ABSTRACT
The time series of the 10-day average (137)Cs volumetric activity concentration in the lower atmosphere measured from 1987 to 1991 in the town of Pripyat, close to the Chernobyl nuclear power plant, was used to construct a model to predict the airborne activity concentration inside the 30-km exclusion zone. For that purpose, individual components of the observed time series were separated by regression analysis and the Group Method of Data Handling. The measured data in Pripyat were divided in two periods. The long-term prediction by the model established using the measured data of the first period, has been validated with the data in the second period with good agreement. The behaviour of the model parameters depending on the length of the periods was also analysed, and the first period of 4.5 y was shown as sufficient for estimating the parameters. Further increase in the length will not significantly enhance the model parameters and the predictive power.
