Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Physics of the Earth and Planetary Interiors 187 (2011) 92–108 Contents lists available at ScienceDirect Physics of the Earth and Planetary Interiors journal homepage: www.elsevier.com/locate/pepi Planetary magnetic fields: Observations and models G. Schubert ⇑, K.M. Soderlund Department of Earth and Space Sciences, University of California, Los Angeles, CA 90095, USA a r t i c l e i n f o Article history: Available online 12 June 2011 Edited by Keke Zhang Keywords: Planetary magnetic fields Planetary cores Spacecraft observations Numerical dynamo models a b s t r a c t The state of knowledge and understanding of planetary magnetic fields is reviewed. All of the planets, with the possible exception of Venus, have had active dynamos at some time in their evolution. The properties and characteristics of the dynamos are as diverse as the planets themselves. Even within the subclasses of terrestrial and giant planets, the contrasting compositions, sizes, and internal pressures and temperatures of the planets result in strikingly different dynamos. As an example, the dynamos in Mercury and Ganymede are likely driven by compositional buoyancy distributions different from that in the Earth’s core. Dynamo models operate far from the parameter regimes appropriate to the real planets, yet provide insight into the dynamics of their interiors. While Boussinesq models are generally adequate for simulating terrestrial planet dynamos, anelastic models that also account for large density and electrical conductivity variations are needed to simulate the dynamos in giant planets. Future spacecraft missions to planets with active dynamos are needed to learn about the character and temporal variability of the planetary magnetic fields. Ó 2011 Elsevier B.V. All rights reserved. 1. Introduction With the possible exception of Venus, all planets in the solar system have or once had internally generated magnetic fields. The Moon also had its own magnetic field in the past, and Jupiter’s satellite Ganymede has a magnetic field at present. Undoubtedly, other bodies in the solar system have generated magnetic fields in the past, e.g., the angrite meteorite parent body (Weiss et al., 2008) and possibly some outer solar system objects other than Ganymede have a magnetic field at present. It is expected that many of the extrasolar planets now being detected have magnetic fields and some observations support this conjecture (e.g., Sanchez-Lavega, 2004; Shkolnik et al., 2005, 2008). While many planets, satellites, and small bodies generate magnetic fields by dynamo action in their interiors, not all objects with the requisite electrically conducting fluid layers have a dynamo. Prominent among such bodies in our own solar system are Venus and Io, which are likely to have completely or at least partially molten metallic cores (e.g., Konopliv and Yoder, 1996; McEwen et al., 1989; Khurana et al., 2011). The cores in Mars and the Moon are also likely partially molten at present (e.g., Yoder et al., 2003; Khan et al., 2004), yet the dynamos that once existed in these bodies have ceased to function. A planetary magnetic field, or the absence of such a field, informs us about the internal structure of a body and its thermal evo⇑ Corresponding author. Tel.: +1 310 825 4577; fax: +1 310 825 2779. E-mail addresses: [email protected] (G. Schubert), [email protected] (K.M. Soderlund). 0031-9201/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.pepi.2011.05.013 lution. Together with the gravitational field, the magnetic field provides a window into the interiors of bodies that we can only probe from a distance. The magnetic fields recorded and preserved in the crusts of planets and satellites also provide a window into the histories of the bodies. Planetary magnetism therefore merits study from many perspectives, including that of understanding the physics of how dynamos work and what they imply about the past and present properties of the body containing the dynamo. In this paper we review what is known about magnetic fields in the solar system, what attempts have been made to simulate the working dynamos, and what we learn about the host bodies from their magnetic fields. 2. Survey of planetary magnetism The planets in our solar system can be classified into three types – terrestrial, gas giant, or ice giant – based on their physical properties, and the magnetic fields show a large diversity in structure and strength. Table 1 summarizes the physical properties and magnetic field characteristics of the planets and select satellites. The mean density and moment of inertia (MOI) of a planet provide constraints on the internal composition and structure. All of the planets and many of the satellites show evidence of differentiation into layers. The terrestrial planets have an outer rocky crust, an intermediate silicate mantle, and a central iron-alloy core (e.g., Riner et al., 2008; Zharkov, 1983; Dziewonski and Anderson, 1981; Longhi et al., 1992). The gas giants, Jupiter and Saturn, have outer molecular envelopes of hydrogen and helium, fluid mantles that are largely 93 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 Table 1 Summary of physical properties and magnetic field characteristics. Overbars indicate mean values. Physical property data taken from 1Lodders and Fegley (1998), 2de Pater and Lissauer (2001), 3Schubert et al. (2007) for the MOI of the Galilean satellites, and 4Iess et al. (2010) for the MOI of Titan. Magnetic field