MULTISCALE COUPLING IN PLANETARY MAGNETOSPHERES
C. T. Russell
Department of Earth and Space Sciences and Institute of Geophysics and Planetary Physics
University of California Los Angeles
ABSTRACT
Processes in planetary magnetospheres occur on a variety of scales. On the largest scales are the plasma circulations
induced in the magnetospheric plasma externally by the solar wind interaction or internally by processes such as
massloading of the jovian magnetosphere by the moon Io. These large-scale processes are influenced by smallscale processes such as particle scattering by waves, gyro motion, and charge exchange. It is not always clear
which of two coupled processes are in control. For example, many believe that the onset of reconnection requires a
microscale instability but it is possible that external forcing leads to the magnetic configuration at which
reconnection eventually occurs, and it is only the existence of the appropriate magnetic configuration over a
sufficiently large region that is required to cause the macroscopic flows. This paper reviews such coupling in the
magnetospheres of the Earth and Jupiter.
INTRODUCTION
There is no question that microscale and macroscale processes in magnetized plasmas are intimately coupled.
The supersonic solar wind blows against the magnetosphere and a standing shock wave forms that deflects the
flow, slows and heats it so it can flow around the magnetosphere. The shock is thin, less than an ion inertial length.
An electric field is set up along the shock normal that slows the solar wind ions and a standing whistler mode wave
is formed that guides the electrons along and through the shock avoiding the full effect of this electric field. The
thin scale of the shock scatters the gyro motion of the ions heating the ions and providing the dissipation required
by the Rankine-Hugoniot relations. The shock sets the properties of the plasma that then flows against the
magnetopause and the properties of the flow controls whether or not reconnection will occur and at what rate. The
reconnection process is another microscale phenomenon on its finest scales yet it is responsible for the large-scale
circulation of plasma through much of the Earth’s magnetosphere.
It is not always clear whether a microscale process is a key player in a magnetospheric process or is just
incidental. For example, waves are regularly seen in conjunction with reconnection. Are these waves enabling or
peripheral? Waves occur during substorms. Some models make the waves the primary agent of change, such as
the ballooning instability that may push flux tubes outward [e.g. Pu et al., 1992; Hurricane et al., 1997;
Bhattacharjee et al., 1998] or as pitch angle scattering that may evacuate a flux tube and lead to collapse of the
nightside magnetosphere [Parks et al., 1972]. Equally well other models put the reconnection process in the
driver’s seat [Russell and McPherron, 1973] and postulate that the waves that appear at substorm onset result from
the accelerated flow and its unsteadiness.
Our examples above are all drawn from fully ionized plasma in which the coupling takes place via collisionless
processes and Coulomb collisions are rare and unimportant. Not all geophysically important plasmas are
collisionless. In particular, planetary ionospheres occur in a neutral atmosphere whose density most frequently
exceeds that of the ionosphere. Charge exchange is possible in these environments and leads to the acceleration of
neutrals when this might not otherwise be expected. This is another example of the microscale process leading to a
macroscale effect, altering the circulation of the neutrals and the plasma.
In this review we examine in detail some of these processes illustrating how the micro and macroscales interact
in the magnetospheres of Earth and Jupiter. We first begin by looking at reconnection in the Earth’s magnetosphere
as revealed by the instruments on the Polar spacecraft.
B X Payload [nT]
4
Z NIF
Magnetosheath
B( )
Magnetosphere
II
III
YNIF
E(
X NIF
)
May 29, 1996
0704:29.449
B Y Payload [nT]
Exhaust
0
-4
Inflow
0
-4
-8
-YNIF
Inflow
I
IV
-X NIF
E(B(-
)
)
Separatrix
0
-4
-1.0
-Z NIF
0
1.0
8
4
0
-1.0
0
1.0
∆X/(c/ω pe )
Exhaust
Fig. 1. Geometry of the reconnection region in the normal
incidence frame, NIF [after Scudder et al., 2001].
4
B Payload [nT]
B Z Payload [nT]
Separator
Fig. 2. Magnetic field components at 54 Hz near a null point on
May 29, 1996 [after Scudder et al., 2001].
