Permanent flare activity in the magnetosphere during periods of low magnetic activity in the auroral zone

Permanent flare activity in the magnetosphere during periods of low magnetic activity in the auroral zone

@332-0633/E% $3.00+0.00 Pergamon Journals Ltd. PERMANENT FLARE ACTIVITY IN THE MAGNETOSPHERE DURING PERIODS OF LOW MAGNETIC ACTIVITY IN THE AURORAL Z...

2MB Sizes 0 Downloads 7 Views

@332-0633/E% $3.00+0.00 Pergamon Journals Ltd.

PERMANENT FLARE ACTIVITY IN THE MAGNETOSPHERE DURING PERIODS OF LOW MAGNETIC ACTIVITY IN THE AURORAL ZONE V. A. SERCEEV*, A. G. YAHNINt, R. A. RAKHMATULIN~, S. 1. SOLOVJEV& F. S. MOZl?R//, D. J. WILLlAMS’i and C. T. RUSSELL** *Institute of Physics, University of Leningrad. Leningrad 198904, U.S.S.R., tPolar Geophysical Institute, Apatity 184200, U.S.S.R., ISibIZMIR, Irkutsk 664033, U.S.S.R., SIKFIA, Yakutsk 677007. U.S.S.R., /ISpace Science Laboratory, University of California, Berkeley, CA 94720, U.S.A., 7 Applied Physics Laboratory, The Johns Hopkins University, Laurel, MA 20707, U.S.A. and **University of California, Los Angeles, CA 90024, U.S.A. (Rrceiurd 26 May

1986)

Abstract-Auroral, magnetic variation and pulsation data from the dense network in the nearmidnight portion of the aurora1 zone are used together with the measurements of suprathermal particles and electromagnetic fields by the IMP-8 and ISEEspacecraft within the plasma sheet to study the characteristics of activity during two magnetically quiet periods on 3 March 1976 and 23 March 1979. Contrary to existing beliefs, we found clear signatures of numerous (5-10 events per hour) transient events, characterized by plasma flows, energetic particle bursts and E-B field variations. A close association of these events in the plasma sheet with the local aurora1 flares (LAFs) in the conjugate sector of the aurora1 zone is established for many events. We conclude that LAF (local aurora1 arc activation with associated Pi pulsations but extremely weak magnetic bays) have the same plasma sheet manifestations (apparently, the same physics) as the individual substorm intensifications during strong substonn expansion events, which differ from the studied quiet periods mainly by the strength and number of these intensifications. These transient phenomena seem to play an important role in the energetics of the quiet time magnetotail. 1. INTRODUCTION

In the past studies of the magnetotail processes, the investigators mainly paid attention to the substormassociated features. It has been indirectly suggested that in the absence of observable (mainly magnetic) activity in the aurora1 zone, the plasma sheet (PS) is not seriously disturbed and is cIose to some equilibrium state. However, few published results seriously disagree with this viewpoint. So, Coroniti et al. (1980) and Yahnin et ul. (1984) presented examples in which non-steady plasma flows and a strongly varied magnetic field within the PS have been observed during prolonged (a few hours in duration) quiet periods. In similar conditions, a few authors detected signatures of transient energy dissipation events, earlier described as features of strong substorm expansion in the magnetotail. This list of signatures includes the observations of rapid (both tailward and Earthward) plasma flow bursts and pIasma energization (Lui et al., 1976: Hones et al., 1976; Nishida et al., 1981), signatures of PS current disruption (Lui, 1978) and X-type neutral lines (Lui and Meng, 1979), transient electric field events (Cattell and Mozer, 1982) etc. Relevance of these signatures to the expansion-like process was, however, questioned (cf. Lui and Meng, 1979; Coroniti rr ul., 1980), because

of the well-known difficulties in the interpretation of the one-point measurements within PS made by a satellite. No regularly observed ground-based signatures of these events have been found yet. Whether these PS events are a rule or exception during quiet intervals is unclear. A spatial locality of these events was often supposed; then no observable response in the ground magnetic variations was explained by the gaps in the station network. However, the facts show that some of the substorm expansion signatures observed on the ground may occur without (with very weak) bay-like magnetic variations. So, in the very beginning of the substorm studies, Akasofu (1965) introduced a special tea-.~pseudo-breakup”-to distinguish those aurora1 break-up events in which no observable magnetic effects have been found immediately beneath the aurora1 bulge. Many such examples have been published before, two of them have been carefully studied by Sergeev and Yahnin (1978). These authors, as well as Rakhmatulin et ai. (1984) and Yahnin er ul. (1984), found Pi-type magnetic pulsations in the aurora1 zone, appearing in both temporal and spatial coincidence with the aurora1 flare. They noted that some Pi bursts appeared during the appearance or brightening of the discrete aurora1 arc, whereas in others the spatial defo~ation of this activated

1169

V. A. SERCEE~

1170

structure brought to it small-scale folding (to the small-scale amoral bulge). Hence, both Pi bursts and the activation of aurora1 arc (we shall use for these hereafter a special term-local uuroral flare--LAF) are still observed in many cases without other ground substorm signatures, as magnetic bays at aurora1 and mid-latitudes and midlatitude PiZs. We may suspect that the difference between the processes responsible for strong break-ups and LAF lies in their intensity. A detailed description of PS phenomena during LAFs is highly desirable to check this suggestion. The main questions studied in this paper are: (1) Whether the LAFs occurring during magnetically quiet periods have counte~rts in the nearly conjugate part of plasma sheet. (2) Whether LAFs have preferred situations in terms of substorm phases (recovery, growth, quiet period). (3) In which respect LAF-associated phenomena are similar or different as compared to the strong substorm activations.

2

OBSERVATIONS

2.1. Data sefection, insfrzimentat~on A basic requirement for such a study consists of the favourable location of a satellite and the ground network. A spacecraft must provide measurements during at least a few hours within the plasma sheet region in the central part of the tail, whereas the dense network of ground observatories must cover the highlatitude (67-72”CGLat) portion of the nightside aurora1 zone and provide the measurements of auroras, magnetic variations and pulsations in the region magnetically conjugated to the spacecraft position. The polar plots of station networks, used in our analysis of two quiet-time periods, are given in Fig. 1. Note that the latitudes of contracted amoral oval (677 7l”CGlat) are properly covered within the longitudinai sector 160-2iO”E (-3.5 h in M.L.T.). The time of the local magnetic midnight is - 15 U.T. thus at 205”E and - 18 U.T. at 160”E meridian; during the studied periods the network covered the nearmidnight portion of contracted oval. This network was in operation during two SYBERIA-IMS campaigns (see Erushenkov et al., 1976 for the details of networks and instrumentation). Two main elements are the meridional chains of stations at - 162”E (Norilsk chain) and at -2OO”E (Yakutsk chain), equipped with magnetometers, zenith photometers and other illstruments. The wide angle (60’ cone) zenith photometers, being sensitive mainly to the violet-green part of the aurora1 spectrum, are especially interesting for our purposes, together

et al.

with the induction magnetometers, installed at some stations of the network. The best coverage is available for the 3 March 1976 events, when four stations gave information from 699 71” latitudes in the sector of interest and data from three all-sky cameras are available. This period is a part of the lo-24 U.T. interval on this day, extensively studied by a special working group. Both the data collection (published in preprints by Sergeev et al., 198 1 and Mishin et al., 1982) and results of analysis of substorm events have been published before. Particularly, the overall review of ground and magnetotail activity has been given by Sergeev (1984) and an analysis of the quiet interval under the study by Yahnin et al. (1984); hereafter we shall refer to them as papers I and II, respectively. During these intervals, the IMP-8 and ISEEsatellites moved in the central part of the magnetotail. Their projections into the ionosphere are given by dotted lines in Fig. 1. Note that the latitudes of the spacecraft projections from such large distances are very sensitive to the details of the true distribution of the magnetotail currents at each moment. Their longitudes are much more conservative and do not change si~ni~cantly when we use different models (Sergeev and Tsyganenko, 1980). So, the poIar plots of s/c projections must be used mainly for qualitative purposes to determine whether or not the spacecraft projection is close to the ground array in longitude. In this study, the basic information about the plasma sheet processes is obtained by using the measurements of the magnetic field and suprathermal both ion and electron fluxes and anisotropy. At IMP8, the magnetic field data with 15.36 s resolution came from the GSFC magnetometer, and the EPD experiment (Roelof et al., 1976) provided the measurements of >50 keV protons and >30 keV electrons in I6 angular sectors of the ecliptic plane in 20.24 s. The magnetic field measurements at the ISEE1 spacecraft (Russell, 1978) are available with 3 s resolution. The detailed description of the MEPI experiment, which measured the energy spectrum and full 3-dimensional angular resolution in 36 s for ions and electrons of >20 keV energies, is given by Williams et al. (1978). The information on the strong plasma flows and the direction of the latter, as well as the information on the acceleration processes, may be extracted from the data on the suprathermal plasma. Additionally, the double probe measurements of electric fields (see Mozer et al., 1978) are available from the ISEEspacecraft with 3 s time resolution. The accuracy of m~surements expected is - l/2 mV m- ’ for the E, component and - 1 mV m - ’ for the E, component.

