aa the tropopause were discounted as a barrier, no reliable indications Bf the Significance of the height of the above-tropopause cloud existed. Fukther- more, even if these factors had been known, the only available long-rknge fallout data on land-surface bursts were limited to a single low-yielBR shot (SUGAR of the 1951 JANGLE series), which meant that the assignment off reasonable values to isodose lines derived from the prediction system wag difficult and possibly inaccurate. both Consequently, although the elliptiical approximations system of fallout forecasting was used on BRAVO to augment Other data, it was limited to discussions of the relative merits of the assumptions and scaling upon which it was based. Since there was no feal basis for assuming the system was valid, it did not override the manyjfother factors involved in shot decisions. Nevertheless, as the tests progrBssed and observed fallout effects answered some of the many questions involved, the system was relied upon to a much greater extent (Reference 16, pp 57 and 58). The elliptical approximations method, or any other extension of Te surface radex area beyond about 6 to 12 hours, was also limited by be ng based on the ground-zero winds. A new method was needed to accommodafe the changes in wind systems as particles drifted farther from ground fero. nies Thus, during CASTLE a new technique was developed that utilized airbofne Particle trajectory analyses and progressive forecasts of wind patter s in accordance with time and displacement aspects of the cloud. Practica] methods were devised to apply such a system to the last three CASTLE #hots for a valid forecast period of H-hour to H+24 (Reference 16, Tab D). Weather Because fallout forecasts depend primarily on forecasts of wind ffelds, the accuracy of the CASTLE fallout predictions was related to the acciracy of weather predictions. CJUTF 7 in Operation Plan 3-53 (Reference 9) ds- signed responsibility to CTG 7.4 to administer and logistically suppogt weather elements to provide meteorological data and to perform radsaf missions required to conduct CASTLE operations. TG 7.4 was responsib]e for a four-island ground weather reporting element, an aircraft weathdr 112