Long-term environmental radioactive contamination of Europe due to the Chernobyl accident - Results of the Joint Danube Survey 2013

Author’s Accepted Manuscript Long-term environmental radioactive contamination of Europe due to the Chernobyl accident - Results of the Joint Danube Survey 2013 Franz Josef Maringer, Claudia Ackerl, Andreas Baumgartner, Christopher Burger-Scheidlin, Maria Kocadag, Johannes H. Sterba, Michael Stietka, Jan Matthew Welch

PII: DOI: Reference:

www.elsevier.com/locate/apradiso

S0969-8043(16)30617-0 http://dx.doi.org/10.1016/j.apradiso.2017.02.035 ARI7806

To appear in: Applied Radiation and Isotopes Received date: 27 August 2016 Revised date: 16 February 2017 Accepted date: 21 February 2017 Cite this article as: Franz Josef Maringer, Claudia Ackerl, Andreas Baumgartner, Christopher Burger-Scheidlin, Maria Kocadag, Johannes H. Sterba, Michael Stietka and Jan Matthew Welch, Long-term environmental radioactive contamination of Europe due to the Chernobyl accident - Results of the Joint Danube Survey 2013, Applied Radiation and Isotopes, http://dx.doi.org/10.1016/j.apradiso.2017.02.035 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Long-term environmental radioactive contamination of Europe due to the Chernobyl accident - Results of the Joint Danube Survey 2013 Franz Josef Maringer1,2,3*, Claudia Ackerl2, Andreas Baumgartner2, Christopher Burger-Scheidlin3, Maria Kocadag3, Johannes H. Sterba3, Michael Stietka2, Jan Matthew Welch3 1 BEV – Federal Office of Metrology and Surveying, Austria, 2 BOKU – University of Natural Resources and Life Sciences Vienna, Austria 3 TU Wien – Technische Universität Wien, Austria * Corresponding author: [email protected]

Abstract In the course of the Joint Danube Survey 3 (JDS3), coordinated by the International Commission for the Protection of the Danube River (ICPDR), laboratory ships travelled 2375 km down the Danube River engaging in sampling, processing and on-board analyses during the summer of 2013. The results of the radiometric analysis of 90Sr, 137Cs and natural radionuclides in recent riverbed sediment are presented. The activity concentrations of 90Sr and 137Cs in Danube sediments have been found below 100 Bq/kg. Keywords: Joint Danube Survey 3, Freshwater radioecology, radionuclide metrology, Chernobyl contamination, 90Sr, 137Cs, natural radionuclides

1. Introduction Since the middle of the 20th century, atmospheric nuclear weapons testing as well as nuclear reactor accidents, most prominently the Chernobyl accidents, in what is now Ukrainian territory, have occurred. These contaminations, resulting from the above-mentioned sources, have had their impact on the environment and this includes, quite prominently, freshwater resources. Research in the Danube region began as early as 1967 (Frantz, 1967, Rank, 1976, Tschurlovits, et al., 1980). Later, naturally occurring radionuclides such as 210Pb, 226Ra, 228Ra and 238U were also included in the monitoring process. Elevated levels of artificial radionuclide concentration in rivers lead to increased health risks for populations drinking processed river water or consuming river fish. The use of river water for irrigation can increase health risks by consumption of the agricultural products produced in the irrigated areas. Therefore, it is of importance to investigate the radioecological status of the Danube River ecosphere to evaluate the impacts of radionuclides on the health of the population living in the basin. Currently, 137Cs activity concentration (half-life of 30.05 ± 0.08 years) in Danube water and riverbed sediments originated primarily from the nuclear power accident in Chernobyl (April-May 1986) and secondarily from atmospheric nuclear weapons testing during the 1950s and 1960s. 90Sr (half-life of 28.80 ± 0.07 years) originates primarily from atmospheric weapons tests and, to a much lesser extent, from Chernobyl fallout. Measuring the activity of 90Sr in soil and food (e.g. milk) is, in many countries, part of the monitoring of environmental radioactivity. Data are also collected on 90Sr in sediments of lakes, like Leman and Taniguchi (1972) did in Lake Superior and Lake Ontario in Canada or Mundschenk (1994) for German inland waters. Nevertheless, this work is the first time that a series of samples (68 in total) taken along the course of a single river, the Danube river, were analysed for 90Sr. This work was carried out within the scope of the Joint Danube Survey 3 (JDS3), coordinated by the International Commission for the Protection of the Danube River (ICPDR). The survey was the world’s biggest river research expedition of its kind in 2013, the UN International Year of Water Cooperation. The ICPDR Joint Danube Survey is carried out every six years – JDS1 was held in 2001 (ICPDR, 2002), JDS2 in 2007 (ICPDR, 2008). JDS3 completed the sampling in summer 2013 to start an extensive ecological, biological, chemical and radiometric analysis stage (ICPDR, 2015). The Joint Danube Survey pursued three main objectives: 1) Collect information on parameters not covered in the ongoing monitoring 2) Acquire data that are readily comparable for the entire river because they come from a single source

