Planetary Magnetospheres

C. T. Russell


Institute of Geophysics and Space Physics and Department of Earth and Space Sciences University of California, Los Angeles California 90024-1567

Originally published in:
Science Progress, 75, 93-105, 1991



The study of magnetospheres is of great scientific interest and practical importance. Magnetospheres provide ideal laboratories in which to study the behavior of plasmas, a state of matter being increasing important in our quest for new energy sources. Moreover, the Earth's magnetosphere is host to numerous expensive satellite systems, and often not a benign host. All planets visited to date have magnetospheres whether the planets have an internally generated magnetic field or not. This article provides a brief overview of these planetary magnetospheres and compares several processes as they occur at different planets.



All planets and comets explored to date have magnetospheres. The existence of these magnetospheres is independent of whether the planet has an internally generated or intrinsic magnetic field, but the nature of these magnetospheres is quite dependent on this fact. For the planets that have no internal magnetic dynamo the solar wind induces a magnetosphere through its interaction with the upper atmosphere and ionosphere. We will distinguish between these two types of magnetospheres by calling them intrinsic and induced magnetospheres according to the source of their magnetic fields.

The study of planetary magnetospheres may at first seem arcane and exotic, and perhaps of little practical importance. The physical processes that take place in magnetospheres involve rarefied gases, often nearly completely ionized, in which collisions seldom occur. Despite the absence of collisions these ionized gases, or plasmas as they are usually called, behave much like collisional gases or fluids, with coherent behavior induced by their electric and magnetic fields. How the analogues of ordinary processes such as diffusion and dissipation take place in these systems are of immense interest to space plasma physicists. Also of great interest are the various phenomena such as magnetic reconnection and Landau damping that have no analogues in ordinary gases. These processes are also of importance to astrophysicists and plasma fusion physicists. In the former case, the planetary magnetospheres provide in-situ data for processes that may occur on a grander scale elsewhere. In the latter case space provides plasmas without wall effects and often with more complete plasma diagnostics.

In addition to these academic motivations there are two very practical reasons to study planetary magnetospheres. First, the Earth has an internally generated magnetic field whose source we still do not understand. We would hope by studying the generation of magnetic fields at the other planets we might better understand our own. In order to determine the characteristics of these fields we must first understand, especially for the weakly magnetized planets, the nature of the external contributions to the magnetic field, those of the planetary magnetosphere. Secondly, we live in an increasingly technological society on a planet with a significant magnetosphere into which we continue to launch sophisticated spacecraft critical to that technological society. The operations of these spacecraft are affected by that environment. In fact several expensive spacecraft have ceased operations because of large magnetospheric disturbances.

The effects of these magnetospheric disturbances are not restricted to the regions well above the surface of the Earth. Power blackouts have been caused by the intense voltage surges induced in long distance power distribution systems. Communication disruptions have been produced as the effects of these disturbances alter the properties of the ionosphere. Finally, these disturbances cause immense auroral displays. At usual times these displays are seen mainly over Siberia, Alaska, Northern Canada and Scandanavia and their southern counterparts but at disturbed times have been seen as close to the equator as Mexico and Japan. It could be asked why we should study these phenomena at planets other than the Earth. The data obtainable from terrestrial satellites must be of higher quality and quantity than those returned by flyby missions to the distant reaches of the solar system. We need to travel abroad in the solar system because the terrestrial magnetosphere presents us with a limited range of boundary conditions and scale sizes with which to test our theories. The solar wind varies greatly with radial distance from the sun and the external conditions it imposes on the various magnetospheres likewise change. These changes and the varying strength of the magnetic fields of the planets also cause the sizes of the magnetospheres to vary greatly. The largest magnetosphere easily could contain the sun and its corona; the smallest could be contained inside the volume of the Earth.

In this review we outline the basic physical processes that occur in both intrinsic and induced magnetospheres, review briefly the highlights of our exploration of these planets and what problems remain, and show some examples of how the same physical process varies as it occurs in different regions of the solar system.


