Search

In recent years, all-sky camera airglow observations of evolving nighttime F-region structures have raised questions regarding the formation and apparent motion of these often wave-like structures. We address these issues using a pseudo-spectral method code developed to numerically solve the Perkins (1973. Spread F and ionospheric currents. J. Geophys. Res. 78, 218–226) moment equations modeling F-region electrodynamics. To aid in interpretation of the results, we utilize a Gaussian shape initial condition of the (geomagnetic field, B, parallel) integrated conductivity under the homogeneous TEC (B-parallel total electron content) condition and a northeastward DC electric field (E-field). We find that the initial Gaussian shape conductivity structure gradually evolves into banded structures oriented along the northwest–southeast direction while the amplitude of the banded structures continues growing and the peak of the structure moves to the northwest due to the E×B drift. The potential distribution corresponding to the initial Gaussian conductivity distribution is more complex but also becomes banded with the same orientation and growing trend as the conductivity. Wave vector domain plots show structure growth in approximately the first and third quadrants and damping in the second and fourth quadrants for both the conductivity and potential, as Perkins predicts—this leads to the orientation of the structures. We note that the evolved banded structures in conductivity and potential are oriented perpendicular to the direction given by half the angle between the DC E-field and east—the direction of maximum instability growth rate predicted by Perkins. The polarization (perturbation) E-field is seen mainly perpendicular to the long axis of the banded structures—i.e., no obvious structure-parallel E-field is observed in the simulation. By tracking the maximum point of the conductivity as a function of time, it is found that the localized structures move northwestward at a nearly constant speed that corresponds to the E×B drift velocity (to within relative errors on the direction and magnitude of ≃4%). We also note that the E×B drift velocity has a dominant effect on the speed and propagation direction of the wave-like bands. The “wave” velocity is the projection of the E×B drift velocity on the line perpendicular to the wave front. Thus, the movement of a northwest–southeast oriented band can be decomposed into two components—parallel (to the band, northwestward) and perpendicular (southwestward) motions. A preliminary comparison of these results with an Arecibo all-sky camera observations shows good agreement.

The Night-time Winter Anomaly (NWA) effect is shown to be a regular phenomenon at the mid-latitude ionosphere in the American and Asian longitude sectors under low solar activity conditions. Total electron content, vertical sounding data as well as topside sounder measurements onboard Alouette indicate a significant higher night-time ionization level in winter than in summer under certain conditions which are discussed in more detail. It is shown that the NWA effect is closely related to the asymmetry of the geographic-geomagnetic relationships in the American and in the Asian longitude sectors, respectively. In the former sector NWA occurs in the northern hemisphere, in the latter in the southern hemisphere. Since the sea-sonal variation of the night-time ionization is similar at conjugated hemispheres, interhemispheric coupling processes are assumed to generate the NWA effect. This idea is consistent with numerical calculations, modelling the system ionosphere-plasmasphere-ionosphere along magnetic flux tubes in the American longitude sector. The model calculations show that interhemispheric fluxes react very sensitive to dynamical forces as f.i. meridional winds in both ionospheres. The derived field-aligned plasma fluxes as well as the resulting higher tube content in December than in June are consistent with the observations.

A model of the ionosphere and plasmasphere is used to study some of the signatures of an idealized SAID (subauroral ion drift) event in the nightside ionosphere. A closed subauroral tube of plasma is considered under solar maximum atmospheric conditions and the westward velocity of 3 km s −1 persists for 30 min. By pursuing, in turn, calculations in which plasma diffusion is suppressed and in which chemical loss of O + is suppressed, the signatures of O + chemistry alone and of field-aligned diffusion alone during the SAID event can be elucidated. Both chemical loss and transport contribute to the decay of the F -layer. Results from full calculations (including both chemistry and transport) demonstrate strong chemical-transport interaction.

This paper presents a preliminary study of ionospheric disturbances at dip latitudes less than 40° as seen by the ion drift meter and the retarding potential analyzer on board Atmosphere Explorer E during a period of low solar activity. The altitude of observation was relatively low, mostly below 300 km. The emphasis is on the midlatitude region, where some features resemble equatorial bubbles; no clear demarcation in latitude could be recognized between the bubbles and other midlatitude disturbances. Excellent evidence was found that the Perkins instability is responsible for very structured disturbances which were frequently observed. In most cases where regions of inward and outward E × B drift were encountered, diffusive motion up or down the field line partially cancelled the vertical component of the cross-field drift. In some cases cancellation was almost perfect, but in others it appeared not to occur at all (probably cases involving rapid changes). Within the larger structures caused by the Perkins instability there were places where the secondary gradient drift instability was also active.

Observations are presented of the perturbation of the mid-latitude bottom-side F-layer by Pi2 geomagnetic pulsations using vertical HF sounding by several well-spaced frequencies simultaneously. It is found that the amplitude and phase of the Doppler velocity oscillation are constant with height. The observations were compared with three mechanisms previously proposed for such perturbations. The detailed numerical model of Sutcliffe and Poole [J. geophys. Res.94, 13,505 (1989)] is interpreted as predicting that, for Pi2 pulsations at night, the dominant mechanism of the three will be vertical electron drift driven by the pulsation electric field. In the framework of this theory, the observed amplitude of Doppler velocity and its height profile show that vertical drift indeed dominates. Simple models, assuming hydromagnetic wave fields uniform with height and neglecting collisions, are developed for each mechanism. These confirm features of Sutcliffe and Poole's model and, taken with the observations presented, again indicate that vertical electron drift dominates for Pi2. The height profile of Doppler velocity phase is consistent with the dominance of vertical drift, but is a poor diagnostic for distinguishing the three mechanisms in the case of Pi2 pulsations.