3 ) Promote the work of the ICPDR and raise awareness of water management. The position along the Danube (in kilometers) is measured from its estuary at the Black Sea and increasing towards its source in the German Black Forest. The Danube can be separated into three large segments: the Northern Alpine Forelands and Vienna Basin (upper section), the Pannonian lowland (middle section) and the Walach lowlands (lower section) (Fig. 1). The upper section reaches from the river’s origin to the Hainburger Gate, the middle section from the Hainburger Gate to the Iron Gate, and the lower section from the Iron Gate to Sulina. The Hainburger Gate is located at the Austrian – Slovak border and the river stretch observed was defined between Danube river kilometres 2857 and 1880. Starting at river kilometre 1880, the middle section goes as far as the Iron Gate, which however, is over 80 km long. Therefor the centre at Danube river km 904 was chosen as a compromise. The last section starts at the Iron Gate and ends at the Black Sea. The main tributaries of the Danube are: the Inn (km 2225; right), Morava (km 1880; left), Drava (km 1383; right), Tisza (km 1215; left), Sava (km 1170; right), Velika Morava (km 1105; right), Iskar (km 637; right), Siret (km 155; left) and Prut (km 134; left). This paper aims to evaluate and visualise the latest measurements of 40K, 90Sr, 137Cs, 226Ra, 228Ra, 228Th, 238 U and 210Pb radionuclides in Danube riverbed sediments, collected in the course of the Joint Danube Survey 3 in summer 2013 and compare results acquired in previous analyses such as the results of the Joint Danube Survey 2 (ICPDR, 2008).

2. Materials and Methods 2.1 Riverbed sediment sampling For six weeks between 13 August and 26 September 2013, the JDS3 laboratory ships travelled 2375 km down the Danube River, through 10 countries, to the Danube Delta. An international Core Team of 20 scientists was responsible for sampling, sample processing, on-board analyses and all survey activities. 68 sites were sampled by the JDS3 Core Team along a 2581 km stretch of the Danube, 15 of which were located in the mouths of tributaries or side arms. Samples from the first two stations in Germany were collected using by car, the remaining 2354 km were sampled by ships. The riverbed sediment samples were taken from the left and right banks of the Danube River by riverbed surface sediment sampling at 1-1.5 m water depth with a sampling net. This was followed by on-board homogenisation and grain size fractioning by wet sieving in order to obtain a 63 µm grain size fraction for radiometric analysis in the laboratory. The activity concentrations of artificial and natural gamma-ray emitting radionuclides and β-particle emitting 90Sr in riverbed sediment samples from the 68 locations along the Danube river were determined by gamma-ray spectrometry and radiochemical analysis. 2.2 Low-level gamma-ray spectrometry of Danube sediment samples The radiometric analysis of the sediment samples (grain size fraction <63 µm; dried at 105°C) was carried out by low-level gamma-spectrometry. By this method, which was carried out by a low-level germanium detector in the specially shielded Low-level Counting Laboratory Arsenal, Vienna (Aiginger et al., 1986), the activity concentration of the gamma emitting radionuclides in the samples were analysed. The radiometric equipment consists of a low-level coaxial broad-energy HP-Ge-detector (BSI, GCD50195X, S/N: 1936-13), Cryostat: ultra low-background U-type, 30 l Dewar, shielding: low-level lead (120 mm), liners: Sn (1 mm) / electrolytic Cu (1.5 mm), crystal: diam. 80.6 mm – length: 40.5 mm, window: carbon epoxy, face deadlier: ~0.3 µm, rel. eff.: 52.2 %, peak to Compton ratio: 66:1, FWHM: 1715 eV at 1332 keV, 678 eV at 122 keV, 481 eV at 14.4 keV, detecting gamma energies between 100 to 2800 keV. Calibration of the detector was carried out by NIST-, IAEA-, and PTB-standard reference materials. The calculation of the radionuclide concentrations and data handling were done by SpectraLineHandy 1.5 (©LSRM Ltd.) and additional software developed by the departmental research group. To show exemplarily the total measurement uncertainties of the applied low-level gamma-ray

