See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/3164937 Tutorial on geomagnetic storms and substorms Article in IEEE Transactions on Plasma Science · January 2001 DOI: 10.1109/27.902214 · Source: IEEE Xplore CITATIONS READS 30 1,932 1 author: A. T. Y. Lui Johns Hopkins University 504 PUBLICATIONS 14,166 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Dipolarization Front And Current Disruption View project All content following this page was uploaded by A. T. Y. Lui on 07 September 2018. The user has requested enhancement of the downloaded file. 1854 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 28, NO. 6, DECEMBER 2000 Tutorial on Geomagnetic Storms and Substorms Anthony T. Y. Lui Invited Paper Abstract—Geomagnetic storms have been studied for more than a century and substorms for nearly four decades. The space era which began in the late 1950s has ushered new discoveries and given new practical importance to this scientific discipline as we continue to amass technological assets in space. Geomagnetic storms and substorms pose hazards to our venture into the final frontier, much like adverse atmospheric weather does to our outdoor activity. This tutorial provides an overview of the main characteristics of these space phenomena and a brief review of the present theories concerning their initiation. For geomagnetic storms, the two prevailing views on their cause are discussed. One considers a storm to be the accumulated effects of a series of substorms while the other considers it to be the result of strong and prolonged enhancement of the global magnetospheric electric field. A synergistic model combining elements from both views is proposed as the likely explanation for the cause of geomagnetic storms. For substorms, there are generally four categories of models. The first proposes some plasma instabilities or externally-imposed reduction in the magnetospheric electric field at the near-Earth region. The second calls for magnetic reconnection occurring in the mid-tail environment. The third focuses on coupling the ionosphere to the magnetosphere while the fourth invokes abstract descriptions in nonlinear dynamics to address some statistical characteristics of substorms. An evaluation of observations and these models suggests that substorm onset may not be uniquely attributed to a single physical process. A synthesis of the essential elements of the present models in all four categories may well yield the answer to the mystery on the initiation of magnetospheric substorms. Index Terms—Magnetic reconnection, plasma instability, storm, substorm. I. INTRODUCTION T HERE is enhanced awareness of space weather due to the coincidence of the new millennium with the approach of the solar maximum. Before the millennium transition, the Y2K problem associated with the two-digit designation of the year by computer software was expected to be a menace. Anticipating a similar menace to our technological society by solar activity reaching the maximum of its cycle, some space scientists had referred to the upcoming solar maximum as the other Y2K problem. While the Y2K problem turned out to be a nonproblem as a result of much resources devoted to minimize Manuscript received January 3, 2000; revised April 27, 2000. This work was supported by the Atmospheric Sciences Division of the NSF to the Johns Hopkins University Applied Physics Laboratory. The author is with the Applied Physics Laboratory,The Johns Hopkins University Laurel, MD 20723 USA. Publisher Item Identifier S 0093-3813(00)10656-3. its impact beforehand, there is no comparable preparation for the upcoming solar maximum. Since the maiden voyage of Sputnik in 1957 which demonstrated the ability to launch artificial satellites, there has been a progressively rapid increase of technological assets in space. With each new endeavor follows an ever-increasing dependence on space and ever-growing need to monitor and forecast the space environment to avoid potential hazards. These developments can justifiably be viewed as the dawning of our venture into the final frontier, our first step into the enormous universe awaiting for our exploration. No longer is space weather just an academic pursuit or a curiosity but a practical necessity. This is analogous to the early days of air travel when the atmospheric environment at airplane altitudes took on a sudden significance in our daily lives. Historically, the discovery of magnetic storms may be contributed to Baron Alexander von Humboldt who reported the observation of the deflection in the local magnetic declination at Berlin on December 21, 1806 [1]. He also noted the simultaneous auroral activity overhead with the magnetic deflection. The general description of a geomagnetic storm in the early days is through the time history of the horizontal or -component of the magnetic field. The official index called Dst (disturbance storm time), constructed on the basis of the -components of four low-latitude magnetic stations, is now routinely used as a measure for the storm intensity. The Dst index during two magnetic storms is shown in Fig. 1. A storm generally begins with a sudden increase of the -component of the geomagnetic field (the storm sudden commencement or SSC), followed by a period of the field remaining enhanced (the initial phase), then followed by a period of a substantially reduced field (as much as 1% of the Earth’s main equatorial field) lasting for several hours (the main phase). The eventual return of the field to its nominal quiet-time value (the recovery phase) usually takes several days but can last for several weeks. Deviations from the general picture as shown in Fig. 1(a) exist, e.g., some storms do not have well defined SSC and initial phase, as illustrated in Fig. 1(b). A long time past from the early report of Humboldt before the source disturbance was traced back to the Sun. The magnetic field depression during the storm main phase was first attributed to electrical currents flowing near the Earth from charged particles originating directly at the Sun [2], [3]. However, the Earth’s intrinsic magnetic field is a very effective barrier to charged particles such that direct entry of solar particles is very limited. Nevertheless, the idea that magnetic storms are caused by a ring of electrical current encircling the Earth called the ring current is 0093–3813/00$10.00 © 2000 IEEE LUI: TUTORIAL ON GEOMAGNETIC STORMS AND SUBSTORMS (a) 1855 Fig. 2. Schematic diagram to illustrate the trapping of charged particles in the Earth’s magnetic field by the magnetic mirror points in strong field regions. [8]. For a charged particle of mass , charge and velocity , its equation of motion in a uniform magnetic field is (1) If we resolve the velocity into components parallel to , we find perpendicular and (b) Fig. 1. (a) Dst index for a classic geomagnetic storm which begins with a sudden storm commencement, followed by the initial phase, the main phase, and the recovery phase. (b) The Dst index for a geomagnetic storm which shows deviation from the classic example, i.e., no recognizable SSC and the initial phase preceding the main phase. The date labels are placed at the mid-day time. essentially a correct one. As will be explained in Section II, the charged particles responsible for the ring current are trapped by the Earth’s magnetic field similar to the energetic charged particles in the Van Allen radiation belts. This tutorial will briefly review the underlying physics concerning