The O+2 dissociative recombination quantum yields, f(1S) and f(1D), that have been deduced from nocturnal rocket measurements on the volume emission rates V(5577) and V(6300) in the equatorial F region have been analysed to see if they are consistent with the results of ab initio computations on dissociative recombination. In the case of the 557.7 nm emission the only serious problem is that the rocket f(1S) values are much larger than expected in the region below about 220 km. The most interesting feature of the apparent f(1D) altitude profile is that it is markedly influenced by the transport of the suprathermal O(1D) atoms produced by the dissociative recombination.

Power spectra of the fluctuations in the East-West electric field in the post-sunset F-region at the magnetic equator (determined from vertical plasma drift measurements) are presented. The spectra reveal that the fluctuations consist of several components with periods varying from a few minutes to several tens of minutes. The dominant fluctuations have periods of the order of a few tens of minutes with the most frequent being about 30 min; the periods show no dependence on the level of magnetic activity. The spectra also indicate that one component of the fluctuations, usually the component with period about 30 min, is amplified during the period of the upward drift enhancement and attains maximum amplitude at around the time of the evening downward reversal. In general, the fluctuations have minimum amplitude at around 22:00 local standard time (L.S.T.). The characteristics of the spectra suggest that the frequently occurring medium scale gravity waves could be a source for the observed fluctuations in the East-West electric fields.

The effects of an imposed westward plasma drift on O(+) and molecular ion behavior in the nightside ionosphere are investigated using a model of the ionosphere and plasmasphere. A closed subauroral tube of the plasma is considered, and the velocity input persists for 30 min. The rapid increase in the F-region ion temperature resulting from ion-neutral frictional heating causes an immediate surge in the O(+) field-aligned velocity, upwards in the topside ionosphere and downwards below the F2-peak, but after about 10 min into the event the surge in the topside disappears. After the event there is a return flow of O(+) from the plasmasphere. The relative abundance of O(+) decreases during the event due to the increased rate of conversion of O(+) into NO(+) and O2(+); the decrease is more marked for greater values of the imposed westward ion drift. The implications of these results for satellite observations of subauroral ion drifts events and on EISCAT incoherent scatter radar observations of ion heating events is discussed.

The quantum yields ƒ( 1 D) and ƒ( 1 S) of the atomic oxygen excited states O( 1 D) and O( 1 S), respectively, from dissociative recombination in the nocturnal F-region are determined utilizing rocket airglow (OI 630 nm and OI 557.7 nm) and electron density data obtained in two experiments that were carried out at Natal (geogr. 5.8°S, 35.2°W), Brazil, on 11 December 1985 and 31 October 1986. Using the 557.7 and 630 nm airglow data from the second experiment Takahashi et al. (1990, Planet. Space Sci . 38 , 547) calculated the ratio ƒ( 1 S)/tf( 1 D). From a different approach we utilize in this paper the airglow volumetric emission rate and electron density to calculate ƒ( 1 S) and ƒ( 1 D), individually, which are compared with the previously published results from in situ and laboratory measurements. Our results show ƒ( 1 S) to be height dependent. In the second experiment it varied from 1.96 × 10 −2 at 190 km to 1.33 × 10 −1 at 315 km with an average value of 0.053. The average value for the first experiment was 0.041. The ƒ( 1 D) magnitude was also found to vary significantly with altitude, its value increasing both upwards and downwards from 250 km where it attains a minimum of about 0.77.

Quarter-hourly ionograms of Kodaikanal (10°4′N, 77°29′E, dip 3°N) are analysed to evaluate the response of equatorial ionosphere to the severe geomagnetic storm (| Dst |max, 176 nT) associated with the transit of an interplanetary magnetic cloud at Earth during January 13–15, 1967. It is shown that during the early stage of the storm initial phase on January 13, which corresponded to the local dusk-midnight sector, the entire bottomside F region over Kodaikanal experienced a sudden and spectacular downward drift for a short-duration of ∼2 h (18:45 – 20:30 L.T.). The decrease in height of F layer peak ( hm ) by 223 km over the interval 18:45–20:30 L.T. implied a gross apparent downward drift of ∼35.4 m s −1 . The height disturbance occured during the recovery phase of an auroral substorm (bay disturbance) that developed at the commencement of the geomagnetic storm. It also displayed an apparent temporal association with a prominent northward swing in IMF Bz and decrease in polar cap potential drop, φ, estimated from IMF parameters. The anomalous and rapid post-sunset descent of F layer was accompanied by an increase in electron density at and below the F layer peak. The evident perturbation in F region height near the dip Equator is interpreted as the signature of a transient westward electric field disturbance generated by the decrease in magnetospheric convection (decrease in polar cap potential) during the recovery phase of the substorm. The values of vertical downward drift derived from the time rate of change of h ′F and corrected for chemical loss effects indicated the maximum amplitude of the westward electric fields to be ∼1.9 mV m −1 . The increase in Nm is shown to be the outcome of the ionization convergence rate induced by the abnormally large downward drift exceeding the chemical loss rate ( βNe ) at altitudes ⩾250 km.

