The Earth's magnetosphere is the region surrounding the planet, above the outer atmosphere and ionosphere (q.v.), which contains and is controlled by the Earth's magnetic field. It extends from an altitude of ∼500 km above the Earth's surface to an outer boundary which is formed by the interaction of the planetary magnetic field with the solar wind, the plasma (charged particle) gas that streams continuously outward from the Sun. On the dayside, the compressive effect of the solar wind confines the field to a region extending ∼10 Earth radii from the Earth, while on the nightside the field is stretched into a long comet‐like tail which extends typically in excess of ∼1000 Earth radii. The Earth's radius, R E ∼6400 km, is thus a convenient measure of magnetospheric spatial scales. Electric currents flowing in these regions, whose magnetic effects can be observed at the Earth's surface, include those flowing at the solar wind‐magnetosphere boundary, those produced by the drift of energetic charged particles inside the magnetosphere, and those associated with the coupling between the magnetosphere and the ionosphere Ohtani et al. (2000). Magnetic perturbations observed at the Earth's surface due to these three sources peak typically at a few tens, a few hundreds, and a few thousands of nanoteslas, respectively, and vary on a range of timescales from minutes to hours and days. In this article we first concentrate on the overall structure and basic physics of the coupled solar wind‐magnetosphere‐ionosphere system, and then consider the dynamic processes which occur, leading to strong temporal variations.
Structure of the magnetosphere
In Figure M158 we show a cross‐section through the Earth's magnetosphere in the noon‐midnight meridian plane, where the arrowed solid lines indicate magnetic field lines. The small symbols indicate the main plasma populations, while the arrows show their principal flows. The small dashes show the solar wind plasma, together with the magnetospheric “boundary layer” plasma populations derived directly from it. The direction of the Sun is toward the left, so that the solar wind flows from left to right as indicated. The solar wind derives from hydrogen gas in the solar atmosphere, which is heated sufficiently strongly in the solar corona, to temperatures in excess of a million degrees, that the atoms are fully broken up by collisions to form a plasma of electrons and protons (plus a few percent α‐particles from helium), which then streams continuously out into the solar system at speeds ∼500 km s−1. At the orbit of the Earth, the proton and electron number densities are ∼10 cm–3, sufficiently low that the particles in the gas are collision‐free, meaning that the mean free path for particle collisions is comparable with or larger than the size of the system. A second source of plasma in the magnetosphere derives from the Earth's upper atmosphere, which is relatively cool (∼1000 K) but is partially ionized by solar far‐UV and X‐rays at altitudes above ∼100 km, forming the ionosphere. These charged particles, consisting of protons and heavy ions, principally singly‐charged oxygen, together with corresponding electrons, can also stream out of the topside ionosphere into the magnetosphere, as indicated by the small crosses in the figure. In the ionosphere, collisions between ions and neutral atoms are significant at altitudes up to ∼500 km. Above this altitude, however, the ionospheric ions also become collision‐free due to falling densities, ∼500 km thus being chosen (rather arbitrarily) above as the “base” of the magnetosphere, where it interfaces with the ionosphere. Like the solar wind, therefore, the plasma of charged particles in the Earth's magnetosphere is also collision‐free. Although the Earth's neutral atmosphere is detectable throughout the magnetosphere to distances of at least ∼10 R E (forming a hydrogen “geocorona” originating from the photodissociation of water vapor lower in the atmosphere), the behavior of neutral and charged particle populations within it are largely decoupled due to the lack of collisions, apart from the occasional charge‐exchange reaction between them.
When collisions can be ignored, the motion of ions and electrons in the plasma is governed by the forces of the large‐scale fields prevailing, and for both solar wind and magnetosphere the most important forces by far are electromagnetic. Solar and planetary gravity is only important in the regions close to the bodies concerned, i.e., in the solar corona near the Sun, and in the ionosphere near the Earth. The motion of charged particles in a large‐scale electromagnetic field consists of a number of components, which take place on generally widely separated timescales. The first is a rapid gyration of the particles around the magnetic field lines in nearly circular orbits whose radius is generally tiny compared with the scale size of the systems. The second is a uniform motion along the magnetic field lines, which is subject to the magnetic mirror force which repels particles from regions of increasing field strength along the field lines (i.e., is directed away from regions where field lines converge). The mirror force is produced by the action of the field strength gradient on the magnetic dipole formed by the gyrating particle. In the quasidipolar field of a planetary magnetosphere, for example, the field strength along a given field line is weakest at the equator and increases greatly toward the planet (see Figure M158). This configuration can therefore trap charged particles almost indefinitely, the mirror force resulting in a bounce motion between mirror points in the northern and southern hemisphere. Only those particles that are directed almost along the field lines at the equator, which thus have small magnetic moments, can reach down to the upper layers of the atmosphere. The third motion consists of large‐scale plasma flows transverse to the magnetic field lines, such as that shown for the solar wind in Figure M158. Such flows are associated with an electric field E directed transverse to the magnetic field B, related to the plasma velocity V by the vector relation . (The effect of additional cross‐field drifts associated with magnetic field inhomogeneity will be discussed further below.) Inspection of the flows indicated in Figure M158 shows that the electric field is directed everywhere out of the plane of the diagram, though it is not everywhere of equal strength. In such “crossed” electric and magnetic fields, charged particles drift perpendicular to the magnetic field with velocity independent of their charge or mass, this expression being exactly equivalent to the vector relation for E given above.