spectrometry technique, the uncertainty budget of sediment sample is given in Tab. 1.

234

Th activity concentration of a typical Danube

2.3 Radiochemical analysis of 90Sr in Danube sediment samples Analysis of the activity of 90Sr in soil is, in many countries, part of the monitoring of environmental radioactivity. This is the first time in Austria, that a series of 68 bottom sediment samples taken along the course of the Danube River have been analysed. Before analysis, the sediment samples had been dried in the laboratory at 105 °C. All results are given for dried samples (grain size fraction <63 µm, dry weight at 105 °C). The yield of strontium extracted from the geological matrix was determined by total reflection X-ray fluorescence (TXRF). The 90Sr was then processed radiochemically, separated with strontium-selective resins and quantified by liquid scintillation counting (LSC). Prior to application of radiochemical separation and liquid scintillation quantification of 90Sr in the Danube sediment samples, the efficiency of strontium extraction from the geologic matrix under acidic conditions was optimized. Inductively Coupled Plasma Mass Spectrometry (ICP-MS), Neutron Activation Analysis (NAA) and Total Reflection X-ray Fluorescence (TXRF) were employed in attempts to assess the amounts of strontium extracted from a representative sediment sample. Extraction duration and process repetition were assessed in order to optimize both quality of 90Sr quantification and the use of costly strontium selective resin in subsequent radiochemical separation steps. The sediment sub-samples (2.00 g) were weighed into a round-bottomed flask (10 ml) and concentrated (14.8M) HNO3 (2.5 ml), distilled water (2.5 mL) and 30% H2O2 (1 ml) were added, a reflux condenser was mounted, and the flask was heated for 30 minutes on a boiling water bath. The hot solution was filtered and the round-bottomed flask was washed 5 times with 1 ml 8M HNO3 which was also subsequently filtered. The filtrate (extractant and washings) was collected and set aside, while the solid residue was scraped from the filter paper into a 10 ml round-bottomed flask. The extraction and washing procedure was repeated twice more, resulting in three liquid fractions of extractant and washing from each cycle in three separate vessels. All samples were yellowish in colour but the first was the most strongly coloured, supposedly from dissolved iron salts. A control solution was also produced with 2.5 ml conc. HNO3, 2.5 ml dist. water and 1 ml H2O2 and all four samples were diluted to the same of 10 ml. The scheme of the complete radiochemical sample preparation for 90Sr analysis is given in Fig. 2. For the separation of the strontium from other radionuclides, crown-ether (4,4'(5')-di-tbutylcyclohexano 18-crown-6) –based strontium-selective resins were used (Horowitz, 1991; Pimpl, 1995). Approximately 350 mg of resin were weighed into a disposable plastic column and preconditioned with 5 ml 8M HNO3 for at least 1.5 hours. Excess acid was then eluted from the column and a frit applied to the top of the column. The solution with the sample was applied to the resin and the beaker containing the sample was washed twice with 1 ml 8M HNO3 which was also loaded onto the resin. The loaded resin was washed with 5 ml of a solution with 3M HNO 3 and 0,05M oxalic acid to elute ions other than strontium. In the next step, the strontium was eluted from the column with 10 ml 0.01 M HNO3 and the resulting fraction was then evaporated to dryness. Due to spectroscopic interference, no sensible results could be achieved by ICP-MS. Spectroscopic interferences are produced after the atomization of the sample. A distinction is made between chemical effects (e.g. the formation of oxides), physical effects (e.g. formation of doubly charged particles) and overlap of masses, through isotopes. In the case of strontium, the formation of FeO2 (the same mass as 88 Sr) causes strong interference. Neutron activation analysis also yielded no useful results, since soil consists of several elements that can be activated. This leads to a complex spectrum of gamma radiation with a large number of overlapping peaks. Matters were complicated further by the fact that 85Sr, which we are interested in, has a gamma line at 514 keV, which is very close to the gamma line of the annihilation peak at 511 keV. In certain circumstances, when there is little strontium in the sample, the 85Sr-peak is concealed by the annihilation peak.