geomagnetic storms and substorms. Section II deals with how electric currents exist in space, which necessitates some basic understanding of how charged particles move in the near-Earth space. Section III addresses the question of where do these charged particles come from and eventually go, which relates to the interaction between the Earth’s magnetic field and the continuous supersonic blast of ionized gas from the Sun. Section IV summarizes the present two competing views for geomagnetic storm onsets. Section V provides a bird’s eye view on the myriad of space disturbances associated with substorms. Section VI discusses theories on onset of substorms, classifying them into basically four schools of thought. The last section concludes by highlighting some future outstanding issues to be solved on these phenomena. II. HOW DO ELECTRIC CURRENTS EXIST IN SPACE? The charged particle population of the storm-time ring current has been measured directly by spacecraft [4], [5]. The dominant population, with 95% of the charged particles in the energy range of 5–500 keV [6], resides in the radial range of 2 to 8 ( km) from the Earth. To understand how these particles can be responsible for the world wide depression of the Earth’s magnetic field requires some basic knowledge on how charged particles move in space. The guiding center concept for charged particle motion introduced by Alfvén [7] laid the foundation for the modern day explanation of this ring current (2) indicating that the particle’s velocity along the magnetic field is constant while its motion perpendicular to the magnetic field , analogous to the basic is circular with a frequency of orbital motion of a planet around the Sun due to the gravitational force. The sense of gyration is dependent on the charge. The frequency of this gyration is called the gyrofrequency and is typically very high. In the geostationary orbit where the magnetic field may be 100 nT, the gyrofrequency is 2.8 kHz for electrons and 1.5 Hz for protons. In addition, the gyration motion of a charged particle generates a circular current loop, equivalent to a small magnet with a magnetic moment . When the magnetic field is nonuniform and is converging, the repelling force (also known as the mirroring force) acting . This force acts to accelerate on a charged particle is the particle away from a strong field region and often leads to the particle being reflected back (mirrored) from the region. A magnetic field with a strong field at both ends of a field line, as in the case for the Earth, traps a charged particle due to this force. This concept is illustrated in Fig. 2. The particle bounces between two points called the mirror points at which its component becomes zero. This is the basic principle by which Van Allen radiation belt particles are trapped in space. Note that the magnetic field does no work on the particle so that the particle’s component detotal energy remains constant, i.e., when the component increases to conserve the total encreases, the ergy. It is customary to refer to a magnetic field line in terms of its shell value. The shell coordinate takes into account the deviation of the magnetospheric field from a perfect dipole and organizes well the charged particle population in space. The shell value is basically the equatorial crossing distance in of the magnetic field line if the magnetic field in space were a pure dipole. The guiding center concept introduced by Alfvén [7] considers the average position and motion of a charged particle 1856 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 28, NO. 6, DECEMBER 2000 by neglecting the fast gyration around the magnetic field. With this approximation (valid when scales of magnetic field inhomogeneities are large compared to the radius of gyration and when the time scale of variation is slow compared to the gyration period), the drift motion of the gyrocenter (the center of the fast gyration) of a charged particle under a given force is (3) Based on this simplified expression, we can readily obtain the drift motions of a charged particle moving in a nonuniform magnetic field. For example, the motion along a curved magnetic is field with a radius of curvature , resulting in a drift associated with a centrifugal force of , as illustrated in Fig. 3(a). given by Since the drift direction is charge dependent, for a particle pop, the curvature ulation with a velocity distribution given by drift gives rise to a current density (4) where is the particle pressure component parallel to the magnetic field with the assumption that the velocity distribution is independent on the phase of gyration, i.e., the number of is particles in the velocity range and . In a magnetic field gradient , the reand the drift pelling force for a magnetic moment is , as illustrated in Fig. 3(b). Similarly, is as the drift direction is charge dependent, the current density associated with gradient drift for a particle population is (a) (b) (c) (5) is the particle pressure component perpendicular to where the magnetic field. Furthermore, if there is a density or temperature gradient of the particle population, then there is a current density which does not involve any actual particle drift. This current density arises from gyration motion of the particles, from crowding of gyration orbits due to a density gradient or a larger gyration speed due to a temperature gradient, as illustrated in Fig. 3(c) and (d). This is called the magnetization current and is given by (6) The combination of all these real and apparent drift motions gives rise to electrical currents which are quite ubiquitous in space. In fact, a weak ring current exists even outside magnetic storm periods. Quantitatively, all these drifts may be added and simplified after some vector algebra to give the equation for the total electric current perpendicular to the magnetic field (d) Fig. 3. Electric current arising from the motion of a charged particle in: (a) a curved magnetic field geometry, (b) a magnetic field gradient, (c) a density gradient (represented by two circles on the higher density side), and (d) a temperature gradient (represented by the larger circle on the higher temperature side). The sense of gyration and the drift motion direction is shown for positive charged particle. Magnetic field in panels (b)–(d) points out of the paper. The force F acting on the particle in panels (a) and (b) is explicitly given with the corresponding guiding center drift v . carried by charged particles in the nonuniform magnetic field of the magnetosphere as (7) ) This expression is valid for a quasisteady state (i.e., and when the gravitational force can be neglected. From in situ measurements, it is found that the first term in (7) generally LUI: TUTORIAL ON GEOMAGNETIC STORMS AND SUBSTORMS 1857 is small compared with the total particle kinetic energy. Several good references now exist for extended discussion on topics covered in this section [12]–[15]. III. SOURCES AND SINKS OF THE RING CURRENT CHARGED PARTICLES Fig. 4. Observations during the main phase of a geomagnetic storm showing the equatorial radial profile of (a) ring current particle pressure components, and (b) current densities (positive and negative values denote westward and eastward currents, respectively). The radial distance is represented by the L shell value. dominates over the second term during storms [9]. As an example, Fig. 4 shows a radial profile of ring current pressure components measured by the