1.8.8 Aircraft Sampling of Clouds. A condition necessary ior use of the gas-particle sam- pling technique for the determination of device partition ls that the samples be collected from a region that is losing material by fallout but not receiving particles from any other section of the cloud. The portions of the cloud that are suitable for this type of sampling are dependent on the wind structure existing at the time of burst. For one type of structure that occurs fairly . frequently at EPG, the top and bottom parts of the cloud are blown off rapidly in different directions, leaving a layer approximately 1 mile thick that experiences only light and variable winds. Hence this stratum, which is located between 50,000 and 60,000 feet, will soon be isolated from the rest of the cloud and may remain fairly stationary above ground zero for a day or more. It is called the light dnd variable wind layer and is satisfactory for coincident sampling, be~ cause it can aot receive fallout from higher cloud levels. In cases where the stratum is not well defined, sample collections can be made from the top of the cloud (provided it can be reached and followed by the sampling aircraft) or from a location selected to minimize the feed-in of falfout from higher altitudes. The theory of this technique has been discussed under Section 1.2.4, and the sampling equip- ment is described in Chapter 2. The operation plan was to fly through the light and variable layer at several intervals between H+2 and H+ 24 hours with B-57D aircraft, equipped both with the coincident samplers and with wing tank particle collectors. The coincident samples were to be analyzed for Kr"*® and Mo”to determine the fallout partition (Section 1.2.4), and the wing tank samples for 10 radionuclides to investigate fractionation with particle size. 1.3.4 Aircraft Sampling of Fallout. The fallout sampling part of the program was intended to provide information supplementary to that obtained from the rocket and aircraft cloudsampling experiments. WB-50 aircraft were scheduled to fly at an altitude of 1,000 feet aad to coliect fallout at various times between H+4 and H +24 hours along height lines that would correspond to the cloud level (about 55,000 feet) sampled by the B-57D’s. Because the cloud is an extended sourceof fallout, the term “height-line sampling, ” as used here, signifies the sampling of a band of material centered on the geometrical height line and having a bandwidth approximately equal to the diameter of the cloud. The wind structure described in the preceding section on the formation of the light and vari- able layer also leads to isolation of the 55,000-foot height line along the eastern periphery of the fallout curtain. This situation is advantageous for height-line sampling, because the air- craft may proceed westward from a position east of the fallout area and collect the first fallout encountered. The samples should contain 55,000-foot fallout alone, uncontaminated by ma- terial from the rest of the cloud. Other types of wind structure would probably not be as favorable for height-line sampling, and the fallout collected likely would contain particles originating from different levels in the cloud. Outward from ground zero along a height line, the particle size of the fallout decreases and the time of arrival increases. However, low-altitude sampling at a given location should pro- vide a sample containing particles of relatively uniform size (used synonomously with falling rate). Hence, by making a series of collections along a height line at different distances from the shot point, advantage can be taken of particle size separation by natural fallout processes. The WB-50 operations were arranged to utilize this situation to obtain a set of samples suitable for an investigationof size-dependent properties. It was planned to use the radiochemical data from these samples to corroborate the composition of local fallout as determined from the rocket experiments, to investigate fractionation with particle size, and to compare the composition of local fallout with worldwide fallout. The data can also be used for determination of device partition Lf the fallout is shown to be highly depleted in a particular fission product. The enrichment of the debris remaining aloft in this fission product will then be related to the fraction of the debris that hase fallen out, in much the same way as has already been described for interpretation of the enrichment of a gaseous fission product in the cloud with respect to particulate debris. 20