Observations of birkeland currents

Recent measurements of precipitating energetic particles and vector magnetic fields from satellites and sounding rockets have verified the existence of geomagnetically-aligned electric currents at high latitudes in the ionosphere and magnetosphere. The spatial and temporal configuration of such currents, now commonly called Birkeland currents, has delineated their role in providing ionospheric closure of magnetospheric current systems, and gross features of these current systems may be understood in terms of theoretical models of magnetospheric convection. The association of Birkeland currents with auroral features on a very small scale suggests that auroral acceleration may result from the current flow.     

Energy flux in the Earth's magnetosphere: Storm – substorm relationship

Three ways of the energy transfer in the Earth's magnetosphere are studied. The solar wind MHD generator is an unique energy source for all magnetospheric processes. Field-aligned currents directly transport the energy and momentum of the solar wind plasma to the Earth's ionosphere. The magnetospheric lobe and plasma sheet convection generated by the solar wind is another magnetospheric energy source. Plasma sheet particles and cold ionospheric polar wind ions are accelerated by convection electric field. After energetic particle precipitation into the upper atmosphere the solar wind energy is transferred into the ionosphere and atmosphere. This way of the energy transfer can include the tail lobe magnetic field energy storage connected with the increase of the tail current during the southward IMF. After that the magnetospheric substorm occurs. The model calculations of the magnetospheric energy give possibility to determine the ground state of the magnetosphere, and to calculate relative contributions of the tail current, ring current and field-aligned currents to the magnetospheric energy. The magnetospheric substorms and storms manifest that the permanent solar wind energy transfer ways are not enough for the covering of the solar wind energy input into the magnetosphere. Nonlinear explosive processes are necessary for the energy transmission into the ionosphere and atmosphere. For understanding a relation between substorm and storm it is necessary to take into account that they are the concurrent energy transferring ways. This revised version was published online in August 2006 with corrections to the Cover Date.     

Auroral flares and solar flares

The morphology of development of auroral flares (magnetospheric substorms) for both electron and proton auroras is summarized, based on ground-based as well as rocket-borne and satellite-borne data with specific reference to the morphology of solar flares.The growth phase of an auroral flare is produced by the inflow of the solar wind energy into the magnetosphere by the reconnection mechanism between the solar wind field and the geomagnetic field, thus the neutral and plasma sheets in the magnetotail attaining their minimum thickness with a great stretch of the geomagnetic fluxes into the tail.The onset of the expansion phase of an auroral flare is represented by the break-up of electron and proton auroras, which is associated with strong auroral electrojets, a sudden increase in CNA, VLF hiss emissions and characteristic ULF emissions. The auroral break-up is triggered by the relaxation of stretched magnetic fluxes caused by cutting off of the tail fluxes at successively formed X-type neutral lines in the magnetotail.The resultant field-aligned currents flowing between the tailward magnetosphere and the polar ionosphere produce the field-aligned anomalous resistivity owing to the electrostatic ion-cyclotron waves; the electrical potential drop thus increased further accelerates precipitating charged particles with a result of the intensification of both the field-aligned currents and the auroral electrojet. It seems that the rapid building-up of this positive feedback system for precipitating charged particles is responsible for the break-up of an auroral flare.     

Energetics of the magnetosphere

Energy flow in various large-scale processes of the Earth's magnetosphere is examined. This energy comes from the solar wind, via the dawn-to-dusk convection electric field, a field established primarily by magnetic merging but with viscous-like boundary interaction as a possible contributor. The convection field passes about 5 × 10¹¹ W to the near-Earth part of the plasma sheet, and also moves the plasma earthward. In addition, 1–3 × 10¹¹ W are given to the complex system of the Birkeland currents: about 4 × 10¹⁰ of this, on the average, goes to parallel acceleration, chiefly of auroral electrons, about 2–3 times that amount to joule heating of the ionosphere, and the rest heats the ring current. The ring current stores energy (mainly as kinetic energy of particles) of the order of 2 × 10¹⁵ J, and this value rises and decays during magnetic storms, on time scales ranging from a fraction of a day to several days. The tail can store comparable amounts as magnetic energy, and appreciable fractions of its energy may be released in substorms, on time scales of tens of minutes. The sporadic power level of such events reaches the order of 3 × 10¹² W. The role of magnetic merging in such releases of magnetic energy is briefly discussed, as is the correlation between properties of the solar wind and magnetospheric power levels.     

