SPECIALSECTION evidence for plate tectonics? Where are the References and Notes E. D. Miner, M. S. Matthews, Eds. (Univ. of Arizona Press, plumes_ vents, where are the plumes_ sources 1. C. C. Porco et al., Science 311, 1393 (2006). Tucson, 1991), p. 410. 18. J. S. Kargel, S. Pozio, Icarus 119, 385 (1996). in the interior, and how are they driven? Does 2. C. J. Hansen et al., Science 311, 1422 (2006). 3. F. Spahn et al., Science 311, 1416 (2006). 19. B. A. Smith et al., Science 215, 504 (1982). Enceladus host hydrated salts or acids, such as 4. J. H. Waite Jr. et al., Science 311, 1419 (2006). 20. J. B. Plescia, J. M. Boyce, Nature 301, 666 (1983). those found on Europa (15, 16)? Where is the 5. J. R. Spencer et al., Science 311, 1401 (2006). 21. S. W. Squyres, R. T. Reynolds, P. M. Cassen, Icarus 53, 319 ammonia, or is it truly absent? Is there a briny 6. R. H. Brown et al., Science 311, 1425 (2006). (1983). 22. R. G. Ross, J. S. Kargel, in Solar System Ices, C. de Bergh, _ 7. M. K. Dougherty, Science 311, 1406 (2006). ocean, such as Europa s and other Galilean M. Festou, B. Schmitt, Eds. (Kluwer, Dordrecht, 8. G. H. Jones et al., Science 311, 1412 (2006). satellites_ (13–15)? Netherlands, 1998), p. 33. 9. R. L. Tokar et al., Science 311, 1409 (2006). If a wet domain exists at the bottom of 23. O. Prieto, J. S. Kargel, Lunar Planet. Sci. XXXIII, abstract _ 10. M. G. Kivelson, Science 311, 1391 (2006). 1726 (2002). Enceladus icy crust, like a miniature Europan 11. J.-P. Lebreton et al., Nature 438, 758 (2005). ocean, Cassini may help to confirm it. Might it be 24. J. S. Lewis, Icarus 16, 241 (1972). 12. S. K. Croft et al.,inNeptune and Triton, 25. D. J. Stevenson, Nature 298, 142 (1982). a habitat? Cassini cannot answer this question. D. P. Cruikshank, Ed. (Univ. of Arizona Press, Tucson, 26. J. S. Kargel, Icarus 100, 556 (1992). Any life that existed could not be luxuriant and 1996), p. 879. 27. O. Prieto-Ballesteros, J. S. Kargel, Lunar Planet. Sci. Conf. would have to deal with low temperatures, feeble 13. M. G. Kivelson et al., Science 289, 1340 (2000). XXXVII, Abstract 1971 (2006). 14. R. T. Pappalardo et al., J. Geophys. Res. 104, 24105 metabolic energy, and perhaps a severe chemical 28. J. I. Lunine, D. J. Stevenson, Icarus 70, 61 (1987). (1999). 29. R. Irion, Science 311, 589 (2006). environment (16). Neverthess, we cannot dis- 15. T. B. McCord et al., Science 280, 1242 (1998). 30. Fig. 1 source images: See http://saturn.jpl.nasa.gov/home/. count the possibility that Enceladus might be 16. J. S. Kargel et al., Icarus 148, 226 (2000). life_s distant outpost. 17. R. Greenberg et al.,inUranus, J. T. Bergstralh, 10.1126/science.1124495

