The sum of the nuclide fractions from the cloud and fallout #hould be | in eack case, pro-
vided that the R-values used are representative of the cloud and fallout ag 2 whole. This seems
to be likely for the fallout where the R-values change only relatively slightly with time but more
doubtful in the cloud as a result of the scatter of the analytical results.
Table 3.9 gives a
comparison between the deposited fractions (from Table 3.5) and airborne fractions (from
Tables 3.2 and 3.8). The agreement is generally aa good as could be expected, considering
the nature of the data.
In Shot Koa, the gaa sample data is very meager. The gas and particulate samples are not
matched well in time and altitude. It is belleved that the Mo" fractions, and consequently the
Sr® and Cs" fractions, as calculated from the Sr’? and Cs" in the cloud and fallout are better
values than those calculated from Kr**,
For Shot Walnut, the late fallout results are limited and not interpretable in obtaining the
fraction airborne; hence, only the gas sample data has been used. This fallout data also leads
to unreasonably large fractions deposited.
In Shot Oak, both fallout and gas samples gave similar values for the fractions deposited and
airborne.
The averages have been used.
3.2 DATA RELIABILITY
3.2.1 Cross-Contamination of Koa Samples. As discussed in Section 2.3.2, a preliminary
examination of the samples from Shot Koa, shortly after their receipt at LASL, indicated that
they might be badly contaminated with debris from Shot Fir. If this were the case, the fission
ratios from the Koa cloud data could not be used for the determination of fallout partition, because they would not be representative of the detonation. To investigate the extent of crosscontamination, the Koa samples were analyzed
Table 3.10 gives a summary of the results of this work.
It is evident
that the Koa samples contained at most a little over 1 percent of material from the Fir cloud,
and generally much less. Hence, the quantities of molybdenum and krypton introduced into the
Koa cloud from Fir were small enough go that they would have a negligible effect
¢
on the fission
ratios.
3.2.2 Accuracy of Radiochemistry. Radionuclide analyses on the particle samples were
accurate to 5 percent on a relative basis, and the gas counting had an accuracy better than
10 percent.
3.2.3 Reliability of Sampling. Certain points on the curves of Figure 3.1 are to be attributed
somewhat less significance than the others because of uncertainties regarding the samples.
On Koa, the fission ratio for Sample 981 R may be off by a factor of 2 as a result of the smail
sample size and high counter background from fallout, which would decrease the counting ac-
curacy. On Walnut, Sample 978 L (27.5 hour) the probe velocity was low, and Kr*® only was
determined. (Probe velocity refers to the pumping speed in the gas particle coincident sampler.) Sample 980 & for Oak has been disregarded because of the very low probe velocity,
which would tend to make the Mo" to Kr® ratio too high.
3.2.4 Particle Fail Rates and Specific Activities. The particle size distributions (and hence
the specific activity as a function of particle size) could have been altered tn a number of ways
before the fall rate studies were made. Among these are breakup of particles by impaction on
the filter, losé of fine particlea in handling, spontaneous breakup of particles in the fatlout process itself due to atmospheric moisture (see Appendix C regarding the behavior of particles in
liquids), and severai other possible means of alteration.
It is possible to calculate what fall rate a particle would need to fall 59,000 feet {n four hours,
i.e., to be collected in Koa Massive Ll. This fall rate is 125 cm/sec. The diameter of a
40