Cusp dynamics and ionospheric outflow

One of the IMAGE mission science goals is to understand the dayside auroral oval and its dynamic relationship to the magnetosphere. Two ways the auroral oval is dynamically coupled to the magnetosphere are through the injection of magnetosheath plasma into the magnetospheric cusps and through the ejection of ionospheric plasma into the magnetosphere. The ionospheric footpoints of the Earth's magnetospheric cusps are relatively narrow regions in invariant latitude that map magnetically to the magnetopause. Monitoring the cusp reveals two important aspects of magnetic reconnection at the magnetopause. Continuous cusp observations reveal the relative contributions of quasi-steady versus impulsive reconnection to the overall transfer of mass, energy, and momentum across the magnetopause. The location of the cusp is used to determine where magnetic reconnection is occurring on the magnetopause. Of particular interest is the distinction between anti-parallel reconnection, where the magnetosheath and magnetospheric field lines are strictly anti-parallel, and component merging, where the magnetosheath and magnetospheric field lines have one component that is anti-parallel. IMAGE observations suggest that quasi-steady, anti-parallel reconnection is occurring in regions at the dayside magnetopause. However, it is difficult to rule out additional component reconnection using these observations. The ionospheric footpoint of the cusp is also a region of relatively intense ionospheric outflow. Since outflow also occurs in other regions of the auroral oval, one of the long-standing problems has been to determine the relative contributions of the cusp/cleft and the rest of the auroral oval to the overall ionospheric ion content in the Earth's magnetosphere. While the nature of ionospheric outflow has made it difficult to resolve this long-standing problem, the new neutral atom images from IMAGE have provided important evidence that ionospheric outflow is strongly controlled by solar wind input, is `prompt' in response to changes in the solar wind, and may have very narrow and distinct pitch angle structures and charge exchange altitudes.     

Steady magnetospheric convection: A review of recent results

Theoretical pressure balance arguments have implied that steady convection is hardly possible in the terrestrial magnetotail and that steady energy input necessarily generates a cyclic loading-unloading sequence, i.e., repetitive substorms. However, observations have revealed that enhanced solar wind energy input to the magnetospheric system may either lead to substorm activity or enhanced but steady convection. This topic is reviewed with emphasis on several recent case studies of the Steady Magnetospheric Convection (SMC) events. In these cases extensive data sets from both satellite and ground-based instruments from various magnetospheric and ionospheric regions were available.Accurate distinction of the spatial and temporal scales of the magnetospheric processes is vital for correct interpretation of the observations during SMC periods. We show that on the large scale, the magnetospheric configuration and plasma convection are stable during SMC events, but that both reveal considerable differences from their quiet-time assemblies. On a shorter time scale, there are numerous transient activations which are similar to those found during substorms, but which presumably originate from a more distant tail reconnection process, and map to the poleward boundary of the auroral oval. The available observations and the unresolved questions are summarized here.The tail magnetic field during SMC events resembles both substorm growth and recovery phases in the neartail and midtail, respectively, but this configuration may remain stable for up to ten hours. Based on observations and model results we discuss how the magnetospheric system avoids pressure balance problems when the plasma convects earthward.Finally, the importance of further coordinated studies of SMC events is emphasized. Such studies may shed more light on the substorm dynamics and help to verify quantitatively the theoretical models of the convecting magnetosphere.     

High-Latitude Particle Precipitation and its Relationship to Magnetospheric Source Regions

Two central issues in magnetospheric research are understanding the mapping of the low-altitude ionosphere to the distant regions of the magnetsphere, and understanding the relationship between the small-scale features detected in the various regions of the ionosphere and the global properties of the magnetosphere. The high-latitude ionosphere, through its magnetic connection to the outer magnetosphere, provides an important view of magnetospheric boundaries and the physical processes occurring there. All physical manifestations of this magnetic connectivity (waves, particle precipitation, etc.), however, have non-zero propagation times during which they are convected by the large-scale magnetospheric electric field, with phenomena undergoing different convection distances depending on their propagation times. Identification of the ionospheric signatures of magnetospheric regions and phenomena, therefore, can be difficult. Considerable progress has recently been made in identifying these convection signatures in data from low- and high-altitude satellites. This work has allowed us to learn much about issues such as: the rates of magnetic reconnection, both at the dayside magnetopause and in the magnetotail; particle transport across the open magnetopause; and particle acceleration at the magnetopause and the magnetotail current sheets.     

