Site-Sensitive Hazards of Potential Airborne Radioactive Release from Sources on the Kola Peninsula

R SE9800283 l-ebruary 1998 ISSN 1104-9154

DEFENCE RESEARCH Scientific report ESTABLISHMENT Site-sensitive Hazards of Potential Airborne Radioactive Release from Sources on the Kola Peninsula

R. Bergman*, L. Thaning*, and A. Baklanov*1 *The Defence Research Establishment (FOA), Umea, Sweden ** Kola Science Centre of Russian Science Academy, Apatity, Russia

KNNP Ground contamination after 5 days Total release: 1 Bq Start971228 00GMT Duration: 1 h Height: 100 m Dry + wet deposition

Bq/m > 10 •" > io13 > io12 > 10"

1OOO 1900 2000 25O0 Start Um« 981228 OOOO - Slop time 9S122B 01 00 HEIGHT 100 m Submarine Ground contamination after 5 days Total release: 1 Bq Start 961228 00 GMT Duration: 1 h Height: 100 m Dry + wet deposition

Bq/m > 10H > 10 " -12 I •11 > 10

1000 1500 2000 2MO Start Urn* J91228 0000 - Stop tfrrw 581228 0100 HDCHT 100 m

Division of NBC Defence 29-4 SE-901 82 UMEA DEFENCE RESEARCH ESTABLISHMENT FOA-R-98-00717-861-SE Division of NBC Defence February 1998 SE-901 82 UMEA ISSN 1104-9154 SWEDEN

Ronny Bergman, LennartThaning, Alexander Baklanov Site-sensitive Hazards of Potential Airborne Radioactive Release from Sources on the Kola Peninsula

Distribution: FOA: Issuing organization Document ref. No., ISRN Defence Research Establishment FOA-R-98-00717-861 -SE Division of NBC Defence Date of issue Project No. SE-901 82UMEA February 1998 E479 SWEDEN Project name (abbrev. if necessary) Long-term consequences of radioactive fallout Author(s) Initiator or sponsoring organization FOA, OCB Ronny Bergman, Lennart Thaning, Alexander Project manager Baklanov Lars Rejnus Scientifically and technically responsible Ronny Bergman Document title Site-sensitive Hazards of Potential Airborne Radioactive Release from Sources on the Kola Peninsula

Abstract An abstract of this paper has previously appeared in Extended abstracts volume 2 from the Third International Conference on ENVIRONMENTAL RADIOACTIVITY IN THE ARCTIC, Troms0, Norway June 1-5, 1997. In this work we focus on cases of airborne releases from some of the sources on the Kola Peninsula - primarily nuclear reactors on submarines and the Kola Nuclear Power Plant (KNPP). The purpose of our study is to illustrate, and discuss some features - dependent on site and release characteristics - of the deposition patterns resulting from assumed unit radioactive releases to the atmosphere from a location at a fjord and from the KNPP in Polyarnye Zori. Using meteorological data for one real weather situation, the analysis is based on simulating the transport in air of assumed radioactive releases and estimating the deposition pattern on local, meso- and regional scales. By allowing unit releases to occur simultaneously from the site at the fjord and from the power plant (and with the same release profile in time) comparisons are made of differences in deposition patterns in and outside the Kola region. In this case study a set of assumed release heights, durations of the release, and particle size distributions are applied to indicate the dependence for the resulting deposition pattern on these parameters.

