Sodium lidar measurements obtained between December 1980 and May 1986 at Urbana, Illinois, are compared with other reported lidar measurements and with recently developed models of the layer. Sodium abundance at Urbana reaches a maximum value in November, December and January which is approximately 4.5 times larger than the June minimum value of 2.15 x 109 cm -2. Also in November and December, the layer centroid height is about 1.5 km lower than the yearly average value of 92 km. There is no significant seasonal variation of the rms layer width whi•ch• has an average value of approximately 4.25 km. The peak-to-peak variations of the centroid height and rms width of the average nocturnal layer are 1 km and 600 m, respectively. The semidiurnal tide is primarily responsible for a 36% peak-to-peak abundance variation in the average nocturnal layer which is compatible with a 20 cm s-x amplitude for the vertical wind velocity. The vertical phase velocity of the density perturbations is approximately 1 m s-x which implies a 45 km vertical wavelength for the semidiurnal tide. Because of gravity waves and tides the structure of the nocturnal layer changes substantially throughout the night. Variations of over 200% in abundance, 2 km in centroid height, and 1 km in rms width have been observed within the time span of a few hours. 109-10 •ø m -3. Before lidar systems became available, sodium measurements were largely restricted to studying resonantly scattered sunlight. Groundbased observations of this type were able to define seasonal variations in column abundance [Blamont and Donahue, 1961; Gadsden and Purdy, 1970], but the sharp layer boundaries were not revealed until rocketborne dayglow measurements were made [Hunten and Wallace, 1967]. Lidar observations of the vertical structure of the layer were first made in England [Bowman et al., 1969]. Since then similar measurements have been reported from a variety of locations including France [Me•lie and Blamont, 1977], Brazil [Simonich et al., 1979], Japan [Aru•la et al., 1974], Illinois [Richter et al., 1981], California [Hake et al., 1972] and at the high latitudes of Franz Joseph Land, USSR [Juramy et al., 1981] and Andoya, Norway [Fricke and yon Zahn, 1985]. During the past few years numerous chemical and dynamical processes have been proposed in an attempt to explain the general characteristics of the seasonal, diurnal and geographical variations of the layer structure. Sodium chloride from the oceans and volcanic eruptions have been discussed as possible sodium sources. However, meteoric ablation is generally regarded as the dominant source of all the alkali metal layers including sodium [Clemesha et al., 1978; Richter and Sechrist, 1979a, b; Je•lou et al., 1985a, b]. Much of the intrinsic character of the layer is apparently governed by a complex chemistry involving many reactions among numerous neutral and ionic species [e.g., Sze et al., 1982; Kirchhoff, 1983; Jegou, 1985a]. The dominant loss mechanisms are believed to include the conversion of neutral sodium to t...

During the early morning of January 16, 1989 the Arecibo 430 MHz incoherent scatter radar and the University of Illinois CEDAR Na lidar were operated simultaneously for several hours at the Arecibo Observatory, Puerto Rico (18øN, 67øW). At approximately 0425 LST a sporadic E (Es) layer began forming near 101 km altitude and moved steadily downward reaching an altitude of 99 km by 0500 LST. The electron abundance of the Es layer increased rapidly from approximately 3x109 cm -2 at 0510 LST to approximately 6x109 cm -2 by 0530 LST. During this same time period a sporadic Na (Na s) layer formed at the same altitude as the Es layer. Both the Es and Na s layers moved upwards together between 0530 and 0600 LST before dissipating at an altitude of 101 km. The high correlation between the altitude and abundance variations of the two layers strongly suggests that the Na s layer formation is related to the Es layer and may involve Na cluster ion chemistry. 104, 1988. von Zahn, U., P. von der Gathen, and G. Hansen, Forced release of sodium from upper atmospheric dust particles, Geophys. Res.

This paper reviews the ground‐based and rocket techniques that are presently being used to determine electron density profiles in the ionospheric D region. Ground‐based techniques include VLF, LF, and MF sounding, differential absorption and differential phase using partial reflections, wave interaction, and incoherent scatter. Rocket techniques include differential absorption and Faraday rotation in association with high‐resolution dc probes calibrated by means of the radio measurements. The characteristics of the aforementioned techniques are presented, including time and height resolution, accuracy estimates, preferred height ranges, and problems encountered. Electron density profiles obtained with these techniques are presented for comparable solar zenith angles and undisturbed solar and geophysical conditions, and variances are described.

Lidar observations of the mesospheric sodium layer often reveal wavelike features moving through the layer. It is often assumed that these features are a layer density response to gravity waves. Chiu and Ching [1978] described the approximate form of the linear response of atmospheric layers to gravity waves. In this paper, their results are used to predict the response of the sodium layer to gravity waves. These simulations are compared with experimental observations and a good correlation is found between the two. Because of the thickness of the sodium layer and the density gradients found in it, a linear model of the layer response is not always adequate to describe gravity wave ‐ sodium layer interactions. Inclusion of nonlinearities in the layer response is briefly discussed. Experimental data is seen to contain features consistent with the predicted nonlinearities.

The photodetachment rates of several negative ions believed to be present in the D region are estimated for sunrise conditions. The photodetachment cross sections of O− and O2− which have been measured in the visible and infrared wavelength ranges are used. Since the photodetachment cross sections of NO2− and NO3− are not yet known, the behavior of their rates could only be estimated. Infrared band absorption by H2O and CO2 is included in the calculations. For atoms and molecules with large electron affinities the ozone ultraviolet absorption is important. The effects of photodetachment on D‐region behavior are discussed and found to be relatively unimportant.