*Institute of Geophysics and Planetary Physics, +Dept.of Earth and Space Sciences,
University of California, Los Angeles, CA 90095-1567
Originally published in:
Science, 280, 1061-1064, 1998.
The oppositely-directed magnetic field in the jovian magnetic tail is expected eventually to reconnect across the current sheet, allowing plasma produced deep inside the magnetosphere near Io's orbit to escape in the antisolar direction down the tail. The Galileo spacecraft has found localized regions of strong northward and southward field components beyond about 50 jovian radii in the post-midnight, predawn sector of the jovian magnetosphere. These pockets of vertical magnetic fields can be stronger than the surrounding magnetotail and magnetodisk fields. They may result from episodic reconnection of patches of the near-jovian magnetotail.
The Galileo spacecraft has since insertion on 7 December 1995, operated at Jupiter in a series of orbits whose line of apsides has rotated from a position behind the dawn terminator to close to midnight. The eighth of these orbits had sufficient telemetry bandwidth that nearly continuous measurements could be obtained through the apojove region and telemetered to Earth. This allowed us to study the temporal stability of the magnetodisk and near magnetotail with the data from the magnetometer (1) from midnight to 3 AM local time and distances from 50 to 100 jovian radii (RJ) from the planet (Fig. 1A). Based on previous observations by Pioneer 10 and 11 (2), Voyager 1 and 2 (3) and Ulysses (4), we expected the magnetospheric field to be radially directed through most of this region except near the current sheet. The tilt of the magnetic dipole axis controls the current sheet location so that the current sheet is in general displaced from the rotational equator in which Galileo orbits so that it can fly by the four Galilean moons (Fig. 1B). The rotation of Jupiter carries the current sheet back and forth across Galileo every 10 hr.
Fig. 1: Orbit of the Galileo spacecraft in the jovian magnetosphere. The orbit remains within 1o of the rotational equator throughout its trajectory.
|Fig. 1A. The eighth orbit projected in the equatorial plane. The data examined herein were obtained between 15 May and 22 June. The average location of the magnetopause in this plane is shown (11).|
|Fig. 1B. The magnetic field lines in the noon-midnight meridian at a time when the northern magnetic dipole axis is tilted away from the sun and the current sheet is below the Galileo orbit.|
The volcanic moon Io supplies up to a ton of ions per second to the jovian magnetosphere. This material is transported radially outward to ultimately be lost down the tail in the antisolar direction (5,6). In a perfectly electrically conducting fluid, the fluid elements carry the magnetic field with them. Since the magnetic flux in the jovian magnetosphere is controlled not by Io but by currents in the interior of Jupiter, this transport process cannot result in a net loss of magnetic flux by the planet. Thus, we expect that, at some distance from Jupiter in the midnight sector, the current sheet associated with the radial field configuration will pinch off, allowing the magnetic field to reconnect across a portion of the current disk and releasing an island of magnetized plasma down the magnetotail (5). The rate of plasma loss through this process on average should roughly balance the supply of plasma to the magnetodisk from Io deep in the magnetosphere. The emptied and shortened flux tubes, rooted to Jupiter, should then return to the inner magnetosphere in response to magnetic and buoyancy forces. This existing paradigm has been built principally on indirect observations. There have been no direct observations of the reconnection process and no indications as to whether reconnection is steady, or episodic as on Earth.
The magnetic measurements obtained from 60 jovian radii (RJ) to apojove and back to 60 RJ on Galileo’s eighth orbit, show an irregular square wave pattern on the radial and azimuthal components as the current sheet crosses back and forth over the spacecraft due to the 10-hr rotation of the planet (Fig. 2). Although not evident on the scale plotted here, the radial and azimuthal components are anticorrelated because the magnetic field lines are swept back out of the meridian plane, in much the same way as is the magnetic field in the solar wind. Unlike the solar wind, the sweptback jovian spiral field accelerates the material in the current sheet in the corotational direction restoring some of the angular velocity lost to the conservation of angular momentum as material moves outward. The magnetic field strength is also modulated by the rotation because the magnetic field is weaker inside the current sheet. On top of these quasi-periodic variations a general unsteadiness is seen that manifests itself in a number of ways. On time scales of several days the magnitude of the magnetic field rises and falls, sometimes gradually and sometimes abruptly. The field direction can also become predominantly unidirectional for several planetary rotations, such as seen from 1 June to 5 June. At these distances and corresponding magnetic pressures the field configuration is sensitive to the solar wind conditions because the typical dynamic pressure of the solar wind at Jupiter is about the pressure exerted by a 10 nT magnetic field. We therefore believe that the field magnitude variations could correspond to variations in solar wind dynamic pressure, and the apparent movement of the current sheet into continual residence above or below the spacecraft could be caused by north and south deviations of the solar wind flow direction.
