(9) Source sise, IVY MIKE indicated that the arrent sethods of predicting fall-out on the assumption of a point source (or at best, 5 very small source) were not acceptable for high ylelds. Consequently, CASTLE forecasting was originally made on the sasumption.cf 9 circular source approximately 1§ miles in radius (i.e. the approximate radius of fall-cut crosavwind and updnd from MIKE). BRAVO results indicated thie to be somewad small, and probably more like an eilipee with a semi-eminor axis of about 25 miles. For the barge shots, the L5-aile eireular radius (actually nore elliptieal than circular) appeared to be adequate. Tienet effeet of the above was to consider the surface RiDEX constructed om a point source type analysis as a generating elawnt. Shifting the point sarce RADEX arcund a source area selected in accordance vith yield end type of burst generated the complete area which shouldbe considered in the surface RiIDEX. 4 similar adjustment wes made to the Air RADEX forecaste with excellent results in defining the areas of contamination waich resulted. f. Construction of RADEXES and Fall-cut Plots, Generally speaking, the RADEX ares (RiMiclogical BXelusion ares) ws considered to be &@ limited fall-out area for the first six hours post-shet in the vicinity of the teat aite amt for the primary use of test personnel and equipment. Area fall-cut plots were conaideryd ta be the entire area of significant fall-out, to’ inelude infinity isodcee lines of at least 50 roentgens, and in seme cases dow to 10. rcentgens. The RADEX was used es a operational devies to define dangercus areas and to deny entry of task force units into Certain areas exsept under spesifie suthorisation, 2 a consequence, it owe way given wide dissemination throughout the task forces for the information