AMPTE/CCE spacecraft and the deduced ring current densities during a geomagnetic storm on September 4, 1984. In general, the ring current is directed east2.5 to 3.5) but ward inward of the peak pressure location (at is directed predominantly westward beyond the peak. Overall, the ring current has a net westward component which causes the world-wide decrease in the geomagnetic field. The magnetic field reduction at the Earth due to the ring current population can be obtained from applying the Biot–Savart law based on the measured electrical current [9]. However, the common practice is to use the Dessler–Parker–Sckopke relation [10], [11] which to the relates the total energy of the ring current particles on the ground which is given as magnetic perturbation (8) is the magnetic field at the Earth’s equatorial surwhere face and is the total magnetic energy of the Earth’s dipole field above the surface. This expression is derived based on the assumption that the magnetosphere is a closed quasistationary system and the total magnetic energy inside the magnetosphere The main issues for geomagnetic storms are the origin of the ring current particles, how they are transported to the ring current region, and their loss mechanisms. There are two sources for ring current particles. The first source comes from the interaction between the Earth’s magnetic dipole and the solar wind which is a supersonic stream of charged particles continuously emitted from the Sun. The Earth’s magnetic field is confined by the solar wind to a volume called the magnetosphere and its outer boundary surface is called the magnetopause (see Fig. 5). A fraction of the solar wind particles enters through the magnetopause to the plasma sheet, creating a large reservoir of charged particles in the Earth’s distorted nightside magnetic field known as the magnetotail. These solar wind particles are transported inward to the ring current region during magnetic storms. The second source is the ionosphere which provides a significant outflow of charged particles to the plasma sheet in association with the physical processes for auroral arc formation [16], [17]. Whether the dominant source for the ring current is the solar wind or the ionosphere depends on a number of parameters such as the storm intensity, the phase of the solar cycle, and the strength of solar ultraviolet emission which influences the ionization rate in the ionosphere. Observations have indicated that ionospheric ions like singly charged oxygen ions may dominant over the usually most abundant protons in intense geomagnetic storms briefly around their main phase [18]. Protons which are most abundant during less intense storms may also be originated from the ionosphere. The storm-time ring current population gradually becomes depleted by a number of processes. A charged particle may collide with a neutral particle (i.e., one which carries no electric charge) in the magnetosphere, exchange the charge with it and escape from the confinement of the magnetic field. A ring current particle may loose its energy by Coulomb collision with another charged particle of lower energy in the magnetosphere. A ring current particle may also get into a trajectory, perhaps through its interaction with plasma waves, which brings it down to the ionosphere where it collides with a dense atmosphere and loses its energy. A ring current particle may diffuse through the magnetosphere due to electric and magnetic fluctuations or be transported by electric field inside the magnetosphere, eventually escaping from the magnetosphere. In fact, this flux leakage is an important source of energetic charged particles outside the magnetosphere [19]. IV. FORMATION OF THE STORM-TIME RING CURRENT The most significant increase in the ring current population during magnetic storms is in the inner magnetosphere within a . At present, there are two basic geocentric distance of views on the storm-time ring current formation. The conventional idea is that a magnetic storm results from the accumulation of many elementary disturbances, termed magnetospheric 1858 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 28, NO. 6, DECEMBER 2000 Fig. 5. Illustration of the various electric currents in space from the interaction between the solar wind and the Earth’s magnetic field. Enclosing the magnetosphere are the dayside and nightside magnetopause currents. The magnetosphere is shaped like a comet with the magnetic field stretched in the nightside to form the magnetotail. A cross-tail current flows in the equatorial plane of the magnetotail within the plasma sheet. This current links with the auroral ionosphere through field-aligned currents. Earthward of the plasma sheet is the ring current encircling the Earth. substorms, manifested most vividly in the polar region [20]. Each substorm produces an enhanced westward electric field in the outer ring current boundary and brings the particles in from the plasma sheet, a phenomenon known as substorm injection [21], as illustrated in Fig. 6(a). Equation (3) indicates that an electric field leads to particle motion called convection which is independent of charge and mass so that all particles move in a coherent fashion (no electric current is generated by convection). Enhanced convection competes in the radial transport of the ring current particles with the curvature and gradient drifts which move the particles azimuthally instead. The plasma sheet population may be modified during the substorm process such that more ionospheric particles contribute to the ring current population as magnetic storms evolve. Subsequently, the ring current particles move further inward by electric field fluctuations as predicted by early theoretical study of ring current formation which describes the ring current formation by a diffusion equation [22]. The time evolution of the distribution funcfor ring current particles with energy on the tion -shell (averaged over their bounce motion and their phase of gyration) can be described by [23] (9) where the radial diffusion coefficient due to electric impulses is given by (10) is the distance to the subsolar dayside magHere is the equatorial magnetic field at the subsolar netopause, is the power spectral density of elecmagnetopause, and tric impulses at the frequency of the particles’ drift motion around the Earth. Note that the radial diffusion coefficient depends on the sixth power of the value so that the diffusion becomes less efficient as the particles approach closer to the Earth. Overall, this view on storm onset implies that the cause and the intensity of magnetic storms may be traced to the factors which determine the frequency and intensity of each substorm. A challenge to the view of a storm as a sum of substorms has surfaced early on and has become more noticeable recently. Observations indicate that magnetic storms arise from a specific solar wind condition which is favorable for setting up a largescale electric field in the magnetosphere to transport charged particles from the plasma sheet to the inner magnetosphere. The favorable condition is long-duration ( 3 h) of strong southward nT) [24]. In addition, interplanetary magnetic field ( LUI: TUTORIAL ON GEOMAGNETIC STORMS AND SUBSTORMS (a) Storm as a sum of frequent substorms. (b) Storm as a prolonged strong convection. Fig. 6. Schematic diagrams to illustrate the two views for the onset of a geomagnetic storm: (a) it is the accumulated effects of frequent intense substorms which