Parameters of the interplanetary magnetic field and solar wind plasma during periods of 163 isolated substorms have been studied. It is shown that the solar wind velocity V and plasma density N remain approximately constant for at least 3 h before substorm onset Т o and 1 h after Т o . On average, the velocity of the solar wind exhibits a stable trend toward anticorrelation with its density over the whole data array. However, the situation is different if the values of V and N are considered with respect to the intensity of substorms observed during that period. With the growth of substorm intensity, quantified as the maximum absolute value of AL index, an increase in both the solar wind plasma velocity and density, at which these substorms appear, is obsreved. It has been found that the magnitude of the solar wind dynamic pressure P is closely related to the magnetosphere energy load defined as averaged values of the Kan–Lee electric field E KL and Newell parameter d Φ/ dt averaged for 1 h interval before Т o . The growth of the dynamic pressure is accompanied by an increase in the load energy necessary for substorm generation. This interrelation between P and values of EKL and d Φ/ dt is absent in other, arbitrarily chosen periods. It is believed that the processes accompanying increasing dynamic pressure of the solar wind result in the formation of magnetosphere conditions that increasingly impede substorm generation. Thus, the larger is P , the more solar wind energy must enter the Earth’s magnetosphere during the period of the growth phase for substorm generation. This energy is later released during the period of the substorm expansion phase and creates even more intense magnetic bays.

Characteristics of isolated substorms selected by variations in the 1-min values of the AL index are analyzed. The substorms were divided into several types with respect to the behavior of the Bz component of the interplanetary magnetic field (IMF) during the expansion phase. The probability of observations of substorms associated with the northward turn of the Bz component of IMF was ~19%, while the substorms taking place at Bz < 0 were observed in 53% of cases. A substantial number of events in which no substorm magnetic activity was observed in the auroral zone after a long (>30 min) period of the southward IMF and a following sharp turn of the Bz component of IMF before the north was detected. The data suggest that a northward IMF turn is neither a necessary nor sufficient condition for generating substorms. It has been shown for substorms of the both types that the average duration of the southward IMF to moment T 0 and the average intensity of the magnetic perturbation in the maximum are approximately the same and amount to ~80 min and–650 nT, respectively. However, for substorms at Bz < 0, their mean duration, including the expansive and recovery phases, is on average 30 min longer than that at a northward turn of IMF. Correlations between the loading–unloading processes in the magnetosphere in the periods of magnetospheric substorms were investigated with different functions that determine the efficiency of the energy transfer from the solar wind to the magnetosphere. It has been shown that the highest correlation coefficient ( r = 0.84) is observed when the function suggested by Newell et al. (2007) is used. It has been detected that a simple function VB S yields a high correlation coefficient ( r = 0.75).

