Observations and modelling of ionospheric and thermospheric disturbances during major geomagnetic storms: A review

The thermosphere is primarily energised by the combination of three sources of energy and momentum. Solar UV and EUV energy is absorbed globally on the dayside within the middle and upper thermosphere. There is a persistent, but highly variable, inflow of energy and momentum from the magnetosphere. These magnetospheric inputs are usually confined to high latitudes, except at times of very large geomagnetic disturbances. Tides and gravity waves upwell from their sources in the troposphere and stratosphere to deposit energy and momentum at levels from the middle mesosphere to the upper thermosphere. Solar EUV radiation between 120 ran and 250 nm photo-dissociates the molecules which dominate the composition of the lower thermosphere, in particular producing atomic oxygen which dominates the composition of the upper thermosphere. The combination of solar EUV radiation at wavelengths shorter than 120 nm, plus energetic (mainly) charged particles from the magnetosphere, also ionise the neutral constituents of the thermosphere, creating the ionosphere. Particularly at high latitudes, within the geomagnetic polar caps and auroral ovals, the energetic, dynamical and chemical coupling and interactions between the thermosphere and ionosphere dominate the structural and dynamical response of both the thermosphere and ionosphere to solar and geomagnetic inputs of energy and momentum.Comparisons between predictions using global thermosphere-ionosphere coupled models and comparable observational sets have shown encouraging agreement during periods of relatively quiet geomagnetic activity. This indicates that the major energetic, ionisation, chemical and dynamical processes and interactions can be described in models with reasonable accuracy. During periods of high geomagnetic activity, and particularly during major geomagnetic storms, large rapid disturbances of the thermosphere occur with extremely rapid variations. These disturbances are observed as large increases of temperature, density, major changes of neutral composition, and with the development of high speed wind flows and large amplitude waves which may propagate to affect the entire globe. Since the ionosphere is formed from thermospheric constituents and affected by thermospherc dynamics, the gross disturbances of the ionosphere during highly disturbed periods are related to contemporary changes of density, composition and flows of the thermosphere, as well as changes of ionisation sources and electric fields. Observations which describe the nature and scale of disturbances of the thermosphere during geomagnetic storms will be used, in combination with appropriate global numerical simulations, to aid interpretation of storm-time ionospheric phenomena. The role of energetic, dynamical and chemical coupling between the thermosphere and ionosphere is emphasised.     

Vertical motions in the thermosphere over Mawson,Antarctica

High time resolution measurements of Doppler shift and broadening of the (OI) >1630 nm emission in the night airglow and aurora have provided determinations of vertical velocities and temperatures in the neutral thermosphere over Mawson, Antarctica. The vertical wind exhibits a large, rapid and complex response to geomagnetic energy input. Upward winds greater than 50 m s⁻¹ are frequently associated with the expansion phase of auroral substorms. Following the disturbance, prolonged periods of downward winds produce temperature enhancements of 200K outside the source region, thus providing a mechanism for the redistribution of geomagnetic energy. Oscillatory behaviour consistent with thermospheric gravity waves is observed during both quiet and disturbed conditions.     

Simulations of the behaviour of the polar thermosphere and ionosphere for northward IMF

