Anthropogenic 129I Traced in Environmental Archives by Accelerator Mass Spectrometry

Full text

(1)Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 448. Anthropogenic 129I Traced in Environmental Archives by Accelerator Mass Spectrometry EDVARD ENGLUND. ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2008. ISSN 1651-6214 ISBN 978-91-554-7238-2 urn:nbn:se:uu:diva-8989.

(2)

(3)

(4)

(5)

(6)

(7)

(8) !" #! $

(9) % %& ' ( ! ( ( )! !" #!

(10)

(11) *

(12)

(13) +

(14) !" +

(15)

(16) +" %&" ,

(17) !

(18) %-. #

(19) +

(20)

(21)

(22) ,! , / 0 " ,

(23)

(24) "

(25)

(26)

(27)

(28) &" 1 " " .023 -4&5-51154%6&5%" 0

(29) !

(30)

(31)

(32) ( !

(33)

(34)

(35) ! -

(36) ( !

(37)

(38) !

(39)

(40)

(41) " #! ! * ! !

(42)

(43) ( !

(44) !

(45) %-. 7#8%9 1"4 /:

(46)

(47) !

(48) + " ,

(49)

(50) (

(51) (

(52) (

(53) ( !

(54)

(55) (

(56) ! %-.

(57)

(58)

(59) ! 5

(60) ! " #! !

(61) ; * ( ! ( !.

(62) ( %-.

(63)

(64) ( * <

(65) 0*

(66)

(67)

(68) <

(69) =

(70)

(71) *

(72) ( *

(73) ! 2 0 " .

(74)

(75) %-.

(76)

(77)

(78)

(79) (

(80) !

(81)

(82) !

(83) 0*

(84)

(85) ! -&6 % !

(86) " #!

(87)

(88) !

(89) !

(90) %-. (

(91) ! -1 !

(92) <

(93) ( !

(94) (

(95) (

(96) 0 ( 7>:

(97) ? 7=

(98) :" , !

(99)

(100)

(101) * ! ! @!

(102)

(103)

(104) ( ( ! 0*! <

(105) ! !! @!

(106) ( " 3

(107) ( ! %-.

(108) ! A1B ( ! ( ! <

(109) ! ;

(110) ( !

(111) ( " #!

(112)

(113) !

(114)

(115) ( ! @!

(116)

(117) ! %-. (" #!

(118) (

(119) !

(120) +

(121) %-.

(122) ! *

(123) (

(124) !!

(125)

(126)

(127)

(128)

(129) !

(130) !

(131) 0*

(132) " + ( %-. (

(133) ! . C%1B

(134)

(135) ! ( " #!

(136) ( %-.

(137) !

(138) !

(139) !

(140) ( !

(141) * < ( !

(142)

(143)

(144)

(145)

(146)

(147)

(148) "

(149) 5%-

(150)

(151) @!

(152)

(153) *

(154) 0 ( ? ,/0 !

(155) ! " #! $ % &'(! ! )*&+,+ ! D + +

(156)

(157) %& .003 E15E% .023 -4&5-51154%6&5%

(158) '

(159)

(160) ''' 5&-&- 7! '88

(161) "<"8 F

(162) 9

(163) '

(164)

(165) ''' 5&-&-:.

(166) List of Papers. This thesis is based on the following papers: I. II. III. IV. V. E. Englund, A. Aldahan, G. Possnert, V. Alfimov (2007). A routine preparation method for AMS measurement of 129 I in solid material. Nucl. Instr. Meth. B 259: 365 - 369 A. Aldahan, E. Englund, G. Possnert, I. Cato, X. L. Hou (2007). Iodine129 enrichment in sediment of the Baltic Sea. Appl. Geochem. 22: 637647 E. Englund, A. Aldahan, G. Possnert (2008). Tracing anthropogenic nuclear activity with 129 I in lake sediments. J. Environ. Radioact. 99: 219 - 229 E. Englund, A. Aldahan, G. Possnert, X. Hou, I. Renberg, T. Saarinen. Modeling fallout of anthropogenic 129 I submitted to Environ. Sci. Technol. April 2008 E. Englund, A. Aldahan, X. Hou, G. Possnert. Time series of 129 I in aerosols, manuscript. Reprints were made with permission from the publishers. In paper I, my part of the work includes installation and purchase of the laboratory for combustion and chemical treatment, design of the combustion system and introduction of new routines for the extraction procedure, as well as writing of the paper. In paper II, part of the laboratory work as well as interpretation was done by me. In paper III, my part includes laboratory work, data processing and calculations, interpretation of the data and writing. In paper IV, I introduced the concept of signal processing to interpret the 129 I data in sediment archives, and did the laboratory work and writing. In paper V, I did laboratory work, data processing, interpretation and writing..

(167)

(168) Contents. 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Iodine in the environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Natural and anthropogenic sources of 129 I . . . . . . . . . . . . . . . . 2 Sampling and analytical procedures . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Sampling sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Lake and Baltic Sea sediments . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Extraction of iodine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.1 The Uppsala accelerator mass spectrometry system . . . . . 2.4.2 Measuring procedure of 129 I by AMS . . . . . . . . . . . . . . . . 3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Sample preparation method . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Chronology of the sediment profiles . . . . . . . . . . . . . . . . . . . . . 3.3 Iodine in sediment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Summary in Swedish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1 1 2 4 7 8 9 9 10 12 13 17 17 18 20 24 27 33 37 39 41 45.

(169)

(170) 1. Introduction. Since the beginning of the nuclear era, starting during the 1940s, large amounts of radioactivity have been released into natural environments. Beside concerns about hazards effects, radioactive isotopes associated with these emissions provide opportunities as anthropogenic tracers of environmental processes. Anthropogenic contributions of the long-lived isotope 129 I (T1/2 = 15.7 Myr) have been studied in fresh and marine water and sediment of north Europe [Rucklidge et al., 1994; Szidat et al., 2000; Buraglio et al., 2001ab; Alfimov et al., 2004; Gallagher et al., 2005; Michel et al., 2005; Aldahan et al., 2007a]. However, a continuous record covering the entire emission period is lacking. Sediments have in general provided high resolution and continuous annual archives for environmental analysis. The small amount of sediment that can be recovered on an annual scale for 129 I extraction has until recently been a limiting factor restricting the utilization of 129 I as a tracer. Accordingly, development of a standard method for extraction of 129 I from small amount of solid materials was needed in order to expose details of past anthropogenic emission of 129 I. This thesis deals with the tracing of 129 I emissions over Scandinavia and Finland since the early 1940s through the use of sediment archives. In addition, atmospheric transport pathways and fallout modes (wet and dry) of 129 I are elucidated through analysis of aerosols samples and modeling.. 1.1. Objectives. The objectives of this thesis include: • Development of a routine sample preparation procedure suitable for extraction of 129 I from small amount of solid material. • Tracing variability in 129 I in northern Europe since the beginning of the nuclear era by using a combination of lake and Baltic Sea sediments. • Providing new data about 129 I concentration in aerosols, that together with available data in precipitation are used to evaluate transport pathways and emission sources in northern Europe. • Depicting relative contributions from the different anthropogenic sources of 129 I by numerical modeling. Before presenting results and discussion related to the above mentioned objectives, a summary of the occurrence of iodine in the environment is given in the next sections.. 1.

(171) 1.2. Iodine in the environment. The ocean, atmosphere, terrestrial and biosphere form mobile iodine pools (Figure 1.1), which are included in a global box model for estimation of total inventories and major circulation patterns among the different compartments [Kocher, 1981] (Figure 1.2). A major part (> 99%) of the iodine is contained in the oceans as iodide (I− ) and iodate (IO− 3 ). Iodine is accumulated in marine algae and phytoplancton, which are linked to the iodine transfer from ocean to atmosphere [Fenical, 1981; Singh et al., 1983; Class and Ballschmiter, 1988]. It is hence concluded that iodine transfer mainly occurs at the biologically active areas of the oceans [Carpenter et al., 1999]. Phytoplancton occurs both at open seas and particularly at the coasts, whereas macro algae are strictly limited to coast areas, which covers only ∼0.5% of the oceans.. Inorganic species, I2, HI, HOI,... Residence time in Atmosphere ~2 weeks. Aerosols I Photodissociation (~1 Week):. Dry. Wet. Organic iodine: CH3I, CH2I2, CH2ClI, ... Sea spray I–, IO3–. In soils:. Organically mediated Particulate. IO3-. Ocean. Washout I–, IO3– , (particles). I-. Organic Metal oxides. Dissolved salts I–, IO3–. Land. Figure 1.1: Iodine in the upper ocean, atmosphere and terrestrial compartment, modified after Aldahan et al. [2006]. Among marine areas with highly elevated concentrations of iodine, considerable amounts of alkyl iodides have been identified in the atmosphere. These species, mainly methyl iodide, CH3 I but also for example CH2 I2 , C3 H7 I and CH2 ClI undergo photodissociation at a time scale of a few days and produce elemental iodine, I, which in turn rapidly reacts with the surroundings to form a variety of inorganic species [Vogt et al., 1999; Carpenter et al., 1999]. The behavior of iodine in the atmosphere is complex, and more than 200 reactions, particularly transitions between the different phases are considered 2.

