Atmosphere, Upper

An experimental investigation of the abundance of upper atmospheric sodium at Boca Raton, Florida, using high resolution absorption spectroscopy, is described. The discussion is based primarily on data taken from January to October, 1968, but W. Lammer's data of September through November, 1967, is used for the seasonal analysis. A description of the data and the method of analysing it are presented, with emphasis on the theory underlying a computer method of analysis.
Measurements indicate a seasonal variation of about a factor of two, with a sharp drop off in late February, 1968. A maximum average monthly abundance of 5.4 ± 2.0 x 10^9 atoms/cm^2 in a vertical column in October, 1967, and a minimum average of 2.6 ± 0.6 x 10 atoms/cm^2 in March, 1968, was measured. The average for the twelve months beginning November, 1967, was 3.6 x 10^9 atoms/cm^2 . In general, uncertainties of the order of 50% are present in the measured abundance determined by a single recorder trace. This value is generally smaller for those traces which are less noisy, and which show a larger amount of terrestrial sodium absorption. No correlation between the abundance and hurricanes has been established. If any correlation exists, it probably is smaller than the uncertainty of our measurements.

An experimental investigation is described in
which a Pepsios high resolution pressure scanning s pectrometer
is used to measure the absorption of light due to
atomic sodium in the upper atmospher e. Design and construction
of the spectrometer will be discussed with attention
given to unique features of the instrument which were
develo ped in this study. An attempt is made to see if any
dependence on the abundance of the sodium can be accounted
for by the proximity to the ocean or by the winds and
extreme turbulence developed by tropical storms and hurricanes.
It is hoped that this data will compliment the
existing data found at other latitudes so that a circulation
dependence may be found.

Measurements of the rate of change of twilight glow were made at a fixed
zenith angle with a Pepsios spectrometer at Boca Raton, Florida and
Fritz Peak, Colorado in 1973. The twilight derivative may be represented
by the convolution of a vertical sodium distribution function and the
derivative of an atmospheric transmission function. Computer calculations
of the convolution were made with artificial sodium density distributions
for comparison with observed intensity changes. Na layer was found to be 88 +- 6 km for the Florida data and 89 +- 8 km for the Colorado data. The computed distributions were similar to those
found by other observers. The average half width from Florida was found
to be 3 km greater than that from Colorado. The 3 km difference was
about two times the standard deviation and it may reflect either a seasonal
or geographic effect.