MAGNETOPAUSE RECONNECTION
Magnetic reconnection occurs when oppositely directed magnetic fields, initially separated by a current sheet
with no interconnection of the magnetic field across the current sheet, establish a connection across that current
sheet. Figure 1 shows the geometry of this process [Scudder et al., 2001]. On either side of the current sheet
plasma moves inward toward the center once reconnection has begun and “reconnected” plasma is exhausted from
the central region allowing the inflow to proceed. The energy in the exhaust flow is provided by the Lorentz or
JxB force of the bent field lines. The boundary between the topologically distinct magnetic field lines is called the
separatrix.
The high resolution plasma and magnetic data from the Polar spacecraft have been used to study this process
concentrating on a few hour period on May 29, 1996 when the spacecraft was near apogee and the solar wind
dynamic pressure was of sufficient strength to keep the magnetopause in the vicinity of the spacecraft [Scudder et
al., 2001]. Figure 2 shows the magnetic field measurements at 54 Hz for 400 msec across a magnetic field null.
The temporal variation has been converted to a spatial variation with the measured plasma velocity. We see that
the scale lengths are of the order of the electron inertial length c/ωpe.
The scale lengths of the magnetic field measured as a fraction of the electron inertial length and the electron
thermal gyro radius are shown in Figure 3 together with the electron beta, all plotted versus the minimum magnetic
field in the depression. It is clear from this display that the scale size is ordered by the ion inertial length and not
the electron thermal gyro radius and also that the electron beta can be extremely large in these field nulls. Thus on
the finest possible scales in the plasma the enabling physics is taking place that controls the flows on the largest
possible scales in the magnetosphere but what controls the rate of reconnection?
THE EFFECT OF THE BOW SHOCK ON RECONNECTION
An important factor in controlling the rate of reconnection is the nature of the plasma in which the reconnection
is taking place. The accelerated plasma reaches the Alfven velocity based on the normal component of the
magnetic field and the local plasma density. The ISEE 1 and 2 measurements at the dayside magnetopause were
the first to show this [Paschmann et al., 1979]. The properties of the plasma in the region depend on the conditions
at the bow shock. Figure 4 shows a sketch of the solar wind interaction with the Earth’s magnetosphere. The shock
front deflects the flow, heats, compresses and slows the solar wind plasma. It also bends the magnetic field. The
structure of the shock depends in part on the orientation of the magnetic field. The shaded region denotes the
foreshock region where heated ions flow upstream against the solar wind flow and interact with the solar wind. As
2
L B -1=d log B /dX NIF
L B /L*
100.00
L*=c/ω pe
1.00
0.01
L B /L*
100.00
-1
L*=w th,e /Ω ce
1.00
X
-1
2
Field Line
Streamline
0.01
1
βe
100.00
Foreshock
Boundary
1.00
0.01
1
10
B min [nT]
Fig. 3. Variation of the dimensionless scale length of the
depression in B in units of the electron inertial length
(top); in units of the electron thermal gyroradius (middle)
and variation of βe with magnetic field strength in the
field depression (lower panel) [after Scudder et al., 2001].
Y
Fig. 4. The flow and field lines in the Earth’s magnetosheath.
Shaded region indicates the foreshock and the region downstream of the foreshock.
illustrated here, the magnetic field typically makes a 45o angle with the solar wind flow. However, at times the
magnetic field can become aligned with the flow and when it does it is not compressed at the subsolar shock. The
plasma that intersects the subsolar point paints the magnetopause and, when the magnetopause is covered with a
weak field, reconnection appears to be slow [Scurry and Russell, 1991].
Another situation that can weaken the magnetic field in the sheath is the occurrence of a high Mach number at
the bow shock that causes a high beta downstream so that the pressure applied to the magnetosphere is due to the
plasma and not the magnetic field. Figure 5 shows a proxy measure of the reconnection rate based on the level of
geomagnetic activity. At low Mach numbers geomagnetic activity is fairly constant but, as the Mach number
increases, geomagnetic activity suddenly decreases at about 8. This is not generally important at Earth where such
Mach numbers are rare but this can be important in the magnetospheres of the outer planets where the Mach
number is typically in this range.
RECONNECTION IN THE JOVIAN MAGNETOSPHERE
The solar wind dynamic pressure establishes a magnetic cavity at Jupiter much like that at the Earth but much
grander in scale. The high Mach number of the solar wind affects the rate of reconnection but in any event the
rapid rotation and vast scale of the jovian magnetosphere causes a circulation pattern that is little affected by the
solar wind except to determine its overall scale size. Nevertheless, there is a circulation pattern in the jovian
magnetosphere that is intimately tied to reconnection but this circulation has an internal driver.