During a campaign of observations in 1990, we used the Very Large Array (VLA) radiotelescope to study ionospheric waves via their transient signatures in the line of sight total electron content (TEC). The data analysis exploited the fact that VLA baselines (⩽35 km) are comparable with the waves‘ horizontal scale-lengths λ/(2π), thus enabling us to measure the waves’ horizontal phase velocity V. The observed waves consist of two main groups: as seen by numerous earlier studies, traveling ionospheric disturbances (TIDs) dominate at long periods ( >103s); their horizontal phase speeds are typically V

The conditions and mechanisms leading to the observation of perturbations in ionospheric records due to downgoing ULF hydromagnetic waves are examined. About 80 such events recorded at L = 2.1 with a vertical-incidence Doppler sounder and ground magnetometer have been analysed in detail. The dependence on magnetic activity, altitude and pulsation frequency are compared with model predictions. Doppler oscillations associated with Pc3–4 pulsations of amplitude ⩽3.0 nT on the ground are most likely due to wave-driven compression and rarefaction of the ionospheric plasma. At larger amplitudes an E × B mechanism is more important. The reverse holds for night Pi2 events. It is also possible to make estimates of the pulsation scale size in the ionosphere. These suggest the magnetic pulsations are due to field line resonances and global cavity mode oscillations. Shortcomings of the existing interaction models are also discussed. These include the assumption of purely transverse Alfven waves incident on the ionosphere. In the real case fast and coupled wavemodes are also likely to be present. The proximity to a field line resonance thus determines the form of the interaction between the downgoing hydromagnetic waves and the ionosphere. Temporal variations in the contributions of the various interaction mechanisms, and in the relative amplitude of the o and x mode electromagnetic waves, further complicate the relationship between the magnetic pulsations and the observed Doppler oscillations.

This paper outlines the principles of the incoherent scatter technique and briefly reviews the ways in which it has been used to study the ionospheric E and F regions. Examples are given of the important scientific advances that have been made over the last thirty years.

Data derived from 5-min interval ionograms of Kodaikanal (10°14′N, 77°29′E, geomag. lat. 0.6°N) on the night of 29–30 August 1957 showed the presence of a marked perturbation in F-region height (h′F) in the midnight-to-morning period. The perturbation is characterized by a sudden decrease in h′F (65 km in 1 hr, with h′F reaching a low value of 200 km) followed by a prominent increase (200 km in 75 min) over a 3-h period. Changes in h′F of essentially the same nature were also noticed at Calcutta (geomag. lat. 12.25°N) and Ahmedabad (geomag. lat. 14°N) in the same longitude sector simultaneous with those at Kodaikanal. Examination of the auroral electrojet (AE) index and magnetograms of high latitude stations widely separated in longitude revealed the prevalence of an isolated substorm of moderate strength at the time of the perturbations in equatorial F-region height. The decrease in h′F is found to occur around the onset of the substorm and the subsequent increase during the substorm recovery phase. The observed F-region height disturbance is interpreted as the signature of a transient composite disturbance in the equatorial east-west (E-W) electric field caused by the prompt penetration of substormrelated perturbations in high latitude electric fields. The nature of the electric field disturbance is in good agreement with the recent modelling results which predict a westward peturbation in E-W field at equatorial latitudes in response to an increase in polar cap potential (around the onset of a substorm) and an eastward one with a decrease in polar cap potential (during the recovery phase) in the midnight-dawn sector.

The non-linear regime of electrostatic perturbations of the equatorial ionospheric F-region generated by Rayleigh-Taylor instability has been discussed, taking into account conductivity along magnetic field lines. A closed non-linear equation has been derived in the stationary limit for the polarization electric field potential. It coincides with the Karman equation of an ideal liquid. To solve the equation, the averaged variational Whitham method has been proposed. Some solutions localized along and across the geomagnetic field, B, as well as quasi-periodic solutions in the transverse direction, have been investigated. Non-linear longitudinal localization of perturbations has been shown to be due to electron-ion collisions.

The monthly median virtual height (h′F) of the F-region was studied for a period of 6 years (1980–1985) from sunspot maximum to minimum, using data from 11 ionosonde stations in the Japanese-Australian longitudinal sector, in an invariant latitude range: 37°N to 54°S. The night-time maximum in the median height progressively decreases equatorwards, particularly in the local winter and spring, while a reverse weak tendency is observed in summer. The median height reaches peak in both hemispheres from 1 to 2 years after sunspot maximum then decreases towards sunspot minimum. A second diurnal maximum in h′F, preceded by a well-defined minimum, was consistently observed over the solar cycle close to the sunrise time at the F-region, mainly at low invariant latitudes (9–20°). The second maximum has a distinct seasonal variation, being most pronounced in winter and diminishing in summer. It is envisaged that the second peak in h′F is associated with the wave disturbance generated by the supersonic motion of the sunrise terminator. Possible effects of the background height variations on the propagation of the magnetic storm-induced travelling ionospheric disturbances are discussed.

The Saskatoon medium frequency radar and Fabry-Perot interferometer (FPI) were operated continuously from September 1988 to February 1989. The radar data are from the new phase-coherent system which provides winds from 55/75 to 100/110 km during daylight/night hours. The yield at the airglow layer's height near 97 km is much improved, and regularly provides hourly mean winds even on quiet nights when electron precipitation associated with auroral activity is minimal. Observations of the green atomic oxygen OI 557.7 nm emission provide Doppler winds from the airglow layer, on the relatively few nights per month (2–5) when there is no contamination due to clouds, moon or aurorae. These data are supplemented with winds from the F-region's red atomic oxygen OI 630.0 nm emission, which are referred to on disturbed nights. Comparisons are made between the wind vectors from both systems on individual days (mainly quiet ones), and the resulting tidal profiles/points from the radar/FPI systems. Generally the agreement is excellent, with solar tides dominating the wind field of both instruments. However, on disturbed nights, while the radar winds show little or no change compared to the quiet neighbouring days, the FPI winds (558 nm) demonstrate non-tidal motions quite similar to those obtained from the 630-nm emission.