Keywords Radioactive, risk, source, Kola, release, airborne, Further bibliographic information Language English

ISSN 1104-9154 ISBN Pages p. 14 Price Ace. to pricelist

Distributor (if not issuing organization) Dokumentets utgivare Dokumentbeteckning, ISRN Försvarets forskningsanstalt FO A-R--98-00717-861 --SE Avdelningen för NBC-skydd Dokumentets datum Uppdragsnummer 901 82 UMEÅ Februari 1998 Projektnamn (ev förkortat) Långsiktiga radiakproblem Upphovsman(män) Uppdragsgivare FOA, ÖCB Ronny Bergman, Lennart Thaning, Alexander Projektansvarig Baklanov Lars Rejnus Fackansvarig Ronny Bergman Dokumentets titel i översättning Lägesberoende risker för Iuftburna radioaktiva utsläpp från källor på Kolahalvön

Sammanfattning Rapporten behandlar fall av utsläpp i luften från några källor av särskilt intresse på Kolahalvön - främst kärnreaktorer på ubåtar och kärnkraftverket vid Imandrasjön. Syftet är att illustrera och analysera huvuddragen - beroende på geografiskt läge och utsläppskaraktäristik - avseende de beläggningsmönster som uppkommer efter antagna atmosfäriska utsläpp av enhetsaktivitet från en plats vid en fjord och från kärnkraftverket Polyarnye Zori.

Beräkning och analys av depositionsmönstret på lokal nivå samt på meso- och regional skala görs med användning av meteorologiska data för en verklig vädersituation och simulering av transporten i luft av det antagna radioaktiva utsläppet. Genom att låta enhetsutsläppen ske samtidigt från läget vid fjorden och vid kärnkraftverket (och med samma utsläppsprofil i tiden) görs jämförelser av skillnader i depositionsmönster i och utanför regionen på Kola. Denna fallstudie utgår från ett antal antagna utsläppshöjder, tidsintervall för utsläppen och partikelfördelningar för att indikera beroendet för det resulterande beläggningsmönstret på dessa parametrar.

Nyckelord Radioaktiv, risk, källa, Kola, utsläpp, luftburen

Övriga bibliografiska uppgifter Språk Engelska

ISSN 1104-9154 ISBN

Omfång 14 Pris Enligt prislista

Distributör (om annan än ovan) INTRODUCTION Several studies dealing with radioactive sources and actual or potential contamination in the whole or part of the Arctic (AMAP 1994, NACC 1995, OTA 1995, Bellona 1996, IIASA 1996) consistently indicate that most Arctic sources of substantial radiological concern are to be found in north-western Russia on the Kola Peninsula, Novaya Zemlya and in the adjacent Barents and Kara Seas. In this paper we focus on cases of airborne releases from some of these sources - primarily nuclear reactors on submarines and the Kola Nuclear Power Plant (KNPP). The purpose with our study is to illustrate, and discuss some features - dependent on site and release characteristics - of the deposition patterns resulting from assumed unit radioactive releases to the atmosphere from a location at a fjord and from the KNPP in Polyarnye Zori. We don not attempt to describe the probability for - and the chain of events leading to significant radioactive releases. However, based on previous assessments of the probable upper level of the inventory - primarily of 137Cs in reactor fuel used in submarines (NACC 1995, Gussgard 1995) - we scale the deposition derived from our calculation of unit release to reflect release of 100 % of the inventory. (The radioactive deposition at various distances associated with other assumptions about the fraction released to the environment may thus easily be obtained by another choice of the scaling factor.) We discuss the hazards indicated for areas in the vicinity and far from these sources, as well as illustrate uncertainties involved in generalisations from case studies like this. SIMULATIONS OF RELEASES TO THE ATMOSPHERE Using meteorological data for one real weather situation, the analysis is based on simulating the transport in air of assumed radioactive releases and estimating the deposition pattern on local, meso- and regional scales. By allowing unit releases to occur simultaneously from the site at the fjord and at the power plant (and with the same release profile in time) compaiisons are made of differences in deposition patterns in and outside the Kola *--- region. In this case study a set of assumed release heights, ;:-Case studies ol release from durations of the release, and particle size distributions are -a flora site anfl uie ivom "> applied to indicate the dependence for the resulting deposition pattern on these parameters. These sets of lsfi?^0^ ^rKnfrs parameters values are illustrated to the right. j •Initial plume rise The deposition patterns based on unit release from a fjord 100, 300, 500, 1000 m site and the location of the Kola Nuclear Power-Plant ex- _ ,. . . _,...... hibit the range of significant deposition subsequent to a 0.3u•Particlm aned siz3uem distribution release under the specific weather conditions prevailing at a (median radius, lognormal) certain date close and far from the source - and at a deposition coefficient of 1.5 mm/s, which is assumed ap- •Deposition velocity propriate for the particulate carrier concerning deposition at 1.5 mm/s and 15 mm/s relatively long distances in the Chernobyl case (Devell I 'Deposition with and 1991), and similar to the standard value of 1 mm/s. L ^Aout jgredpitation Regional transport For the simulation on the regional scale, the MATHEW/ADPIC model has been used. MATHEW/ADPIC was developed at the Lawrence Livermore National Laboratory (LLNL), USA (Sherman, 1978, Lange, 1978, Foster, 1992) and has been adjusted at the Defence Re- search Establishment (FOA) for the FOA-environment (Thaning & Naslund, 1991, Naslund & Holmstrom, 1993). In ADPIC, the dispersion is described by a Particle-in-Cell model. The 3D model wind field, in which the particles are advected, is mass consistent and produced by in- terpolating real wind data from the European Centre for Medium-Range Weather Forecasts (ECMWF), Reading, UK into the model grid. The FOA version of the model has been compared with the ETEX-1 Full-scale Experiment, and has been proved to be capable of producing a realistic evolution in time for the concentration (Thaning and Johanson 1995, Noclop 1997). The calculations for regional scale have been carried out for the region of Scandinavia, Finland and Northwest Russia including the Kola Peninsula at a distance up to 2000 km. The release parameter variations for the model calculations include the particle size distribution, height of release, duration of release, wet or dry deposition and release site. Figure 1 illustrate the accumulated deposition pattern over north-western Russia and Fenno-Scandia five days after the start of a 1 hour release at a height of 100 m from the Kola Nuclear Power Plant and from a fjord site on the Kola Peninsula, respectively. These two cases constitute "reference cases" for the subsequent comparisons of deposition patterns dependent of specific combinations of parameter values appealing in figures 2 - 6. KNNP Ground contamination after 5 days Total release: 1 Bq Start 971228 00 GMT Duration: 1 h Height: 100 m Dry + wet deposition