Total Reflection X-ray Fluorescence (TXRF) was the most effective way to determine the amount of strontium extracted from the samples. However, this method was complicated by the large amount of iron in the samples which necessitated dilution. The samples were diluted such that the iron signals did not oversaturate the detector but, on the other hand, the strontium signals remained detectable. The results of the TXRF showed that after one hour of heating 89 % of the strontium is extracted, after two hours 98.5 % and after three hours 98.6 %. The radiochemical separation largely followed the procedure of Steinhauser et al., 2013, and Kocadag et al., 2013, used to identify the activity concentrations of 90Sr in soil and vegetation samples from Japan after the Fukushima accident. The principal difference is the extraction time since Steinhauser et al., 2013, boiled their samples for 30 minutes, in order to dissolve all materials completely. As long as the sediment samples were heated, complete dissolution was not necessary because the 90Sr is thought to be confined to the particle surfaces and has not migrated into the soil particles. To measure the efficiency of the strontium extraction, the authors followed the same procedure as Steinhauser et al., 2013. 30 mg of Sr(NO3)2 was activated in the TRIGA Mark Reactor II at the Atominstitut, TU Wien, and dissolved in 30 ml water. Then the procedure of strontium extraction was performed as described in Fig. 2. Since 85Sr is a gamma-emitting radionuclide, gamma-spectroscopy of the intermediate steps could be performed easily. The recovery of the strontium was about 87 % whereas the recoveries of the extraction of Steinhauser et al., 2013, were in a range of 92.4 - 98.2 %. LSC measurement and colour quench correction Since the samples prepared have a yellowish colouration, a correction for the effects of this colouration on LSC counting efficiency must be made. In this case K2CrO4 or K2Cr2O7 may be used to imitate the colouration of the samples and account for the effect of colour quenching on the LSC results. A correction for colour quenching may be applied by measuring a series of standard samples of known activity of 90Sr with different amounts of colourant. When each sample has been counted, a quench curve is produced showing the relation between tSIE (transformed Special Index of External standard spectrum) and count efficiency (on the basis of the known activity). The quench curve produced shows that counting efficiency is only effected by very strong colouration (Fig. 3). Since none of the samples of sediment extract was strongly coloured, colour quenching does not significantly influence the results of sample counting. The first five samples were counted for 1000 minutes, but it was decided to decrease the measuring time to 700 minutes such that two samples per day could be measured with relative measurement uncertainty due to counting statistics below 2 % (k=1). Uncertainty analysis Since the radiochemical extraction and separation processes are laborious with several intermediate steps, it is important to discuss the uncertainties of each step. The first step was the extraction of strontium from the sediment, in which we found by TXRF-measurements, that with an extraction time of two hours almost complete extraction (98.6 %) of the strontium from the samples was possible. Therefore, uncertainty introduced by the extraction step is minimal, so we focus on the intermediate steps of the radiochemical process. To be sure that all of the strontium is dissolved, the soil needs to be heated in acid for longer than two hours until complete dissolution, which is not reasonable for so many samples. The next steps were filtration and evaporation. To make a precise statement regarding the efficiencies, the whole process was carried out with 85Sr. In these steps, a recovery of 87.2 % ± 2.4 is reached, the loss of strontium occurs mainly in the filter. The strontium resins used for the separation had a very good efficiency of about 99 % at the first use. All resins used for the analysis of the samples were used only one time. The final step was the LSC-measurement. Since long measuring times of 12 hours were used, the counting statistic’s uncertainty was very low (< 2 %). 3. Results and discussion The naturally-occurring potassium radionuclide 40K is found all over the planet and represents only a very small fraction of the total amount of potassium in the environment (0.012 %; NIST, 2016). Fig. 4 shows the results acquired by measurements of this radionuclide. An increasing concentration of radioactive potassium can be observed the closer the sampling locations were to the Black Sea. A

