JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 106, NO. A9, PAGES 18,843–18,855, SEPTEMBER 1, 2001 Substorm and convection bay compared: Auroral and magnetotail dynamics during convection bay V. A. Sergeev,1 M. V. Kubyshkina,1 K. Liou,2 P. T. Newell,2 G. Parks,3 R. Nakamura,4 and T. Mukai5 Abstract. Using observations from eight spacecraft and a ground network, we study two subsequent bay-like disturbances on December 10, 1996, initiated by southward interplanetary magnetic field intervals, one being a classic substorm and another one a convection bay. Both events showed enhanced convection and Dst decreases as well as Pi2 pulsations in the auroral zone. Contrasting to the well-defined substorm signatures of the first event (poleward auroral expansion, substorm current wedge, strong particle injection to 6.6 R E ) resulting from energy loading/unloading and near-Earth reconnection in the tail, these signatures were virtually absent during the convection bay (CB). Distinctive features of the CB event were the same as those during the Steady Magnetospheric Convection intervals: (1) wide double oval at the nightside; (2) thick plasma sheet, relaxed lobe field, and enhanced magnetic flux closure (large B z ) and multiple bursty earthward flows (BBFs) in the midtail; (3) sporadic narrow soft injections to 6.6 R E ; (4) auroral streamers associated with both BBFs and narrow injections. We emphasize the development of multiple and sporadic auroral streamers which start at the poleward oval boundary, propagate equatorward (in 3– 8 min) and end with a long-duration bright spot in the equatorward oval. We conclude that the plasma sheet and auroral dynamics during the convection bay was formed by sporadic narrow (a few R E wide) plasma streams (plasma bubbles) which transported the plasma sheet material from the distant magnetotail reconnection regions to the inner magnetosphere and may significantly contribute to the magnetospheric circulation on the nightside. We modeled the nightside tail configuration using magnetotail magnetic observations and low-altitude particle boundaries to show that at the beginning of the convection bay the increase of magnetic flux tube volume with distance was small in the midtail. Therefore the “pressure crisis” in the tail was significantly reduced during the convection bay, and the efficient earthward transport by sporadic narrow plasma streams was probably able to balance the magnetospheric circulation to avoid the large-scale instability of the magnetotail. 1. Introduction The magnetosphere is a driven system powered by the energy supplied from the solar wind. Different types of the magnetospheric response to the enhanced energy input to the magnetosphere have been identified. The substorm (the most studied and most frequent type of response, timescale about 1–2 hours) represents a large-scale magnetotail instability. It starts with a global slow magnetic energy storage in the tail. To get this substorm growth phase, there should be an imbalance between the magnetic flux transport from dayside to the tail and the return earthward convection in the plasma sheet. The 1 Institute of Physics, St. Petersburg State University, St. Petersburg, Russia. 2 Applied Physics Laboratory, The Johns Hopkins University, Laurel, Maryland. 3 Space Sciences Division, Geophysics Program, University of Washington, Seattle, Washington. 4 Max-Planck-Institut fur extraterrestrische Physik, Garching, Germany. 5 Institute of Space and Astronautical Science, Sagamihara, Kanagawa, Japan. Copyright 2001 by the American Geophysical Union. Paper number 2000JA900087. 0148-0227/01/2000JA900087$09.00 reason of this imbalance could be the “pressure crisis” in the tail [e.g., Erickson, 1992] which makes impossible the earthward contraction (convection) of closed plasma sheet tubes in the standard tail configurations. This loading phase of a substorm usually turns suddenly to the explosive energy release as the magnetic reconnection starts at the near-Earth neutral line [Baker et al., 1996]. The basic ground signatures of this expansion phase are the poleward auroral expansion, Pi2 pulsations, and the development of DP 1–type current system (westward auroral electrojet and associated specific variations outside the auroral zone) [Rostoker et al., 1980]. The latter feature is due to the growth of the substorm current wedge connecting the active sector of ionosphere with the current disruption region in the tail by the field-aligned currents [Baker et al., 1996]. The DP 2 (twinvortex) current system accompanies moderate and strong substorms during all phases manifesting the enhanced magnetospheric convection [Kamide and Kokubun, 1996] as long as the intense energy supply from the solar wind is continuing. Such behavior is sometimes not observed, and the observations reported show a variety of deviations from the classic substorm scheme. During continuous strong energy input the magnetotail may stay in nearly steady state for a long time, exceeding a few substorm timescales. Such time intervals are known as Steady Magnetospheric Convection (SMC) events 18,843 18,844 SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED (see a review by Sergeev et al. [1996a]). The magnetic configuration during SMC events was argued to be stable but specific, with less “pressure crisis” in the midtail portion of the tail allowing plasma to convect earthward more easily. The observations also showed that nightside convection is not a laminar process: a lot of transient mesoscale features including multiple bursty earthward flows (BBFs) in the plasma sheet (that is typical for all conditions [e.g., Baumjohann, 1993]), magnetic impulsive variations and propagating convection bursts in the nightside auroral zone have been repeatedly identified during the SMC events [Sergeev et al., 1996a]. The full picture of these transient phenomena is not quite understood, particularly, because the information about the auroral signatures of these transients is yet