natural assumption that the moon is roughly PERSPECTIVE spherical, one can predict the form of the mag- netic perturbations along flyby trajectories in Does Enceladus Govern different parts of the interaction region. The first flyby (17 February 2005) occurred È1259 km north of the moon, and indeed, the observed Magnetospheric Dynamics at Saturn? magnetic perturbations were consistent with the Margaret Galland Kivelson sling-shot analogy. On the closer (È497 km) flyby of 9 March Instruments on the Cassini spacecraft reveal that a heat source within Saturn’s moon Enceladus 2005, however, the field bent in a direction dif- powers a great plume of water ice particles and dust grains, a geyser that jets outward from the ferent from that anticipated for a compact quasi- south polar regions and most likely serves as the dominant source of Saturn’s E ring. The spherical moon; as a result, the magnetometer interaction of flowing magnetospheric plasma with the plume modifies the particle and field team concluded that an extended atmosphere environment of Enceladus. The structure of Saturn’s magnetosphere, the extended region of space must be present, localized near the south pole. To threaded by magnetic-field lines linked to the planet, is shaped by the ion source at Enceladus, and test this inference, the altitude of the third flyby magnetospheric dynamics may be affected by the rate at which fresh ions are created. (14 July 2005) was decreased to È168 km. This optimized pass enabled the remote sensing in- arly in 2005, the Cassini spacecraft passed of neutral O and OH present in Saturn_sinner struments (1, 3, 5, 7, 9) to detect not an atmo- Saturn_s moon Enceladus, providing a magnetosphere (11, 12). sphere but a rather narrow plume jetting water Ewealth of new data on this curious sat- In an accompanying Perspective (13), Kargel vapor and dust particles (6) above the south ellite, as the mission teams report in this issue discusses implications of the observations for our polar regions. The plume originates in the region (1–9). A tiny moon (diameter È500 km) with understanding of the interior and physical proper- of heated and distinctively colored surface linea- an exceptionally bright icy surface, Enceladus ties of Enceladus. The newest discoveries also ments referred to as Btiger stripes[ and is the orbits Saturn at a distance of 4.89 RS (where provide critical insight into the processes that source of a localized cloud of ions and electrons 0 _ RS is a Saturn radius 60,268 km). A textbook drive Saturn s magnetospheric dynamics. (4, 8) whose effects on the magnetic field forecast on planetary sciences (10) published shortly In particular, the plume ejecta affect local its presence. Slowing and diversion of the plas- before Cassini_s arrival in the Saturn system properties of the magnetospheric plasma, such as ma flow began at a distance of 27 Enceladus radii B [ describes it as remarkable and enigmatic, mass density and flow patterns. A magnetized (RE) from the moon, consistent with a widely conjectures that its interior may be partially plasma flowing toward an electrically conducting distributed source of neutral material (8). liquid, and considers that it may be the source moon or a moon surrounded by a cloud of ions The particulates that maintain Saturn_sE-ring of particles forming the E ring. Cassini_s will be slowed and diverted (14, 15). The local properties now can be fully identified. The tra- recent discoveries of warm surface features magnetospheric magnetic field, in turn, responds jectories and fluxes of dust particles in the plume near the south pole (7) and of an extended as if it were frozen into the flowing plasma; the appear adequate to maintain the E ring (6). The plume of water ice particles and dust (5, 6) field strength and plasma density increase in the ice particles and water-group atoms and mol- provide insight into the enigma and begin to regions of slowed flow. Much of the plasma ecules supplement an E-ring source, and through resolve the question of the source of the bright avoids encounter with the moon; field lines charge exchange and impact ionization they con- material on the surface, the E ring, and the torus bow out in response to the diversion of the tinuously supply heavy ions to the inner mag- flow. Because the plasma slows first near the netosphere. Tokar et al.(8)estimatearateof moon and only with some delay at locations on mass loading (Q100 kg/s) compatible with earlier field lines far above and below the moon, the estimates based on Hubble Space Telescope Institute of Geophysics and Planetary Physics and Depart- ment of Earth and Space Sciences, University of California, interaction bends those field lines as if they observations (16). Los Angeles, CA 90095–1567, USA. E-mail: mkivelson@ were the strings of sling shots draped around It is also clear that not only Enceladus but igpp.ucla.edu projectiles. With this picture in mind and the also its plume absorb energetic particles that drift

www.sciencemag.org SCIENCE VOL 311 10 MARCH 2006 1391 CASSINI AT ENCELADUS

of the internal magnetic field of Saturn. Aspects of this picture relate to the magnetic anomaly model introduced to describe some Jovian mag- netospheric processes (30). Variations of ionospheric conductance at fixed latitude could arise from high-order multipoles of the internal field (undetectable at spacecraft altitudes) or from nonuniform atmospheric struc- ture. A complete account of the process would have to explain the source of periodicity and also the variability of the period of kilometric radio emissions (labeled as SKR in Fig. 1) over decades (31, 32). Although many of these specific ideas have not yet been tested, we can be sure that the dynamics of Saturn_s magnetosphere are domi- nated by responses to plasma introduced by Enceladus.