(172) [Vogt et al., 1999]. All of the listed iodine species are chemically unstable with the exception of particular bounded iodate. Although gaseous iodine is dominant in the atmosphere, the fraction of particulate and gaseous iodine varies with geographical locations [Rahn et al., 1976; Kitto et al., 1988; Wershofen and Aumann, 1989; Gäbler and Heumann, 1993]. Special interest has been devoted to iodine in the atmosphere, for its role in the destruction of ozone [Chameides and Davis, 1980; Barrie et al., 1988; Solomon et al., 1994] and cloud formation [Saiz-Lopez et al., 2006]. Removal of iodine from the atmosphere occurs through dry and/or wet deposition. The considered pathways are (1) fallout through precipitation (wet deposition), either below cloud level (i.e. washout) or by incorporation into water droplets at cloud level, or (2) direct deposition of aerosol particles (dry deposition) [Whitehead, 1984; Baker et al., 2001; Buseck and Schwartz, 2003]. The residence time of iodine in the atmosphere is estimated to be a few weeks [Rahn et al., 1976; Kocher, 1981]. 1.2 x 1011 g/yr. Ocean Atmosphere. Land atmosphere. 2.0 x 1010. 8.3 x 1010 g. 5.7 x 10 9. 5.7 x 10 9 g/yr. Terrestrial biosphere. g. g/yr. 1.9 x 1012. 2.0 x 1012. g/yr. 1.0 x 1011. 5.7 x 1010. g/yr. g/yr. g/yr. Ocean mixed layer. 5.7 x 1010. Surface soil region. g/yr. 7.2 x 1013. g/yr. g/yr. Deep ocean 8.1 x 1016 g. 1.8 x 1011 g/yr. 1.8 x 1011 g/yr. 5.7 x 1010 g/yr. 4.2 x 1014 g. 1.4 x 1015 g 7.2 x 1013. 3.0 x 1011 g. 1.5 x 1010 g/yr. 1.5x 1010 g/yr. Shallow subsurface region 1.4 x 1013 g. 2.9 x 10 8 g/yr. 2.9 x 10 8 g/yr. Recent ocean sediments. Deep subsurface region. 8.9 x 1017 g. 1.1 x 1013 g. Figure 1.2: The compartment model of global iodine cycle [Kocher, 1981]. The major pool of iodine on land is soils with concentrations varying between 0.5 and 40 mg kg−1 . Within the soils the organic fraction is the main carrier of iodine [Sutcliffe and Davidson, 1979; Whitehead, 1979; Wenlock et al., 1982]. Soil iodine has a tendency to volatilize under certain conditions, e.g. if there is a low pH and low content of organic material or otherwise low 3.

(173) biological activity [Whitehead, 1984]. Concentrations of iodine in the soil are furthermore influenced by proximity to the sea and weathered parent material. The iodine concentration of soils may also show a decreasing trend with depth suggesting removal through leaching [Ernst et al., 2003].. 1.3 129 I. Natural and anthropogenic sources of 129I. is produced spontaneously in nature by spallation of xenon in the upper atmosphere under influence of cosmic rays, and as a fission product of 238 U in the earth’s crust [Fabryka-Martin et al., 1985]. Before the nuclear era, which started with the Manhattan project in the 1940s, the natural ratio of 129 I/127 I was in spatial and temporal equilibrium at ∼ 10−12 , corresponding to a total amount of about 100 kg of 129 I (Table 1.1). Presently, the reprocessing plants in Sellafield (UK) and La Hague (France) are by far the most significant global sources of 129 I. The total amount of liquid released 129 I is estimated to ∼5000 kg (Irish Sea and English Channel), and 230 kg as gaseous releases (to the atmosphere) by the year 2006 [Aldahan et al., 2007a]. Anthropogenic 129 I has during the last decades increased the natural 129 I/127 I ratio to as much as 8 orders of magnitude in the vicinity of the reprocessing plants [Rucklidge et al., 1994]. The nuclear weapon tests are believed to have contributed with 50 - 150 kg 129 I, during the 1950s and 1960s. In addition to the above mentioned globally relevant sources of 129 I, the Chernobyl accident in 1986 affected parts of Europe. The amount of 129 I released was relatively small, about 6 kg, and the fallout was well defined in both time and space. For details and comprehensive description of the distribution of 129 I in the environment the reader is referred to the following references [Chamberlain, 1991; Raisbeck and Yiou, 1999; Aldahan et al., 2007a].. 4.

(174) Release form. Referenceb. ∼100 50 - 150 2-6. Gaseous Gaseous. 1 1, 2, 3 4, a. 260 68 157 1420 70 3320 Published data not found. Gaseous Gaseous Gaseous Liquid Gaseous Liquid -. 5 6 7, 8, 9 7, 8 7, 10 7, 10. Source. 129 I. Natural Nuclear weapons testing 1945 - 1980 Chernobyl accident, 1986 Nuclear reprocessing facilities: Hanford, 1944 - 1972, USA Marcoule, 1988 - 1997, France Sellafield, 1952 - 2006, UK La Hague, 1966 - 2006, France Others (Russia, Japan, India, China). (kg). a Estimated by data of 129 I/131 I from Mironov et al. [2002] and the released 131 I from UNSCEAR [2000] b References: 1. Raisbeck and Yiou [1999], 2. Eisenbud and Gesell [1997], 3. Wagner et al. [1996], 4. Paul et al. [1987], 5. Hanford [1997], 6. Cogema [1997], 7. Lopez-Gutierrez et al. [2004], 8. RIFE [2001 - 2006], 9. Reithmeier et al. [2006], 10. Cogema [2001 - 2006]. Table 1.1: Amount of 129 I and release form from nuclear sources, modified after Aldahan et al. [2007a].. 5.

(175)

(176) 2. Sampling and analytical procedures. Materials used in this study include sediment cores taken from three lakes and the Baltic Sea, as well as aerosol samples (Figure 2.1). Site description, sampling procedures and analytical techniques are presented in this section.. F B Finland. A C Sweden D E G Sellafield. Chernobyl La Hague. Figure 2.1: The map shows sampling sites, major 129 I sources and general route of 129 I in the North Sea and Baltic Sea. Triangular dots mark lake or marine sediment sites, circular dots mark aerosol sampling stations. Individual sampling sites are A Lake Nylandssjön, B - Lake Lehmilampi, C - Lake Loppesjön, D - Baltic Sea, core 1, E - Baltic Sea, core 2, F - Kiruna, G - Ljungbyhed.. 7.

(177) 2.1. Sampling sites. Sediment cores were collected from two lakes in Sweden and one in Finland, details of each site are given in Table 2.1. These sites were selected for their undisturbed sediment sections, relatively high sedimentation rates and defined chronology. Two of the sites, Lake Nylandssjön and Lake Lehmilampi have varved sediment sequences (Figure 2.2). The varve structure typically consists of an organic rich part deposited during the biologically active seasons (spring, summer and autumn) associated with a higher mineralogical content during spring and autumn floods. At winter, fine grained material is dominating showing dark color and different texture.. Figure 2.2: Varved sediment from Lake Nylandssjön. Lake Nylandssjön Coordinates Annual precipitation Lake area Drainage area Max depth Altitude a.s.l.. ◦. ◦. 62 57’ N, 18 17’ E 800 mm 0.28 km2 0.95 km2 17.5 m 34 m. Lake Lehmilampi ◦. ◦. 63 37’ N, 29 6’ E 700 mm 0.15 km2 ∼1 km2 11.6 m 150 m. Lake Loppesjön 61◦ 42’ N, 16◦ 48’ E 500 mm 0.28 km2 4.1 km2 14 m 97 m. Table 2.1: Site information of the studied lakes. Sediments cores were collected from two locations in the Baltic Sea at coordinates (59◦ 4’N, 19◦ 7’ E) and (59◦ 13’ N, 18◦ 40’ E). The Baltic Sea receives its marine input through the Skagerrak and Kategatt basins that modulate water exchange between the North and the Baltic Seas (Figure 2.1). Aerosol filters from two stations, one from northern Sweden (Kiruna, 67◦ 50’ N, 20◦ 20’ E) and the other from southern Sweden (Ljungbyhed , 56◦ 5’ N, 13◦ 14’ E) were provided by the Swedish defense research institute (FOI). The station in northern Sweden is located at an altitude of 408 m a.s.l. and is characterized by mountainous topography and sparse vegetation. The station in southern Sweden (altitude 43 m) is located in an agricultural and forested 8.