Permanent

flare activity

in magnetosphere

1171

3 March 1976

FIG.

I.

POLAR PLOTS (IN CORRECTED GEOMAGNETIC COORDINATES)SHOWING THE DISTRIBUTION OF GROIJNDBASED OBSERVATORIES DURING TWO STUDIED INTERVALS.

The used instruments are coded using the legend at the top of the figure. The standard abbreviations of the station names are used. Dotted lines (with marked hour marks) show the ionospheric projections of the spacecrafts along the magnetic field lines. A quiet state variant of cislunar model (Sergeev and Tsyganenko, 1980) is used for IMP-8, a “K, = 1” variant of the empirical model by Tsyganenko and Usmanov (1982),

for ISEE- projection. The dipole tilt effects are taken into account properly. GSM coordinates changes between [-36.5, 6.9, 0.11 and r-36.7, 5.6, 1.21 for 14-17 U.T. interval on 3 March 1976 (IMP-8, -4.2,3.2]for 15-20 ionospheric projection at -22 h M.L.T.), andbetween [- 19.7, - 1.9,4.9] and[-21.2, U.T. interval on 23 March 1979 (ISEE-1, ionospheric projection at -01 h M.L.T.).

3 March 1976, 14-17 U.T. According to the analysis, presented in papers I and II, this interval falls between very strong (until 2 1000 nT in AE) substorm events, whose expansion phases started at 11.51 U.T. and 17.15 U.T. respectively. Some residuals of the first substorm, which ended before 13 U.T., are still seen in the data. They include the diminished amplitude of the aurora1 zone currents (see AE index in Fig. 2a) and the recovery of the previously large fluxes of energetic electrons, which faded to the threshold values near 15 U.T. (Fig. 2b). At the same time (from 13.45 until 15 U.T.), the thermal plasma in the plasma sheet, previously energized during a substorm, was replaced by the new (cold but dense) plasma, in the direction from the outer part towards the central part of PS (paper I). The period 2.2.

14-15 U.T., therefore, has some properties of the recovery phase (although this term does not yet have an undisputable definition). The next hour, 15-16 U.T., was almost quiet in the sense of the observed magnetic variations (Fig. 2a). Finally, after 16.20 U.T., a growth phase of the next strong substorm started to develop (see paper I for complete information, including auroral, magnetic, polar cap convection and tail data). As regards the state of interplanetary medium, there were no measurements in the near-Earth region of solar wind. However, since the large solar wind velocity values have been published by King (1979) during the first hours of that day, we are sure that the solar wind had a comparatively large (presumably >600 km s-‘) velocity during the period of interest.

1172

V.

14 00

A.

15-00

SERGEEV

et al.

16.00

UT

Frc.2a.To~ BLOCK:COMMONSCALEBOTH AE(l2) INDEXANDSUPERIMPOSEDH-COMPONENTMAGNETOGRAMS OFFOURREPRESENTATIVESTAT~ONS (STO,KOT,CCS,UTA), CHARACTERlZEDTHEMAGNETIcACTIVlTY1NTHE SECTOR OF INTEREST (CGLat 67-71 N, CGLong 160-200 E).

Central and bottom parts-amoral intensity recordings, made by zenith photometers at Norilsk and Yakutsk meridional chains between 65 and 71 CGLat. The units are arbitrary, the photometers were intercalibrated only at Norilsk meridian. Vertical bars denote the onsets or intensifications ofpulsationsin Pi2 range, recorded at any station of network, shown in Fig. 1.

Permanent

flare activity

Electrons

Protons

(30

(50

Down - and

14:00

FIG.~~.

COUNT

in magnetosphere

1173

keV -=Zs.6.< 90

keV=C E, < duskword

200

keV)

flUx

I

,

15 00

16 00

“I,,

I u T.

RATESOF SUPRATHERMAL ELECTRONS AND PROTONS AND GSM COMPONENTSOF FIELD,MEASURED BY IMP-8 SPACECRAFT.

MAGNETIC

Count rates of electron detector are averaged over the satellite spin, whereas the count rates of proton detector are given for four directions in the ecliptic plane separately. For each of them the count rates are averaged over 3 adjacent 22.5 sectors; Earthward- and dawnward-directed fluxes are given by dotted lines. Similar to the Fig. 2a, vertical lines denote the onsets of Pi2 pulsations in the aurora1 zone.

An analysis of the low-latitude magnetograms showed the absence of pronounced SC or SI events; these data will be presented below. The zenith photometer recordings in the nearmidnight sector, presented in Fig. 2a, display during all this period the complicated variations with many bursts having a typical duration of 3-5 min. From the data of the Norilsk chain, for which the photometer recordings are scaled in the same units,

the aurora1 emissions are more intense at 69”CGLat (STE) than at 71” and 67” (ISA and UTA stations, respectively), as is expected for a quiet period. The bursty variations are mostly concentrated at 67-70” CGLat with a few exceptions of bursts, registered at Tixie Bay (TIX). In many cases, individual bursts are recorded at only one or two photometers and clearly represent the localized structures. The all-sky pictures, analysed for this period in paper II, showed the

1174

V. A. SERGEEV et al.

enhancements of brightness of individual discrete aurora1 arcs in association with most of these bursts. These activations were sometimes followed by the arc folding or a formation of small-scale surges. However, the correspondence between the mentioned arc intensifications and the spikes in the photometer recordings was not complete, partly because of the limited field of view of the all-sky cameras. Note also, that due to the wide cone angles of the photometers, variations of relatively weak but widespread diffuse glow often produced a more significant input than the intense brightening of very narrow aurora1 arcs. In this respect we must use the pulsation data to support our determination of aurora1 flares. The vertical bars in Fig. 2a denote the onsets or clear intensifications of magnetic pulsations in Pi2 range, recorded by at least one fluxmeter from our chain. In three-quarters of all bursts these moments coincide with the aurora1 intensifications. A correspondence is especially clear for the strongest spikes: 1, 5-7,8, 1 l14, 15-19. The strongest pulsations have been measured at 14:(15-19) and 16:(15-25) U.T. Only during these two intervals have Pi2 pulsations been detected at mid-latitudes (see records from Yakutsk in Fig. 4) and distinct negative magnetic bays with amplitudes 50-100 nT have been found at points STO, DUN and DUN, UTA respectively (all points are at -67” CGLat). As a result, this 3-h long interval covered a recovery, quiet period and initial stage of the growth phase of the next substorm, which displayed numerous local activations of the amoral arcs and Pi pulsations (numerous LAF events) in the absence of pronounced magnetic disturbance at the same stations. Next, let us consider IMP-8 observations-Fig. 2b. Concerning the magnetic field variations, two conclusions can be done without thorough analysis of the data. First, considering the small magnitude and multiple changes of the &-component sign, we see that for most of the time until 17:30 U.T. a spacecraft was deeply within PS; this is supported by the data on thermal plasma-see paper I. Second, an obvious feature of this interval is the large variability, seen in all three magnetic field components. This sharply contrasts with the absence of observable magnetic variations in the aurora1 zone. The fluxes of suprathermal protons and electrons also exhibit bursty variations. The proton fluxes are anisotropic all the time, but especially during the bursts. A dominant direction of anisotropic proton fluxes in the ecliptic plane is tailward along the tail and duskward across it. Whereas the burst character of the magnetic field variations and energetic particle fluxes agree qualitatively with the picture, inferred before from the

conjugate aurora1 regions, there are many difficulties in a one-to-one comparison of these data. Indeed, most of the ground-based inferred activations (vertical lines) fall closely to the times of PS bursts; however, this correspondence is not simple and exact. We emphasize here that the two strongest particle bursts occurred shortly after 14:15(l) and 16:15( 15) U.T., in coincidence with the strongest LAF events, seen in the aurora1 zone. These and a few other most exciting examples will be considered in more detail at the end of this section, after presentation of the data for the next interval. 2.3. 23 March 1979, 14:30-20:00 U.T. During this period, the Earth was embedded within the high-speed stream in the solar wind, whose frontal part produced strong disturbances, studied by the CDAW-6 working group. At the time of interest, the solar wind velocity was N 500 km s-l, the IMF magnitude - 10 nT, and B, (negative) was a dominant component of IMF. Using the IMF observations, made by ISEE- (235 R, upstream of the Earth), we calculated the GSM field components and the resulting variation of e-index (Akasofu, 1981) in the near-Earth region (delayed by 52 min-an average delay for propagation of the solar wind). This information is given in Fig. 3a together with the AE (13) index (Kamei and Maeda, 1982) and ISEEmeasurements within the plasma sheet. As for the first interval, this interval is bounded by two substorms, which started at 14:47 and 19:05 U.T. However, these substorm events are of small intensity (ZOC~O nT) and very different in character. A first event occurs without a significant input of solar wind energy into the magnetosphere (at least, as indicated by the E index). It was apparently triggered by a comparatively weak (lo-15 nT) SIf event, observed at midlatitude observatories at 14:37 U.T. and 2 min later at the ISEEposition (X = - 19 R, in the magnetotail). Ten minutes later, an expansion onset is evident in AE index and in many other phenomena (including aurora1 break-up, Pi2 pulsations and weak midlatitude bays, observed along the Yakutsk meridian). Only after 15:02 U.T. did ISEE- 1 enter into the expanding plasma sheet and detect here many substorm-associated phenomena, including enhanced energetic proton and electron fluxes and strong transient electric fields-Fig. 3a. Quite similar behaviour was observed during the second substorm, which started after 19:05 U.T. In both cases, the proton fluxes were highly anisotropic with the dominant plasma flow in the Earthward direction. As important feature, we mention here a clear multistep pattern in the development of both substorms,

