Coordination Of The Geosynchronous And Auroral Substorms

Studies using data from the ATS-5 geosynchronous spacecraft revealed a clear relationship between midnight region injection events near the spacecraft and auroral displays near the ATS magnetic conjugate point (Hones et al., 1971a; Mende et al., 1972; Eather et al., 1976; Mende and Shelley, 1976). A comparison of ATS-5 particle and magnetic field data with all-sky photographs taken at the conjugate point, Great Whale River, indicated that an injection at geostationary orbit generally corresponded to the brightening of the onset arc when the spacecraft was in the midnight sector (Akasofu et al., 1974). Results such as this whetted the collective appetite. How closely can the initial onset and injection be related to one another in time, do the onset and injection start on the same field field line, does the westward propagation of dipolarization correspond to the westward surge, can one relate the fine structures of the auroral expansion and the dipolarization? As time passed, increasingly precise answers have been given to these and similar questions, and auroral and geosynchronous substorm phenomenology has become more tightly integrated. In this chapter, we sample some of the evidence that supports this statement. The GEOS 2 spacecraft was stationed with its magnetic conjugate point near Kiruna, Sweden, so that the conjugate aurora could be studied with the extensive network of ground-based observatories in Scandinavia (Knott, 1975; Knott et al., 1979). In the first part of this chapter, we review some of the correlation studies carried out in the GEOS 2 project. In one particular series of four substorms, it was found that the dipolarization occurred at the same time as the aurora brightened and expanded poleward over the ground conjugate region (Section 14.2). In another case, a dispersionless injection at GEOS 2 corresponded to an intensification of the auroral X-ray band in Scandinavia (Section 14.2). Westward surges at the auroral conjugate point were associated with dipolarization at the spacecraft on a statistical basis (Section 14.3). Finally, the close relationship between both the auroral and geostationary substorm phenomena was extended to small spatio-temporal scales.

The Geosynchronous Substorm

The reconnection model of substorms deals with the large-scale changes in the structure of the magnetosphere and tail as convection intensifies following a sudden increase in the dayside reconnection rate. The model has difficulty making statements relevant to the small scales that characterize auroral onset. However, there has been considerable progress in assembling high-resolution observations of the events in space that now appear to be tightly coupled to the dramatic auroral events that first defined the term substorm. We will call this clear and consistent ensemble the geosynchronous model of substorms, since most of it was first conceived from observations made on geostationary spacecraft. We will also include in this ensemble the recent observations made using the quasigeostationary spacecraft, AMPTE/CCE, and so, by the geosynchronous substorm, we really mean the substorm as it appears on the earth's nightside typically between 6 and, say, 10 RE downtail. The earth’s magnetic field at geosynchronous orbit is about 100 nT, some three times larger than in the tail lobes. Study of quiet field intervals singles out the dependence of the geosynchronous field on solar wind dynamic pressure, since the modulation due to changes in the direction of the interplanetary field is presumably negligible during quiet conditions. The periodic variations in the quiet field depend on local time, season, and orientation of the earth’s dipole axis relative to spacecraft location (McPherron and Barfield, 1980; Rufenach et al., 1992). Superposed on the quiet field are perturbations up to about 50 nT due to several magnetospheric current systems, including the magnetopause current, the ring current, and the cross-tail current; the most striking are due to changes in the cross-tail current system. Observations from geosynchronous orbit were the first to indicate that the nightside magnetic field becomes more “tail-like” during substorm growth phase, and more dipolar during the expansion phase. This simple observation is the foundation on which today’s elaborate geosynchronous substorm model rests. The geosynchronous field becomes progressively more “tail-like” as the cross-tail current system intensifies and/or moves earthward during the substorm growth phase (McPherron et al., 1975; Coleman and McPherron, 1976; McPherron, 1979; Kauffmann, 1987).