Induced Magnetospheres

The sun emits a constant stream of electrons and protons in all directions at speeds well above the speed of "sound". This supersonic ionized gas, or plasma, called the solar wind carries with it a magnetic field and a frame dependent electric field. The frame-dependence arises due to the high electrical conductivity of the plasma and its magnetic field. In the frame moving with the plasma the electric field under most circumstances is zero. There is no electric field parallel or perpendicular to the magnetic field. In a frame not moving with the plasma there is an electric field perpendicular to the magnetic field and to the velocity vector proportional to both the magnetic field and the component of velocity perpendicular to the magnetic field. This electric field is very important for the removal of a planetary atmosphere from an unmagneticed planet.

Solar extreme ultraviolet radiation ionizes the upper atmospheres of all planets to varying degrees. If the thermal pressure of this ionosphere exceeds the solar wind momentum flux or dynamic pressure, a quantity proportional to the density times the square of the velocity, then the ionosphere can stand off the solar wind and it remains unmagnetized. A magnetic lid or cap forms on the ionosphere called the magnetic barrier and this barrier in turn deflects the solar wind. The solar wind as mentioned above is supersonic and thus this deflection must involve the formation of a detached bow shock. This bow shock, which interestingly forms without the aid of collisions in the gas, slows, heats and deflects the solar wind. Figure 1 shows a cross section of this interaction.


The behavior of the ionosphere in such an interaction is quite unexpected. Although the thermal pressure of the ionosphere may be strong enough to hold off the solar wind, still small magnetic filaments or magnetic flux ropes sink from the magnetic barrier into the ionosphere, providing an opportunity to study, in- situ, a phenomenon otherwise seen only remotely on the solar surface. When the solar wind dynamic pressure is high and exceeds that of the thermal ionosphere magnetic field and plasma is pushed downward into the ionosphere and it acquires a steady global magnetic field.


The induced magnetosphere has one more very important feature. The solar wind moves past the planet at supersonic speed carrying its magnetic field with it. Near the planet the flow is slowed. The magnetic field that connects the fast and slow regions must perforce be distorted as shown in Figure 2 leading to the generation of a magnetic tail. The interaction can pick up mass from the ionosphere, and through ionization from the atmosphere. This further slows the flow near the planet and increases the magnetic flux in the tail. The bend in the magnetic field and gradients in field strength act to accelerate the plasma in the antisolar direction. Much plasma can reach escape velocities by this mechanism.


Another route for atmospheric loss is the electric field of the solar wind. If particles are ionized in the magnetized flow, they will be quickly accelerated by the electric field and if the direction of the acceleration is correct they can spiral out into the solar wind as illustrated in Figure 3. The combined effect of the electric and magnetic fields of the solar wind acts to remove atmospheric gases from the unmagnetized planets only some of which is replenished by the absorption of the incoming solar wind.

The magnetic moment of Venus is less than one hundred thousandths of that of the Earth and plays no role in the solar wind interaction with the planet. Venus has been extensively explored in the Soviet and American programs with the Mariner 2, 5 and 10 flyby missions, the Venera 2, 4, 6, 8-14 landers;the Venera 9 and 10 orbiters and the Pioneer Venus atmospheric probes and orbiter. The orbiter missions especially have revealed much of the understanding outlined above. Nevertheless we still do not know how much atmosphere is being lost to the solar wind, nor do we understand many of the phenomena found to occur in the ionosphere such as the formation of magnetic flux ropes.

The precise size of the magnetic field of Mars is not known but its strength is probably much less than one ten thousandths of that of the Earth and like Venus the intrinsic magnetic field is not significant for the solar wind interaction. The Martian magnetosphere has been studied by the Mariner 4 flyby mission and the Mars 2, 3, 5 and Phobos orbiters. The ionosphere is thought to be magnetized because the solar wind dynamic pressure exceeds the thermal pressure of the ionosphere but no measurements have been made to confirm this hypothesis. Other features, such as the bow shock and magnetotail, are very similar to those of Venus. We have better measurements of the loss of the Martian ionosphere due to the solar wind interaction taken on the Phobos mission but at this writing these data are not yet fully reduced.

