Papadakos G N, Karangelos D J, Petropoulos N P, Anagnostakis M J, Hinis E P, Simopoulos S E
Nuclear Engineering Laboratory, National Technical University of Athens, 15780 Athens, Greece.
Nuclear Engineering Laboratory, National Technical University of Athens, 15780 Athens, Greece.
J Environ Radioact. 2017 May;171:234-245. doi: 10.1016/j.jenvrad.2017.02.009. Epub 2017 Mar 10.
Within the presented study, soil samples were collected in year 2007 at 20 different locations of the Greek terrain, both from the surface and also from depths down to 26 cm. Sampling locations were selected primarily from areas where high levels of Cs deposition after the Chernobyl accident had already been identified by the Nuclear Engineering Laboratory of the National Technical University of Athens during and after the year of 1986. At one location of relatively higher deposition, soil core samples were collected following a 60 m by 60 m Cartesian grid with a 20 m node-to-node distance. Single or pair core samples were also collected from the remaining 19 locations. Sample measurements and analysis were used to estimate Cs inventory and the corresponding depth migration, twenty years after the deposition on Greek terrain. Based on these data, the uncertainty components of the whole sampling-to-results procedure were investigated. A cause-and-effect assessment process was used to apply the law of error propagation and demonstrate that the dominating significant component of the combined uncertainty is that due to the spatial variability of the contemporary (2007) Cs inventory. A secondary, yet also significant component was identified to be the activity measurement process itself. Other less-significant uncertainty parameters were sampling methods, the variation in the soil field density with depth and the preparation of samples for measurement. The sampling grid experiment allowed for the quantitative evaluation of the uncertainty due to spatial variability, also by the assistance of the semivariance analysis. Denser, optimized grid could return more accurate values for this component but with a significantly elevated laboratory cost, in terms of both, human and material resources. Using the hereby collected data and for the case of a single core soil sampling using a well-defined sampling methodology quality assurance, the uncertainty component due to spatial variability was evaluated to about 19% for the Cs inventory and up to 34% for the Cs penetration depth. Based on the presented results and also on related literature, it is argued that such high uncertainties should be anticipated for single core samplings conducted using similar methodology and employed as Cs inventory and penetration depth estimators.
在本研究中，2007年在希腊境内20个不同地点采集了土壤样本，包括地表和深度达26厘米的土壤。采样地点主要选自在1986年及之后雅典国立技术大学核工程实验室已确定切尔诺贝利事故后铯沉积量较高的区域。在一个沉积量相对较高的地点，按照60米×60米的笛卡尔网格、节点间距20米采集土壤岩芯样本。还从其余19个地点采集了单个或成对的岩芯样本。利用样本测量和分析来估算希腊境内沉积20年后的铯存量及相应的深度迁移情况。基于这些数据，研究了整个从采样到得出结果过程中的不确定度分量。采用因果评估过程来应用误差传播定律，并证明合成不确定度中占主导地位的重要分量是由于当代（2007年）铯存量的空间变异性所致。确定的第二个虽不那么重要但也很显著的分量是活度测量过程本身。其他不太重要的不确定度参数包括采样方法、土壤现场密度随深度的变化以及样本测量准备情况。采样网格实验借助半方差分析，还能对因空间变异性导致的不确定度进行定量评估。更密集、优化的网格能为该分量返回更准确的值，但在人力和物力方面实验室成本会显著增加。利用此处收集的数据以及采用明确采样方法质量保证的单芯土壤采样情况，对于铯存量，因空间变异性导致的不确定度分量评估约为19%，对于铯穿透深度则高达34%。基于所呈现的结果以及相关文献，有人认为，对于采用类似方法进行的单芯采样并用作铯存量和穿透深度估算器的情况，应预料到会有如此高的不确定度。