Dayside Proton Aurora: Comparisons between Global MHD Simulations and IMAGE Observations

The IMAGE mission provides a unique opportunity to evaluate the accuracy of current global models of the solar wind interaction with the Earth's magnetosphere. In particular, images of proton auroras from the Far Ultraviolet Instrument (FUV) onboard the IMAGE spacecraft are well suited to support investigations of the response of the Earth's magnetosphere to interplanetary disturbances. Accordingly, we have modeled two events that occurred on June 8 and July 28, 2000, using plasma and magnetic field parameters measured upstream of the bow shock as input to three-dimensional magnetohydrodynamic (MHD) simulations. This paper begins with a discussion of images of proton auroras from the FUV SI-12 instrument in comparison with the simulation results. The comparison showed a very good agreement between intensifications in the auroral emissions measured by FUV SI-12 and the enhancement of plasma flows into the dayside ionosphere predicted by the global simulations. Subsequently, the IMAGE observations are analyzed in the context of the dayside magnetosphere's topological changes in magnetic field and plasma flows inferred from the simulation results. Finding include that the global dynamics of the auroral proton precipitation patterns observed by IMAGE are consistent with magnetic field reconnection occurring as a continuous process while the IMF changes in direction and the solar wind dynamic pressure varies. The global simulations also indicate that some of the transient patterns observed by IMAGE are consistent with sporadic reconnection processes. Global merging patterns found in the simulations agree with the antiparallel merging model, though locally component merging might broaden the merging region, especially in the region where shocked solar wind discontinuities first reach the magnetopause. Finally, the simulations predict the accretion of plasma near the bow shock in the regions threaded by newly open field lines on which plasma flows into the dayside ionosphere are enhanced. Overall the results of these initial comparisons between global MHD simulation results and IMAGE observations emphasize the interplay between reconnection and dynamic pressure processes at the dayside magnetopause, as well as the intricate connection between the bow shock and the auroral region.     

Electrodynamic coupling of the magnetosphere and ionosphere

The auroral zone ionosphere is coupled to the outer magnetosphere by means of field-aligned currents. Parallel electric fields associated with these currents are now widely accepted to be responsible for the acceleration of auroral particles. This paper will review the theoretical concepts and models describing this coupling. The dynamics of auroral zone particles will be described, beginning with the adiabatic motions of particles in the converging geomagnetic field in the presence of parallel potential drops and then considering the modifications to these adiabatic trajectories due to wave-particle interactions. The formation of parallel electric fields can be viewed both from microscopic and macroscopic viewpoints. The presence of a current carrying plasma can give rise to plasma instabilities which in a weakly turbulent situation can affect the particle motions, giving rise to an effective resistivity in the plasma. Recent satellite observations, however, indicate that the parallel electric field is organized into discrete potential jumps, known as double layers. From a macroscopic viewpoint, the response of the particles to a parallel potential drop leads to an approximately linear relationship between the current density and the potential drop.The currents flowing in the auroral circuit must close in the ionosphere. To a first approximation, the ionospheric conductivity can be considered to be constant, and in this case combining the ionospheric Ohm's Law with the linear current-voltage relation for parallel currents leads to an outer scale length, above which electric fields can map down to the ionosphere and below which parallel electric fields become important. The effects of particle precipitation make the picture more complex, leading to enhanced ionization in upward current regions and to the possibility of feedback interactions with the magnetosphere.Determining adiabatic particle orbits in steady-state electric and magnetic fields can be used to determine the self-consistent particle and field distributions on auroral field lines. However, it is difficult to pursue this approach when the fields are varying with time. Magnetohydrodynamic (MHD) models deal with these time-dependent situations by treating the particles as a fluid. This class of model, however, cannot treat kinetic effects in detail. Such effects can in some cases be modeled by effective transport coefficients inserted into the MHD equations. Intrinsically time-dependent processes such as the development of magnetic micropulsations and the response of the magnetosphere to ionospheric fluctuations can be readily treated in this framework.The response of the lower altitude auroral zone depends in part on how the system is driven. Currents are generated in the outer parts of the magnetosphere as a result of the plasma convection. The dynamics of this region is in turn affected by the coupling to the ionosphere. Since dissipation rates are very low in the outer magnetosphere, the convection may become turbulent, implying that nonlinear effects such as spectral transfer of energy to different scales become important. MHD turbulence theory, modified by the ionospheric coupling, can describe the dynamics of the boundary-layer region. Turbulent MHD fluids can give rise to the generation of field-aligned currents through the so-called -effect, which is utilized in the theory of the generation of the Earth's magnetic field. It is suggested that similar processes acting in the boundary-layer plasma may be ultimately responsible for the generation of auroral currents.     