This “ drift” has a very special property, first discovered by Alfvén in 1942 (see Alfvén's theorem ), that particles whose centers of gyration lie initially on the same field line as each other, remain on the same field line as each other for all time, thus showing why plasma structures are strongly organized by magnetic field lines, as indicated by various features in Figure M158. We may picture the collective behavior as one in which the magnetic field and the plasma are “frozen” together. However, we can think either of the plasma as moving and carrying the field lines, or equivalently of the field lines as moving and carrying the plasma. Which of these pictures is the most appropriate for a given system depends upon the relative energies in the plasma and magnetic field. Thus, for example, since the plasma energy exceeds the magnetic energy in the solar wind, the solar magnetic field is carried outward “frozen” into the expanding plasma flow, forming a large‐scale interplanetary magnetic field (IMF) that pervades the entire solar system (see Figure M158). The strength of the IMF at the Earth's orbit is typically ∼5–10 nT, directed, on the average, in the ecliptic plane at an angle of ∼45° to the Earth‐Sun line. The latter tilt is due to the rotation of the Sun, which winds the IMF into a spiral form as the field lines are carried out into the solar system by the plasma flow. On the other hand, in the quasidipolar field regions of the magnetosphere and ionosphere, where the magnetic energy dominates, we instead think of the field lines as moving, transporting the plasma.
We may apply this “frozen‐in” concept to the interaction between the solar wind and the planetary magnetic field. Since the solar wind and IMF are frozen together, as well as the planetary field and planetary plasma (e.g., from the ionosphere), then when these two media interact they will not mix. Instead, the solar wind confines the planetary field to a cavity surrounding the planet, around which it flows, as first deduced by Chapman and Ferraro in 1931 (see Chapman, Sydney ). This magnetic cavity is the planet's magnetosphere, whose outer boundary, shown by the dashed line in Figure M158, is called the “magnetopause.” A “bow shock” forms ahead of the cavity, also shown by a dashed line in Figure M158, due to the fact that the magnetosphere represents a blunt obstacle in the supersonic solar wind flow. Across the shock the solar wind is slowed, compressed, and heated, forming the turbulent “magnetosheath” layer located between the shock and the magnetopause boundary.
The size of the magnetospheric cavity is set by the condition of pressure balance at the boundary. A simple estimate of the distance of the equatorial boundary at noon (the minimum distance in the direction facing the Sun) can be made by equating the ram pressure of the solar wind on one side of the boundary, with the magnetic pressure () of the compressed planetary field on the other. With typical solar wind values the radial distance of the boundary is estimated to be ∼10 R E on this basis, as observed, a position which may vary by factors of up to two in either direction under extreme solar wind conditions. The strength of the compressed planetary field just inside the boundary is typically ∼60 nT, representing a planetary dipole field of ∼30 nT enhanced by a factor of ∼2 by the electric current flowing in the magnetopause boundary. This latter current is termed the “Chapman‐Ferraro” current, and inspection of Figure M158 with Ampere's law in mind shows that it flows out of the plane of the diagram in the equatorial region on the dayside, closing over the magnetopause into the plane of the diagram over the polar regions and on the nightside. These rings of magnetopause current produce a perturbation magnetic field in the near‐Earth magnetosphere whose strength is typically a few tens of nanoteslas directed northward (i.e., upward in Figure M158), hence enhancing the horizontal field at low and middle latitudes at the Earth's surface.
Frozen‐in behavior of the field and plasma associated with the drift is not, however, a universally valid description. In a collision‐free medium it is broken in particular by the presence of additional particle drifts, the most significant of which cause ions and electrons to drift in opposite directions across the field (thus producing a current) due to the presence of gradients in the strength and direction of the magnetic field. These “field inhomogeneity” drifts are proportional to the field gradients, and also to the particle energy, thus being more important for particles of higher energy. When the field gradients are weak, these additional drifts are important only for particles in the high‐energy tail of the energy distribution, and frozen‐in motion then represents a useful organizing concept for the bulk of the plasma population. This limit applies essentially throughout the solar wind, and through most of the Earth's magnetosphere. However, when the field gradients are very strong, such that the motion of the bulk of the plasma particles is affected, then the frozen‐in picture breaks down. One such place where this happens is the magnetopause boundary, where the field strength and direction in general switch rapidly from magnetosheath to magnetosphere values across the magnetopause current sheet. The simplest theoretical description of the consequence of frozen‐flux breakdown is that the magnetic field diffuses through the plasma, locally, in the region of the strong gradient. As first pointed out by Dungey in 1961, this then allows magnetic field lines to become joined across the boundary, producing “open” magnetic field lines which pass from the solar wind at one end, through the magnetopause, to the Earth's polar regions at the other. This process is called “magnetic reconnection.” Two newly reconnected open field lines are shown in Figure M158 passing through the dayside magnetopause shortly after reconnection has taken place near the equator. Sharply bent magnetic field lines exert a tension force on the plasma like the force of rubber bands (the force per unit volume being , where j is the current density in the plasma), in this case accelerating the boundary plasma poleward away from the equator, such that the field lines also contract poleward, releasing energy to the plasma and allowing further reconnection to proceed at the equator. Subsequently, the open field lines are carried downstream frozen into the magnetosheath flow, and are stretched out into a long cylindrical comet‐like tail. This tail consists of two lobes, D‐shaped in cross‐section, one connected to the northern polar region at Earth, the other to the southern, as indicated in Figure M158. Observations show that the tail lobe field lines remain open typically for a few hours, such that with a downstream speed of ∼500 km s–1, the tail is typically ∼1000 R E long.