The dynamics and structure of the polar thermosphere and ionosphere within the polar regions are strongly influenced by the magnetospheric electric field. The convection of ionospheric plasma imposed by this electric field generates a large-scale thermospheric circulation which tends to follow the pattern of the ionospheric circulation itself. The magnetospheric electric field pattern is strongly influenced by the magnitude and direction of the interplanetary magnetic field (IMF), and by the dynamic pressure of the solar wind. Previous numerical simulations of the thermospheric response to magnetospheric activity have used available models of auroral precipitation and magnetospheric electric fields appropriate for a southward-directed IMF. In this study, the UCL/Sheffield coupled thermosphere/ionosphere model has been used, including convection electric field models for a northward IMF configuration. During periods of persistent strong northward IMF B_z, regions of sunward thermospheric winds (up to 200 m s⁻¹) may occur deep within the polar cap, reversing the generally anti-sunward polar cap winds driven by low-latitude solar EUV heating and enhanced by geomagnetic forcing under all conditions of southward IMF B_z. The development of sunward polar cap winds requires persistent northward IMF and enhanced solar wind dynamic pressure for at least 2–4 h, and the magnitude of the northward IMF component should exceed approximately 5 nT. Sunward winds will occur preferentially on the dawn (dusk) side of the polar cap for IMF B_y negative (positive) in the northern hemisphere (reverse in the southern hemisphere). The magnitude of sunward polar cap winds will be significantly modulated by UT and season, reflecting E-and F-region plasma densities. For example, in northern mid-winter, sunward polar cap winds will tend to be a factor of two stronger around 1800 UT, when the geomagnetic polar cusp is sunlit, then at 0600 UT, when the entire polar cap is in darkness.     

Simulations of the seasonal variations of the thermosphere and ionosphere using a coupled,three-dimensional,global model,including variations of the interplanetary magnetic field

The University College London Thermospheric Model and the Sheffield University Ionospheric Convection Model have been integrated and improved to produce a self-consistent coupled global thermospheric/high latitude ionospheric model. The neutral thermospheric equations for wind velocity, composition, density and energy are solved, including their full interactions with the evolution of high latitude ion drift and plasma density, as these respond to convection, precipitation, solar photoionisation and changes of the thermosphere, particularly composition and wind velocity. Four 24 h Universal Time (UT) simulations have been performed. These correspond to positive and negative values of the IMF BY component at high solar activity, for a level of moderate geomagnetic activity, for each of the June and December solstices. In this paper we will describe the seasonal and IMF reponses of the coupled ionosphere/thermosphere system, as depicted by these simulations. In the winter polar region the diurnal migration of the polar convection pattern into and out of sunlight, together with ion transport, plays a major role in the plasma density structure at F-region altitudes. In the summer polar region an increase in the proportion of molecular to atomic species, created by the global seasonal thermospheric circulation and augmented by the geomagnetic forcing, controls the plasma densities at all Universal Times. The increased destruction of F-region ions in the summer polar region reduces the mean level of ionization to similar mean levels seen in winter, despite the increased level of solar insolation. In the upper thermosphere in winter for BY negative, a tongue of plasma is transported anti-sunward over the dusk side of the polar cap. To effect this transport, co-rotation and plasma convection work in the same sense. For IMF BY positive, plasma convection and co-rotation tend to oppose so that, despite similar cross-polar cap electric fields, a smaller polar cap plasma tongue is produced, distributed more centrally across the polar cap. In the summer polar cap, the enhanced plasma destruction due to enhancement of neutral molecular species and thus a changed ionospheric composition, causes F-region plasma minima at the same locations where the polar cap plasma maxima are produced in winter.     

Plasma drifts inferred from thermospheric neutral winds and temperature gradients observed at low latitudes

Low-latitude plasma drifts (zonal and meridional) in the F-region are inferred from the observed night-time thermospheric neutral wind velocities and temperature gradients, together with models for the neutral density (MSIS-86 model) and the electron density (IRI model). The thermospheric neutral winds and temperatures are derived from measurements of Doppler shifts and widths of the Oi 630.0 nm airglow emission line, respectively, using a Fabry-Perot interferometer at Cachoeira Paulista (23°S, 45°W), Brazil. The equations considered are the ideal gas law and the momentum equation for the thermosphere, which includes the time variation of the neutral wind, the pressure gradient which is related to the temperature and density gradients and the ion drag force. The present method to infer the night-time plasma drift using observed neutral parameters (time variation of neutral wind velocities and temperature gradients) showed results that are in reasonable agreement with our calculated plasma drifts and those observed in other low-latitude locations. On the other hand, it is surprising that sometimes the winds flow from the observed coldest sector to the hottest part of the thermosphere during many hours, suggesting that plasma drift can drive the neutral winds at low latitudes for a period of time.     