Bq/m •u > 10 > 10 •12 > 10 > 10

1O0O 1500 2000 2500 Stort t!m« 981228 OOOQ - Stop tlm. 991228 0100 HEICHT 100 m

Submarine Ground contamination after 5 days Total release: 1 Bq Start 961228 00 GMT Duration: 1 h Height: 100 m Dry + wet deposition

Bq/m > 10 •13 > 10 > 10' > 10

1OOO 1500 2000 2500 Start tlm. 961221 0000 - Stop time 96122B 0100 HEIGHT 100 m [km] Fig 1. Release assumed to occur from the Kola Nuclear Power Plant (KNPP) or from the fjord site (Submarine). Deposition pattern (cumulative deposition of Cs by wet and dry-deposition) for the case of a unit (lBq) 1 hour release at 100 m height. DEPOSITION KNPP Duration 1 hour 0.1 - 1 urn V = 0.15 cm/s 1 - 10 um V.= 0.15 cm/s

1000 1600 2000 2500 1000 1600 2000 2SO0 Start tim. 881228 0000 - Stop tim. 581228 0100 HEIGHT 100 m Start lirrt. 881228 OOOO - Step t!m« 581228 0100 HEIGHT 100 m M [km] 0.1-lum V = 1.5 cm/s 0.1 - 1 um V = 0.15 cm/s

1000 1500 2000 2500 1000 1500 2000 2300 Start time 96122S 0000 - Stop tlm. 981228 0100 HBCHT 100 m Start Um. 9B1228 0000 - Stop Brno SS122B 0100 HEIGHT 100 m [km] [km]

-14 -13 > 10 10 ^> 10'12 p> 10'" (Bq/m2)/Bq Figure 2 a (upper left) and 2b (upper right): Deposition pattern (cumulative deposition of 137Cs by wet and dry-deposition) for the case of a unit 1 hour release at 100 m height from the Kola Nuclear Power Plant (KNPP) under the assumption of 2a: Log normal panicle size distribution in the range 0.1-1 |mn, 2b: 1 - 10 urn, median range 0.3 |im and 3 um respectively; Figure 2c (lower left): deposition velocity assumed to be 1.5 cm/s, i.e. ten times higher* than the standard case; Figure 2d (lower right): no wet deposition is allowed, only dry deposition is operative.