The open field lines form magnetic pathways along which the magnetosheath plasma may enter the magnetosphere. Such plasma thus flows along newly opened field lines to form a boundary layer adjacent to the dayside magnetopause, and the “cusp” population as it then moves down toward the Earth (see the magnetospheric dashed regions in Figure M158). The majority of the particles, however, are repelled by the magnetic mirror force as the field strength increases near the Earth, and hence move back out again toward the outer magnetosphere. Due to the antisunward motion of the open field lines, however, the cusp plasma flows back out into the lobes of the tail, in the region adjacent to the magnetopause, forming the “plasma mantle” population. As the open field lines are carried down the tail, so the field lines and mantle plasma sink in toward the center plane of the tail, followed by further entry of antisunward flowing magnetosheath plasma at the tail magnetopause, such that the mantle grows wider and with increasing density at larger distances. Plasma from the Earth's polar ionosphere also flows into the lobes (cross symbols in Figure M158), but because of its low velocity along the field lines (∼10 km s−1), it does not reach far down the tail on the few‐hour timescale that the lobe field lines remain open. Overall, the plasma density in the inner part of the tail lobes is very low, ∼0.01–0.1 cm–3, and with temperatures typically of order a few tens to a few hundred electronvolts, most of the system energy resides in the lobe magnetic field. (Note that while temperatures in the solar and terrestrial ionized atmospheres are generally quoted in Kelvin, as above, the temperatures of magnetospheric plasmas are usually indicated by the mean or typical energy of the particles W, in eV. To convert between them, we note that for a near‐Maxwellian velocity distribution . Thus, for example, typical hot magnetospheric plasma of ∼1 keV mean energy corresponds to a temperature of ∼107 K.)
The residency of the oppositely directed open field lines in the two tail lobes is terminated when they sink into the center plane of the tail and reconnect in the equatorial current sheet that separates them, as shown by the X‐shaped field configuration on the right side of Figure M158. On the tailward side of the tail reconnection site the lobe field lines become disconnected from the Earth, and the (rubber band) tension force accelerates the tail plasma rapidly away from the Earth, where it eventually rejoins the solar wind. On the Earthward side, the process forms new “closed” field lines, connected to the Earth at both ends, which similarly contract rapidly earthward, compressing and heating the lobe plasma as they do so. This hot plasma, containing both solar wind and ionospheric contributions, is termed the “plasma sheet” population, and is shown by the dotted region in Figure M158. In the near‐Earth tail it is characterized by densities ∼0.1–1 cm–3, and electron and ion temperatures of ∼0.5 and ∼5 keV, respectively, but both density and temperatures increase as the plasma flows earthward into the quasidipolar inner magnetosphere and is further compressed and heated. Throughout these regions the hot plasma sheet population carries a current, due to the field‐inhomogeneity effects mentioned above, which is directed westward, i.e., out of the plane of Figure M158 on the nightside of the Earth, and into the plane of the diagram on the dayside. On the nightside within the tail, the current paths reach the tail boundary, and close over the magnetopause to form two D‐shaped solenoids lying back‐to‐back, as required by the lobe magnetic field structure. Nearer the Earth, the current paths may close wholly round the Earth to form a plasma “ ring current ” (q.v.). Overall, these currents produce a perturbation field at the Earth, which is directed southward (i.e., downward in Figure M158), and is usually a few tens of nanoteslas in amplitude. However, as we discuss below, these currents become greatly enhanced during magnetic storms (q.v.), producing perturbation fields up to an order of magnitude larger.
The closed field lines and hot plasma produced in the tail subsequently flow sunward around the Earth in the quasidipolar magnetosphere, eventually reaching the dayside magnetopause where they become open again, allowing the hot magnetospheric plasma to escape into the magnetosheath, as indicated by the dotted region outside the magnetopause. Overall, therefore, reconnection at the magnetopause and in the tail results in the formation of a large‐scale cyclical flow of field lines and plasma through the Earth's magnetosphere, first discussed by Dungey in 1961. The overall cycle time is ∼12 h, of which the field lines spend ∼4 h open in the tail lobes, and ∼8 h closed flowing sunward through the central magnetosphere. The solar wind is not, however, the only source of momentum that drives magnetospheric flow. A second source resides in the rotation of the Earth, and the coupled rotation of the Earth's upper atmosphere. Collisions between ions and neutral atmospheric particles in the ionosphere produce a frictional force on the feet of the magnetospheric field lines which tends to spin them up toward rotation with the planet, a flow called “corotation.” In general, the overall magnetospheric flow results from a summation of the effects of the planetary and solar wind driving forces. The corotation effect tends to produce a flow in the equatorial plane increasing linearly with the distance from the planet (i.e., a flow with the planetary angular velocity), while in the simplest picture of a dipole magnetic field and uniform cross‐system electric field, the equatorial Dungey‐cycle flow increases as the cube of the distance (i.e., as the inverse of the field strength). Reflection on these relative dependencies then shows that corotation will dominate close‐in, and Dungey‐cycle convection at larger distances. For Earth, the boundary between these flow regimes is located typically at ∼5 R E in the equatorial plane, about half‐way to the dayside magnetopause. The hot plasma flowing in from the tail is therefore excluded from this inner core of corotating field lines, as indicated in Figure M158, which is instead filled to relatively high densities (∼100–1000 cm–3) by cold (∼1–10 eV) hydrogen and oxygen plasma from the ionosphere. This population is termed the plasmasphere, and is indicated by the near‐Earth region of dense crosses in Figure M158.