characteristics are calculated using Eqs. (2)– (4) and data in Uno et al. (2009) for Mercury, IAGA Division V-MOD geomagnetic field modeling website for Earth, Yu et al. (2010) for Jupiter, Kivelson et al. (2002) for Ganymede, Burton et al. (2009) for Saturn, and Holme and Bloxham (1996) for the ice giants. MOI is normalized with respect to the product of the mass and the square of the radius of the body. Planet Mercury Venus Earth Moon Mars Jupiter Io Europa Ganymede Callisto Saturn Titan Uranus Neptune Mass1 Radius1 (km) Density1 (kg/m3) MOI24 (1024 kg) 0.33 4.87 5.97 0.07 0.64 1900 0.09 0.05 0.15 0.11 570 0.13 87 100 2440 6052 6371 1738 3390 69,911 1821 1565 2634 2403 58,232 2575 25,362 24,624 5.4 5.2 5.5 3.3 3.9 1.3 3.5 3.0 1.9 1.9 0.7 1.9 1.3 1.6 0.33 0.33 0.33 0.39 0.37 0.25 0.38 0.35 0.31 0.35 0.21 0.34 0.23 0.23 composed of metallic hydrogen and helium, and perhaps central rocky cores (Guillot, 1999a). The ice giants, Uranus and Neptune, have outer molecular envelopes of hydrogen, helium, and ices, intermediate ionic oceans of water, methane, and ammonia, and possibly central rocky cores (Guillot, 1999a). However, the layer thicknesses are not well-constrained for planets other than the Earth. Outer solar system satellites exhibit a wide variety of sizes, from tens of kilometers to larger than the planet Mercury, and have mean densities that are icy to rocky. Ganymede is the largest satellite in the solar system and also has the lowest moment of inertia of all known planetary bodies (Schubert et al., 2007). This indicates that the satellite is fully differentiated, with a thick water-ice shell overlying an intermediate rocky mantle and a central metallic core (Schubert et al., 2007). Thus, Ganymede has a terrestrial-style internal structure beneath the ice shell. Magnetic fields originate in the electrically conducting fluid regions of planets and satellites and are likely driven by dynamo action (e.g., Elsasser, 1939; Bullard, 1949; Stevenson, 2003). These regions are expected to be unstable to thermochemical convection in order to remove the heat of formation and radioactive decay (Stevenson, 2010). Convection of an electrically conducting fluid drives electrical currents and generates a magnetic field; the process whereby kinetic energy is converted into magnetic energy is known as dynamo action. While convection is the most probable dynamo driver, it has been argued that precession, libration, nutation, and tidal interactions may also promote dynamo action (Malkus, 1994; Tilgner, 2007; Dwyer et al., 2010). Intrinsic planetary magnetic fields diffuse as a potential field through the overlying electrically insulating or less electrically conducting regions. This field can be expressed as B = rU, where the magnetic potential U can be decomposed into spherical harmonics: ( lþ1 1 X l X m RP Uðr; h; /Þ ¼ RP g l cosðm/Þ r l¼1 m¼0 m þhl sinðm/Þ Pm ðcoshÞ l ð1Þ where r is radius, h is colatitude, / is longitude, RP is the planet radius, l is spherical harmonic degree, m is spherical harmonic order, Pm l are associated Legendre polynomials with partial Schmidt normalization (see Appendix B of Merrill et al. (1996)), and g m l and m hl are the gauss coefficients. The gauss coefficients are used to describe planetary magnetic fields. The radial magnetic field for a given radius r is obtained through Br ðrÞ ¼ rU ^r ¼ Surface jBr j (lT) Dipolarity 0.30 – 38 [100 [0.1 550 – – 0.91 – 28 – 32 27 0.71 – 0.61 – – 0.61 – – 0.95 – 0.85 – 0.42 0.31 lmax X l X ( ðl þ 1Þ l¼1 m¼0 m þhl sinðm/Þ Pm l ðcoshÞ Dipole tilt (°) 3 – 10 – – 9 – – 4 – <0.5 – 59 45 lþ2 m RP g l cosðm/Þ r ð2Þ where lmax is the maximum harmonic degree of the observations. The two radial levels most frequently considered are the planet surface, r = RP, and the top of the dynamo-generating region, r = RD. The strength of the dipole relative to the total field strength is the dipolarity, P1 m m jg 1 j þ jh1 j : f ¼ Plmaxm¼0 Pl m m m¼0 jg l j þ jhl j l¼1 ð3Þ Perfectly-dipolar magnetic fields have f = 1. The fields are considered to be dipole-dominated when f > 0.5. The tilt of the dipole relative to the rotation axis, hd, is given by 0 1 1 2 1 2 1=2 g þ h B C 1 1 B C hd ¼ tan1 B C: 0 @ A jg 1 j ð4Þ 2.1. Active dynamos 2.1.1. Earth Earth’s magnetic field is at least 3.5 billion years old as recorded by crustal rocks that were magnetized as they cooled (Hale and Dunlop, 1984). These paleomagnetic records indicate that the dynamo has almost always had a strong axial dipole component with a field strength that has not significantly varied over time (Hulot et al., 2010). The present field has a mean surface strength of about 40 lT and the dominant dipole component is offset 10° from the rotation axis. However, episodic reversals in field polarity have been recorded as magnetic stripes in oceanic crust over the past hundred million years (Johnson et al., 2003). During the 10,000 years over which reversals occur, the total field strength decreases and the field becomes multipolar (Bogue and Merrill, 1992). Observatories and satellites have greatly improved our understanding of the Earth’s magnetic field and its evolution over short timescales. The Magsat (1979–1980), Ørsted (1999–), CHAMP (2000–2010), and SAC-C (2000–2004) satellites have provided continuous measurements of the field over a range of altitudes for more than ten years. The SWARM satellite constellation will be launched in 2011 to continue these efforts. Fig. 1 shows the radial magnetic field at the surface and the core-mantle boundary plotted 94 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 up to spherical harmonic degree 13. Higher order harmonics of the internal field are masked by crustal remnant magnetism. The dynamic nature of the geodynamo is revealed through secular variation, where changes in the strength and spatial structure of the field occur over timescales of months (Olsen and Mandea, 2008) to centuries (Bloxham et al., 1989) to millenia (Cande and Kent, 1995). The most notable aspects of secular variation are polarity reversals, westward drift of the non-axisymmetric field, anticyclonic field motions at high latitudes, growth of the South Atlantic anomaly where the field is unusually weak, and geomagnetic jerks (Finlay et al., 2010). The physical mechanisms that lead to these behaviors, however, are not yet understood. 