STATUS OF GLOBAL FALLOUT PREDICTIONS 371 However, this simple picture may not always be found. The upper and lower parts of Fig. 1 illustrate two exceptions. In the upper part the ground-level 185W air concentrations measured by the Naval Research Laboratory (NRL) along the 80th meridian are plotted schematically, with some liberties being taken with the observed data because of dif- ferences in station elevation.’ The peak concentration is displaced to the subtropical latitudes of each hemisphere. Machta’ has argued that bomb debris placed initially in the upper equatorial troposphere will be carried poleward by the return poleward circulation of the Hadley Cell to subtropical latitudes. The lower part of Fig. 1 shows fallout at stations of the Health and Safety Laboratory (HASL) pot and ion-exchange network and along the NRL 80°W meridian network during the U.S.S.R. 1961 nuclear test series. The peak deposited ®*Sr radioactivity occurs 20 to 30 degrees south of the source, and the air-concentration maxi- mum is over 40 degrees south of the source, In part, this pattern may result from limited or poor sampling, especially near the pole, More likely it is real and results from meteorological processes. Material injected into the upper polar troposphere will mix downward along isentropic surfaces to lower levels in temperature latitudes in days or weeks, Thus debris is more readily availablefor precipitation scavenging and final mixing to the ground temperature latitudes than in polar latitudes for sources originally in the upper polar troposphere, Current Forecasting The specific trajectory, spread of debris, and precipitation- scavenging areas must be knownto predict radioactive fallout after an atmospheric injection. It is believed that “true” trajectory predictions are useful to two days and that after-the-fact reconstructed trajectories are likely to be useful to perhaps five or ten days. Beyond these periods the average error may be too large for the results to contribute to the understanding of the fallout distribution. The past operational forecasts by the Weather Bureau were based on subjectively predicted or reconstructed constant-level or constant- pressure trajectories. A swath rather than a line was used to incor- porate both the diffusive spread and the uncertainties of meteorology. An example of this kind of an after-the-fact forecast is given in Fig. 2 along with ground-level verification over the United States and Canada. In this case, the 30,000-ft, or 300-mb, level was chosen becauSe it was believed that most of the tropospheric fission products from the test lay near the tropopause. Rain areas through which the nuclear cloud might pass were also given with the forecast or reconstructed trajec- tories since it is in these regions where greater deposition should be expected. The shaded areas delineate first arrival of the debris in ground- level air at the station shown by the dots. The verification shows a de-