Magnetosphere‐ionosphere coupling
The magnetospheric flows discussed above produce corresponding motions of the field and plasma in the polar ionosphere, resulting in large‐scale current systems flowing between these regions. Figure M159 shows a view looking down on the North Polar Region, with noon (the direction toward the Sun) at the top of the diagram, dusk to the left, dawn to the right, and midnight at the bottom. The central circular dashed line shows the outer boundary of the “open” field region (located typically at magnetic latitudes of ∼75° at noon and ∼70° at midnight), while the arrowed solid lines show plasma streamlines. The open field lines within the dashed line boundary, mapping to the northern tail lobe, flow antisunward from noon to midnight at speeds typically of a few 100 m s–1, while the closed field lines at lower latitudes return sunward. This twin‐vortex flow is the signature of the Dungey‐cycle in the ionosphere. The electric field associated with the flow is directed from dawn to dusk across the open field line region as indicated, reversing in sense to poleward at dusk and equatorward at dawn. The small symbols also show the magnetospheric plasma populations into which the ionospheric field lines map, the symbol code corresponding to that in Figure M158. Ions and electrons from the cusp (dashes) and plasma sheet (dots) precipitate into the upper atmosphere, excite the atoms (typically at altitudes ∼100–300 km), and cause them to emit photons, giving rise to one component of auroral light emissions (see Auroral oval ) Paschmann et al. (2002) and Sandholt et al. (2001). They also ionize the neutral atoms and increase the plasma density of the ionosphere, a process that is especially important at night when the solar production mechanism is inoperative.
The Dungey‐cycle flow drives two components of electric current in the lower ionosphere, both resulting from the effects of collisions between ionospheric ions and atmospheric neutral particles. At altitudes below ∼120 km, these collisions become sufficiently frequent that the ions do not move with the field lines at all, but are essentially tied to the generally smaller motions of the upper atmosphere (i.e., the winds in the neutral thermosphere). However, the electron motion remains almost unaffected by collisions, frozen to the field line motion throughout the ionosphere, to altitudes below ∼100 km. Consequently, in the layer between ∼100 and 120 km, ionospheric electrons flow around the streamlines shown in Figure M159, carrying a current in the opposite direction, while the ions in the layer remain almost motionless, tied to the neutral atmosphere. This current, directed transverse to both the electric and magnetic fields is thus termed a Hall current, and with a height‐integrated ionospheric conductivity of ∼10 mho and typical ionospheric flows of several hundred meters per second, it produces equal and opposite magnetic perturbations on either side of the ionospheric Hall layer which are typically ∼100–200 nT in strength. At the Earth's surface in the northern hemisphere these perturbation fields are roughly in the direction of the overhead ionospheric electric field (opposite in the southern hemisphere). These currents are nondissipative (i.e., ), and in principle can close wholly within the ionosphere (for uniform conductivity), flowing round the plasma flow streamlines.
The second current component flows at a slightly higher altitude of ∼120–150 km, where the ion motion is affected by collisions with neutrals, but is not yet wholly dominated by them. In this layer collisions provide the ions with some mobility in the direction of the electric field, while continuing to move approximately with the electrons along the plasma streamlines. An ion current thus flows in the direction of E, termed the Pedersen current, and its magnitude is such that the force of the current, directed along the plasma streamlines, just balances the frictional drag on the ionospheric plasma due to ion‐neutral collisions in the ionosphere. A key fact about the Pedersen current is that it cannot, in principle, close within the ionosphere itself, but is instead part of a large‐scale current system which flows between the magnetosphere and ionosphere, which imposes the flow of the former onto the latter in the presence of ionospheric ion‐neutral drag. It can be seen in Figure M159, for example, that the Pedersen currents, directed along E, flow away from the open‐closed field line boundary on both sides at dawn, while flowing toward the boundary from both sides at dusk. Current continuity then requires the presence of field‐aligned currents (FACs), which flow down the field lines from the magnetosphere to the ionosphere at the dawn boundary, and up the field lines from the ionosphere to the magnetosphere at the dusk boundary. These are called the “region 1” FACs, whose direction is indicated in Figure M159 by the circled dots and crosses for upward and downward currents, respectively. These FACs close across the field lines at large distances in the plasma which is the source of the momentum producing the flow, i.e., in the magnetosheath plasma at the tail magnetopause. The magnetic perturbations produced by the overall current circuit bends the magnetic field lines in such a way that the force slows the magnetosheath plasma, while transferring the momentum to the ionosphere to maintain the flow against collisional frictional drag. Overall, this current system is solenoidal in nature, such that the main magnetic effects are contained within the effective current solenoid between the ionosphere and magnetopause. Thus although the ionospheric height‐integrated Pedersen conductivity is generally comparable with the height‐integrated Hall conductivity, such that the two height‐integrated current components are of similar intensity in the ionosphere, the combined contribution of the ionospheric Pedersen current and the associated FACs approximately cancels underneath the ionosphere, and produces a much weaker magnetic perturbation on the ground than that of the Hall current. Above the ionospheric Pedersen layer, however, the perturbation field produced by the Pedersen‐FAC system is directed approximately opposite to the plasma streamlines in the northern hemisphere, and along the streamlines in the southern hemisphere. Typical amplitudes in the region just above the Pedersen layer are again a few 100 nT. The total current flowing in the “region 1” FAC system is typically ∼2–3 MA.