progressively build up the storm-time ring current, and (b) it is an enhanced and prolonged magnetospheric convection event. as mentioned earlier, the Dst index is commonly used to measure the storm intensity. From a study of the substorm effect on the Dst index, it was found that the rate of increase in the magnitude of the Dst index during magnetic storms does not become greater with substorm development [25]. In other words, if we take the Dst index to be a true indicator on the buildup of the storm-time ring current, then a substorm tends to cause the ring current to develop less rapidly than before the substorm onset. Furthermore, it is estimated that the induction electric field associated with substorms is not effective in the radial transport of where the ring current particles to geocentric distances population increases most significantly during magnetic storms [26], [27]. These points are used to justify the new view that sustained enhancement of magnetospheric electric field rather than frequent occurrence of intense substorms is the reason for the ring current formation. In other words, a magnetic storm is a result of sustained strong magnetospheric convection, as illustrated in Fig. 6(b). However, this new view is not without its own challenges. Numerical simulations have shown that the buildup of the ring current is effective from an enhanced convection only initially (for the first 3 h) [28], [29]. Continual presence of a steady convection gives no further increase in the ring current thereafter. This is understandable because strong convection also removes particles from the magnetosphere through the dayside magnetopause. On the other hand, the ring current can be continuously built up after this initial rise by electric field fluctuations, which may be substorm-related or variations in the convection electric field. Nevertheless, there is little doubt that each substorm injects particles into the ring current region and thus is expected to contribute to the strengthening of the ring current. The lack of clear indication of ring current buildup from substorm development in the Dst index may well be a defect of the Dst index in portraying the true ring current intensity [31]. Note that the 1859 ring current gives a negative perturbation of the -component on the ground while the current system associated with a substorm (to be discussed later) gives a positive perturbation [30]. As a result, the true strengthening of the ring current may well be masked by the substorm current system. A better index than the Dst index in monitoring the ring current population needs to be developed to overcome this deficiency [31], [32]. The argument that substorms are ineffective in radial transport can be countered by evidence that substorm onsets during storms occur at very low latitudes on the ground, i.e., very close to the Earth in the magnetosphere, so that substorm-energized particles need not be transported from far to contribute to the ring current population. A synthesis of the old and new ideas may well be the correct scenario for storm development. Both steady convection and substorms play important roles in storm-time ring current formation. The former provides the initial rise of the ring current population while the latter provides the particle source as well as the continual buildup after the initial rise from convection by diffusion due to electric field impulses generated by substorm activity. While solar wind fluctuations may generate convection electric field fluctuations, the ones generated by substorms are larger in magnitude and thus are more efficient in inward transport of particles. On the other hand, numerical simulations show that transport (convective or diffusive) cannot account for a large storm unless there is a stormtime enhancement of the plasma sheet distribution [28], [29], [33]. This suggested scenario of ring current buildup is consistent with numerical simulations which can isolate these different factors and is also compatible with observations indicating that larger storms generally have two or more distinct episodes of ring current enhancements [34]. V. SUBSTORM PHENOMENA Substorm disturbances are seen over many regions in the near-Earth space. Three different phases of a substorm have been defined: growth, expansion, and recovery [35], [36]. Substorms may occur so frequently that they overlap. Therefore, the best way to identify clearly substorm features in each of these phases is to examine an isolated substorm with observations from multiple ground stations and satellites distributed within the magnetosphere [37]. From such investigations, the key substorm phenomena are identified. The very quiet magnetospheric condition corresponds typically to northward interplanetary magnetic field (IMF) and a small polar cap encircled by the global distribution of auroral arcs. The polar cap is the region at Earth which contains magnetic field lines connected to the IMF. Growth phase usually begins with the start of southward turning of the IMF. In the ionosphere, the polar cap size increases as the ionospheric electrojets (currents) in the polar region intensify. In the magnetosphere, the cross-section of the magnetotail enlarges. Plasma sheet thinning develops in the near-Earth magnetotail ( 5–15 downstream), altering the dipolar field there to become taillike. Fig. 7 illustrates some key disturbances attributed to substorm expansion activity. At substorm expansion onset one of 1860 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 28, NO. 6, DECEMBER 2000 netic stations (opposite in effect to the storm-time ring current). Referring back to Fig. 7(e), thinning of the plasma sheet commences at further downstream distances in the mid-tail. Transient fast plasma flows are detected in the central plasma sheet, along with impulsive and highly fluctuating electric and magnetic fields. Recovery phase begins when the poleward expansion of the auroral bulge halts. The last episode of expansion usually exhibits a major auroral arc brightening over an extended portion of the poleward front of the auroral bulge. In the mid-tail region ( 15–80 downstream), plasma sheet suddenly thickens with fast plasma flows, predominantly field-aligned, at the plasma sheet boundary. In the distant tail region [Fig. 7(f)], magnetic field shows a northward then southward swing. This is typically accompanied by tailward streaming of energetic electrons at the plasma sheet boundary. These features are interpreted as signatures of a plasmoid in which energized particles are enclosed by closed loops of magnetic field. VI. SUBSTORM THEORIES Fig. 7. Diagram to illustrate several key substorm phenomenon. The substorm onset time is indicated by the vertical dashed lines in the panels. The ground activity is shown by the H -component of the magnetic field at high latitude stations (H -high) and at mid-latitude stations (H -mid). The micropulsation at onset is indicated by the H trace. The increased fluxes of energetic particles at the geostationary altitude and in the near-tail region are indicated by the J traces. The magnetic field dipolarization in these regions is indicated by the B trace. Magnetic turbulence at the current sheet is often seen prior to dipolarization in the near-tail at substorm onset. Plasma sheet thinning in the mid-tail region is often seen by drops in number density (N trace) and temperature (T trace). Plasma flow (V trace) may occasionally be tailward before dropout and