3 to October 1952 for any United States stations except Salt Lake City, However, the values reported may be taken as a satisfactory approximation of the total gamma dose delivered to date. TL.fallout prior to October 1952 was of a low order compared with that during the sub- sequent period. As expected, the high accumulations of mixed fission products (Fig. 1) exist in the vicinity of the proving grounds in Nevada and in the vicinitv of the Pacific test area. The highest recorded deposition, Grand Junction, Colo., is 740 milli- curies per square mile as of | January 1956. The values are lower everywhere else, the next highest deposition being 290 millicuries per square mile at Iwo Jima in the Western Pacific Ocean. The lowest value is Lagos, Nigeria, with 33 milli- curies per square mile. In general, the mixed fission products data are characterized by a high degree of regional uniform- ity in which, with few exceptions, adjacentstations agree within + 50 percent. The estimates of strontium-90 are shown in Tables | and 2. Outside of the United States, the observed range is from 1.8 millicuries per square mile in Peru to 24 millicuries per square mile at Iwo Jima. In the United States, the accumulation varies from 2.1 at San Francisco to 23 at Salt Lake City. For most of the world covered by this network, the range of 2 to 5 millicuries per square mile may be taken as a representative estimate of the distribution of strontium-90. With respect to the gamma dose, the average value for the United States is higher thanit is for the rest of the world. The range of values in the United States is relatively narrow, 6 to 49 millirads, except for Salt Lake City (160), Grand Junction (120), and Albuquerque, N.M. (110). The representative dose for east- ern United States is about 15 to 20 milli- rads, with slightly higher values in the Middle West and lower values on the West Coast. The cumulative gamma dose at the foreign stations is in the range of 4 to 23 millirads, except for some of the Pacific Table 2. Fallout at stations within continental United States, October 1951 to September 1955. . Station Mixed fission proce) Sr* (me/mi’) Albuquerque, N.M. Atlanta, Ga. Billings, Mont. Binghamton, N.Y. 400 120 160 61 20 3.8 5.7 2.2 110 16 24 7.8 Chicago, Il. 140 5.3 2+ Boise, Idaho Dailas, Tex. Des Moines, lowa 160 170 170 4.2 18 3.3 8.4 4.9 3.6 21 120 13 49 18 17 New York, N.Y. Philadelphia, Pa. Pittsburgh, Pa. Rapid City, $.D. 110 110 100 150 4.2 4.6 4.1 6.1 17 16 13 25 St. Louis, Mo. Salt Lake City, Utah San Francisco, Calif. 200 680 47 6.0 23 2.1 27 {60 5.8 Scottsbluff, Neb. Washington, D.C. 200 86 6.3 3.0 26 il New Orleans, La. Rochester, N.Y. Seattle, Wash. 170 for collection of total fallout activity is taken to be 63 percent, and all of the data have been corrected accordingly. A source of error which is more diffhcult to assess is that inherent in the assumption that the decay of the radioactivity is proportional to ¢-1-?, Application of this decay law requires that the age of the debris be known with some cer- tainty, and this has become increasingly difficult during the past 2 years, Prior to 1954, it was possible to predict the decay characteristics of a sample accurately because at any given time the debris was known to haveoriginated from the most recent test series. After each series, the daily fallout would diminish rapidly and would ordinarily be unde- 1952, the yields from detonations were relatively low, and the bulk of the debris where fallout is greatly hastened by precipitation and other factors, In contrast, the detonation of devices having yields noted previously, the over-all efficiency stable region of the earth’s atmosphere, efficiency of the tance in interpreting reported values. As 13 11 the period October 1951 to September 1955. gummed film is of fundamental impor- . collection 3.5 27 * Gammadose for the period October 1952 to Seprember 1955, except for Salt Lake City, which covers equivalent to megatons of TNT produces clouds of radioactive debris which pierce the tropopause and becomedistributed in The 3.7 89 was distributed below the tropopause, Discussion 5.7 99 was deliberately conservative. tion was intended to provide a tentative upper limit of the estimated dose and 29 28 140 740 88 200 130 110 These gamma values are somewhat lower than the average estimate of 100 milliroentgens for the United States as reported by Dunning (5) whose calcula- millirads. 6.1 6.2 16 Detroit, Mich. Grand Junction, Colo. Jacksonville, Fla. Memphis, Tenn. Minneapolis, Minn. New Haven, Conn. tectable before the next series of tests started. The rapid diminution in fallout from tests conducted prior to the spring of 1954 can be explained by the fact that. except for Operation Ivy in November islands, where the rangeis from 13 to 150 9.2 Gammadose* (mrad) the stratosphere. From this relatively the particles descend slowly, and fallout to the earth’s surface occurs over a period of time which is measured in years rather than weeks or months. Thetraces of rela- tively dld debris from high-yield devices become mixed with the debris of subse- quent detonations. This being the case, neither the decay characteristics of a sample nor the relative abundance of the long-lived isotopes can be predicted from theory. In recent months, the procedure has been changed and, where necessary,either individual or pooled samplesare followed for decay. There is no practical way in which the accuracy of the estimates of accumulated mixed fission products can be tested experimentally. However, it is possible to test the validity of the strontium-90esti- mates by direct radiochemical analyses of the soils from thevicinity of the sampling stations. In October 1955, immediately following the sampling period covered by this report, soil samples were collected from 17 widely scattered locations in the United States. The relationships between the predicted and measured values are given in Figure 2. It is seen that the data consist of two groupings, in which [4 points show a reasonably satisfactory re- gression of measured on predicted values. The correlation coefficient for these data is 0.70. The otherthree points forgations near the Nevadatest site show low ratios