very scarce. Sometimes during nearly steady energy input into the magnetosphere one can observe the sequence of substorms, but these could be separated by comparatively long time intervals ⱖ2–3 hours without loading of magnetic energy into the tail [e.g., Yermolaev et al., 1999]. Enhanced bursty earthward convection is seen at these time in the plasma sheet with no signatures of large-scale changes in the tail [Nakamura et al., 1999; Yermolaev et al., 1999]. Like during the SMC events, the magnetosphere could be in a “directly driven” mode during these episodes between successive substorms. A shorter type of events (⬃1–2 substorm timescales) has been named convection bay [Pytte et al., 1978]. That kind of event did not attract so much attention as the substorms and our knowledge of convection bays is very incomplete and, in some parts, confusing. The identification of convection bays was always based on the absence of substorm expansion signatures (during enhanced convection-related DP 2 current system) rather than on some specific signatures inherent to the convection bays. Pytte et al. [1978] reported a ⬃3– 4 hours long convection bay event following after a substorm expansion phase. It was not a pure SMC-like case because the authors presented evidence of changing large-scale configuration (magnetic field stretching) at the end of this event. Shorter (⬍1 hour) isolated bays in AE index have also been reported [Pellinen et al., 1982; Sergeev et al., 1998]. Both these events showed intense growth (energy loading) phase, which, however, was concluded only with very weak expansion phase (pseudobreakup). The growth-type related current system (DP 2) provided the main contribution to the ground magnetic effects during the above two events. No short-duration event was yet reported in which the magnetosphere was basically in the directly driven mode without significant growth phase effects. One such event will be presented in this paper. Obviously, a large variety of observed responses put the questions about the origin of such variant behavior of the magnetotail and about additional factors (processes) which control this variance under seemingly similar external conditions. One way to look at this problem is to compare the observational signatures of subsequent isolated substorm and convection bay events observed one after another with the same extensive network of spacecraft and ground instruments. It is just the purpose of our paper to study such a unique sequence. In this paper we emphasize two aspects. First, we pay special attention to the global auroral dynamics during the convection bay observed by Polar ultraviolet imager (UVI) instrument to find the distinctive features of convection bays, to provide a linkage between spacecraft observations in different tail regions and to form a picture of transient processes in the plasma sheet. Second, as discussed above, an important controlling variable may be the magnetospheric configuration, particularly, the radial variation of the magnetic flux tube volume in the midtail. This aspect will be addressed using observations from Geotail and Interball spacecraft in the tail and energetic particle boundaries observed by DMSP- and NOAA-type spacecraft, which all are used as the input for the modeling of the magnetic field configuration at the beginning of the convection bay. 2. 2.1. Observations Overview of Observations According to the Wind spacecraft measurements in the solar wind on December 10, 1996, the solar wind velocity changed between ⬃500 km s⫺1 (1600 UT) and ⬃600 km s⫺1 (2200 UT) and there were two pressure pulses at 1658 and 1952 UT (see Figure 1). (Here and afterwards the timing of Wind data corresponds to the corrected time taking into account propagation from Wind position [73, ⫺42, 0] to X ⫽ 10 R E at solar wind velocity). The IMF was large and positive before 1600 UT, B z component changed to small negative values after 1600 UT. It finally turned to the north at 2110 UT. There were two 0.5–2 hour long episodes of strong southward interplanetary magnetic field (IMF) B z and of resulting enhanced energy input to the magnetosphere (see Eps3 ⫽ V B sin3 (/2) parameter in Figure 1, which is known to give the best correlation with the cross-tail potential drop [Boyle et al., 1997]). As a consequence, there were two bay-like enhancements of magnetic activity in the auroral zone commencing at ⬃1710 UT (weak growth phase actually started after 1610 UT) and at ⬃1900 UT (see a stack plot of magnetograms in Figure 1). The ring current also showed the same response; the decreases of Dst index are comparable in both events. The sudden impulses corresponding to pressure jumps in the solar wind can be noticed at 1658, 1941, and 2035 UT. Both events were marked by significant Pi2 pulsation activity in the nightside auroral zone (Lovozero) and midlatitudes (see Kakioka data in Figure 2). However, there were also considerable differences allowing to classify them as a classic substorm (first event) or as a convection bay (second event). The tail observations show the most drastic differences. The total pressure in the plasma sheet (from Geotail located near [⫺25, 0, ⫺2.5]R E GSM, most of time spent in the plasma sheet) as well as the lobe magnetic pressure (from Interball 1 spacecraft at [⫺27, 8, 3]R E GSM while being in the lobe) consistently show the strong loading/ unloading of magnetic energy in the tail during the first event; see Figure 1 (bottom). A few strong tailward flow bursts (accompanied by north-south bipolar excursions of the magnetic field) at Geotail which changed to the strong earthward flow bursts and fluctuating positive B z in the dipolarized plasma sheet after 1753 UT provide evidence of near-Earth reconnection in this event which progressed tailward during the substorm expansion phase (see a more detailed treatment of this classic substorm event by Nagai et al. [1998], Nakamura et al. [1998], and Haaland et al. [1999]). On the contrary, no significant loading/unloading