We also note from Figure M159 that Pedersen current continuity at the equatorward boundary of the Dungey‐cycle flow (typically located at magnetic latitudes of ∼70° at noon and ∼65° at midnight), also requires the presence of upward FACs at dawn and downward FACs at dusk. These are called the “region 2” FAC system, in which the upward currents at dawn close in the downward currents at dusk via field‐perpendicular currents flowing westward in the inner edge of the equatorial plasma sheet population in the nightside magnetosphere (and then via ionospheric Pedersen currents, the “region 1” FAC, and the tail magnetopause). The total current flowing in the “region 2” FACs is a little less than that in the “region 1” FACs, due to the fact that the latter currents are fed by Pedersen currents from both open and closed field line regions. We also note that regions of upward‐directed FAC are often associated with bright “discrete” auroras (e.g., structured curtain‐like forms), thus forming another component of the auroral emissions. This is because upward currents are carried primarily by warm plasma sheet electrons, which flow down the field lines into the ionosphere. In order to produce a downward electron flux which is sufficient to carry the upward current, the electrons may be accelerated downward by an electric field directed upward along the field lines (this being another scenario in which the frozen‐in picture breaks down). Typically the required field‐aligned voltages are a few kilovolts, such that the precipitating electrons carry a sufficient energy flux to produce a bright aurora.
Magnetosphere dynamics
The above discussion has centered on the plasma physics principles which govern the structure and properties of the coupled magnetosphere‐ionosphere system, focusing particularly on the nature of the current systems that flow within it. For clarity of exposition we have viewed the system as being steadily driven, principally by the solar wind, but also by planetary rotation. However, it is now important to emphasize that the Earth's plasma environment is almost never in such a steady state, but generally varies strongly with time Cowley et al. (2003). There are two reasons for this. The first is that the interplanetary medium which impinges on the Earth's field itself varies strongly with time, on timescales from a few minutes to the ∼27‐day rotation period of the Sun. There are also variations over the 11‐year period of the solar cycle. The second factor is that even when the solar wind is relatively steady, the processes in the key interaction regions at the magnetopause and in the tail show a propensity for pulsed behavior. We will now outline these behaviors, beginning here with those which are driven essentially directly by variations in the interplanetary medium.
Although the density and velocity of the solar wind determine the size of the magnetospheric cavity and the speed with which open flux tubes are transported to the tail, by far the most important upstream parameter which determines the overall nature of the magnetospheric interaction with the solar wind, and hence the dynamics of the Earth's plasma‐field environment, is the direction and strength of the IMF. This determines how much flux is reconnected at the magnetopause per unit time, where on the magnetopause it is reconnected, and how the newly formed open tubes move into the tail. Although the IMF on average lies in the ecliptic plane at the ∼45° spiral angle mentioned above, pointing either toward or away from the Sun along this direction, variable north‐south fields are also produced by a variety of effects, such as waves and other disturbances propagating in the wind outward from the Sun. Open flux production is largest, and the Dungey‐cycle flow and related currents strongest, when the IMF is directed southward opposite to the Earth's equatorial field (i.e., points downward in Figure M158, as drawn in the figure), and is weak when the IMF is directed northward. From Faraday's law the overall magnetic flux transfer in the Dungey‐cycle, in Wb s–1, is equal to the voltage of the cross‐system electric field associated with the flow, in volts. This cannot be measured directly in the magnetosphere, but can routinely be measured at ionospheric heights using a network of ground‐based radars, which determine the flow in the polar ionosphere. These observations show that when the IMF, of strength ∼5–10 nT, points south, the voltage associated with the flow is ∼100 kV, i.e., the flux throughput in the Dungey‐cycle is ∼100 k Wb s–1. This compares with voltages in the solar wind across the magnetospheric diameter of about five times this value. Reconnection is thus ∼20% efficient during such southward field intervals, in the sense that ∼20% of the interplanetary magnetic flux which is directed toward the magnetosphere in the solar wind becomes reconnected with the terrestrial field. At the same time, of course, ∼80% of the magnetic flux is deflected around the sides of the magnetosphere, such that most of the impinging solar wind flow is so deflected, as envisaged in Chapman and Ferraro's early considerations. Nevertheless, it is the breakdown of perfect deflection at the ∼20% level which is critical to magnetospheric dynamics. As the IMF direction rotates away from southward, however, reconnection becomes less efficient. For example, when the IMF lies in its average direction in the ecliptic plane, the voltage values fall to ∼50 kV, while when it rotates to northward they fall to “background” values of ∼20 kV or less. The Dungey‐cycle flow in the magnetosphere‐ionosphere system is therefore strongly modulated by the sense and magnitude of the north‐south component of the IMF, and is found to respond rapidly to changes in this component on timescales down to a few minutes.
The form of the flow is also found to change not only in response to the north‐south field component of the IMF, usually referred to as the z component (with B z positive northward), but also to the component which points into or out of the plane of the diagram in Figure M158, referred to as the y component (with B y positive out of the plane of the diagram). When dayside reconnection takes place in the presence of IMF B y , the force on newly opened flux tubes pulls the open tubes “sideways” (i.e., east‐west) as well as poleward, oppositely in the two hemispheres. This is illustrated schematically in Figure M160, which shows a view of the dayside magnetosphere looking from the direction of the Sun shortly after reconnection has taken place with an IMF having a positive B y component (pointing from left to right in the figure). The (rubber band) tension force pulls the newly opened field lines westward (to the left) in the dayside cusp in the northern hemisphere, and simultaneously eastward (to the right) in the dayside cusp in the southern hemisphere. The open field lines are correspondingly carried asymmetrically into the tail lobes, preferentially into the dawn side of the northern lobe, and into the dusk side of the southern lobe, an effect that tends to twist the tail structure on the nightside. The corresponding ionospheric flows in the northern hemisphere are shown in the upper row of diagrams in Figure M161, for near‐zero or negative IMF B z and various IMF B y as indicated. Again noon is at the top of each diagram and dusk to the left. In this case the region of open field lines is relatively expanded and the twin cell Dungey‐cycle flows well‐developed due to the negative IMF B z , while the IMF B y component produces dawn‐dusk asymmetries in the flow with newly opened cusp field lines flowing west for B y positive (as indicated in Figure M160), and east for B y negative. The simultaneous east‐west flows are opposite in the southern hemisphere. These variations of the flow in the cusp region (and beyond) produce corresponding variations in the Hall and Pedersen‐FAC current system that can be sensed both on the ground and above the ionosphere according to the general principles outlined above. The perturbations observed at magnetic observatories lying under the cusp (at ∼75° magnetic latitude near noon) can be used in particular to reconstruct the large‐scale structure of the IMF over long intervals from historic magnetic records.
