In freshwater, radionuclide transfers depend both on exchanges between liquid and solid phases and on mass transfers between suspended and bottom sediment. Whereas the former ones depend on chemical processes (sorption/desorption, complexation ...), the latter ones are regulated by hydrological and sedimentary considerations (dispersion, erosion, deposit ...) closely related to the interactions between flow, suspended matter and bed sediment. Previous studies highlight the need to consider the matter heterogeneity and its specific sediment dynamics to correctly report the in homogeneity of fluxes in time and in space. These considerations lead us to develop an experimental method to distinguish the different particle groups present in natural water, mainly according to their settling kinetics. In this context, this paper presents the experimental protocol TALISMEN to characterize a natural bulk suspension according to its main settling kinetics groups. In a first step, this identification is achieved by the use of a settling tank that allows the monitoring of the suspended solid concentrations at various depths, combined to a settling model applying a multi-class approach. In a second step, the particle groups are isolated and characterized. The sorption property of each group toward 137Cs is determined by estimation of its distribution coefficients (kd) and sorption kinetics.

The objectives of this work are to study the behaviour of 238U and its daughters in the continental environment and to determine the radiological impact of U mining past activities in the Haute-Vienne French administrative division. The distribution of radionuclides among different size fractions was investigated upstream and downstream mines. Particulate (>0.45μm) and dissolved (<0.45μm) fractions were separated by filtration and analysed by gamma spectrometry. A preconcentration step by evaporation was elaborated and validated for liquid samples. Results indicate that radiological impacts of U mines are the enhancement of radionuclides activities and a change of their distribution between particulate and dissolved fractions at 0.45μm. While activity levels depend on the input intensity, the distribution is controlled by the mine water treatment witch reduces drastically the contribution of mine radionuclides in streams. Analysis of particles from traps and sediments confirms that the distribution of mine radionuclides promote their accumulation in lacustrian sediments. Nevertheless, the impact of the past activities is very limited in space. Indeed, it becomes weak at 10 km downstream the inputs and is no more detectable at 30 km downstream.

Field experiments in the Chernobyl exclusion zone have been carried out to determine the behaviour of the radioactive halogens 36Cl and 125I (as a surrogate for 129I) in the soil-plant system and along human food chains. The investigations on the migration in four types of soil of the two radionuclides showed that the vertical transport of chlorine was the most rapid: 9 months after the contamination the residual fractions of 36Cl in the 20 cm arable layer varied from 1% to 3% according to soil type. For iodine this percentage was less than 3% in any case. The soil to plant transfers were quantified for various agricultural crops (cereals, grass, root vegetables, leaves vegetables and fruit vegetables) and for several kinds of soils. The root uptake of 36Cl by plants was particularly important. As a result of this very high biological assimilation about 60% of the contamination was extracted from the soils by the plants after one vegetation period. For iodine the concentration ratios values are 2 or 3 order of magnitude lower than for chlorine.

Some 3×109 kg of phosphogypsum (PG) wastes are annually generated by two fertiliser-production factories in Huelva (south-western Spain). PG has relatively high concentrations of 226Ra and other radionuclides, with an special concern due to the 222Rn emissions. These wastes could be used to improve the fertility of agriculture soils in a large former marsh area of the Guadalquivir river. Thus, it is interesting to study the levels and behaviour of natural radionuclides within this system to evaluate the radioactive impact of this amendment. An agronomical test is being conducted by one of the authors in an experimental farm in Lebrija (Seville). The soils are treated with 13 and 26 t ha-1 of PG, 30 t ha-1 of manure. Each treatment was repeated twice and continued for two years with beetroot and cotton plant production. We are measuring 226Ra (by alpha counting and gamma spectrometry) and U isotopes (by alpha spectrometry and ICP-MS analysis) in drainage waters, soils and vegetal-tissues samples. The PG used in the treatment has 620 ± 70 Bq kg-1 of 226Ra. The drainage waters have 226Ra contents similar to those from non-contaminated natural waters, but the uranium concentrations are one order of magnitude higher. Our results are suggesting that the major uranium input comes from the application of phosphate-fertiliser. No significant levels of radionuclides were found in the vegetal tissues.