become earthward at plasma sheet recovery. Observation in the distant tail often show a transient increase in the total magnetic field magnitude (B trace) accompanied by a north-then-south excursion of the B component and tailward plasma flow (V trace). These are interpreted as signatures of a plasmoid being ejected down the tail. 1 the nightside auroral arcs, typically the most equatorward one in the midnight sector, brightens suddenly, breaks up, and expands poleward [Fig. 7(a)]. This activity builds up to an auroral bulge in which its westward end exhibits one or multiple surge forms. The -components at high-latitude ground magnetic stations show large negative excursions while those at midlatitude and equatorial stations show small positive excursions [Fig. 7(b)]. Micropulsations with periodicities in the 40–150s range, called Pi2, are often observed in these magnetograms. In the near-tail region [Fig. 7(c)] and at the geostationary altitude [Fig. 7(d)], the stretched tail-like configuration developed during the growth phase relaxes abruptly to a dipolar-like field geometry. Earthward injection of energetic particles and thickening of the plasma sheet are seen. This field dipolarization is often preceded by large magnetic fluctuations suggestive of a turbulent state. The change in the current system can be regarded as a rerouting of a portion of the cross-tail current from the tail to the ionosphere, as illustrated in Fig. 8. This current system is called the substorm current wedge and produces positive perturbation in the -component at mid-latitude and equatorial mag- There are many theories for substorm. In general, they may be classified into four categories: (a) those based on some internally triggered or externally driven processes in the near-Earth region, (b) those based on magnetic reconnection occurring in in the tail, (c) those the mid-tail region downstream of 20 based on coupling between the ionosphere and magnetosphere, and (d) those based on abstract descriptions of nonlinear dynamics without specifying the physical process. Fig. 9 shows a schematic diagram for the first category of substorm theories. The most equatorward arc is associated with an intense cross-tail current and an electric field which enhances the local current density to excite a current-driven instability. The onset of the instability and its ensuing turbulence disrupt the cross-tail current and diverts its path to the ionosphere. The disruption injects particles earthward and thickens the plasma sheet. Tailward of the disruption, a rarefaction wave is launched which causes further thinning of the plasma sheet. Magnetic reconnection develops later at a current disruption site which then produces overall thickening of the plasma sheet and tailward ejection of a plasmoid. The possible candidates for current disruption are the cross-field current instability [38], [39], the ballooning instability [40]–[42], or the convection electric field reduction due to external solar wind condition [43]. Fig. 10 illustrates the basic physics principle behind the cross-field current instability which encompasses the modified two-stream and the Ion-Weibel instabilities [44]. The existence of an intense current indicates a large relative drift between ions and electrons which is a form of free energy available to excite waves. An obliquely propagating wave (generated out of thermal noise) may be excited through its interaction with both ions and electrons. Since electrons have small gyroradius and are tied to the magnetic field line, they have to go a long distance along the field to interact with the ions, as illustrated in Fig. 10(a). This long path essentially makes electrons behaving like the heavy ions in the sense that their apparent inertia in interacting with the waves is like that of ions. This effect allows them LUI: TUTORIAL ON GEOMAGNETIC STORMS AND SUBSTORMS 1861 Fig. 8. Diagram of the equatorial cross-section of the magnetosphere to illustrate the diversion of the cross-tail current during substorm expansion. Current paths are drawn with arrows indicating the current flow direction. The times below each panel refers to the time after substorm expansion onset. Fig. 9. Main features of the category of substorm models which places the substorm onset in the near-Earth magnetotail. The sequential developments are numbered from 1 to 6 accordingly. to exchange momentum with the ions efficiently through the wave. As a result, the relative drift between ions and electrons is reduced, i.e., current disruption. In cold plasma theory, the instability is described by (11) Fig. 10. (a) Modified two-stream instability. The excited wave propagates obliquely to the magnetic field as indicated by the wavenumber vector k. Ions have large gyroradius and can interact with the wave readily while electrons have small gyroradius and have to travel a long distance along the magnetic field line to interact with the wave. (b) The Ion–Weibel instability. The wave propagating along the magnetic field causes bunching (n >0) of the current-carrying ions which in turn amplify the wave. For both instabilities, the free energy from the relative drift between ions and electrons, i.e., electric current, is reduced in exciting the wave. Here, and plasma frequency and gyrofrequency for electrons, respectively; ion plasma frequency; relative drift between ions and electrons; angle between magnetic field and wave propagation direction; wavenumber; complex quantity in which the real part is the wave frequency and the imaginary part is the growth or damping rate. For the Ion–Weibel instability, the ion response to a wave propagating along the magnetic field can be written as (12) represents the relative drift between ions and elecAgain, , the wave magnetic trons, as illustrated in Fig. 10(b). At accelerates the ions upward while at a distance of half field wavelength away, the wave magnetic field changes direction and accelerates the ions there downward. This causes a bunching 1862 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 28, NO. 6, DECEMBER 2000 Fig. 12. Diagram to illustrate the idea that a substorm current wedge can be set up by reduction in convection and differential drifts of particles in the plasma sheet [43]. Fig. 11. Ballooning instability is analogous to the Rayleigh–Taylor instability for a heavy fluid resting on a light one. The normal orientation of the magnetosphere is rotated 90 to express the centrifugal force in ballooning instability being analogous to the gravitational force in the Rayleigh–Taylor instability. The centrifugal force acts on the particles within the magnetic flux tube and displaces them tailward when the particle pressure is sufficiently high. Particle pressure is relieved by inflating the weakest point in the magnetic flux tube. of ions in layers (parallel to the -plane in the figure) half a wavelength apart. Since the ions are carrying the current, the ion bunching causes an enhanced current layer which further amplifies the initial wave magnetic field. The ensuing turbulence of this current filamentation extracts the free energy from the relative drift of ions and electrons and disrupts the current. In cold fluid theory, this instability is a purely growing mode with , where is the velocity of light. growth rate The basic physics of the ballooning instability can be gained by drawing its analogy to the Rayleigh–Taylor instability for a heavy fluid resting on a light one. In the ballooning instability, the