The Auroral Substorm

Around the time the steady convection model was being developed, Akasofu (1964) was arranging ground-based magnetometer and all-sky camera observations of the complex time dependence of nightside auroral activity into the central phenomenological conception of tune-dependent magnetospheric physics—the auroral substorm. In this chapter, we assemble a description of a substorm from modern observations. We will see that observations of electric fields, auroral X rays, cosmic noise absorption, ionospheric density, and geomagnetic micropulsations have also been successfully ordered by the substorm paradigm. At the same time, it will become clear that each individual substorm has its own irreducible individuality, and that our summary description is really a list of effects that anyone thinking about substorms ought to consider. No real substorm will look exactly like the one described here. Spacecraft observations of auroral light, precipitation, currents, and fields from polar orbit have held out high promise for unified understanding of the development of the auroral substorm around the entire oval. Without truly global auroral observations, it would be difficult to establish decisive contact with observations of large-scale convection and the associated changes in magnetospheric configuration. Despite the high promise and the many other successes of spacecraft observations of the aurora, synthetic understanding of the time development of the auroral substorm at all local times, dayside and nightside, evening and dawn, has been slow in emerging, perhaps because a stringent combination of field of view, sensitivity, space and time resolution, and multispectral capability is required. One needs images of the whole oval with sufficient space resolution to identify important arc structures (50-100 km or better) in a temporal sequence that can articulate the evolution of activity on better than the 10-minute time scale on which polar cap convection develops. Only recently has it been possible to observe auroral activity at all local tunes around the auroral oval simultaneously and follow its time development from the beginning of the growth phase until well into the expansion phase. This amplification of the original paradigm is the subject of Sections 12.2 and 12.3.

Convection For Northward Interplanetary Field

Besides common sense, a number of results suggest that we can learn more about the slow “viscous” flow state by studying the magnetosphere during northward interplanetary field conditions. In particular, statistical studies have consistently identified a “residual” state of magnetospheric and ionospheric convection in northward field conditions. The integrated potential across the high latitudeionosphere does not drop below a certain resting value of about 20 kV even when the interplanetary field has been due north for several hours. There appears to be a similar residual component of geomagnetic activity that is independent of the direction of the interplanetary field (Scurry and Russell, 1991). Its correlation with the dynamic pressure of the solar wind strengthens our suspicion that it is related to viscosity. Will we be able to prove the convection in this residual state is driven by viscosity? Does the flow in northward field conditions resemble the underlying irregular flow state of the plasma sheet found at other times? Does the magnetosphere approach the teardrop configuration during prolonged intervals of northward interplanetary field? These are but a few of the questions that whet our interest in convection during northward field conditions. One does not arrive at the state of pure viscous convection immediately after the interplanetary field swings northward. Dungey (1963) was the first of many to argue that a northward magnetosheath field line will reconnect with an open tail lobe field line to create one that is connected to the ionosphere at one end and draped over the dayside magnetopause at the other. The sudden reconfiguration of stress will lead to sunward convection on the newly reconnected field lines. In the ionosphere, this superposes a “reverse” two-cell convection pattern in the central polar cap upon the two “direct” convection cells. If and when the draped reconnected field line finds a partner in the opposite tail lobe with which to reconnect, a newly closed field line will form. Dungey had imagined that the same magnetosheath field line would reconnect simultaneously with both tail lobes, in which case the rate at which open magnetic flux is closed depends upon the rate of tail-lobe reconnection.

The Bell-Like Magnetosphere

The fact that the geomagnetic field “pulsates” was known a century before the space age opened. The century of ground-based observations did lead to an effective empirical classification of the pulsations based on period, wave form, and geographical distribution (Section 3.1), but why the magnetic field of an astronomical body should oscillate on short time scales was a first-class scientific puzzle that could only by solved in the space age. Low-frequency hydromagnetic waves were first observed in the distant magnetosphere on Explorer 6 (Judge and Coleman, 1962). The task for space research was to relate the oscillations of plasma and fields in deep space to the ground observations using the refined theoretical languages of magnetohydrodynamics and plasma physics. There have been two critical issues. The first was to understand how plasma instabilities generate some of the observed pulsations. The second, the subject of this chapter, has been to understand how motions of the magnetopause induced by the variability of the solar wind are communicated to the interior of the magnetosphere. The breakthrough came when it was understood that the MHD fast mode can cross field lines and couple resonantly to localized standing Alfven waves. What is seen on the ground is due primarily to the resonant Alfven waves (Section 3.3). In Section 3.4, we provide basic theoretical information about the eigenmodes of the “MHD box” as a conceptual framework for the observations of oscillating fields and particles in the magnetospheric cavity. Space observations provided convincing evidence for the existence of standing Alfven waves shortly after the fast-wave coupling theory was proposed (Section 3.5). The next issue was which standing wave harmonics are excited (Section 3.6). Multiharmonic excitations now seem to be a semipermanent feature of the dayside magnetosphere, attesting to the constant activity at the magnetopause. There have been a few observations of the “global mode,” the low-frequency, radially standing compressional wave that may be responsible for discrete frequency resonant oscillations (Section 3.7).