Kinetic transfer coefficients are important parameters to understand and reliably model the behaviour of non-conservative radionuclides in aquatic environments. This report pertains to a series of radiotracer experiments on 239Pu uptake in natural unfiltered waters from the Hueznar reservoir (southern Spain). The experimental procedure ensures the preservation of environmental conditions and the dissolved Pu activity was measured by liquid scintillation technique. The data suggest that the main pathways for Pu uptake consist of two parallel and reversible reactions. We studied the effects of suspended matter concentrations and the specific surface area on the kinetic transfer coefficients.

The protection of biota from ionising radiations needs reliable predictions of radionuclide dynamics in wild animals. Data specific for many wild animal - radionuclide combinations is lacking and a number of approaches including allometry have been proposed to address this. However, for 14C and 3H, which are integral components of animal tissues and their diets, a different approach is needed. Here we propose a metabolically based model which can be parameterised predominantly on the basis of published metabolic data. We begin with a metabolic definition of the 14C and OBT loss rate (assumed to be the same) from the whole body and specific organs. The mammalian body is conceptually partitioned into compartments (body water, viscera, adipose, muscle, blood and remainder) and a simple model defined using net maintenance and growth needs of mammals. The model is tested with data from studies using rats and sheep. It provides a reliable prediction for whole body and muscle activity concentrations without the requirement for any calibration specific to 3H and 14C. Predictions from the model for representative wild mammals) are presented. Potential developments of a metabolic model for birds and the application of our work to human foodchain modelling are also discussed.

In order to perform radiological impact assessments of radioactive waste disposals, a compartmental model taking into account generic data for biosphere transfer parameters values is usually used. To improve its performance assessments calculations, Andra (French national radioactive waste management agency) decided to adapt this type of biosphere model to its sites conditions. On one hand, specific models have been developed such as model for 36Cl based on isotopic dilution and on local stable chlorine contents. On the other hand, biosphere transfer factors values are issued from experimental studies adapted to Andra context: soil characteristics, climate, agricultural habits and species. These specific data are useful only if main factors controlling mobility have been determined previously for each radionuclide, which implicates good knowledge of environmental radionuclides behaviour. This site-specific approach allows to reduce uncertainty and range of variation of impact calculations and ensure a consistent model. It helps also to achieve scientific community approval by showing mastery of radionuclides environmental processes and public understanding by focusing demonstration on real local conditions. The methodology is fully presented. Two examples for uranium and chlorine and the limits of this approach are given.

In a period between 1943 and 1974 radioactive wastes were temporary buried on a special site on the territory of RRC “Kurchatov Institute". The site monitoring held since the end of the 80s showed that contaminants are located not only on the surface and in the ground but they have also spread in groundwater. The paper presents preliminary results of the work on development of numerical models of radioactive contamination migration the waste disposal site. The objectives of the work were to evaluate the existing contamination plume, to determine mechanisms of contaminant migration on the site and to develop a numerical model of radioactive contamination transport that would allow correctly predicting further plume spreading for making necessary engineering decisions. Based on laboratory findings and radiation monitoring data obtained at the waste disposal site and its adjacent areas, there were determined the site hydrogeological structure and parameters, and a geoinformation database was developed. Three-dimensional numerical models of groundwater flow (using the MODFLOW code) and mass transport (using the MT3DMS code) were built and verified against field measurements. Using these models, preliminary predictions of radionuclide migration from the waste disposal site were made.

An accurate model of atmospheric dispersion of radionuclides over the complex terrain of La Hague reprocessing plant (North Cotentin, France) has been developed by COGEMA, in partnership with Paris VI University, and has been yet adapted for an other industrial site (Marcoule). This model, called CEDRAT, takes into account areas typically outside the validity range of Gaussian models: topography and building influence, short-distance (beyond 500 m from the release point) and various atmospheric stability conditions. CEDRAT allows to reach an effective compromise between rapidity (30 min on a 6000 nodes grid, on a standard PC), robustness and accuracy, coupled with a user-friendly interface. The validation process consisted first of a comparison with the 3D reference model MERCURE, developed by EDF, and then, results of both models were compared on release scenarios thanks to tracer data (Krypton-85, an inert gas released from a height of 100 m) collected at five monitoring stations around the site. Then a second level of comparison, presented here, was made in collaboration with the IRSN Cherbourg Radioecological Laboratory, on ground and altitude (collected with a tethered balloon) field data.