(178) area. There is a large difference in the climate between the two stations where arctic conditions dominate at the northern station while temperate conditions are found in southern Sweden.. 2.2. Aerosols. Collection of air filters has been conducted since 1957 as part of the Swedish surveillance program for radioactivity in airborne particulate matters. The filters, made of glass fiber with a total area of 0.58 m×0.58 m, were exposed at ground level to a daily air flow of ∼24000 m3 (flow rate = 84 cm s−1 ) and were changed once per week [Vintersved and De Geer, 1982]. A 4 cm2 to 16 cm2 of the aerosol filters were used in this study at selective years between 1983 and 2000 with ∼4 samples for each year. Filters were stored in closed envelopes in ∼ 14 × 14 cm2 pieces at +4◦ C. Visual inspection of filters shows that those from southern Sweden were brown-black in color, while the filters from northern Sweden were white to light gray. Collection efficiency for the used filters indicate that > 99% of small particles ( ∼ 0.3 μ m) is captured at a flow rate between 10 and 300 cm s−1 [Suchny, 1968, the filter of this study is labeled FOA-1-484].. 2.3. Lake and Baltic Sea sediments. The sampling devices utilized for sediment coring were (1) gravity corer in Lake Loppesjön and the Baltic Sea, and (2) freeze corer in Lake Nylandssjön and Lake Lehmilampi. After slicing, the sediments were dried and measured for 137 Cs. Selective samples were set aside for isotopic and lithological measurements described in the following sections. Sediment cores have been collected continuously from Lake Nylandssjön since the 1970s for environmental and paleoclimatic purposes. Samples used in this study for 129 I and total iodine were cored in winter 2007 at a water depth of 17.5 m, with a core length of 20 cm and an area of ∼ 10 cm2 . Slicing of the core was performed in a freeze room after removing the outer parts with a hand plane, making the varves clearly visible. Individual varves were thereafter cut with a scalpel with a varve thickness typically 3 - 4 mm. In addition to varve counting, the chronology of the sediments has been verified by X-ray imaging analysis as well as direct control of annual sedimentation of cores taken in previous years. Sediment cores were collected from Lake Lehmilampi in winter 2006. The same sampling procedure applied for cores of Lake Nylandssjön was also used for the sediments from Lake Lehmilampi. About 50 - 100 mg of sediment was sampled, covering 1 - 5 years for each subsample, down to the year 1968. Sediment cores from Lake Loppesjön, each 30 cm long and ∼6 cm in diameter, were collected in winter 2004 from the deepest part of the lake (13.5 14 m). A total of six cores were collected. The sediments had silt-clay texture 9.

(179) with a dark grayish-brownish color without visible laminations. Slicing of the cores was done either in the field or in the laboratory with a resolution of 0.3 - 1 cm. The pore water was extracted by centrifugation of 3 cm thick intervals (i.e. 3 adjacent slices). The two sediment cores from the Baltic Sea , were 55 cm and 50 cm long and were collected at depths of 144 m and 78 m respectively, during autumn 1997 by the Geological Survey of Sweden. The sediment cores were sliced at interval of 1 cm and stored in a freezer immediately after the collection. The topmost sediment was comprised predominately of black, homogeneous post glacial clay while the deeper part exhibited gradual color change from gray to black. Additional analyzes were performed on the sediments, which include organic matter content in Lake Loppesjön, Lake Nylandssjön and the Baltic Sea and grain size analysis, 210 Pb and 14 C in Lake Loppesjön.. 2.4. Extraction of iodine. Summary of the procedure used for extraction of iodine from small solid material is shown in Figure 2.3 [Englund et al., 2007]. A mixture of a known amount of iodine carrier (∼2 mg) and sample material was combusted at 800◦ C in a stream of oxygen (in case of total iodine measurements the carrier was omitted). The liberated iodine was collected in a trapping solution and precipitated as silver iodide, which was mixed with niobium (Figure 2.4) for the accelerator mass spectrometry (AMS) at the Uppsala facility. Considering that some supplementary 129 I is added during the chemical procedure, the resulting 129 I/127 I ratios of the AMS-targets (Figure 2.4) can be described from intrinsic, carrier and background iodine according to 129 I target 127 I target. =. 129 I 129 I intrinsic + background 127 I 127 I + intrinsic carrier. Assuming that the amount of intrinsic iodine (127 Iintrinsic) in samples is much less than the iodine added from carrier, the above ratio can be approximated by 129 I 129 Iintrinsic 129 Ibackground target ≈ + 127 127 I 127 I Icarrier target carrier This approximation is adequate when 2 mg (2000 μ g) of iodine carrier is added to samples that contains less than 20 μ g of intrinsic iodine, which apply to conditions in this study. The last term is approximated by procedural blank values and is further described in section 2.4.2. Extraction of 129 I from pore and lake water was done according to the procedure in Buraglio et al. [2000]. For extraction of 129 I from the aerosol samples, the filters were mixed with 35 ml of 0.5 M KOH and 0.05 M Na2 SO3 10.

(180) Pump. Thermal ribbons Bypass connection. Sample + carrier Quartz wool. Furnace. O2 Trapping solution Figure 2.3: Principal sketch of the combustion system.. 100 – 1500 mg dried and pulverized sample + 2 mg dissolved iodine carrier Stepwise heating: 200° – 300° C, 45 – 120 min. 680° C, 15 min. 800° C, 30 min. Collection of iodine in trapping solution (KOH + Na2SO3) Acidification to pH 2, Precipitation of AgI by AgNO3 AMS-targets, AgI + Nb Figure 2.4: Flow chart of the combustion procedure.. 11.

(181) solution together with 1 - 2 mg of iodine carrier, and stirred for ∼2 days. The mixture was filtered and iodine was precipitated as AgI. Procedural blank values of (2 − 5) × 10−13 in 129 I/127 I were at least three times lower than the ratio of 10−12 to 10−10 measured in sediments, water and aerosol samples. A summary of measured samples for 129 I is given in (Table 2.2). Site. Material. Interval. Resolution. Lake Nylandssjön. Sediment. 1942 - 2007. Lake Nylandssjön Lake Lehmilampi. Lake water Sediment. 2 - 15 m 1968 - 2006. Lake Loppesjön core 1 core 2 core 3 Baltic Sea core 1 core 2 Kiruna. Lake water Sediment Sediment Pore water. 0.5 m - bottom 0 - 30 cm 0 - 17 cm 0 - 24 cm. 1 year, annual to biannual 2 samples 1 - 5 years, entire interval 4 samples 1 cm, entire interval 0.3 - 1 cm, 21 samples 3 cm, 3 samples. Sediment Sediment Aerosols. 0 - 51 cm 0 - 45 cm 1983 - 2000. Ljungbyhed. Aerosols. 1983 - 2000. 1 cm, 12 samples 1 cm, 11 samples ∼4 weeks/year, 10 selective years ∼4 weeks/year, 7 selective years. Table 2.2: Data related to the measured 129 I samples.. Measurement of total iodine in sediments was carried out at the University College Dublin following the procedure of Ohashi et al. [2000] and at the Risø National Laboratory for Sustainable Energy, Denmark, using the ICPMS method. For aerosol and water samples only the ICP-MS method was employed.. 2.4.1. The Uppsala accelerator mass spectrometry system. The Uppsala tandem accelerator facility used for the measurements of 129 I is shown in (Figure 2.5). The injector is equipped with a Cs+ sputter ion source for negative ion production that can be loaded with up to 20 samples, a 90◦ double-focusing electrostatic deflector and a 90◦ double-focusing sector magnet. The vacuum chamber in the injector magnet is electrically isolated and an external power supply is applied for generating square-type voltage pulses for rapid interval selection of the 127 I and 129 I isotopes (∼10 Hz). Ion beam acceleration is achieved by using a 5 MV tandem accelerator (NEC 15SDH2 Pelletron ). The analyzing and switching magnets at the high energy side makes separation of heavy elements feasible. A dedicated AMS beam line is located after the switching magnet followed by a quadrapole doublet magnetic 12.

