from the rising cloud when their fall rate is equal to the rise rate of the cloud. Results from the small barge are not inconsistent with this assumption, since, during the interval from H+ 0.5 to H+1 minute, cloud rise rates of 1,700 to 454 feet per minute were measured for the upper and lower parts of the cloud. © This is not the case for the YCV barge at Station S-6, however. For the interval be- tween 0.1 and 0.6 minute, the lowest rise rate associated with the visible part of the cloud was 570 ft/min, which is over twice the value of fay: Still, the large difference between the amountsof activity collected by the two adjacent air samplers suggests that a large error could be associated with the fall-rate calculation, large enough even to accommodate the above assumption. It must be concluded that these air sampler results are too uncertain for anything more than speculation. 3.10 FALLOUT COMPARISONS In the first chapter, estimates were made of fallout to be expected from a surface burst by scaling results from other shots. Figure 3.14 compares these estimates with estimates based on measurements from Shot Fig. Values from Fig fall considerably below estimates from higher yields. The scaling techniques of Reference 1, which depend on assumptions enumerated in Section 1.2.2, lead to gross overestimates of downwind extent for H+1 hour dose rates of 100 r/hr and lower. Also, the percentage of total activity deposited within the Fig array is estimated by this source to be 12 percent as compared to the 4 percent estimate based on measurements. Results from the 1.2-kt surface detonation of Operation Jangle form the basis for estimates in Reference 1, together with the so-called cube-root scaling procedure. It is interesting to speculate on the reasons for failure of the scaling method to predict better the fallout from Shot Fig. Cloud dimensions for the two events, Jangle Surface and Fig, did not scale as the cube root of yield. The ratio of cloud heights, for example, was about 2 to 1, This would cause cube~-root scaling to provide an overestimate of intensities. Also, in the case of Shot Fig, the cloud drifted across the entire instrumentation array before stabilization occurred. Since the amount of time required for cloud stabilization is independentof yield, drift during cloud rise could cause failure of scaling methods that do not account for it, especially if, as in this case, scaling is to be made for distances that are not large comparedto the drift of the cloud during stabilization. To gain a better indication of the significant differences in fallout from Fig and Jangle Surface, a comparison was made of the percentage of activity deposited within comparable distances for the two shots, accounting for actual cloud heights and drift of the cloud during its rise. d = _ Riturp i The downwind distance d of a particle is the vector sum of the distance covered during rise and the distance covered during fall as follows: (3.1) Where: Ug = effective mean wind velocity during particle rise uFo effective mean wind velocity during particle fall h = altitude where the particle with fall rate f starts free fall at time t Since cloud rise is not constant, Up is not necessarily equal to Up . 44

Select target paragraph3