Observations of thermospheric neutral winds by the UCL Doppler imaging system at Kiruna in northern Scandinavia

Since the 1982/1983 winter, the UCL group, in collaboration with the Swedish Institute for Space Physics (previously Kiruna Geophysical Institute), has operated a Doppler imaging system at the high latitude station of Kiruna (67°N, 22°E). The Doppler imaging system is an imaging Fabry-Perot interferometer of 13.2 cm aperture. This instrument has been operated on a ‘campaign’ basis for mapping thermospheric winds using the OI emission at 630 nm (240 km altitude) from a region up to about 400 km radius about Kiruna. In November 1986, the performance of this wide-field Doppler imaging system was augmented by improvements to the detector and all-sky optics. We present data from December 1986, obtained during periods with both clear skies and active auroral and geomagnetic conditions. Maps of the neutral wind flow within the auororal oval during disturbed conditions and near magnetic midnight show continuous and rapid changes of thermospheric winds. The typical scale sizes of eddies observed within the mean flow around magnetic midnight are 100–300 km, with fluctuations at all time scales resolved by the 10 min between successive Doppler images. The local and short period fluctuations appear to be a filtered response of the thermosphere to rapid local variations of the convection and precipitation patterns, within a background of global scale changes     

A numerical model for gravity wave dissipation in the thermosphere

Two simplified models of internal gravity wave dissipation due to viscosity, thermal conduction and ion-drag, in a multilayered, isothermal thermosphere are developed. Each of these models uses the WKB approximation, ray theory and the time-averaged equation of energy conservation, but whereas one of the models (A) employs all of the gravity wave equations appropriate to a dissipative atmosphere, the other (B) does not. Results derived from these models for one particular wave are compared to each other and also to some previously published results of Klostermeyer, which employed a full-wave, model. A breakdown of the WKB approximation in the lower, non-isothermal thermosphere leads to models A and B underestimating the total dissipation there. In the middle thermosphere model A estimates the dissipation reasonably well, while model B grossly overestimates the dissipation. In the upper thermosphere model A underestimates the total upward energy flux, probably as a result of the neglect of coupling into the dissipative waves at these levels, while no energy remains in model B. Results from model A show that when dissipation due to viscosity and thermal conduction are included correctly and simultaneously, the dissipation due to viscosity can exceed that due to thermal conduction by a factor of three. It is argued that ray theory may either overestimate or underestimate the energy flux reaching the upper boundary of a thermospheric model depending on both its height and the particular thermospheric model used.     

A study of the thermospheric forces at a high latitude site on two days of differing geomagnetic activity

Data from the Fabry-Perot Interferometer and Dynasonde at Halley (75.5°S, 26.6°W, L ∼ 4.2), Antarctica, have been used to calculate the forces acting on the high latitude thermosphere. Two case studies of the forces have been undertaken to study why the thermospheric zonal wind speeds are typically so different on nights with different geomagnetic activity. One case study analyses the forces on a geomagnetically active night and the other analyses them on a geomagnetically quiet night. Even on the geomagnetically active night, it is found that the ion drag force is not necessarily the largest force at any one time. Simple comparison of the magnitudes of the forces does not make it very clear which ones dominate in controlling the motion of the thermosphere. This can be seen more clearly by rewriting the momentum equation so that the neutral velocity is expressed in terms of the ion velocity, and the other forces normalized by the ion density. It then becomes clear that, in the evening, the differences in the neutral velocity are due to increases in both ion density and ion velocity, while in the morning, only changes in ion density are important. Thus, although the ion drag force is often not the largest force, it appears that changes in it can account for the variations in neutral velocity between the two nights that were studied.It has also been shown as part of the analysis that whether or not the viscosity needs to be considered when calculating the ion drag force at an altitude of 240 km depends on the ion density profile. If the profile has a single peak then it is only necessary to consider the ion density at 240 km. It is, however, possible that just considering the ion density at this altitude may lead to an underestimate of the effective ion drag force if more than one peak is present.     