Observations indicate that reconnection at the dayside magnetopause tends to occur preferentially in regions where the magnetospheric and magnetosheath field adjacent to the magnetopause are antiparallel, such that the “magnetic gradients” across the boundary are largest. When the IMF turns from southward to northward, these regions migrate from low to high latitudes, such that reconnection can then take place between northward IMF and open field lines in the tail lobes that were produced by earlier intervals of southward field. Such “lobe reconnection” does not change the amount of open flux in the tail, but causes it to circulate within the region of open field lines, as the newly reconnected field lines flow around the dayside magnetopause and back into the lobe. The patterns of lobe circulation produced by this effect are shown in the lower row of diagrams in Figure M161, which also indicate that the amount of open flux present is usually smaller under these conditions. Again, these patterns of ionospheric flow are associated with corresponding patterns of magnetic perturbations above and below the ionosphere, produced by the corresponding Hall and Pedersen‐FAC currents that are driven.
Magnetospheric substorms
The above sections have discussed the flows and currents that are driven in the coupled magnetosphere‐ionosphere system by the Dungey‐cycle, viewed as a quasisteady process that is strongly modulated by variations in the strength and direction of the IMF. Observations show, however, that even when interplanetary conditions are relatively steady, the driving processes in the key interaction regions at the magnetopause and in the geomagnetic tail are not. Reconnection on the dayside, in particular, often occurs as waves that propagate east‐west over the magnetopause from the dayside toward the tail, which recur on timescales of ∼5–10 min. These “flux transfer events” (FTEs) give rise to pulsed magnetic and plasma signatures at the magnetopause, and pulsed flow, current, plasma injection, and auroras in the cusp ionosphere on the dayside. Due to the typically overlapping nature of the effect of these pulses, however, the overall flow and current modulations are generally modest.
The pulsed behavior in the tail takes place on longer timescales of around 1–2 h, however, and produces major perturbations of the flow and current in the nightside magnetosphere‐ionosphere system, termed magnetospheric substorms (see Storms and substorms ). Substorms are initiated by intervals of southward‐directed IMF, typically southward fields of a few nanoteslas lasting for several tens of minutes. Such intervals, which happen rather frequently (often several per day), are sufficient to produce enhanced open flux production at the dayside magnetopause which in turn enhances the radius and field strength of the lobes of the tail. The current carried by the plasma sheet at the center of the tail is also correspondingly intensified, while the thickness of this current layer is also found to decrease, within the near‐Earth tail, to only a few thousand kilometers. After ∼30–40 min of such tail development, termed the “growth phase” of the substorm, a disruption is observed to occur within the plasma sheet. The origin and development of the disruption remain controversial, but it involves the formation within a minute or two of a new reconnection region within the plasma sheet in the central tail at down‐tail distances of 20–30 R E. Reconnection at this site first pinches off the tailward portion of the preexisting plasma sheet, forming a closed‐loop plasmoid field structure, illustrated schematically on the right‐hand side of Figure M158, which is expelled tailward into the solar wind at speeds in excess ∼500 km s–1. Subsequently, open field lines in the tail lobes are also reconnected, reducing the amount of open flux in the system. Earthward of the new reconnection region, new closed field lines collapse toward the Earth, “dipolarizing” the extended tail field lines, and heating and compressing the plasma sheet plasma. Precipitation from this hot plasma into the nightside ionosphere produces an expanding patch of bright active auroras, called the “substorm auroral bulge,” and also considerably enhances the density of the nightside ionosphere. Consequently the height‐integrated Hall and Pedersen conductivities in this “bulge” are also strongly enhanced, from ∼1 mho or less in the presubstorm nightside ionosphere, to values of ∼10–100 mho. This has the effect of suppressing the flow and electric field within the perturbed region, but even so intense currents flow within the region, consisting principally of a westward‐directed Hall current or “substorm electrojet.” The total current flow is typically ∼0.5–1 MA, yielding magnetic perturbations on the ground underneath the electrojet of several hundred nanoteslas. The disturbance field is directed equatorward, weakening the horizontal component of the planetary dipole field in both hemispheres. Current continuity is maintained by FACs which flow from the tail plasma sheet into the ionospheric electrojet at its eastern end, and return from the ionosphere to the plasma sheet at its western end, as shown in Figure M162, forming the “substorm current wedge”. The FACs then close in the cross‐tail current on either side of the dipolarized region, and from thence, over the tail lobe magnetopause as shown. It can therefore be seen that the current wedge is associated with a reduction in the cross‐tail current within the dipolarized region, the FACs being associated with the shears in the tail field that occur in the interface between the dipolarized field within the wedge, and the (as yet) undipolarized taillike field outside.