Figure 6 shows a sketch of the interaction of Io with the jovian magnetosphere. Jupiter rotates rapidly and
drags its magnetospheric plasma with it because of the very high electrical conductivity of the magnetosphere along
field lines. This is similar to the corotation of the Earth’s plasmasphere. Io orbits at 5.9 jovian radii (RJ) at an
orbital speed of 17 km/s while the magnetospheric plasma moves at 74 km/s. The net difference in velocity of 57
km/s means that newly formed ions from the exosphere of Io will experience an electric field, VxB, that rapidly
accelerates the ion and causes it to drift with the other plasma. These ions build up in intensity to form a torus of
about 4000 ions/cm3 density extending out to about 7.7 RJ. At Io the massloading is strong enough to bend the
magnetic field, leading to field-aligned currents that close in the ionosphere. The massloading process as discussed
at greater length in a later section involves the production of fast neutrals that cross the magnetic field lines and
help spread the torus over the region observed.
3
20
∆am adj /∆VB tan [nT/mV-m-1]
B
J ll
Fast
Neutrals
10
VIO
17 km/s
VCO ~
~74 km/s
0
-10
B
0
2
4
6
8
10
V
E = -V x B
J ll
Io/Neutral Rest Frame
Pickup Ion
Magnetosonic Mach Number
Fig. 5. The efficiency of the interplanetary electric field in
generating geomagnetic activity as a function of the solar
wind magnetosonic Mach number.
Fig. 6. The massloading process at Io in Jupiter’s
magnetosphere. Io orbits Jupiter at 17 km/s and the
magnetospheric plasma corotates at 74 km/s. This interaction
bends the magnetic field lines leading to Jǁ.
The inexorable outward march of the iogenic plasma allows us to examine the radial evolution of the structure
of the plasma sheet as a temporal evolution. Figure 7 shows a sequence of crossings of the current sheet in the
jovian magnetodisk beginning at 25.8 RJ [Russell et al., 1999]. The coordinate system has been chosen so that the
magnetic field component crossing the current sheet is in the b direction. The a direction is the direction that
reverses across the sheet. There is always some high frequency structure in the current sheet but as one moves
outward the component across the average sheet direction begins to become more variable. Part of this variability
is due to the bending of the current sheet but part of it must also be due to the growth of tearing in the current. This
is especially evident at 54.6 RJ where sharp downward spikes are seen.
Figure 8 shows a 21-day period near midnight where the current sheet has rocked back and forth across the
spacecraft because of the tilt of the dipole [Russell et al., 1998]. The Bs component here is the component across
the average current sheet. A dipolar field would have a positive (southward) direction in this coordinate system.
There are frequent glitches in this component. Some of these are very large and all of them are real, not telemetry
dropouts.
Figure 9 shows an enlargement of one of these events. One hour of data is shown. The magnetic field goes
strongly vertical, reaching a magnetic field strength over three times that seen in the lobes above and below the
current sheet. This rapid and extremely strong event is interpreted to be reconnection. We believe the reconnection
event is so strong because reconnection has reached the very low plasma density conditions in the lobes. In 45
minutes the magnetodisk has returned to normal.
Figure 10 shows the magnetospheric circulation pattern that we believe is consistent with this reconnection
[Vasyliunas 1983]. Throughout the inner magnetosphere, the plasma corotates and slowly moves outward. In the
distant nightside magnetosphere reconnection occurs making a magnetic island that contains ions but the net
magnetic flux is zero. The flux tube on the closed side of the reconnection region has been emptied of much of its
plasma and returns to the inner magnetosphere floating against the outgoing flow. The emptied tubes appear to
breakup into slender tubes that can move inward rapidly, reaching the Io orbit and repeating the massloading circuit
[Russell et al., 2000].