On the global and regional behaviour of the mid-latitude ionosphere

Data from a chain of seven ionosondes in the range of 56–38 N and 1–38° E geographic coordinates were analysed to illustrate the global and regional behaviour of the mid-latitude F-region for some selected geomagnetic storms that occurred during the solar cycle 21. It was found that there are different spatial scales in the response of the mid-latitude ionosphere to the disturbance in the magnetosphere-ionosphere thermosphere system. The physical mechanisms and processes are discussed in relation to the relevance of various theories in the understanding of the dynamics of ionospheric storms.     

Wind and temperature effects on F-region medium-scale gravity waves estimated using a multi-layer atmospheric model

Diurnal variations in the propagation direction of atmospheric gravity waves, and the travelling ionospheric disturbances to which they give rise, have been observed in many experimental observations and several modelling studies have demonstrated that this is primarily due to the corresponding diurnal rotation in the direction of the thermospheric wind. Other variations have been attributed to seasonal or other effects, but the effects of variations in the thermospheric temperature have not previously been analysed in detail. We present results from a study of the propagation of gravity waves through a layered atmosphere in which the thermospheric wind and temperature are derived from a three-dimensional time-dependent model. The analysis has been carried out for a range of wave speeds and periods, and for a range of times, seasons and propagation azimuths. Results suggest that a significant diurnal variation in the transmission coefficient for waves propagating through the thermosphere exists with seasonally dependent maxima. Transmission increases for increasing wave period up to about 50 min, after which it remains approximately constant. Maximum transmission occurs for wave phase speeds around 200–250 m/s and falls to zero for speeds less than about 100 m/s. An exception to this rule occurs for waves with periods less than 40 min and speeds less than 50 m/s for which significant transmission appears to be theoretically possible.     

Effects of geomagnetic storms in the lower ionosphere,middle atmosphere and troposphere

Geomagnetic storm effects at heights of about 0–100 km are briefly (not comprehensively) reviewed, with emphasis being paid to middle latitudes, particularly to Europe. Effects of galactic cosmic rays, solar particle events, relativistic and highly relativistic electrons, and IMF sector boundary crossings are briefly mentioned as well. Geomagnetic storms disturb the lower ionosphere heavily at high latitudes and very significantly also at middle latitudes. The effect is almost simultaneous at high latitudes, while an after-effect dominates at middle latitudes. The lower thermosphere is disturbed significantly. In the mesosphere and stratosphere, the effects become weaker and eventually non-detectable. There is an effect in total ozone but only under special conditions. Surprisingly enough, correlations with geomagnetic storms seem to reappear in the troposphere, particularly in the Northern Hemisphere. Atmospheric electricity is affected by geomagnetic storms, as well. We essentially understand the effects of geomagnetic storms in the lower ionosphere, but there is a lack of mechanisms to explain correlations found deeper in the atmosphere, particularly in the troposphere. There seem to be two different groups of effects with possibly different mechanisms—those observed in the lower ionosphere, lower thermosphere and mesosphere, and those observed in the troposphere.     

On tidal variability and the existence of planetary wave-like oscillations in the upper thermosphere—I. Observations of tidal variability

The evidence for the existence of tidal variability as observed in the meridional thermospheric wind (approx. 300 km height) is presented for a set of eight ionosonde locations (three in the northern hemisphere and five in the southern hemisphere). The data set corresponds to a full year (1984) of hourly values. The detected variability can be seen in the tidal components of the meridional wind. The diurnal and semidiurnal components are spectrally analysed. The quarterly spectra show that the tidal amplitudes oscillate with periods between 2 and 60 days. The more important oscillations have periods from 15 to 3 days. No direct link between solar and magnetic activity indices was detected. Possible reasons for the observed tidal variability are discussed in the light of the current theory developed for the mesosphere and lower thermosphere.     