2.1.2. Mercury Mercury has the weakest intrinsic magnetic field in the solar system with a mean surface strength of only about 0.3 lT. This field was first measured during the Mariner 10 flybys in 1974 and 1975 (Ness et al., 1975, 1976) and again during the MESSENGER flybys in 2008 and 2009 (Anderson et al., 2008, 2010; Uno et al., 2009; McNutt et al., 2010; Solomon, 2011). These observations suggest that the field is dipole-dominated and nearly aligned with the rotation axis. The surface radial magnetic field plotted up to spherical harmonic degree three is shown in Fig. 2a. Compared to the Earth, the field is about 100 times weaker at the surface and about 500 times weaker at the core-mantle boundary, assuming Mercury’s core has a radius of 1900 km. It is not yet understood why Mercury has such a weak magnetic field. The source of the field is not agreed upon in the literature (e.g., Schubert et al., 1988; Stevenson, 2010), although dynamo action is the general consensus because libration measurements by Margot et al. (2007), gravity measurements of MESSENGER (Smith et al., 2010), and thermodynamic phase equilibrium simulations of Malavergne et al. (2010) support the presence of a liquid core. Other possibilities include remnant magnetism (Aharonson et al., 2005) or thermoelectric currents (Stevenson, 1987; Giampieri and Balogh, 2002). Further, MESSENGER found that the internal field interacts with currents in the planet’s magnetosphere (McNutt et al., 2010), but the dynamical implications of this coupling are not well constrained (Glassmeier et al., 2007; Kabin et al., 2008; Gomez-Perez and Solomon, 2010; Gomez-Perez and Wicht, 2010). 2.1.3. Ganymede The Galileo spacecraft revealed that the Jovian satellite Ganymede also has an intrinsic magnetic field (Kivelson et al., 1996). As shown in Fig. 2b, these measurements suggest that the field is nearly axially-aligned and dipole-dominated with a mean surface strength of about 1 lT. However, it might also be appropriate to consider the value of Ganymede’s magnetic field at the ice-rock boundary since the body has a terrestrial-style internal structure beneath the ice shell. Assuming an ice shell thickness of about 900 km (Schubert et al., 2007), the mean magnetic field is 3 lT at the ice-rock boundary. This is about an order of magnitude larger (smaller) than Mercury’s (Earth’s) surface field. The field resolution, however, is limited to the dipole and quadrupole components and additional data are needed to further characterize the magnetic field. Since Ganymede is thought to have a metallic core with a fluid component, the field is likely driven by dynamo action (Schubert et al., 1996; Sarson et al., 1997; Kivelson et al., 2002). The size of the core is poorly constrained, and the energy source for maintaining a liquid or partially liquid core in such a small body is not fully understood. 2.1.4. Jupiter The magnetic field of Jupiter was discovered when Burke and Franklin (1955) showed that the planet was a strong radio source. This was the first detection of a planetary magnetic field besides that of the Earth. The existence of the field was confirmed when the planet was visited by the Pioneer 10 flyby in 1973 (Smith et al., 1974). Additional measurements were obtained by the Pioneer 11 flyby in 1974 (Smith et al., 1975), the Voyager 1 and 2 flybys in 1979 (Ness et al., 1979), the Ulysses flybys in 1992 and 2004 (Balogh et al., 1992), the Galileo orbiter between 1995 and 2003 (e.g., Connerney, 1993; Yu et al., 2010), and the New Horizons flyby in 2007 (e.g., McComas et al., 2007; McNutt et al., 2007). These measurements show that the magnetic field is dominated by the axial dipole component and is the strongest field in the solar system with a mean surface strength of 550 lT. The surface of gaseous planets is taken to be the 1 bar pressure level. The surface radial magnetic field including up to the octupole component is shown in Fig. 2c. The spatial and temporal resolution of the field is a significant limitation. While Jupiter’s magnetic field has been measured by several missions, the orbits do not extend inward of 1.6 Jupiter radii and have longitudinal data gaps (Russell and Dougherty, 2010). Consequently, the observations are only resolved to spherical harmonic degree three. Jupiter has not exhibited unambiguous secular variations in field strength or dipole tilt in the 20 years between the Pioneer and Galileo missions (Yu et al., 2010). (b) Earth CMB (a) Earth surface -80 µT 0 80 µT -800 µT 0 800 µT Fig. 1. Radial magnetic field of Earth at (a) the surface and (b) the core-mantle boundary (CMB). The colors represent field intensity where purple (green) indicates outward (inward) directed field. A mollweide projection is used in which the horizontal lines indicate constant latitude. IGRF-11 coefficients are used and are available on the IAGA Division V-MOD geomagnetic field modeling website. 