Statistical characteristics of the spatial distribution of territorial contamination by radionuclides from the chernobyl accident

Abstract

Random (statistical) processes played a significant role in the formation of the spatial distribution of radioactive contaminations owing to the Chernobyl accident, both at the stage of fallout and subsequent migration of radionuclides. In particular, a detailed study of the distribution of the surface density of soil contamination, for example, performed by mass -r-ray spectrometric analysis of soil samples obtained on the territory of a specific inhabited area, should reveal fluctuations resulting from the nonuniformity of th e precipitation produced in the fallout by medium-scale - hundreds of meters atmospheric turbulence. For brevity, we shall term statistical phenomena with this scale intermediate statistics. When the territory is subdivided to dimensions of up to 10 m or smaller, local fluctuations of the contamination density, which are associated with the heterogeneity of radionuclide migration (primarily owing to the local microrelief) and which we term local statistics, should be expected. In the radiochernical determination of the content of 9°Sr and plutonium isotopes a small portion (several percent) of the sample is, as a rule, analyzed, and this results in the appearance of an additional random component. Our objective in the present work is to clarify the following interrelated questions: the character of the statistical distributions determined by local and intermediate statistics; the number of samples which must be analyzed in order to determine with prescribed accuracy the surface density (activity excess) of radioactive contamination over some arbitrary area as a function of the dimensions of the area; and, the degree to which all existing information on soil contamination in the inhabited areas of Bryanskaya oblast [1] (chosen because the maximum possible amount of information is available for this region) satisfy statistical laws and to determine the parameters and structure of these laws. We note that statistical fluctuations of the density of radioactive fallout density with a spatial scale of approximately several kilometers was observed back in 1957 after the accident at Windscale (Great Britain) - they were observed in the determination of the 131I concentration in milk [2]. Among the studies of the consequences of the Chernobyl accident, we single out [3], where the distribution of the 137Cs contamination density was estimated for separate inhabited areas and regions in the territory of Russia and Belorus. In [3] a significant spread was noted in the contamination density within specific inhabited areas and it was shown that the 137Cs contamination density is described satisfactorily by a log-normal distribution. In [3] the spread in the values of the activity of separate soil samples is associated to the heterogeneity of atmospheric precipitations. In order to reveal the effect of the local statistics, in 1991-1992 we collected soil samples on 8 × 8 m sections with a step not exceeding 2 m on four specially selected test areas (level sections, which were known not to be farmed after the Chernobyl accident and having I37Cs surface contamination of 1110-1850 kBq/m 2 or 30-50 Ci/km2), situated on the territory of the Narovlyansk region of the Gomel' oblast of Belorus and the Novozybkov and Zlynkov regions of Bryanskaya oblast in Russia. In order to specify the pattern within these sections additional samples were taken near points with maximum activity on2 × 2m squares. The soil samples consisted of 140 mm in diameter and 50 mm high cylindrical core samples. The 137Cs content was measured using both a precise spectrometer with a Ge(Li) semiconductor detector and a scintillation spectrometer with a NaI(T1) crystal. It follows from Fig. 1 that the activities of the samples differ by a factor of 3-5. In order to estimate the effect of local nonuniformities in the distribution of radioactive contamination on the results