notable result was obtained at river kilometre locations 1170 and 1159, where the level of 40K was significantly lower than anywhere else on the entire length of the Danube. At km 1159 the result was below the decision limit of 51.1 Bq/kg. In general, the 90Sr activity concentrations were found below the 10 Bq/kg decision limit apart from a few exceptions (Fig. 5). Most of the 90Sr activity was found in the upper half of the Danube river. As shown in Fig. 5, there are four sampling locations with 90Sr activity concentration above the decision limit: km 2120 Asten, (Austria) with 30 ± 4 Bq/kg, km 1790 Moson (Hungary) with 57 ± 9 Bq/kg, km 1384 Dravski Kut (Croatia) with 48 ± 7 Bq/kg and km 1252, Novi-Sad (Serbia) with 19 ± 3 Bq/kg (k=1; grain size fraction < 63 µm; dry weight 105°C) . The measured 90Sr activity decreases along the river, which results from the flow of the river but also the contamination from 1986. According to UNSCEAR, 2000, the Danube states, Austria, Germany and Romania were mostly affected by radioactive deposition after the Chernobyl accident. Unfortunately there are no results by UNSCEAR 2000 for Serbia (former Yugoslavia) due to the Baltic war, but a higher contamination can be expected there, since on the Romanian side of the border a higher activity concentration was also measured. This could be a possible reason for the high 90Sr values, when one bears in mind that the maxima migrate with time and with the flow of the river. Due to the inflow of several tributaries and therefore increasing water volume and low-contaminated soil particles input, the 90Sr may be diluted, as confirmed by the results. The measured 137Cs activities can be traced to the nuclear accident at the Chernobyl reactor in 1986 as well as to the atmospheric nuclear tests conducted during the 1950s and 1960s. A decrease in 137Cs levels can be seen from a linear regression of the data in Fig. 6. This is due to the development and distribution of the radioactive plume in the days after the Chernobyl disaster with winds blowing west towards central and northern Europe and the subsequent precipitation (UNSCEAR, 2000). Another contributing factor is the dependence on the mineralogical composition and the grain size distribution of the riverbed sediment (Maringer et al., 2015). Both 228Ra and 228Th can be found in the 232Th series with 228Ra decaying via a beta (β—) decay to 228Ac and another beta decay to 228Th. As a result, similar outcomes were expected when comparing the two nuclides (Fig. 7). Because of the comparably long half-life of 228Ra it is often used as a monitoring tracer in water quality assessment, since many consecutive isotopes have relatively short half-lives and their presence is hard to measure accurately without significant efforts. 228Th is one of the radioactive elements in the sequence which takes longer to decay and can thus be detected more easily. The elements 238U, 226Ra and 210Pb (Fig. 8) are part of the natural uranium-radium decay series. The radioactive elements from the series have become important in understanding erosion and chemical weathering. The activities and ratios between nuclides can be used as an indicator of the previously stated processes (Dossetoa, et al., 2008). In Fig. 9 the activity concentration ratio in which 228Ra and 226Ra are present on the length of the Danube is shown. This ratio is commonly used as indicator for the origin of sediment particles, since the two radionuclides originate from two different decay series, the thorium series and the uranium-radium series, respectively (Maringer et al., 2009). Generally, a tendency towards an increase in the ratio can be found when advancing downstream. Values between 0.8 - 1.0 in the middle section of the Danube can be observed, with some higher ratio results towards the mouth of the river. Fig. 10 incorporates data from 4 different research projects from the years 1988, 2004, 2007 and 2013 showing 137Cs activity concentration levels including the results of the European joint research project Aquaterra (2010). Between 1988 and the subsequent data of 2004, a decrease of one order of magnitude is notable. The results from 2007 align with the previous measurements, but some decrease in activity can be observed. It seems the processes may be best described by an exponential decay.