1175

Permanent flare activity in magnetosphere 23 March I%‘9

I

Substorm

1

I

Substorm

Ef activity

HEP bursts 15bo

1700

1900

UT

FIG. 3a. FROM TOP TO THE BOTTOM :&-INDEX CALCULATED FROM ISEE- SPACECRAFT DATA; AU-AL(13) ~N~EX~FROMKAMEIANDMAE~A (1982): GSM B,-AUTJ B~~~M~~E~~I~ THETAIL(ISEE-I);IN~ERVALSOF I)tJRiNGAT LEAST 15s LONG INTERVALS); O~~I~IRECT~ONAL FLlJXES OF E, EVENTS (iE,l>lSmVm~' SUPRATHERMALPROTONS (Pl AND P4 CORRESPONDSTO 34 keV AND I20keV ENERGIES)ANDELECTRONS (31 keV AND 98 keV ENERGIES IN El AND E~CHANNELS); APPEARANCEOFTHE ENERGETICPARTICLERURSTSIS SHOWNRYLEGENDATTHEBoTTOMOFFIGURE. The bursts have been identified when the count rates rise by more than factor 2 in < 5 min; the vertical broken lines indicate the burst onsets; the black spaces show the intervals of fast rise in the count rate.

seen in a similar way in the ground-based observations. This is especially clear for the first substorm. Note, the maximal count rates of energetic particles as well as maximal positive B, values within PS have been achieved only at 16:15 U.T., 1.5 h after the expansion onset and -_ 1h after the onset of

recovery in AL index. Whether or not all this interval 14:47-X:15 may be considered as one substorm event is debatable. Similarly to the period on 3 March 1976, the weakly disturbed period from 16:30 till 19 U.T. can bedivided into three parts. The first one-16:30-17:30 U.T.-we

V. A. SERGEEVet al.

1176

HE portlcle burst activity (plasma sheet) I

II II

I

n

u

:

I

‘I

:

I’:

I

1 1

1600

1700

I800

1900

UT

FIG. 3b. FROMTOPTO BOTTOM:(a) HE PARTICLE BURSTACTIVITYCHART,RETAINED FROMFIG. 3a; (b) A PLOT, SHOWINGBY BLACKAREASTHE DURATIONAND INTENSITY0~ Pi1 (FREQUENCY RANGE0.3-3 HZ) PULSATION BURSTS,RECORDEDBY FLUXMETERS AT THREESTATIONSOF NORILSKCHAIN. After 19:15 U.T. the signal is contaminated by continuous PiC pulsations. (c) RECORD OF ZENITH PHOTOMETERSINSTALLEDAT~~~CGLAT.(~)MAGNETOGRAMSOFNORILSKMERIDIONALCHAINOFSTATIONS. For the information on the position of stations see Fig. 1. Note that vertical lines in Fig. 3b denote the onsets

of LAFs, recorded

just at Norilsk

chain,

denote a recovery stage, taking into account the tendencies of recovery of El and P4 energetic particle fluxes and BZ within the plasma sheet. Note that Pl channel measured protons, whose energy is close to (only 2-10 times higher than) the average energy of

and are not the same as in Fig. 3a.

thermal protons, respond mainly to the changes in their number density. On the contrary, El and P4 channels correspond to the high energy tail of the particle distribution. The next period, 17:30-18:20 U.T., can be characterized as a “simply” quiet period,

Permanent flare activity in magnetosphere the E parameter changes this time around a rather low value, - lOi erg s-i. The last period, after 18:20 U.T., occurs during the steady rise of both E and AUAL indices. The B, component in the tail has a tendency to increase (with the exception of a doubled sudden traversal of the neutral sheet); after 18 :56 U.T. ISEE- 1 gradually leaves the plasma sheet. At the same time, a clear indication of the equatorward expansion of the aurora1 oval is found from the magnetic data of zonal stations (cf. Fig. 3b). Thus, this interval (- 18:20 till 19:OS U.T.) has all the signatures of the growth phase. During all the period from 16:30 until 18:56 U.T., the spacecraft stays within the PS, as evidenced by large P 1 fluxes. A large number of the rapid increases in the fluxes of energetic particles is found and marked in Fig. 3a by vertical lines. We shall concentrate our attention on those bursts which occurred after 16:30 U.T. The bursts are clearly seen mostly in P4 and El channels; in half of the cases the bursts appear in both of them. Only in one case (10) is the HEP increase prominent in the 100 keV electron channel (E3). The examples of both Earthward and tailward flow bursts have been found during the more detailed analysis of MEPI three-dimensional data. So, the enhanced Earthward and duskward directed flows were dominant during bursts 11-12; the examples of strong tailward flows during bursts 13-14 will be given below. During processing of the E-field measurements we used here as “E-events” only cases with long and intense signals in the E, component (2 1.5 mV m-i during > 15 s). These events tend to appear during the strongest energetic particle bursts (mainly, during substorms). Only in two cases (1415) do they occur during a studied quiet period. As regards the behaviour of the magnetic field within the PS, it behaves more smoothly as compared to the IMP-8 observations at a further distance during the 3 March 1976 interval. The sudden changes are evident in association with few HE bursts (cf. numbers 8, 10, 11, 13, 14 and 15). In most cases B, stays in the positive range with the exception of burst 13, studied in detail later. The two center panels of Fig. 3b show a number of low-amplitude LAFs, seen during the quiet period and growth phase both as Pi1 bursts and small enhancements of aurora1 emissions at Cape Sterlegovo (STE). Note that during our quiet period, the small magnetic disturbances are most intense just over this station, as well as most of the Pi 1 bursts. The numbers, which mark these bursts, are retained the same as in Fig. 3a, if the LAF onset is within 4 min of HEP burst onset; the vertical lines here correspond to the LAFs’ onsets. In this case, where LAFs are not so frequent,

1177

as during the 3 March 1976 period, mutual correspondence of LAFs to the HEP bursts is very pronounced. One LAF, marked by letter (A), has no relevant HEP burst, the second new one (B) occurred at the break-up onset, when the satellite was outside the PS. Note, at 17 U.T. the satellite projection falls 2 h M.L.T. eastward of the Norilsk meridian and occurs at this longitude at 19 U.T. (Fig. 1). The amplitude of aurora1 brightening, recorded by the photometer during LAFs, is approximately 5-10 times lower than during main break-up (which produced a 300 nT negative bay) or during small substorm intensification at 16 U.T. (which produced only 100 nT bays at the nearest positions and no response in AL index). The LAF associated magnetic disturbances are hard to be identified in any case of weak LAF, if we use the usual magnetic field records (sensitivity 10 nT mm - 1 and record speed 40 mm h- ’ in our case). The last thing we must mention in this section concerns Pi2 pulsations. Only during the strongest events as the first and second substorm, have they been identified at midlatitudes. Their amplitudes were not strong (only 34 nT at 52” CGLat), but their period was unusually large (approx. 200 s in comparison to the nominal 40-150 s). As a result we failed to identify them surely on the existing background variation in the aurora1 zone and used here mainly the highfrequency (Pil) part of Pi pulsations.

2.4. Relevance of LAFs to DCF-current variations The sensitive (-0.5 nT mm-‘) and rapid-run (90 mm h-‘) magnetic records from two midlatitude stations (Yakutsk and Sverdlovsk, their magnetic midnights are at 16 and 20 U.T., respectively) are given in Fig. 4 for both studied intervals. In spite of the large distance between the stations, the H-component variations repeat one another in many details. Since similar short-term variations have been found on most of the low-latitude stations in different MLT sectors, their origin may be naturally attributed to the variations of DCF (Chapman-Ferraro) currents, induced by the changes of solar wind parameters. Recall that in both cases the solar wind was disturbed and had an enhanced (> 500 km s ‘) speed. As regards the role of these DCF changes in the LAFs’ generation, the direct influence of these solar wind inhomogeneities hardly plays a main role here. A frequency of LAFs is greater than a dominant temporal scale of DCF-variation; LAFs are more or less uniformly distributed over the intervals of both DCF rise and fall. Nevertheless, the inhomogeneities of IMF and solar wind shown by these observations

V. A.

1178

SERGEEV

et al.