(182) Gas ionization detector. AMS beamline. Switching magnet. Post accelerator analyzing magnet. Electrostatic deflector. Accelerator tank. 0. 1. Ion source. Electrostatic deflector. Injector for I-129. Magnetic deflector. 10 m. lens, +20◦ electrostatic deflector and energy defining slits in front of a gas ionization energy detector. Setups for beam diagnostics consist of a Faraday cups, beam profile monitors and viewers in order to tune and optimize the ion beam transport.. Figure 2.5: The Uppsala tandem accelerator.. 2.4.2. Measuring procedure of 129I by AMS. The measuring scheme includes measurements of 5 - 10 samples followed by a measurement of a standard sample with a known isotopic ratio of 1.05 × 13.

(183) 10−11 (±3%), obtained from dilution of the standard material NIST SRM 4949C. This sequence is repeated a number of times until required accuracy is achieved, normally 1 to 4 repetitions. In order to obtain standard and Poisson errors for 129 I/127 I ratios, the time interval for a single sample measurement, normally 300 seconds, is partitioned into 10 second subintervals. In the presentation of 129 I/127 I ratio and error estimation, the following notation are used Sample cycle. Time interval for a single sample measurement, normally 300 s.. Sample subcycle. A sample cycle is partitioned in time intervals of 10 s.. Standard cycle. The sequence of 5 - 10 sample cycles, initialized and terminated with measurement of a standard sample.. Ratio and error over a sample subcycle is expressed as ri Niri N Ri = si σRi = Q Qsi where N ri describes the number of 129 I atoms detected (ri = rare isotope) and Qsix represents the integrated current of 127 I (si = stable isotope). The standard deviation, σRi , is obtained by Poisson statistics. Consequently, ratio R over a sample cycle, containing i sample subcycles, is obtained by the formula R=. ∑ik=1 Nkri ∑ik=1 Qsik. The corresponding Poisson and standard error is calculated as i Nkri ∑ k=1 ∑ik=1 (Rk − R)2 Standard σRPoisson = i σ = R (i − 1) ∑k=1 Qsik. (2.1). In following calculations, error for a sample cycle is taken as the maximum of the Poisson and standard error. σRsmp = max (σRPoisson , σRStandard ) Sample ratios (i.e. 129 I/127 I ratio obtained during a sample cycle) are normalized with standard 129 I/127 I ratios, which are obtained by time interpolated values of the standard samples that initiate and terminates a standard cycle. Hence, the sample ratio R for a single standard cycle is achieved as R=. 14. Rsmp , w1 Rstd (i) + w2 Rstd (i + 1).

(184) with start time of Rsmp total time of the standard cycle w1 = 1 − w2. w2 =. The error associated with the time interpolated standard value, σinterpl , is. σinterpl = max (σRPoisson , (i) std. σRPoisson , (i+1) std. σRStandard , (i) std. σRStandard , (i+1) std. .... 1 [Rstd (i) − Rstd(i + 1)]) (2.2) 2. Using the error propagation rule, the error for the sample ratio is expressed as 2 σRsmp 2 σinterpl σR = R + Rsmp w1 Rstd (i) + w2 Rstd (i + 1) Here, the σRsmp and σR are determined by the maximum error achieved in std equation (2.1). The ratio of a sample over the entire measurement procedure is then summarized as the weighted average of each R at the m times the sample is processed Rk ∑m k=1 σ 2 R R = m 1k ∑k=1 σ 2 Rk. and the corresponding weighted Poisson and standard error are ∑m (Rk −R)2 k=1 σR2 1 k σRPoisson = , σRStandard = . m 1 (i − 1) ∑k=1 σ12 ∑m k=1 σ 2 R Rk. k. Background correction is performed by use of procedural blank values Rbl (including the chemical procedure as well), to obtain the corrected ratios R R = R − Rbl , σR = σR2 + σR2 bl. Consequently, sample 129 I/127 I ratio corrected for background is given as 129 129 I I = R × 127 127 I I std smp. 15.

(185) The latter term is the 129 I/127 I ratio of the standard sample. Using the error propagation rule, the corrected error is ⎛ σ ⎞2 129 I 129 σ 2 127 I I ⎜ ⎟ R σ 129 I = R × 127 × + ⎝ 129 std ⎠ I I std R 127 I 127 I smp std. 16.

(186) 3. Results. The presentation of the results is divided into three sections, one devoted to the experimental sample preparation methodology, the other to sediments (i.e. the profiles of lake sediments in Sweden, Finland as well as sediments from the Baltic Sea) and the third to aerosols.. 3.1. Sample preparation method. The sample preparation method has been tested for a variety of environmental materials with a wide range of 129 I concentrations (Figure 3.1) [Englund et al., 2007]. The preparation scheme was setup with frequently measurements of blanks, which showed a negligible memory effect between preparations. Repeated combustions show reproducible concentrations and absolute concentrations were validated by measurements of IAEA standard soil # 375. Results of five consecutive measurements of the IAEA standard soil sample were (11.4 ± 0.4) × 108 atoms g−1 , which is within the confidence interval given from IAEA of (9.7 − 14.7) × 108 atoms g−1 [Strachnov et al., 1996]. 10. 129. I atoms g−1. 9. 10. Surface soil Central Sweden. 8. Maple leaves. 10. 7. 10. Soil "sample 10". 129. IAEA #375. I [108 atoms g−1]. 10. 15 Nominal value 10. 5. 0 6. 10. (a) Range and reproducibility for different materials. . nvbwgkmq nvbwgkmq nv wgkmq et al tudy t ss Laboratory code ida Thi z S. (b) IAEA standard soil # 375. Figure 3.1: Performance of the preparation method used in this study for different materials (a) including standard soil IAEA # 375, maple leaves and soil "sample 10" from the interlaboratory comparison [Roberts et al., 1997]. (b) shows measurement of this study compared with data from the different AMS laboratories in [Roberts and Caffee, 2000]; the three clusters represent different chemical preparation procedures.. 17.

(187) 3.2. Chronology of the sediment profiles. In order to establish absolute chronologies of the sediment profiles, isotopic measurements were used together with the conventional varve counting. Sediments of both Swedish lakes have a marked enhancement in 137 Cs at the 1986 year varve (Lake Nylandssjön) and at 9 cm depth (Lake Loppesjön), which clearly are linked to the Chernobyl accident (Figure 3.2a-b). Total 137 Cs inventory in the sediments are 4.6 Bq cm−2 in Lake Nylandssjön and 3.7 Bq cm−2 in Lake Loppesjön, which are 44% and 37% higher than the Chernobyl fallout at the respective site, i.e. 3.2 Bq cm−2 and 2.7 Bq cm−2 respectively [Edvarson, 1991]. The 1986 137 Cs peak in Lake Nylandssjön is higher than in Lake Loppesjön, which reflects the higher activity of ∼50 Bq g−1 in Lake Nylandssjön compared to ∼10 Bq g−1 in Lake Loppesjön. In contrast with the Swedish lakes, the 137 Cs profile in Lake Lehmilampi, Finland, shows constantly low values of < 2 Bq g−1 (Figure 3.2c), which is in accordance with the low Chernobyl 137 Cs fallout of <0.05 Bq cm−2 at the area [Paatero et al., 2007].. 137. Cs [Bq g−1]. 0. 2. 4. 6. 137 8. 10 2010. 0. 0. 10. 20. [Bq g−1] 30. 40. 50. 60. 2000. Depth [year]. Depth [cm]. 5 10 15 20. 1990 1980 1970 1960 1950. 25. Core 1 Core 2. 1930. (a) Lake Loppesjön 137 2010. This study Appleby 1994. 1940. 30. 0. 1. 2. (b) Lake Nylandssjön 137. [Bq g−1] 3. −1. Cs [mBq g ]. 4. 5. 0. 0. 20. 40. 60. 80. Core 1 Core 2. 2005. Depth [cm]. Depth [year]. 2000 1995 1990 1985 1980. 5. 10. 1975 1970. 15. 1965. (c) Lake Lehmilampi. (d) Baltic Sea. Figure 3.2: Concentrations of 137 Cs in lake and marine sediments (dry weight). Measuring accuracy is within 5%.. 18.