Comets are much smaller objects than planets if only their nuclei are considered. Their much smaller mass means that gravity is not a factor in the solar wind interaction. The size over which the cometary gas can spread in the solar wind is thus controlled by the speed of expansion of the cometary gas (about one km/s) and the ionization time (about a day at 1 AU from the Sun). Their product is about 105 km which is much larger than the size of the interaction regions at Venus and Mars. Not only does the interaction cover greater territory but it is much more gradual. Thus, for example, the bow shock is much weaker at a comet because much of the ionization forms ahead of the region where the bow shock forms so the solar wind is slowed prior to the shock. Measurements at and near comet Halley were made by five spacecraft Vega 1 and 2, Giotto, Sakigake and Suisei. Measurements from the smaller comet Giacobini- Zinner, were obtained from the ISEE-3 spacecraft. In no case were measurements made in the fully developed cometary tail. The data returned by these missions provided interesting insights into the physics of cometary magnetospheres but mainly whetted the appetites of cometary physicists. A mission that matches trajectories with a comet and can take long-term measurements is needed before the processes occurring at a comet are fully understood.


Intrinsic Magnetospheres

For the magnetized planets, those with intrinsic magnetic fields, the obstacle to the solar wind is the planetary magnetic field and the size of the magnetosphere is governed by the relative strengths of the magnetic field and the solar wind at the planet. The strength of a planetary magnetic field is given by its dipole magnetic moment, the equatorial surface field strength times the cube of the planetary radius. The dipole magnetic field falls off as the cube of the radius of the planet. Since the pressure balance is established between the magnetic pressure and the solar wind dynamic pressure at the subsolar point and since magnetic pressure is proportional to the square of the magnetic field strength, the sizes of planetary magnetospheres are proportional to the sixth root of the dynamic pressure. The table lists the dipole magnetic moments for all of the planets, the average solar wind dynamic pressure for each planet which decreases as the square of the distance from the sun and the expected location of the pressure balance point along the subsolar direction. Only one planet, Jupiter, fails to follow this simple relation. At Jupiter part of the outward pressure is supplied by rapidly rotating plasma supplied by the volcanoes of Io. As the table shows the magnetosphere of Mercury is clearly the smallest and that of Jupiter is by far the largest.

               Distance        Magnetic        Solar wind     Magnetopause
Planet         from sun (AU)   moment (ME)     pressure       distance

Mercury         0.4               4 x 10-4       20 nPa       1.5 Rm
Earth           1.0               1.0           3.0 nPa        10 Re
Jupiter         5.2               1.8 x 104     0.1 nPa        70 Rj
Saturn          9.5             580              30 pPa        21 Rs
Uranus         19.2              50               8 pPa        27 Ru
Neptune        30.1              24               3 Ppa        26 Rn

The magnetosphere of the Earth is of course the magnetosphere that has been most thoroughly studied. Because the properties of this magnetosphere generally lie in the middle of the range of properties found in the solar system we can regard the terrestrial magnetosphere as typical. Figure 4 shows a cut away drawing of the magnetosphere. The outer boundary of the magnetosphere is called the magnetopause, upon which flows the magnetopause current, a large current vortex which separates the magnetic field of the Earth and the solar wind. Behind the Earth are the two lobes of the magnetic tail, the top one pointing to the Earth and the bottom one pointing away. These magnetic field lines enter and leave the Earth in oval shaped regions known as the polar caps. These polar caps vary in size as solar wind conditions vary. This variation plays a very important role in energy transfer into the magnetosphere and will be discussed in greater detail below. Between the two tail lobes flows the neutral sheet current which is simply part of the magnetopause current vortex and also the plasma sheet a hotter and denser plasma than in the surrounding regions. The production of this plasma sheet is one of the areas of most intense study at the present time.


Deeper in the magnetosphere we find the plasmasphere, a region of dense cold plasma which is the upper extension of the ionosphere. The plasmasphere extends out to about 5 Earth radii. Within this distance magnetic flux tubes fill up with cold plasma from the ionosphere below. Outside this distance the filling time is long compared to the transport and loss time so the magnetic flux tubes do not fill up with cold plasma.

