127I ratios in Scottish coastal surface sea water: Geographical and temporal responses to changing emissions

Applied Geochemistry Applied Geochemistry 22 (2007) 619–627 www.elsevier.com/locate/apgeochem

129

I/127I ratios in Scottish coastal surface sea water: Geographical and temporal responses to changing emissions Christoph Schnabel a,*, Vale´rie Olive a, Mariko Atarashi-Andoh b, Andrew Dougans a, Robert M. Ellam a, Stewart Freeman a, Colin Maden a, Martin Stocker c, Hans-Arno Synal c, Lukas Wacker c, Sheng Xu a a

Scottish Universities Environmental Research Centre, East Kilbride G75 0QF, United Kingdom b Japan Atomic Energy Agency, Tokai, Ibaraki 319-1195, Japan c Institute of Particle Physics, ETH Zurich, CH-8093 Zurich, Switzerland Available online 22 December 2006

Abstract This work constitutes the first survey of I isotope ratios for Scottish sea water including the first data for the west of Scotland. These data are of importance because of the proximity to the world’s second largest emission source of 129I to the sea, the Sellafield nuclear reprocessing plant, because of the increasing importance of the sea to land transfer of 129I and also as input data for dose estimates based on this pathway of 129I. 129I/127I ratios in SW Scotland reached 3 · 106 in 2004. No strong variation of I isotope ratios was found from 2003 to 2005 in Scottish sea waters. Iodine isotope ratios increased by about a factor of 6 from 1992 to 2003 in NE Scotland, in agreement with the increase of liquid 129I emissions from Sellafield over that time period. It is demonstrated that 129I/127I ratios agree better than 129I concentrations for samples from similar locations taken in very close temporal proximity, indicating that this ratio is more appropriate to interpret than the radionuclide concentration.  2007 Elsevier Ltd. All rights reserved.

1. Introduction 129

I is a long-lived (T1/2 = 15.7 Ma) fission product which is released in large quantities from nuclear reprocessing plants. About 1330 kg 129I were discharged to the Irish Sea from Sellafield (1952–2004) (Atarashi-Andoh et al., 2007) and about 2320 kg have entered the English Channel from La Hague (1966–2000) (Lo´pez-Gutie´rrez et al., 2004). A com* Corresponding author. Tel.: +44 (0) 1355 270188; fax: +44 (0) 1355 229898. E-mail address: [email protected] (C. Schnabel).

parison of these quantities with about 100 kg of 129 I calculated for the pre-nuclear mixed marine surface layer of 100 m depth (Raisbeck et al., 1995) demonstrates the dominance of these emissions for the global marine 129I budget. Such discharges have increased the pre-nuclear marine 129I/127I ratio of about 1.5 · 1012 (Moran et al., 1998) by several orders of magnitude. Raisbeck and co-workers have demonstrated the use of 129I/127I as a ‘‘pathway tracer’’ in the North Atlantic and Arctic Oceans (e.g. Raisbeck et al., 1995; Raisbeck and Yiou, 1999). The large difference in 129I/99Tc ratios released to the sea from La Hague and Sellafield can also be used

0883-2927/$ - see front matter  2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.apgeochem.2006.12.007

620

C. Schnabel et al. / Applied Geochemistry 22 (2007) 619–627

to determine relative contributions of sea water currents passing both reprocessing sites at the sampling site (Yiou et al., 2002). Liquid 129I emissions from both reprocessing plants show strong increases during the 1990s which have been tracked by Raisbeck and Yiou (1999) through seaweed analysis at two locations on the Norwegian coast. The 129I/127I ratios at Utsira (southern Norway) increased by a factor of 9 from 1980 to 1994. Raisbeck and co-workers also analysed 129I/127I ratios in sea water or seaweed samples at many locations in European seas (Yiou et al., 1994; Raisbeck et al., 1995) covering a time period from 1985 to 1993. More recent 129I/127I ratios close to the dominating source La Hague have been studied by Fre´chou et al. (2001) in the English Channel and by Szidat et al. (2000) and Ernst (2004) on the German and Danish North Sea coast. The I isotope ratios were about 1 · 105 in the English Channel and about 1.5 · 106 for the North Sea coast samples. For regions relatively close to the Sellafield emission source the only published data are for two locations in the Irish Sea sampled in 1992 and one location in NE Scotland also from 1992 (Yiou et al., 1994). However, liquid emissions from Sellafield have increased by a factor of 12 from the 1992 minimum to the maximum in 2002 or by a factor of nearly 5 from 1991 to 2002. This increase should be reflected in increased temporary 129I/127I ratios in the Irish Sea and along the Scottish coast compared to 1992. Consequently, high 129I/127I ratios can be expected now and new analyses of sea water samples have been carried out by Atarashi-Andoh et al. (2007) for the Irish Sea in 2004 and 2005 and in the present study for the Scottish coastlines from 2003 to 2005. The new data also constitute input parameters to study the sea to land transfer of 129I and for dose calculations considering this pathway. The relative importance of this type of transfer of 129I has increased drastically since the early 1990s because the gaseous emissions from both reprocessing plants remained either constant or even decreased, in strong contrast to the aforementioned strong increase in liquid discharges. It has to be mentioned that contemporary high 129 127 I/ I ratios are not considered a health risk, even if they were present in human thyroids. An I isotope ratio of 1.2 · 103 would have to be reached in a human thyroid to create a dose of 900 lSv a1, which is the maximum acceptable single organ dose according to German radiation protection legislation (Schmidt, 1998).

