Simultaneous measurement of ion and neutral motions by radar and optical techniques

The results of simultaneous thermospheric neutral wind and ionization drift measurements from near College, Alaska (L = 5.6, A = 65 ø) are presented. The neutral wind data were obtained by observing the Doppler shift of the 6300 A atomic oxygen line with the 15-cm Fabry-Perot interferometer of the Michigan Airglow Observatory which is located temporarily at Ester Dome, Alaska. Ionization drifts were measured by the Chatanika incoherent scatter radar facility. These simultaneous measurements indicate that in the premidnight sector both the neutral wind and the ionization drift are generally westward. This westward ionization drift is consistent with the general magnetospheric convection pattern but the measured neutral wind is in a direction opposite to the diurnal pressure gradients and thus must be driven by ion drag. In the postmidnight sector the ionization drift tums eastward while the neutral wind direction tums south. Again, the ion drift is consistent with previously published results; the reasons for the absence of significant zonal neutral winds and the significant southward meridional wind in the postmidnight sector are not well understeod at this time, but are probably a combination of a decrease in the ion drag force following magnetic midnight, Coriolis force, and pressure gradients due to both the diurnal and auroral heat sources. calculations based


INTRODUCTION
During the last few years it has become clear that upper atmospheric motions play an important role in determining the structure, chemistry, and energetics of the thermosphere. Rishbeth [1972] has recently reviewed and summarized our present understanding of these thermospheric winds. Theoretical calculations of the global neutral wind systems have most often been based on solutions of the neutral gas momentum equation, using pressure gradients obtained from models [e.g., Geisler, 1967;Kohl and King, 1967;Challinor, 1970.; Blum and Harris, 1973]. The equation of motion for the neutral gas involves the ion velocity; thus, more recent calculations have attempted to solve the simultaneous momentum equations for both the neutrals and the ions. The ion velocity is strongly influenced by the presence Copyright ¸ 1974 by the American Geophysical Union. of electric fields; however, until very recently little was known about these fields so they were ignored in most 'calculations. Assuming typical convection electric field values, Fedder and Banks [1972] and Banks [1972] calculated the influence of electric fields, perpendicular to the magnetic field lines, on the neutral winds. They neglected pressure gradients and their results are probably only representative of the initial response of the atmosphere to an applied electric field. Nevertheless, their work [also see Cole, 1971] clearly indicates that thermospheric winds are greatly influenced by electric fields of magnetospheric origin; therefore, realistic wind calculations must include these effects. Recent experimental data [Hays and Roble, 1971;Rees, 1971;Meriwether et al., 1973] indicate that the actual thermospheric wind system is not described by presently published simplified calculations; the major discrepancies are related to the direction and magnitude of the wind at middle 315 and high latitudes (e.g., measured winds blow to the west in the evenings, which is opposite to the direction obtained from calculations based on model pressure gradients).
The purpose of this brief paper is to present some results of simultaneous measurements o.f ionization drifts and of neutral winds in the thermosphere. By measuring these quantities simultaneously in the same locality we are attempting to determine the factors which control the respective motions. Accordingly, our discussion focuses attention on the gross features of the results, with detailed discussions to follow in a subsequent paper.

RESULTS AND METHOD OF MEASUREMENTS
The neutral wind measurements were made by observing the Doppler shift of the 6300 A line of atomic oxygen with the six-in. Fabry-Perot interferometer of the Michigan Airglow Observatory, which is located temporarily at Ester Dome, near College, Alaska (L = 5.6, A --65ø). The basic facility [Roble, 1969;Hays et al., 1969] and the general observing technique for midlatitudes have been described previously [Hays and Roble, 1971]. The method of scanning the line profile with the interferometer was changed in 1972 to a discrete wavelength (density) scan [McWa'tters et al., 1973] to provide improved data quality and time resolution. Details of the modified system and data reduction method will be described in a later paper; it will suffice for the present to point out that typically it took about ten minutes to measure the wind component in a given direction. Additional explanation of the data-taking procedure accompanies the results to follow.
Motions of the ionospheric plasma were measured by the radar incoherent scatter technique. Underlying theory and general nature of this operation have been reviewed by Evans [1969] and Evans [1972]. The observations reported here were carried out with the Chatanika Radar Facility located near College, Alaska. Both the facility [Leadabrand et al., 1972] and the measurement technique [Doupnik et al., 1972;Banks et al., 1973] have been described previously in the literature. From Figure 8, data from March 22, 1973, one sees the neutral wind vector following a regular rotation pattern, westward before magnetic midnight, then to the south. But the pattern of observed ionization drifts was not that commonly recorded; the westward drifts in the premidnight sector intensified around midnight rather than reversing. It should be noted that a 1000 ¾ negative bay was observed at 0100 UT on the College, Alaska magnetogram H trace.

DISCUSSIOl•
The major features of our simultaneous measurements of neutral thermospheric winds and ionization drift velocities are the following.
(1) In the premagnetic midnight sector the observed ionization drifts are generally directed toward the west. Around magnetic midnight a reversal to eastward drifts is commonly seen. This pattern accords with the generally accepted view of magnetospheric convection and with other radar observations [e.g., Doupnik et al., 1972;Banks, et al., 1973].
(2) In .the premagnetic midnight sector the neutral zonal winds blow toward the west, apparently driven by the ionization motion. At the same time neutral meridional winds do not appear to be controlled by the ionization drifts which are considerably smaller than the corresponding east-west drifts. After magnetic midnight the neutral zonal winds are not  such investigations are becoming available. For example, we have recently turned toward the model neutral atmosphere generated from OGO-6 neutral mass spectrometer data [Hedin et al., 1973] and have found that one would expect neutral winds to blow toward the south beginning around magnetic midnight and increasing thereafter (B. Hinton, private communication, 1973). This prediction corresponds to the data shown, for example, in Figure 3 and is based on the total pressure gradients one computes for the College, Alaska location from the OGO-6 model atmosphere, which includes perturbation (auroral) heating and long-term average behavior.
To conclude, simultaneous measurements of neutral winds in the thermosphere and of ionization drift velocities, when properly interpreted, can be used to identify the driving forces behind the motions. Such interpretations are complicated, however, and must utilize relevant theory of neutral atmosphere dynamics, electrodynamics, and other experimental data as available, e.g., records of magnetic activity and other optical observations of auroral activity.