References and Notes 1. R. H. Brown et al., Science 311, 1425 (2006). 2. M. K. Dougherty et al., Science 311, 1406 (2006). 3. C. J. Hansen et al., Science 311, 1422 (2006). 4. G. H. Jones et al., Science 311, 1412 (2006). 5. C. C. Porco et al., Science 311, 1393 (2006). 6. F. Spahn et al., Science 311, 1416 (2006). Fig. 1. Schematic of Saturn’s magnetosphere showing the plume at Enceladus as the source of plasma 7. J. R. Spencer et al., Science 311, 1401 (2006). and outflowing plasma on the night side in a region linked magnetically to a region of lower-than- 8. R. L. Tokar et al., Science 311, 1409 (2006). average ionospheric conductance. Red arrows indicate kilometric radiation emitted from Saturn’s polar 9. J. H. Waite Jr. et al., Science 311, 1419 (2006). region (SKR). Diagram is not to scale. 10. I. dePater, J. J. Lissauer, Planetary Sciences (Cambridge Univ. Press, Cambridge, 2001). 11. D. E. Shemansky et al., Nature 363, 329 (1993). 12. A. Eviatar, J. D. Richardson, Ann. Geophys. 10, 511 (1992). toward them. Changes in absorption signatures of Currents carried in the plasma establish the 13. J. S. Kargel, Science 311, 1389 (2006). energetic electrons from orbit to orbit suggest that magnetic structure of the middle and outer mag- 14. F. M. Neubauer, J. Geophys. Res. 85, 1171 (1980). 15. D. J. Southwood, M. G. Kivelson, R. J. Walker, J. A. Slavin, the outgassing rate of the plume varies on time netosphere. Beyond roughly 10 RS,thethermal scales of days or weeks (4). energy density (ºp) becomes greater than the J. Geophys. Res. 85, 5959 (1980). º 2 16. S. Jurac et al., Geophys. Res. Lett. 29, 2172 (2002). With the water ion source in the inner mag- magnetic energy density ( B /m0), where B is 17. T. W. Hill, Planet. Space Sci. 24, 1151 (1976). netosphere better understood, we can think about the field magnitude, m0 is the permeability of 18. D. J. Southwood et al., J. Geophys. Res. 106, 30109 (2001). how this plasma contributes to magnetospheric vacuum, and p is the thermal pressure. Under 19. M. G. Kivelson et al., Geophys. Res. Lett. 24,2127(1997). structure and dynamics (Fig. 1). Electromagnetic these conditions, plasma currents cause the 20. Leisner et al., Geophys. Res. Lett. 32, L14S08, 10.1029/ forces accelerate the newly ionized material magnetic field to bulge out radially near the 2005GL022652 (2005). 21. D. T. Young et al., Science 307, 1262 (2005). until it roughly corotates with Saturn, draining equator (referred to as ballooning) and the plasma 22. J. E. P. Connerney, M. H. Acuna, N. F. Ness, J. Geophys. angular momentum from Saturn_s ionosphere. expands outward, forming a plasma disk (22, 23). Res. 88, 8779 (1983). This rotating plasma exerts centrifugal stress on Probably the most puzzling aspect of magneto- 23. M. K. Dougherty et al., Science 307, 1266 its surroundings. Beyond 1.9 R from the spin spheric dynamics is that the field configuration 10.1126/science.1106098 (2005). S 24. S. M. Krimigis et al., Science 307, 1270 (2005). axis, the stresses related to rotation dominate the (23) and particle fluxes (24, 25) vary at the plan- 25. C. Paranicas et al., Geophys. Res. Lett. 32, L21101, gravitational force. If the integrated mass of ions etary rotation period, as does the radio emission 10.1029/2005GL023656 (2005). on magnetic flux tubes crossing the equator of in the kilometric band (26). There is, as yet, no 26. M. L. Kaiser et al.,inSaturn, T. Gehrels, Saturn near the orbit of Enceladus exceeds the consensus on how periodicity is imposed by the M. Shapley-Matthews, Eds. (Arizona Press, Tucson, AZ, 1984), pp. 378–415. integrated mass of ions on flux tubes crossing rotation of a nearly axially dipole field (27). The 27. M. H. Acuna, J. E. P. Connerney, N. F. Ness, J. Geophys. further out, the system is unstable to flux tube mechanism proposed by Espinosa et al.(28)re- Res. 88, 8771 (1983). interchange—a process in which mass-loaded quires a Bcamshaft[—some anomaly in the inner 28. S. A. Espinosa, M. K. Dougherty, D. J. Southwood, flux tubes change places with less heavily loaded magnetosphere that launches outward-moving J. Geophys. Res. 108, A2, 10.1029/2001JA005084 (2003). 29. D. J. Southwood, M. G. Kivelson, paper presented at the flux tubes (17). Mass is thus transported outward, pulsed perturbations. (A camshaft is a structure American Geophysical Union Fall Meeting, San Francisco, possibly with little effect on the magnetic con- that converts rotational motion into linear mo- CA, 5 to 9 December 2005. figuration. The process is analogous to that found tion.) The high level of symmetry of Saturn_s 30. T. W. Hill, A. J. Dessler, C. K. Goertz, in Physics of the in gravitationally bound atmospheres, wherein a magnetic field, however, made it difficult to Jovian Magnetosphere, A. J. Dessler, Ed. (Cambridge cold dense element of gas embedded in a hot identify the form of the anomaly. Univ. Press, New York, 1983), pp. 353–394. 31. A. Lecacheux, P. H. M. Galopeau, M. Aubier, in Planetary tenuous environment at fixed pressure falls Cassini magnetometer data are consistent with Radio Emissions IV, H. Rucker, S. Bauer, A. Lecacheux, Eds. toward the surface. The rotational interchange the camshaft model, with the modulated rate of (Austrian Academy of Science, Vienna, 1997), pp. 313–325. process, which spontaneously carries mass out- outflow of plasma from the inner magnetosphere 32. D. A. Gurnett et al., Science 307, 1255 (2005). ward, is less often described in the magneto- possibly linked to varying rates of interchange 33. I thank D. J. Southwood for sharing insights into the dynamics of the Saturn’s magnetosphere and M. K. Dougherty for spheric context, but its signatures have been and ballooning of the flux tubes linked to dif- permission to work with her team. I acknowledge support identified at Earth (18), Jupiter (19), and most ferent positions around Saturn (29). Modulated from NSF under grant ATM 02-05958. recently at Saturn (20, 21), where they are found transport would then impose periodicity on both over a large range of radial distances. plasma and field signatures despite the symmetry 10.1126/science.1124494

1392 10 MARCH 2006 VOL 311 SCIENCE www.sciencemag.org