2. Experimental 2.1. Sampling Sea water samples of between 250 mL and 1 L were collected from coastal surface waters with the smaller volumes taken closer to the emission source. The samples were preserved by addition of 200 mg NaOH and stored in the dark until processing to avoid losses through elemental I2 formation during storage. For the fresh water samples 1 L samples were collected and preserved in the same way. Fig. 1 depicts the sampling locations of this study as well as two sampling locations in SW Scotland from the work of Atarashi-Andoh et al. (2007) and the locality where Yiou et al. (1994) analysed a 1992 sample at Lossiemouth, NE Scotland. The locations include all Scottish coastlines as well as the Western Isles and the Orkney Islands. The alkaline samples were filtered using a 0.45 lm membrane filter and split into a small subsample for 127I analysis (about 30 mL) and the remaining main part of the sample, which was analysed for 129I. 2.2.

127

127

I analysis

I concentrations were determined using a VG Elemental PlasmaQuad 2Plus Quadrupole mass spectrometer at SUERC. The aliquot taken for 127 I analysis was diluted to about 1–2 lg/kg I. A reducing alkaline matrix was used: 0.08% tetramethylammonium hydroxide (TMAH), 0.01 M NaHSO3, 0.02 M NH2OH Æ HCl. Cs (10 ng/mL) was chosen as internal standard to correct for sample matrix suppression effect and time dependent changes in sensitivity. A standard sample was measured before and after every unknown sample, which were each measured at least 3 times. Washing between samples was done with 3% HNO3 followed by cleaning with the matrix solution. The count rate of the matrix solution was used for the blank correction of the unknowns. Typical standard uncertainties for the I concentration ranged from 3% to 5%. The measurement procedure was validated by inter-comparison measurements with ETH Zurich. This aimed to prove identical sensitivity in ICPMS measurements for different chemical compounds of I, and to check standard material used at SUERC by comparing to a standard prepared at ETH (see Table 1). This check of standards is

C. Schnabel et al. / Applied Geochemistry 22 (2007) 619–627

Sampling sites

621

12 13 14

5˚W

3˚W

58˚N

10 9 19

7 6

11 ABERDEEN

57˚N 57˚N

8

5

DUNDEE E

56˚N

56˚N

GLASGOW

4

3 Ayrshire

16

2 55˚N

Sellafield

56˚N

North Channel 0 km

50 km 5˚W

Dumfries & Galloway

15 1

18 17

Fig. 1. Sampling locations of this work (locations 1–14 sea water and 15–16 fresh water), Atarashi-Andoh et al., 2007 (locations 17–18) and Yiou et al., 1994 (location 19). The inset shows the location of Sellafield and Parton, Heysham and Millom are located very close to Sellafield on the inset map. The sampling locations shown are: 1: Garlieston; 2: Girvan; 3: Troon; 4: Sannox Bay (Arran); 5: Sanna Bay (Ardnamurchan); 6: Vatersay; 7: Pollachar (South Uist); 8: Barra; 9: Gruinard Bay; 10: Dornoch; 11: Aberdour Bay; 12: Orkney Mainland NW; 13: Orkney Mainland NE; 14: South Ronaldsay; 15: Clatteringshaws Loch; 16: Rowantree Burn; 17: Brighouse Bay (AtarashiAndoh et al., 2007); 18: Loch Ken (Atarashi-Andoh et al., 2007); 19: Lossiemouth (Yiou et al., 1994).