Research results about planetary-scale auroral distributions are presented in a historical retrospective, beginning with the first "maps of isochasms" – lines of equal visibility of auroras in the firmament (Fig. 2) – up to "isoaurora maps" – lines of equal occurrence frequency of auroras in the zenith (Fig. 4). The exploration of auroras in Russia from Lomonosov in the 18th century (Fig. 1) until the start of the International Geophysical Year (IGY) in 1957 is shortly summed up. A generalised pattern of discrete auroral forms along the auroral oval during geomagnetically very quiet intervals is presented in Fig. 5. The changes of discrete auroral forms versus local time exhibit a fixed pattern with respect to the sun. The auroral forms comprise rays near noon, homogeneous arcs during the evening, and rayed arcs and bands during the night and in the morning. This fixed auroral pattern is unsettled during disturbances, which occur sometimes even during very quiet intervals. The azimuths of extended auroral forms vary with local time. Such variations in the orientation of extended forms above stations in the auroral zone have been used by various investigators to determine the position of the auroral oval (Fig. 9). Auroral luminosity of the daytime and nighttime sectors differ owing to different luminosity forms, directions of motion of the discrete forms, the height of the luminescent layers, and the spectral composition (predominant red emissions during daytime and green emissions during the night). Schemes that summarise principal peculiarities of daytime luminosity, its structure in MLT (magnetic local time) and MLat (magnetic latitude) coordinates, and the spectral composition of the luminosity are presented in Figs. 15 and 19. We discuss in detail the daytime sector dynamics of individual discrete forms for both quiet conditions and auroral substorms. The most important auroral changes during substorms occur in the nighttime sector. We present the evolution of conceptions about the succession of discrete auroral forms and their dynamics during disturbance intervals. This ranges from Birkeland's polar elementary storms, over the prospect of a fixed auroral pattern up to the auroral substorm model. The classic schemes of the spatial distribution and motion of discrete auroral forms during single substorms are shown in Fig. 20 (expansive and recovery phases) and Fig. 21 (creation, expansive and recovery phases). In this review we discuss various models of bulge formation, in particular as a result of new formation of arcs about 50–100 km poleward of previously existing auroral structures (Fig. 24). Discrete steps in the development of an expanding bulge are separated by 1–3 min from each other. The model of successive activations confines only to a ~40° longitudinal portion of the magnetotail (Fig. 28). We consider differences in the development of single substorms and substorms during magnetic storms. The structure and dynamics of auroras during steady magnetospheric convection (SMC) periods are dealt with in Sect. 8. A generalised scheme of the auroral distribution during SMC periods is shown in Fig. 34. Separate sections describe discrete auroras in the polar cap (Sect. 5), and the diffuse luminosity equatorward of the auroral oval (Sect. 9). Visual observations of diffuse auroral forms at midlatitudes suggest that the whole latitudinal interval between the auroral oval and the stable auroral red (SAR) arc is filled up with diffuse luminosity. SAR arcs with intensities of several tens of Rayleigh enclose systematically the region of diffuse luminosity; they are positioned at the border of the plasmasphere.

On the basis of observations for the IGY period (visoplots) it is shown, that during magnetic storms diffuse glow is detected at all latitudes between the lowest latitude of the visually observed auroral glow at the zenith and the auroral oval. The diffuse glow region spatially coincides with the region of soft electron precipitation extending equatorward from the boundary of the oval to the latitude of the plasmopause projections along the magnetic force lines to the ionosphere. Using published materials on the diffuse glow dynamics and SAR arcs at the Yakutsk meridian, as well as simultaneous measurements of the DMSP F9 satellite, we discuss the contribution from low-energy electron precipitation transfered via convection toward Earth from the magnetosphere's plasma sheet to excitation of 630.0 nm emission in low-intensity (<1.0 kR) SAR arcs.