|Fig. 2. The radial, theta, and azimuthal component of the magnetic field on orbit 8. The radial component is positive outward from Jupiter, the theta component positive southward, and the azimuthal component positive in the direction of Jupiter's rotation.|
Through this period, the theta component (perpendicular to the rotational equator and positive southward) appears to be relatively steady, averaging about 1 nT (Fig. 2). Occasionally after 26 May there are several apparent glitches in the theta component. These are not telemetry noise. Rather they are the signatures of strong negative and positive (northward and southward respectively) turnings of the field, some so strong that they more than double the background field strength. We examine in greater detail two of these events indicated by arrows in Figure 2.
The largest of these events was a southward turning of the magnetic field that occurred at 1219 UT on 17 June 1997, as the spacecraft was crossing the current sheet from north to south at a distance of 74 RJ and a local time of 0245 (Fig. 3A). The initial crossing of the current layer, as indicated by the reversal in the radial component, is brief compared to other crossings suggesting that the current layer is locally thin or moving rapidly. The initial multiple reversals in the radial component may simply be due to oscillations in the position of the sheet caused by the onset of the event. The azimuthal component reverses across the current sheet as we expect for the sweptback field geometry, but becomes stronger than expected and does not reverse its sign when the radial component does. Thus, after the initial transient behavior, the field becomes "swept-forward" out of the meridian plane. After a short dip the theta component rises abruptly and decays almost exponentially with time over the next 40 minutes. The field strength reaches 20 nT, over a factor of three greater than the immediate ambient field magnitude and more than twice the field strength observed when the spacecraft was completely out of the current sheet.
Fig. 3: The radial, vertical and azimuthal components of the magnetic field for one hour surrounding the transient event. The vertical dashed lines indicate the time of the zero crossing of the north-south component.
17 June 1997|
14 June 1997|
A smaller event occurred at 0406 UT on 14 June 1997, when Galileo was at a distance of 85 RJ at a local time of 0220 (Fig. 3B). At the onset of this event the spacecraft was not in the current sheet, nor did it cross the sheet during the event because the radial component of the field remains outward at all times. In this instance the field turns impulsively northward and remains northward for 66 min. The transient behavior in the azmuthal component 6 min after the initial transient appears to be a second but weaker "impulsive change". As with the previous event the field strength and the azimuthal field begin to change before the onset of the northward turning. The azimuthal component reverses again after the northward turning. Here the azimuthal component is negative and the radial component is positive. Thus the field returns to a swept-back configuration. We discuss this point below.
These two events are interpreted as transient reconnection in a rapidly rotating magnetized plasma. The opposite directions of turning in the two events illustrated indicate that the events occurred on either side of the null point in the center of the current sheet (Fig. 4). We believe that the onset of reconnection occurs when the current sheet becomes locally thin, analogous to the paradigm for a terrestrial substorm (7). The abrupt increase in the strength of the vertical component occurs when the magnetic field that has reconnected at some radial distance away from the location of the spacecraft is brought rapidly to the vicinity of the spacecraft by the accelerated plasma. The black hemispheres denote the relatively massive current disk in a region where thinning has not taken place (Fig. 4). As in the case of reconnection in the terrestrial magnetotail, reconnection may proceed slowly at first in the denser current sheet and become much more rapid when it reaches the low-density lobes. Activity prior to the sudden southward or northward turning is present in both events.
|Fig. 4. A slice through the current sheet in the twisted magnetic meridian showing the reconnection point and the magnetic field piled against the magnetodisk plasma both inside and outside the reconnection point. Asterisks mark the inferred locations of the spacecraft for the two events in Fig. 3a and 3b.|
The rotation of Jupiter produces important differences between the terrestrial magnetotail and the jovian magnetodisk. Angular momentum conservation imposes predictable perturbations in the azimuthal component of the magnetic field as the magnetized plasma is convected either toward or away from Jupiter by the reconnection process. The angular velocity of the plasma initially increases inward of the reconnection point producing a corotation lead. Outward from this point it decreases and the field is swept backward out of the meridian plane. This backward sweep of the field began prior to the sudden northward turning on 14 June and lasted about 4 minutes. Then the field reverts to the usual anti-correlated radial and azimuthal fields characteristic of quiet times. The 17 June event has the anti-correlated radial and azimuthal fields for the entire period of enhanced vertical field.
