Collective models in environmental sciences

The results of models for predicting the migration of toxic substances through the environment are, generally, affected with large uncertainties that are due to the complexity of the modelled environmental systems and to the difficulty of obtaining reliable values of the model parameters. Surprisingly, some models for predicting the migration of radioactive substances in water bodies, despite the complexity of the involved processes, are characterised by small levels of uncertainty. In the present paper, it was demonstrated that some of the parameters used in such sort of models show small variability, which is due to the statistical or ‘collective’ effects caused by the mutual compensation of some processes in the environment. As a consequence, the values of the ‘collective’ parameters vary within small ranges despite the wide variability of the environmental circumstances. This occurrence implies that collective models show relatively small levels of uncertainty.