During three days of quiet geomagnetic conditions in August 1988, both the EISCAT UHF radar and the NOAA digital ionospheric sounder (Dynasonde) were operated from 12:00 to 22:00 U.T. Sporadic-E layers were observed on all three days, controlled by the action of the semi-diurnal tide. Gravity wave activity was very evident in both data sets, and particularly visible in the changes in intensity and position of the layers. By combining the two sets of measurements it is possible to build up a spatial picture of the phase fronts passing over Tromso, with sporadic-E layers acting as tracers for them in the E-region. It is observed that structures in the electron density associated with gravity waves descend from the F-region through the E-region to merge with the increases in the layer intensity. The enhancements in the layers, which we suggest are produced by horizontal convergence, are ribbon-like and aligned along the gravity wave phase fronts in the E-W direction, travelling southwards at 60 km h−1 and about 50 km apart.

Hourly values of the virtual height enhancements (Δh'F) of the F-region, from 15 ionosonde stations in both hemispheres, were compared with the corresponding hourly values of the auroral electrojet (AE) index for the year of sunspot maximum (1980). It is evident that surges in the auroral electrojet generate trains of large scale travelling ionospheric disturbances (LSTIDs) in both hemispheres, which subsequently propagate equatorwards causing substantial height enhancements (Δh'F) from sub-auroral to equatorial latitudes. A high degree of correlation exists between the current surges (represented by the variability of AE-index) and Δh'Fin the night-time sector, with an average correlation coefficient of 0.7 and on some occasions reaching 0.9. The optimum correlation between these two different parameters exists for an average time lag of 1.6 h between the onset of an auroral surge and the corresponding value of Δh'F in mid-latitudes. The ratio of the distance (from the auroral oval to a mid-latitude location) of travel of LSTIDs to the time lag corresponds to an average equatorward velocity (500 m s-1) of the disturbance. The correlation between AE and Δh'F depends strongly on solar illumination condition and is absent in the daytime sector. The amplitude of LSTIDs, as represented by the magnitude of Δh'F, tends to be largest in sub-auroral locations near the source region. It rapidly decreases in mid-latitudes and remains fairly constant from mid to low latitudes, with a slight tendency to increase at the equator.

A model of the ionosphere and plasmasphere is used to investigate the effects of an imposed westward plasma drift on H+ and He+ behaviour in the nightside topside ionosphere. A closed subauroral tube of plasma is considered and the velocity spike, of maximum value 2 km s−1, persists for 10 min. The upward surge of O+ flow (caused by frictional heating of the ions in the F-region) gives rise to rapid upward H+ and He+ flows. The H+ and He+ velocities in the topside are comparable to those of O+ but the transient H+ and He+ fluxes are not significant contributors to the magnetospheric light ion population. Above 500 km altitude the ionospheric H+ and He+ densities decrease during the westward plasma drift.

Spectrophotometric observations of the near-u.v. O + emissions from night-time bright auroras peaked in the E -region, as well as from subvisual mid-day cusp auroras peaked in the F -region, are presented. These measurements indicate that the night-time auroral emissions around 3727 A consist mostly (>90%) of the O + line from the allowed transition (3s 4 P 3 2 → 3p 4 S 3 2 ) . On the other hand, in mid-day cusp auroras similar measurements show that the spectral feature observed around 3726 A represents predominantly (about 90%) the forbidden transition of O + [ 2 D 0 3 2 → 4 S 0 3 2 ] . The measured brightness of the latter in the high altitude (>200 km) cusp auroras is (45 ± 12) R. Numerical model calculations of this emission, based on all currently known sources and sinks of O + [ 2 D], yield 0.5 R for the night-time low altitude auroras wherein the population of the long-lived O + [ 2 D] ions is severely depleted through non-radiative loss channels. Similar calculations for the mid-day cusp auroral emissions, excited in the F -region where the collision frequency is relatively much smaller, yield 45 R for the O + [ 2 D 0 3 2 → 4 S 0 3 2 ] 3726.18 line, in agreement with the measured value of (45 ± 12) R. Such a close agreement between measurements and theory indicates that the 40% efficiency value for the production of O + [ 2 D] ions in auroras through electron impact on O, and the rate of 8 × 10 −10 cm 3 s −1 for the quenching O + [ 2 D] by N 2 [to produce N + 2 (A, v′ = 1) and O( 3 P ) ], used in numerical models are most likely appropriate for terrestrial auroral conditions. It also points to these two reactions as the dominant source and non-radiative sink of O + [ 2 D].