centrifugal force associated with magnetic field line curvature takes up the role of gravitational force in the Rayleigh–Taylor instability, as illustrated in Fig. 11 [45]. The buildup of current density at the inner edge of the plasma sheet produces a large pressure gradient, similar to the inflation of a balloon. If the tailward displacement of particles is energetically favorable, the system relaxes by releasing its pressure tailward, thus reducing the pressure gradient [and the current density, see (7)] locally. The consequence of this instability actually leads to a thinner plasma sheet with a higher current density tailward of the disruption site than before. For the idea of externally driven reduction in the magnetospheric convection electric field, it is postulated that the interplanetary electric field can readily penetrate into the magnetosphere. Northward turning or reduction of the dawn–dusk component of the IMF could reduce magnetospheric convection. Once the convection electric field is reduced sufficiently, higher energy ions will drift more rapidly westward than the lower energy ions due to the energy dependence of the curvature and gradient drifts. The change of field-aligned current is (13) where is the particle pressure and is the thermal energy. The dawn–dusk profile of the thermal energy due to the paron the dusk side and ticle drifts would lead to on the dawn side. As a result, a current system consistent with the substorm current wedge is set up. The development under this scheme is illustrated in Fig. 12 [43]. Fig. 13 depicts the model in the second category which is known as the near-Earth neutral line model [46], [47]. In essence, a near-Earth X-type magnetic neutral line is formed a few minutes at the downstream distance beyond 20 before substorm expansion onset on the ground. Fast Earthward plasma flows are created Earthward of the X-line. Braking of this flow in the near-Earth region creates an eastward inertial current in the equatorial plane which closes in the ionosphere to form the substorm current wedge. Tailward of the X-line, the plasma sheet is severed and a plasmoid is formed and ejected downstream. The neutral line stays in the near-Earth region until substorm recovery, at which time it moves downstream, causing the plasma sheet Earthward of the neutral line to thicken. The process responsible for the X-line formation is still an open question. Early studies suggested tearing mode instability [48], [49] but later studies found the mode to be heavily damped in a realistic current sheet whenever a small magnetic field normal to the current sheet is present [50], [51]. This instability appears to require imposing a large nonlinear tearing mode perturbation on the tail current sheet to drive it to a nonlinear stage or to require the aid of another process [52], [53]. The third category of substorm models emphasizes on the amplification of field-aligned currents by Alfvén waves bouncing between the ionosphere and the magnetosphere [54]–[56]. Due to a mismatch of impedance between the two regions, each reflection of Alfvén waves at the ionosphere amplifies the associated field-aligned current, thus producing a large positive feedback mechanism to generate the substorm current wedge. No specific prediction about the plasma sheet dynamics emerge from this approach. For instance, there is no definite prediction about thinning and thickening of plasma sheet, magnetic turbulence, plasma flows, and plasmoid formation. LUI: TUTORIAL ON GEOMAGNETIC STORMS AND SUBSTORMS 1863 Fig. 13. Main features for the near-Earth neutral line model. Magnetic reconnection in the mid-tail creates earthward plasma flows which brake at the near-Earth region. An eastward inertial current is set up which closes in the ionosphere by driving field-aligned currents downward at its eastward end and upward at its westward end [38]. The fourth category of substorm models uses abstract descriptions from nonlinear dynamics to obtain the statistical characteristics of substorms [57]–[61]. Under this approach, the power law slope found in distributions of energy dissipation, size, and duration associated with substorms could be understood in terms of a simple avalanche system. However, these models do not provide any specific physical process for substorm onsets which can be evaluated with observations. A synthesis of these models seems likely to be the most probable scenario of substorm development [62]. This is illustrated in Fig. 14 in which the columns depict, respectively, the auroral activity, the magnetospheric activity in the noon–midnight cross-section and at the equatorial projection during the different phases of a substorm. The location of substorm onset to be linked with the most equatorward auroral arc is due to the electric field associated with an auroral arc and the intense cross-tail current associated with the most equatorward arc [63]. The substorm expansion onset occurs when the intense cross-tail current built up in the near-Earth region suddenly becomes reduced drastically within a narrow longitudinal sector, as illustrated in the second row of the middle column. This current disruption arises from the onset of internal instabilities or convection reduction as described in the first category of substorm models, producing magnetic turbulence in the disruption region. The ionosphere plays a significant role in restraining or promoting further current diversion, incorporating an essential feature of the third category of models. The magnetosphere also plays a role in determining the eventual outcome of the activation by the free energy content in the activity site. If the free energy content is limited, even a favorable ionospheric response will not lead to a large dissipation. Current disruption, which can occur in multiple sites, leads to substorm injection and plasma sheet thickening. One of these current disruption sites may eventually develop to a large-scale magnetic reconnection site as envisioned by the second category of substorm models. This is one possible sequence, probably the most common scenario for substorms. Since solar wind condition can be very different for different substorms, it is quite conceivable that substorm development under some solar wind condition and past history of magnetospheric activity may deviate from the above scenario and resemble the sequence described by the second category. Therefore, it may not be meaningful to attribute one physical process to be responsible for all substorm onsets or a fixed sequence of for all substorm evolution [64]. Ideas developed along the category of current disruption are consistent with this philosophy. For instance, the latest theoretical work attributes turbulence associated with current disruption to more than one plasma instability, kinetic ballooning and cross-field current instabilities [65]. Furthermore, the threshold of current disruption depends crucially on the local condition and can be triggered over a wide range of downstream distance in the magnetotail. Therefore, current disruption occurring first in mid-tail region for some substorms is not unexpected. In addition, current disruption at one site instigates current disruption at the adjacent regions [38], [39] and therefore fits in well with the nonlinear dynamics treatment of substorms as a kind of avalanche event [66], [67]. VII. SOME OUTSTANDING QUESTIONS We conclude this tutorial by highlighting some important issues to be resolved in