or fast tailward flows were observed in the second event. Here the earthward flow bursts (BBFs) and relatively large (ⱖ5 nT) positive (but fluctuating) B z component are persistently observed at both growth and declining phase of auroral magnetic Figure 1. Overview of activity on December 10, 1996. (a) Solar wind dynamic pressure (Pd) and (b) dayside merging electric field estimate computed from solar wind velocity and magnetic field (Eps3 ⫽ V B sin3 (/2)); the universal time is corrected by the propagation at the solar wind speed. (c) Dst proxy (SYM index); the onsets of sudden impulses are marked by triangles. (d) Stack plot of AE station magnetograms. (e) Total pressure at Geotail and magnetic pressure at Interball (while in the lobe) compared with tail lobe pressure (P 30 ) predicted at r ⫽ 30 R E using Fairfield and Jones [1996] regression, and identification of plasma sheet regimes encountered by Geotail based on plasma  parameter, IPS when  ⱖ 1 and LOBE/BLPS when  ⱕ 0.1. (f, g) B z magnetic field component and V x component of the plasma flow measured at Geotail. Vertical dashed lines show approximate onsets of enhancements of auroral electrojets and of the Dst decreases. 18,846 SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED Figure 2. Band-pass filtered (40 –200 s) H component magnetograms (December 10, 1996) from Kakioka (34⬚ corrected geomagnetic latitude, local midnight at 1500 UT) and Lovozero (63⬚ corrected geomagnetic latitude, local midnight at 2100 UT). activity. The plasma beta parameter was permanently large ( ⱖ 1), indicating a thick plasma sheet. The auroral observations show strongly enhanced precipitation in both events as well as considerable differences in their morphology. As seen from auroral keograms made from Polar UVI observations at two different local times (Plate 1), the auroral oval looks rather narrow at the beginning of first event. The main precipitation was observed between 66⬚ and 69⬚ before the classic poleward expansion phase (from low to high latitudes) in this event. In the second case the oval was wide before the main activity (between 65⬚ and 77⬚) and the main site of activity was its poleward boundary. The situation here resembles the recovery phase of the first (substorm) event between 1820 and 1850 UT. The active auroras show some southward motion from the poleward boundary, just opposite to the substorm case. The details of auroral dynamics will be studied in the next section. Two geostationary Los Alamos National Laboratory (LANL) spacecraft 1991-080 and 1994-084 (hereinafter referred to as L80 and L84) were favorably located at the nightside. They both observed (Figure 3) a strong energetic particle injection in the first substorm (preceded by deep dropout at L84 just at midnight), but only a few weak, soft and short injections during the second event. At keV energies (Magnetospheric Plasma Analyzer (MPA) instrument) in both cases one observes the increases of plasma pressures which could correspond to the inward motion of the plasma sheet during enhanced convection. The magnetic field inclination () evaluated from the main axis of the electron anisotropy tensor (reliable measurements are labeled with circles) showed a strong magnetic field stretching (⌬⌰ ⱖ 20⬚) during the first event, whereas only slight stretching (ⱕ10⬚) was noticed at the beginning of the second event. The midlatitude magnetograms (Figure 4) show the disturbance which is also different in both events. A few episodes of substorm current wedge (SCW) growth are clearly seen in the first event, starting near the meridian of Gnangara station (GNA) at 1700 and 1716 UT and then expanding both west- ward (at 1735 and 1748 UT) and eastward (1748 UT) (compare with the poleward expansion in Figure 3) [see Haaland et al., 1999]. The SCW growth is recognizable as sharp positive H excursion in the central SCW sector accompanied by positive (negative) D excursions westward (eastward) of this meridian. No such evident excursions are seen during the second event where positive deviations of H component are weak (those at 1944 and 2038 UT actually correspond to the pressure increases in the solar wind, Figure 1). Positive D excursion dominates at near-midnight stations (GNA, AMS, CZT), and this corresponds to the effects of the “partial DR current” (convection-related) system (see, e.g., Sergeev et al. [1998] for the discussion). To conclude, the first event is a classical substorm with all its signatures, including the auroral expansion, growth of substorm current wedge and Pi2 activity, energetic particle injections to 6.6 R E , loading/unloading of magnetic energy in the tail as well as the formation of the near-Earth neutral line and its subsequent tailward retreat. The second event did not show any of these substorm signatures (except for Pi2 pulsations in the auroral zone). In the next section we study the auroral dynamics emphasizing the localized sporadic auroral streamers, which give us a global view and will allow to link together other observations available. 2.2. Auroral Streamers and Related Phenomena The auroral UV morphology during the most active half hour of convection bay is illustrated in Plate 2 using a full time resolution of the UVI instrument. (Movies showing auroral UV dynamics for the whole event are available at 具http://sdwww.jhuapl.edu/Aurora/mpg_samples.html典.) The event started as the activation of a small bright surge form in the poleward half of the premidnight oval (at ⬃75⬚) at 1917 UT, but it faded soon and the following activity differs much from the typical substorm behavior. Being sporadic in time and very structured in space (in local time), it was formed by specific localized structures, auroral streamers, which started to develop from the poleward oval and propagated into the equa- SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED 18,847 Figure 3. Summary of observations at two nightside geostationary LANL spacecraft. (a) Polar angle (theta) of main axis of electron