378 MACHTA Inventory Any prediction of stratospheric fallout requires quantitative data on the initial distribution of radioactivity in the stratosphere. Sources of data, analytical techniques, and errors in the inventories have been re- ported elsewhere in detail.® Briefly, between the tropopause and about 70,000 ft, aircraft sampling sponsored by the Defense Atomic Support Agency (DASA) and the U. S. Atomic Energy Commission (AEC) provides the data; between 65,000 and 105,000 ft (occasionally to 120,000 ft) the AEC’s balloon sampling program supplies the information. The radionuclides *’Sr and 4c are the two which are adequately measured in the atmosphere to obtain inventories. Results of measurements of these nuclides reported by Machta, List, and Telegadas,® together with a few more recent values, are given in Tables 1 and 2. Note that of the items, the oceanic uptake values of C and the local fallout of Sr have been estimated rather than observed and that the value for the deposited Sr has been obtained from the AEC—USDA soil-sampling program and the HASL pot and ion-exchange network, A random error of the total inventory of either the “Sr or '‘C of the order of 40% at the 90% confidence level is quoted. Unfortunately, there is evidence of anunknown systematic error possible due to greater fallout of **Sr over the oceans than over land. The tables also list, in the last lines, the amounts of "Sr and 4C calculated to have been produced from nuclear testing on the basis of the announced fission and fusion yields, Although the conversion from energy yield to amount of Sr and, particularly, of Mo may be questionable, it is of interest to note that the observed inventories fall slightly short of those computed. Part of the difference might be due to the presence of Sr and “C above 100,000 ft, and part of the *Sr dis- crepancy might be due to the larger oceanic than land fallout. Since the differences between observed and computed inventories are far less than a factor of 2 (indeed, in January 1964 it was only 25%), the two techniques reenforce confidence in the total accountable amountof these nuclides. History of Forecasting Methods Soon after Operation Castle at the Pacific Proving Grounds in 1954, it was realized that the continued worldwide fallout could be accounted for best by assuming a stratospheric reservoir of Sr with a residence time measured in years. Later, aircraft and balloon measurements confirmed the presence of a large “Sr and “C content in the lower strato- sphere. Libby!’ was the first to propose a rational forecasting technique for predicting the rate of exit of this debris. Libby considered that a Simple exponential removal rate could explain the observed fallout data and be used to predict future fallout. The assumption of a simple ex-