Exceptionally large auroral disturbances, associated with the sunspot maximum, occurred on 12–14 March 1989, culminating in severe substorm activity between 12:00 and 15:00 U.T., on 13 March 1989 when the planetary magnetic Kp-index was 8 + and auroral electrojet index (AE) reached a value of 2000 nT. The global effect of the substorm was particularly evident from large rises of the virtual height of the F-region (h′F) and from large ionization depletions found from the critical frequency (foF2) decreases of the layer, using data from 46 ionosonde stations. The fluctuations in h′F showed a significant correlation with auroral surges, deduced from the AE index. The evidence of sequential enhancements in h′F and associated ionization depletions, indicated the presence of large-scale travelling disturbances (LSTIDs) launched during the disturbed period and propagating to low latitudes. LSTIDs moved with velocities between 460 and 530 m s−1, originating equatorwards of the auroral zone at the shell with an L-value = 2.3. The exceptionally large and rapid enhancements in h′F, with the height increments exceeding 200 km and reaching 400 km, as well as the shift of the origin of LSTIDs to higher mid-latitudes, are the unusual aspects of the present disturbances, distinguishing them from other cases of LSTIDs studied on the global scale. The storm-induced height rises were strongly affected by the diurnal conditions and only the ionosonde stations positioned in the night-time sector of the Earth indicated the presence of pronounced increments in h′F from high to low latitudes.

Observations during the re-entry into the denser atmosphere of a mother-daughter payload equipped with a powerful electron gun are reported. The behaviour of the payload potential, the flux of returning electrons and the propagation of an HF signal differed drastically at heights below approx. 130 km from what was observed in the F-region; in particular the payload potential remained well below the accelerator voltage. A 10 MHz signal transmitted between daughter and mother showed variations both in phase and amplitude, whereas no such signatures were seen earlier in the flight when the two bodies were closer to each other. The most likely explanation is based on an increase in plasma density near the payloads in denser regions of the atmosphere. The dynamics of the effects are discussed in terms of ion plasma waves, although no firm conclusions can be drawn.

The upper thermosphere and F -region ionosphere system at 43°N is modelled for equinox and moderate solar conditions via a series of iterative calculations employing a thermospheric wind model and a one-dimensional ionospheric model which are mutually coupled. Several feedback loops within the system involving F 2 -layer peak height, F 2 -layer peak density, zonal wind, meridional wind, and Coriolis force are investigated to better understand the interactive aspect of ionosphere-thermosphere coupling. The interplay of primary importance involves the night-time ascent/descent of the F -layer due to equatorward/poleward neutral winds, the resulting changes in ion drag presented to the meridional and zonal wind fields, and the Coriolis force modification of the ion drag coupling. Wind shear and plasma profile shape are not significantly coupled. For magnetically undisturbed conditions, self-consistent treatment of these effects modifies a non-interactive “control” calculation by 20–35 m s −1 in the wind field. During geomagnetically disturbed periods interactive processes play a more crucial role in determining thermospheric and ionospheric storm responses. Our calculations reveal wind enhancements of up to 100 m s −1 associated with the lifting and negative-phase depletion of the F -region for prolonged magnetic disturbance conditions, the former mechanism accounting for a major portion of the effect.

A comparative study is carried out between a semi-empirical low-latitude ionospheric model (SLIM), including calculated OI 630 nm airglow emission intensities predicted by the SLIM model, with ionospheric and OI 630 nm airglow measurements carried out at Cachoeira Paulista (22.7°S, 45.0°W; geomag. 11.9°S), Brazil. A long series of regular measurements of ionospheric F -region peak electron densities and peak heights, and regular zenith measurements of the OI 630 nm airglow emission intensities, made during the period 1975–1982, are used for comparison with corresponding parameters predicted by the SLIM model. Large discrepancies between the observed and model-predicted OI 630 nm intensities have been found for high solar activity, mainly due to an overestimate of the F -region peak heights by the SLIM model. Use of SLIM profiles normalized by actual F -region peak height and density observations essentially removed the disagreements. The reasons for the height discrepancies are discussed in terms of the longitudinal dependence of the geomagnetic control on the E × B vertical plasma drifts and the thermospheric neutral wind effects on the F -region. The effects of the magnetic declination angle on both the E × B ionospheric plasma drifts and thermospheric neutral winds at low-latitude suggest that ionospheric models need to address separately the longitudinal zones of magnetic declination either positive, negative or zero.

A number of case studies of large, transient, field-aligned ion flows in the topside ionosphere at high-latitudes have been reported, showing that these events occur during periods of frictional heating and/or intense particle precipitation. This study examines the frequency of occurrence of such events for the altitude range 200–500 km, based on 3 years of incoherent scatter data. Correlations of the upgoing ion flux at 400 km with ion and electron temperatures at lower altitudes are presented, together with a discussion of possible mechanisms for the production of such large flows. The influence of low-altitude electron precipitation on the production of these events is also considered.

The Utah State University Ionospheric Model was run to obtain diurnally reproducible ionospheric densities and temperatures for summer and winter conditions using both distorted two-cell and three-cell convection patterns. Differences due to the different convection patterns manifest themselves in the depth and location of polar holes in the F-region electron density. While the total depth of the model holes is a characteristic of the diurnally reproducible pattern, the features appear and are recognizable within 0.5 h. Langmuir probe data from 41 DE-2 passes, during which the IMF Bz component was northward, have been qualitatively checked against the model predictions. The cross polar cap electron density profiles of a large majority of the passes more closely conform to the distorted two-cell runs for both polarities of the IMF By component. This test can be generalized to rule out proposed convection patterns based on the presence/absence and position of polar electron density holes.