temperature anisotropy tensor; values with large anisotropy (T ⬜ /T 㛳 ⱖ 1.13 or ⱕ 0.9) are marked by circles, B field orientation predicted with T89 models are shown for comparison. (b) Anisotropy ratio (T ⬜ /T 㛳 ). (c) Partial pressures of hot protons (0.13– 43 keV) and electrons (0.03–24 keV). All these data are from L80 spacecraft. (d) and (e) Energetic electron (50 –315 keV) fluxes at L80 and L84 geosynchronous spacecraft, respectively. The triangle in Figure 3d shows the data point used in the magnetospheric modeling. torial oval. About 10 very distinct streamers have been identified during the convection bay event. Some of them were clustered in time (like during the most active time period between 1945 and 2000 UT), some appeared isolated, their timing results are given in Table 1. There were also a few confusing events in which stages B or C (see below) were missing or could not be resolved for sure. Three different stages could be identified during the streamer development. Quite often its equatorward propagation is delayed by a few minutes after the initial brightening, which takes place near the poleward border (phase A). Then its active phase (B) follows when the streamer propagates toward the equatorward oval. The bright spot in the equatorward oval is formed after the streamer hits the equatorward oval. This spot can be eventually observed for as long as 10 –20 min (for example, the rests of streamer f could be recognized until 2022 UT), and this is the phase C in our definition. The duration of the active phase B varied between 3 and 8 min in this event. During the convection bay the Geotail spacecraft was right near midnight; however, the time period of strong auroral activity between 1945 and 2030 UT (Plate 1) does not look more active in the spacecraft records. Geotail observed strong flow bursts (with Ey ⱖ 2 mV m⫺1) at 1920, 1943, 2011, 2022, 2116, and 2148 UT. Five of these six bursts had associated streamers in phase A or B (streamers a, c, h, i, j), as shown by horizontal bars in Figure 5. As seen from Table 1 and Plate 2, these streamers were localized within ⬃1 hour from the Geotail foot point (at 00 MLT). More detailed comparison will be presented in a separate paper. From Figure 5 one could clearly see the density and pressure drops and sharp increases of B z in the central plasma sheet associated with impulsive flux transport events. 3. Energetic Particle Observations and Magnetic Configuration at the Beginning of Convection Bay and Substorm Between 1943 and 1948 UT, at the beginning of the convection bay, the NOAA 14 polar spacecraft crossed the southern oval near the midnight meridian (see Figure 6). The fluxes of both trapped and precipitated energetic particles were measured together with the precipitated energy flux of auroral 18,848 SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED Figure 4. Stack plot of midlatitude magnetograms of H component (solid lines) and D component (dashed lines). D component sign of stations in the Southern Hemisphere was inverted to facilitate the comparisons. The stations are arranged according to the magnetic local time; their magnetic coordinates (latitude and longitude) are Chambon-la-Foret (CLF, 50⬚, 86⬚), Crozet (CZT, ⫺51⬚, 112⬚), Martin de Vives (AMS, ⫺47⬚, 144⬚), Gnangara (GNA, ⫺42⬚, 188⬚), and Canberra (CNB, ⫺43⬚, 226⬚). Magnetic midnight is shown by the circles under the magnetograms. particles integrated over 0.3–20 keV. These observations carry useful information concerning the magnetotail structure. The regions of empty loss cone ( J 㛳 /J ⬜ ⬍⬍ 1) or isotropic ( J 㛳 /J ⬜ ⬃ 1) pitch angle distributions of energetic electrons and protons may show us the mapping of regions in the equatorial current sheet where the ratio R C / is ⱖ8 (adiabatic motion) or ⱕ8 (nonadiabatic pitch angle scattering), respectively. (R C is the magnetic field curvature and is the particle gyroradius. See Kubyshkina et al. [1999] for more complete discussion and references on that subject.) The equatorwardmost transition, the isotropic boundary, corresponds to the ratio value R C / ⫽ 8. This gives us a formal rule to compute the isotropic boundary position from the magnetospheric magnetic field model for the particle of given mass, energy, and charge. In the current sheet, R C / ⬃ B 2Z /j, so the equatorial position of the isotropic boundary is mostly sensitive to the equatorial magnetic field (B Z component) and, to less extent, the local current density. This makes the isotropic boundaries to be an important source of information for the magnetospheric modeling. Positions of isotropic boundaries are marked by dashed lines in Figure 6. Although the detector which measured the trapped flux of energetic protons was corrupted by penetrating radiation, the isotropic boundary can still be determined with a large confidence as the position of maximal precipitating flux. (This is always the case when both precipitated and trapped proton fluxes are measured; this was specially studied by Newell et al. [1998].) We modeled the magnetic configuration at 1945 UT by using a special technique (Hybrid Input Algorithm [Kubyshkina et al., 1999]) which allows one to use as input the observations of different kind if the formal rule is known how to compute them from the model. We used magnetic field observations from both Interball/Tail spacecraft (in the lobe) and Geotail (in the central plasma sheet). We also used the isotropic boundaries from NOAA 14 (the formal rule to compute it in the tail neutral sheet was R C / ⫽ 8) and the orientation of electron anisotropy tensor at geostationary L80 spacecraft (supposed to be a proxy of the local magnetic field orientation). By varying the intensity of both ring and tail currents as well as by allowing the localized thinning of tail current and the tilt of plasma sheet to be varied, we obtained a best fit model shown in Figure 7. The model parameters included the current sheet thinning at x ⫽ ⫺14 R E (D ⫽ 0.35), enhanced ring current (I RC ⫽ 