As indicated above, substorm tail reconnection and field dipolarization begin in a localized region near the center of the tail (generally displaced to the dusk side of midnight), but then spreads in both directions across the tail. The bright auroral and electrojet current region in the ionosphere correspondingly starts as a small oval region in the premidnight sector, located typically at ∼65° magnetic latitude, and then spreads toward dusk and dawn, as well as poleward as flux continues to be reconnected and plasma compressed and heated in the nightside tail. This is called the “expansion phase” of the substorm, and was first described by Akasofu in 1964. Sometimes the auroral bulge expands to cover much of the nightside polar ionosphere, though it does not generally reach to the magnetic pole itself. After 20–30 min the auroral bulge reaches its maximum size, and the auroral intensity and the currents then decline, signaling the start of the substorm “recovery phase” which typically lasts a further 30–40 min. In the tail, “recovery” is associated with the down‐tail propagation of the new substorm reconnection region to large distances from Earth, such that the plasma sheet re‐forms in the near‐Earth tail. On the ground underneath the bulge, where the magnetic effects are dominated by the overhead electrojet current, the magnetic depression in the horizontal field typically grows rapidly and impulsively during the expansion phase, and then declines somewhat more gradually during recovery, giving rise to a signature which is called a “magnetic bay” in high‐latitude magnetic records. Substorm signatures are also observed in nightside midlatitude magnetic fields, but in this case both the ionospheric currents and the FACs of the current wedge contribute to the form of the overall disturbance.
Several substorm cycles often occur each day, driven by individual episodes of modest southward field in the IMF, producing magnetic disturbances in the polar region under the electrojet of ∼100–1000 nT which grow and decay over intervals of ∼1 h. During rarer extended intervals of strong southward IMF, however, impulsive electrojet activity is essentially continuously present in the high‐latitude nightside ionosphere, with variable magnetic disturbances on the ground underneath peaking at ∼1000–2000 nT. These high‐latitude perturbations occur simultaneously with worldwide field depressions of typically ∼50–250 nT which are produced under the same conditions. These are termed geomagnetic storms and will be discussed in the next section. The largest high‐latitude disturbance observed since systematic records began in 1957 was of ∼3000 nT, corresponding to ∼5% of the planetary field, this being the largest magnetic effect produced at the Earth's surface due to external currents. At other times, however, the IMF may remain small and northward‐pointing for extended intervals, leading to prolonged intervals of “magnetic quiet” on the ground, when only the variations due to the effects of the Sq‐current system are present, driven by the daily thermal‐tidal motions of the upper atmosphere.
Geomagnetic storms
Usually the north‐south component of the IMF is a few nanoteslas in magnitude and fluctuates in sense on timescales of minutes to a few tens of minutes, as indicated above. This drives variable Dungey‐cycle flows and substorms in the magnetosphere, as discussed in the previous sections. However, under some rather rare but specific circumstances the IMF can become very strong (e.g., several tens of nanoteslas), and can remain southward‐pointing for extended intervals of several hours. Under these circumstances a geomagnetic storm is produced at the Earth in which the horizontal (northward) field is depressed globally over intervals of hours and days, the depression typically maximizing at a few tens to a few hundreds of nanoteslas. These effects are masked at high‐latitudes, however, by the stronger and more variable magnetic perturbations associated with magnetosphere‐ionosphere coupling discussed in the previous section.
For reasons that will be outlined below, the occurrence of magnetic storms tends to follow the 11‐year solar sunspot cycle, as first noted by Sabine in 1852, but on average there are roughly 10 storms each year whose midlatitude field depression exceeds 50 nT, and one that exceeds 250 nT. The largest depression observed since systematic records began in 1957 was of ∼600 nT during a storm in March 1989, which occurred near solar maximum. After a variable “initial phase” to be discussed below, the magnetic depression typically grows fairly gradually over an interval of several hours (∼2–12 h), termed the “main phase,” and then decays even more gradually over the following several days (∼1–5 days), termed the “recovery phase.” These effects result from a prolonged enhancement of the Dungey‐cycle flow, which is produced by the prolonged interval of strong southward IMF. During such intervals the boundary between corotating and Dungey‐cycle flow in the inner equatorial magnetosphere shrinks significantly in size, such that the outer layers of the preexisting plasmasphere are stripped off and flow out to the dayside magnetopause. Correspondingly, the hot plasma produced in the tail flows inward to replace it, significantly enhancing the westward‐directed “ring current” produced by the differential drift of energetic ions and electrons in the inhomogeneous quasidipolar planetary magnetic field, as outlined above. The perturbation field produced by the hot inner plasma is a major contributor to the “main phase” field depression at Earth. However, the picture is complicated by the fact that the inflowing hot plasma is usually asymmetric in its distribution around the Earth, leading to partial current “rings” that close in the ionosphere via “region 2”‐type FACs. These FACs also contribute to the overall magnetic disturbance, as do the combined “fringing” fields of the enhanced plasma sheet and tail currents on the nightside. During the recovery phase, however, Dungey‐cycle convection is reduced, so that these current systems also decline. Hot “ring current” plasma now marooned within the newly expanded corotating region slowly decays due to wave‐scattering followed by precipitation into the midlatitude atmosphere, and due to charge‐exchange with neutral atoms of the geocorona. The outer newly corotating flux tubes also refill with cold plasma from the ionosphere on timescales of a few days, thus re‐forming a more extended plasmasphere.