IOOO 1500 2000 2500 1000 1500 2000 2500 Start tlrnn 961223 0000 - Stop timo 961228 0100 HEIGHT 100 m Start timo 951228 0000 - Slop time 961228 0100 HEIGHT 300 m [km] [km] Release height 500 m Release height 1000 m

1000 1500 2000 2500 1000 1300 2000 2500 Stort tlm« 961228 0000 - Slop Urrw 981228 0100 HDOHT 500 m [km] Stort tlrrn 961228 0000 - Stop time 96122S 0100 HEIGHT 1000 m

13 > 10 > 10' > 10" (Bq/m2) / Bq Figure 3 a-c: Dependence of the deposition pattern on assumed height of release at the fjord site (Submarine case). Release characteristics as in the reference situation (Fig. 1) except for release height, which is: 3a (upper left) 100m; 3b (upper right) 300 m; 3c (lower left) 500 m; 3d (lower right) 1000 m. DEPOSITION KNPP Duration 1 hour Release height 100 m Release height 300 m

1000 1500 2000 2500 1000 1300 2000 2300 Start Uma 961220 0000 - Slop time 961228 0100 HEIGHT 100 m 5tort tlm. 961228 0000 - Stop tlma 961228 0100 HEIGHT 300 m [km] [km] Release height 500m Release height 1000 m

1000 1500 2000 2500 2000 Start time 9812Z8 0000 - Stop time 981228 0100 HEIGHT 500 m [km] Start tfrni 961228 0000 - Stop tl™ 961228 0100 HUGH I 1000 I [Km]

10 M> 10 M> 10 M> 10 (Bq/m2) / Bq Figure 4 a-d: Dependence of the deposition pattern on assumed height of release** from the Kola Nuclear Power Plant (KNPP case). Release characteristics as in the reference situation (Fig. 1) except for release height, which is: 4a (upper left) 100m; 4b (upper right) 300 m; 4c (lower left) 500 m; 4d (lower right) 1000 m.

** The effective mixing height used in the simulations was 600 m. DEPOSITION KNPP Submarine Duration 1 hour

1000 1500 2000 2500 1000 1500 2000 2500 Stort tim. 961228 0000 - Stop Urn. 961226 0100 HEICHT 100 m Stort time 961226 0000 - Stop time 981228 0100 HEICHT 100 m [km] [km]

Without wet-deposition Without wet-deposition

1000 1500 2000 2500 1000 1500 2000 2500 Start time 9612Z8 0000 - Stop time 961228 0100 HEIGHT 100 m Start time 961228 0000 - Stop time 961228 0100 HEIGHT 100 m [km] [km] -14 -13 "12 535™ -11 2 > 10 > 10 10 \> 10 (Bq/m)/Bq Figure 5. Effects on the resulting deposition pattern when transfer by the actual wet deposition is included (upper left and right) or excluded (lower left and right) for the cases of release from the Kola Nuclear Power Plant, KNPP case, (left half); and the fjord site, Submarine case, (right half). Release characteristics are for the rest the same as in figure 1. DEPOSITION ^ KNPP Submarine Duration 1 hour Duration 1 hour

rV- i I i , , i I 1DO0 1500 2D00 2500 1000 1500 2000 2500 Start time 961228 0000 - Stop tlmo 961228 0100 HEIGHT 100 m Start tlmo 961228 OOOO - Stop time 961228 0100 HEIGHT 100 m [km] [km] Duration 48 hours Duration 48 hours

