Mabit L, Bernard C
Soil Science Unit, FAO/IAEA Agriculture and Biotechnology Laboratory, International Atomic Energy Laboratories Seibersdorf, A-2444 Seibersdorf, Austria.
J Environ Radioact. 2007;97(2-3):206-19. doi: 10.1016/j.jenvrad.2007.05.008. Epub 2007 Jul 30.
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 x 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.
在介绍了地质统计学的概念及其在环境科学，特别是在沉降放射性核素（FRN）空间化中的应用后，本文给出了一个在田间尺度上使用地质统计学研究铯 - 137（(137)Cs）再分布的案例研究。在加拿大的一块农田上，地质统计学与地理信息系统（GIS）相结合，用于测试三种不同的插值技术[普通克里金法（OK）、一次反距离加权法（IDW1）和二次反距离加权法（IDW2）]，以创建(137)Cs地图并建立放射性同位素预算。在优化变差函数参数后，绘制了实验半变异函数图以确定(137)Cs的空间依赖性。它被调整为一个范围为30米且块金效应非常小的球形各向同性模型。该(137)Cs半变异函数图显示出良好的自相关性（R(2)=0.91）且结构良好（“块金值与基台值”比率为4%）。它还表明采样策略足以揭示(137)Cs的空间相关性。通过普通克里金法和反距离加权法估计(137)Cs的空间再分布，以生成等高线图。为所研究的2.16公顷农田建立了放射性同位素预算。据估计，约2×10(7)贝克勒尔的(137)Cs缺失（约占初始总沉降量的30%），并通过物理过程（径流和侵蚀过程）从研究区域输出。交叉验证分析表明，在空间结构化数据的情况下，对于评估潜在放射性污染和/或污染，普通克里金法是比IDW1或IDW2更好的插值方法。