The trouble with thermospheric vertical winds: geomagnetic,seasonal and solar cycle dependence at high latitudes

The vertical wind component is frequently used to determine the zero-velocity baseline for measurements of thermospheric winds by Fabry-Perot and other interferometers. For many of the upper atmospheric emission lines from which Doppler shifts are determined, for example for the OI 630 nm emission, available laboratory sources are not convenient for long-term use at remote automatic observatories. Therefore, the assumption that the long-term average vertical wind is zero is frequently used to create a baseline from which the Doppler shifts corresponding with the line-of-sight wind from other observing directions can then be calculated. A data base consisting of 1242 nights of thermospheric wind measurements from Kiruna (68°N, 20°E), a high-latitude site, has been analysed. There are many interesting short-term fluctuations of the vertical wind which will be discussed in future papers. However, the mean vertical wind at Kiruna also has a systematic variation dependent on geomagnetic activity, season and solar cycle. This means that the assumption that the average value of the vertical wind is zero over the observing period cannot be used in isolation to determine the instrument reference or baseline. Despite this note of caution, even within the auroral oval, the assumption of a zero mean vertical wind can be used to derive a baseline which is probably valid within 5 ms⁻¹ during periods of quiet geomagnetic activity (K_p < 2), near winter solstice. During other seasons, and during periods of elevated geomagnetic activity, a systematic error in excess of 10 ms⁻¹ may occur.     

Optical interferometer measurements of thermospheric temperature at Kavalur (12.5°N, 78.5°E), India

First results on the behaviour of thermospheric temperature over Kavalur (12.5°N, 78.5°E geographic; 2.8°N geomagnetic latitude) located close to the geomagnetic equator in the Indian zone are presented. The results are based on measurements of the Doppler width of O(¹D) night airglow emission at 630 nm made with a pressure-scanned Fabry-Perot interferometer (FPI) on 16 nights during March April 1992. The average nighttime (2130-0430 IST) thermospheric temperature is found to be consistently higher than the MSIS-86 predictions on all but one of the nights. The mean difference between the observed nightly temperatures and model values is 269 K with a standard error of 91 K. On one of the nights (9/10 April 1992, Ap = 6) the temperature is found to increase by ~250 K around 2330 IST and is accompanied by a ‘midnight collapse’ of the F-region over Ahmedabad (23°N, 72°E, dip 26.3°N). This relationship between the temperature increase at Kavalur and F-region height decrease at Ahmedabad is also seen in the average behaviour of the two parameters. The temperature enhancement at Kavalur is interpreted as the signature of the equatorial midnight temperature maximum (MTM) and the descent of the F-region over Ahmedabad as the effect of the poleward neutral winds associated with the MTM.     

Variabilities in the thermospheric temperatures in the region of the crest of the equatorial ionization anomaly—a case study

Results on spectroscopic measurements of thermospheric temperatures made from a low latitude station, Abu (24.6°N, 72.7°E, geographic; 18°N dip latitude), India, situated in the crest region of the equatorial ionization anomaly (EIA), are presented. On many occasions, these measurements reveal large deviations frorn the predictions of the neutral atmospheric model, MSIS-86, bringing out its limitations as applied to the equatorial and low latitude thermosphere. The role played by large-scale geophysical processes like the EIA, equatorial spread F (ESF) and the midnight temperature maximum (MTM), all of which influence the thermal structure of the upper atmosphere, is examined in the context of explaining the differences between the measured temperatures and model predictions. It has been conclusively shown that Joule heating associated with ESF irregular electric fields is not solely responsible for the observed deviations, and the possibility of a significant role by the EIA related processes is indicated.     