MESSENGER: Exploring Mercury’s Magnetosphere

The MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER) mission to Mercury offers our first opportunity to explore this planet’s miniature magnetosphere since the brief flybys of Mariner 10. Mercury’s magnetosphere is unique in many respects. The magnetosphere of Mercury is among the smallest in the solar system; its magnetic field typically stands off the solar wind only ∼1000 to 2000 km above the surface. For this reason there are no closed drift paths for energetic particles and, hence, no radiation belts. Magnetic reconnection at the dayside magnetopause may erode the subsolar magnetosphere, allowing solar wind ions to impact directly the regolith. Inductive currents in Mercury’s interior may act to modify the solar wind interaction by resisting changes due to solar wind pressure variations. Indeed, observations of these induction effects may be an important source of information on the state of Mercury’s interior. In addition, Mercury’s magnetosphere is the only one with its defining magnetic flux tubes rooted beneath the solid surface as opposed to an atmosphere with a conductive ionospheric layer. This lack of an ionosphere is probably the underlying reason for the brevity of the very intense, but short-lived, ∼1–2 min, substorm-like energetic particle events observed by Mariner 10 during its first traversal of Mercury’s magnetic tail. Because of Mercury’s proximity to the sun, 0.3–0.5 AU, this magnetosphere experiences the most extreme driving forces in the solar system. All of these factors are expected to produce complicated interactions involving the exchange and recycling of neutrals and ions among the solar wind, magnetosphere, and regolith. The electrodynamics of Mercury’s magnetosphere are expected to be equally complex, with strong forcing by the solar wind, magnetic reconnection, and pick-up of planetary ions all playing roles in the generation of field-aligned electric currents. However, these field-aligned currents do not close in an ionosphere, but in some other manner. In addition to the insights into magnetospheric physics offered by study of the solar wind–Mercury system, quantitative specification of the “external” magnetic field generated by magnetospheric currents is necessary for accurate determination of the strength and multi-polar decomposition of Mercury’s intrinsic magnetic field. MESSENGER’s highly capable instrumentation and broad orbital coverage will greatly advance our understanding of both the origin of Mercury’s magnetic field and the acceleration of charged particles in small magnetospheres. In this article, we review what is known about Mercury’s magnetosphere and describe the MESSENGER science team’s strategy for obtaining answers to the outstanding science questions surrounding the interaction of the solar wind with Mercury and its small, but dynamic, magnetosphere.     

Magnetic structure of the boundary layer

It now appears that magnetospheric convection is driven by both magnetic reconnection and ‘viscous’ dragging of closed flux tubes acting in varying proportions but with reconnection being, on the average, the predominant cause. But the action of the closed flux tubes of the low latitude boundary layer seems predominant in driving system 1 field aligned currents and discrete auroras. A search of ISEE boundary layer data for the magnetic shear effects associated with field aligned currents has revealed, on some occasions, a ‘reverse draping’ of boundary layer field lines whose cause might be plasma entry around the cusps, a north-south asymmetry in current flow to the polar regions, or magnetic reconnection at high latitudes. Observed asymmetries in the nature and distribution of boundary layer encounters suggest that the boundary layer and/or the plasma mantle may differ substantially at the dawn and dusk sides of the magnetosphere and that there may be seasonal dependencies of their properties.     

Flow Bursts in the Plasma Sheet and Auroral Substorm Onset: Observational Constraints on Connection Between Midtail and Near-earth Substorm Processes