95 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 The magnetic field is produced by convectively-driven dynamo action in the highly electrically conducting metallic hydrogen mantle and in the less electrically conducting region near the base of the molecular envelope (Stanley and Glatzmaier, 2010). The exact location and extent of the dynamo region is poorly constrained. Nellis (2000) argues that the dynamo can extend up to 95% of the planet radius. Consequently, the dynamo region is expected to exhibit significant changes in density and electrical conductivity, the effects of which are relatively unknown. The upcoming Juno mission will critically improve our understanding of Jupiter by resolving the magnetic field up to spherical harmonic degree 14, detecting short timescale secular variation, and constraining the planet’s internal structure (Connerney, private communication). 2.1.5. Saturn Saturn was revealed to have a magnetic field during the Pioneer 11 flyby in 1979 (Smith et al., 1980), and subsequent observations were made by Voyager 1 in 1980 (Ness et al., 1981) and Voyager 2 in 1981 (Ness et al., 1982). The Cassini spacecraft arrived at Saturn in 2004 and is still collecting data (Dougherty et al., 2005; Burton et al., 2009). These observations have mapped the large-scale radial magnetic field, shown in Fig. 2d, and suggest that no resolved (b) Ganymede (a) Mercury -0.6 µT 0 0.6 µT -1.5 µT (c) Jupiter -1200 µT 0 0 1.5 µT (d) Saturn 1200 µT -60 µT (e) Uranus -100 µT 0 0 60 µT (f) Neptune 100 µT -100 µT 0 100 µT Fig. 2. Radial magnetic field at the surfaces of (a) Mercury, (b) Ganymede, (c) Jupiter, (d) Saturn, (e) Uranus, and (f) Neptune. Data taken from Uno et al. (2009) for Mercury (with spectral resolution l, m 6 3), Kivelson et al. (2002) for Ganymede (l, m 6 2), Yu et al. (2010) for Jupiter (l, m 6 3), Burton et al. (2009) for Saturn (l, m 6 3), and Holme and Bloxham (1996) for the ice giants (l, m 6 3). 96 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 surface field strengths near 30 lT, comparable to that of the Earth and Saturn. These multipolar magnetic fields were a surprise to many scientists, and it is still not understood why these bodies produce remarkably different fields from the other planets. The fields are thought to be generated in the planets’ ionic oceans (Cavazzoni et al., 1999; Lee et al., 2006). However, the internal structures and dynamics of the ice giants are poorly constrained. (a) Mars 60 Latitude 30 0 -30 2.2. Extinct dynamos: Mars and the Moon -60 0 90 -200 nT 180 Longitude 270 0 0 200 nT (b) Moon 60 Latitude 30 0 -30 -60 0 90 -10 nT 180 Longitude 0 270 0 10 nT Fig. 3. Radial magnetic fields of (a) Mars and (b) the Moon. The colors represent field intensity where red (blue) indicates outward (inward) directed field. A Mercator projection is used, and circles indicate impact basins. Adapted from Langlais et al. (2010). secular variation has occurred (Burton et al., 2009). The spatial resolution is limited to spherical harmonic degree three. However, this is due to uncertainty in the planet’s rotation rate (see Gurnett et al. (2007) and Anderson and Schubert (2007)) rather than insufficient measurements (Burton et al., 2009). The field is strongly dipole-dominated with a mean surface (1 bar) strength of about 30 lT and has a remarkably small offset of less than 1°. No other observed magnetic field has such a small dipole tilt. According to Cowling’s theorem, no axisymmetric magnetic field can be generated by dynamo action, so the field must have unresolved nonaxisymmetric components. It has also been proposed that a secondary mechanism operating in the overlying regions may be responsible for axisymmetrizing the field (e.g., Stevenson, 1980, 1982, 1983). The mechanisms that lead to a strongly axisymmetric magnetic field, however, are not yet well-understood. 2.1.6. Uranus and Neptune The magnetic fields of Uranus and Neptune were discovered by the Voyager 2 flybys in 1986 and 1989, respectively (Ness et al., 1986, 1989; Holme and Bloxham, 1996). The spacecraft passed within 4.2 Uranian radii and 1.2 Neptunian radii, allowing their dipole, quadrupole, and octupole components to be mapped as shown in Fig. 2e and f. The ice giants’ magnetic fields are fundamentally different from other known planetary fields since they are multipolar, with f = 0.42 for Uranus and f = 0.31 for Neptune. The dipole component is tilted about 59° and 47° from the rotation axis for Uranus and Neptune, respectively. Both planets have mean Mars Global Surveyor (MGS) showed that Mars does not presently have an active magnetic field, yet discovered magnetic anomalies associated with the magnetization of crustal rocks due to an ancient dynamo (e.g., Ness et al., 1999; Acuña et al., 1998, 1999; Connerney et al., 1999, 2001; Purucker et al., 2000; Langlais et al., 2004; Mitchell et al., 2007; Lillis et al., 2004, 2008a). Similarly, the Moon does not presently have an active magnetic field, but likely had one in the past. Crustal magnetization of lunar rocks was first detected during the era of Apollo exploration (Coleman et al., 1972), and has since been mapped in detail by Lunar Prospector in 1998 and 1999 (Binder, 1998; Purucker, 2008). The internal structures of these bodies are consistent with the inferred dynamo mechanism since there is strong evidence that