.-rarar'.,.._.- 1000 1500 2000 2500 1000 1500 2000 2500 Start tim« 981228 OOOO - Stop tfm« 961230 OOOO HEICHT 100 m 5tort time 961228 OOOO - Stop tlmo 961230 OOOO HEIGHT 100 m [Km] [km] •14 -13 •12 > 10 (Bq/m2)/Bq Figure 6. Effects on the resulting deposition pattern of release during 1 or 48 hours for the cases of release from the Kola Nuclear Power Plant, KNPP case, (left half); and the fjord site, Submarine case, (right half). Release characteristics are for the rest the same as in figure 1.

10 SOURCE STRENGTH, POTENTIAL RELEASE AND RESULTING DEPOSITION A conspicuous feature in the calculated deposition patterns is the similar levels of deposition arising in most pails of Northern Europe exposed to the radioactive release. It is reasonable to expect that the notable characteristics of relative high deposition at long distances will prevail also in most other weather conditions. The Chernobyl case and results from several calculations based on other weather data support this presumption. However, to analyse the possible radiological consequences of such a release, among other the source-strength of the most important radioactive nuclides, the fraction of the inventory released in accidents and the time-course of the release need to be considered. Submarines in operation or awaiting decommissioning Although ship reactor accidents may lead to serious environmental consequences, some case studies - in principle similar to those performed for accidental releases from nuclear power plants - indicate that any potential naval rector accident will not be nearly as severe as the Chernobyl accident (NACC1995). Some calculations of airborne transport, deposition and exposure have been made based on an assumed release of 1 PBq 137Cs (NACC 1995). It is not obvious that such quantities would actually be released in an accident, even if they are available in the core, and the relative uncertainty of the release is estimated to be a factor of ten. The maximum content of I37Cs in a first and second generation Russian ship reactor is estimated to 5 PBq (Gussgard 1995), while 10 PBq may be accumulated during operation of a typical modern ship reactor - i.e. with a power level slightly less than 200 MW. Our calculations for unit release in the fjord case indicate that large areas as far as in southern Scandinavia will obtain deposition of a fraction of the order of 10~13 per m2 (see illustration below). This corresponds to a deposition of 100 Bq/m2 per 1 PBq of airborne release of 137Cs in the accident, i.e. the level used in the NACC (1995) study. The extreme case of a release of the whole inventory of I37Cs in a modern submarine, i.e. 10 PBq according to Gussgard (1995) would give rise to levels of the order of 1 kBq/m2. On the Kola Peninsula and in the adjacent regions of Norway and Finland one to two orders of magnitude higher deposition density may be expected. One PBq of airborne l37Cs release corresponds in our fjord case to 1-10 kBq/m2.

pistance from point of release Fraction of unit release I deposited per m2 pn the Kola Peninsula 11 1A-l2 in- &nd in adjacent regions of Norway and Finland JLU • JLU

1000-2000 km >10"13 (100 Bq/m2 per 1 PBq of airborne release of 137Cs)

Source inventory of I37Cs ;Submarines 3-10 PBq The oldest unit of the Kola NPP ca 150 PBq

11 The Kola Nuclear Power Plant Unit release from the Kola Nuclear Power Plant is used for comparisons with the submarine case, primarily with focus on possible site-related effects on the deposition pattern locally and regionally. The reactor core inventory of caesium isotopes in the oldest Unit 1 of the VVER- 440/230 type of the Kola NPP is u4Cs ca 150 PBq, and 137Cs ca 200 PBq (IIASA 1996) respectively, i.e. 20 - 50 times the inventory in the different generations of submarine reactors. Scenarios in previous studies of severe accidents cover a broad range of releases - from around 3% to about the same fraction as in the Chernobyl case (ca 30%) or even higher (IIASA 1996). A one percent release would correspond to about 1.5 PBq of 134Cs and 2 PBq of 137Cs.