regarded as important due to the lack of standard reference materials for I analysis in aqueous matrix. In Zurich, I, IO 3 and C2H4IOH (as an example of organic I compounds) solutions prepared from their own chemicals were compared with IO 3 prepared at SUERC. The measurements were performed at EAWAG on a Perkin–Elmer Elan 5000 with a micro concentric glass nebuliser. The procedure was slightly different to the one used at SUERC:TMAH (0.08%; no reducing agent added) was chosen as the matrix and 100 ng/mL Br was used as internal standard (instead of Cs). The sys-

tem was rinsed with 0.08% TMAH between samples. At SUERC only I solution prepared at ETH and I solution prepared at SUERC were measured against IO 3 solution prepared at SUERC. One might presume identical sensitivities for solutions prepared from different chemical I compounds in a reducing matrix because equilibration to I should be complete. However, it was found at SUERC that the likelihood of memory effects is smaller for a solution prepared from IO 3 in that reducing matrix than for one prepared from I. This is reflected in

622

C. Schnabel et al. / Applied Geochemistry 22 (2007) 619–627

Table 1 Quality control of 127I ICP-MS measurements: sensitivities (count rate per ng/mL I) for different chemical I compounds prepared at ETH or SUERC normalized to the sensitivity obtained for a solution prepared from IO3 at SUERC Substance (prepared at)

Measurement location

Matrix

Sensitivity ratio

Iodide (SUERC) Iodide (ETH) Iodide (SUERC) Iodide (ETH) Iodate (ETH) C2H4IOH (ETH)

SUERC SUERC ETH ETH ETH ETH

Reducing TMAH Reducing TMAH 0.08% TMAH 0.08% TMAH 0.08% TMAH 0.08% TMAH

1.00 ± 0.08 1.04 ± 0.05 0.94 ± 0.05 0.97 ± 0.05 0.94 ± 0.05 0.99 ± 0.05

TMAH stands for tetramethylammmonium hydroxide.

shorter washing times for the samples prepared from IO 3 . Table 1 shows that within measurement uncertainties, all chemical compounds were measured with the same sensitivity. 2.3. Analysis of

129

I

Between 7 and 15 mg of Woodward I as I were added to the remaining solution after taking the aliquot for 127I analysis. The carrier masses were chosen to aim for 129I/127I ratios of 1 · 109 for the AMS measurements at ETH and the early measurements at SUERC (Freeman et al., 2004), whereas the carrier mass was chosen to result in I isotope ratios of only 3 · 1010 for the later measurements at SUERC (samples taken in 2005). To equilibrate all I species as IO 3 , 5 mL 2.5% NaClO3 solution were added to the solution and shaken for 15 min. The combination of NaOH (present in the solution before the chlorate addition) and chlorate(V) should also ensure the conversion of  organic I compounds to IO 3 . Reduction to I was carried out with sodium sulfite and hydroxylammonium hydrochloride. The pH was adjusted to 5–6 by addition of HNO3. Addition of sodium nitrite and more HNO3 resulted in the oxidation to I2, which was extracted into chloroform in a separatory funnel. Reduction and back-extraction into the aqueous phase was carried out with sodium sulfite solution. More NaOH (200 mg) was added to the aqueous phase to avoid evaporative losses of I during volume reduction to about 30 mL. Saturated Ba(NO3)2 (99.999%, Alfa Aesar) solution was added to remove S and associated Te. High purity Ba(NO3)2 is necessary to minimise count rates in the detector of the AMS measurement caused by Sr2+. Addition of AgNO3 resulted in the precipitation of AgI, which was centrifuged and washed with diluted ammonia solution and water. After drying the AgI at 70C it was mixed with high-purity Ag (99.95%, Alfa Aesar) using a ratio 1:2 wt/wt AgI:Ag.

The samples collected in 2003 were measured for I with the 0.6 MV compact tandem accelerator at ETH Zurich. 490 kV was chosen as terminal voltage and the charge state I4+ was chosen for detection. The interference of 97Mo3+ (e.g. injected as MoO 2 Þ was separated in the detector. This measurement was carried out using a dilution of the NIST 4949B standard reference material (D2New) as standard. The samples taken in 2004 and 2005 were measured for 129I with the 5 MV NEC Pelletron accelerator at SUERC. The terminal voltage was set at 3 MV and I3+ was selected. 129I/127I ratios of typical samples were 3 · 1010 to 1 · 109. Blank values were determined by memory effects in the ion source and were always <1 · 1012 (first samples) and generally between 1 · 1013 and 3 · 1013. Details of the measurement procedure are given in Maden et al. (accepted for publication). Measurements at SUERC were carried out relative to the Z94-0595 standard from PRIME laboratory. The D2New standard used in Zurich was found to agree within 5% with the PRIME laboratory standard through measurements at SUERC.