(188) DΔ14C (per mille) 500. 14. 0. 1000 2000. 5. 1990. 90. Depth [cm]. C activity [pM] 95 100. 0. 105. 1980. 10. Year. The 137 Cs profiles in the Baltic Sea exhibit a marked increase in concentrations from 14 cm to 12 cm and 13 cm to 11 cm in the two sediments cores respectively, corresponding to maximum values of ∼ 0.08 Bq g−1 (Figure 3.2d). Activity at more recent sediments (above 11 cm), remains at a relatively high level, although decreasing towards the top to ∼ 0.04 Bq g−1 . The 137 Cs profiles demonstrates impact from the Chernobyl accident (the single outstanding source of 137 Cs). Consequently, the layers at 11 cm and 13 cm respectively are dated to 1986. Carbon-14 activity associated with the atmospheric nuclear weapon tests [Naegler and Levin, 2006] was traced in six samples from Lake Loppesjön in order to mark the increased signal around 1963-65 (Figure 3.3). The 14 C enhancement is clearly observed at 15 - 16 cm depth defining the absolute chronology at this interval. As a reference, measured 14 C fallout in the atmosphere is shown in figure 3.4.. 1970 15. 1960. 20. 1950. Figure 3.3: 14 C activity in sediments from Lake Loppesjön.. Figure 3.4: Atmospheric [Levin et al., 1985].. 210. log( 3. 3.5. 4. 14 C. fallout. −1. Pb) [mBq g ] 4.5. 5. 5.5. 6. 0. Depth [cm]. 5 10 15 20 25 30. Figure 3.5: Excess. 210 Pb. in sediments of Lake Loppesjön. Sediment accumulation rate in Lake Loppesjön was determined by the time marker of 137 Cs and 14 C, associated to the Chernobyl accident in 1986 and 19.

(189) the nuclear weapon tests during the 1960s respectively. In addition, the 210 Pb (Figure 3.5) was considered for determination of sedimentation rates by using the Constant Rate of Supply model (CRS) and Constant Initial Concentration model (CIC) approaches of [Appleby and Oldfield, 1992]. The different approaches generated similar results of ∼50 mg (cm2 y)−1 . Apparently, the sediment accumulation rate has relatively small variations with depth, since (1) the methods of the 210 Pb dating show small variations with depth and (2) the time marker determinations results in similar accumulation rates. The sediment accumulation rates in Lake Nylandssjön vary between 16 and 37 mg (cm2 y)−1 , whereas the variability in Lake Lehmilampi is 29 and 62 mg (cm2 y)−1 , but is associated with high uncertainty. 129. 8. 129. −1. 0. 5. 10. 15. 20. 0. 25. −1. 5. 10. 15. 20. 25. 40. 50. 2010. 2000. 2000. 1980. Depth [year]. Depth [year]. 8. I [10 atoms g ]. I [10 atoms g ]. 1960 1940. 1980 1970 1960 1950. Core 1 Core 2. 1920. 1990. 1940. (a) Lake Loppesjön. (b) Lake Nylandssjön 129. 129. I [108 atoms g−1]. 0. 5. 10. 10. 20. −1. 30. 0. 2010. 10. Depth [cm]. 2000. Depth [year]. 8. I [10 atoms g ]. 0. 15. 1990 1980 1970. 20 30 40 50. Baltic 1 Baltic 2. 60. 1960. (c) Lake Lehmilampi. (d) Baltic Sea. Figure 3.6: Concentrations of 129 I in lake and Baltic Sea sediments.. 3.3. Iodine in sediment. The distribution of 129 I in the sediment profiles from the Swedish lakes, Lake Nylandssjön and Lake Loppesjön (Figure 3.6) exhibits similar patterns. In both profiles, there appears a constantly low 129 I concentrations ∼ (0.2 − 0.3) × 108 atoms g−1 that continue until ∼1960. This trend is followed by 20.

(190) a gradual increase until 1986 where a marked enhancement in concentrations over a narrow interval is observed. In the succeeding interval 129 I of Lake Loppesjön shows an approximately constant level of 10 × 108 atoms g−1 , whereas the profile of Lake Nylandssjön is varying at (14 − 24) × 108 atoms g−1 . At the topmost part of the sections, 129 I concentrations increase toward the sediment-water interface in both lakes. The sediment section in Lake Lehmilampi, Finland, exhibits a general increase in the 129 I concentrations throughout the profile, with maximum a value observed at the top of the section (Figure 3.6). The pre-1986 129 I concentrations of (2 − 3) × 108 atoms g−1 are in fair agreement among all the three lakes. There is, however, no corresponding increase in 129 I concentrations in the Finnish lake as is seen in the Swedish lakes around 1986. Similar to the trends observed in the lake sediment profiles, the 129 I concentrations in the Baltic Sea sediments show the lowest values at the bottom of the cores and an increasing signal towards the sediment-water interface (Figure 3.6), with exceptions at depths of ∼17 cm and the topmost layer. The 129 I concentrations are generally higher than in lake sediments, ranging between (0.5 − 45) × 108 atoms g−1 . The amount of 129 I in pore water in sediments of Lake Loppesjön ranged at ∼ (17 − 30) × 105 atoms ml−1 at three depth intervals between the top and 24 cm depth (Table 3.1). By calculating concentrations in atoms (ml wet sediment)−1 , the pore water 129 I concentrations vary between 1% at top and 28 % at bottom at the respective depths in sediment (Table 3.2). The amount of 129 I in the pore water is, however, negligible compared to the integrated amount of 129 I in the sediment.. Lake water 0.5 m 8m 13 m Bottom Sediment Top water 0 - 3 cm 10 - 13 cm 21 - 24 cm. Volume (wet sed.). Volume (pore water). [ml]. [ml]. 129 I. in Pore water/ lake water [105 atoms ml−1 ] 1.8 ± 0.1 2.0 ± 0.3 1.7 ± 0.2 7.6 ± 0.3. 94 94 94. 36 35 30. 19.8 ± 0.7 29.6 ± 1.3 27.0 ± 1.0 16.9 ± 1.1. Table 3.1: 129 I in lake water and pore water in Lake Loppesjön.. The 129 I concentrations in the lake water, ∼ 2 × 105 atoms ml−1 in Lake Loppesjön and ∼ 5 × 105 atoms ml−1 in Lake Nylandssjön, are considerably less than in the pore water (Figure 3.7). Bottom water in Lake Loppesjön 21.

(191) 129 I. in pore water. Partition. [106 atoms (ml wet sediment)−1 ]. Interval in sediment 0 - 3 cm. 1.1 ± 0.1. 1%. 10 - 13 cm. 1.0 ± 0.1. 12%. 21 - 24 cm. 0.5 ± 0.05. 28%. Table 3.2: Amount of 129 I contained in the pore water and related sediments of Lake Loppesjön.. (accumulation bottom) also shows enhanced concentrations relative to the lake water. 5. Depth in lake water. 0 0.5 m. 5. −1. 10 atoms (ml of WATER) 10 15 20 25 (2 m). 30. 35. Lake Loppesjön Lake Nylandssjön. 8m. Depth in sediment [cm]. 13 m (15 m) ~1 dm above bottom ~1 cm above bottom Water−sediment interface 0−3 10−13 21−24. Figure 3.7: 129 I concentration in lake water and pore water of the sediment in Lake Loppesjön together with lake water concentrations in Lake Nylandssjön.. Total iodine (127 I) concentration is higher in Lake Nylandssjön compared to Lake Loppesjön (Figure 3.8). The iodine concentration in the sediments from the Baltic Sea range between 40 and 80 μ g g−1 , comparable with the concentrations in Lake Nylandssjön. Total organic content varies between 11% and 22% in Lake Nylandssjön, 8% - 12% in Lake Loppesjön and 4% - 10% in the Baltic Sea . Despite visual correlation between total iodine and organic content, only the data from Lake Nylandssjön permit estimation of a reliable correlation coefficient of r2 = 0.85.. 22.