CONTAMINATION IN FOOD-CHAINS Radioactive contamination in food-chains often exhibit a considerable seasonal variation related to when deposition has occurred. Generally, agricultural systems based mainly on cultivation of grass crops and on pastures for feeding of animal herds belong to the most sensitive. Early after deposition milk from dairy cows remaining on pasture exposed to the radioactive deposition is particularly affected, and if the deposition contains 131I this nuclide will probably be the radioecologically most important factor in the early phase after a fallout. Direct deposition of 131I or 137Cs on pasture and crops at levels falling below 1 kBq/m2 is not likely to lead to restrictions in normal agricultural practice or formal acceptance for commercial use of the food-products, even in the period during summer when the resulting contamination will attain its maximum. The concentration of radioactive caesium ( Cs and 137Cs, ) in reindeer is high, when feeding during winter on lichens exposed to fallout. The ratio between activity concentration (Bq/kg) in reindeer meat and deposition density (Bq/m2) will be close to 1 (kg/m2) in the latter half of the first winter season after the fallout (Ahman and Ahman 1994). This implies contamination about one order of magnitude higher than concerning direct deposition in the sensitive food-chain over grass-cow-milk. Intake of radioactive caesium in groups or populations consuming large quantities of reindeer meat is therefore expected to be particularly high. This is accentuated by the persistence and high availability during several years over the lichen pathway for 137Cs, Case study as well as the fact that reindeer herding is extensive, and constitutes an important factor in 1 PBq release of 137Cs the economies in northern Fenno-Scandia and on from a fjord site the Kola Peninsula. or the Kola Nuclear Power Plant Assuming a release of 1 PBq of 137Cs in our case study, the levels of 137Cs attained in reindeer Areas affected by deposition meat during winter slaughter are expected to be r: on me Kola Peninsula of the order of 1 kBq/kg in large parts of the areas affected by deposition on the Kola and adjacent regions of Fenno-Scamlia Peninsula, and one order of magnitude higher still Levels in reindeer : U7 over somewhat smaller areas ranging several 1-10 kBq/kg of Cs hundreds of kilometres from the point of release. With other wind directions similar deposition H: Further SOUth in Scandinavia patterns are likely to occur within the same range from the source, but over other parts of Fenno- Levels in reindeer : 137 Scandia and Kola. of the order of 100 Bq/kg of Cs

12 CONCLUSIONS The parameter variations (i.e. particle size distribution, height of release, duration of release, wet or dry deposition, release site) illustrate that: • PARTICLE SIZE DISTRIBUTIONS chosen to resemble that found in Western Europe and Scan- dinavia after the Chernobyl accident (mean radius 0.3 urn), or of a size distribution with a ten times higher mean (3 |jm) yield similar deposition patterns;

• PREDOMINANT RADIONUCLIDES: Radionuclide composition of the modelled release is an important characteristic for different type of risk objects and is different for the local and regional scales. For a NPP release during the first days, iodine dominates in the exposure for a local/meso scale, whereas at the end of the acute phase 137Cs starts to dominate for the regional scale especially;

• INITIAL PLUME RISE ("release height") below the mixing height affects only slightly the results outside the local scale, while plume rise above that level leads to radically changed patterns with relatively low depositions on the local and mesoscale ranges;

• DURATION OF RELEASE: The longer the duration the more extensive the areas covered by relatively high deposition;

• RELEASE SITE : The patterns associated with our cases of release from a source at the fjord site in comparison to release at the site of the nuclear power plant, at the same date and time, exhibit clearly a strong site dependence of rainfall and wind direction on the resulting deposition on the regional scale. Besides possible direct radiological consequences primarily of relevance for exposure in populations close to the source, the deposition in our case studies - under the assumption of a 1 PBq release of 137Cs - involves risk for long-time adverse economical effects. This is indicated by the vulnerability to the impact of Chernobyl-fallout exhibited by some regional economies in Fenno-Scandia strongly dependent on certain boreal food-chains - notably rein-deer herding. Thus, it may also likely lead to negative social and cultural side effects. Despite the much smaller inventory of radioactive nuclides in naval nuclear reactors, as compared to the reactors at the Kola Nuclear Power Plant, a closer study of the probable source strength in various release scenarios is thus motivated for submarine reactors at sites along the shoreline of Kola, and - depending on the outcome - to be pursued with further analyses to elucidate the potential impact in radiological, as well as socio-economical and cultural dimensions.