3 March 1976

23 March 1979

17

16

1.3

UT

FIG. 4. HIGH-SPEED (90 mm h _ ‘) ANDSENSITIVE (0.5 nT mm- I) RECORDS OFMAGNETIC FIELD I%-COMPONENT FROMMIDLATITUDE (52” CGLat) STATIONS~VERDLOVSKANDYAKUTSK. The onsets of LAFs are indicated over the traces according to their determination in Figs 2a and 3b.

may play some role by disturbing

the plasma sheet in

the studied cases. 2.5.

Examples

individual LAF

of plasma sheet manifestations

during

events

Below we shall present some individual cases, in which: (a) the distinct examples of LAFs have been observed by the ground stations in the proximity of the ionospheric projection of the satellite; (b) plasma sheet signatures were distinct and illustrative; and (c) there were no problems in the timing of the event, which is well separated from the other ones. 2.5.1. A case of strong HE particle acceleration. A clear Pi2 event, 100 nT magnetic bay at nearby location and aurora1 break-up at 66”~$, (in the northern sky of Tixie all-sky camera), was observed at 14:15 U.T. This event has been discussed in detail in Paper II, so we shall only summarize its main features. The break-up and all associated ionospheric phenomena were of only 4-5 min in duration. Besides the strong aurora1 brightening evident in the record of photometer at Tixie (event l-Fig. 2a), the active aurora1 structure experienced a folding, which resulted in the small aurora1 bulge (with NS extent l”4,). A pronounced HE particle spike is evident at IMP-8 (Fig. 2b). The event started at 14:16.3 U.T., firstly as the abrupt rise of HE electron fluxes, especially at higher energies (until MeV range). Simultaneously, the beginning of both the changes of the ambient magnetic field and the burst of anisotropic HE protons were recorded (Fig. 2b); the dominant proton anisotropy was in the tailward direction. An estimate of the equivalent bulk flow speed by the method of Roelof et nl. (1976) gives for the maximum of the

proton event (14:19 U.T.) the equivalent speed I/~800 kms-‘. The main peculiarity of this event is the very hard spectrum and high intensity of energetic electrons. So, the integral flux of > 0.2 MeV electrons increased by approx. two orders in magnitude and exceeded the value 10’ (cm s sr))‘. Approximately the same maximal value was achieved during two neighbouring substorm events, in which the AE index exceeded 1000 nT. Since before and during this short-lasting event the &-component of magnetic field was 5 nT in magnitude and changed its sign both at 14:17.5 and 14:20 U.T., the observations have been made surely in the central part of PS. The energetic electrons during the burst had nearly isotropic pitch-angle distribution. This is an example of very efficient transient particle acceleration, which occurred in association with a distinct LAF event. PS signatures are found here with 1 min delay against LAF onset. BZin the central part of the PS turns briefly southwards at 14:20 U.T. for 1 min. 2.5.2. Weak and strong LAFs, observed in the central part of plasma sheet. LAF N14 in Figs 2 and 5 represents a case of a weak event, having no observable effect both in Pi2 at midlatitudes and in the magnetic variations in the aurora1 zone. On the contrary, these two signatures are pronounced for the group of successive LAFs N15-19, which caused the strong enhancement of the aurora1 brightness, observed by photometers (Fig. 2a), Pi2 at Yakutsk (Fig. 4) and 80-100 nT magnetic bay at UTA and DUN stations. Since the satellite was nearly in the same position (in the central part of the PS, since lBx\ 5 5 nT) at the initial phase of these events, this

Permanent

Etectrons

(3O
2 ,oo _ Protons (5OC ED<200

flare activity

in magnetosphere

keV1

keV)

e

5 0 $ s ; 4! z Fi f0

-5

5 0 -5

5 0 -5 16 05

16 00

FIG. 5.

AN EXAMPLEOFWEAK

16 IO

1620

16 15

(N14) AND MULTIPLESTRONG

325

UT

(NN 15-19)LAF SIGNATURE&OBSERVED OFPLASMASHEETON 3 MARCH 1976.

BY

IMP-8 IN THECENTRAL PART The particle count rates are averaged for all directions

in the ecliptic plane. The cases of tailward streaming electron angular distributions are marked by triangles.

gives us an opportunity to compare the PS effects during weak and strong LAFs. The main difference of these events is in the intensity of the associated HE particle fluxes, whereas the magnetic field behaviour and the character of the proton anisotropy (tailward in both events) are not very different. In association with the increase of particle flux at LAF onsets (lines 14, 15), B, turns southwards and then fluctuates together with other field components. Such behaviour of 8, is common for the substorm-associated tailward plasma flows (Caan et ~11.,1978; Nishida et a!., 1981). The temporal behaviour of HE proton and electron fluxes is different during the group of intense overlapping intensifications 15-18. The main increase of energetic etectron count rates occurs after the

maximal phase of the proton burst during the satellite traversal of the outer part of PS. A significant tailward streaming anisotropy of HE electrons (with back-toforth ratio exceeding 10) was observed during two sequences of measurements. Such behaviour of energetic particle fluxes and their anisotropies has been reported for a substorm-associated burst event by Kirsch et al. (1981). According to the observed anisotropy of energetic protons, our estimate gives the maximal flow speeds -500 and _ 1000 km s-* for events 14 and 15-18, respectively. Quite large bulk flows of the thermal protons are evident in the data of LASL plasma analysator (E. W. Hones, private communication). According to them, the large tailward flows occurred between 16:14.5 and 16:22 U.T. (during the main

1180

V. A. SERGEEVet al

proton burst) with the magnitude changing between 300 and 1000 km s- ’ . One measurement gave the tailward flow of 300 km s- ’ during the weak LAF event N14, this occurred at 14:ll U.T. near the maximal stage of associated HE proton burst. The time delays between the ground and PS signatures are not significant for events N14 and 1.5. 2.5.3. Weak LAF event with tailward flows at 20 R,. A distinct PilB spike was observed at the Nor&k chain at 18:27-18:28 U.T. (Fig. 3b, 6). The largest intensity of pulsations and the center of electrojet were found this time over UTA station, no associated aurora1 intensity increase was observed by the photometer installed northwards at STE station (Fig. 3b). The associated negative bay at UTA, if found, does not exceed 20 nT. This event apparently corresponds to the initial stage of the growth phase, since the tendency ofthe southward drift of the aurora1 electrojet is seen after 18 U.T.; the similar tendency is also indicated by the comparison of Pi-intensity pattern in the top block of Fig. 3b. This event is well defined in the data on both the magnetic field and HE plasma from the ISEEsatellite-Fig. 6. The first signature was the enhancement of the electron flux, which started approx. at 18:23 U.T. and existed during 4 min. The first 36 s frame of MEPI measurements (see the ecliptic cut at 18:23:39 in Fig. 6) indicates the tailward streaming anisotropy ofelectrons. The onset of the HE electron spike is followed by the magnetic field variations and slow increase of the tailward and duskward flowing protons. The maximal anisotropy during this event corresponds to 400 km s-i flow speed, if kTi = 6 keV is assumed. Note again the southward B, excursion in with the appearance of tailward association anisotropy of 30 keV protons. The ground-based signatures of LAF in this event are delayed by 4 min against the onset of HE particle burst in the plasma sheet, but occurred before the time when maximal anisotropy of protons (plasma flow speed) was achieved. 25.4. Tailward and Earthward plasma jlows and E,events during rapid traversals ofplasma sheet at 20 Re. A clear enhancement of Pi 1 noise-like pulsations has been observed by all three fluxmeters of the Norilsk chain at 18 h 42 m (55 _t 10) s on 23 March 1979. This time, a 20% increase of aurora1 luminosity was recorded by the photometer; a 40 nT negative magnetic bay started at STE and 50 nT positive bay at UTA (see Fig. 3b). ISEEprior to 18:43 U.T. was located in the northern boundary part of the PS, as indicated by the magnetic field and Pl channel data in Fig. 3a. Just at 18:43 ISEEstarted its 10 min