future work to gain better understanding of onsets of storms and substorms. The evidence for storms not due to the accumulated effects of substorms relies heavily on how well the Dst index portrays the true intensity of the ring current. The reliability of the Dst index in monitoring the ring current strength has been in question for a while [31]. There is another means to measure the ring current intensity which is being investigated only recently. This is through the monitoring of the energetic neutral atoms emissions [68]–[72] which could 1864 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 28, NO. 6, DECEMBER 2000 Fig. 14. Schematic diagram to illustrate the major elements in the substorm synthesis model. The substorm onset occurs in the near-Earth region with current disruption. The ionosphere plays a major role in enhancing or quenching further disruption of the cross-tail current. Current disruption spreads tailward and a near-Earth neutral line eventually develops at one of the current disruption sites. This synthesis incorporates key features in many substorm models. be used to infer the total energy content of the ring current ion population. This method will offer a way to determine the ring current intensity without the interference from the tail current system which the Dst index suffers. Another important issue on storms is to address the relative importance of different loss processes for the storm-time ring current during all the phases of a storm. For substorms, a large number of the onset mechanisms are not well understood. For instance, the near-Earth neutral line model calls for the formation of a neutral line in the mid-tail region. How is this crucial region formed? What is the non-MHD process operating in the diffusion region of magnetic reconnection? How is the mid-tail reconnection linked to field line of the most equatorward arc in the near-Earth region? How are magnetic fluctuations in current disruption events formed in this model? For models with near-Earth instabilities as the onset mechanisms, what is the nonlinear evolution of these instabilities? How do these instabilities couple to the large-scale evolution of the magnetosphere? Most of these questions are quite challenging, but their answers would definitely advance significantly our present understanding of storms and substorms. LUI: TUTORIAL ON GEOMAGNETIC STORMS AND SUBSTORMS ACKNOWLEDGMENT The author is grateful to M. W. Chen and E. E. Hume, Jr., for discussions on the early version of the manuscript. REFERENCES [1] B. T. Tsurutani, W. D. Gonzalez, Y. Kamide, and J. K. Arballo, “Magnetic storms,” in AGU Monograph 98 Washington, DC, 1997. [2] A. Schmidt, “Erdmagnetisumus,” in Enzyklopädie der mathematischen wissenschaften: Band VI, Leipzig, Germany, 1917. [3] S. Chapman, “An outline of a theory of magnetic storms,” Proc. R. Soc. London, Ser. A, vol. 95, p. 61, 1918. [4] L. A. Frank, “On the extraterrestrial ring current during geomagnetic storms,” J. Geophys. Res., vol. 72, p. 2753, 1967. [5] P. H. Smith and R. A. Hoffmann, “Ring current particle distributions during the magnetic storms of December 16–18, 1971,” J. Geophys. Res., vol. 78, p. 4731, 1973. [6] D. J. Williams, “Ring current composition and sources: An update,” Planet. Space Sci., vol. 29, p. 1195, 1981. [7] H. Alfvén, Cosmical Electrodynamics, Oxford, U.K.: Oxford University Press, 1950. [8] S. F. Singer, “A new model of magnetic storms and aurorae,” EOS Trans. AGU, vol. 38, p. 175, 1957. [9] A. T. Y. Lui, R. W. McEntire, and S. M. Krimigis, “Evolution of the ring current during two geomagnetic storms,” J. Geophys. Res., vol. 92, pp. 7459–7470, 1987. [10] A. J. Dessler and E. N. Parker, “Hydromagnetic theory of geomagnetic storms,” J. Geophys. Res., vol. 64, p. 2239, 1959. [11] N. Sckopke, “A general relation between the energy of trapped particles and the disturbance field near the earth,” J. Geophys. Res., vol. 71, p. 3125, 1966. [12] L. R. Lyons and D. J. Williams, Quantitative Aspects of Magnetospheric Physics. Dordrecht, The Netherlands: Reidel, 1984. [13] G. K. Parks, Physics of Space Plasmas. Redwood City, CA: AddisonWesley, Reidel, 1991. [14] M. G. Kivelson and C. T. Russell, Eds., Introduction to Space Physics, Cambridge, U.K.: Cambridge Univ. Press, 1995. [15] T. I. Gombosi, Physics of Space Environment, Cambridge, U.K.: Cambridge Univ. Press, 1998. [16] C. R. Chappell, T. E. Moore, and J. H. Waite, Jr., “The ionosphere as a fully adequate source of plasma for the Earth’s magnetosphere,” J. Geophys. Res., vol. 92, p. 5896, 1987. [17] A. W. Yau, E. G. Shelley, and W. K. Peterson, “Energetic auroral and polar ion outflow at DE 1 altitudes: Magnitude, composition, magnetic activity dependence, and long-term variations,” J. Geophys. Res., vol. 90, p. 8417, 1985. [18] D. C. Hamilton, G. Gloeckler, F. M. Ipavich, W. Studemann, B. Wilken, and G. Kremser, “Ring current development during the great geomagnetic storm of February 1986,” J. Geophys. Res., vol. 93, p. 14 343, 1988. [19] S. M. Krimigis, D. Venkatesan, J. C. Barichello, and E. T. Sarris, “Simultaneous measurements of energetic protons and electrons in the distant magnetosheath, magnetotail and upstream in the solar wind,” Geophys. Res. Lett., vol. 5, p. 961, 1978. [20] S.-I. Akasofu, Polar and Magnetospheric Substorms. Norwell, MA.: Reidel, 1968, p. 5. [21] C. E. McIlwain, “Substorm injection boundaries,” in Magnetospheric Physics, B. M. McCormac, Ed. Hingham, MA: Reidel, 1974, p. 143. [22] L. R. Lyons and M. Schulz, “Access of energetic particles to stormtime ring current through enhanced radial diffusion,” J. Geophys. Res., vol. 94, p. 5491, 1989. [23] C.-G. Fälthammar, “Effects of time dependent electric fields on geomagnetically trapped radiation,” J. Geophys. Res., vol. 70, p. 2503, 1965. [24] W. D. Gonzalez and B. T. Tsurutani, “Criteria of interplanetary parameters causing intense magnetic storms ( 100 nT),” Planet. Space Sci., vol. 35, p. 1101, 1987. [25] T. Iyemori and D. R. K. Rao, “Decay of the field of geomagnetic disturbance after substorm onset and its implication to storm-substorm relation,” Ann. Geophysicae, vol. 14, p. 608, 1996. [26] R. A. Wolf, J. W. Freeman, Jr., B. A. Hausman, R. W. Spiro, R. B. Hilmer, and R. L. Lambour, “Modeling convection effects in magnetic storms,” in Magnetic Storms, B. T. Tsurutani and Y. Kamide, Eds. Washington, D.C.: AGU, 1996, p. 161. [27] M.-C. Fok, T. E. Moore, and D. C. Delcourt, “Modeling of inner plasma sheet and ring current during substorms,” J. Geophys. Res., vol. 104, p. 14 557, 1996. Dst < 0 Dst 1865 [28] M. W. Chen, M. Schulz, and L. R. Lyons, “Simulations of phase space distributions of stormtime proton ring current,” J. Geophys. Res., vol. 99, p. 5745, 1994. [29] M.-C. Fok , T. E. Moore , and M. E. Greenspan, “Ring current development during storm main phase,” J. Geophys. Res., vol. 101, p. 15 311, 1996. [30] G. Rostoker, W. Baumjohann, W. D. Gonzalez, Y. Kamide, S. Kokubun, R. L. McPherron, and B. T. Tsurutani, “Comment on field of geomagnetic disturbance after substorm ‘Decay of the onset and its implication to storm-substorm relation’,” Ann. Geophys., vol. 15, p. 848, 1997. [31] W. H. Campbell, “Geomagnetic storms, the ring-current myth and lgonormal distributions,” J. Atm. Terr. Phys., vol. 58, pp. 1171–1181, 1996. [32] W. Sun and S.