The Teardrop Magnetosphere

In this chapter, we try to infer from magnetohydrodynamic reasoning and observation how the magnetosphere might look and behave if the magnetopause were inactive. Since there probably never has been an occasion when both viscosity and reconnection were absent, all we can do is array observations of phenomena that do not depend on either mechanism for their existence. As a result, we end up focusing on how the magnetosphere arrives at a balance of pressure with the solar wind. How it responds to changes in its confining pressure will be the topic of the next chapter. All discussions of the magnetosphere start with the magnetopause, and, indeed, the first models of the magnetosphere were calculations of the shape of the magnetopause. Without reconnection and without viscosity, the magnetopause would be given by the Chapman-Ferraro model on the dayside and close due to the reexpansion of the finite-temperature solar wind on the nightside (Section 2.2). This magnetosphere has a teardrop shape. After the dependence upon the interplanetary field via the reconnection process is taken into account, the average position and shape of the dayside magnetopause is in general accord with the Chapman-Ferraro model (Section 2.3). Because the magnetopause is always in motion, the early estimates of its thickness were uncertain until the first twospacecraft observations were made (Section 2.4). The magnetopause current layer proved to be several ion Larmor radii thick, significantly thicker than the electron inertial length. Once the average position of the magnetopause is specified, the position of the bow shock can be calculated using methods first employed for hypersonic flow around blunt bodies, which are easily extended to a weak-field MHD regime. The measured average positions of the bow shock and magnetopause agree once variations in solar wind dynamic pressure are taken into account (Section 2.5). While weak-field MHD does a good job with the bow shock, it fails in the subsolar magnetosheath, where a plasma depletion layer forms just upstream of the magnetopause (Section 2.6). Full MHD theory suggests that as many as three shocks could be standing in the flow enclosing the magnetosphere, a fast bow shock, an intermediate shock, and a slow shock.

Epilogue

The reader who has endured to the end, and even the one who skips to the end to see why he should read what is in the middle, deserves an answer to certain questions. Why, for example, did we employ a writing style so at odds with prevailing standards of scientific exposition? Of what use is it to read so much “anecdotal science”? Why, most of all, why did we place such a light emphasis upon theory, or for that matter, quantitative experimental results? The reader need only look as far as the comparison between the rates of transport due to viscosity and reconnection for an answer. These estimates were made with “back-of-theenvelope accuracy” in the 1960s, they are still made the same way on the eve of the twenty-first century, and we still do not know how the two systems of convection interact when they are time dependent. Magnetospheric physics, despite thirty-five years of the most intense, world wide activity, is still in the stage of paradigm clarification. It is, contrary to what one hears in many quarters, not a mature subject at all. There had been such optimism when the space age opened. A few rocket flights, a few spacecraft, and we thought we would know, really know, what the earth’s space environment was like. And in a sense, we learned, we really did learn. Standing out there in the solar wind was a bow shock and a magnetosphere of convecting plasma connecting to highly time-variable auroral displays. These first broad brush strokes of the picture were created by a confident, mid-twentiethcentury command technology, and it seemed that only a few more strokes would fill the canvas. Yet as time passed, the pattern seemed no clearer than it had at the beginning. It turned out that we were painting our picture as a pointillist would, each spacecraft, each ground experiment a single dot. Seurat, as he painted, had a clear idea of the picture he wanted to create, but we had to wait for our picture to emerge, one dot a time, dots of one color at a time.