[1] The paper by Akasofu et al. [2010, hereinafter Paper 1] gives an ambiguous interpretation of the auroral substorm observations. The authors consider their conclusion as a new result but it does not follow from the observations described in the paper. On the other hand, the phenomenon discussed was investigated earlier in papers not cited in Paper 1. [2] The concept of auroral substorm was proposed by Akasofu [1964] for interpretation of the morphology and dynamics of auroras on the night side of the Earth (from 1800 to 0600 MLT). The change from quiet to disturbed conditions and back to the original state is a cyclic process. According to him each cycle consists of two phases: expansion phase (0 < T < 30 min) and recovery phase (30 min < T < 2 h). The time T = 0 (or T0) denotes the onset of a substorm and is the beginning of the expansion phase. [3] For years in the 1950s, Y. Feldstein and G. Starkov were conducting aurora observations at the Arctic Observatory on the Dixon Island. Since the morphology and dynamics of the auroras observed fitted the Akasofu's concept, they supported it without reserve. Аkasofu [2002, p. 59] notices “Many auroral scientists who have actually little experience in observing the aurora simply followed the experienced ones. Thus, it was hard to convince anyone about the validity of the concept of the auroral substorm. The only exception at that time was Feldstein, who strongly supported my finding.” [4] The original scheme of the auroral substorm was supplemented by Akasofu [1968, hereinafter Paper 2] with the distribution of auroras in the daytime sector obtained for substorm interval first by Feldstein and Starkov [1967a, 1967b]. The new scheme covering all MLT hours is represented in Paper 1, Figure 1 or Akasofu [2010, Figure 2]. This scheme of evolution of the auroral substorm has gained ample recognition in the scientific community. Often, it is erroneously referred to as the publication of 1964 (like it is in Paper 1, see Akasofu [2010, Figures 1 and 2]), while, in fact, it was published only by Akasofu [1968]. Perfection of two-phase auroral substorm model was continued later with supplement of the third phase (creation or growth phase) for interval −60 < T < 0 min by Feldstein and Starkov [1970] and Starkov and Feldstein [1971, hereinafter Paper 3]. The term growth phase was defined by McPherron [1970] based on magnetic field variations analyze and so far used for many phenomena before T = 0. For auroral activity before T = 0 we offered a term creation phase. The most distinct manifestation of this phase is an equatorward shift of auroral oval in evening and premidnight sectors before T = 0. [5] Using Akasofu's [1968] scheme as a working definition of the planetary dynamics of auroras in the substorm period, Starkov et al. [1971] analyzed ascafilms from Chelyuskin (Φ′ = 71°.2), Dixon (Φ′ = 68°), and Murmansk (Φ′ = 65°.1) stations for a large number of substorms. The criterion for the selection of substorms was the appearance of negative bay-like magnetic disturbances after the quiet background at premidnight hours. This selection technique leaves out pseudo equatorial motions of auroras in the evening sector due to propagation of the large-scale folding structure (WTS according to Akasofu [1964, 1968]). The result of the analysis was the scheme of evolution of an isolated auroral substorm consisting of three phases (creation, expansion, and recovery), which is represented in Figure 1 [Feldstein and Starkov, 1970; Starkov and Feldstein, 1971]. Below, we discuss the rightfulness of the new conclusion drawn by Akasofu et al. (Paper 1) and its compatibility with the schemes of the auroral substorm by Akasofu (Paper 2) and Feldstein and Starkov (Paper 3). [6] Based on meridian scanning photometer (MSP) observations, Akasofu et al. [2010, paragraph 23] made a conclusion that “The equator half of the oval (EQ), not the whole oval, shifts equatorward prior to onset; this is a new result. The poleward arc(s) remains approximately in the same location. This is a new observation.” MSP observations for the isolated substorm of 12 January 1997, which started after a few magnetically quiet hours, are shown in Figure 4a of Paper 1 to illustrate a typical case justifying such conclusion. The variations of AL index of geomagnetic activity represented in Figure 2 (left) agree with the classical scheme of a magnetic substorm proposed by McPherron [1970]: a slow decrease from −25 nT at 0630 UT to −100 nT at 0720 UT during the growth phase followed by a steep fall down to −700 nT (substorm commencement, T0). In the evening sector of the auroral oval, there usually exist from one to three auroral arcs. As is usually the case under relatively quiet geomagnetic conditions, the oval in Figure 4a of Paper 1 is represented by a single arc. Until the WTS passage at about 0810 UT, the arc is moving equatorward. It is not only the equatorial part that moves but the oval as a whole. According to the MSP data of Figure 4a, there was only very faint luminosity poleward of the arc, which remained in the poleward sky. It did not have the characteristics attributed to the poleward arc on the evening oval due to very low intensity and absence of auroral forms in luminosity. [7] The other examples of the auroras dynamics by Akasofu et al. [2010, paragraph 12], from which “one can see more clearly the presence and independence of poleward arc(s) from EQ,” are given in section 3 of Paper 1 (Figures 4b, 4c, and 4d). In these examples, the magnetic field is strongly disturbed, and the distribution of auroras in the evening sector is typical for the expansion phase, that is, largely, due to the WTS passage from the nighttime sector. We shall restrict our consideration to the event of 25 January 2003. The MSP data for this event are given in Figure 4b of Paper 1 and the AL index in Figure 2 (right). A few intensive substorms manifested in the evening auroras were recorded in the magnetic field during the time interval 0430 UT–0900 UT. The first substorm was accompanied by aurora enhancement at about 0500 UT. The onset of the next one at ∼0545 UT resulted in a fast shift of the equatorial boundary of the oval to lower latitudes and its essential expansion. The sawtooth oscillations of the boundary position reflect the irregular nature of the magnetic disturbance. In this case, the equatorward motion occurs after the moment T0, i.e., in the expansion phase. The high-latitude arc in Figure 4b of Paper 1 is part of the WTS, which has its poleward edge in the expansion phase located usually at even higher latitudes than the arc. The arc in Figure 4b keeps its latitude nearly constant, since it is located at the maximum zenith angles of the photometer near the poleward horizon of the observation point. [8] The Poker Flat observations (Φ′∼65°) are obviously unsuitable for unambiguously determining the morphology and structure of the substorm auroras on the poleward side of the oval, because the auroral bulge in the late evening sector extends as far as Φ′ > 72°, i.e., out of sight of MSP. So, the data from the stations located at 68° < Φ′ < 72° must be invoked. Such observations have been carried out since the 1970s on Chelyuskin and Dixon stations in the Eastern hemisphere and by the meridian chain of all-sky cameras in the Western hemisphere, and their results have been published, e.g., by Starkov et al. [1971], Vorobjev et al. [1976], Zverev et al. [1976] and by Snyder and Akasofu [1972], respectively. We shall dwell on the paper by Snyder and Akasofu [1972], which was written in coauthorship with the first author of the paper we are commenting here. [9] The Alaskan meridian chain of station “scans” the polar sky once a day between geomagnetic latitudes of 60° and 80°. The paper presents for five selected periods in 1969–1970 the dynamic behavior of the