(192) Total iodine [ μg g−1] 0. 4. −1. Total iodine [ μg g ]. 6. 8. 0. 10. 20. 30. 40. 50. 60. 70. 2010. 0. Carbon, core 1 Carbon, core 2 T. iodine, core 3. 2000. Depth [year]. 5. Depth [cm]. 2. 10 15 20 25. 1990 1980 1970 1960 Total iodine Carbon. 1950. 30. 1940. 0. 5. 10. 0. 15. 5. 10. 15. 20. Carbon content [%]. Carbon content [%]. (a) Lake Loppesjön. (b) Lake Nylandssjön Total iodine [ μg g−1]. 0. 20. 40. 60. 80. 0 Carbon, core 1 Carbon, core 2 T. iodine, core1. Depth [cm]. 10 20 30 40 50 60 0. 2. 4. 6. 8. 10. Carbon content [%]. (c) Baltic Sea. Figure 3.8: Total iodine and carbon contents in sediments from Swedish lakes and Baltic Sea.. 23.

(193) 3.4. Aerosols. The concentrations of total iodine in aerosols vary between 0.34 and 2.31 ng m−3 (median 0.58 ng m−3 ) in southern Sweden and between 0.14 and 0.49 ng m−3 (median 0.29 ng m−3 ) in northern Sweden. The data do no show any clear temporal trends (Figure 3.9) considering the annual standard deviation of 0.45 ng m−3 and 0.1 ng m−3 for southern and northern Sweden respectively. 2.5 0.5. 2. ng m−3. ng m. −3. 0.4 1.5 1. 0.3 0.2. 0.5. 0.1. 0. 0 1985. 1990. 1995. 2000. 1985. 1990. Year. 1995. 2000. Year. (a) Southern Sweden (Ljungbyhed). (b) Northern Sweden (Kiruna). Figure 3.9: Total iodine in aerosols in southern and northern Sweden. The concentrations of 129 I in aerosols show a higher range in southern (4 − 203) × 104 atoms m−3 compared to northern (0.7 − 39) × 104 atoms m−3 Sweden (Figure 3.10). Despite differences in amplitude, mean annual concentrations at the two sampling sites are correlated, r2 = 0.68. 50. 250 Winter Summer. 40. 10 atoms m. 150. 30 20. 4. 100. 4. 10 atoms m. −3. −3. 200. 10. 50. 0. 0 1985. 1990. 1995. 2000. 1985. 1990. (a) Southern Sweden (Ljungbyhed). 1995. 2000. Year. Year. (b) Northern Sweden (Kiruna). Figure 3.10: 129 I in aerosols in southern and northern Sweden. Isotopic ratios of 129 I/127 I [atoms/atoms] varied between 1.3 × 10−8 and 23.9 × 10−8 (median 5.9 × 10−8 ) in southern Sweden and between 0.6 × 10−8 and 7.1 × 10−8 (median 2.3 × 10−8 ) in northern Sweden (Figure 3.11). 24.

(194) [atoms/atoms]. 20. I 10. −8. 15 10. 127. [atoms/atoms] −8. 25. I/. 5. 129. 129. I/. 127. I 10. 30. 0 1985. 1990. 1995. 2000. Year. (a) Southern Sweden (Ljungbyhed). 9 8 7 6 5 4 3 2 1 0 1985. 1990. 1995. 2000. Year. (b) Northern Sweden (Kiruna). Figure 3.11: 129 I/127 I ratios in aerosols at southern and northern Sweden. 25.

(195)

(196) 4. Modeling. The 129 I distribution patterns in the lake sediment archives were used to numerically model the contributions of the different sources to the studied regions. Modeling of 129 I fallout was accomplished by taking into account relative influence from emission rates, possible transport pathways and the influences of the lake system. Major sources of emission were considered to be: 1) Liquid and gaseous releases from the reprocessing facilities in Sellafield and La Hague. The two facilities are here regarded as a single source, although different modes of release form (gaseous or liquid) is addressed. 2) Fallout from the Chernobyl accident in 1986 and 3) Fallout from the nuclear weapon tests during the late 1950s and early 1960s. A comprehensive description of the model is provided in Paper IV. Influences from the different sources are described by the vectors in Table 4.1.. Model vector. Description. Chern (t) NW (t). Influences from the Chernobyl accident in 1986 Influences from the nuclear weapon tests in late 1950s and early 1960s Influences from gaseous emissions from the facilities in Sellafield and La Hague Influences from the cumulative liquid emissions from the facilities in Sellafield and La Hague. G (t) LΣ (t). Table 4.1: Description of model vectors. Chern (t) is defined as a single spike with height 1 at the year t = 1986 and zero elsewhere, and NW (t) is proportional to the modeled fallout values from the nuclear weapon tests at the late 1950s and early 1960s in central Europe [Reithmeier et al., 2006]. G (t) is proportional to the summed gaseous releases from the facilities in Sellafield and La Hague, and LΣ (t) describes the cumulative liquid emissions in the English Channel/ Irish Sea /North Sea, and is defined as follows. Let L (t) [kg] designate the liquid release rate of 129 I in absolute amounts [kg] (Figure 5.2). The parameter tlag corresponds to the time interval when the emissions from the reprocessing facilities are supposed to affect the sampling sites. Accordingly, the total amount of liquid 129 I that 27.

(197) will impact the investigated area at time t is described as LΣ (t) =. t. τ =t−tlag. L ( τ ) d τ. [kg]. The modeling vector LΣ (t) is proportional to LΣ according to the normalization described below. Fallout at the sampling sites are described as a weighted sum of the normalized vectors (LΣ (t), G (t),Chern (t), NW (t)) multiplied with the fallout constant φF129 Fallout (t) = φF129 [wL LΣ (t) + wG G (t) + wChernChern (t) + wNW NW (t)] , (4.1) with the conditions wL + wG + wChern + wNW = 1 and. ∑ LΣ (t) = ∑ G (t) = ∑ Chern (t) = ∑ NW (t) = 1 t. t. t. t. L (t). so that LΣ (t) = ∑ ΣL (t) and so forth. With this definition, the weights (wL , wG , t Σ wCh and wNW ) designate the total contribution (as a fraction between 0 and 1) of 129 I in fallout from a specific source during the measured time interval. Influences from the lake system are described with the convolution Dep (t) = φIF ×. n. ∑. (ai f allout (t − i)). (4.2). i= −m. where the sequence {ai }ni=−m denotes the impact of the lake system on 137 Cs or 129 I, stretching over a time interval of (m + n + 1) years. The constant φIF (impact factor) denotes the relation between 129 I fallout and deposition into the sediments. In case of 137 Cs, the values of the coefficients {ai }ni=−m are proportional to the 137 Cs sediment profile itself and are calculated as Dep 137 (1986 + i) , ∑τ Dep 137 (τ ) m = 1986 − 1942 = 44 n = 2006 − 1986 = 20. ai =. (4.3). The response function, Φfallout → sed , for 129 I (Figure 4.1 and 4.2) is assumed to follow 137 Cs and is hence expressed on basis of the coefficients ai in equa-. 28.

(198) tion (4.3) as min(n,2006−t). Dep129 (t) = φ IF129 ×. ∑. ai Fallout129 (t − i) =. i=−m˜ min(n,2006−t). φ IF129 φ F129 × CAmp. ∑. ai [wL LΣ (t − i) + wG G (t − i)+. i=−m˜. wChernChern (t − i) + wNW NW (t − i)] (4.4) Because of assumed impact from the nuclear weapon tests, the sequence ai was here truncated for i > m˜ = 16, representing an effect from the year 1969 and earlier in the 137 Cs profile, and again normalized.. 0.1. 0.4 0.35. 0.08. ai. 0.25. a. i. 0.3 0.06. 0.2 0.04. 0.15 0.1. 0.02. 0.05 0 20. 15. 10. 5. 0. −5. −10. −15. 0 20. 15. 10. i. 5. 0. −5. −10. −15. i. (a) Lake Nylandssjön and Lake Lehmilampi. (b) Lake Loppesjön. Figure 4.1: 129 I response functions for Lake Nylandssjön, Lake Lehmilampi and Lake Loppesjön.. The relation between deposition at the lake bottom and concentration in sediment with steady state diagenesis is assumed to (137 Cs or 129 I) Dep (t − x) = where t x acc (t − x) dia (x) Dep (t − x) Conc (x,t). Conc (x,t) × acc (t − x) dia (x). (4.5). Time of sampling Depth measured in years Dry mass accumulation rate Diagenetic influence Deposition Concentration. The model was applied to 129 I profiles of lake sediments at Lake Nylandssjön, Lake Lehmilampi and Lake Loppesjön. In case of the latter sampling site 29.