4
20
25.8 R J
Delta B
Delta B
Ba
0
41.4 R J
Ba
10
0
Normal
20
-10
10
Bb
Current
0
20
Bc
0
40
20
Magnitude
Current
Normal
-20
Magnitude
Magnetic Field in Current Sheet Coordinates [nT]
20
Bmag
Bb
0
10
Bc
0
20
Bmag
10
0
0
1400
1500
1600
1700
0900
1000
1100
June 24, 1996
12
51.9 R J
6
Delta B
Delta B
Ba
0
54.6 R J
Ba
6
0
Normal
6
0
12
6
Bb
Current
0
6
Bc
Magnitude
Current
Normal
-6
Magnitude
Magnetic Field in Current Sheet Coordinates [nT]
September 5, 1996
Bmag
0
0600
-6
3
0
Bb
0
-3
Bc
Bmag
6
0
0700
2000
0800
2100
September 2, 1996
Universal Time
2200
Universal Time
September 1, 1996
Fig. 7. Radial evolution of the structure in the jovian current sheet. Component a shows the magnetic field
that changes across the current sheet; b shows the component that is approximately constant and normal to
the current sheet; c shows the component that is approximately zero and in the direction of the “azimuthal”
current [Russell et al., 1999]. With increasing radial distance, increasing structure in the component
crossing the current sheet, Bb, is observed. At 54.6 RJ the structure looks like tearing islands.
B
Radial
Bs
Bt
0
Theta
Magnetic field [nT]
0
A
Outward
Inward
-8
8
South
0
North
Azimuthal
10
-10
0
16
4
0
With Rotation
Opposite Rotation
-10
10
0
June 1
16
B
Magnitude
Radial
Southward
Tangential
Br
0
-10
Magnitude
Magnetic Field in Corotational Coordinates [nT]
10
A
June 6
June 11
June 16
8
0
1200
June 21
1997
1230
Universal Time
1300
June 17, 1997
Fig. 9. An expansion of event A in Figure 8. Coordinate
system is the same as in Figure 8 [after Russell et al., 1998].
Fig. 8. Twenty-one days of magnetic measurements in the
night side jovian current sheet. The radial, Br, southward, Bs
and tangential or azimuthal, Bz, components are shown. The
glitches in the north-south component show the presence of
the jovian equivalent of substorms [after Russell et al., 1998].
5
Magnetic
X-Line
4
Magnetopause
4
3
Magnetic
0-Line
3
2
2
1
1
Fig. 10. Schematic diagram of the circulation of plasma in the jovian magnetosphere [after Vasyliunas, 1983].
The left-hand panel shows the circulation in the equatorial plane and the right-hand panel shows the plasma
circulation and magnetic configuration in four planes whose locations are given in the left-hand panel. When the
flow reaches plane 2, reconnection begins producing an X-O neutral line pair that grows as the flow advances to
earlier local times.
A LESSON FOR EARTH
Returning to the Earth’s magnetosphere for a moment we can draw a lesson for substorm genesis from both the
Earth’s dayside magnetopause and the jovian magnetodisk. One of the puzzles of substorm onsets is what causes
the substorm expansion phase. Other questions include why northward turnings of the IMF trigger substorms, why
sometimes there is a pseudo breakup when a substorm seems to slow and then stops before it proceeds far and
double onset substorms. Figure 11 shows a sketch of a model substorm that uses the fact that the reconnection
depends on the Alfven velocity at the reconnection point [Russell, 2000]. In this model a distant neutral point is
reconnecting magnetic flux in the tail lobe while flux is being added by dayside reconnection. A second neutral
point has formed near the Earth. It proceeds slowly and cannot reach the open flux of the lobes because the distant
neutral point provides plasma sheet that keeps the Alfven velocity low at the inner neutral point. The formation of
the near Earth neutral point signals the first onset of dual onset substorms and/or the psuedo-breakup onset. If the
IMF turns northward the distant neutral point will stop and the near Earth neutral point will continue to reconnect
and eventually reach the lobes. There the reconnection rate jumps and the main expansion phase can occur. If the
IMF returns southward before the near Earth neutral point reaches the open field lines, the rate of reconnection can
be slowed at the near Earth neutral point and a psuedo-breakup result.
MULTISCALE COUPLING IN A PARTIALLY IONIZED GAS
Returning to Jupiter we take a closer look at the Io massloading process as shown in Figure 12. The torus
plasma flows into the Io exosphere and, because it is strongly magnetized, there is a strong electric field radially
outward from Jupiter. When exospheric neutrals are ionized they rapidly accelerate and gyrate and drift in a
cycloidal path. Some fraction of these new ions can charge exchange with exospheric neutrals and then travel
rapidly across the magnetic field until photoionization or some other process takes the neutral back to the ionized
state at which time it gyrates and drifts as before. This process creates a disk of massloading in the direction
downstream from Io, both inward and outward, but not upstream [Wang et al., 2001].
