624
VOLCHOK
effect that the meteorological factors have on the relative air concentrations. It is suggested that,whether or not atmospheric testing is be-
ing carried out, the main peaks and valleys of relative air concentrations will otcur. We hope to substantiate this conclusion by looking at
data from past years and at future collections.
To simplify examination of these results for annual oscillations,
Figs. 7 and 8 were prepared. In these the relative values for the two
years were
averaged;
all areas under the curves that had values
greater than unity were shaded. In Fig. 7 no obvious symmetry can be
seen between the hemispheres. It was found by trial and error that by
offsetting the hemispheres by four months the best symmetry (under
the limitations of these data) is achieved. This is illustrated in Fig. 8.
The first appearance of substantially higher than average air concentrations is in January or February in the northern hemisphere and in
September or October in the south. The southern hemisphere seems to
exhibit a geographically broader but lower intensity peaking in this
initial period. In this and also in the next period, it is interesting to
note that the highest values are at fairly low latitudes, probably be-
tween the equator and 20°. In the final two periods shown in the lower
third of the figure, the hemispheres are remarkably symmetrical, with
high relative activities at latitudes above 20°.
Since this method of analysis eliminates the effects of the absolute
differences of the source concentrations, it would appear that the particular hemispheric characteristics represent true meteorological differences. Both hemispheres exhibit annual fluctuations with higher than
average relative concentrations at some latitudes for two-thirds of the
year. The initial appearance of the high is at low latitudes followed by
increased values and a poleward shift. The asymmetry between the
hemispheres occurs mostly in the first half of the season, whereas the
southern hemisphere values do not vary to as great a degree as do
those of the northern hemisphere.
DEPENDENCE OF AIR CONCENTRATIONS UPON RAINFALL
It has been suggested that precipitation characteristics may play
an important role in determining air concentrations of "Sr. We at-
tempted to study this by utilizing the rain data from the worldwide
HASL pot and column network, These data were averaged over 10° latitudes bands and normalized in the same manner as the Sr values, i.e.,
bimonthly rainfall amounts are expressed as fractions of the annual
bimonthly average.
The best fit with the relative Sr data was found by offsetting the
precipitation one month, i.e., by comparing August and Septemberrain
concentrations with September and October air concentrations. The re-
sult of this comparison for the northern hemisphere is shown in Fig. 9.