The closed, dipolar field lines in the magnetosphere provide efficient magnetic mirrors in which to trap energetic particles. Close to the Earth these radiation belts are very stable and can remain constant for hundreds of years but in the outer regions the belts are subject to frequent disturbances and change from day-to-day. Particles from the outer regions can cross the field lines by diffusion and convection. Diffusion is a slow process which relies on fluctuations of the magnetic and electric fields. Convection refers to the drifts induced by the large scale electric field in the magnetosphere. It is important only for low energy particles and only in the outer parts of the magnetosphere.

If one pushes or pulls on the outer parts of the magnetosphere, one would expect the stresses created by that action to affect the plasma in the Earth's ionosphere for the ionosphere is where the magnetosphere is coupled to the Earth. The magnetosphere communicates this stress through field-aligned currents. Figure 4 shows the paths of some of these currents.

The magnetic moment of Mercury is about one 1/3000th of the terrestrial magnetic moment. The equatorial surface magnetic field strength is about 250 nT. Mercury has been explored by only one spacecraft Mariner 10 which passed by Mercury 3 times in 1974 and 1975. On two of these passes the spacecraft passed through the wake of the planet encountering a mini-magnetosphere much like that of the Earth. These two passes gave us only a brief glimpse of the nature of the Mercury magnetosphere. This glimpse was not enough to precisely determine the strength of the magnetic moment of the planet. It did however suggest that the magnetosphere more efficiently extracts energy from the solar wind than does the Earth's magnetosphere. Scientists hope to revisit Mercury in the future with one or more orbiting spacecraft, but presently it is expected that this will not happen until early in the 21st century.

The equatorial surface field of the Earth is about 31,000 nT. It is strong enough to activate rudimentary magnetic compasses and has been used as a navigational aid for at least 1000 years. The investigation of the earth's magnetic field began in about the 16th century but reached its zenith in the space age when it could be more fully explored with spacecraft. The spacecraft which have examined the Earth's magnetosphere are too numerous to name and have been launched by all the spacefaring nations. At present the most active area of research in magnetospheric physics is energy transfer from the solar wind to the magnetosphere. In the mid-1990's a consortium of space agencies (ESA, Intercosmos and NASA) are going to launch a flotilla of spacecraft into the magnetosphere to study this problem. This program is called the International Solar Terrestrial Program and will consist of over 15 different spacecraft.

The magnetic moment of Jupiter, as befitting the largest planet in the solar system, is also the largest of the planetary system over 10,000 times that of the earth. Its equatorial surface field is over 10 times that of the Earth. The strength of its magnetic field combined with the weakness of the solar wind at Jupiter produces a magnetosphere that is enormous. The sun could easily fit inside the magnetosphere. Its tail is thought to extend past Saturn, over 5 AU away. If Jupiter's magnetosphere could be seen from Earth it would appear to be larger than the Earth's moon.

Deep inside the jovian magnetosphere orbit the Galilean satellites. One of these, Io, has a volcanically produced atmosphere that is constantly being bombarded by the intense radiation belts of jupiter. This bombardment knocks atoms out of the atmosphere of Io into the magnetosphere of Jupiter where they become ionized. This process produces a torus, or doughnut, of hot ions circling Jupiter near Io's orbit. This torus together with the enormous electrical and magnetic forces in the Jovian magnetosphere leads to intense radiation belts and radio emissions. These emissions can be detected from Earth and were the first indication of Jupiter's enormous magnetic field well before the first interplanetary spacecraft were launched.

Jupiter has been visited four times by spacecraft: Pioneer 10 in 1973; Pioneer 11 in 1974; and Voyager 1 and 2 in 1979. Each of these spacecraft were on flyby trajectories. At this writing the Galileo spacecraft is on its way to Jupiter when it will be injected into an elliptic near equatorial orbit in 1995.