The measurements conducted in 2013 in the course of JDS3 show an intermingled progression. The overall results are significantly lower than data acquired six years prior. Again, the values obtained are lower downstream than at the river source. Notably, the data indicates values have reached activity concentrations of less than 10 Bq / kg in the lowest section of the river. 4. Conclusions Eight radionuclides were analysed from Danube riverbed sediment samples by low-level gamma-ray spectrometry and radiochemical analysis. The 68 riverbed samples were collected at the Joint Danube Survey 3 (JDS3) in summer 2013. Radionuclide activity concentrations show different distributions, most likely influenced by the soil contamination after the Chernobyl accident in Europe and the variety of rock present in the river basin as well as many geological, geographical, and anthropogenic influences that exceed the scale of this paper. 27 years after the environmental contamination due to the Chernobyl accident, the 90Sr contamination level of the Danube riverbed sediment does not exceed the 10 Bq/kg decision limit with only a few exceptions. 40K, 90Sr, 137Cs, 226Ra, 228Ra, 228Th, 238U and 210Pb have been radiometrically analysed by lowlevel gamma-ray spectrometry. Generally, the 137Cs activity concentrations of the riverbed sediment samples are below 100 Bq/kg. A significant decrease in activity concentration of sediment along the river was observed. This set of observations is of great importance since it is the first complete radiometric analysis of riverbed sediments for 90Sr along the entire length of the Danube River. Notably, the chronological development of the 137Cs levels give insight into the ecological decay of the element and makes it possible to apply a theoretical model that may give a projection for possible development in the future. As part of JDS3, the evaluation of the sediment samples should provide a representative snapshot of the current activity concentration levels in the Danube as well as permit further research to be conducted with comparable data in order to be able to detect and interpret changes in years to come, not just for 137 Cs but also for many of the other nuclides.

Acknowledgements The authors wish to thank all of those who supported and assisted in carrying out the riverbed sediment sampling and sample preparation at JDS3, including Core Team members, national teams and supporting laboratories, as well as sponsors from the private sector. Their support was crucial in carrying out this up-to-date radioecological study of the Danube River.

References Aiginger, H., Maringer, F.J., Rank, D., Unfried, E., 1986. A New Laboratory for Routine Low-LevelMeasurements (BVFA Arsenal, Wien). Nucl. Instr. Meth. B17, pp 435-437. Aquaterra, 2010. BASIN – Integration of Research in AquaTerra River Basins, Chapter 7. In: Advanced Tools and Models to Improve River Basin Management in Europe in the Context of Global Change – AquaTerra. Ed. M. Finkel, P. Grathwohl, and J. Barth. IWA Publishing, London UK. Horwitz P. E., Dietz M. L., Fisher D.E, Separation and Preconcentration of Strontium from Biological, Environmental, and Nuclear Waste Samples by Extraction Chromatography Using a Crown Ether , Anal. Chem. 1991, 63, pp 522-525. Kocadag M., Musilek A., Steinhauser G., 2013. On the Interference of Pb-210 in the Determination of Sr90 Using a Strontium Specific Resin. Nuclear Technology & Radiation Protection 28(2), pp 163-168.