129

3. Results and discussion Table 2 summarises 127I and 129I concentrations as well as 129I/127I ratios determined for the Scottish water samples in this work; the results of AtarashiAndoh et al. (2007) for 2 water samples from SW Scotland; 3 Irish Sea samples taken in 2004 and 2005; and the only published analysis for Scotland before these recent studies – one sample from NE Scotland collected in 1992 (Yiou et al., 1994). 3.1.

127

I concentrations

Concentrations of stable I determined in sea water samples in this study ranged from 21 to 52 lg/kg. Raisbeck et al. (1995) found a concentra-

C. Schnabel et al. / Applied Geochemistry 22 (2007) 619–627

623

Table 2 127 I and 129I concentrations as well as 129I/127I atom ratios of the water samples of this study and of samples for comparison (AtarashiAndoh et al., 2007 for SW Scotland as well as the Irish Sea samples Parton, Heysham and Millom, and Yiou et al., 1994 for NE Scotland) Location (Sample code in Fig. 1)

Latitude Longitude

Sampling date

127

129

129

Garlieston (1) Garlieston (1) Girvan (2) Girvan (2) Troon (3) Troon (3) Sannox Bay (4) Sannox Bay (4) Sanna Bay (5) Vatersay (6) Pollachar (7) Barra (8) Gruinard Bay (9) Gruinard Bay (9) Gruinard Bay (9) Dornoch (10) Dornoch (10) Dornoch (10) Aberdour Bay (11) Orkney Mainland NW (12) Orkney Mainland NE (13) South Ronaldsay (14) Clatteringshaws Loch (15) Rowantree Burn (16) Brighouse Bay (17) (Atarashi-Andoh et al., 2007) Loch Ken (18) (Atarashi-Andoh et al., 2007) Lossiemouth (19) (Yiou et al., 1994)

5446 0 N 5446 0 N 5514 0 N 5514 0 N 5532 0 N 5532 0 N 5540 0 N 5540 0 N 5645 0 N 5655 0 N 5706 0 N 5656 0 N 5752 0 N 5752 0 N 5752 0 N 5753 0 N 5753 0 N 5753 0 N 5740 0 N 5908 0 N 5855 0 N 5844 0 N 5504 0 N 5511 0 N 5444 0 N

0422 0 W 0422 0 W 0452 0 W 0452 0 W 0440 0 W 0440 0 W 0510 0 W 0510 0 W 0611 0 W 0731 0 W 0724 0 W 0722 0 W 0528 0 W 0528 0 W 0528 0 W 0401 0 W 0401 0 W 0401 0 W 0212 0 W 0318 0 W 0247 0 W 0258 0 W 0416 0 W 0434 0 W 0408 0 W

June 2004 January 2005 June 2004 January 2005 August 2003 January 2005 April 2003 November 2004 October 2004 June 2003 June 2003 May2005 July 2003 July 2004 July 2005 July 2003 July 2004 July 2005 January 2003 July 2005 July 2005 July 2005 January 2005 June 2004 December 2004

n.d. 29.3 ± 1.2 32.9 ± 1.5 36.9 ± 1.5 50.3 ± 2.7 34.9 ± 1.7 47.2 ± 2.4 35.3 ± 1.6 43.9 ± 1.8 34.5 ± 1.4 52.1 ± 2.6 38.5 ± 1.5 49.6 ± 2.0 21.5 ± 0.9 35.9 ± 1.4 49.2 ± 5.4 42.3 ± 1.3 41.1 ± 1.6 34.8 ± 1.0 39.9 ± 1.6 32.2 ± 1.3 29.9 ± 1.2 1.72 ± 0.07 n.d. 25.1 ± 1.7