The magnetosphere of Saturn is quite benign compared to that of Jupiter. Since Saturn is a smaller planet, its conducting core in which the planetary magnetic field is generated is smaller, and so is the planetary magnetic field. The magnetic moment of Saturn is 580 times that of the Earth but its equatorial surface magnetic field strength is about equal that of the Earth. In stark contrast to the magnetic fields of all the other planets, the Saturnian dipole moment is not tilted with respect to the rotation axis of the planet. This observation was a great surprise to those studying planetary magnetic dynamos. Saturn's ring system absorbs radiation belt particles so that the radiation belts are weaker than at Jupiter and none of Saturn's moons exhibits volcanic activity similar to that of Io. As a consequence Saturn's radiation belt resemble more those of the Earth than those of Jupiter and few radio emissions are produced.

Saturn has been visited by 3 spacecraft Pioneer 11 in 1979, Voyager 1 in 1980 and Voyager 2 in 1981. Each of these were on flyby trajectories. Currently, NASA and ESA are working on an orbiter/probe mission called Cassini/Huygens which is scheduled to arrive at Saturn early in the 21st century.

Uranus and Neptune
The magnetic fields of Uranus and Neptune are quite unlike those of the other planets. The magnetic fields are quite irregular and cannot be well represented by a simple dipole field. When a dipole moment is fit to the flyby data available from Voyager 2 which flew by these planets in 1986 and 1989 respectively, a very large tilt angle between the rotation axis and the dipole axis is found, about 50o. The magnetic fields are also much weaker than those found at Jupiter and Saturn. The magnetic moments are about 40 times that of Earth and their surface magnetic fields slightly less than the terrestrial field. The reason for this weakness and the irregularity may be that the magnetic field is generated, not in a deep molten core like the Earth's, but in salty ice/water oceans closer to the surface. The radiation both of Uranus and Neptune are quite weak. There are no present plans to explore these planets further.


Comparative Magnetospheres

The magnetospheres of the planets differ both in size and internal energy sources but also in the strength of the solar wind flow past their surfaces. Thus, the interaction of each of the magnetospheres with the solar wind differs in some degree from the others. Herein we examine how some of these processes vary from planet to planet.

The Bow Shock
The bow shock is a standing wave in front of a magnetosphere at which the supersonic solar wind is slowed, heated, and deflected around the planet. The strength of this shock depends on the flow velocity of the solar wind relative to the velocity of compressional waves in the plasma. This latter velocity decreases with increasing distance from the sun while the former remains quite constant. As a result, the strength or Mach number of the bow shock increases markedly from the inner solar system to the outer solar system. At Mercury the bow shock has a Mach number of about 4 but at Neptune it is about 20. At low Mach numbers the shock is found to be quite smoothly varying or laminar in appearance but at high Mach number the shock becomes very turbulent.

Upstream Waves
The bow shock represents an obstacle to some of the solar wind particles and they are reflected back upstream along the magnetic field. These counterstreaming particles cause waves to grow in the solar wind. These waves cannot propagate upstream against the solar wind and are blown back against the planetary shocks. The number of particles reflected by a planetary bow shock increases with the strength of the bow shock. Thus the strength and characteristics of the upstream waves change with heliocentric distance. Figure 5 shows one such property of the waves, their frequency. As one moves outward in the solar system the frequency of the waves change in proportion to the field strength as would be expected if the waves were associated with a gyro resonance with the reflected solar wind ions.


Another process that appears to be influenced by the Mach number is the phenomenon known as reconnection. In this process magnetic field lines in the solar wind link up with those of the planetary magnetosphere, thereby increasing the tangential stress on the magnetosphere and adding magnetic energy to the magnetotail. Under solar wind conditions typical of those in the inner solar system this process is controlled principally by the direction of the solar wind magnetic field relative to the direction of the planetary magnetic field. When these directions are antiparallel, reconnection takes place readily and, when they are parallel, it does not take place at all. However, when solar wind conditions change to those typical of the outer solar system reconnection seems to cease. This is illustrated in Figure 6 which shows the reconnection efficiency judged from terrestrial records of geomagnetic activity versus solar wind Mach number. It shows that about a Mach number of 7 the reconnection rate appears to go to zero. Thus reconnection is expected to be more important in the inner solar system where the Mach number is typically 7 or less than in the outer solar system where it is often 10 or greater.


