The material left in the troposphere is thought to remain 3 -.: up to 40 days and to circle the earth a few times before reaching ground level. It deposits :~ -elatively narrow bands, centered on the detonation latitude, with little evidence of diffusion acroas the stable air barrier located in the troposphere north of the equator. It is probably brought down largely by the scavenging effect of rainfall or other precipitation (Reference 24). Those particles which do not fall out within the first few weeks will remain suspended in the atmosphere for a prolonged period, which is frequently described by the term “half-residence time.” This is the time during which the amount of material 30 suspended will be depleted by one-half. The halt-residence times for the stratosphere vary from 6 months to 5 years depend- ing on the latitude and altitude of injection. Polar shots like those of the USSR in October 1958 gave about a 6-month half-residence time. The equatorial shots similar to those of Hardtack, which stabilized in the lower stratosphere, have a half-residence time of about 1 year. Clouds that stabilize in the higher stratosphere like those from Shot Bravo during Operation Castile and Shot Orange during Operation Hardtack may have a half-residence time of up to 5 years. The particle size of the material in the stratosphere is extremely small, much of it being less than 0.1 micron (Reference 25). It ia distributed by the stratospheric winds in the east-west or west-east direction, and there is also thought to be a slow circulation toward the poles. Movement into the troposphere can take place by slow settling or by seasonal changes in the altitude of the tropopause. The exchange may be most prevalent at the break in the tropopause near the middle latitudes. Once transfer from the stratosphere is completed, the material will be deposited relatively quickly in the same manner ag intermediate fallout (Reference 24). 1.2.4 Procedures for the Determination of Fallout Partition. The hazards of nuclear testing are associated primarily with worldwide fallout, inasmuch as local fallout can be controlled by selection of the test site and the proper winds aloft so that its area of deposition will be of minor consequence to the population of the world. However, local fallout has regional ecological consequences that are not negligible. It may spread over considerable areas of as much as a million square miles (Reference 26). Introduction of radionuclides, such as sr™ into the human environment via worldwide fallout has a potential effect on the whole population, and the significance of such nuclides has been studied in great detail (Reference 27). These studies led to the conclusion that certain radionuclide levels at the earth’s surface can be tolerated and that these levels can be maintained within acceptable limits by restrictions on the rate of nuclear testing. This is based on the concept that a condition of equilibrium is reached in the strato- sphere at which the rate of injection of radioactive debris will be equal to the decay plus deposition rate. The fraction of the device appearing in glooal fallout has usuaily been estimated indirectly by measuring the fallout in the local area and subtracting from unity. The methods used for the determination of local fallout have involved measurement of gamma ray field contours or representative sampling of the material arriving at the surface of the earth (References 28 and 29). The total amount of radioactive debris in the fallout area may be calculated if the relation be- tween dose rate and surface density of radioactive material is known. Similarly, samples representing a known area of the fallout field may be analyzed for amount of weapon debris, and all Such areas summedto give the total local fallout. A combination of fallout sampling and analysis plus gamma radiation measurements has also been used (Reference 29). These procedures are subject to a number of difficulties and uncertainties, not only with regard to making adequate sample collections and radiation fleld measurements but also in data interpretation. The establishment of accurate gamma contours requires an extensive and costly field program, because radiation intensity measurements must be made over areas up to tens of thousands of square miles. When the fallout is deposited mainly over the surface of the ocean, the original patterns are distorted continuously by settling of the particles and by ocean currents. The collection of samples at the earth’s surface, which are truly representative of the area sampled and free from coilector bias, presents problems that have not been fully solved to date. 15