Abstract. This paper presents a summary of 630.0 nm emission observations made by the Cornell All-Sky Imager that have revealed an abundance of structure in the midlatitude thermosphere. Some events were so bright that the weaker 557.7 nm thermospheric line was readily visible and produced sharper images because of the shorter excitation lifetime. Global Positioning System observations show that the airglow features are traveling ionospheric disturbances (TIDs). The remarkable feature of the data is the overwhelming tendency for these low-velocity TIDs to develop with a northwest to southeast orientation and to propagate in the southwest direction. Speeds ranged from 50 to 170 m/s, and wavelengths ranged from 50 to 500 km. The Perkins instability is investigated as a possible explanation for the structures. The linear theory, including both winds and electric fields, predicts a positive but small growth rate. However, the real part of the dispersion relation gives the wrong sign for the wave propagation. Furthermore, the growth rate seems too small to amplify a seed gravity wave significantly during one period of neutral gas oscillation. We conclude that this class of low-velocity TID is not yet explained theoretically.

Observations of midlatitude spread F irregularities have shown that the most violent eruptions are often periodic and may be related to modulations of the F region bottomside. Similar aspects of equatorial spread F have been conclusively shown to be the result of seeding by atmospheric gravity waves. In this paper, we pursue a nonlinear description of the midlatitude F region instability first proposed by Perkins (1973) and show that, by itself, this instability saturates at approximately 1%. With gravity wave seeding, however, the growth of the instability is significantly enhanced and progresses until it is limited by third‐ or fourth‐order nonlinearities. In the course of the analysis we expand the Perkins linear instability study, particularly with regard to the wave propagation angles suitable for stimulating linearly unstable, and eventually, nonlinear turbulent upwellings. Although gravity waves themselves have limited spatial extent, we also show that gravity waves which propagate perpendicular to the Earth's magnetic field create electric fields which map along the field lines to considerable distances. This transmission of electric fields is dependent on both the particular characteristics of the ionosphere along the field lines and the wavelength of the perturbation, but under certain circumstances, gravity wave‐induced large‐scale electric fields can map to the F region from either the E or conjugate F region and thus influence F region electrodynamics. Finally, we note that our results are consistent with observations of midlatitude spread F irregularities by the MU radar in Japan and large‐scale perturbations of the F region over Arecibo.

Although it is generally accepted that equatorial spread F (ESF) is due to nonlinear evolution of the Rayleigh‐Taylor instability, a number of important properties of the process remain unexplained. In particular, we investigate two as yet unexplained features of ESF: the common dominance of very large scale features (≥20 km) and their large amplitude. Although associated for years with spread F we show here for the first time that atmospheric gravity waves can initiate the Rayleigh‐Taylor instability. In agreement with other analytical theories and computer simulations we find that the initiated instability will be saturated by nonlinear coupling of unstable modes to damped modes if the amplitude of the seed gravity wave is small. However, if the amplitude of the seed gravity wave is large enough, the relative plasma density perturbations can reach 50% or more, implying essentially no saturation. The required initial amplitudes are not unreasonable. In the latter case, significant enhancements and depletions of the plasma density occur within several hundred seconds. The analysis presented here demonstrates the possibility that large‐scale spread F is triggered by gravity waves with the Rayleigh‐Taylor instability as a source of amplification. In a separate calculation we also show that large‐amplitude plasma perturbations can be produced by an explosive instability of the Rayleigh‐Taylor modes. It is found that the condition for explosive growth of the Rayleigh‐Taylor modes can be satisfied in the ionospheric F region. We propose that the explosive mode coupling of the Rayleigh‐Taylor instability is a possible mechanism for production of large‐amplitude bottomside sinusoidal structures.

We have used a computer simulation to study gravity wave modulation of midlatitude sporadic E (Es) layers. It is shown that a horizontally stratified Es layer may be deformed by gravity waves and become a large‐scale wavelike structure. Spatial resonance is not required for significant modulations to form. For a southward propagating gravity wave the south side sections of the wavelike deformed Es layer may overturn and appear as field‐aligned features stretching about 10 km in altitude. The deformed Es layer drifts at a velocity larger than neutral wind velocity but smaller than the phase velocity of the gravity wave. The scale length of the wavelike Es layers is determined by the horizontal wavelength of the gravity wave. If there are two or more Es layers modulated simultaneously by a gravity wave, the height dependence of the amplitude of the wave must be taken into account. The Es layer at lower altitude is weakly disturbed if the gravity wave has only a small amplitude there. In contrast, the Es layer at higher altitude may be deeply modulated and appear as field aligned, since the amplitude of the gravity wave has increased greatly in comparison with the amplitude at lower altitude. The numerical results may be used to explain the midlatitude E region field‐aligned irregularities observed by the Middle and Upper atmospheric (MU) radar.

No abstract