both objects have metallic cores. The exact radii and densities of the cores are not well-known, but are constrained by the measured values of the moment of inertia of both objects (e.g., Folkner et al., 1997; Konopliv et al., 1998). The physical states of the cores are also largely unknown, although there is evidence that both cores are at least partially molten (Yoder et al., 2003; Khan et al., 2004). The crustal magnetic fields of Mars and the Moon are mapped in Fig. 3. The Martian crustal field is essentially confined to the southern hemisphere of the planet and is further concentrated into a relatively small region of the southern highlands between about 120° and 240° East longitude (Terra Cimmeria and Terra Sirenum). Notably, many major impact basins lack magnetic signatures. Intensities of local flux patches at spacecraft altitudes tend to be one to two orders of magnitude larger than the Earth’s crustal magnetization of about 20 nT (Langlais et al., 2010). The lunar crustal field is distributed unevenly across the surface with flux patches that tend to be at least an order of magnitude weaker than those of Mars (Langlais et al., 2010). The timing of the onset and cessation of dynamo activity in Mars and the Moon is debated. For Mars, it has been argued that the dynamo was active in the early Noachian (first Martian geologic period) and ceased to operate prior to the formation of the Hellas impact basin about four billion years ago (Acuña et al., 1999; Acuña et al., 2001). However, Schubert et al. (2000) have proposed that the onset of dynamo activity could have postdated the formation of Hellas with continued operation into the Hesperian or Amazonian periods (see also Milbury and Schubert (2010)). Still another possibility is that the Martian dynamo was active for part of the Noachian prior to the Hellas impact and was re-activated after the Hellas event (Lillis et al., 2005). There are similar uncertainties for the lunar dynamo. Paleointensity data from Apollo samples imply that the dynamo was long-lived and turned off about four Gyr ago (Garrick-Bethell et al., 2009). This early dynamo scenario is supported by modeling efforts. Stegman et al. (2003) propose that changes in mantle convection early in the Moon’s history enhanced core heat flux enough to power a dynamo, although short-lived. In addition, Takahashi and Tsunakawa (2009) argue that a heterogenous core-mantle boundary heat flux due to hemispherically-asymmetric mantle convection can also power an early lunar dynamo. Alternatively, Dwyer et al. (2010) have suggested an early nutation-driven dynamo, while Russell et al. (1977) have proposed that the Moon’s G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 dynamo turned on only when the body cooled sufficiently for a solid inner core to freeze out. Why then did the dynamos in Mars and the Moon turn off? For both bodies, one of two scenarios seems most likely. Either the cores of these bodies cooled to the point where thermal convection and dynamo action were no longer sustainable or they cooled to the point of growing such large solid inner cores that dynamo action could no longer take place in the surviving liquid outer core. 2.3. No dynamos 2.3.1. Venus Measurements by the Mariner spacecraft in the 1960s and Pioneer Venus Orbiter in 1978 showed that Venus does not have an intrinsic magnetic field (Bridge et al., 1969; Russell et al., 1980; Phillips and Russell, 1987). Why not? The planet likely has a molten metallic core similar in size to Earth’s core (Zhang and Zhang, 1995). Two plausible explanations have been discussed. One possibility is that Venus has not cooled sufficiently for an inner core to have yet formed but enough that a purely thermally-driven dynamo cannot operate. This is consistent with Venus’ slightly smaller size compared with Earth (Stevenson et al., 1983). It also attributes a major role to compositional buoyancy release upon inner core solidification in driving dynamo action. A second explanation is that Venus has no plate tectonics and, therefore, is inefficient in cooling its core (Nimmo and Stevenson, 2000). A small heat flow out of Venus’ core could mean that the core is not convective, a state without the possibility of dynamo action. Venus may have had a dynamo in the past, perhaps if a sudden mantle overturn resurfaced the planet and rapidly cooled the core as has been proposed for the Moon (Stegman et al., 2003). Venus’ high surface temperature makes it unlikely (though not impossible) for crustal magnetization to have preserved the signature of a past dynamo. Absence of plate tectonics on Mercury and Ganymede, however, has not inhibited these bodies from generating magnetic fields in their cores. 2.3.2. Io Io, a moon of Jupiter about the size of Earth’s Moon, does not have a magnetic field (Jia et al., 2010). Yet it has a large metallic core (Schubert et al., 2007). This is perhaps surprising in that its close neighbor, Ganymede, has a magnetic field and a virtually identical internal structure below its thick ice shell. What accounts for the lack of dynamo activity in Io’s core? The answer appears to be that the satellite is too hot. There is so much heat produced in the mantle by tidal dissipation that it cannot cool its core. With no core cooling, there is no core thermal convection and no dynamo activity (Wienbruch and Spohn, 1995; Khurana et al., 2011). This is basically the same explanation as offered above for why Venus has no magnetic field, though the reason for inefficient cooling of the core is different in the two cases. 