1.6), and unchanged tail current of the starting T89 Kp ⫽ 2 model (terminology is the same as given by Kubyshkina et al. [1999]). The additional tilt of the plasma sheet was 0.8⬚. The modeling results give us a specific configuration with nonmonotonous B z profile with local minimum of B z ⬃ 1 nT at 12.5 R E , where the current density increased up to 5 nA m⫺2, partly, because of thinning of the current sheet down to 0.8 R E . One important observation which was not tried as input can be used as an independent test of the model. There exists a second region of anisotropic (empty loss cone) energetic electrons which was observed poleward of 66⬚ corrected geomagnetic latitude (Figure 6). Whereas the isotropic distributions between 65⬚ and 66⬚ may correspond to a small B z (and/or strongly enhanced current), the more poleward anisotropic region implies a somewhat larger B Z down the tail (or depressed current or thick current sheet) so that energetic electrons are again nearly adiabatic and no more scattered in the current sheet. This feature (including the poleward boundary of isotropic region) agrees well with the model prediction as shown in Figure 7. This assures us that the nonmonotonous B z profile in the tail (with B Z minimum at 12 R E ) is a realistic feature. 4. Discussion 4.1. Tail Magnetic Configuration as a Factor Controlling the Stability of Magnetotail Steady convection time intervals exist when convecting magnetosphere operates without global changes of configuration, that is, under the balanced flux transport from dayside to the tail and back (in the plasma sheet). Such a mode has been observed in the three-dimensional MHD simulations of magnetosphere under the continuous southward IMF made by Ogino et al. [1994]. The logics of the “pressure crisis” concept [Erickson, 1992] implies that the global growth of the tail current (energy loading) occurs when the return earthward convection is braked in Plate 1. Keograms of auroral luminosity in LBHL (long) band constructed from Polar UVI imager data at ⬃22 MLT and ⬃24 MLT meridians, December 10, 1996. SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED 18,849 18,850 SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED Plate 2. Time series of Polar UVI images at high time resolution for the time period of intense streamer activity between 1940 and 2004 UT. Two subsequent frames in LBH (short) filter (LBHS) image follow after two frames obtained with LBH (long) filter (LBHL), thus forming the cycle. Magnetic midnight (dawn, dusk) is on the right (up, down) side of each image, respectively. SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED Table 1. Development of Auroral Streamers and Their Ground Magnetic and Geostationary Effectsa Poleward Activation, a b c d e f g h i j UT (CGLat, deg. MLT, hours) Active Phase B, UT Phase C, MLT, hours 1921 (74,23) 1936 (75,00) 1945 (74, 23.5) 1945 (72,22) 1948 (76, 00.5) 1951 (76,22) 2009 (74, 00.5) 2009 (71,23) 2022 (71,00) 2115 (75,23) 1925–1930 1936–1941 1948–1951 1949–1955 1949–1954 1957–2001 2010–2013 2010–2016 2022–2030 2115–2121 00.5 01 00.5 22 01 23 01 00 00.5 23.5 a Abbreviations are CGLat, corrected geomagnetic latitude; MLT, magnetic local time. the tail plasma sheet by strong earthward plasma pressure gradients, which eventually leads to the accumulation in the tail of the large magnetic flux transported from dayside magnetosphere. The tail-like magnetic configuration, 18,851 namely, the fast increase of flux tube volume (V) with distance, is considered to be the main factor leading to the convection crisis. The alternative configurations with small radial gradient of V may be constructed, but they can not give a realistic solution globally, because in that case of large magnetic flux closure through the plasma sheet the length of the tail will be unrealistically short [e.g., Schindler and Birn, 1982]. However, the “pressure crisis” problem is basically sharp between 10 and 30–40 RE, and, if the configuration is “crisisless” is that region, that will considerably relax the problem of getting intense earthward convection. The modeling results in Figure 7 indicate that this was the case during the convection bay considered. Comparing the pressure increase at the midnight meridian in the neutral sheet computed from the pressure balance (by integrating grad P ⫽ jyBZ from starting point at 30 RE, Pmod) and that from the adiabatic state equation (PV␥ ⫽ const, ␥ ⫽ 5/3, Padiab), we find nearly the same pressure increase values between 30 RE and 12 RE for the convection bay configuration. Oppositely, for the standard T89 model, such comparison gives ⌬Padiab ⬃ 10⌬Pmod, which is usually taken as indication of strong “convection crisis.” Our results indicate Figure 5. Plasma sheet parameters measured at Geotail spacecraft: Horizontal bars shown in Figure 5b corresponds to the observation of auroral streamers in the MLT sector of Geotail. Vertical dashed lines mark the onset of flow bursts with [V ⫻ B] y ⬎ 2 mV m⫺1. Note the sharp decreases of both density and plasma pressure following the flow burst onset. Both the total Vx (V TOT) and its part perpendicular to B (V PERP) are shown on the V x panel. On the bottom panel, P PLASMA and P TOT are plasma pressure and plasma plus magnetic pressure, correspondingly. 