Causality between near-Earth and midtail substorm processes is one of the most controversial issues about the substorm trigger mechanism. The currently most popular model, the outside-in model, assumes that near-Earth reconnection is initiated in the midtail region before substorm onset and that the associated flow burst causes tail current disruption in the near-Earth region. However, there remain some outstanding issues that may serve as critical tests of this model. The present article reviews recent satellite and ground observations addressing three such critical issues with a focus on substorm-related auroral features. First, near-Earth reconnection, even if it reaches the lobe magnetic field, does not necessarily trigger a global substorm, but it is often related to a pseudobreakup. This fact suggests that there is an additional or alternative condition for substorm development. Secondly, although there appears to be one-to-one correspondence between flow bursts in the plasma sheet and equatorward-moving auroral structures (auroral streamers), no such auroral feature that can be associated with the fast plasma flow can be identified prior to auroral breakups. On the other hand, the flow burst is widely regarded as a manifestation of reconnection and therefore, according to the outside-in model, should be created in the near-Earth plasma sheet before substorm onset. Finally, auroral arcs poleward of a breakup arc are not affected until the front of auroral intensification reaches those arcs. The last two points suggest that if substorm is triggered as the outside-in model describes, the ionosphere is electromagnetically detached from the magnetosphere, which, however, has not been addressed theoretically. Thus, it should be crucial for a better understanding of the substorm trigger process to implement the magnetosphere-ionosphere coupling in future modeling efforts and to address those basic issues as a guide for critically evaluating each model.     

The magnetotail and substorms

The tail plays a very active and important role in substorms. Magnetic flux eroded from the dayside magnetosphere is stored here. As more and more flux is transported to the magnetotail and stored, the boundary of the tail flares more, the field strength in the tail increases, and the currents strengthen and move closer to the Earth. Further, the plasma sheet thins and the magnetic flux crossing the neutral sheet lessens. At the onset of the expansion phase, the stored magnetic flux is returned from the tail and energy is deposited in the magnetosphere and ionosphere. During the expansion phase of isolated substorms, the flaring angle and the lobe field strength decrease, the plasma sheet thickens and more magnetic flux crosses the neutral sheet.In this review, we discuss the experimental evidence for these processes and present a phenomenological or qualitative model of the substorm sequence. In this model, the flux transport is driven by the merging of the magnetospheric and interplanetary magnetic fields. During the growth phase of substorms the merging rate on the dayside magnetosphere exceeds the reconnection rate in the neutral sheet. In order to remove the oversupply of magnetic flux in the tail, a neutral point forms in the near earth portion of the tail. If the new reconnection rate exceeds the dayside merging rate, then an isolated substorm results. However, a situation can occur in which dayside merging and tail reconnection are in equilibrium. The observed polar cap electric field and its correlation with the interplanetary magnetic field is found to be in accord with open magnetospheric models.     

Relationship Between Substorm Auroras and Processes in the Near-Earth Magnetotail

Although the auroral substorm has been long regarded as a manifestation of the magnetospheric substorm, a direct relation of active auroras to certain magnetospheric processes is still debatable. To investigate the relationship, we combine the data of the UV imager onboard the Polar satellite with plasma and magnetic field measurements by the Geotail spacecraft. The poleward edge of the auroral bulge, as determined from the images obtained at the LHBL passband, is found to be conjugated with the region where the oppositely directed fast plasma flows observed in the near-Earth plasma sheet during substorms are generated. We conclude that the auroras forming the bulge are due to the near-Earth reconnection process. This implies that the magnetic flux through the auroral bulge is equal to the flux dissipated in the magnetotail during the substorm. Comparison of the magnetic flux through the auroral bulge with the magnetic flux accumulated in the tail lobe during the growth phase shows that these parameters have the comparable values. This is a clear evidence of the loading–unloading scheme of substorm development. It is shown that the area of the auroral bulge developing during substorm is proportional to the total (magnetic plus plasma) pressure decrease in the magnetotail. These findings stress the importance of auroral bulge observations for monitoring of substorm intensity in terms of the magnetic flux and energy dissipation.     

Ionosphere-magnetosphere coupling

The large-scale electrical coupling between the ionosphere and magnetosphere is reviewed, particularly with respect to behavior on time scales of hours or more. The following circuit elements are included: (1) the magnetopause boundary layer, which serves as the generator for the magnetospheric-convection circuit; (2) magnetic field lines, usually good conductors but sometimes subject to anomalous resistivity; (3) the ionosphere, which can conduct current across magnetic field lines; (4) the magnetospheric particle distributions, including tail current and partial-ring currents. Magnetic merging and a viscous interaction are considered as possible generating mechanisms, but merging seems the most likely alternative. Several mechanisms have been proposed for causing large potential drops along magnetic field lines in the upper ionosphere, and many isolated measurements of parallel electric fields have been reported, but the global pattern and significance of these electric fields are unknown. Ionospheric conductivities are now thoroughly measured, but are highly variable. Simple self-consistent theoretical models of the magnetospheric-convection system imply that the magnetospheric particles should shield the inner magnetosphere and low-latitude ionosphere from most of the time-average convection electric field.     