2.3.3. Callisto and Titan Callisto and Titan, satellites of Jupiter and Saturn, respectively, might be expected to have magnetic fields based on their comparable sizes to Ganymede. However, they do not (Jia et al., 2010). The lack of a magnetic field is easy to understand in the case of these bodies because they are only partially differentiated, rock from ice, and do not have metallic cores (Schubert et al., 2004; Iess et al., 2010). 97 among the fields. Given the diverse spectrum, the fundamental goal is to determine what controls the strength, morphology, and evolution of the magnetic fields. Since dynamos are intimately linked to convective fluid motions driven by thermal and compositional gradients, the coupling between magnetic fields, fluid flow, and heat/mass transfer must also be understood. Further, it is important to understand why some large terrestrial bodies do not have, and may have not had, intrinsic magnetic fields. This raises the question of whether a purely thermally-driven dynamo can exist or if compositional buoyancy accompanying inner core freeze-out is needed. For example, Earth’s dynamo is at least partially driven by compositional buoyancy associated with growth of the solid inner core. Are the dynamos of Mercury or Ganymede driven only by thermal gradients? Though this seems unlikely, we will probably not be able to observationally confirm or deny the existence of solid inner cores in these bodies for the foreseeable future. 3. Fundamentals of dynamo theory 3.1. Basic equations The governing equations of dynamo theory are the induction equation, conservation of mass, conservation of momentum, conservation of energy, and the equation of state. The unknowns are the magnetic field B, the velocity field u, and three thermodynamic variables, e.g., pressure p, temperature T, and density q. There are nine equations for these nine variables. The induction equation is @B ¼ r ðu BÞ r ðgr BÞ @t ð5Þ where g = 1/lor is the magnetic diffusivity, lo = 4p 107 H/m is the permeability of free space, and r is the electrical conductivity. The conservation of mass is @q þ r ðquÞ ¼ 0; @t ð6Þ and the conservation of momentum is q Du 1 þ 2qX u ¼ rp qX ðX rÞ þ ðr BÞ B Dt lo 1 þ qg þ qmrðr uÞ þ qmr2 u; 3 ð7Þ where X is the rotation rate of the reference frame, r is the position vector, g is the acceleration of gravity, and m is the kinematic viscosity. In (7), terms from left to right are inertial acceleration, Coriolis acceleration, pressure gradient, centrifugal force, Lorentz force, buoyancy, and two viscous diffusion terms where the Stokes assumption to neglect the bulk viscosity has been made. The conservation of energy equation is DT aT Dp U þ r ðkrTÞ ¼ ; Dt qC p Dt qC p ð8Þ where a is the thermal expansivity, Cp is the specific heat at constant pressure, k is the thermal conductivity, and U is the dissipation function including both viscous and ohmic heating. The second term on the left of (8) represents heating (cooling) due to adiabatic compression (expansion). The equation of state relates density to the temperature and/or pressure. 2.4. Comparative planetology While there are outstanding questions about each planetary magnetic field, it is also valuable to consider the comparative approach which attempts to explain the similarities and differences 3.1.1. Boussinesq approximation In most geodynamo studies, Eqs. (5)–(8) are simplified using the Boussinesq approximation. Here, the material properties are assumed to be constant and the density is treated as constant except 98 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 in the buoyancy force of the momentum equation. The induction equation simplifies to @B ¼ r ðu BÞ þ go r2 B; @t ð9Þ the conservation of mass simplifies to r u ¼ 0; ð10Þ and the conservation of momentum simplifies to Du 1 þ 2X u ¼ rP þ ðr BÞ B ao T 0 g þ mo r2 u: Dt lo q o ð11Þ The buoyancy force on the right side of (11) arises from density differences with respect to the density of the motionless background state (usually assumed to be constant density qo). These perturbations, q0 , are produced by temperature variations, T0 , from the background state temperature profile. The equation of state is then given by q = qo + q0 = qo(1 aT0 ), and the gravity term becomes qg = qog qoaT0 g. The mean gravitational component is hydrostatic and can be absorbed into the pressure gradient term. Since the centrifugal force can be written as the gradient of a potential, this term is also incorporated into the pressure gradient. As a result, P represents the effective pressure relative to the hydrostatic pressure of a motionless background state, i.e., P is associated with the velocity field. The energy equation also simplifies in the Boussinesq approximation. Heating associated with adiabatic compression and due to viscous and ohmic dissipation are assumed negligible (Hewitt et al., 1975), and (8) becomes DT ¼ jr2 T Dt been studied using the Boussinesq approximation with some success (e.g., Heimpel et al., 2005a; Stanley and Bloxham, 2006). 