Spatial Distribution of Caesium-137

Abstract

ConclusionIt has been shown that 137Cs inventories have a coefficient of variation of approximately 20% at reference sites (Pennock, 2000), and that this variability is probably because of random spatial variability (Owens and Walling, 1996). At reference sites it is recommended that a grid-sampling network is employed, with the collection of three replicate cores within 1 m of each sampling point. The cores may be bulked for 137Cs analysis (Owens and Walling, 1996). While the number of samples needed to estimate the reference value within the required statistical limits will depend on the variability found during analysis, between fifteen and thirty samples will usually be necessary for an accurate estimate of central tendency (Pennock, 2000).Owens and Walling (1996), suggest that Eq. 6.2 is used to determine a “reference range” rather than a single value, which might imply a greater accuracy than is really the case. This range can then be used to determine points of net soil loss and gain within the landscape.Sampling for spatial variability of 137Cs may take the form of transects or, preferably, a grid (Chapter 2). There has been no standardization of sampling frequency, as Tables 6.2 and 6.3 reveal. Rather, each study’s methodology has been established according to the study aims, the perceived field-variability of the isotope (e.g. has cultivation mixed the soil, or is micro-topography a factor in the redistribution of 137Cs?), and the number of samples that can be analysed in the laboratory. It is, however, recommended that at least twenty samples be taken along transects or in a grid pattern. Samples can be taken by core to a depth that will incorporate the total 137Cs profile, previously determined by depth sampling, although the deposition of sediment particularly at the slope-base may over-thicken the profile in areas affected by water erosion. Excavation, soil auger or other suitable means should be used to sample such sites to greater depths (Chapter 3).It is recommended that at least some basic soil properties be measured (e.g. texture, organic matter, bulk density and horizon/plough layer depth) to aid 137Cs interpretation (Chapter 2). Topographic characteristics can be used to group landform elements, and land-use types can be grouped for 137Cs comparisons. The selection of appropriate models to transform 137Cs data into estimates of soil losses and gains that can be related to controlling factors such as soil erodibility, slope characteristics, and land use and management are discussed in Chapter 7.