On short time scales when the reconnection rate is high, the inertia of the mass of the thicker part of the current sheet represents an obstacle to the reconnected plasma. The newly reconnected magnetized plasma piles up against the current sheet and creates a thick region of plasma surrounding it. Evidence for the thickening can be seen in the noise and weakened radial field in the smaller event. The noise and weakened radial field here are signs of hot plasma, far from the center of the current sheet (~7 RJ) as estimated from a model (8).
The vertical transients in the magnetic field (Fig. 2) occurred throughout June 1997 when Galileo was at local times greater than 0040 and radial distances 50 RJ, but few were as strong as our two examples. Similar behavior is seen in the shorter segments of data on other Galileo orbits in the region post-midnight, prior to 3 AM local time and greater than 50 RJ from the planet. These orbits also contain energetic particle bursts possibly associated with a reconfiguration of the magnetic field (9). Voyager 2 was the only previous mission to pass through this region and did not observe such magnetic events (3). The absence of similar magnetic signatures in the Voyager 2 data may relate to its greater radial velocity so that Voyager spent much less time in the active region.
It is difficult with a single satellite to determine how large is the affected region, especially in the radial direction. The disturbed vertical field lasts about 30 min during which time the planet has rotated about 20o. The arc of rotating plasma that moves past Galileo in 30 min at these distances is about 25 RJ if the plasma is nearly corotating. We would expect that the radial extent might be similar. Thus these disturbances appear to be large but not global events and may represent only a fraction of the reconnection taking place in the magnetodisk (10). The rapidity of the onset is not surprising given that the reconnection should accelerate the plasma to the order of the Alfven velocity that may be higher than 5000 km/sec in the tail lobes. Thus a radial displacement of 12.5 RJ might occur in a time as short as 3 min, not inconsistent with the onsets of our two observed events.
1. M. G. Kivelson, K. K Khurana, J. D Means, C. T. Russell, R. C Snare, Space Sci. Rev., 60, 357 (1992).
2. E. J Smith, L. Davis, Jr., D. E Jones, in Jupiter, ed. T. Gehrels, pp.788-920, U. Arizona Press, Tucson (1979).
3. M. Acéna, K. W. Behannon, J. E. P. Connerney, in Physics of the Jovian Magnetosphere pp. 1-50, Cambridge U. Press, London (1983).
4. A. Balogh, M. K. Dougherty, R. J. Forsyth, D. J. Southwood, B. T. Tsurutani, N. Murphy, Burton, M. E., Science, 257, 1515 (1992).
5. V. M. Vasyliunas, in Physics of the Jovian Magnetosphere, pp.395-453, Cambridge U. Press London (1983).
6. T. W. Hill, A. J. Dessler, C. K. Goertz, in Physics of the Jovian Magnetosphere, pp.353-394, Cambridge U. Press, London (1983).
7. C. T. Russell and R. M. McPherron, Space Sci. Rev., 15, 205 (1973).
8. K. K. Khurana, J. Geophys. Res., 102, 11295 (1987).
9. N. Krupp, A. Lagg, J. Woch, B. Wilken, S. Livi, D. J. Williams, Eos, Trans., AGU 78(46), Fall Meeting Suppl. F470 (1997).
10. Similar to the two neutral points in the Earth’s tail, one close and transitory and the other distant and more permanent, there may exist a more steady-state neutral line at a greater radial distance in the jovian tail (5).
11. D. E. Huddleston, C.T. Russell, M. G. Kivelson, K. K. Khurana, L. Bennett, J. Geophys. Res., in press (1998).
12. This work was supported by NASA through the Jet Propulsion Laboratory.