excursion into the southern outer part of PS, apparently caused by some large-scale (possibly, flapping) plasma sheet motion. The square-pulse B, variation associated with this NS then SN transition of satellite through PS, seen in Fig. 3a, indicates the tipping of plasma sheet with respect to the GSM equatorial plane. Approximatley 10” inclination angle is enough to explain both 5 nT amplitude of this B, square-pulse variation and the crossing of the PS center at ZN = +2.2 R, spacecraft position (a hinging distance of 8 R, was supposed in the calculation of this distance with respect to the neutral sheet plane). These effects must be considered in the interpretation of the BZ variation. According to the three-dimensionai pitch-angle information from the Pl channel of the MEPI experiment, prior to 18:43 U.T. (Fig. 7) proton fluxes had a moderate anisotropy, with maximal fluxes directed tailwards and duskwards and equivalent flow speed 100-200 km s-l. An order of value increase of both the omnidirectional fluxes and anisotropy (with approx. the same orientation) suddenly started at 18:43 :OO.A traversal of ISEE- 1 into the southern part of the PS started 12 s later, simultaneously with the appearance of impulsive electric field variation. Two minutes later, ISEEfinished its traversal into the southern boundary part of the PS, a weak field-aligned (tailward directed) anisotropy remained until the next PS crossing. During next traversal of the PS, taking place between 18:51 and 18:54 U.T., a new burst of PI count rates and increase of the anisotropy are evident. A reversal of anisotropy from tailward to Earthward direction is a remarkable feature of this period; the new anisotropy direction remained thereafter until the count rates fell down to very low values. When comparing the behaviour of E, and B, component variations during intervals of comparatively large E, values, their detailed anticorrelation during the first pass and correlation during the second one are clearly seen in Fig. 7. Such behaviour has been found in many substorm-related cases, with the sign of correlation related to the direction of the fast plasma flows (NC for tailward flows and PC for Earthward ones-see Nishida et al., 1983; Cattell and Mozer, 1984). Thus, the strong impulsive enhancements of plasma flows and the change of their direction from tailward to the Earthward ones are clearly evidenced by these data, taken together. An estimate of the Xcomponent of convection speed, calculated from measured E,- and ~~-component and inferred E,component [by assuming (E-B) = 01. gives for both traversals the value V, N 300 km s-r but in the opposite directions, being consistent with the

1181

“lnTn 18128

18:25

FIG&TOP

UT

El COUNTRATES AND PITCH-ANGLES OF THE MEASURED MEPI EXPERIMENT IN THE ECLIPTW PLANE (ISEE- 1).The vertical bars indicate the 90 local pitch angles, letters D. T, E mark the measurements of particles moving in duskward, tailward and Earthward directions. CENTRAL PART--B,- AND B,-COMPONENTS PART-THE ANGULAR DISTRIBUTIONSOF THE Pl. PARTICLES DURING THE SAME AZIMUTHAL SCANSOF

OF MAGNETIC FIELD IN THE SPACECRAFT COORDINATE SVSTEM(ISEE-I). BOTTOM PART-REWRD PULSATIONS f0.3-3 Hz) MADE AT UST-TAREYA (UTA) STATIUN.

OF THE Pil

The vertical line indicate the onset of Pi burst. anisotropy of proton fluxes. Note, this estimate gives only the cross-field component of flow. whereas the field aligned flow is often the strongest component {Caan et al., 1978). In this respect, our estimatesfor the maximal taifward and Earthward flow speeds, -500 to 600 km s-r, made by using the information on the PI proton anisotropy in the ecliptic plane, agree well with the inferred convection speeds for both their value and direction along the tail. By comparing GSM B= values during two neutral sheet crossing (say, when /B,/ -=z3 nT), we found them

within the ranges -0.2 to - 3.3 nT (first NS crossing, 7 points, (B,) = - 2.2 nT) and + 0.2 to + 5.7 nT (secand crossing, 22 points, (Bz) = t-3.6 nT). A predominance of negative B, during tailward flow and positive BZ during Earthward flow of plasma agrees with their usual distribution in substorm-related plasma jets (Caan et a/., 1978). During both crossings, a signiticant negative GSM B, component was found in the plasma sheet. Particularly it ranged within -9 to - 18 nT (first crossing) and -2 to - 10 nT (second crossing) in the

1182

V. A.

SERGEEV

et al.

1000 BOO 600 400 200 O 1800

60"

FIG. 7.

ELECTRIC

AND

MAGNETIC

FIELD

1853

16.50

1645

1642

MEASUREMENTS

(IN SPACECRAFT

COORDINATE

UT SYSTEM)

AND

SOME

Pl CHANNELOF MEPI EXPERIMENT(B~TH ON BOARD ISEE-1). Top part represents some azimuthal distributions of Pl count rates in the ecliptic plane (see Fig. 6 for explanation); below, the high resolution information is given. The vertical bars denote the maximal and minimal count rates during each 3-s spin ofdetector (P 1) as well as maximal and minimal pitch-angles (PAR) of measured particles during each scan. The rectangles show the intervals, when negative (NC) or positive (PC) correlation is evident between the changes of Bz and E, components. The Pi 1 bursts have been observed in the conjugate aurora1 zone at 18:42:50 and 18:56 U.T. INFORMATION

FROM

proximity of the neutral sheet. An appearance of such a large value of B,, especially during the first crossing (B, magnitude is close to the magnitude of lobe field), gives evidence on the significant perturbations of the PS magnetic configuration, at least of local significance. The observed sequence of the fast plasma flows and

the magnetic variations within PS resembles the sequence seen often during substorms. The source initially was somewhere between lo-20J7,. It produced the fast tailward plasma flow; the positive B, excursion during the first minute (see Fig. 7) may be associated with the frontal part of the plasmoid, whereas in its wake B, is mainly in the southward

Permanent flare activity in magnetosphere range, even in the PS center. Thereafter, the direction of the flow is suddenly reversed to the earthward one together with the increase of B,. The main difference is the absence of significant thinning and expansion of the PS, whose signatures in our case may be contaminated by the large scale PS motions of the flapping type. The important peculiarity of the studied event is the extreme weakness of the processes, leading to the acceleration of particles in the high-energy tail of particle distribution. The associated bursts are seen only for 30 keV electron energies (El in Fig. 3a), but not for both protons and electrons of _ 100 keV energy (P4 and E3). Whereas the first event occurred in exact coincidence with a well-defined LAF in the conjugate aurora1 zone, the next LAF was detected only L5 min after the second neutral sheet crossing at 18:56 U.T. (N15 in Fig. 3b). Its relevance to the second event in the plasma sheet is not clear, although the tailward movement (or jump) of the source of fast plasma flow during the second event may help to understand this increasing time delay.

3. DISCUSSION

3.1. Summary of observations During two studied periods, characterized by both the weak global magnetic activity in the aurora1 zone (AE changed around 50-100 nT) and the absence of well-defined substorm events, numerous aurora1 flares (activations of aurora1 arcs accompanied by Pi pulsation bursts) have been identified in the nearmidnight portion of aurora1 zone. In most cases, Pi2 pulsations have not been identified at midlatitudes and no (or very weak) magnetic bays have been found immediately beneath the LAF region. Clear signatures of LAF-associated activity have been observed. According to the results of our study and of the companion paper by Yahnin et al. (1984), the main findings may be summarized as follows: (1) LAF is a short-lasting (usually 3-5 min in duration) and localized (sometimes < 40” CGLong) phenomenon. Five to ten events per hour have been identified in our cases. (2) LAFs occur similarly during periods of both growth phase and recovery phase, as well as during the intermediate quiet periods. No direct relationship between the appearance of LAFs and DCF-variations are found. (3) A clear relationship between individual LAF events in the aurora1 zone and signatures of short-lived dissipation events within PS are found in many cases during the observations in the nearly conjugate portions of PS and aurora1 zone. This relationship is

1183

quite clear for the second interval (23 March 1979), when individual LAFs are more isolated in time. The onsets of LAFs and associated PS phenomena do not exactly coincide, but usually lie within 24 min one from the other. (4) The mentioned signatures of LAF-associated dissipation events include bursts in both the plasma flow (a few hundred km s - ‘) and the flux of accelerated energetic particles. The concurrent short-term magnetic variations and, sometimes, bursts of electric field, are also observed. The observations of southward magnetic fields are not uncommon within the tailward flowing plasma, even within the central part of PS. Both the isotropic and, rarely, streaming energetic electron angular distributions (only in the outer part of the PS) have occurred during the bursts. (5) According to the character of the observed anisotropy of suprathermal protons, the sources of LAF-associated events usually lie earthwards of 30 R, distance in the nightside plasma sheet (the tailward flowing protons during the bursts dominate at IMP-8 distance). The cases of both the earthward and tailward proton anisotropy have been found at 20 R, in the magnetotail plasma sheet. When interpreting the time variations of the 30-100 keV suprathermal particle fluxes, we must bear in mind the following difference between the main sources for both sorts of particles. Since the kinetic energy of proton (ion), moving with the speed of a few hundred km s-l, is comparable to the average thermal proton energy, strong plasma flows of such magnitude bring to the rise of both the omnidirectional fluxes and the angular anisotropy of suprathermal protons (Roelof et al., 1976). LAF-associated proton bursts in our cases typically display a strong anisotropy, whose association with rapid thermal plasma flow and rapid convective (E x B) flow have been verified directly for few cases. For this reason, we will refer to the highly anisotropic proton bursts simply as “plasma flows”, although the density gradients and changes in true energy spectrum (in the reference system tied to the moving plasma) can also contribute to them. At the same time, plasma flows themselves cannot change seriously the measured fluxes of suprathermal electrons. In the numerous cases where the satellite stayed in the central part of the plasma sheet both prior to and during the burst, the appearance of an energetic electron burst signifies the switch on of some acceleration process. Typically their pitch angle distributions are very close to the isotropy. In more than half of all bursts in Figs 2b and 3a (see also Figs 5, 6), the bursts in suprathermal electrons and protons have appeared together. The simultaneity of the responsible for both the bulk flow processes,