45.8 ± 1.4 46.7 ± 1.4 11.1 ± 0.3 16.9 ± 0.5 11.7 ± 0.3 18.9 ± 0.6 7.02 ± 0.27 13.9 ± 0.4 6.23 ± 0.19 1.40 ± 0.13 2.19 ± 0.16 2.28 ± 0.06 3.01 ± 0.15 0.83 ± 0.02 2.36 ± 0.05 2.68 ± 0.16 1.63 ± 0.05 1.39 ± 0.07 1.49 ± 0.04 1.84 ± 0.04 1.12 ± 0.04 1.02 ± 0.02 0.382 ± 0.012 1.02 ± 0.03 33.1 ± 2.4

n.d. 33,600 ± 1700 7130 ± 390 9670 ± 480 4900 ± 290 11,400 ± 650 3130 ± 200 8330 ± 450 2990 ± 150 855 ± 85 886 ± 78 1250 ± 60 1280 ± 80 808 ± 40 1380 ± 60 1150 ± 140 814 ± 41 713 ± 40 901 ± 45 974 ± 44 734 ± 42 717 ± 32 4680 ± 120 n.d. 27,800 ± 700

5458 0 N 0403 0 W

November 2004

4.95 ± 0.19

0.58 ± 0.03

2480 ± 700

ca. 5743 0 N 0317 0 W 5434 N 0335 0 W 5403 0 N 0254 0 W 5418 0 N 0325 0 W

April 1992

57.8 ± 2.9

March 2004 December 2004 June 2005

33.5 ± 1.9 26.5 ± 2.0 32.7 ± 1.3

Parton (Atarashi-Andoh et al., 2007) Heysham (Atarashi-Andoh et al., 2007) Millom (Atarashi-Andoh et al., 2007)

I (lg/kg)

I (1010/kg)

0.4 (calculated) 128 ± 26 103 ± 8 87.5 ± 1.8

I/127I (1010)

160 ± 20 80,800 ± 15,600 82,000 ± 1900 56,400 ± 2500

n.d. = not determined.

tion range from 36 to 120 lg/kg at 15 sampling locations. Other comparative data are the results from Michel and co-workers (Ernst, 2004) of 28– 102 lg/kg for 8 analysed samples. The review article by Wong (1991) gives 60 lg/kg as typical I concentration in sea water. Truesdale (1994) analysed total I concentrations in the Irish Sea in September 1977. For the south coast of Dumfries and Galloway (location Nos. 1 and 17 on Fig. 1) they found 42 lg/kg total dissolved I. For the north coast of Wales their total dissolved I concentrations were even lower, 38 lg/kg. This comparison shows that the 127I sea water concentrations determined in this work are in the range of published concentrations, but are on average lower than those data. The lack of concentra-

tions above 60 lg/kg does not seem to be of concern because the second highest concentration determined in Raisbeck’s work was also lower than 60 lg/kg. More importantly it seems there is an apparent time dependence of the 127I concentrations determined in this work. The 2003 samples were on average higher than the later samples. The authors have checked the standard solution used for the later samples with newly prepared solutions and have confirmed the analyses. The most likely interpretation of the time dependence is a slight change in the sample preparation procedure. It seems possible that the hydroxide precipitate that occurred after addition of NaOH was washed more thoroughly with water for the 2003 samples than for the later ones. However, any losses of I at this stage would

624

C. Schnabel et al. / Applied Geochemistry 22 (2007) 619–627

occur with the isotope ratio of the original sample because the carrier had not yet been added. Consequently both the 127I and 129I concentrations determined in this work can be considered as lower limits, whereas the 129I/127I ratios determined for the water samples should be correct within the uncertainties given. This is of great importance because this ratio is easier to interpret than 129I concentrations as the isotope ratio is insensitive to dilution with fresh water. 3.2.

129

I concentrations

In general, 129I concentrations determined in this work decreased along the Scottish coastline with increasing distance from the emission source, but were also lower for Island samples further away from the coastline of the Scottish mainland. The radioiodine concentrations found in Dumfries and Galloway (Garlieston No. 1 and Brighouse Bay No. 17) were a factor of 2.5–3 lower than those determined close to the source (Parton and Heysham in Cumbria, very close to Sellafield in the inset in Fig. 1) by Atarashi-Andoh et al. (2007). 129I concentrations decrease by a factor of 4 when the water currents reach the Ayrshire coastline (Girvan No. 2 and Troon No. 3). This strong decrease could be attributed to dilution of the radionuclide concentration by passing through deeper zones in the North Channel and to dilution of the radionuclide concentration with water masses from the Atlantic Ocean which enter through the North Channel. As explained above, I isotope ratios are more appropriate to compare than 129I concentrations. For this reason, the 129 I results are not discussed in detail here. However, attention is drawn to two Hebridean samples taken in 2003, Pollachar (South Uist, No. 7) and Vatersay (south of Barra, No. 6). A factor of 1.5 is noted between the 129I concentrations of these samples which one would expect to be affected in a very similar way by the emissions from Sellafield. This will be revisited when the iodine isotope ratios are discussed. The 129I concentrations of the Orkney Island samples (locations 12–14 on Fig. 1) can be compared to a surface sample taken between Orkney and the Shetland Islands (5924 0 N 154 0 W) in June 1999 by Alfimov et al. (2004). This surface sample (Station No. 10 in Alfimov et al., 2004) was part of a transect analysed for 129I from the North Atlantic to the Baltic Sea. The present results of (10– 18) · 1010 at 129I/kg are much higher than the concentration of 2.6 · 109 at 129I/L for the transect