Thermospheric and mesospheric temperatures during geomagnetic storms at 23°S

Night-time thermospheric temperatures, T63o, and mesospheric rotational temperatures, T(OH) and T(O₂), have been measured at Cachoeira Paulista (23°S, 45°W, 16°S dip latitude), located in both the equatorial ionospheric anomaly and the South Atlantic Geomagnetic Anomaly, with a Fabry-Perot interferometer and a multi-channel tilting filter-type photometer, respectively. The thermospheric temperatures are obtained from the Doppler line broadening of the OI 630.0 nm emission and the mesospheric rotational temperatures from the OH(9,4) and O₂A(0,1) band emissions. Measurements made during three geomagnetic storms showed that the nocturnal mean values of T₆₃₀ during the recovery phase of the storms were lower than those observed during quiet time and from model predictions. Also, the nocturnal mean value of the T₆₃₀ soon after the SSC event on 27 June 1992 was higher than the quiet time and model predictions. The observed mesospheric nocturnal mean rotational temperatures, T(O₂) and T(O₂), were unaffected by the storms. A comparison of the night-time observed temperatures T₆₃₀, T(OH) and T(O₂) with those calculated using the MSIS-86 model is also presented.     

Influence of fluctuations in neutral gas density height distribution on F1 -layer dynamics

Model ionograms are analyzed, corresponding to conditions of disturbances to the diffusive-equilibrium gas density distribution. The results obtained show that the F1-layer clearly demonstrates a response to the dynamics of the vertical thermospheric structure. Features of the model virtual height-frequency curves correspond to the character of experimental ionograms recorded in the case of internal gravity waves.     

Comparison of ionospheric electrical conductances inferred from coincident radar and spacecraft measurements and photoionization models

Height-integrated electrical conductivities (conductances) inferred from coincident Sondrestrom incoherent scatter radar and DMSP-F7 observations in the high-latitude ionosphere during solar minimum are compared with results from photoionization models. We use radar and spacecraft measurements in combination with atmospheric and ionospheric models to distinguish between the contributions of the two main sources of ionization of the thermosphere, namely, solar UV/EUV radiation and auroral electron precipitation. The model of Robinsonet al. (1987, J. geophys. Res.89, 3951) of Pedersen and Hall conductances resulting from electron precipitation appears to be in accordance with radar measurements. Published models of the conductances resulting from photoionization that use the solar zenith angle and the solar 10.7-cm radio flux as scaling parameters are, however, in discrepancy with radar observations. At solar zenith angles of less than 90°, the solar radiation components of the Pedersen and Hall conductances are systematically overestimated by most of these models. Geophysical conditions that have some bearing on the state of the high-latitude thermosphere (e.g. geomagnetic and substorm activity and a seasonal variation of the neutral gas distribution) seem to influence the conductivity distribution but are to our knowledge not yet sufficiently well modelled.     

Non-migrating tides

Recently, two interesting advancements in the study of non-migrating tides have occurred. There are currently two distinct approaches to this subject. One. based on mechanistic models that consider heating due to non-uniform global distribution of water vapour or heating only over lands, solves the primitive equations. One model of this approach that shows insolution absorption of the non-uniformly distributed water vapour produces non-migrating tides with 15% of the migrating ones in the lower thermosphere. The planetary boundary layer heating can explain very enhanced tides over land masses and those with short vertical wavelengths in the stratosphere. The other approach uses a general circulation model (GCM) simulation. This model can produce tides globally and in many details. The model predicts enhancement of the non-migrating modes as eastwards traveling modes with a wave number 3 and westwards traveling modes with a wave number 5 that is in a surprisingly good agreement with observation at sea-level, at 700 mb and at 300 mb.Whilst the GCM simulation is to be developed so as to include thermospheric tides, the mechanistic model should consider more realistic situations so as to include winds. Observation of tides at many heights and locations is essential in future studies.