1184

V. A.

SERGEEV

acceleration of plasma and its energization, follows naturally from this coincidence. The list of plasma sheet signatures of LAFs [point (4)], documented by our data, which covers all essential findings of substorm-like phenomena during quiet periods, appeared before in the literature (Lui et al., 1976; Hones et al., 1976; Lui and Meng, 1979; Nishida et al., 1981; Cattell and Mozer, 1982). In most cases, authors have presented only the individual short-term events as examples of the phenomenon, which occurred unexpectedly during the quiet time. Much more interesting is the comparison with the results by Coroniti et al. (1980), who documented the different observations, made by IMP-7 within the PS during a 4-h quiet period on 27 October 1972. The satellite position and the aurora1 magnetic disturbance were nearly the same as in our 3 March 1976 interval; the main findings of Coroniti et al. (1980) agree in all essential features with our observations. In a similar way (see their Fig. 3), they observed the variable patterns of both the magnetic field and plasma flow. The bulk flow of thermal plasma exhibited the bursty character with peak values 300-700 km s-l, a few bursts of energetic protons and electrons are seen also in the data presented. Most of these flow bursts were in the tailward direction, but dawnward and earthward flows have been found in few cases. A typical flow burst duration of lo-15 minis cited in this paper. Two intervals analysed in detail in our paper give a strong support to the conclusion of Coroniti et al. (1980), that “even at.quiet times the plasma sheet is apparently quite dissipative”. An example of the thermal plasma measurements within the PS made during the inbound pass of IMP-6 at distances 20-30 R, during the very quiet period of 4-5 October 1971 has been given by Hones rt al. (1976). Besides the example of the event (starting at 01:20 U.T.), which displayed strong plasma energization and earthward flows - 1000 km s-l, their Fig. 8 shows a large number of more weak (maximal amplitude - 300-500 km s- ‘) flow bursts, directed both tailwards and earthwards. As follows from these comparisons, the plasma flows during the bursts may be in either direction at 20-30 R, distance, but the tailward flows are much more frequent at -35 R, in the plasma sheet. According to these findings, we feel that the most probable location of the source of these flows is distance. Only one apparently at 15-30 R, measurement of flow burst (earthward flow) is available at 11 R, in the magnetotail (Lui et a/., 1976). The past evidence on the ground-based manifestations during these short-lived dissipation events are scarce, since most of the cited authors limited themselves by considering only the standard

et al.

magnetograms and AE index showing no response. In a few papers authors reported P-2 type pulsations, occurring concurrently with the most exciting bursts (Hones et al., 1976; Nishida et al., 1981). More detailed information was presented for the strong dissipation event on 17 November 197 1, seen in the PS at 11 R, (Lui et al., 1976). Simultaneously with the onset of this event, the appearance (intensification) of the aurora1 arc was observed at _ 70”$, (Inuvik) in the region being within 1.5 h M.L.T. of the satellite projection; also the large-period (300 s) Pi-pulsations have been found both at aurora1 and midlatitudes at the nightside (Saito, 1979). These results agree with our observations and results of paper II, showing that both the irregular pulsations and the aurora1 arc activations are the typical ground manifestations of these short-lived PS dissipative processes. Note that all cases reported before, which had associated Pi2 response at midlatitudes, belong to the class of rather strong LAF’s, for which V > 400 km s-l A correspondence between the intensities of these PS and ground signatures will be discussed later. Lui and Meng (1979) failed to find any aurora1 break-up signatures during the observations of looplike magnetic configurations (negative BZ in the proximity of neutral sheet) at - 30 R, in quiet periods. Two reasons could lead to this negative conclusion. Firstly, the used DMSP photos have been obtained at different time, displaced by more than 10 min from the PS observations, whereas the typical duration of LAFs is - 5 min (cf. Fig. 2a). Secondly, as has been documented by Yahnin et al. (1984) and Rakhmatulin et al. (1984), the signatures of aurora1 bulge, searched by Lui and Meng (1979), are not the common features of LAF-associated aurora1 activations. The arc appearance or activation is often the only aurora1 signature of these events, as in the case reported by Lui et al. (1976). Such types of activated aurora1 structures are not uncommon during very quiet periods. In four of five representative examples of DMSP photos, presented by Makita and Meng (1984), we see the localized active portions of aurora1 arcs without surge-like deformations, having typically 5 2 h M.L.T. length in longitude. According to the statistical results of Berkey and Kamide (1976), such structures are usually observed in the near-midnight M.L.T. sector even for very low activity (say, AE < 25 nT). A few examples of the small-size surges have been also presented by the latter authors, for one of them (10 December 1972) no magnetic response occurred at the station (Dixon), which was immediately beneath the small surge. Most of the discussed localized active structures appeared at the latitudes between 67 and 7 1‘4,

Permanent flare activity in magnetosphere (Berkey and Kamide, 1976) and fall into the equatorward part of the region of electron precipitation, apparently into the hard zone (Makita and Meng, 1984). Thus the field tubes, containing LAF structures, seem to be associated with the nearEarth portion of plasma sheet. This conclusion does not contradict the recent empirical models of the magnetospheric magnetic field (Tsyganenko and Usmanov, 1982): according to their “K, = 0” model, the point at 15 R, distance at the midnight equator is projected to 70”#,, into the ionosphere. The small difference between these results of field line projection and the inferred range 15-30 RE of LAF sources in the magnetotail is not a serious disagreement, since the results of the projection are sensitive to the real momentary value of the current density in the tail. The difference in the projections from case to case within &2” is possible due to the variations of dynamic pressure of the solar wind during quiet periods (A. V. Usmanov, private communication). A good correspondence of both the aurora1 arc activation and generation of Pi pulsation to the transient dissipation events in the magnetotail is not amusing, since both of them are thought to originate due to the Alfven waves, generated by impulsive changes of EM field and current in the plasma sheet (cf. Baumjohann and GlaRmeier, 1984; Sato, 1982). The time for a propagation of the Alfven wave to the ionosphere together the propagation of a disturbance within the PS from the source point to the spacecraft point will form the time delay between the registration of ground-observed and PS signatures. According to our calculations, based on the semiempirical magnetospheric models (Sergeev and Tsyganenko, 1980). the time for travelling of the Alfven wave from the equator to the ionosphere is within 0.5-2 min for N = 0.1 cm-’ and 1.56 min for N = 1 cm-’ (these are extremal PS densities for quiet periods), when the source lies in the range X = - 15 to -30 R,. Thus the observed delay (5 4 min) falls well within this range of possible time delays. The nearly simultaneous registration of both PS and ground signatures, as in the cases of Figs 5 and 7, is possible when both time delays (source-ionosphere, source-spacecraft) are comparable (the source lies between the Earth and spacecraft). A slight discrepancy is evident concerning the mentioned duration of LAFs (335 min, point 1) and lo--l5 min value, cited by Coroniti et al. (1980). One possible explanation is that often a few LAFs are grouped in time, as evident in many examples in Fig. 2. Completing this section, we conclude that the frequent occurrence of short-lived dissipation events in the plasma sheet and LAF structures in the near-

118.5

midnight portion of aurora1 oval during quiet times, their mutual correspondence, as well as the concrete features of both these phenomena seem to be well established and supported by much evidence. 3.2. intensity relutionship between difirent LAF- and substorm-associated signatures LAF-associated dissipation events are quite localized phenomena, which appeared sporadically in different parts of near-midnight portion of the aurora1 oval (cf. paper II and Fig. 2a, for example). The variability of magnetospheric signatures, seen during consecutive events in our cases, is possible due to the different position of the spacecraft with respect to the source region. In spite of these difficulties, our data apparently allow us to establish the relationship between the strength of magnetospherjc and ionospheric manifestations of LAFs and substorms. Two events on 3 March 1976 (Nl and 15-18) produced the strongest flows and HEP bursts. These two, disucssed in detail in sections 2.4.1-2.4.2, differ from other events also in ground signatures. Only these ones produced 50-100 nT magnetic bays under the LAF regions and strongest Pi pulsations in the aurora1 zone. seen even at midlatitudes. The associated aurora1 luminosity spikes are also the strongest ones. A clear illustration of the difference between weak and strong LAFs is given by Fig. 5, where both events (14 and 15) have been observed in the same central part of the PS. Similarly, the event N5-7 on 23 March 1979, which also produced a 100 nT bay under the LAF region, is intense in all kinds of data, as compared to the subsequent LAFs N 9-15. The difference in intensity of these LAFs as compared to the substorm activations on that day is also quite pronounced in all kinds of data, although these substorms (especially the first one) were of rather low intensity. Thus, we feel that the intensities of both the magnetospheric and ionospheric phenomena change in parallel in the range covered by local am-oral Bares and substorm activations of different strength. The reasons why not all known substorm signatures are seen during LAFs and some substorm intensifications, are explained by the different signalto-noise ratio in the different phenomena. The midlatitude magnetic bays, produced by the current wedge system, are apparently the most crude indicator of substorm-like events. Since at 4 - 40” a bay amplitude is N l/20 of the negative bay amplitude in the aurora1 zone, this signature fails to serve already at 100-200 nT amplitude of AL-index increase, since the corresponding effects are in the range of the background variations, like those in Fig. 4. Both Pi2 at