auroral oval as delineated by the all-sky photographs (film). Substorm phases and moment Т0 were determined by all-sky camera (ASC), the magnetic records chain of stations and magnetic activity indices AL and AU. The substorms occurred at near-midnight hours. Among several features of auroral morphology presented by Snyder and Akasofu [1972] we quote the following: [10] 1. Before the onset of an auroral substorm (till moment Т0) enhanced equatorward drift of 2 or 3 auroral forms occurs which constitutes evening-premidnight oval sector. All forms drift equatorward. There are no differences in drifts between equatorward and poleward halves of the oval. The speed of the equatorward auroral motions may be the same magnitude as the speed of the poleward expansion motions that occur after T0. Equatorward drift motions of auroras are a common feature and therefore, according to opinion of authors, cannot be ground for existence of the growth phase; [11] 2. At the time T0 the sky poleward the oval becomes clear from auroral forms. Such a clearing, along with the equatorward drift motions, makes the oval thinner, so that, at the time T0, it may be represented by a single arc; [12] 3. The boundary of the auroral bulge in the near-midnight sector reaches the latitudes Φ′∼75°–77°; [13] 4. There are no unambiguous characteristic in the AU or AL indices that can be associated with the enhanced equatorward drift of auroras before the onset of an auroral substorm in the midnight sector. They note additionally that the method used to isolate the growth phase by McPherron [1970] relies on the data from insufficiently dense magnetic observatories network in the auroral zone, which makes the identification of the growth phase as a separate, independent phase of a substorm difficult. [14] Thus, observations of the Alaskan meridian chain of stations did not reveal the division of the auroral oval prior to T0 into the equatorial and poleward parts with different types of drift according to Paper 1. All arcs, that form the oval in the evening and premidnight sectors, drift in one direction, i.e., toward the equator. The drift velocity increases 1.5 h before T0. It is strange enough that after a few decades of doubt about the existence of the creation (growth) phase as a typical part of the auroral substorm, the authors of Paper 1 explain it just like Feldstein [1974] did it 36 years ago as can be seen below: [15] From Paper 1 [2010, paragraph 11]: “Note that the speed of this southward shift of EQ [equatorward half of the auroral oval] is much faster than the apparent shift of the oval because of its eccentricity with respect to the geomagnetic pole (1° in gm. latitude/1 hour).” An abrupt increase in the equatorward drift velocity of auroral forms occur within 1.5 h prior to T = 0. Such increase in the velocity is due solely to the subsequent substorm, for the other days at the same hours of UT in the absence of substorms the equatorward drift velocity of the auroral form was considerably lower. A ∼5° latitude shift of the luminosity region for a time of∼1.5 h cannot be explained by the Earth's rotation beneath the oval. [17] The velocities of the natural drift of arcs in different time sectors in three phases of a substorm were determined by Vorobjev et al. [1976]. The drift of discrete forms due to the Earth rotation under the auroral oval was excluded. [18] Let us interpret the aurora dynamics before and after the beginning of the substorm active phase described in Paper 1 using the observations summed up in the substorm scheme shown in Figure 1. The scheme comprises the substorm creation (or, growth) phase (−1 h < T < 0), when during the minimum variation in the luminosity intensity the oval nighttime sector shift for a typical substorm by ∼5° equatorward down to Φ∼65°. This motion explains why, according to Akasofu [1968], the substorm starts at midnight at precisely this latitude. The model in Figure 1 allows us also to understand the particularities of luminosity distribution illustrated in Figure 4a of Paper 1 during the creation phase. The photometer records display a weak diffuse luminosity in the evening sector poleward of the discrete forms (oval), which is interpreted in Paper 1 as the poleward part of the oval. However, according to the model in Figure 1, a diffuse luminosity occurs in the creation phase poleward of the auroral oval arcs. It differs from the discrete forms in the auroral oval and cannot belong to the high-latitude part of the oval. These two types of the luminosity differ also by their relation to the plasma domains in the magnetosphere: the nighttime sector of the auroral oval is projected by the magnetic field lines onto the central plasma sheet in the magnetosphere tail, while the diffuse luminosity poleward of the oval is projected onto the plasma sheet boundary layer. The morphology of nighttime auroral luminosity, including diffuse aurora poleward of the auroral oval, its characteristics and connection with plasma structure of the magnetosphere is stated in the review by Feldstein and Galperin [1985]. [19] Model luminosity distribution shown in Figure 1 has a characteristic peculiarity: in the substorm active phase the structured forms are disintegrated in the central part of the oval. Such regions of diffuse glow are readily revealed as a dark area in MSP data (Figures 4b, 4c, and 4d of Paper 1) during the substorm expansion and recovery phases. This peculiarity is one of differences between substorm development schemes by Akasofu et al. [2010, Figure 1] and in Figure 1. [20] Based on ASC observations of auroras Akasofu et al. [2010, paragraph 22] claim that “poleward arc(s) brightens only after the initial brightening and shows only moderate activities.” It follows from Figure 1 that, in the case of two arcs, AE and AP existing in the midnight sector, the time T0 is associated with the equatorial arc (AE) brightening and splitting (or appearance of a new arc immediately on its poleward side), and a fast poleward motion of its more intensive higher-latitude part. The weaker lower-latitude part remains in the old place. The initial poleward arc (AP) does not take active part in the development of the auroral substorm and is absorbed by the poleward auroral bulge that originates from AE. Among the poleward and equatorward arcs that are formed from the original AE, the main role in the substorm evolution belongs to the poleward arc, which is at the high-latitude boundary of the bulge (of the auroral oval), while the equatorward arc, which is at the equatorward boundary of the oval, is less active (see Figures 2c and 2d of Paper 1). [21] Main conclusions: [22] 1. Paper 1 differs from the earlier paper by the same author [Snyder and Akasofu, 1972]. [23] 2. Conclusions of the earlier paper were based on observations covering simultaneously bigger longitude and latitude intervals (all-sky cameras), than MSP in Paper 1. The conclusions of the earlier paper were consistent with the results obtained by this comments author in 1970–1971 years. [24] 3. We believe that Figure A of Feldstein and Starkov [1970] still holds as the most complete summary of the auroral substorm development. [25] The scheme of evolution of the auroral substorm with two phases was proposed by Akasofu [1964] and was modified by Feldstein and Starkov [1970] and Starkov and Feldstein [1971], who introduced the third phase. Now, 40 years later, it's probably, high time to modify it again. Large aurora observations data, both from spacecraft and from ground are available now. The new auroral substorm scheme should incorporate and generalize them. We hope that the new scheme would be dated to the year of creation rather than the years of publication of the first (1964) or second (1970) versions. [26] Robert Lysak thanks the reviewer for his assistance in evaluating this paper.

