From analysis of the orbiter atmospheric drag (OAD) data obtained from the orbital decay of the Pioneer Venus orbiter from December 9, 1978, to August 7, 1979, atmospheric densities have been determined and tabulated near 16°N latitude between 140 and 190 km for all times of day. Maximum daytime densities on Venus are approximately 8 × 10−13 g cm−3 at an altitude of 150 km (dropping by a factor of 24 at night) and 7 × 10−14 g cm−3 at 170 km (dropping by a factor of 82 at night). Comparative atmospheric densities on earth at 150 km are a factor of 3.5 higher during the day with only a 1% diurnal variation. An atmospheric composition, temperature, and density model based on OAD vertical structure measurements is presented. The inferred exospheric temperatures and atomic oxygen concentrations are surprisingly insensitive to model assumptions. This model indicates that atomic oxygen is the major species in the Venus atmosphere above 145 km at night and above 160 km during the day with mixing ratios near 140 km in excess of 0.1. Atomic oxygen concentrations determined from drag measurements near 170 km range from 1 × 109 cm−3 during the day to 3 × 107 cm−3 at night. Exospheric temperatures inferred from OAD measurements are found to vary from 100°K at night to approximately 300°K during the day. The very low exospheric temperatures discovered at night, lower than temperatures near 100 km, are inconsistent with the concept of a planetary thermosphere, in which temperature increases with altitude. We have adopted the term ‘cryosphere’ to define this region of cooling where temperature decreases with altitude. The phase of the diurnal temperature variation is consistent with theoretical models of an atmosphere with mean motion to the west more rapid than planetary rotation superimposed on solar‐antisolar motion. The very low nighttime temperatures and our observation of an absence of the predicted nighttime atomic oxygen bulge suggest substantial downward eddy transport of heat as well as of atomic oxygen. Finally, the OAD data indicate that the neutral upper atmosphere of Venus is unexpectedly insensitive to both solar extreme ultraviolet variations and variations in the solar wind. Thus, the Venus upper atmosphere may not be controlled by the major heating mechanisms generally assumed for planetary thermospheres.

Analysis of global ozone variations for the period April 1970 to December 1975 was performed by using the reprocessed Nimbus 4 backscattered ultraviolet (BUV) measurements of total ozone. A correlation coefficient of 0.97 is found between the 6‐month running mean of global mean total ozone (filtered for mean semiannual, annual, and quasibiennial variations) and the 10.7‐cm solar activity index. Correcting ozone for a time‐dependent latitudinal bias relative to Dobson ozone measurements reduces to 2–3% the global mean ozone variation over the solar cycle. The solar ultraviolet variability required in a one‐dimensional time‐dependent radiative photochemical model to account for the observed ozone variation appears to be consistent with recent solar UV observations.

A gas beam consisting of a mixture of atomic and molecular oxygen has been directed at the ion source of a mass spectrometer like those used in sounding rockets for determining the neutral composition of the lower thermosphere. The loss of atomic oxygen on mass spectrometer surfaces was evaluated by flagging the beam in several ways and comparing the experimental results with predicted values. The results obtained suggest that in rocket flights using similar instruments the atomic oxygen densities computed assuming no-loss conditions may be low by a factor of 2.5. Studies made using a beam containing tracer indicate that carbon dioxide observed when atomic oxygen enters the source is formed in a reaction involving atomic oxygen from the beam and carbon monoxide from the surfaces bombarded. Although atomic oxygen is one of the mostimportant constituents of the lower thermosphere, there is little agreement as to its absolute abundance. Recently reported values for any given altitude vary by a factor of 5 or. more. Since the loss of atomic oxygen in conventional mass spectrometer ion sources is not known, uncorrected densities are reported and are believed to be low [Hedin and Nier, 1966; Kasprzak e.t al., 1968; Krankowsky et al., 1968; Mauersberger .et al., 1968; Gross et al., 1968; Bitterberg et al., 1970; Hickman. and Nier, 1972]. Densities deduced from optical absorption measurements [Hall et al., 1963, 1965, 1967] tend to be higher, as are those reported from incoherent scatter measurements [Bauer et al., 1970; Cox and Evans, 1970; Giraud et al., 1972]. Von Zahn [1970], in a comparison of atmospheric densities at 150 km as determined by mass spectrometer composit. ion measurements and by satellite drag measurements, concluded that the values could be reconciled ß if, along with several other adjustments, one assumed that the mass spectrometrically found atomic oxygen results were low by a substantial factor, presumably owing to the loss of atomic oxygen in mass spectrometer ion sources. This view seems to be supported by the results of O•er-Copyright (•) 1973 by the American Geopkysical Union. mann and yon Zahn [1971] who, with a heliumcooled mass spectrometer ion source, obtained an appreciably higher value for the n(O)/n(02) ratio at 120 km than had previously been reported. On the other hand, in a direct comparison between density scale heights below 180 km as found by mass spectrometry and by drag measurements with the low-altitude satellites OV1 15 [Champion et al., 1970a] and OV1 16 [Cham.pio• et al., 1970b], Nier [1972] con-

Although the assumption of complete and immediate recombination of atmospheric atomic oxygen at the surfaces of the open ion source of the AEROS‐A mass spectrometer is reasonable below 300 km, application of this assumption at higher altitudes results in a substantial underestimation of the ambient density of atomic oxygen. A two layer surface model seems necessary to explain the observations. Parameters for several surface reaction processes are estimated.

No abstract