The Nightside Auroral Oval

The basic structure of the auroral oval was pieced together from relatively local magnetometer measurements and all-sky photographs taken on the ground. The all-sky cameras picked out relatively intense features whose intensities exceeded roughly one kilorayleigh. Their fields of view had a 500-1000 km radius at auroral altitudes, and so extended over 5-10 degrees of latitude and about 90 minutes of local time. Had the aurora been stationary and time-independent, this would have been enough, and it was enough to spot the existence of substorms. It was not enough to solve the substorm problem. As the instruments to study auroral phenomena grew in sophistication and comprehensiveness, so also did our understanding of the concept of the auroral oval. This chapter is dedicated to communicating some of this modern understanding as a prelude to the discussion of substorms in the next chapter. Ground instruments can follow the time development of events within their fields of view but have difficulty separating changes in space and time on scales longer than an hour of universal time or local time, because the observing station rotates with the earth to a local time sector where the aurora may differ. This difficulty can be offset to some extent by airplane flights that remain at a constant local time. However, the real breakthrough came with auroral imaging from space. In the 1970s, optical wavelength imaging from low-altitude polar orbit provided snapshots of the aurora over several thousand kilometer scale portions of the oval on each polar pass of the spacecraft (Shepherd et al., 1973; Anger et al., 1973; Lui and Anger, 1973; Pike and Whalen, 1974; Snyder and Akasofu, 1974). And the spacecraft could detect the precipitating particles responsible for the auroral light emitted from the magnetic footprint of the field lines along its path. The results from the first generation of auroral imaging experiments have been summarized in excellent reviews (Akasofu, 1974, 1976; Hultquist, 1974; Burch, 1979). Ultraviolet imaging allows one to see the dayside aurora.

Bimodal Plasma Sheet Flow

How does the plasma sheet respond to the complex pattern of waves coming over the poles from bursty magnetopause reconnection events, or to the vortices and other irregular perturbations coming around the flanks of the magnetosphere in the low-latitude boundary layer? It is probably too much to expect that the complex input from the dayside will sort itself out into a steady flow on the nightside, but there has been a seductive hope that, on a statistical basis, the observations of the plasma sheet could be rationalized using steady convection thinking. This hope depends on the belief that the average magnetic field configuration in the plasma sheet actually is compatible with steady convection. The first doubts on this score were raised by Erickson and Wolf (1980), and were subsequently elaborated by Tsyganenko (1982), Birn and Schindler (1983), and Liu and Hill (1985); the“plasma sheet pressure paradox” they posed is the subject of Section 9.2. Theoretical arguments are one thing, measurements are another; the truly important issue is whether the real plasma sheet manifests steady flow. Several groups have searched large data sets to see whether the statistically averaged flow in the central plasma sheet resembles the flow predicted by the steady convection model. This effort has led to a growing but still incomplete comprehension of the statistical properties of plasma sheet transport. Results obtained using ensembles of data acquired by ISEE 1 and AMPTE/IRM will be reviewed in Section 9.3. The unusual distribution of bulk flow velocities suggests that the plasma sheet flow is bimodal, alternating between a predominant irregular low-speed state and an infrequently occurring state of high-speed earthward flow. In search of steady plasma sheet flow, one could also look into substormfree periods of stable solar wind properties. One of the best such studies, in which great care was taken to find periods of exceptionally stable solar wind and geomagnetic conditions, is reviewed in Section 9.4. Even this study found highly irregular and bursty flow.

On The Relation Between Convection And Substorms

Why after 30 years of research have we not settled the relationship between substorms and convection? Why after 30 years are there still substantially different, competing models of substorm onset? Why has the progression from phenomenological to fully quantitative understanding not occurred? Answers to such questions will probably only become clear in retrospect, after the transition to quantitative understanding has taken place; in the meantime, we can only express our individual views. In our view, there have been three central difficulties. Contradictory pictures of plasma sheet transport have been able to flourish side-by-side because this fundamentally unsteady and highly spatially structured system was and is undersampled. As a result, people can still argue about when and where tail reconnection occurs in the substorm sequence, and other issues equally fundamental. Moreover, it is still not possible to connect the substorm onset in the auroral ionosphere to an event in space. For one thing, there is the timing problem: Waves communicate events’ existence across the magnetosphere and to the auroral ionosphere within a minute or two. We are only beginning to study the magnetoionosphere system globally with the required time resolution, and until we do so, there will be a “chicken and egg” problem. Finally, two key measurements are missing. No plasma sheet signatures of auroral arcs have been identified, so we do not know when a spacecraft is connected to a potential auroral onset region. We do not yet have an accepted ionospheric diagnostic of plasma sheet reconnection and the plasmoid formation; there is no auroral data display that illustrates at a glance the relationship of plasma sheet events to the onset and expansion of the substorm. Despite all this, there is real cause for optimism. The sheer magnitude of the observational effort and the volume and diversity of the results produced over the years have finally enabled us to perceive how complex the behavior of the magnetosphere really is. As our perceptions have evolved, we have designed multi-instrument, multi-spacecraft studies of ever-increasing comprehensiveness, articulation, and resolution, which further clarified our perceptions. All this effort is beginning to pay off.