This paper reports some results derived by studying statistical characteristics of irregularities which are produced as a result of plasma mixing by random drifts. It is found that as time progresses, the irregularity amplitude grows in proportion to √ t . Characteristics of the irregularity spectrum undergo more violent changes. For example, the spectrum width grows according to a law close to the exponential one. Thus, relatively large-scale random motions (of the size of ~ 10 km, for example) rapidly generate a broad spectrum of small-scale irregularities as small as meter scales. A broadening of the spectrum is stopped by the transverse diffusion or by the longitudinal diffusion and recombination. An assessment is made of the influence of these processes. Equations are obtained, which describe the evolution of the correlation function of plasma density fluctuations for given statistical properties of irregular mixing motions in the case of low-intensity mixing. For the Gaussian spectrum of random drifts, numerical calculations of spectra of irregularities appearing from the initial state, which contain only a smooth large-scale gradient of plasma density, are made. In the initial stage the calculated irregularity spectrum is also close to the Gaussian one and does not involve small scales. A gently-sloping short-wavelength part is then found, where the dependence of spectral power on the wave vector is close to a power law with index — 1 (for a one-dimensional spectrum). Subsequently, the character of the spectrum remains unchanged. The formation mechanism for irregularities investigated in this paper can operate effectively under conditions of middle and high latitudes.

The O(1S) quantum yield in O2+ dissociative recombination f(1S) in the nocturnal F-region and processes O+ + O2 → O + O2+(ν ⩽ 7) and O2+ (ν) + O → O2+(ν′ < ν) + SpaceSci. 36, 621) and Yee et al. (1989, in Dissociative Recombination: Theory, Experiment and Applications, World Scientific, New Jersey) have earlier concluded the values of f(1S) obtained from the Visible Airglow Experiment do not seem explicable in terms of the ab initio results of Guberman (1987, Nature327, 408) on the α(1S, ν) direct dissociative recombination coefficients. It is argued that the discrepancy cannot be attributed to an overlooked O(1S) source in the F-region. Doubt is thrown on the correctness of the results of the Visible Airglow Experiment by a recent determination of the f(1S)/f(1D) ratio by Takahashi et al. (1990, Planet. Space Sci.38, 547) from equatorial F-region profile measurements. These are consistent in general with theoretical expectation. However, discrepancies appear in detailed facets of the data. If they are real they may conceivably be due to the computed α(1S, ν) being less accurate than has been supposed. The possibility is raised that α(1S, 0), in particular, is affected by the neglected indirect dissociative recombination. Laboratory measurements on f(1S) by Queffelec et al. (1989, J. chem. Phys.91, 5335) are considered. They conflict with so much independent evidence that they must be regarded with reserve.

OI 6300 A (1D-3P) and OI 5577 A (1S-1D) airglow emission profiles were measured in the equatorial region. Natal (5.8°S, 35.2°W), by photometers on board a sounding rocket. The two emissions showed similar profiles in the F-region, peaking at around 230 km, with widths of 50–60 km. The overhead zenith intensities below the thermospheric emission layers were 52 Rayleighs (R) for 01 5577 A and 310 R for OI 6300 A, giving an intensity ratio of 0.17. The ratio between the volume emission rates was found to vary with height from 0.25 at 200 km to 0.09 at 310 km. The ratio between the quantum yields f(1S) and f(1D), also found to be a function of height and varying from 0.02 at 200 km to 0.08 at 270 km, showed a good correlation with the simultaneously observed electron density profile. This suggests that the quantum yield f(1S) of the O(1S) production channel of the dissociative recombination process is related in some degree to the electron density.

Measurements from the ionospheric sounding rocket POLEWARD LEAP have been analyzed in order to study the nature of wake phenomena. Some outstanding events, which apparently resulted from the interaction between the incoming particles and the F-layer plasma, are discussed. It is shown that the suprathermal electron fluxes observed during the auroral particle precipitation were clearly spin modulated, and that they usually occurred in the wake region of the rocket. In addition, background measurements of both the auroral particle precipitation and the ambient plasma are presented.

A theoretical multiconstituent model (including O, N2, and O2) which describes the interactions between neutral winds, dynamo electric fields, and ion drifts is used to interpret observations that revealed a dominance of the fundamental diurnal tide in the upper thermosphere and at equatorial latitudes, and its effect on the thermospheric circulation. The model is shown to reproduce reasonably well the magnitudes of the neutral winds, ion drift velocities, and the ratio between the two. A solution for the neutral winds in which the dynamo electric field is forced to zero shows that the dynamo-induced ion drift is very important in accelerating the neutral atmosphere at higher altitudes. The dynamo interaction primarily affects the curl component of the field; its effect on the temperature and density perturbations is small.

Macroscopic instability in the nocturnal ionospheric F -region at midlatitudes is examined when eastward external electric fields exist. The instability growth rate is calculated considering plasma in the entire flux tube. Ion density distribution is obtained by a modeling approach, in which the ion continuity equations for O + and H + are solved. The gravitational Rayleigh-Taylor mode is stable at midlatitudes higher than the crests of the equatorial anomaly, while the E × B mode is unstable at all latitudes when the electric fields are eastward. Net instability growth rate is negative in the narrow region around the magnetic field line which crosses the equator at 1100 km and penetrates the F -layer peak near 18° in magnetic latitude. In lower and higher latitude regions, the ionosphere becomes unstable. In the lower latitude region the instability accounts for generation of plasma bubbles, while in the higher latitude region it accounts for generation of plasma ducts, plasma blobs observed in the topside ionosphere, and midlatitude bottom- side spread F .