An associated phenomenon is that known as the Flux Transfer Event which appears to be the signature of temporally and spatially varying reconnection. These features have been observed at the magnetopauses of Mercury, Earth, and Jupiter. At Mercury these events are of short duration, about 1 s and occur frequently about every 30 s. At Earth these features last about 30 s and occur about every 5 minutes. At Jupiter the signature is similar to that at the Earth. This observation suggests that the small size of the Mercury magnetosphere affects the generation of Flux Transfer Events. However, at Earth and Jupiter the size of Flux Transfer Events may be controlled by some other property of the magnetosphere such as the thickness of the magnetopause which is the same at both planets.



In the sections above we have outlined the general features of planetary magnetospheres. Some of these magnetospheres are induced and some intrinsic. Both types stand off the solar wind flow and cause planetary bow shocks. The variation of the solar wind with distance from the shock together with other planets to planet differences causes a spectrum of responses to the solar wind flow. These differences in turn allow us better to understand the processes taking place. The space missions to these planets over the last 2 decades have returned a wealth of data about their magnetospheres, data through which we are still sorting. Many mysteries have been answered with the acquisition of these data, yet many mysteries remain. Thus we look forward to the upcoming missions such as the International Solar Terrestrial Program, Galileo, Cassini and Mercury Orbiter to help solve these problems.



The preparation of this report was supported by the National Aeronautics and Space Administration under research grant NAS2-501.



  1. Hundhausen, A. J. (1972) Coronal Expansion and Solar Wind, Springer- Verlag, Berlin.

  2. Luhmann, J. G. (1986) Solar wind interaction with Venus, Space Sci. Rev., 44, 241.

  3. Luhmann, J. G. and Cravens, T. A. (1991) Magnetic fields in the ionosphere of Venus, Space Sci. Rev., 55, 201-274.

  4. Russell, C. T., Elphic, R. C. & Slavin, J. A. (1980) Limits on the possible intrinsic magnetic field of Venus, J. Geophys. Rev., 85, 8319.

  5. Schwingenschuh, K. et al. (1991) Adv. Space Res., in press.

  6. Grewing, M., Praderie, F. & Reinhard, R. (eds) (1988) Exploration of Halley's Comet, Springer-Verlag, Berlin.

  7. Russell, C. T. (1987) Planetary magnetism, In: Geomagnetism, Vol. 2, (ed. by J. A. Jacobs) p 457, Academic Press, New York.

  8. Russell, C. T. (1987) The magnetosphere, In: The Solar Wind and the Earth, (ed. by S-I. Akasofu and Y. Kamide) p 73, Terra Scientic Publishing Co., Tokyo.

  9. Russell, C. T., Baker, D. N. & Slavin, J. A. (1988) The magnetosphere of Mercury, in Mercury, (ed. by F. Vilas, C. Chapman and M. Matthews), p 514, Univ. Arizona Press, Tucson.

  10. Dessler, A. J. (ed.) (1983) Physics of the Jovian Magnetosphere, Cambridge University Press, Cambridge.

  11. Russell, C. T. (1985) Planetary bow shocks, In: Collisionless Shocks in the Heliosphere: Reviews of Current Research, (ed. B. T. Tsurutani and R. G. Stone), p 109, American Geophysical Union, Washington, D. C.

  12. Russell, C. T., Lepping, R. P. & Smith, C. W. (1990) Upstream waves at Uranus, J. Geophys. Res., 95, 2273.

  13. Scurry, L. & Russell, C. T. (1991) Proxy studies of energy transfer to the magnetosphere, J. Geophys. Res., 96, in press.

  14. Russell, C. T. & Elphic, R. C. Initial ISEE magnetometer results: Magnetopause observations, (1978) Space Sci. Rev., 22, 681.

  15. Russell, C. T. & Walker, R. J. Flux transfer events at Mercury, (1985) J. Geophys. Res., 90, 11067.

  16. Walker, R. J. & Russell, C. T. (1985) Flux transfer events at the Jovian magnetosphere, J. Geophys. Res., 90, 7397.

Back to CT Russell's page More On-line Resources
Back to the SSC Home Page