Assessment of spatial distribution of fallout radionuclides through geostatistics concept

摘要

After introducing geostatistics concept and its utility in environmental science and especially in Fallout Radionuclide (FRN) spatialisation, a case study for cesium-137 (~(137)Cs) redistribution at the field scale using geostatistics is presented. On a Canadian agricultural field, geostatistics coupled with a Geographic Information System (GIS) was used to test three different techniques of interpolation [Ordinary Kriging (OK), Inverse Distance Weighting power one (IDW1) and two (IDW2)] to create a ~(137)Cs map and to establish a radioisotope budget. Following the optimization of variographic parameters, an experimental semivariogram was developed to determine the spatial dependence of ~(137)Cs. It was adjusted to a spherical isotropic model with a range of 30 m and a very small nugget effect. This ~(137)Cs semivariogram showed a good autocorrelation (R~2 = 0.91) and was well structured ('nugget-to-sill' ratio of 4%). It also revealed that the sampling strategy was adequate to reveal the spatial correlation of ~(137)Cs. The spatial redistribution of ~(137)Cs was estimated by Ordinary Kriging and IDW to produce contour maps. A radioisotope budget was established for the 2.16 ha agricultural field under investigation. It was estimated that around 2 × 10~7 Bq of ~(137)Cs were missing (around 30% of the total initial fallout) and were exported by physical processes (runoff and erosion processes) from the area under investigation. The cross-validation analysis showed that in the case of spatially structured data, OK is a better interpolation method than IDW1 or IDW2 for the assessment of potential radioactive contamination and/or pollution.