Figure 13 shows the neutral cloud produced by following the process in a simulation starting from a state with
no neutral cloud present. As time proceeds and Io orbits, a steady state evolves in which the characteristic neutral
cloud of Io is seen. This neutral cloud is being eroded by ionization with a much longer time scale. The picked up
ions rotate faster than Io and eventually form a complete ring about Jupiter as shown in Figure 14.
6
Φ Lobe
Solar
Wind
M
RD
Φ Day
ΦPM
Φ NPS
ΦDPS
RN
P NL
Flux Transport Rate
S
R
RN
M
RD
Φ NPS Φ Lobe
ΦPM
Thin
0
ΦDPS
Total Magnetic Flux
Φ Day
C
0
Exp
Time
Fig. 11. Schematic diagram illustrating how the presence of two neutral points can produce many of the
subtleties of substorm phenomena such as pseudo-breakups and northward turning triggering of substorm
expansions [Russell, 2000]. The top panel shows the magnetic configuration in the noon-midnight meridian
of the Earth’s magnetosphere well after the interplanetary has turned southward and two neutral points have
formed. The middle panel shows the flux transport rates: reconnection at the magnetopause, M; at the
distant neutral point, RD; at the near-Earth neutral point, RN; and the convection to the dayside, C. The five
dashed lines mark the southward turning of the IMF, S; the onset of plasmoid formation, P, when the nearEarth neutral point forms; the northward turning, N; the time at which the near-Earth neutral point reaches
the tail lobe, L: and the time at which the recovery phase ends, R. The bottom panel shows how the
magnetic flux varies in each region during the substorm in the closed dayside magnetosphere, Day; the
open lobe, Lobe; the plasmoid, PM; and the distant plasma sheet, DPS.
7
Atmosphere Neutrals
Ionization
Neutralization
distant torus
pick up ions
near torus
pick up ions
hν
hν
To Jupiter
hν
slow neutral
Io
Sun light
fast neutral
v
B
E
Fig. 12. Schematic diagram of the details of the massloading process at Io showing how the transport of
neutral particles across the field lines increases the extent of the massloading region. Ions produced near Io
are accelerated in the electric field of the corotating magnetosphere and then neutralized by charge
exchange close to Io. On the Jupiter side of Io these fast neutrals collide with Io but on the anti-Jupiter side
they spread out to form a disk of massloading particles mainly outward and downstream of Io.
Fig. 13. Simulation of the production of the cloud of neutrals
around Io [Wang et al., 2001]. Simulation is started with no
neutral cloud at t=0. After 9.3 hours the neutral cloud,
eroded by ionization is close to steady state.
Fig. 14. Simulation of the production of the Io torus from the
neutral cloud [Wang et al., 2001]. The ions produced in the
neutral cloud shown in Figure 13 drift with the corotating
magnetospheric plasma to eventually produce a complete torus.
8
Average Gyro Velocity (km/s)
150
100
Io
50
0
2
4
6
8
10
12
Radial Distance (R J )
Fig. 15. Gyro velocity of newly born ions as a function of jovicentric distance in the model of
Wang et al. [2000]. The inner torus is produced “cool” and the outer torus “warm” in this model.
This process emplaces ions both inside Io’s orbit and outside. The gyro velocity of this process varies with
radius as illustrated in Figure 15. The plasma produced inside has a much lower gyro velocity than outside. Since
this gyro velocity produces the perpendicular pressure in the plasma, this emplacement results in an initially cooler
plasma inside than outside Io. These temperatures from the simulation are generally warmer than those observed in
the torus but the full corotation velocity was used in the pickup process and some cooling is expected due to
Coulomb collisions with the neutrals. The important point here is that a microscale process has set the large scale
thermal structure of the torus.