Foliar transfer of transuranium elements had not received the same attention as other long-lived radionuclides, such as the well-known classical fission products Cs and Sr. Thus, the aim of this study was to compare the foliar transfer parameters of 241Am and $^{239,240}$Pu with those of 137Cs and 85Sr. Bean plants (Phaseolus vulgaris) at flowering development stage were contaminated by soaking their first two trifoliate leaves for 3 hours in a solution containing 241Am, $^{239,240}$Pu, or 137Cs and 85Sr. Results showed that transuranium elements were more retained by leaves than Cs or Sr. The mean leave-to-pod translocation factor of Am was of the same order as that of Sr, showing that the behaviour of Am was close to the behaviour of Sr. Translocation of Pu was significantly lower than for Cs, Sr or Am and occurred preferentially in leaves. Differential mobility of Am, Pu, Cs and Sr were compared to those used in related studies. Comparison showed that, when specific values of representative radioecological parameters are lacking, the behaviour of Am and Pu towards foliar transfer may be conservatively described as “Sr-like".

A dynamic compartment model was developed to evaluate the transport of accidentally released radionuclides onto rice-fields. In the model, the surface water compartment and shoot-base absorption were introduced to account for the effect of an irrigation which is essential to rice cultivation. Allowance was made for the effect of a soil mixing by a plough and an irrigation before transplanting the rice. The rate of root-uptake and shoot-base absorption were modeled in terms of the function of the biomass. In order to test the validation of the model, some 137Cs deposition experiments were carried out for several years while cultivating rice in a greenhouse using soils sampled from rice-fields around Kori, Yonggwang and Ulchin nuclear power plants in Korea. The model prediction agreed well to the experimental results within one order of magnitude.

Long-lived radionuclides 137Cs and 90Sr have been regularly monitored since the 1960's in dairy milk in Finland. The aim of this work was to study regional differences in long-term behaviour of 137Cs and 90Sr in dairy milk including the nuclear weapons tests period and the period after the Chernobyl fallout. In the milk production areas in Finland soil types vary from clayish to humus and peat soils. Although total 137Cs deposition from nuclear weapon testings was approximately the same in all regions, 137Cs concentrations in dairy milk during the 1960s ranged between 5 and 35 Bq/l being highest in regions with high proportion of peat soils. The uneven deposition after the Chernobyl accident before the start of the growing season gave rise to 137Cs concentrations in dairy milk ranging from 3 Bq/l to 60 Bq/l. In the milk production areas dominated by organic soils the ecological half-lives for decrease of 137Cs were higher in the regions with clayish soils. Half-lives of 90Sr for longer time intervals were considerably longer than those for 137Cs. The long-term data on behaviour of 137Cs and 90Sr in dairy milk can be utilized in testing food chain models.

The investigation wells for a control of the underground water contamination were bored after the cratering explosions at the Semipalatinsk Test Site, now they are restored partially. The analysis of the retrospective information of the Institute of Global Climate & Ecology (Moscow, Russia) give a possibility to choose wells and terrains for the successful study of radionuclide migration with the underground water. The epicentral zone, the crater and the territory with radius 1,5 km around the explosion “1003" were investigated under the ISTC project K-810. Underground water and soil samples were taken at the two expeditions of 2003. The chemical extraction methods taking into account the water mineral composition, gamma-spectrum methods, methods of liquid scintillation spectrometry and methods of alfa-spectrometry were used. The modern radionuclide content (3H, 90Sr, 137Cs, 239+240Pu, 241Am etc.) of the underground water is presented and compared with a radionuclide content of soils around crater. The retrospective information has been added by these modern data. The vertical radionuclide distrioin soils is presented.

