was expelled at 260,000 feet and had Largefall races in che firse several
thousand feet, wita che resule chat horizontal speeds can be shown to be less
reliadle in this cone thao in otters.
Direction should be accurate excepr ia
Cases of extreme fall rates.
(3) Most observations were terainated between 220,000 and 230,000 feet;
the longest fell co 211,000 feet.
Below 230,000 feet, radar operators aoted
an apparent rapid dispersion of the chaff target.
This accelerated dispersion
may result from strong turbulence and vertical shears in this zone, or it is
csncaivable chet it is a relatively constant function of time after release.
{&) Altbougb it is possible to prove the existence of 12+ and 24-hour
periodicities in these measurements, their correlation with atmospheric tides
was difficult to substantiate because of data scarcity.
This lack of obdser-
wations precluded use of haraonic analysis which would have allowed a direct
detersination of phase and amplitude of possible oscillations.
Inoatead, a
sioe function of two harmonics was fitted to these data; its significance
was proven by the F test, and these computed pericficities were then conpared
with semidiurmal theoretical cidal motions as provided by Stolov.'3
The
west-to-east components (u) and the south-to-north components (v) of the
winds were thus fitted at 240,000, 250,000, and 260,000 feet.
Conparison
with theoretical computations indicated an average phase ditference near 50
degrees and an amplitude ratio of approximately two to ome, with the fitted
curves larger.
Since theory is not well verified at lower latitudes, this
agreement is considered good.
Also, theory omits lunar influence, which
tends to decrease the difierence in phase by an unknown amount.
Justification
for belief that theoretical anplicudes should be greater is found in recent
observations in equatorial regions which show larger tidal effects thao anticipated in the E and F layer virtual height variations. !*
(5) A plot of altitude versus average u and v components for the Mé-hour
periods begianing and ending at 0600, local standard cise, is given in Fiz.
5.19.
Here all observations within either lé-hour period sere sucmed to ob-
tain the average.
This plot indicates a saall but apparently definite de-
crease nocturcall,y in speeds and the eagerness of the cross-latitudiunal
flow ve.
Finally. short period variations: of less thas @ cers sppesr ts $s
small on the basis of the three observations of July 31.