because of the Fukushima Dai-ichi NPP accident can be compared with the most important nuclear accidents and the discharges of nuclear reprocessing fuel plants. The releases of 90Sr from Chernobyl were quantified in 10PBq (IAEA, 2003), of which 277TBq were discharged to the Black Sea through the Pripyat and Dnieper rivers (Vakulovsky et al., 1994). However, the direct inputs to the ocean were considered to be negligible as the radioactivity release was mainly of atmospheric origin (IAEA, 2005). On the contrary, 90Sr releases into the Pacific Ocean Biogeosciences, 10, 3649–3659, 2013 www.biogeosciences.net/10/3649/2013/ N. Casacuberta et al.: 90Sr and 89Sr in seawater off Japan 3657 Table 3. Amounts of 90Sr released into the sea from the Fukushima NPP by June 2011, in TBq(×1012 Bq), calculated using the published estimates of 137Cs releases and a 90Sr/,137Cs ratio of 0.0265±0.0006. The leakage 90Sr occurred in December 2011 is not included in these estimates. Reference 137Cs (PBq) 90Sr (TBq) Range Average Range Average Tsumune et al. (2011) 3.5±0.7 93±19 Bailly du Bois et al. (2012) 10–34. 22 265–901 583±13 Rypina et al. (2013) 9.5–22 16.2±1.6 252–583 429±43 Charette et al. (2013) 11–16. 292–424 Fig. 7. Fukushima-derived 90Sr/137Cs ratio in surface waters. Background concentrations of both 137Cs (2.1±0.5Bqm−3)and 90Sr (1.2±0.2Bqm−3)have been subtracted. 137Cs values are taken from Buesseler et al. (2012), which were collected simultaneously with the samples for 90Sr presented in this study. The average 90Sr/137Cs ratio is 0.0265±0.0006. due to the nuclear accident in Fukushima are comparable with discharges from the nuclear reprocessing plants in Sellafield and Cap de la Hague. Indeed, the total discharges of 90Sr to the oceans (decay corrected to year 2000) due to the releases comprised between 1970 and 1998 were of 3195 and 744TBq for Sellafield and La Hague, respectively (IAEA, 2005). Thus, the inventory of Fukushima-derived 90Sr in the oceans represents between 3 and 23% of the total amount due to direct discharges. It is important to note here that the Fukushimadischarges,although comparableto the onesfrom Sellafield, have occurred over a much shorter time scale. Compared to the ocean inventory of 90Sr due to global fallout, of about 116PBq estimated in 2011 (IAEA, 2005), the Fukushima-derived 90Sr directly released into the sea would represent less than 1% of 90Sr of the total inventory. Fig. 8. The 90Sr/137Cs ratio as a function of 137Cs specific activities. The grey dashed line represents the oceanic background ratio decay corrected to 2011. Also shown are the 90Sr/137Cs ratios corresponding to the oceanic background ratio, the direct discharges from Fukushima Dai-ichi and the atmospheric deposition. 4.5 Direct releases of 90Sr into the sea versus atmospheric deposition The amount of radiostrontium measured in the study area during the KOK cruise could be a result of two main sources: (i) the localized direct discharge of contaminated coolant waters into the near-shore ocean, and (ii) the fallout of airborne discharge released into the atmosphere during the Fukushima NPP explosions, with subsequent deposition onto the ocean. Strontium is much less volatile than other fission products such as cesium and iodine isotopes, and thus smaller amounts of 90Sr compared to 137Cs were expected to be released into the atmosphere. Indeed, TEPCO published a 90Sr/137Cs ratio in soils of 0.001, which is far below the averageratio measuredin seawater in thisstudy (i.e.,0.0265). If the 90Sr/137Cs ratio of liquid discharges in April–May 2011 was 0.0265 and the ratio from atmospheric release was 0.001, it is clear that the ocean samples had a much stronger influence of the direct discharges of cooling water into the sea www.biogeosciences.net/10/3649/2013/ Biogeosciences, 10, 3649–3659, 2013