During the Chernobyl accident, large areas in Europe, particularly forests, were contaminated by radiocaesium. Soil-plant transfer is often the first step by which 137Cs enters to food chains by incorporating the forest biogeochemical cycle. The present work is devoted to assess the soil-plant transfer of radiocaesium in weakly contaminated forest area, in the Franche-Comté region (16 202 km2) in North Eastern France. This region has been markely contaminated after the Chernobyl accident. Twelve sites, with various functional features and high diversity of ecological contexts were selected to sample forests trees. Two species (Picea abies and Fagus sylvatica) and two organs, (leaves or needles and young branches) were sampled for radioactivity measurements. Radiocaesium activities in soils varied between 61 to 280 Bq.kg-1DW at 0-5 cm depth and between 14 to 224 Bq.kg-1DW at 10-15 cm. Radiocaesium activities in leaves, needles and branches displayed high variability. We reported a large variability of Aggregated Transfer Factors (Tag) values with respect to vegetation type, organ type and soil features. The activity concentration in the leaves and needles were generally higher than those measured in the branches. The Tag values could be explained by the radiocaesium bioavailability variability in relation to soil type, soil features and environmental parameters.

A linked chain of 1D river model RIVTOX and 3D estuary model THREETOX was used to assess impact of the previous and potential releases from the nuclear installations in the basins of Ob' and Yenisey rivers on radioactive contamination of the rivers and the Kara Sea. The RIVTOX includes the model of river hydraulics, suspended sediment and radionuclide transport in river channels. THREETOX includes a set of sub-models: a hydrodynamics sub-model, ice dynamics-thermodynamics sub-model, suspended sediment transport and radionuclide transport sub-models. The radionuclide transport model simulates processes in the water, suspended sediments and in bottom sediments. The simulations of the flux of 90Sr and 137Cs through the Ob' River and estuary from activities at the Mayak Production Association for the period 1949-1994 and through the Yenisey River and estuary from activities at the Mining, Chemical Combine for the period 1959-1994 were carried out. Based on scenarios, simulations have been performed in order to assess the potential risk of contamination from existing and potential sources of radionuclides into the Kara Sea through the Ob' and Yenisey rivers.

Two types of plants (lettuce and radish) were contaminated by dry deposition of radioactive aerosols (137Cs, 85Sr, 133Ba and 123mTe). Due to the rain efficiency in decreasing radiological contamination of crops, a rain scenario was applied on the two types of plants. These experiments aimed at improving the prediction of the consequences on agricultural products of a nuclear accident occurring on a PWR, and at improving the understanding of the different processes occurring between contamination and harvest. For a plant species at a given stage of its growth cycle, the interception of the four radionuclides was found to be identical and varied from 68% for radish to 75% (at the middle of ripeness stage) or 78% (at the mature plant stage). Using a conceptual model, the predominance between the two processes - washing-off the leaves and absorption through the leaves cuticles allowing translocation - was evidenced. For Cs, Sr and Ba, for both the lettuce and the radish, washing-off on the foliar cover was the most significant during the first rain. The absorption process inside the plant became dominant only from the third rain event. Tellurium had a particular behaviour: it was non-mobile and stayed put on the leaves. The values of the washing-off and absorption coefficients were estimated. The global transfer factor values were dependant on both the radionuclides and the plant species; nevertheless, a higher value was obtained for cesium, regardless of the plant and the rainfall (from 0.06 m2.kg$_{\scriptsize\textrm{fresh}}$-1 for radish to 0.1 m2.kg$_{\scriptsize\textrm{fresh}}$-1 for a whole lettuce).

