A probabilistic evaluation of fallout effects associated with nuclear air bursts

摘要

Over the past quarter of a century a great deal of theoretical and empirical knowledge concerning atmospheric processes has been developed. Many models attempting to predict atmospheric dispersion effects have been devised, but few have endured, and thus far none has achieved universal applicability, although some have enjoyed widespread popularity. Perhaps among the reasons why most models were or are of limited operational utility is that they failed to adequately incorporate the fact that meteorological and geophysical phenomena are inherently stochastic in nature, and the behavior of such phenomena can be predicted only in statistical terms. Until recently, atmospheric transport models have tended to be deterministic in nature, using estimates of "average" or "worst case" values for computational elements, and have provided for little measure of the probability of occurrence of the predictions. The approach summarized in this paper describes an attempt to remedy that shortcoming, in that itexplicitly recognizes the inherent variability of these natural phenomena and attempts to incorporate that variability into a scheme which explicitly quantitates the uncertainty of prediction. It attempts to go a bit further than other approaches to evaluating environmental effects, in that it not only predicts effects but also allows evaluation of the likelihood of occurrence of the predictions.