18,852 SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED Figure 6. Particle fluxes measured in the Southern Hemisphere at NOAA 14 spacecraft. (a) Total energy flux of 0.3–20 keV protons and (b)–(e) differential fluxes of energetic particles (protons ⱖ 80 and ⱖ 250 keV and electrons ⱖ 100 and ⱖ 300 keV). Vertical dashed lines on Figures 6b– 6e show the positions of the isotropic boundaries. that there is virtually no such problem during the convection bay event. Comparison of classic substorm and convection bay is consistent with that view. Indeed, the major difference between two events was in their initial states. Before the substorm the configuration was very stressed as indicated by enhanced total pressure in the tail (Figure 1), by the inclined magnetic field and dropout of energetic particle flux (Figure 3), and by the narrow auroral oval (Plate 1). However, because of the action of the preceding substorm, the midtail was dipolarized and the configuration was still crisisless before the convection bay. All basic features of configuration we had during the convection bay are similar to those found previously when studying the steady convection events when the tail globally should be in nearly steady state (see a summary in the review by Sergeev et al. [1996a]). Indeed, the large width of the auroral oval, thick plasma sheet, and large B z in the midtail plasma sheet (all indicating a large magnetic flux closure through the midtail plasma sheet) and the isolated region of energetic electron precipitation corresponding to the local B z minimum at r ⬃ 12 R E are also the essential features of steady convection events. A reduced total pressure in the midtail (smaller than lobe pressure estimated from the solar wind dynamic pressure alone) could be one more key signature of such configuration (this is true in our case; compare total pressure and P 30 at the bottom of Figure 1). Such reduced pressure has also been noticed by Nakamura et al. [1999] and Yermolaev et al. [1999] during episodes of intense plasma sheet convection continued for 2–3 hours after substorms without any energy loading/ unloading signatures. This set of features of steady convection episodes in the plasma sheet indicate a direction in which one could search for a formal parameter to predict whether the tail respond in a steady or time-varying mode to the enhanced driving by the solar wind. However, this may only be a necessary (but not sufficient) condition. During both convection bay studied and during steady convection events the convection appears to be very bursty and localized, which demonstrate that two requirements (locally crisisless configuration in the midtail and intense bursty convection) may both be important to close the global circulation pattern in its plasma sheet segment. 4.2. Auroral Streamers and BBFs The north-south oriented (NS) active auroral forms are known for a long time [Akasofu, 1968]. Previously, the NS auroras have been identified as the features of early expansion phase (when the westward surge is formed [Nakamura et al., 1993]) or, more frequently, during the late expansion or recovery phase when the double oval is well formed [e.g., Rostoker et al., 1987; Elphinstone et al., 1996; Henderson et al., 1998]. The term “auroral streamer” was introduced by Elphinstone et al. [1996] to capture the specific dynamics of these forms (propagation from poleward to equatorward oval) rather than their orientation which may differ from case to case. This may be an important distinction since sometimes these forms are oriented more east-west than north-south (see, e.g., the streamer d in Table 1 and Plate 2) and, oppositely, in ground observations the observed (NS-oriented) part of the large-scale fold of an auroral arc can be incorrectly identified as a streamer. As follows from this discussion, the streamers can be observed at vastly different conditions including different substorm phases (see above), during convection bays (this study), as well as during the steady convection time periods (recently, we studied a number of SMC time intervals to confirm this; the results will be presented in following publications). Following Henderson et al. [1998] and Sergeev et al. [1999], SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED 18,853 Figure 7. (top) Magnetic field configuration in XZ plane modeled at 1945 UT with superimposed projections of the spacecraft (stars) and equatorial mapping of isotropic boundaries (dark circles). (bottom) Modeled and standard (T89 model, Kp ⫽ 3) configurations are compared. Blank and hatched blocks in equatorial B z panel show the observed and predicted positions of strong loss cone filling of 100 keV electrons. we interpret the auroral streamer in terms of sporadic and narrow fast-flow channel (fast plasma stream) which is initiated in the distant tail (phase A), propagates earthward over large distance (phase B), and finally intrudes (stops) in the nearEarth magnetosphere leaving here the locally enhanced pressure region (stage C). The physical processes as well as those controlling the formation of auroras may differ at these three stages reflecting very different plasma regimes in the distant tail (A), through the midtail region (phase B) and in the inner magnetosphere (phase C). Briefly, the magnetic reconnection can be responsible for the stage A and be the origin of fast plasma streams (BBFs and/or bubbles [e.g., Chen and Wolf, 1993, 1999; Sergeev et al., 1996a, 1996b]). The modified pressure profile formed after the intrusion of fast stream into the near tail may be the origin of mesoscale field-aligned current (FAC) system (via the [grad P grad V] mechanism and of modified precipitation. A more detailed discussion of mecha- nisms, including mechanisms to form the auroral precipitation, will be given elsewhere. There exists a number of direct observations confirming this picture. The most important element is the association of auroral streamers with the midtail fast flows. In detail this was recently confirmed for the streamers h and i of our convection bay event where three spacecraft (Interball/Auroral, Geotail, and geosynchronous L80) were favorably connected to the same auroral streamer to follow the propagation of fast stream over the distance of ⱖ30 R E which took about 8 min [Sergeev et al., 2000]. General temporal and spatial connection between the fast plasma sheet flows and auroral activations have been recently discussed by Nakamura et al. [1998], Lyons et al. [1999], and Fairfield et al. [1999]. Narrow injections to 6.6 R E at the stage C of streamer development have been shown by Henderson et al. [1998] and, in more detail, by Sergeev et al. [1999]. 