We have calculated spectra of small-scale electric fields generated by atmospheric turbulence in the E-region and transmitted into the F-region of the ionosphere. In the case of a power law spectrum of atmospheric turbulence the calculated spectra of electric field also exhibit a power law part in the region of long wavelengths (over several kilometers). In the short wavelength region the spectrum decreases faster as compared with the power law. It is shown that these electric fields produce electron density irregularities with 1–10% amplitude in the F-region. The calculated spectra of the irregularities exhibit a strong elongation of the irregular structure along the geomagnetic field. A typical longitudinal size of the irregularities is of the order of 100 km, and the transverse size is at least several kilometers. The amplitude of electron density fluctuations obtained from estimations agrees with the observed ; however, the calculated spectra do not show short wavelength irregularities with the size of less than or of the order of 1 km, which disagrees with experimental data. The most likely reason for this discrepancy is that the assumption about the wave-like character of turbulence in the upper atmosphere is untenable (we have used this assumption when calculating plasma density fluctuations).

A new Alfven waveguide physical mechanism for the region above the ionospheric F-layer is proposed. This waveguide is due to variation of the ionic masses, connected with the altitudinal increase of light ions' fraction in the chemical compound of the ionsopheric magnetoplasma.

The International Reference Ionosphere 1990 (IRI-90) is described. IRI described monthly averages of the electron density, electron temperature, ion temperature, and ion composition in the altitude range from 50 to 1000 km for magnetically quiet conditions in the non-auroral ionosphere. The most important improvements and new developments are summarized.

Midlatitude F-region neutral winds and temperatures determined from Fabry-Perot interferometer measurements of the doppler shifts and widths of nightglow 630.0 nm line profiles are presented for the priority regular world day 14 August 1980. They exhibit, in many respects, the observed behavior for other summer, geomagnetically quiet nights at solar maximum. The neutral temperature decreases from ∼ 1500°K after sunset (21 h LT) to a minimum of ~ 1200°K before dawn (05 h LT), except to the north of the observatory. The zonal winds are eastward at sunset at ∼ 50 m/sec, decrease to zero at ∼ 02 h LT and are westward just before dawn. The meridional winds are zero just after sunset and reach a maximum equatorward value of ∼ 50–70 m/sec at local midnight but do not decrease as predicted; instead, they remain at roughly these values towards dawn. The NCAR thermospheric general circulation model (TGCM) is used to predict the global upper atmospheric temperature and circulation patterns for this world day. The model predictions agree with the measured neutral temperatures and exhibit qualitative similarities to the measured neutral winds. It is concluded that inclusion in the model of ion drift at midlatitudes should improve the agreement with observations.

Measurements of night-time 6300 A airglow intensities at the Arecibo Observatory have been compared with dissociative recombination calculations based on electron densities derived from simultaneous incoherent backscatter measurements. The agreement indicates that the nightglow can be fully accounted for by dissociative recombination. The comparisons are examined to determine the importance of quenching, heavy ions, ionization above the F-layer peak, and the temperature parameter of the model atmosphere. Comparable fits between the observed and calculated intensities are found for several available model atmospheres. The least-squares fitting process, used to make the comparisons, produces comparable fits over a wide range of combinations of neutral densities and of reaction constants. Yet, the fitting places constraints upon the possible combinations; these constraints indicate that the latest laboratory chemical constants and densities extrapolated to a base altitude are mutually consistent.

Sharp density gradients coupled with field-aligned currents can give rise to short wavelength (≲15 m) drift waves due to collisional effects in the F -region of the auroral ionosphere. In this wavelength range, ion-ion collisions at altitudes of 300–450 km render the ions unmagnetized and a field-aligned current can drive a drift wave, propagating almost transverse to the magnetic field, unstable due to the resistance in electron parallel motion arising from electron collisions.

The generation of growing electrostatic electron waves by electron beams in the ionosphere and magnetosphere is investigated. The auroral F -region, the high latitude exosphere, the auroral acceleration region around 1 R e , altitude, the outer plasmasphere and the plasmasheet are treated. It is found that auroral electron beams can amplify electrostatic waves in all these regions but in different k -ranges. The growth rate, in terms of ω i / ω , generally increases outward. The propagation direction range of the waves discussed varies from a narrow cone around the magnetic field lines to all directions except close to the perpendicular. Strong cyclotron resonance effects at propagation angles close to 90 are not dealt with. The method used can easily be applied to any plasma system where free energy is available in the form of an electron beam, including laboratory plasma.

An unusual SAR arc observed during the growth phase of the ring current is described. Proton precipitation was observed, with electron temperature enhancements throughout the region, and an F-region trough present at the equatorward boundary; a high flux of low energy at the SAR arc location and a 'slot' in the ambient electron density are characteristic of this event. Comparisons are made with S3-A spacecraft observations made in the equatorial region at the same time and with Isis-II observations of a more normal SAR arc; the kinetic Alfven process described by Hasegawa and Mima (1978) appears to account for the acceleration of these low energy electrons, although an auroral-type acceleration process cannot be excluded.

Data from the Topside sounder aboard ISIS II are used to calculate the latitudinal distribution of thermal plasma at 1400 km when the satellite moves along a line of constant geographic longitude. The distribution of thermal plasma is related to the latitudinal regime of the auroral electrojets as inferred from data from a meridian line of magnetometers. It is found that thermal plasma at high altitude tends to be found above the poleward portion of the auroral electrojet. This finding is explained in terms of the spectrum of precipitating electrons across the auroral oval. It is found that the thermal plasma distribution in the post-noon sector is distinctly different from that in the premidnight sector despite the fact that both quadrants feature clear latitudinally confined eastward current flow. This difference can be used to define whether or not the polar cleft penetrates into the local time sector traversed by the satellite. On the nightside, the peak in thermal plasma poleward of the ionospheric trough can be used to identify the magnetic field lines which map to the boundary between the tail lobe and the plasma sheet.