Lerman A. and Taniguchi H., Strontium-90 — diffusional transport in sediments of the Great Lakes, Journal of geographical Research, 1972, Vol.77, pp 474-481. Mundschenk H., Verfahren zur Bestimmung von Strontium-98 und Strontium-90 in Oberflächenwasser im Normalfall, Leitstelle für Oberflächenwasser, Schwebstoff und Sediment in Binnengewässern, 1994, ISSN 1865-8725. Pimpl M., 89Sr/90Sr-Determination in soils and sediments using crown ethers for Ca/Sr-separation, Journal of Radioanalytical and Nuclear Chemistry, 1995, Vol. 194, Issue 2, pp 311-318. Steinhauser G., Schauer V., Shozugawa K., 2013, Concentration of Strontium-90 at Selected Hot Spots in Japan. PLOS ONE, 8(3):e57760. Doi:10.1371 /journal.pone.0057760. ICPDR, 2002. JOINT DANUBE SURVEY -Technical Report of the International Commission for the Protection of the Danube River. International Commission for the Protection of the Danube River, Vienna. ICPDR, 2008. Joint Danube Survey 2. Final Scientific Report, Joint Danube Survey 3. A Comprehensive Analysis of Danube Water Quality. International Commission for the Protection of the Danube River, Vienna. ICPDR, 2015. Joint Danube Survey 3. Final Scientific Report, Joint Danube Survey 3. A Comprehensive Analysis of Danube Water Quality. International Commission for the Protection of the Danube River. ISBN 978-3-200-03795-3, Vienna. Dossetoa, A., Bourdonb, B. & Turnera, S. P., 2008. Uranium-series isotopes in river materials: Insights into the timescales of erosion and sediment transport. Earth and Planetary Science Letters Volume 265, Issues 1–2, 15 January, pp 1-17. Frantz, A., 1967. The radioactivity of the Danube. In: Limnology of the Danube, Volume IV. E. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart. pp 84- 96. Maringer, F. J., Gruber, V., Hrachowitz, M., Baumgartner, A., Weilner, S., Seidel, C., 2009. Long-term monitoring of the Danube river - Sampling techniques, radionulide metrology and radioecological assessment. Applied Radiation and Isotopes 67, pp 894-900. Maringer, F.J., Baumgarter, A., Seidel, C., Stietka, M., 2015. Radioactivity in the Danube. In: The Handbook of Environmental Chemistry 39. The Danube River Basin. Ed. Igor Liska. Springer. pp 271-284. NIST, 2016. Basic Atomic Spectroscopic Data. National Institute of Standards and Technology, 2016. [Online] Available at: http://physics.nist.gov/PhysRefData/Handbook/Tables/potassiumtable1.htm [Accessed 30 September 2016]. Rank, D., 1976. Measurement of tritium and radio-carbon in the environment. Proceedings of the OESRAD-Meeting 1976, Arsenal. Austrian Radiation Protection Association, Vienna. Tschurlovits, M., Buchtela, K., Sas-Hubicki, J. & Unfried, E., 1980. Determination of Cs-137, Sr-90 and gamma spectroscopy of water samples from the Danube river. IAEA-TECDOC-229, Vienna. pp 9-22. UNSCEAR, 2000. Report to the General Assembly, with scientific annexes, Sources and effects of ionizing radiation. Vol. II, Effects, Annex J. United Nations Scientific Committee on the Effects of Atomic Radiation. UNSCEAR, Vienna.

Fig. 1. Geography of the Danube River

Fig. 2. Schematic sequence of extraction, separation and measurement of 90Sr in sediment samples

Fig. 3. Colour quench correction curve

Fig. 4. 40K activity concentration of Danube riverbed sediment samples, August/September 2013 (grain size fraction <63 µm; dry weight 105°C)

40

K

Siret

Tisza

K

lin. regr.

800

600

Sava

Velika Morava

1000

200

1200

Inn

Drava

Iskar

400

Danube river km

0

200

400

600

800

1400

1600

1800

2000

2200

2400

0 2600

Activity concentration (Bq/kg)

Morava

40

Prut

1000

Fig. 5. 90Sr activity concentration of Danube riverbed sediment samples, August/September 2013 (grain size fraction <63 µm; dry weight 105°C)

80 Sr

Siret

Prut

Tisza

Morava

60

Iskar

Velika Morava

Drava

20

Sava

40

Inn

Activity concentration (Bq/kg)

90

decision limit

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

2400

2600

0

Danube river km

Fig. 6. 137Cs activity concentration of Danube riverbed sediment samples, August/September 2013 (grain size fraction <63 µm; dry weight 105°C)

60

40

Prut

30

20

Siret

Activity concentration (Bq/kg)

50

Cs lin. regr. 137Cs

Tisza

Morava

137

0

200

400

Iskar

Velika Morava

1000

600

Sava

1200

1600

1800

2000

2200

2400

2600

0

800

Drava

1400

Inn

10

Danube river km

Fig. 7. 228Ra and 228Th activity concentration of Danube riverbed sediment samples, August/September 2013 (grain size fraction <63 µm; dry weight 105°C)

100 228

Ra

228

Th

Prut

228

Th

lin. regr.