(199) [Arbitrary unit] 0. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 0 2. A. 4. Depth [years]. 6 8. B. 10 12 14. C. 16 18. Fallout Total respons Individual responses. 20. Figure 4.2: Illustration of how three hypothetical point releases (A, B and C) are distributed in the sediment according to the response function for Lake Nylandssjön.. (Lake Loppesjön), a time lag ρ = 3 years related to processes in the lake system was incorporated into the model, according to the observed shift of ∼1 cm between the 129 I and 137 Cs profiles [Englund et al., 2008], and equation (4.4) transforms to min(n,2006−t). Dep129 (t + ρ ) = φ IF129 ×. ∑. ai Fallout129 (t − i). i=−m˜. In case of Lake Lehmilampi, which lacks a clear 137 Cs time marker from the Chernobyl accident, the response function is approximated with that of Lake Nylandssjön. This is motivated by the similar conditions at the two lakes with respect to lake and drainage areas and the fact that both lakes has laminated sediments. The diagenetic behavior of 129 I is approximated by that of carbon, Closs , in Lake Nylandssjön [Gälman et al., 2008] in all sampling sites according to 1 Closs (t), (4.6) dia (t) = 1 − 1.2 100 The cumulative effect of the liquid 129 I releases, determined by tlag , was constrained to an upper limit of 3 years related to the transit times between the reprocessing facilities and the North Sea/Norwegian coast waters boundary. The model parameters, CAmp , tlag , wL , wChern and wNW (the fourth weight, wG , comes from the condition wL + wG + wChern + wNW = 1), have been optimized to minimize the objective function Q = ∑y εt2 , with the norm εt defined as ε (t) = log (Dep129 (t)measured) − log (Dep129 (t)modeled) in order to meet the criteria [Fletcher, 2000; Xu, 2001] • {ε (t)} is independent of {ε (t + k)}, for any k ≥ 1 30.

(200) • {ε (t)} has no trend component and constant variance when plotted against either Dep129 (t) or t In addition to the above criteria, the set of {ε (t)} is also normally distributed (Kolgorov-Smirnoff test), which allows to infer standard deviations of the model parameters. The optimization is performed by the interior-reflective Newton method [Coleman and Li, 1994; Coleman and Yuying, 1996] used in the MATLABTM environment.. 31.

(201)

(202) 5. Discussion. The sample preparation method presented has proven to be reproducible and is independently validated. However, to achieve a high chemical yield, special precautions have been taken regarding the combustion. Stepwise heating, slow oxygen gas flow rate and continuous warming of the whole system to avoid condensation were among the critical processing steps of significance. 6. 0. −1. 6. 80. 2000. 2000. 1990. 1990. 1980 1970 1960 1950. (a). Measured deposition Modeled deposition Liquid derived fallout Gaseous derived fallout. 2. −1. 10 atoms (cm y) 20 40 60. 0. Year. Year. 2. 10 atoms (cm y) 20 40 60. 80. 1980 1970 1960 1950. (a) Lake Nylandssjön. (b). (b) Lake Lehmilampi 106 atoms (cm2 y)−1 50 100. 0 2000. Year. 1990 1980 1970 1960 1950. (c). (c) Lake Loppesjön. Figure 5.1: Modeled and measured and Finland. 129. I deposition in lake sediments from Sweden. Despite different geographical location, drainage areas, lake surface areas and precipitation rates between the sites (Table 2.1), there is a rather similar range in the 129 I deposited within the sediment for the period 1950 to 1986 (Figure 5.1). The profile of the pre nuclear era (before 1950s) is shown in Lake 33.

(203) Loppesjön and Lake Nylandssjön. This period is characterized by close to natural background 129 I concentrations. At the beginning of the 1950s, the 129 I flux to the sediments increases indicating contributions from anthropogenic emissions related to the Sellafield facility (Figure 5.2) and from atmospheric nuclear bomb tests [Schink et al., 1995; Oktay et al., 2000]. A further increase reflecting emissions from La Hague facility is also marked at the late 1960s. Liquid 129I emissions [kg] 0. 100. 200. 300. Gaseous 129I emissions [kg] 400 0. 2. 4. 6. 8. 10. 2000. Year. 1990. 1980. 1970. 1960 Sellafield La Hauge Total 1950. Figure 5.2: Historical liquid and gaseous emissions from the reprocessing facilities in Sellafield and La Hague.. The sharp increase in the 129 I profile of both Lake Loppesjön and Lake Nylandssjön during 1986 coincides with the Chernobyl accident. This pattern is not observed in Lake Lehmilampi and thus clearly demonstrates the response of the sediments to the varying intensity of the Chernobyl fallout over north Europe. Use of Chernobyl 137 Cs distribution as a response function to model 129 I deposition seems to agree well with measured data. However, the mechanisms that control 129 I transport throughout the system (ocean/atmosphere/land) are not directly described by the model parameterization. Accordingly, the main outcome of this approach is a quantification of the cumulative rather than individual effects of each process. The positive correlation observed between 129 I and carbon (r2 = 0.87) and total iodine and carbon (r2 = 0.85), show that most of the 129 I is associated with the organic material in the sediment. Despite these correlations the 129 I flux pattern cannot be explained by changes in the organic matter content, due 34.

(204) to the relatively much lower variation in the organic content (×2 difference) compared to 129 I (×100 difference). The same argument can also be applied to infer only small impact of diagenesis on the 129 I flux pattern observed since the total iodine concentrations vary only by a factor of two (Figure 3.8). Post-Chernobyl 129 I fluxes differ somewhat among the sites, (40 − 60) × 106 atoms (cm2 y)−1 in Lake Nylandssjön, (60 − 80) × 106 atoms (cm2 y)−1 in Lake Loppesjön, and (20 − 70) × 106 atoms (cm2 y)−1 in Lake Lehmilampi, without significant trend at the former two sites. The flux of 129 I to the sediment depends on both dry and wet fallout of the isotope. Estimate of the wet 129 I fallout was based on precipitation data from Uppsala (59◦ 51’ N, 17◦ 38’ E) and Abisko (68◦ 21’ N, 18◦ 49’ E) [Persson, 2007]. The average annual flux between 2001 and 2005 from precipitation data ranges between 60 × 106 atoms (cm2 y)−1 and 30 × 106 atoms (cm2 y)−1 , without significant temporal trend. This implies that a major part of the 129 I fluxes to the studied sites is associated with wet fallout.. Liquid emissions Gaseous emissions Chernobyl accident Nuclear weapon tests Liquid time delay (years) ∗. Lake Nylandssjön. Lake Lehmilampi. Lake Loppesjön. 70 ± 6% 16 ± 7% 10 ± 4% 4 ± 1% 1 l.b.∗∗. 71 ± 18% 29 ± 20% < 8% 3 u.b.∗. 50 ± 7% 28 ± 9% 21 ± 6% 1 ± 1% 1 l.b.∗∗. Upper boundary value Lower boundary value. ∗∗. Table 5.1: Model outcome on the relative contributions of 129 I from different sources, derived from the weights wL , wG , wCh and wNW in equation (4.1) and (4.4). Modeled 129 I deposition into the sediment archives at the three studied sites (Figure 5.1) depicts contribution from liquid emissions as the dominating fallout source, accounting for 50% to 71% of the total inventories (Table 5.1). A transfer rate of 129 I from sea to atmosphere is derived for pertinent sea areas (English Channel, Irish Sea and North Sea) and is estimated at 0.04 to 0.21 y−1 . This range is higher than the 0.003 y−1 to 0.01 y−1 values previously derived by [Schnabel et al., 2001; Reithmeier et al., 2006] based on 127 I. The discrepancy between the two estimates may be explained by different chemical speciation of 129 I and 127 I and corresponding efficiency in sea-atmosphere transfer. A study of the surface waters in the North Sea clearly shows that 129 I and 127 I occur in different chemical forms in the same water parcel [Hou et al., 2007]. Dry deposition as well as washout of iodine is much dependent on the associated particle size distribution. Estimating dry fallout based on average 129 I concentrations in aerosols and using settling velocity within a range of 0.001 m/s to 0.02 m/s [Baker et al., 2001], indicate values of (2 − 600) × 105 atoms (cm2 y)−1 and (5 −900) ×104 atoms (cm2 y)−1 for southern and northern Sweden respectively (during the years 1983 - 2000). In comparison with wet fall35.