18,854 SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED A comment is required concerning how narrow the auroral streamers are and how much transport they can provide. Plate 2 clearly indicates that the streamers during active stage B are about (or less than) 0.5 hour MLT wide in longitude. The separation of streamers which develop simultaneously can be as small as ⬃0.5–1 hour MLT (e.g., compare streamers c and e). These two characteristics indicate a small cross-tail scale of the magnetospheric source of auroral streamer, being about 2–4 R E if mapped to the magnetosphere. This nicely agrees with a ⬃3 R E cross-tail extent of bursty flows obtained in rare observations with the pair of closely spaced magnetospheric spacecraft [Sergeev et al., 1996b; Angelopoulos et al., 1997]. The electric field is so large (a few mV m⫺1) during the BBFs that with 3 R E cross section of the plasma stream the potential drop (⬃60 kV for E ⫽ 3 mV m⫺1) is able to provide the total expected flux transport in the tail in one stream at a time [Sergeev et al., 2000]. A noteworthy feature is that the streamer-associated earthward bursty flows observed during convection activity invariably show the signatures of the plasma bubbles, for example, the decrease of plasma pressure and density accompanied by a sharp increase of both B z component and B inclination if observed in the high- plasma sheet (see Figure 5 and Kauristie et al. [2000] for other examples of BBFs observed at ⬃12 R E during the continuously disturbed period). In fact, the bubbles have been suggested as another way to realize the earthward convection in the tail-like configurations containing some “pressure crisis” [Chen and Wolf, 1993, 1999]. That is why we believe that possibility of steady (balanced) convection in the tail is favored by the combination of both specific (less crisis) magnetic configuration plus the action of plasma bubble mechanism realizing the earthward convection. 5. Conclusions We studied the convection bay event in which the magnetotail responded to the enhanced solar wind driving in a directly driven mode, with very little change of global magnetospheric configuration, hence, with the balanced flux transport on both dayside-to-tail and plasma sheet segments of the global circulation pattern. We modeled the magnetospheric configuration and argued it to be of the same type as during previously studied steady convection events, with reduced lobe magnetic flux, increased magnetic flux closure through the midtail plasma sheet, stretched field lines in the near tail, and a local B z minimum at ⬃12 R E . The pressure crisis is significantly reduced in such a configuration, and this seems to be a necessary condition for the magnetosphere to respond in the directly driven mode. On the other hand, we found that most distinct morphological feature of the convection bay was the sporadic appearance of multiple intense auroral streamers which are associated with bursty fast plasma flows in the plasma sheet. Following previous work, we argue that these streamers are optical manifestations of sporadic narrow fast plasma sheet streams. These narrow streams transporting plasma from the distant tail to the inner magnetosphere have the characteristics similar to the plasma bubbles. They could be the main process supporting the earthward convection in the plasma sheet, and therefore this activity can be a one more necessary condition to realize the driven mode. Comparison of signatures of the substorm and of convection bay indicates a difficulty to distinguish between these two, so different types of magnetospheric response when using the standard monitoring tools, like AE index, auroral zone magnetograms, and Pi2 pulsations. Both types of events produced the drops of similar magnitude in the Dst index. Other tools, including global auroral imaging and tail lobe magnetic field, are required to do the identification of magnetotail response more reliable. Acknowledgments. We thank S. Kokubun (principal investigator of Geotail MGF instrument), R. Lepping and K. Ogilvie (principal investigators of Wind MFI and SWE instruments), R. Belian and D. McComas (principal investigators of SOPA and MPA instruments at LANL spacecrafts), and S. Romanov (principal investigator of Interball MIF instrument) for their data available on Internet through DARTS, CDAWeb, and Interball-Tail databases. Magnetic field data from Kakioka and Lovozero were made available through WDC-C (Kyoto) and Polar Geophysical Institute, and particle data from NOAA 14 spacecraft (D. Evans and H. Sauer, principal investigators of TED and MEPED instruments) were provided through WDC-A (Boulder). V. Sergeev and M. Kubyshkina thank the DFG for support during their stays in MPE, Garching, as well as partial support by the RFBR grants 98-05-04114, 00-05-64885, Russian Universities grant N 992709 and INTAS grant N99-000078. The work at APL was supported by the NASA grant NAG5-7724. Michel Blanc thanks Susumu Kokubun and another referee for their assistance in evaluating this paper. References Akasofu, S.-I., Polar and Magnetospheric Substorms, D. Reidel, Norwell, Mass., 1968. Angelopoulos, V., et al., Magnetotail flow bursts: Association to global magnetospheric circulation, relationship to ionospheric activity and direct evidence for localization, Geophys. Res. Lett., 24, 2271, 1997. Baker, D. N., T. I. Pulkkinen, V. A. Angelopoulos, W. Baumjohann, and R. L. McPherron, The neutral line model of substorms: Past results and present view, J. Geophys. Res., 101, 12,975, 1996. Baumjohann, W., The near-Earth plasma sheet: An AMPTE/IRM perspective, Space Sci. Rev., 64, 141, 1993. Boyle, C. B., P. H. Reiff, and M. R. Hairston, Empirical polar cap potential, J. Geophys. Res., 102, 111, 1997. Chen, C. X., and R. A. Wolf, Interpretation of high-speed flows in the plasma sheet, J. Geophys. Res., 98, 21,409, 1993. Chen, C. X., and R. A. Wolf, Theory of thin-filament motion in Earth’s magnetotail and its application to bursty flows, J. Geophys. Res., 104, 14,613, 