CHAPTER 6 FEASIBILITY OF DEVELOPING FALLOUT-PREDICTION TECHNIQUES FOR OPERATIONAL APPLICATION A number of fallout-prediction systems have been developed in response to a variety of needs. These are widely used in damage assessment and training exercises, in scientific, engineering, and military studies, and in the prediction of fallout during the conduct of nuclear tests. The general aspects of such prediction systems is that they predict fallout patterns before the fact, using assumed or known yields, heights of burst, locations, and winds. None of them use reports of fallout intensity as a basis for fallout predictions at locations further downwind. Civil defense can use existing prediction systems only in planning and not operationally, since these systems require the inherently unknowable details of the enemy's plans for attack, and their accuracy is limited by uncertainties in weather parameters. What civil-defense authorities can hope to do operationally during and after the attack is to give the best possible advice to the population on where the fallout is, where it will go, when it will get there and at what levels, and where to move to--if that is a viable alternative. The kind of prediction system needed to do these things is quite different from existing systems. In an effort to meet this need, a monitoring and prediction method based on observation of the unfolding fallout event was developed and tested by the Research Directorate of DCPA in the undocumented RESEX I exercise. The method utilized available weather data and techniques to predict the fallout sector once the location and general magnitude of detonation were established. Information on eertain fallout parameters-time when the exposure rate became 0.5 R/hr, time of peak exposure rate, time when exposure rate exceeded or decreased to 50 R/hr, etc,-- were reported by operating areas to higher headquarters (county, state, and regional EOC's) where the data were plotted and extrapolated in time and distance to provide warning and the same fallout parameters for locations farther downwind. The existence of the RESEX 1 exercise shows that a real-time extrapolative prediction of fallout is to some extent feasible. However, there is a question whether such a system could be made to work in the attack situation, what with its critical dependence on the ability to receive data from the field and to disseminate information back. The questions have not been resolved to the Subcommittee's satisfaction. Nevertheless, it is self-evident that a system using current and real data is preferable to before-the-fact prediction. a | 4 we ws bs >? an) Ri } m4 re Fs | wag Ps = pra 23