3.1.2. Anelastic approximation The ratio of dynamo region thickness to density scale height is not small in the outer planets; consequently, the Boussinesq approximation is not valid. Instead, the anelastic approximation should be adopted. This simplification involves replacing the continuity equation with r ðquÞ ¼ 0; ð13Þ where q is the basic density state. Here and below, the overbar refers to the basic state which can vary with radius. Approximating (6) by (13) essentially eliminates sound waves from the problem. In this approximation, pressure variations from the basic state also contribute to buoyancy in the momentum equation and adiabatic, viscous, and ohmic heating must be accounted for in the energy equation: Du 1 þ 2X u ¼ rP þ ðr BÞ B þ vp0 aT 0 g þ mr2 u; ð14Þ Dt lo q DT a U ðu gÞT 0 ¼ jr2 T þ : ð15Þ Dt C p qC p Here v is the isothermal compressibility and p0 is the pressure variation from the basic state. The induction equation is given by (5) with g. For simplicity, a polytropic equation of state is often adopted (e.g., Glatzmaier and Gilman, 1981; Clune et al., 1999; Jones and Kuzanyan, 2009). ð12Þ where j is the thermal diffusivity. The validity of the Boussinesq approximation requires that the relative density perturbations produced by temperature and pressure variations be small compared with unity. The thermally induced relative density perturbations are proportional to aT0 , while the smallness of the pressure induced relative density perturbations requires that flow speeds be much less than the speed of sound. These requirements are likely satisfied in planetary cores, which are expected to be nearly-adiabatic with subsonic flows (e.g., Smylie and Rochester, 1980; Guillot, 2005). In addition, this approximation further requires that the thickness of the convecting region be small compared with the density scale height. Schubert et al. (2001) give a more detailed analysis of the Boussinesq approximation and its applicability. The Boussinesq approximation is likely reasonable for the cores of the terrestrial planets and moons of the outer solar system, but it is probably not adequate for the giant planets. In the Earth’s core, the density increases by 23% across the fluid layer (Boehler, 1996), and the layer thickness corresponds to 0.2 density scale heights. The number of density scale heights in a layer is given by Nq = ln(qbot/qtop), where qbot (qtop) is the density at the bottom (top) of the region (e.g., Evonuk, 2008). Since the Earth is likely more strongly stratified than the other terrestrial bodies, the Boussinesq approximation is also practical for the cores of terrestrial planets and moons of the outer solar system. However, density increases rapidly with depth in the outer solar system planets. The gas giant internal structure model of Guillot (1999b) predicts the density to increase by a factor of about 100 between the 1 kbar and 10 Mbar levels; this region then extends about 5 density scale heights. The ice giant internal structure models of Hubbard et al. (1995) predict their convecting regions to extend at least one density scale height. Because of the significant density stratification in the outer solar system planets, it is likely necessary to adopt the anelastic approximation for these bodies. However, we note that deep convection and dynamo models of the giant planets have 3.1.3. Boundary conditions The dynamo equations are usually solved in the spherical shell geometry appropriate to planets. Velocity boundary conditions are impenetrable and either stress-free or no-slip surfaces. No-slip boundaries are relevant to the liquid cores in terrestrial planets, while stress free boundaries may be more relevant to dynamos in the outer planets. Temperature boundary conditions are either isothermal or insulating surfaces. The magnetic field must satisfy the continuity of the normal component of B and the continuity of the tangential component of B on the spherical shell boundaries. The latter condition assumes no surface currents and no change in permeability across the boundaries. The magnetic field outside the dynamo region must therefore be determined or specified consistently with the field in the dynamo region. 3.2. Non-dimensional parameters Non-dimensional parameters are meaningful ways to simply represent the dynamics and geometry of a system. Table 2 summarizes the parameters often used in the planetary dynamo community. The relative influences of the terms in the governing equations can be evaluated by taking their ratios and assuming typical scales for length D, velocity U, and magnetic field strength B. For example, the Ekman number, E, estimates the ratio of the viscous to the Coriolis forces: E = (mr2u)/(2X u) (mU/D2)/ (XU) = m/XD2. The Rayleigh number, Ra = agDT D3/mj, characterizes the ratio of buoyancy to diffusion. The Prandtl numbers, Pr = m/ j and Pm = m/g, characterize the ratio of viscous to thermal and magnetic diffusivities, respectively. The Rossby number, Ro = U/XD, characterizes the ratio of inertial to Coriolis forces. The Reynolds number, Re = UD/m = Ro/E, characterizes the ratio of inertial to viscous forces, and the magnetic Reynolds number, Rm = UD/g = RePm, characterizes the ratio of magnetic induction to magnetic diffusion. The Elsasser number, K = B2/qlog X, estimates the ratio of Lorentz to Coriolis forces. The depth of the dynamo region relative to the planet radius is given by RD/RP, and the 99 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 Lee et al., 2006); these values are used for the ice giants. The characteristic rms magnetic field strengths at the tops of the dynamo regions are calculated by downward continuing the measured radial magnetic fields using (2). This provides a lower bound for the Elsasser number since the actual magnetic field strength will be stronger than the estimated value due to the unresolved field components. In addition, the Elsasser number will be larger still within the dynamo region since toroidal field components are