In the period of the International Geophysical Year (IGY), almost the entire planet was covered for the first time by ground-based geophysical observations. Their analysis led to two fundamental results: the existence of the auroral oval and auroral (magnetospheric) substorm. At the final stage of the IGY, satellite explorations of the near-Earth space began. The auroral luminosity appeared to be related to the plasma structure of the magnetosphere. That opened new possibilities for parameters diagnostics of the Earth’s magnetosphere on the basis of ground-based aurora observations. The concepts of auroral oval and magnetospheric substorm became paradigms of the new science of solar-terrestrial physics.

The optical observations on Heiss Island (Φ′ = 75.0°) have been used to study the characteristics of auroras in the near-noon MLT sector after abrupt increases in the solar wind dynamic pressure at negative and positive polarity of the IMF B z component. It has been found out that the 427.8 and 557.7 nm emission intensities considerably increased at B z < 0 both equatorward of the dayside red luminosity band and within this band. The value of the emission intensities at a red luminosity maximum (I 6300/I 5577 ∼ 0.5) indicates that energetic electron precipitation is of the magnetospheric origin. At B z > 0, fluxes of harder (E > 1 keV) precipitating electrons were superimposed on the soft spectrum of precipitating particles in the equatorial part of the red luminosity band. This red band part was hypothetically caused by the low-latitude boundary layer (LLBL) on closed lines of the geomagnetic field, the estimated thickness of which is ∼3 R e . The 557.7 nm emission intensity increased during 3–5 min after SC/SI and was accompanied by the displacement of the red band equatorward boundary toward lower latitudes. The displacement value was ∼150–200 km when the dynamic pressure abruptly increased by a factor of 3–5. After SC/SI, the 630.0 nm emission intensity continued increasing during 16–18 min. It is assumed that the time of an increase in the red line intensity corresponds to the time of saturation of the magnetospheric boundary layers with magnetosheath particles after an abrupt increase in their density.