Contaminations of the terrestrial gastropods Helix aspersa maxima by direct deposition or labelled food ingestion of 137Cs, 85Sr, 133Ba and 123mTe were carried out under laboratory conditions. The aim of this study was to compare the two contamination pathways: direct and trophic, in terms of individual mortality and radionuclide uptake, depuration and tissular distribution. A first group of 30 snails (2-years old) was exposed to radioactive aerosols during a twenty-hour period. These aerosols were assumed to be representative of those that would be released during a nuclear accident occurring in a PWR. A second group of 40 snails (same age) was submitted to an ingestion of food contaminated by the same aerosols, twice-a-week for 10 days (flour at a feeding rate of about 0.2g). During the 21 day-observation period, a comparison between the two groups and the reference group (not contaminated by radionuclides) was performed. No significant difference between the three groups was observed in the growth or in the mortality. One day after deposition, cesium was the most bioavailable element, distributed rather homogeneously through the whole body (from 13 to 28% of the total Cs in organs other than the digestive system and the muscle, respectively). Strontium accumulated in the shell (about 70%). Barium was found in the muscle (20%) and the shell (65%). Tellurium was mainly present in the shell (70%) and in the digestive system (20%). After a few days, this element was mostly present in the faeces. As regards contamination by ingestion, it was mainly accumulated in the digestive system.

The need to increase the predictive power of risk assessment for large tritium releases implies a process level aproach for model development. Tritium transfer for atmosphere to plant and the conversion in organically bound tritium depend strongly on plant characteristics, season, and meteorological conditions. In order to cope with this large variability and to avoid also, expensive calibration experiments, we developped a model using knowledge of plant physiology, agrometeorology, soil sciences, hydrology, and climatology. The transfer of tritiated water to plant is modelled with resistance approach including sparce canopy. The canopy resistance is modelled using Jarvis-Calvet approach modified in order to directly use the canopy photosynthesis rate. The crop growth model WOFOST is used for photosynthesis rate both for canopy resistence and formation of organically bound tritium, also. Using this formalism, the tritium transfer parameters are directely linked to known processes and parameters from agricultural sciences. The model predictions for tritium in wheat are closed to a factor two to experimental data without any calibration. The model also is tested for rice and soybean and can be applied for various plants and environmental conditions. For sparce canopy the model uses coupled equations between soil and plants.

Aerosol dry deposition onto agricultural and forest areas is investigated. Its estimation is still uncertain, particularly for rough surfaces and sub-micronic aerosols. In fact models used in radiological risk or air quality assessment studies are highly empirical and fail to reproduce the results of the most recent measurement campaigns. Therefore a theoretical framework, based on a mechanistic description, has been developed. The proposed approach consists in two steps. First, interaction between aerosols and foliar surface is formulated by using a set of parameters, which are defined on the local scale of one foliar element. In the second step, the collective effect of the foliage is taken into account through statistical distribution of these parameters. The model integrates the three main factors governing aerosol dry deposition. These are local aerodynamic characteristics of the flow within the canopy, aerosol mechanisms governing the deposition, and canopy structural and morphological properties. Physical processes considered in the study are inertial impaction, gravitational settling and Brownian diffusion. The canopy characteristics considered are spatial distribution, orientation and micro-structure of foliar surfaces. The applicability of the model is demonstrated on an existing measurement campaign.

In the context of protecting the environment from the effects of ionizing radiation associated with uranium mining, the alpha-emitting radium 226 is the uranium progeny of most interest as it is known to bioaccumulate and confers a high effective dose to biota. For the Ranger uranium mine in northern Australia, total radium 226 activity concentrations were determined in water samples from Magela Creek both upstream and downstream of the mine over two years. Those data were evaluated within three different frameworks and were found to comply with respective limits: human dose estimations, site-specific guideline `trigger' values for ecosystem health (derived from the distributional properties of values from a reference (upstream) dataset) and estimations of absorbed dose to freshwater mussels residing in Magela Creek (long-lived organisms with high rate of uptake and exposure). We propose that the currently-used, human-dose limit of 10 mBq.L-1 be applied to the difference of the wet season arithmetic means calculated for the upstream and downstream locations. In addition, the statistically derived trigger values (e.g. a moving 80$^{\scriptsize\textrm{th}}$ percentile of upstream data) could be used to assess if radium 226 measured downstream of the uranium mine has resulted in an additional dose to biota.