Etudes des mécanismes de mobilisation et d'entraînement des radionucléides au cours de l’érosion hydrique.

David CLAVAL thèse de doctorat de l'Université de Paris XI-Orsay, soutenue le 28/11/2000, 239 p. N°ordre : 6349

Water erosion is responsible for the spread of radioactive contaminant through the agricultural catchment area, its redistribution, possible concentration and export towards the aquatic environment, and also for the post-accident contamination of plants through splash of water. The aim of this work is to offer or improve values for the proper functioning of codes of calculation destined for operational use, These codes of calculation do not take into account the spread of contamination after its release.

Modelling these transfers requires initially the acquisition of the capacity for spread of the contamination at decimetric scale. To this end, we have reproduced deposits of dry and wet radioactive aerosols of cesium and strontium at various stages in the hydric and structural history of the soil in function of a range of degrees of slope of the terrain. The exchanges of cesium and strontium at the water-soil interface have been quantified by means of millimetric experimentation in batchs, during which the various phases of the mobilisation of soil particles have been reproduced on calibrated blocks of aggregates.

At the end of this study, it was apparent that the cesium migrates horizontally at an appreciably faster rate and in greater proportions than the strontium that fixes little on the particles and infiltrates into the first centimeter of the soil. The key parameter to take into account for the model would appear to be the initial humidity of the soil. In addition, the migration of the cesium appears to be particularly sensitive to the development of sedimentary crusts. Where the soil is dry or slaking crusted at the time of contamination, the first run?offs are very efficient carriers of contamination. The sedimentary crust favours the exportation of cesium and strontium by preventing its infiltration, while at the same time trapping part of the contamination on its immobile particles. On steep slopes, the structure of the sedimentary crusts is less continuous, they are less extensive, the macro?aggregates then persist at the surface ; by inhibiting run?off, they favour the dispersal of the contamination, In the case of wet discharge, the contamination of cesium tends to disperse.

An operational expert study model should require a minimum of prior knowledge of the catchment area, the time of calculation should be reasonably short and the spatial resolution of the results is not necessarily of great interest. By applying coefficients of solid-liquid distribution of contaminants to the results of models of erosion such as WEPP, the results of transport in the run-off water may prove satisfactory. In contrast, for the development of a research model, a more conceptual model with a finer time scale and better taking into account of surface obstacles would no doubt provide a basis for better anticipating possible concentrations of contamination in the zones of deposit in the catchment area. An illustration of this type of model is the model LISEM.