REFERRENCES AMAP 1994. Radioactivity in the Arctic, Arctic Monitoring and Assessment Programme: Radioactivity Assessment. Report 94:1. Baklanov, A., Burman, J. & Naslund, E. 1996: Numerical modelling of three-dimensional flow and pollution transport over complex terrain during stable stratification. Scientific report, FOA, Umea, Sweden (FOA- R-96-00376-4.5-SE); Bellona 1994: Nilsen T., B0hmer N. Sources to radioactive contamination in Murmansk and Arkhangelsk counties. Bellona report vol 1:1994. The Bellona Foundation. Bellona 1996: Nilsen T., Kudrik I., Nikitin A. The Russian Northern Fleet: Sources of radioactive contamination. Bellona report vol. 1:1996. The Bellona Foundation. Devell L. 1991. Composition and properties of plume and fallout materials from the Chernobyl accident. In: The Chernobyl Fallout in Sweden, Ed. L Moberg, The Swedish Radiation Protection Institute, 29-46. Forster, C. S., Editor, 1992. User's guide to the MATHEW/ADPIC models. UCRL-MA-103581 Rev 1, Lawrence Livermore National Laboratory (LLNL), USA. Gussgard, K. 1995. Is spent nuclear fuel at the Kola Coast and dumped in waters a real danger? In: Environmental Radioactivity in the Arctic,Eds. P. Strand and A. Cooke, 0steras, 400-404,

13 IIASA 1996: Baklanov A, Bergman R, Segerstahl B. Radioactive sources in the Kola region: Actual and potential radiological consequences for man, International Institute for Applied Systems Analysis. Lange, R., 1978. ADPIC - a three dimensional particle-in-cell model for the dispersal of atmopheric pollutants and its comparison to regional tracer studies. J. Appl. Meteorol., 17, 320-329. NACC 1995. Cross-border environmental problems emanating from defence-related installations and activities. Vol 1. Phase 1: 1993-1995. Report no 204. North Atlantic Treaty Organisation. Nodop K., 1997. Etex symposium on long-range Atmospheric transport, model verification and emergency response 13-16 May 1997, Vienna, Austria. Office for Official Publications of the European Communities. Naslund, E. and Holmstrom, H., 1993. Inclusion of a three-dimensional washout coefficient in ADPIC, UCRL - ID-114054, Lawrence Livermore National Laboratory, Livermore, CA. OTA 1995. Nuclear Wastes in the Arctic. An Analysis of Arctic and Other Regional Impacts from Soviet Nuclear Contamination. Washington, Office of technology Assessment. Congress of the United States. Sherman C.A., 1978: MATHEW: A Mass-Consistent Model for Wind Fields Over Complex Terrain. J. Appl. Meteorol., 17,312-319. Spalding, D.B., 1981: 'A general purpose computer program for multi-dimentional, one- and two-phase flow', Mathematics and Computers in Simulation, XXIII, pp. 267-276. Thaning, L. and E. Naslund, 1991. A Simulation of radioactive fallout using the MATHEW/ADPIC model. FOA C 40292-4.3 Thaning, L. and P.-E. Johansson, 1995. Calculations performed by the Swedish Defence Research Institute (FOA). In: Report of the Nordic Dispersion-/Trajectory Model Comparison with the ETEX-1 Fullscale Experiment /Editors U. Tveten and T. Mikkelsen, Riso-R-847(EN), NKS EKO-4(95)1, p. 49-60. Ahman B. and Ahman G. 1994. Radiocaesium in Swedish reindeer after the Chernobyl fallout: Seasonal variation and long-term decline. Health Phys. 66(5): 503-512.