Siret

Tisza

Ra

lin. regr.

80

60

40

Sava

Velika Morava

1200

1000

Iskar

Drava

1400

20 Inn

Activity concentration (Bq/kg)

Morava

228

0

200

400

600

800

1600

1800

2000

2200

2400

2600

0

Danube river km

Fig. 8. 238U, 226Ra and 210Pb activity concentration of Danube riverbed sediment samples, August/September 2013 (grain size fraction <63 µm; dry weight 105°C)

180 U

lin.regr. 226

210

U

226

Ra

210

Pb

Pb

lin. regr.

Prut

120

238

Ra

lin. regr.

140

Siret

100 80 60

0

200

400

Iskar

600

Velika Morava

1000

1400

1600

1800

2000

2200

2400

2600

0

800

Sava

Inn

20

1200

40 Drava

Activity concentration (Bq/kg)

Tisza

Morava

238

160

Danube river km

Fig. 9. Ratio of activity concentrations a(226Ra) / a(226Ra) of Danube riverbed sediment samples, August/September 2013 (grain size fraction <63 µm; dry weight 105°C)

Prut

Tisza

Ra / 226Ra lin. regr. 228Ra / 226Ra

Siret

228

Morava

1.6 1.4 1.2 1.0 0.8 0.6

Sava

Velika Morava

1200

1000

Iskar

Drava

0.2

1400

0.4 Inn

Activity concentration ratio a (228Ra) / a (226Ra)

1.8

0

200

400

600

800

1600

1800

2000

2200

2400

2600

0.0

Danube river km

Fig. 10. 137Cs of Danube riverbed sediment samples on the entire length of the Danube River, data from 1988 to 2013 (grain size fraction <63 µm; dry weight 105°C)

10000 1988 (IAD) Danube lin. regr. 1988 (IAD) Danube 2004 (AQUATERRA) Danube lin. regr. 2004 (AQUATERRA) Danube 2007 (JDS2) Danube lin. regr. 2007 (JDS2) Danube 2013 (JDS3) Danube lin. regr. 2013 (JDS3) Danube

Tisza

Cs

Morava

Siret Prut

1000

100

Sava

Velika Morava

1200

1000

Iskar

Drava

1400

10

Inn

Activity concentration (Bq/kg)

137

0

200

400

600

800

1600

1800

2000

2200

2400

2600

1

Danube river km

Table 1: Uncertainty budget of 234Th activity concentration of a typical Danube sediment sample analysed by routine gamma spectrometry at the Low-Level Counting Laboratory Arsenal, Vienna, Austria 234

Th activity concentration (63.3 kev) a = (xxx ± yy) Bq.kg-1 Relative standard uncertainties (k=1) ui x 102 evaluated by method Contribution due to A B Sample mass 1.5 Counting statistic 2.9 Dead time 0.2 Counting time 0.2 Background 0.2 Half-life 0.1 Gamma emission probability 2.1 Efficiency calibration 7.4 Sample dimension 0.2 Sample bulk density 0.3 Coincidence summing correction 0.3 Quadratic summation 8.0 2.6 Relative combined standard uncertainty (k=1), uc 8.4

Highlights  Radioactivity analysis results of the ICPDR Joint Danube Survey 3  Radioactivity analysis of Danube river bottom sediment samples  90Sr, 137Cs, 40K, 210Pb, 226Ra, 228Th, 228Ra and 238U are analysed  Traceability and uncertainty budget of the measurement methods are established  Significant decrease in 137Cs activity in sediment along the river is shown