-I. Akasofu, “On the formation of the storm-time ring current belt,” , submitted for publication. [33] J. U. Kozyra, V. K. Jordanova, J. E. Borovsky, M. F. Thomsen, D. J. Knipp, D. S. Evans, D. J. McComas, and T. E. Cayton, “Effects of a high-density plasma sheet on ring current development during the November 2–6, 1993, magnetic storm,” J. Geophys. Res., vol. 103, p. 26 285, 1998. [34] N. Yokoyama and Y. Kamide, “Statistical nature of geomagnetic storms,” J. Geophys. Res., vol. 102, no. 86, p. 189, 1981. [35] S.-I. Akasofu, “The development of the auroral substorm,” Planet. Space Sci., vol. 12, pp. 273–282, 1964. [36] R. L. McPherron, C. T. Russell, and M. Aubry, “Satellite studies of magnetospheric substorms on August 15, 1968, 9, Phenomenological model for substorms,” J. Geophys. Res., vol. 78, p. 3131, 1973. [37] A. T. Y. Lui, D. J. Williams, R. W. McEntire, S. Ohtani, L. J. Zanetti, W. A. Bristow, R. A. Greenwald, P. T. Newell, S. P. Christon, T. Mukai, K. Tsuruda, T. Yamamoto, S. Kokubun, H. Matsumoto, H. Kojima, T. Murata, D. H. Fairfield, R. P. Lepping, J. C. Samson, G. Rostoker, G. D. Reeves, A. S. Rodger, and H. J. Singer, “Multi-point study of a substorm on February 9, 1995,” J. Geophys. Res., vol. 103, pp. 17 333–17 343, 1998. [38] A. T. Y. Lui, C.-L. Chang, A. Mankofsky, H.-K. Wong, and D. Winske, “A cross-field current instability for substorm expansions,” J. Geophys. Res., vol. 96, p. 11 389, 1991. [39] A. T. Y. Lui, “Current disruption in the Earth’s magnetosphere: Observations and models,” J. Geophys. Res., vol. 101, p. 13 067, 1996. [40] A. Roux, S. Perraut, P. Robert, A. Morane, A. Pedersen, A. Korth, G. Kremser, B. Aparicio, D. Rodgers, and R. Pellinen, “Plasma sheet instability related to the westward traveling surge,” J. Geophys. Res., vol. 96, p. 17 697, 1991. [41] G. M. Erickson, “Substorm theories: United they stand, divided they fall,” Rev. Geophys., vol. 33, p. 685, 1995. [42] W. W. Liu, “Disruption of thin current sheets: A two-fluid theory,” J. Geophys. Res., vol. 102, p. 14 331, 1997. [43] L. R. Lyons, “Substorms: Fundamental observational features, distinction from other disturbances, and external triggering,” J. Geophys. Res., vol. 101, p. 13 011, 1996. [44] A. T. Y. Lui, P. H. Yoon, and C.-L. Chang, “Quasi-linear analysis of ion Weibel instability in the Earth’s neutral sheet,” J. Geophys. Res., vol. 98, p. 153, 1993. [45] S.-I. Ohtani and T. Tamao, “Does the ballooning instability trigger substorms in the near-Earth magnetotail,” J. Geophys. Res., vol. 98, p. 19 369, 1993. [46] E. W. Hones, Jr., J. R. Asbridge, S. J. Bame, and S. Singer, “Substorm = 6 to variations of the magnetotail plasma sheet from = 60 ,” J. Geophys. Res., vol. 78, p. 109, 1973. [47] K. Shiokawa, G. Haerendel, and W Baumjohann, “Azimuthal pressure gradient as driving force of substorm currents,” J. Geophys. Res., vol. 25, p. 963, 1998. [48] B. Coppi, G. Laval, and R. Pellat, “Dynamics of the geomagnetic tail,” Phys. Rev. Lett., vol. 16, p. 1207, 1966. [49] K. Schindler, “A theory of the substorm mechanism,” J. Geophys. Res., vol. 79, p. 2803, 1974. [50] R. Pellat, F. V. Coroniti, and P. L. Pritchett, “Does ion tearing exist?,” Geophys. Res. Lett., vol. 18, p. 143, 1991. [51] M. Brittnacher, K. B. Quest, and H. Karimabadi, “A study of the effect of pitch angle and spatial diffusion on tearing instability using a new finite element based linear code,” J. Geophys. Res., vol. 103, p. 4587, 1998. [52] A. P. Kropotkin, O. O. Trubachev, and A. T. Y. Lui, “Nonlinear instability of the geomagnetotail current sheet combining the features of tearing and cross-field current instabilities,” J. Geophys. Res., vol. 104, p. 371, 1999. Dst Dst X 0 R X 0R 1866 [53] M. I. Sitnov and A. T. Y. Lui, “Cross-field current instability as a catalyst of the explosive reconnection in the geomagnetotail,” J. Geophys. Res., vol. 104, p. 6941, 1999. [54] P. L. Rothwell, L. P. Block, M. B. Silevitch, and C.-G. Falthammer, “A new model for substorm onsets: The pre-breakup and triggering regimes,” Geophys. Res. Lett., vol. 15, p. 1279, 1988. [55] J. R. Kan, “A global magnetosphere-ionosphere coupling model of substorms,” J. Geophys. Res., vol. 98, p. 17 263, 1993. [56] R. L. Lysak, “Feedback instability of the ionospheric resonant cavity,” J. Geophys. Res., vol. 96, p. 1553, 1991. [57] T. S. C. Chang, “Low dimensional behavior and symmetry breaking of stochastic systems near criticality—Can these effects be observed in space and in the laboratory?,” IEEE Trans. Plasma Sci., vol. 20, p. 691, 1992. [58] G. Consolini, M. F. Marcucci, and M. Candidi, “Multifractal structure of auroral electrojet index data,” Phys. Rev. Lett., vol. 76, no. 21, p. 4082, 1996. [59] A. J. Klimas, D. Vassiliadis, D. N. Baker, and D. A. Roberts, “The organized nonlinear dynamics of the magnetosphere,” J. Geophys. Res., vol. 101, p. 13 089, 1996. [60] G. Consolini and P. D. Michelis, “Non-Gaussian distribution function of AE-index fluctuations: Evidence for time intermittency,” Geophys. Res. Lett., vol. 25, p. 4087, 1998. [61] S. C. Chapman, N. W Watkins, R. O. Dendy, P. Helander, and G. Rowlands, “A simple avalanche model as an analogue for magnetospheric activity,” Geophys. Res. Lett., vol. 25, p. 2397, 1998. [62] A. T. Y. Lui, “A synthesis of magnetospheric substorm models,” J. Geophys. Res., vol. 96, p. 1849, 1991. [63] A. T. Y. Lui and J. S. Murphree, “A substorm model with onset location tied to an auroral arc,” Geophys. Res. Lett., vol. 25, p. 1269, 1998. [64] A. T. Y. Lui, “Magnetospheric substorms,” Phys. Fluids B, vol. 4, no. 7, p. 2257, 1992. [65] C. Z. Cheng and A. T. Y. Lui, “Kinetic ballooning instability for substorm onset and current disruption observed by AMPTE/CCE,” Geophys. Res. Lett., vol. 25, p. 4091, 1998. [66] G. Consolini et al., “Sandpile cellular automata and magnetospheric dynamics,” in Cosmic Physics in the Year 2000, Aiello et al., Eds, Bologna, Italy: SIF, 1997, vol. 58. View publication stats IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 28, NO. 6, DECEMBER 2000 [67] S. C. Chapman, N. W. Watkins, R. O. Dendy, P. Helander, and G. Rowlands, “A simple avalanche model as an analogue for magnetospheric activity,” Geophys. Res. Lett., vol. 25, p. 2397, 1998. [68] E. C. Roelof, D. G. Mitchell, and D. J. Williams, “Energetic neutral atoms ( 50 keV) from the ring current IMP-7/8 and ISEE-1,” J. Geophys. Res., vol. 90, p. 10 991, 1985. [69] A. T. Y. Lui, D. J. Williams, E. C. Roelof, R. W. McEntire, and D. G. Mitchell, “First composition measurements of energetic neutral atoms,” Geophys. Res. Lett., vol. 23, p. 2641, 1996. [70] A. Milillo and S. Orsini, “Low-altitude energetic neutral atoms imaging of the inner magnetosphere: A geometrical method to identify the energetic neutral atoms contributions from different magnetospheric regions,” J. Geophys. Res., vol. 101, p. 27 123, 1996. [71] M. G. Henderson, G. D. Reeves, H. E. Spence, R. B. Sheldon, A. M. Jorgensen, J. B. Blake, and J. F. Fennell, “First energetic neutral atom images from polar,” Geophys. Res. Lett., vol. 24, p. 1167, 1997. [72] A. M. Jorgensen, H. E. Spence, M. G. Henderson, G. D. Reeves, M. Sugiura, and T. Kamei, “Global energetic neutral atom (ENA) measureindex,” Geophys. Res. Lett., ments and their association with the vol. 24, p. 3173, 1997. E Dst Anthony T. Y. Lui received the Ph.D. degree from the University of Calgary, Calgary, AB, Canada, in 1974. He is a Principal Professional Staff at the Applied Physics Laboratory, The Johns Hopkins University, Baltimore, MD. His main interests are substorms, magnetotail, and aurora in which he has performed data analyses and developed theories to understand these natural phenomena.