Search

An analysis of density measurements of Ar, N2, O, and He made at 280 km with the gas analyzer aboard the polar-orbiting satellite Esro 4 has yielded a global model of the variations in temperature, density, and composition that occur in the disturbed thermosphere. In the model the increase of temperature over quiet conditions is a nonlinear function of the planetary geomagnetic index, its latitude profile being approximated by a fourth-power sin phi law, where phi is the 'invariant' magnetic latitude. A density wave proceeding from high latitudes is approximated by a fourth power cos phi law. A strong nonlinearity in the relation between the temperature variations and the variations in the height of the homopause explains a previously found behavioral difference in the variation of atomic oxygen during magnetic storms and during periods of sustained geomagnetic activity.

The seasonal-latitudinal and the diurnal variations of composition observed by mass spectrometers on the OGO 6 satellite are represented by two simple empirical formulae, each of which uses only one numerical parameter. The formulae are of a very general nature and predict the behavior of these variations at all heights and for all levels of solar activity; they yield a satisfactory representation of the corresponding variations in total density as derived from satellite drag. It is suggested that a seasonal variation of hydrogen might explain the abnormally low hydrogen densities at high northern latitudes in July 1964.

Densities of N2, O, Ar, and He at a height of 280 km, obtained from gas analyzer measurements aboard Esro-4, were analyzed to establish a global pattern for variations that are associated with geomagnetically induced disturbances. It is shown that during longer periods of sustained geomagnetic activity, thermospheric temperature increases with geomagnetic latitude, reaching a maximum in the region of the magnetic pole, while during short-lived magnetic storms, the maximum temperature may be reached in the auroral zones. The observed variations of composition in middle and high latitudes can be explained by assuming that temperature variations are accompanied by variations in the height of the homopause. In the equatorial region the four gases studied vary in phase with similar amplitudes during a magnetic storm, indicating a density wave proceeding from higher latitudes.

The drag analysis of low-perigee satellites such as Injun 3 and Explorer 17 at a time of low solar activity has revealed appreciable fluctuations in upper-atmosphere temperature during geomagnetically undisturbed intervals. These temperature variations are very nearly proportional to the variations in the Kp index, in contrast with the temperature variations during magnetic storms, which show proportionality with the ap index. When Kp

The density variations that accompany geomagnetic disturbances have been studied by analyzing the drag of three satellites with high orbital inclination (Injun 3, Explorer 19, and Explorer 24) and one with moderate inclination (Explorer 17). The average time delay between the peak of the geomagnetic perturbation and that of the atmosphere is 6.7±0.3 hours. While there seems to be no significant dependence on the time delay on the intensity of the perturbation and on the geographic location with respect to the sun, there appears to be some dependence on latitude. For latitudes greater than 55° (average: 65°) the mean time delay is 5.8±0.5 hours, and for latitudes smaller than 55° (average: 25°) it is 7.2±0.3 hours. All three high-inclination satellites give consistently smaller delay times at high latitudes. The observed density changes are interpreted as being caused by changes in temperature. For smaller perturbations (Kp < 5) the temperature T shows a nearly linear dependence on Kp, and for latitudes lower than 55° the rate of change ΔT/ΔKp is about 28°. For latitudes above 55° (average: 65°) ΔT/ΔKp seems to be about 15–25% greater. For more intense disturbances (Kp ≥ 5), ΔT/ΔKp is systematically larger, confirming the nonlinearity of the relation between T and Kp, when considered over its total range; there is also a good indication that some atmospheric perturbations are enhanced in the auroral zones more than others.

Past attempts to represent the semiannual density variation in the heterosphere as a consequence of temperature variation have run into difficulties in two height regions: below 200 and above 1000 km. The main argument in favor of the temperature variations was the dependence of their amplitude on the solar cycle; it can be shown, however, that this dependence is spurious, being caused by the variation of the density change dρ/dT with the temperature T. An analysis of the semiannual density variations at different height levels fails to show a dependence of the amplitude with the sunspot cycle. All difficulties are removed if we assume that the semiannual density variation is not a direct consequence of temperature variations.

The analysis of satellite accelerations leads to an empirical formula that relates the product ρH½ (ρ = atmospheric density, H = scale height) to the geometric height z, the 20-cm solar flux F20, and the angular distance ψ′ from the center of the diurnal bulge. Once the numerical parameters of this formula have been established, tables of ρ and H are computed and a separate formula is derived to represent ρ in function of the same variables.

The ratio r = Tmax/Tmin of the maximum to the minimum exospheric temperature considered on a global scale was derived by means of static models using the densities obtained from the atmospheric drag of three low-inclination satellites. It is found that r varies with the solar cycle, lagging more than a year behind the variations of the 10.7-cm solar flux; on the other hand, the variations of r are in phase with K¯p, which is obtained by averaging the planetary geomagnetic index Kp over the whole year. The obvious conclusion is that, in the mechanism of the diurnal temperature variation in the thermosphere, the solar wind may be at least as important as, if not more important than, solar EUV. No appreciable lag is found in the variation of the minimum nighttime temperature when compared with the variations of the 10.7-cm solar flux.

A search for lunar tides in densities derived from satellite drag has yielded spurious oscillations with the period of one lunar day, caused by the '27-day' variation connected with the solar rotation. The real tides remain undetected, with an amplitude smaller than 1 or 2%.

Ondoh and Hakura [1964] have pointed out that if solar plasma velocities are computed for the time interval between a flare and the corresponding geomagnetic disturbance, and then linearly correlated to the geomagnetic indices Kp and Ap, the resulting regression lines do not agree in slope with the corresponding regression lines derived from solar wind data supplied by Mariner 2 [Snyder et al., 1963]. The Mariner 2 velocities refer mostly to quiet or mildly disturbed days (0 < Kp < 6), whereas the velocities determined by Ondoh and Hakura come from magnetic storms (3 < Kp < 9). If Kp is taken to be the independent variable, the Mariner (‘wind’) slope is smaller than the storm slope; if Ap is the independent variable, the situation is reversed. Also, the Mariner velocities correlate better with Kp, and the storm velocities show a better correlation with Ap.

Upper atmosphere heating in auroral zones during geomagnetic perturbations - analysis of atmospheric drag data from injun iii satellite