Оптические наблюдения на о. Хейса ( = 75.0°) использованы для изучения характеристик полярных сияний в околополуденном секторе MLT после резких увеличений динамического давления солнечного ветра при отрицательной и положительной полярности Bz компоненты ММП. При Bz 0 в экваториальной части полосы красного свечения на мягкий спектр высыпающихся частиц наложились потоки более жестких (E > 1 кэВ) высыпающихся электронов. Предположительным источником этой части красной полосы является низкоширотный граничный слой (LLBL) на замкнутых силовых линиях геомагнитного поля, толщина которого оценена в 3 Re. Увеличение интенсивности эмиссии 557.7 нм после SC/SI наблюдалось в течение 3–5 мин и сопровождалось смещением экваториальной границы красной полосы в более низкие широты. Величина смещения составляла 150 200 км при резких увеличениях динамического давления в 3–5 раз. Увеличение интенсивности эмиссии 630.0 нм после SC/SI продолжалось в течение 16–18 мин. Предполагается, что время роста интенсивности красной линии соответствует времени насыщения граничных слоев магнитосферы частицами переходного слоя после резкого увеличения их концентрации.

The position of the auroral luminosity equatorward boundary during the interaction between the Earth’s magnetosphere and isolated solar wind streams from different solar sources has been statistically studied based on the ground and satellite observations of auroras. These studies continue the series of the works performed in order to develop the technique for predicting auroras based on the characteristics of the interplanetary medium and auroral disturbances. The dependences of the minimal position of the auroral luminosity equatorward boundary (Φ′) on the values of the azimuthal component of the interplanetary electric field (Ey) and AL indices of magnetic activity, averaged over 6 and 24 h, are presented. The distribution limits for each type of isolated solar wind streams on the Φ′-Ey and Φ′-AL planes have been determined.

The studies of auroras at Russian Antarctic observatories in the Southern Hemisphere began in 1957 during the second Complex Antarctic expedition and performed almost continuously up to 1993 during more than 30 years. Many observers of auroras and scientists that analyzed obtained results participated in these studies. Members of the Arctic and Antarctic Research Institute (AANII), Russian Committee on Hydrometeorology (Rosgidromet); Pushkov Institute of Terrestrial Magnetism, Ionosphere, and Radiowave Propagation, Russian Academy of Sciences (IZMIRAN); Vernov Institute of Nuclear Physics, Moscow State University (NIIYaF MGU); Polar Geophysical Institute, Russian Academy of Sciences (PGI); St. Petersburg State University (SPbGU); Schmidt Institute of Physics of the Earth (IFZ); Shafer Institute of Cosmophysical Research and Aeronomy, Siberian Branch, Russian Academy of Sciences (IKFIA SB RAN); and other institutions made an enormous contribution to the studies of Antarctic auroras. The main results of the studies of Antarctic auroras, obtained by Russian scientists, are reviewed in this work.