sample. This discrepancy is not expected based on 238 Pu and 239+240Pu concentrations determined by Murray et al. (1978) close to the Orkney Islands and close to Shetland Islands, who found concentrations close to the Orkney Islands less than a factor of 5 higher than those close to the Shetland Islands. 3.3.

129

I/127I ratios

The I isotope ratios are first examined in a rather descriptive way by comparing the values determined around the Scottish coastline to one another and to data from the Irish Sea only for samples taken from 2003 to 2005 without discussion of possible developments with time. In a second step the data obtained in this work are checked for temporal changes of the I isotope ratio from 2003 to 2005. Then the development of I isotope ratios from the early 1990s to 2003 are estimated by comparing the Aberdour Bay sample (No. 11) to the Lossiemouth sample (No. 19) analysed by Yiou et al. (1994) in 1992. Finally, I isotope ratios in Scottish sea water are compared to other regions of Europe. The 129I/127I ratio determined at Garlieston (No. 1) of about 3 · 106 agrees well with the other result for the Dumfries and Galloway coastline determined by Atarashi-Andoh et al. (2007) at Brighouse Bay (No. 17). Both ratios are less than a factor of 3 lower than those observed in the Irish Sea by Atarashi-Andoh et al. (2007), Parton, Heysham and Millom (in Cumbria close to Sellafield). Moving from Dumfries and Galloway to Ayrshire (location Nos. 2 and 3) I isotope ratios decrease by about a factor of 3 to about 1 · 106. Iodine isotope ratios decrease further to about 1 · 107 for the mainland coastline of NW and NE Scotland. The quite good agreement between the ratios observed at Gruinard Bay (No. 9) and Dornoch (No. 10) is surprising because further dilution of the radionuclide concentration should be expected for water currents moving around the north coast of Scotland. However, 99 Tc concentrations analysed in these regions (Leonard et al., 1997) showed a dilution of 99Tc concentrations by less than an order of magnitude. Due to the limited number of samples taken at Gruinard Bay and Dornoch the authors do not draw the conclusion that additional input of 129I from La Hague is necessary. Iodine isotope ratios determined for Hebridean (location Nos. 6–8) and Orkney Island samples (locations 12–14) are about 1 · 107. Excellent agreement is noted the between the I isotope ratios

C. Schnabel et al. / Applied Geochemistry 22 (2007) 619–627

determined at Pollachar (No. 7) and Vatersay (No. 6) in 2003, which in combination with the 129I and 127 I concentrations analysed for these samples demonstrates that the isotope ratio is easier to interpret than the radioiodine concentration because it is insensitive to dilution with fresh water. Annual liquid 129I emission data from Sellafield until 2000 have been published by Lo´pez-Gutie´rrez et al. (2004). Atarashi-Andoh et al. (2007) updated these data until 2004. Based on these annual liquid emission data one would not expect a strong time dependence for I isotope ratios in Scottish sea water samples from 2003 to early 2005. The decrease in isotope ratios from 2003 to 2004 at Gruinard Bay and Dornoch reflects a decrease in liquid emissions from 2002 to 2003, which is consistent with the expected transit time of the water masses from the source region of at least one year for these locations. This expectation is based on transit times from Sellafield to the North Channel of 6–9 months determined by Leonard et al. (1997) using 99Tc. To interpret the different tendencies at both stations in 2005 one would possibly need emission data at better resolution (e.g. 3 months resolution). Nevertheless, no strong trend is seen at either location in contrast to the strong increase in I isotope ratios observed by Raisbeck et al. (1995) for the Norwegian coastline in the early 1990s. 120