MHD ANALYSIS OF PETSCHEK-TYPE RECONNECTION IN NON-UNIFORM FIELD AND FLOW GEOMETRIES

Analytical studies of reconnection have, for the most part, been confined to steady and uniform current sheet geometries. In contrast to these implifications, natural phenomena associated with the presence of current sheets indicate highly non-uniform structure and time-varying behaviour. Examples include the violent outbursts of energy on the Sun known as solar flares, and magnetospheric phenomena such as flux transfer events, plasmoids, and auroral activity. Unlike the theoretical models, reconnection therefore occurs in a highly dynamic and structured plasma environment. In this article we review the mathematical tools and techniques which are available to formulate models capable of describing the effects of reconnection in such situations. We confine attention to variants of the reconnection model first discussed by Petschek in the 1960s, in view of its successful application in predicting and interpreting phenomena in the terrestrial magnetosphere. The analysis of Petschek-type reconnection is based on the equations of ideal magnetohydrodynamics (MHD), which describe the large-scale behaviour of the magnetic field and plasma flow outside the diffusion region, which we determine as a localised part of the current sheet in which reconnection is initiated. The approach we adopt here is to transform the MHD equations into a Lagrangian or so-called 'frozen-in' coordinate system. In this coordinate system, the equation of motion transforms into a set of coupled nonlinear equations, in which the presence of inhomogeneous magnetic fields and/or plasma flows gives rise to a term similar to that which appears in the study of the ordinary string equation in a non-homogeneous medium. As demonstrated here, this approach not only clarifies and highlights the effects of such non-uniformities, it also simplifies the solution of the original set of MHD equations. In particular, this is true for those types of problem in which the total pressure can be considered as a known quantity from the outset. To illustrate the method, we solve several 2D problems involving magnetic field and flow non-uniformities: reconnection in a stagnation-point flow geometry with antiparallel magnetic fields; reconnection in a Y-type magnetic field geometry with and without velocity shear across the current sheet; and reconnection in a force-free magnetic field geometry with field lines of the form xy = const. These case examples, chosen for their tractability, each incorporate some aspects of the field and flow geomtries encountered in solar-terrestrial applications, and they provide a starting point for further analytical as well as numerical studies of reconnection.     

Towards a unified theory of discrete auroras

Recent theoretical developments in auroral research are reviewed. In spite of the great complexity involved, a unified theory begins to emerge. The framework of this unified theory is provided by the imperfect magnetosphere-ionosphere coupling due to enhanced magnetospheric convection. The auroral potential structures in the imperfect coupling state result from the loss cone constriction on enhanced upward field-aligned currents following enhanced magnetospheric convections. Field-aligned potential drops for the inverted V precipitations are supported by a combination of the double layer process, the pitch-angle anisotropy process, the wave-particle interaction (anomalous resistivity) process, the electrostatic shock and the thermoelectric process in descending order of importance. Quasineutrality along inverted V field lines is maintained primarily by the trapped and back-scattered electrons produced respectively by wave-particle interactions in the magnetosphere and electron-neutral collisions in the ionosphere. Additional electron energization due to the ion cyclotron turbulence may lead to striations along the inverted V field lines to produce thin auroral arcs imbedded in the inverted V precipitation. The formation of discrete auroras can now be understood as a manifestation of the imperfect magnetosphere-ionosphere coupling due to enhanced magnetosphere-ionosphere interaction.     

A Multiscale Model for Substorms

Most substorm researchers assume substorms to be caused by a unique large-scale process. However, a critical evaluation of substorm observations indicates that a new paradigm is needed to understand the substorm phenomenon and the magnetospheric dynamics in general. It is proposed here that substorms involve a number of physical processes covering over a wide range of spatial and temporal scales. Potential candidates include the kinetic or shear ballooning instability, the Kelvin-Helmholtz instability, the cross-field current instability, the tearing instability, and magnetic reconnection. An observational constraint on the qualified process for substorm onset is that it must be associated with magnetic field lines of auroral arcs since substorm onsets start with brightening of a pre-existing auroral arc. Which particular process dominates in a given substorm depends on the present and past states of the magnetosphere as well as the external solar wind. The magnetosphere is almost perpetually driven by the solar wind to be near a critical point and in a metastable state. Magnetospheric disturbances occur sporadically in multiple localized sites. A substorm is realized when the combined effect of these localized disturbances become global in extent, much like the system-wide activity in a sandpile or avalanche model.