Assessment of radiation-induced cancer risks from the Chernobyl fallout in Finland

A. Servomaa,T. Komppa,M. Suomela

Abstract:Application of detailed radiation risk models to populations affected by radiation doses from the Chernobyl fallout allows forecasting and estimation of the consequences of the accident in countries far from the place of the accident, and comparison of the model estimates with epidemiological observations in low-dose conditions among large populations. Both tasks need timedependent estimates of the radiation doses caused by the fallout, including future doses, and both have also severe problems and statistical limitations. Problems in the forecasting originate from the large uncertainties in the risk estimates, especially in the details and in the basic assumptions included in the models. In a large population, the predictions may show considerable numbers of exposureinduced cancers; the numbers may still remain under the statistically significant levels if they are a small fraction of all cancers in the population. One may ask whether it is too confusing to present such uncertain estimates of non-significant size. On the other hand, similar estimates are implicitly applied and generally used in radiation protection, and, even with the limitations, the model predictions may be the only piece of information available in advance while potential latent cancers are developing among the population during tens of years.