129

Transit times from the emission source to Girvan (No. 2) and Troon (No. 3) in SW Scotland are expected to be slightly less than a year (again based on Leonard et al. (1997)). Consequently, the increase in emission from 2003 to 2004 is reflected in an increase in I isotope ratios at these locations over the same period. Fig. 2 depicts changes in 129I emissions from Sellafield from 1980 to 2004 and I isotope ratios analysed in NE Scotland at Lossiemouth in 1992 (Yiou et al., 1994) for comparison with the nearby Aberdour Bay locality in 2003 and the locality of Dornoch in 2003–2005. The graph shows that an increase in I isotope ratio by about a factor of 6 over this time period is associated with the increase in liquid 129I emissions. This agreement becomes even stronger when transit times of about two years are taken into account. The highest I isotope ratios found in Scottish sea water of about 3 · 106 are higher by a factor of 2 than ratios at the German and Danish North Sea coast (Szidat et al., 2000; Ernst, 2004). However, they are more than a factor of 3 lower than isotope ratios reported for the English Channel (Fre´chou et al., 2001). The I isotope ratios for the lake samples in Dumfries and Galloway, Clatteringshaws Loch (No. 15) and Loch Ken (No. 18; Atarashi-Andoh et al., 2007) were more than a factor of 6 lower than the 14

Liquid 129I emissions Gaseous 129I emissions Lossiemouth (Yiou et al., 1994) Aberdour Bay Dornoch

12

10 -8

I (10 )

80

6

I/

8 60

129 127

I emissions (kg)

100

625

40 4 20

2

0 1980

1985

1990

1995

2000

0 2005

Year 129

Fig. 2. Annual liquid and gaseous I emissions from Sellafield (Atarashi-Andoh et al., 2007) and 129I/127I ratios measured at locations in NE Scotland in 1992 (Yiou et al., 1994) and 2003–2005 (this work).

626

C. Schnabel et al. / Applied Geochemistry 22 (2007) 619–627

129

I/127I ratio determined for the south coast of Dumfries and Galloway (Garlieston No. 1 and Brighouse Bay No. 17). Based on these two samples it cannot be decided whether the sea to land transfer is the dominant source for 129I in these lakes or whether there is a substantial contribution of the gaseous emissions from Sellafield. The fact that the Loch Ken sample (No. 18) has a factor of 3 higher 127I concentration than that of Clatteringshaws Loch (No. 15) but the lower I isotope ratio indicates a contribution from the gaseous emissions for these lake samples. 4. Summary

The first survey of 129I/127I ratios in Scottish sea water is presented. Based on a comparison of locations in NE Scotland (Aberdour Bay No. 11 in this work and Lossiemouth No. 19 from Yiou et al., 1994) these ratios increased by a factor of about 6 from 1992 to 2003, apparently in response to changes in liquid 129I emissions from Sellafield over this time period. Iodine isotope ratios in SW Scotland reached 3 · 106 in 2004, more than a factor of 4 higher than results from the Irish Sea for samples from 1992 (Yiou et al., 1994) and only a factor of 3 lower than Irish Sea samples from 2004 (Atarashi-Andoh et al., 2007). No strong trend is seen in I isotope ratios in Scottish sea water from 2003 to 2005, in accordance with the lack of drastic variations in liquid emissions from Sellafield. Samples from similar locations taken very closely in time agree better in their I isotope ratio than in their 129 I concentration, as expected because of the insensitivity of the ratio to dilution with fresh water. 129I concentrations obtained in this study can be used as input parameters for dose calculations based on sea to land transfer, which may be important especially for SW Scotland. Acknowledgements NERC is acknowledged for financial support (salary of C.S.). The authors are grateful to M. Caffee (PRIME Laboratory at Purdue University) for donating standard material for AMS measurements. S. Waldron (University of Glasgow) and R. McGill (SUERC) took several water samples for this study. J. Beaumont (SAMS, Oban) is thanked for taking a sample close to Barra in 2005. M. Migue´ns-Rodrı´guez (SUERC) took part in ICP-MS measurements of 127I. A. Donnachie