1186

V. A. SERGEEVet al.

midlatitudes and the magnetic bays at the proper place in the aurora1 zone are still working. According to our results, the flares of aurora1 arc intensity and Pi pulsations in the proximity of LAF region still provide the important information and serve as the indicator of short-lived dissipation events even for those weak events, when both Pi2 at midlatitudes and magnetic bay-like variations could not be distinguished surely. 3.3. Similarity and difference between LAF and substorm expansion Since the main physical process responsible for the phase is not identified substorm expansion undoubtedly yet, and no adequate quantitative physical model exists for it, we must rely in our discussion only upon its main characteristics, known from the experiment. Among the most important ones are the temporal and spatial scales of the compared phenomena and the lists of their manifestations in both the plasma sheet and the ionosphere. The main differences between LAFs and expansion phases of the strong substorm events are believed to be in the intensity and in both the spatial and temporal scales, since the lists of their local manifestations do not differ drastically. The difference in intensity exists really, but we shall further be interested in the qualitative comparisons of both processes. The changes in the properties of plasma and in the configuration of the magnetic field during substorms are thought to occupy the whole magnetosphere and occur during a larger time scale-about 1 h (Baker et al., 1984). However, these event do not develop monotonously, but in a step-like fashion. These steps, known as substorm intensifications (Rostoker et (II., 1980; Baker et al., 1984), occur clearly in the localized portions of the plasma sheet and the aurora1 zone. The total global changes seem to be realized as the accumulation of the effects of these localized and short events (Sergeev and Tsyganenko, 1980; Sergeev, 1981). Thus, we must compare LAFs with those “elementary events” that constituted the substorm expansion. In the recent detailed studies of substorm signatures (both in the PS and in the ionosphere), an “elementary” time scale of intensification was inferred to be as short as l-3 min (the so-called impulsive structure of expansion-see Sergeev and Yahnin, 1979; Sergeev, 1981). Whether or not this impulsive structure is discernible in the data depends mainly on the used time resolution. position of spacecraft, and abilities to differentiate the variations of temporal and spatial origin (the dense ground network is especially important for this purpose). A list of ionospheric and plasma sheet signatures, referred to the “impulse” of

explosion process during a substorm, is almost the same as was found for LAF in our paper (see the recent results and discussion on this topic in Sergeev et al., 1986a,b). It seems quite interesting to note that occasionally the LAF-like intensifications (aurora1 brightening without strong arc deformation, Pi2 pulsations without associated magnetic bay on the ground, magnetic field variation and energetic particle burst in the PS) are imbedded between the usual strong impulses of expansion (see, for example, impulse E in the paper by Sergeev et al., 1986a). As follows from this comparison, the concrete manifestations of LAFs and “‘elementary intensification” during substorms have no drastic qualitative differences, as well as their temporal and spatiai scales. It is quite probable that their main differences lie in different intensity and occurrence frequency of these “elementary intensifications”. The larger intensity and frequency of these events during strong substorms will result in the higher total energy dissipation and, probably, in the accumulation of the effects, which produces global changes in both the configuration and properties of magnetic field and plasma in the magnetotail. An accumulation of the effects may produce the new observed phenomena in the ionosphere, like the magnetic effects of the current wedge system, the formation of aurora1 bulge etc. Thus, the main differences between the basic processes responsible for LAFs and substorms may be rather quantitative, but not qualitative, ones. 3.4. Permanent occurrence of LAFs during quiet periods und its consequences According to our data and the results, presented by Coroniti et al. (1980), LAF-associated events are more or less evenly distributed over the studied quiet periods, covering similarly the growth and recovery phases and intermediate intervals. The appearance of LAF-like active parts of aurora1 arc is a typical feature of aurora1 displays, photographed by DMSP satellites during quiet periods (Berkey and Kamide, 1976; Makita and Meng, 1984). Many examples of LAFevents, occurring during the growth phase of substorms, may be found in the literature (cf. Untiedt et al., 1978 and Morse et al., 1983 for most excited examples). The appearance of these events does not depend crucially on the level of the magnetospheric convection (say, on the level of AE index), since the similar events are reported to occur for such consistently low AE values as 20 nT (Berkey and Kamide, 1976; Lui et al., 1976; Hones er al., 1976). The reported LAF-like events provide nonnegligible energy dissipation within the PS. Coroniti et nl. (1980) noted that the plasma energy transport rate

1187

Permanent flare activity in magnetosphere associated with the flow bursts may reach 0.1 erg cm-’ s-r during the bursts. For estimates, we choose V = 400 km s-r and N = 0.5 cm-3; then we have l/2 n.m;V3 = 3 x lo-’ erg cm-’ s-r. The minimal longitudinal extent of aurora1 flare (30”) corresponds to -5 R, cross-tail dimension of the LAF region in the magnetotail. By multiplying. this value of the energy flux by the cross-section 5 R, (PS width) x 5 R, (cross-tail size) occupied by the flow, we obtain 3 x 10” erg s-l energy dissipation rate. For 5 bursts per hour and 5 min duration of each burst we have a total energy dissipation 0.5x lO*l erg h-r. This estimate represents rather the lower limit, since we did not take into account the energy carried by the opposite directed jet and the energy associated with the plasma energization. Therefore, the energy dissipation associated with LAF events is clearly the very important factor in the energetics of quiet-time magnetosphere, where the total energy dissipation must be significantly lower than lo*’ erg h-‘, characteristic for substorms (Akasofu, 1981; Baker et al., 1984). The source of the energy for these events (like during substorms) may be the magnetospheric dynamo. As we know, the electric potential difference across the polar cap or aurora1 regions never falls below l(r20 keV (cf. Reiff et al., 1981), even during the quiet periods. One of the real consequences of these phenomena may be the formation of the hard zone of precipitated electrons having the average energy 0.5 keV. This narrow zone is seen during each quiet time pass of it delineates the equatorward polar satellites, boundary of the wide region of soft (- 100 eV) electron precipitation (Makita and Meng, 1984). Spatially, the hard zone at the nightside nearly coincides with the zone of LAF appearance. The new feature following from our analysis concerns the new approach to the problem of the mechanism of the short-lived dissipation events. These events seem to play a similarly important role by controlling both the dynamics and energetics of the plasma sheet processes during substorms, as well as during any other states of the magnetotail (including quiet periods). So far, we arrive at the view of the substorm expansions as the periods with similar basic physics but which differed by the larger intensity (and/or higher frequency rate) of the short-lived dissipation events. This view is drastically different from the existing threshold-type models of substorms (Galeev, 1982). The occurrence frequency of the LAFs during different quiet intervals is not the same, as well as their intensity (compare two studied intervals). The factors controlling these parameters must be clarified in the future studies.

4. CONCLUSIONS

The explosion-like energy dissipation events in the magnetotail, which resulted in transient electromagnetic field variations, plasma flow and acceleration, are generally believed to be related to the substorm periods, but not to the periods of the magnetic calm in the aurora1 zone. Contrary to these expectations, during quiet periods we found evidence of numerous such events, which occurred in the plasma sheet at 2040Re distance. The close association of these short-lived events to the local aurora1 flares (aurora1 arc activation with associated transient Pi pulsations) is indeed well established. These events are evenly distributed over the quiet intervals with (or without) the recovery-type and growth phase-type features, seen in the magnetotail and in the ionosphere. The LAF-associated processes may play a significant role in the energetics of the quiet time magnetosphere. A close similarity in the manifestations (indeed, in the physics) of both LAFs and elementary events of the substorm expansion phase may lead us to the simplified view of the substorms as the periods with enhanced frequency and intensity of these short-lived explosions, which results in the enhanced global energy dissipation rate in the magnetotail and in the large-scale changes of the configuration of the magnetotail plasma sheet and magnetic field. At the same time, a close association of the tail energy dissipation events (plasma flows, particle bursts etc.) with the appearance or intensity variations of the aurora1 arcs during quiet periods, indeed shows an important role of these plasma sheet processes in the generation of the aurora1 structures. The processes in the quiet time plasma sheet are not properly understood today; this area still awaits a future thorough study.