SUMMARY AND CONCLUSIONS
There are many examples of multiscale coupling in space plasma. Many more have been proposed than
discussed herein. Sometimes the microscale process is driven by the macroscale process such as when ULF waves
in the magnetosphere are generated by particles accelerated in a substorm. At times microscale processes have
been thought to control the threshold for some instability when the process was actually controlled by macroscale
processes. For example, the breaking of field line connectivity in reconnection is a microscale process but its rate
may be governed by large scale boundary conditions that set the geometry and plasma conditions at the
reconnection site.
In many places in planetary magnetospheres the properties are dependent on microscale processes. The
magnetosheath is produced by the shock and its properties are determined by microscale processes. These flows in
turn may control the reconnection process at the subsolar point through altering the conditions in the plasma. At
the reconnecting magnetopause the partner-shifting mechanism occurs on a microscopic scale resulting in
reconnection flows that are global. In one example we showed how an externally driven reconnection picture could
explain many of the features found for substorms on Earth, even though internal microscopic processes must be
acting at the reconnection point itself.
Finally, we examined a partially ionized gas and showed how the charge-exchange process could lead to
significant mass transport at Io and help set in place the properties of the Io torus. The intimate connection of
ionized and neutral gases that the charge-exchange process enables needs to be considered in many settings in
magnetized plasmas, not just at Jupiter.
ACKNOWLEDGMENTS
This work was supported by the National Aeronautics and Space Administration under research grants NAG58064 and NAG5-8938.
9
REFERENCES
Bhattacharjee, A., Z. W. Ma and X. Wang, Dynamics of thin current sheets and their disruption by ballooning
instabilities: A mechanism for magnetospheric substorms, Physics of Plasmas, 5, 2002-2009, 1998.
Dessler, A. J., Mass-injection rate from Io into the Io plasma torus, Icarus, 44, 291-295, 1980.
Hurricane, O. A., B. H. Fong and S. C. Cowley, Nonlinear magnetohydrodynamic detonation. I. Physics of
Plasmas, 4, 3565-3580, 1997.
Parks, G. K., G. Laval and R. Pellat, Behaviour of outer radiation zone and a new model of magnetosopheric
substorm, Planet. Space Sci., 20, 391-408, 1972.
Paschmann, G., B. U. Ö. Sonnerup et al., Plasma acceleration at the Earth's magnetopause: Evidence for
reconnection, Nature, 282, 243-246, 1979.
Pu, Z. Y., A. Korth and G. Kremser, Plasma and magnetic field parameters at substorm onsets derived from GEOS
2 observations, J. Geophys. Res., 97, 19,341-19,349, 1992.
Russell, C. T., How northward turnings of the IMF can lead to substorm expansion onsets, Geophys. Res. Lett., 27,
3257-3259, 2000.
Russell, C. T. and R. L. McPherron, The magnetotail and substorms, Space Sci. Rev., 15, 205-266, 1973.
Russell, C. T., K. K. Khurana, D. E. Huddleston and M. G. Kivelson, Localized reconnection in the near Jovian
magnetotail, Science, 280, 1061-1064, 1998.
Russell, C. T., D. E. Huddleston, K. K. Khurana and M. G. Kivelson, Structure of the magnetodisk current sheet:
Initial Galileo observations, Planet. Space Sci., 47, 1101-1109, 1999.
Russell, C. T., M. G. Kivelson, K. K. Khurana and D. E. Huddleston, Circulation and dynamics in the jovian
magnetosphere, Adv. Space Res., 26, 1671-1676, 2000.
Russell, C. T., X. Blanco-Cano and R. J. Strangeway, Ultralow frequency waves in the jovian magnetosphere:
Causes and consequences, Planet. Space Sci., 49, 291-301, 2001.
Scudder, J. D., F. S. Mozer, N. C. Maynard, P.A. Puhl-Quinn, Z. W. Ma and C. T. Russell, Fingerprints of
collisionless reconnection I: Evidence for Hall MHD scales, J. Geophys. Res., in preparation, 2001.
Scurry, L. and C. T. Russell, Proxy studies of energy transfer in the magnetosphere, J. Geophys. Res., 96, 95419548, 1991.
Vasyliunas, V. M., Plasma distribution and flow, Physics of the Jovian Magnetosphere, edited by A. J. Dessler,
395-453, Cambridge University Press, London, 1983.
Wang, Y. L., C. T. Russell and J. Raeder, The Io mass-loading disk: model calculations. J. Geophys. Res, 106, 26,243-26,260,
2001.
10