and E. McKay (both SUERC) participated in the chemical preparation of some samples. S. Bollhalder and D. Kistler (both EAWAG) took part in the ICP-MS measurements at EAWAG. G. Snyder (Rice University) and J. Moran (LLNL) gave advice on 127I measurements with ICP-MS. A.B. MacKenzie and G.T. Cook (both SUERC) are acknowledged for discussions. Finally, we thank 3 anonymous reviewers for their constructive reviews that improved this manuscript. References Alfimov, V., Aldahan, A., Possnert, G., Kekli, A., Meili, M., 2004. Concentrations of 129I along a transect from the North Atlantic to the Baltic Sea. Nucl. Instr. Meth. Phys. Res. B 223/224, 446–450. Atarashi-Andoh, M., Schnabel, C., Cook, G., MacKenzie, A.B., Dougans, A., Ellam, R.M., Freeman, S.P.H.T., Maden, C., Olive, V., Synal, H.-A., Xu, S., 2007. 129I/127I ratios in surface waters of the English Lake District. Appl. Geochem. 22, 628– 636. Ernst, T., 2004. Anthropogenes Iod-129 als Tracer fu¨r Umweltprozesse – Ein Beitrag zum Verhalten von Spurenstoffen bei der Migration in Bo¨den und beim atmospha¨rischen Transport. Ph.D. Thesis, Univ. Hannover. (). Fre´chou, C., Calmet, D., Bouisset, P., Piccot, D., Gaudry, A., Yiou, F., Raisbeck, G., 2001. 129I and 129I/127I ratio determination in environmental biological samples by RNAA, AMS and direct c-X spectrometry measurements. J. Radioanal. Nucl. Chem. 249, 133–138. Freeman, S., Xu, S., Schnabel, C., Dougans, A., Tait, A., Kitchen, R., Klody, G., Loger, R., Pollock, T., Schroeder, J., Sundquist, M., 2004. Initial measurements with the SUERC accelerator mass spectrometer. The SUERC AMS Laboratory after 3 years. Nucl. Instr. Meth. Phys. Res. B 223/224, 195–198. Leonard, K.S., McCubbin, D., Brown, J., Bonfield, R., Brooks, T., 1997. Distribution of technetium-99 in UK coastal waters. Mar. Pollut. Bull. 34, 628–636. Lo´pez-Gutie´rrez, J.M., Garcia-Leon, M., Schnabel, Ch., Suter, M., Synal, H.-A., Szidat, S., Garcia-Tenorio, R., 2004. Relative influences of 129I sources in a sediment core from Kattegat area. Sci. Total Environ. 323, 195–210. Maden, C., Anastasi, P.A.F., Dougans, A., Freeman, S.P.H.T., Kitchen, R., Klody, G., Schnabel, C., Sundquist, M., Vanner, K., Xu, S. SUERC AMS ion detection. Nucl. Instr. Meth. Phys. Res. B (Proc. AMS-10 Conf), accepted for publication. Moran, J.E., Fehn, U., Teng, R.T.D., 1998. Variations in 129I/127I ratios in recent marine sediments: evidence for a fossil organic component. Chem. Geol. 152, 193–203. Murray, C.N., Kautsky, H., Hoppenheit, M., Domian, M., 1978. Actinide activities in water entering the northern North Sea. Nature 276, 225–230. Raisbeck, G.M., Yiou, F., 1999. 129I in the oceans: origins and applications. Sci. Total Environ. 237/238, 31–41. Raisbeck, G.M., Yiou, F., Zhou, Z.Q., Kilius, L.R., 1995. 129I from nuclear fuel reprocessing facilities at Sellafield (U.K.)

C. Schnabel et al. / Applied Geochemistry 22 (2007) 619–627 and La Hague (France); potential as an oceanographic tracer. J. Marine Syst. 6, 561–570. Schmidt, A., 1998. 129I und stabiles Iod in Umweltproben. Ph.D. thesis, Univ. Hannover. (). Szidat, S., Schmidt, A., Handl, J., Jakob, D., Botsch, W., Michel, R., Synal, H.-A., Schnabel, C., Suter, M., Lo´pez-Gutie´rrez, J.M., Sta¨de, W., 2000. Iodine-129: sample preparation, quality control and analyses of pre-nuclear materials and of natural waters from Lower Saxony, Germany. Nucl. Instrum. Meth. Phys. Res. B 172, 699–710.

627

Truesdale, V.W., 1994. Distribution of dissolved iodine in the Irish Sea, a temperate shelf sea. Estuar. Coast. Shelf Sci. 38, 435–446. Wong, G.T.F., 1991. The marine geochemistry of iodine. Rev. Aquat. Sci. 4, 45–73. Yiou, F., Raisbeck, G.M., Zhou, Z.Q., Kilius, L.R., 1994. 129I from nuclear fuel reprocessing: potential as an oceanographic tracer. Nucl. Instr. Meth. Phys. Res. B 92, 436–439. Yiou, F., Raisbeck, G.M., Christensen, G.C., Holm, E., 2002. 129 127 I/ I, 129I/137Cs and 129I/99Tc in the Norwegian coastal current from 1980 to 1998. J. Environ. Radioact. 60, 61–71.