and to the meteorological problem of stratospheric transport, diffusion and
removal.
be solved.
The basic problems of particle size and filter efficiency have to
It may be that an indirect method,
such as
comparison with air-
eraft or whole air collection, will be the only way to provide a solution to
the calibration of the balloon sampling.
In that event, a number of compar ‘sons
at various altitudes will be necessary to establish this calibration.
It will
also be necessary to perform these comparisons for each of the isotopes of
interest.
If the mechanical problems related to the collection of debris can be
solved, the question of the quality of the radiochemical analyses becomes
even more important.
The large unexplained fluctuations in the ratios of the
various isotopes suggests that much of the variability arises from this source.
If it turns out that the variations observed in the past data are indeed real,
then the four-station network sampling each altitude once a month now in
existence would be inadequate.
However, information presently available con-
cerning the stratosphere suggests that the variations are improbable, especially
when viewed in light of the estimate that most of the debris injected into the
stratosphere before 1958,
came from the Castle series in the Spring of 195),
three to four years prior to the measurements disuussed here.
It is also possible that the relatively low stratospheric content indicated by the balloon data, and the decrease between 1957 and 1958,
represents
the best estimate, and that the models of injection and stratospher‘c-tropospheric
exchange that have been employed are grossly in error.
This would imply a much
0
vy
faster exchange than has been postulated or smaller initial injections or both.