STATUS OF GLOBAL FALLOUT PREDICTIONS 385 Table 4—STRATOSPHERIC Sr COMPARED WITH NORTHERN-HEMISPHERE FALLOUT, MEGACURIES year Predicted Observed MON W © RPreoco eo January 1959 May 1960 November 1960 January 1962 January 1963 January 1964 Fomwmocnoeo +I oo =) at 30 to 90°N during following NHOSS below 16.8 km (55,000 ft) Fallout deposited m WO Stratospheric content Nm Oo — from the tropopause to 55,000 ft and 30 to 90°N. Except for the year following May 1960 or the year following November 1960, the past verification suggests that the northern-hemisphere deposition as measured by the HASL pot and ion-exchange collection network will be slightly more than the “Sr content of the polar lower stratospheric box. There is no explanation for the failure of the method for the year following May 1960 and the year following November 1960. The last line shows the prediction submitted to the Federal Radiation Council in mid-1964 for the 1964 fallout prediction. The previously mentioned DASA prediction method agrees with this forecast, Empirical models similar to those just described leave much to be desired in terms of the confidence of the forecast. The models fail to explain why the prediction should be correct; rather they depend on past experience for justification. However, the source configuration rarely duplicates itself, and there is evidence that year to year differences in removal rates exist, It is for these reasons that many groups have been working on realistic models of stratospheric transport. Theoretical Models Meteorological models of stratospheric transport may arise from many considerations. First and foremost, they may be started on the basis of certain preconceived rates of vertical and horizontal mixing and organized meridional circulations with the objective of reproducing the patterns of the tracer experiments such as the radiotungsten or the ozone distributions. A second and more fundamental approachis to allow the diffusive and meridional circulation to develop from the heating and cooling of the atmosphere with appropriate boundary conditions in a computer model. Both kinds of considerations are being actively pursued by various investigators. The Air Resources Laboratory of the Weather Bureau in collabora- tion with Ben Davidson of New York University and the group at Iso-