Modeling radiocesium fixation in upland organic soils of northwest England

Two models of time-dependent Cs-137 adsorption and fixation by soil were fitted to solution activity data from incubated organic upland soils, over a period of 709 days. Both model fits were highly (p < 0.01) or very highly (p < 0.001) significant. Both models differentiated between specifically adsorbed and nonspecifically adsorbed 'labile' radiocesium. Transfer to a nonlabile pool was either by a kinetic step (model I) or a diffusive process (model II). Although fits for model II were comparatively worse in some soils, the second model had the dual advantage of being more mechanistic in structure and requiring fewer parameters. Parameters estimated for models I and II compared favorably with those derived experimentally; some dependence on soil mica was evident. Model I was used to predict the change in bioavailability of radiocesium in soils following documented inputs from Windscale (1957), fallout from weapons testing (1960s), and the Chernobyl release (1986). For all but one of the five organic soils, long-term predictions of relative bioavailability generated by model I were similar to the limited range of published values for Cs-137 uptake by vegetation. It is concluded that the Cs fixation models presented account for a substantial part of the change in availability of Cs-137 observed in field studies.