(205) out during the period 2001 - 2005, which is estimated to 3 × 108 atoms (cm2 y)−1 and 3 × 107 atoms (cm2 y)−1 for southern and northern Sweden respectively [Persson, 2007], the dry fallout constitutes at most 1/4 of the total 129 I fallout. However, the upper limit of the settling velocity (0.02 m s−1 ) reflects particles > 1μ m in diameter, whereas the main atmospheric carrier of 129 I seems to be smaller particles [Wimschneider and Heumann, 1995; Vogt et al., 1999]. Hence, the contribution from dry fallout is relatively small, which is consistent with the agreement between wet fallout and fluxes in sediment [Paper IV]. There is a large difference in the concentration of both 129 I and 127 I in aerosols with respect to the two samplings sites (Figure 3.9 and 3.10) where the higher values are found in southern Sweden. However, the difference is even larger in the 129 I values, ∼9 times, compared to ∼2 times in 127 I concentration. A possible explanation for this feature is the different distances to the emission sources, regarding both the marine surface waters with high 129 I concentrations and the reprocessing facilities (Sellafield and La Hague). Such difference in 127 I concentration in the aerosol has also been observed in precipitation [Persson, 2007]. The 129 I/127 I values in aerosols apparently follow an increasing temporal trend during the years 1983-2000 (Figure 3.11) that partly reflect the liquid releases (Figure 5.2). The contribution of 129 I atmospheric fallout to the Baltic Sea sediment is negligible compared to the supply by sea current from the Kattegatt- Skagerrak [Aldahan et al., 2007b]. Accordingly, the lag time in the received liquid emissions to the studied sites in the Baltic Sea from the reprocessing facilities is expected to be considerably delayed in comparison to the releases occasion. This factor induces difficulties in linking the 129 I trends found in the sediments of the Baltic Sea with those of the lakes. Furthermore dilution of the Chernobyl 129 I contribution by the vast Baltic Sea basin may have caused the relatively weaker 129 I signal in the sediments profiles compared to those of the lakes.. 36.

(206) 6. Conclusions. The principal findings of this thesis are summarized as follows, • A routine sample preparation procedure for extraction of 129 I from small amount (at mg level) of solid materials has been developed. The reproducibility and accuracy of the procedure were evaluated through measurements of international standard reference materials. • The first documentation of anthropogenic 129 I profiles in lake sediments from northern Europe and the Baltic Sea covering the last 80 years is presented. • Imprints of emissions from the nuclear fuel reprocessing facilities in Sellafield (UK) and La Hague (France) are identified in the sediments since the start of operation in 1952 and 1966 respectively. • Signals of 129 I introduced from the atmospheric nuclear weapon tests during the 1950s and 1960s are apparently mixed and overwhelmed in the studied sediments by contributions from the reprocessing activity. • The 129 I associated with the Chernobyl accident is well identified in the sediments from the Swedish lakes, which are located in areas affected by high Chernobyl fallout. The Chernobyl 129 I can also be traced in regions with low fallout as shown by the sediment profiles from the Baltic Sea and Finland. The major pathway of 129 I into the Baltic Sea goes through currents from the North Sea and the associated liquid releases from the reprocessing facilities. • Numerical modeling of the 129 I deposition indicated that more than 50% of the total inventory of the lake sediments are related to the liquid emissions from the reprocessing facilities. The modeling also reasonably simulates the contribution of the Chernobyl event to the total 129 I flux. • The novel time series from northern Europe on 129 I in aerosols show about one order of magnitude higher concentration in northern Sweden compared to southern Sweden. Estimate of 129 I dry fallout based on the aerosol data suggests <25% contribution of the total fallout. • The distribution of 129 I in the sediment archives demonstrates the potential of the isotope as a new time marker for chronological and environmental investigations.. 37.

(207)

(208) 7. Acknowledgments. First of all, I want to thank my supervisors, Göran Possnert and Ala Aldahan for supporting me in finishing this work. In particular, I want to express my gratitude for the intensive cooperation during the latest time, when both of you have spent a lot of time with me to achieve a satisfactory result. Göran, thanks for spending time listen to my ideas, and sharing your analytical skills with me. Thanks Ala, for your enthusiasm, collaboration and response to all my scientific writing stuff and also introducing a humoristic perspective of the world of science. I would like to express my gratitude to my co-authors Ingmar Renberg, Xiaolin Hou and Timo Saarinen for a fruitful collaboration and constructive comments. Special thanks to Jan-Erik Wallin for the help with slicing sediments from Lake Nylandssjön. Thanks Vasily, for introducing me to the laboratories at Geocentrum and Ångström, and for spending time on discussing my questions, despite that you had so many other things to do. Your friendly and never ending positive attitude have encouraged me in my work. I am also very thankful to you people who have supported me in all the practical stuff concerning the experimental setup of the 129 I extraction procedure. Thanks Inge, for constructing all my versions of the combustion furnaces. Sören, for helping me to improve the laboratory at Geocentrum, and Inger, for introducing me to the chemical procedures concerning the 129 I extraction. Bengt and Rogerio, you have been excellent teachers at the tandem laboratory! Thanks Susanna, my companion in the 129 I research field, for great discussions and being a nice friend. I want to thank you all at the tandem laboratory and the Division of Ion Physics for nice company during my time here. Thanks Geotryckeriet for the help with the printing of this thesis! Finally, I want to thank my family for the encouragement during the years. 39.

(209)

(210) 8. Summary in Swedish. Allt sedan början av den nukleära tidsepoken, som startade under 1940-talet då kärnvapenutvecklingen under andra världskriget initierades, har vår miljö belastats med stora mängder radioaktivitet. 1950 och 1960-talen kännetecknades av det kalla krigets kapprustning och ett antal atmosfäriska kärnvapentestsprängningar utfördes som på ett dramatiskt sätt ökade radioaktiviteten i miljön främst i norra hemisfären. Under senare delen av 1900-talet är det i första hand den fredliga användningen av uran för energiproduktion i fissionsrektorer som dominerat bidragen av antropogen radioaktivitet, dels direkt genom utsläpp från reaktorbyggnaderna, dels indirekt vid hanteringen av det utbrända reaktorbränslet. Av speciell betydelse är också haveriet i den sovjetiska reaktorn i Tjernobyl 1986 eftersom utsläppet vid den händelsen är väl karakteriserat i både tid och rum. Den här avhandlingen är inriktad mot studier av den långlivade radioisotopen jod-129 (T1/2 = 15.7 miljoner år) som förekommer naturligt i mycket låga koncentrationer till en följd av fission av uran i jordskorpan och kärnreaktioner mellan den kosmiska strålningen och atmosfärens xenon. Idag har dock den mänskliga användningen av uran som kärnbränsle medfört att den naturliga förekomsten av jod-129 i stort sett är försumbar jämfört med den mängd som introducerats i naturen från upparbetningen av kärnbränsle. De globalt sett mest relevanta utsläppskällorna i detta sammanhang är upparbetningsanläggningarna i Sellafield (England) och La Hague (Frankrike) som togs i bruk under 1950 och 1960-talen. Merparten av det jod-129 som släppts ut är vätskeburen och utsläppen sker främst till Irländska sjön och Engelska kanalen (ca. 5000 kg fram till 2006). En mindre del av den upparbetade aktiviteten avges direkt till atmosfären som gas (ca. 200 kg fram till 2006). Undersökningar av den stabila isotopen av jod (jod-127) som förekommer i relativt stora mängder i havet visar att endast en liten del transporteras från havet till kontinenternas landområden via atmosfären och att inverkan av den biologiska aktiviteten hos alger och plankton i ytvattnet är av betydelse. Den biologiska påverkan tillsammans med att jod kan förekomma i ett antalet möjliga oxidationstillstånd (-1 till +7) möjliggör en antal olika jodföreningar i miljön. Transportmekanismer och spridning av jod är därför komplicerad och idag endast delvis kartlagd. Utöver den grundläggande geokemiska forskningen om jod i miljön har bl.a. dess inverkan på klimatet i samband med nedbrytning av marknära ozon i atmosfären ägnats speciellt intresse, liksom dess produktion av kondensationskärnor som har en central betydelse vid molnbildning.. 41.

No results found