as a barrier, there were no reliable indisstions of the height of super tropopauas cloud to consider as significant. Further, ever if the ebeve factors had been imown, the fact that the caly available long-range information on land surface turete as lintted to one low yield shot, left mich ta be desired in assigning reascnable mubers to isodove lines derived from the system. Consequently, ag 8 general statement, it should be noted that, although this system of fall-cut forecasting wae used of BRAVO to auguent other dats, it was presented and Limited to discussions of the relative merits ofthe aseumptions and sealing upon which it was based. There wee no real basis for assuming it wes valid and as & Consequence confidence could not be pushed to the point of owersriding the many other factors involved in the shot decision. As the operstion progressed, and the cbeerved fall-cat effects provided ome mecsure of answers to the nany questions involved, the syrtenwas relied upon to a mahgreater extant. The major aodification in the use of the method ws to assign Liniting altitudes af .the eloud to use in applying the sethod. ‘th general, the post= shot evidence of the finely divided sercecl-type Gloudfrom the barge shots indicated @ probable significant’ eat-off height at the bese of the high easterlies (i.e. approxinately tropopause height, or about 60,000 feet). Tt appeared that the tropopeuse trapping astion probably has a si gificant effect on very mali particles. Gn theother hand, it appeared from ERwVO that cloud heights for land surface burste should be considered up to at Least 60,000 feet and possibly to 80,000 feet. The fact that the ARVO fireball iteelf went up to the tropopause say sceount ‘for the difference. Many particlee could possibly have been formed at altitude as well as 58