The regularities of the variations in the IMF B-z component have been studied based on the data on the solar wind streams and their solar sources. Isolated solar wind streams such as magnetic clouds and shock layers before them, undisturbed heliospheric current sheets (HCSs), leading edges and bodies of high-speed streams from coronal holes (HSSs from CHs) have been considered. It has been revealed that each type of isolated streams in the interplanetary medium has it own features in the variations in the value and direction of the B-z component related to the stream immanent properties and conditions of propagation in the interplanetary plasma. The appearance of the southward B-z component is obligatory for all these streams which are, therefore, geoeffective.

More than 3800 measurements of the 630.0, 557.7, and 427.8 nm emission intensities have been statistically manipulated, and the dependences of the I 630/I 427.8 and I 557.7/I 427.8 ratios on the I 427.8 nm emission intensity have been obtained. The I 630/I 427.8 ratio decreases from 2 to 0.4 when the I 427.8 nm emission intensity increases from 0.1 to 3 kR. In the I 427.8 nm emission range 0.1–1.8 kR, the I 557.7/I 427.8 ratio tends to increase and takes the values 4.2–6.4. The experimental results have been confirmed by theoretical calculations. The obtained I 557.7/I 427.8 ratios suggest that the NO density at a maximum of its height profile is on the average 108 cm−3 in typical nighttime auroras.

The behavior of equatorward boundaries of auroral luminosity for different types of nonstationary solar wind streams is investigated. The average values of equatorward boundary locations of auroral luminosity in the midnight sector make: for solar flare-generated streams Ф

The ascafilms from Vostok observatory (phi' = -83.3 degrees) and the data on precipitating auroral particle fluxes measured on board the DMSP F6-F9 satellites from 1985 to 1990 were used to correlate the polar-cap auroral-arc position with different regions of particle injection. Seventeen simultaneous events are considered. Auroral arcs, observed at the geomagnetic latitudes from -82 degrees to -84 degrees in the morning setter (06:00-09:00 MLT) under quiet magnetic conditions, are shown to be present predominantly in the poleward part of the region of injection of particles whose characteristics correspond to those of plasma mantle particles.

Morphological features of magnetic impulsive events and the dynamics of associated ionospheric perturbances are examined from the magnetic records of high-latitude stations. Most impulses chosen for investigation display a bipolar variation in all the magnetic field components. Event occurrence patterns versus magnetic local time for different latitudes reveal the existence of large-scale and small-scale events. The large-scale magnetic events were observed typically throughout the wide latitude range from 67 degrees to 77 degrees corrected geomagnetic latitude (CGL). The occurrence pattern of events versus magnetic local time (MLT) shows a single maximum near 0900 MLT. Only about 12% of the impulsive perturbations were generated in the afternoon sector from 1200 to 1500 MLT. These events first moved westward as well as sunward up to noon and then tailward. The small-scale impulsive magnetic perturbations are generally registered in the midday sector from 700 to 74 degrees CGL. The small-scale events, like the large-scale traveling convectin vortices (TCVs), move westward with an average velocity of about 3.7 km/s. IMP-8 observations of interplanetary medium conditions were investigated for the large-scale events. The events occur mainly during periods when the B-x component of the interplanetary magnetic field (IMF) was northward. The events do not tend to occur for the highspeed solar wind flows or the flows with unusually high plasma densities. The IMF was frequently oriented close to the ecliptic plane and was almost radial along the Sun-Earth direction. Magnetic events were caused generally by the solar wind dynamic pressure variations with increases in amplitude from 20 to 140%. We suggest that pressure impulses may be amplified by a factor of 2 by the simultaneous change of the foreshock geometry from lying upstream of the prenoon magnetosphere toward dawn, when the IMF cone angle changes from <30 degrees to >30 degrees.