Acknowledgements-The magnetic field measurements on board the IMP-8 spacecraft were obtained by courtesy of Drs N. F. Ness and R. P. Lepping. We are grateful to Dr E. W. Hones (LANL) and Dr A.T.Y. Lui (APL/JHU) for the information from their thermal plasma and high-energy plasma experiments on this spacecraft. The ground-based observations in Eastern Syberia during SYBERIA-IMS campaigns, used in this paper, have been made available due to the participation of many colleagues from IKFIA and SibIZMIR institutions. We thank also Dr E. J. Smith fJPL) for kind permission to use magnetic field data from the I‘SEE: 3 spacecraft. We are grateful to Drs B. Klayn and L. Ivanova for their permission to use the magnetic records from Borok and Sverdlovsk observatories, and to Mrs L. Nemtseva, T. Roldugina and M. Holeva for their assistance in the preparation of the manuscript.

V. A.

1188

SERGEEV

REFERENCES

Akasofu, S.-I.(198 1) Energy coupling between the solar wind and the magnetosphere. Space Sci. Rev. 28, 121. Akasofu, S.-I. (1965) Dynamic morphology of auroras. Space Sci. Rev. 4, 498. Baker, D. N.,Akasofu, S.-I., Baumjohann, W., Bieber, J. W., Fairfield, D. H., Hones, E. W., Mauk, B., McPherron, R. L. and Moore, T. E. (1984) Solar-terrestrial physicspresent and future, in Subsronns in the Magnetosphere (Edited by Papadopoulos, K. and Butler, D.), Chapter 8. NASA, Washington, D.C. Baumjohann, W. and GlaDmeier, K.-H. (1984) The transient response mechanism and Pi2 pulsations at substorm onset. Review and outlook. Planet. Space Sci. 32, 1361. Berkey, F. T. and Kamide, Y. (1976) On the distribution of global auroras during intervals of magnetospheric quiet. J. geophys. Res. 81, 4701. Caan, M. N., Fairfield, D. H. and Hones, E. W. (1978) Magnetic fields in flowing magnetotail plasmas and their significance for magnetic reconnection. J. geophys. Res. 84, 1971. Cattel1.C. A. and Mozer, F. S. (1982) Electricfields measured by ISEE- within and near the neutral sheet during quiet and active times. Geophys. Res. Lett. 9, 539. Coroniti, F. V., Frank, L. A., Williams, D. J., Lepping, R. P., Scarf, F. L., Krimigis, S. M. and Gloeckler, G. (1980) Variability of plasma sheet dynamics. J. geophys. Res. 85, 2957. Erushenkov, A. I., Ponomarev, E. A., Rakhmatulin, R. A. and Urbanovich, V. D. (1976) Issled. po geomagn. Aeron. i phys. Solntsa (in Russian) 39, 137. Galeev, A. A. (1982) Magnetospheric tail dynamics, in Magnetospheric Plusma Phys. p. 143. Reidel, Tokyo. Hones, E. W., Bame, S. J. and Asbridge, J. R. (1976) Proton flow measurements in the magnetotail plasma sheet made by IMP 6. J. geophys. Res. 81, 227. Kamei, T. and Maeda, H. (1982) Aurora1 electrojet indices (AE)‘for January-June 1979. WDC-C2 for Geomagnetism, Data Book N 5, April 1982. King, J. H. (1979) Interphnetary medium duta hook, Supplement I. NSSDC/WDC-A. Kirsch, E., Krimigis, S. M., Sarris, E. T. and Lepping, R. P. (198 1) Detailed study on acceleration and propagation of energetic electrons and protons in the magnerotail during substorm activity. J. geophys. Res. 86, 6727. Lui, A. T. Y. (1978) Estimates of current changes in the geomagnetotail associated with a substorm. Geophys. Res. Lett. 5, 853. Lui, A. T. Y., Akasofu, S.-I., Hones, E. W., Bame, S. J. and McIlwain, C. E. (1976) Observation of the plasma sheet during a contracted oval substorm in a prolonged quiet period. J. geophys. Res. 81, 1415. Lui, A. T. Y. and Meng, C.-I. (1979) Relevance of southward magnetic fields in the neutral sheet to anisotropic distribution ofenergetic electrons and substorm activity. J. geophys. Res. 84, 5817. Makita, K. and Meng, C.-I. (1984) Average electron precipitation patterns and visual aurora characteristics during geomagnetic quiescence. J. geophys. Res. 89,2861. Morse, T. H. and Romick, G. J. (1982) The fluctuation and fading of aurora1 arcs preceding aurora1 substorm onsets. Geophys. Res. Lett. 9, 1065. Mozer, F. S., Torbert, R. B., Fahleson, U. V., FBlthammar, A. and Pedersen, A. (1978) C.-G., Gonfalone, Measurements of quasistatic and low frequency electric

et al.

fields with spherical double probes on the ISEEspacecraft. ISEE Tram Geosci. Electron. GE-16, 252. Nishida, A., Hayakawa, H. and Hones, E. W. (1981) Observed signatures of reconnection in the magnetotail. J. geophys. Res. 86, 1422. Nishida, A., Tulunay, Y. K., Mozer, F. S., Cattell, C. A., Hones, E. W. and Bim, J. (1983) Electric field evidence for tailward flow at substorm onset. J. geophys. Res. 88,9109. Rakhmatulin, R. A., Parkhomov, V. A. and Lukovnikova, V. I. (1984) On the appearance of irregular Pi pulsations during the growth phase of a substorm. Magnetospheric Res. (in Russian), 5, 111. Reiff, P. H., Spiro, R. W. and Hill, T. W. (198 I) Dependence of polar cap potential drop on interplanetary parameters. J. grophys. Res. 86, 7639. Roelof, E. C., Keath, E. P., Bostrom, C. 0. and Williams, D. J. (1976) Fluxes of 50 keV protons and 30 keV electrons at 35 RE. 1. Velocity anisotropies and plasma flow in the magnetotail. J. geophys. Res. 81, 2304. Russell, C. T. (1978) The ISEE 1 and 2 fluxgate magnetometers. IEEE Trans. Geosci. Electron., GE-16, 239. T. (1979) Pi2 pulsations associated with Saito, magnetospheric substorms, in Mugnetospheric Study, 1979 (Edited by Japanese IMS Committee, Tokyo), p. 104. Sato, T. (1980) Aurora1 plasma physics, in Magnetospheric PIusma Physics (Edited by Nishida, A., Center for Academic Pub]., Japan). Reidel, Tokyo. Sergeev, V. A. (1981) High-time resolution correlation between the magnetic field behaviour in the magnetotail plasma sheet and ground phenomena during substorm expansion phase. J. Geophys. 49, 176. Sergeev, V. A. (1984) Generalcharacteristicsofthe substorms on March 3, 1976. Magnetospheric Res. (in Russian) 5, 10. Sergeev, V. A. and Yahnin, A. G. (1979) The features of aurora1 bulge expansion. Planet. Space Sci. 27, 1429. Sergeev, V. A. and Yahnin, A. G. (1978) Correspondence of expansion phase signatures. Geomagn. Res. (in Russian) 24, 78. Sergeev, V. A. and Tsyganenko, N. A. (1980) The Eurth’s Magnetosphere. Nauka, Moscow. Sergeev, V. A., Pellinen, R., Basinger, T., Baumjohann, W., Stauning, P. and Lui, A. T. Y. (1986a) Spatial (and temporal) characteristics of impulsive structure of magnetospheric substorm. J. Geophys. (accepted). Sergeev, V. A., Semenov, V. S., Sidneva, M. V. and Biisinger, T. (1986b) Impulsive reconnection in the magnetotail during substorm expansion. J. Geophys. (submitted). Tsyganenko, N. A. and Usmanov, A. V. (1982) Determination of the magnetospheric current system parameters and development of experimental geomagnetic field models based on data from IMP and HEOS satellites. Planet. Space Sci. 30, 985. Untiedt, J., Pellinen, R. J. and Kiippers, F. (1978) Observations of the initial development of an aurora1 and magnetic substorm at magnetic midnight. J. Geophys. 45, 41. Williams, D. J., Keppler, E., Fritz, T. A., Wilken, B. and Wibberenz, G. (1978) The ISEE 1 and 2 medium energy particles experiment. IEEE Trans. Geosci. Electron. GE-16, 270. Yahnin, A. G., Sergeev, V. A., Ievenko, I. B., Solovjev, S. I. and Rakhmatulin, R. A. (1984) Characteristics of phenomena, accompanied the local flares of discrete auroras. Mugnetospheric Res. (in Russian) 5, 93.