As indicated at the beginning of this section, the extended intervals of strong southward IMF which excite geomagnetic storms, are associated with specific phenomena in the interplanetary medium which link storms with processes on the Sun and the solar cycle. Two principal mechanisms are responsible. In the first, a large loop‐like field structure located in the solar corona suddenly becomes unstable and is ejected away from the Sun, forming a “coronal mass ejection” (CME). The ejection speed is sometimes very rapid, exceeding ∼1000 km s−1, such that the CME plasma and frozen‐in field plow into the slower solar wind ahead of it. This creates a giant shock wave, which propagates through the solar system compressing and heating the ambient solar wind medium ahead of the CME “driver gas”. Behind the shock the IMF is also compressed to high field strengths. When such a shock impinges on the Earth, the magnetosphere is impulsively compressed, such that the horizontal field at the Earth's surface is suddenly increased, typically by a few tens of nanoteslas, corresponding to the effect of the increased Chapman‐Ferraro current at the magnetopause. What happens after this depends on the direction of the enhanced IMF in the compressed solar wind and CME driver gas. These fields may have any orientation depending on the individual circumstances of the event, and may also fluctuate in direction during the event. A magnetic storm “main phase” occurs only if the enhanced IMF points southward during some extended interval after the shock wave has passed. If this occurs (typically in about one in six events), the impulsive field enhancement at the Earth's surface produced by the shock is called the “storm sudden commencement” (SSC), while the variable interval between the SSC and the onset of the main phase (when the upstream IMF happens to turn southward) is called the “initial phase” of the storm. If such an enduring interval of southward field does not occur, and with it no “main phase” field depression, then the shock‐related compression event is simply termed a “sudden impulse”. The connection between these events and the solar cycle results from the fact that ∼3 CMEs occur each day near solar cycle maximum (only a small number of which impinge on Earth), reducing to one every several days during solar cycle minimum. The connection with solar flares, first noted by Carrington in 1860, results from the fact that flares often occur near the sites in the solar corona where CMEs have formed.
During the declining phase of the solar cycle, extended intervals of strong southward IMF can also be produced by another interplanetary mechanism, which thus can also generate geomagnetic storms. The Sun typically produces two types of quasisteady solar wind outflow (as opposed to CMEs), a “slow” variable wind of ∼300–500 km s–1 from regions surrounding closed field regions of the solar corona, and a “fast” steady wind of ∼800 km s–1 from open‐field “coronal holes.” During solar maximum, the solar field is typically very disordered, and with it the solar wind outflow. During the declining phase of the cycle, however, large‐scale off‐axis coronal holes tend to form, which migrate slowly toward the poles of the Sun as activity decreases toward solar minimum. Initially, however, these coronal holes may extend down to the solar equator, such that during each solar rotation the Earth experiences a pattern of fast and slow solar wind output that may persist in form for many solar rotations. The corotating variable solar wind source regions on the Sun thus give rise to “corotating interaction regions” (CIRs) in the interplanetary medium, in which fast plasma outflow from an equatorial coronal hole “runs into” slower plasma that was emitted earlier into the same direction from a noncoronal hole source region on the rotating Sun. This compresses the plasma and may enhance the frozen‐in field by factors of ∼5–10 above the usual values. As the solar wind velocity subsequently drops, however, after the passage of the material from the coronal hole, a rarefaction region is created in the solar wind, and with it regions of weak IMF occur. Geomagnetic storms can be excited by the enhanced fields of the compression phase of CIRs, during intervals in which the solar wind speed is increasing locally at the Earth, if the enhanced field happens to point southward for a significant interval. Such storms, however, are typically not as intense as those generated by CMEs, because the field directions tend to fluctuate more during these events, as opposed to the more ordered field structures produced by CMEs. In addition, these events are not associated with interplanetary shocks, such that no SSC and “initial phase” occurs prior to the onset of the main phase.
Summary
Overall, it can be seen from this discussion that a variety of interlinked and rather complex plasma physical processes affect the magnetic field observed at the Earth's surface, and throughout the magnetospheric region containing the Earth's magnetic field. The current systems involved are those concerned with the magnetopause boundary of the magnetosphere, the hot plasma that flows inside it, and the current systems that link the magnetosphere and ionosphere and are associated with the auroras. These produce peak perturbation fields at the Earth's surface of typically a few tens, a few hundreds, and a few thousands of nanoteslas, respectively, varying on a range of timescales from minutes during substorms to hours and days during worldwide geomagnetic storms. These current systems are linked to the Sun and solar activity through the solar wind outflow, which forms the Earth's magnetosphere and conditions its dynamics. The variable output of the Sun similarly produces effects on a wide range of timescales, from the minute scales associated with interplanetary shocks, up to the ∼11 year timescales of the solar activity cycle.
Bibliography
Cowley, S.W.H., Davies, J.A., Grocott, A., Khan, H., Lester, M., McWilliams, K.A., Milan, S.E., Provan, G., Sandholt, P.E., Wild, J.A., and Yeoman, T.K., 2003. Solar wind‐magnetosphere‐ionosphere interactions in the Earth's plasma environment. Philosophical Transactions A, 361: 113–126.
Ohtani, S.‐I., Fujii, R., Hesse, M., and Lysak, R.L. 2000. Magnetospheric Current Systems, Geophysical Monograph 118. Washington, DC: American Geophysical Union.
Paschmann, G., Haaland, S., and Treumann, R. (eds.), 2002. Auroral Plasma Physics, Space Science Review, Vol. 103.
Sandholt, P.E., Carlson, H.C., and Egeland, A., 2001. Dayside and Polar Cap Aurora. Dordrecht: Kluwer Academic Publisher.
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2007 Springer-Verlag
About this entry
Cite this entry
Cowley, S.W.H. (2007). Magnetosphere of the Earth. In: Gubbins, D., Herrero-Bervera, E. (eds) Encyclopedia of Geomagnetism and Paleomagnetism. Springer, Dordrecht. https://doi.org/10.1007/978-1-4020-4423-6_205
Download citation
DOI: https://doi.org/10.1007/978-1-4020-4423-6_205
Publisher Name: Springer, Dordrecht
Print ISBN: 978-1-4020-3992-8
Online ISBN: 978-1-4020-4423-6
eBook Packages: Earth and Environmental ScienceReference Module Physical and Materials ScienceReference Module Earth and Environmental Sciences