Measurements of electron content ( N T ) near the crest of the equatorial ionosphere anomaly in South America have been made and analysed to investigate N T variations with solar hour, solar rotation and geomagnetic storms. The annual mean of diurnal ratio, defined as the ratio of the maximum to the minimum electron content of the day is found to be 5.0. Anomalous increases in night time electron content are observed with maxima around 2100 LMT and 2300 LMT during summer and equinoctial months. These increases are found to be linked with vertical motion of the F -layer. Spatial resonance in equatorial F -layer plasma appears to be the possible cause of these increases.

The cooling of thermal electrons in the ionosphere by excitation of the fine structure energy levels of ground-state atomic oxygen is considered. From the evidence of theoretical studies of the fine structure transition strengths and from modelling of plasma temperatures in the ionosphere, it is concluded that the fine structure cooling must be included for calculation of reliable values of electron temperature, particularly under sunspot minimum conditions. Debate continues on the precise value for the cooling rate. The rate depends on whether or not the energy levels are in thermodynamic equilibrium; ionospheric simulations of electron temperature are hampered by uncertainties in the heating rate and in other cooling rates.

A series of barium release experiments have been conducted at altitudes near 160 km from the R.A. Hebrides Range ( L = 3.5). A predominantly westward neutral drift has been observed for all the releases conducted during evening twilight. The development of the ionized clouds are characterized, in this situation, by a lack of separation from the neutral release material and by a rapid onset of the formation of structure. The implications of these features are discussed and numerical modelling used to describe the steepening and striation onset in the ion clouds. It is concluded that the most likely striation sizes are those which are of sufficiently small scale to be electrostatically isolated from other regions of the ionosphere but not so small as to be diffusively dispersed.

High quality optical data, obtained by using a high-resolution Fabry-Perot interferometer, were used to derive the thermospheric neutral temperatures and meridional wind velocities at a mid-latitude station, Albany (42.7°N, 73.8°W), New York, during the period 1978–1982. A total of 24 geomagnetically active nights, when the K p index was more than 3, were selected during the observation period. The neutral temperatures varied between about 900 and 1600 K, with a mean value of 1276 ± 54 K. Thermospheric temperatures were also estimated for these selected nights, using the 83 and 86 versions of the MSIS thermospheric model to compare with the experimental values. The experimentally determined temperature variations at the F -region heights were found to be in qualitative accord with the predicted variations based on the thermospheric models. However, the MSIS-83 model gave temperatures smaller than the measured values by about 250 K, while the difference was less than 200 K with the MSIS-86 estimates. In summary, the results of this study show a fair agreement between the observed temperature variations and the estimated variations using the MSIS models for geomagnetically active periods, and there is an increasing need to resolve the quantitative differences. The meridional wind velocities are generally equatorward and they increase from about 125 m s −1 shortly after sunset, to a maximum value of around 200 m s −1 before local midnight, and then decrease to less than about 100 m s −1 toward the early morning hours and become poleward by morning twilight. The mean meridional wind is essentially equatorward in agreement with the thermospheric circulation theory and also in general accord with other mid-latitude optical wind observations.

Calculations are carried out of the elastic scattering and excitation exchange cross-sections in collisions of O(D-1) and O(P-3) atoms, which determine the degree of thermalization of the O(D-1) atoms produced by dissociative recombination in the thermosphere. The effective elastic scattering and excitation exchange cross-sections are calculated to be 1.55 x 10 to the -15th and 6.25 x 10 to the -16th sq cm, respectively, at a relative collision energy of 1.0 eV. The mutual diffusion coefficient between O(D-1) and O(P-3) atoms is also presented.

With an instrument on board the Japanese satellite EXOS-A electron temperatures of more than 1000 K above the “normal” values have been observed in the night-time topside F -region above the geographic region where the total magnetic field is below 20000 nT. Simultaneously enhanced wave emissions on 45 kHz, 2 MHz and 3 MHz were found and an increase in particle precipitation. The observations are described in detail and several mechanism are discussed which can explain the results.

Assessment is made of the effect of the assumed form for the ion velocity distribution function on estimates of three-dimensional ion temperature from one-dimensional observations. Incoherent scatter observations by the EISCAT radar at a variety of aspect angles are used to demonstrate features of ion temperature determination and to study the ion velocity distribution function. One form of the distribution function which has recently been widely used In the interpretation of EISCAT measurements, is found to be consistent with the data presented here, in that no deviation from a Maxwellian can be detected for observations along the magnetic field line and that the ion temperature and its anisotropy are accurately predicted. It is shown that theoretical predictions of the anisotropy by Monte Carlo computations are very accurate, the observed value being greater by only a few pecent. It is also demonstrated for the case studied that errors of up to 93% are introduced into the ion temperature estimate if the anisotropy is neglected. Observations at an aspect angle of 54.7°, which are not subject to this error, have a much smaller uncertainty (less than 1%) due to the adopted form of the distribution of line-of-sight velocity.

High resolution spectra of the 6300 A and 5200 A regions of the night sky have been obtained using a 1 m spectrometer. Typical errors in measurements of O ( 1 D ) 6300 A and N ( 2 D ) 5200 A intensities due to contanimation by overlapping OH emissions have been calculated for a fixed-filter photometer, a tilting-filter photometer and a spectrophotometer. The importance of careful selection of certain instrumental parameters in order to minimize measurement errors is emphasized.