1999. Elphinstone, R. D., J. S. Murphree, and L. L. Cogger, What is a global auroral substorm?, Rev. Geophys., 34, 169, 1996. Erickson, G. M., A quasi-static magnetospheric convection model in two dimensions, J. Geophys. Res., 97, 6505, 1992. Fairfield, D. H., and J. Jones, Variability of the tail lobe field strength, J. Geophys. Res., 101, 7785, 1996. Fairfield, D. H., et al., Earthward flow bursts in the inner magnetotail and their relation to auroral brightenings, AKR intensifications, geosynchronous particle injections and magnetic activity, J. Geophys. Res., 104, 355, 1999. Haaland, S., et al., Magnetospheric and ionospheric response to a substorm: Geotail HEP-LD and Polar PIXIE observations, J. Geophys. Res., 104, 28,459, 1999. Henderson, M. G., J. S. Murphree, and G. D. Reeves, Are north-south aligned auroral structures the ionospheric manifestations of bursty bulk flows?, Geophys. Res. Lett., 25, 3737, 1998. Kamide, Y., and S. Kokubun, Two-component auroral electrojet: Importance for substorm studies, J. Geophys. Res., 101, 13,027, 1996. Kauristie, K., V. A. Sergeev, M. V. Kubyshkina, T. I. Pulkkinen, V. Angelopoulos, T. Phan, R. P. Lin, and J. A. Slavin, A conjugate study of Wind and ground-based observations during transient plasma sheet flows, J. Geophys. Res., 105, 16,677, 2000. Kubyshkina, M. V., V. A. Sergeev, and T. I. Pulkkinen, Hybrid Input Algorithm: An event-oriented magnetospheric model, J. Geophys. Res., 104, 24,977, 1999. Lyons, L. R., T. Nagai, G. T. Blanchard, J. C. Samson, T. Yamamoto, T. Mukai, A. Nishida, and S. Kokubun, Association between Geotail SERGEEV ET AL.: SUBSTORM AND CONVECTION BAY COMPARED plasma flows and auroral poleward boundary intensifications observed by CANOPUS photometers, J. Geophys. Res., 104, 4485, 1999. Nagai, T., R. Nakamura, S. Kokubun, T. Yamamoto, and T. Mukai, Plasma sheet dynamics during substorms with Geotail observations, in Substorms-4, International Conference on Substorms-4, edited by S. Kokubun and Y. Kamide, p. 137, Terra. Sci., Tokyo, 1998. Nakamura, R., T. Oguti, T. Yamamoto, and S. Kokubun, Equatorward and poleward expansion of the auroras during auroral substorms, J. Geophys. Res., 98, 5743, 1993. Nakamura, R., S. Kokubun, T. Mukai, T. Yamamoto, M. Brittnacher, and G. Parks, Temporal and spatial relationships between midtail substorm disturbances and auroral substorm onsets, in Substorms-4, International Conference on Substorms-4, edited by S. Kokubun and Y. Kamide, p. 179, Terra. Sci., Tokyo, 1998. Nakamura, R., L. F. Bargatze, T. Mukai, T. Nagai, K. B. Baker, M. R. Hairston, P. H. Reiff, A. A. Petrukovich, M. Nozdrachev, and O. A. Troshichev, Response of the midtail electric field to enhanced solar wind energy input, J. Geophys. Res., 104, 17,299, 1999. Newell, P. T., V. A. Sergeev, G. R. Bikkuzina, and S. Wing, Characterizing the state of the magnetosphere: Testing the ion precipitation maxima latitude (b2i) and the ion isotropy boundary, J. Geophys. Res., 103, 4739, 1998. Ogino, T., R. J. Walker, and M. Ashour-Abdalla, A global magnetohydrodynamic simulation of steady magnetospheric convection, in Substorms 2, Second International Conference on Substorms, edited by J. R. Kan, J. D. Craven, and S.-I. Akasofu, p. 544, Geophys. Inst., Fairbanks, Alaska, 1994. Pellinen, R., W. Baumjohann, W. Heikkila, V. A. Sergeev, A. G. Yahnin, G. Marklund, and A. O. Mel’nikov, Event study on presubstorm phases and their relation to the energy coupling between solar wind and magnetosphere, Planet. Space Sci., 30, 371, 1982. Pytte, T., R. L. McPherron, E. W. Hones, and H. I. West, Multiplesatellite studies of magnetospheric substorms: Distinction between polar magnetic substorms and convection-driven negative bays, J. Geophys. Res., 83, 663, 1978. Rostoker, G., S.-I. Akasofu, J. Foster, R. A. Greenwald, Y. Kamide, K. Kawasaki, A. T. Y. Lui, R. L. McPherron, and C. T. Russell, Magnetospheric substorms: Definitions and signatures, J. Geophys. Res., 85, 1663, 1980. Rostoker, G., et al., North-south structures in the midnight sector 18,855 auroras as viewed by the Viking imager, Geophys. Res. Lett., 14, 407, 1987. Schindler, K., and J. Birn, Self-consistent theory of time-dependent convection in the Earth’s magnetotail, J. Geophys. Res., 87, 2263, 1982. Sergeev, V. A., R. J. Pellinen, and T. I. Pulkkinen, Steady magnetospheric convection: A review of recent results, Space Sci. Rev., 75, 551, 1996a. Sergeev, V. A., V. Angelopoulos, C. A. Cattell, and C. T. Russell, Detection of localized plasma bubbles in the plasma sheet, J. Geophys. Res., 101, 10,817, 1996b. Sergeev, V. A., et al., Short duration convection bays and localized IMF structures on November 28, 1995, J. Geophys. Res., 103, 23,593, 1998. Sergeev, V. A., K. Liou, C.-I. Meng, P. T. Newell, M. Brittnacher, G. Parks, and G. D. Reeves, Development of auroral streamers in association with localized impulsive injections to the inner magnetotail, Geophys. Res. Lett., 26, 417, 1999. Sergeev, V. A., et al., Multiple-spacecraft observation of a narrow transient plasma jet in the Earth’s plasma sheet, Geophys. Res. Lett., 27, 851, 2000. Yermolaev, Y. V., V. A. Sergeev, L. M. Zelenyi, A. A. Petrukovich, J.-A. Sauvaud, T. Mukai, and S. Kokubun, Two spacecraft study of plasma sheet convection during continuous external driving: Convection jet and global substorm effects, Geophys. Res. Lett., 26, 177, 1999. M. V. Kubyshkina and V. A. Sergeev, Institute of Physics, St. Petersburg State University, St. Petersburg, 198904, Russia. ([email protected]) K. Liou and P. T. Newell, Applied Physics Laboratory, The Johns Hopkins University, Laurel, MD 20723. T. Mukai, Institute of Space and Astronautical Science, Sagamihara 229, Kanagawa, Japan. R. Nakamura, Max-Planck-Institut fur extraterrestrische Physik, Postfach 1312, Garching, D-85741, Germany. G. Parks, Space Sciences Division, Geophysics Program, University of Washington, Seattle, WA 98195. (Received February 7, 2000; revised May 31, 2000; accepted June 2, 2000.) 18,856
© Copyright 2025 Paperzz