also present. Dynamo action occurs only when the fluid velocities are fast enough that magnetic induction overcomes magnetic diffusion; thus, the magnetic Reynolds number must exceed its critical value: Rm J 102 (Christensen and Aubert, 2006). Using this criterion, we infer the minimum flow speeds in terms of the Reynolds and Rossby numbers. In the Earth’s core, secular variations suggest core speeds of U 0.5 mm/s (Bloxham and Jackson, 1991), yielding Rm 103. Assuming this value is valid for all of the planets, the flow speeds are likely at least an order of magnitude larger than the lower bound estimate. It is difficult to determine the superadiabatic heat flux and, thus, the Rayleigh number. For the Earth’s core, estimates range between 1022 < Ra < 1030 (Jones, 2000; Gubbins, 2001). Additional complications arise since compositional buoyancy should also be considered. Given this uncertainty, we do not make Ra estimates for the other planets. These parameters, given in Table 4, indicate that a range of dynamo region geometries and convective dynamics are possible. The geometries can range from a sphere for Ganymede to a thick spherical shell for Earth to possibly a thin spherical shell for Mercury. Ganymede also has the most deep-seated dynamo. This may explain the satellite’s strongly-dipolar surface magnetic field. However, recall that Ganymede consists of a water-ice shell, rocky mantle, and metallic core. In order to compare against other terrestrial bodies, we also consider the radius ratio of the top of the core to the top of the mantle. This ratio is about 0.5 assuming a mantle thickness of 900 km (Schubert et al., 2007), similar to the RD/ RP = 0.55 ratio of the Earth. Interestingly, Uranus and Neptune appear to have different dynamo region geometries. Recent thermal evolution models suggest that this difference occurs because Neptune is fully convecting, while Uranus may only be convecting in a thin layer (Fortney et al., 2011). Both of the gas giants have thick spherical shell geometries, assuming a central core is present. Further, the dynamo region of Jupiter extends up to a relatively shallow depth. This occurs because the electrical conductivity varies strongly with radius and dynamo action can also occur in the semi-conducting region overlying the metallic hydrogen layer (Liu et al., 2008). The low Ekman and Rossby numbers indicate that the Coriolis force plays a strong role in the interiors of the planets, while viscous and inertial forces are substantially weaker. The inertial force may not be negligible, however, since all of the planets are strongly turbulent. This is especially evident for the ice giants since they have the largest Reynolds numbers in the solar system. Inertial effects have also been linked to multipolar magnetic field generation Table 2 Non-dimensional parameter definitions and physical interpretations. Symbols are defined in the text. Parameter Definition Physical interpretation Ekman Number Rayleigh Number E = m/XD2 Ra = agoDT D3/mj Pr = m/j Pm = m/g Viscous/Coriolis forces Buoyancy/Diffusion Ro = U/XD Re = U D/m Rm = U D/g Inertial/Coriolis forces Inertial/Viscous forces Magnetic induction/Magnetic diffusion Lorentz/Coriolis forces Dynamo region depth Dynamo region geometry Prandtl Number Magnetic Prandtl Number Rossby Number Reynolds Number Magnetic Reynolds Number Elsasser Number K = B2/qlogX RD/RP RDI/RD Viscous/Thermal diffusivities Viscous/Magnetic diffusivities geometry of the dynamo region is given by RDI/RD. In these definitions, X is rotation rate, D = RD RDI is the thickness of the dynamo region, RD is the outer radius of the dynamo region, RDI is the inner radius of the dynamo region, m is kinematic viscosity, j is thermal diffusivity, g is magnetic diffusivity, a is thermal expansion coefficient, g is gravitational acceleration, DT is superadiabatic temperature contrast, q is density, lo is magnetic permeability of free space, U is characteristic rms velocity, and B is characteristic rms magnetic field strength. In order to estimate the non-dimensional parameters, we need to make assumptions about the dynamo regions, the fluid properties, the flow speeds, and the magnetic field strengths. This is a difficult task since many of these values are poorly constrained; the best estimates are given in Table 3. Dynamo region rotation frequencies are taken from Lodders and Fegley (1998). Note that the giant planet rotation rates are assumed to coincide with the rotation rates of their magnetic fields. The magnetic rotation rate is inferred from the periodicity of radio emissions that result from charged magnetospheric particles (e.g., Carr et al., 1981, 1983). Radii of the dynamo regions are taken from Williams et al. (2007) for Mercury, from Dziewonski and Anderson (1981) for Earth, from Nellis (2000) and Stevenson (1982) for the gas giants, from Schubert et al. (2007) for Ganymede, from Lee et al. (2006) and Podolak et al. (1991) for Uranus, and from Lee et al. (2006) and Stevenson (1982) for Neptune. The inner dynamo region radii especially have a large degree of uncertainty. Based on the estimated physical properties of the Earth’s core given in Stacey (2007), we assume m 106 m2/s, g 1 m2/s, and j 105 m2/s for all bodies with liquid iron cores. Transport properties of metallic hydrogen are estimated to be m 107 m2/s and j 106 m2/s (Stevenson et al., 1977) and g 1 m2/s (Nellis, 2000); these values are used for the gas giants. Transport properties of high-pressure water are estimated to be m 106 m2/s (Abramson, 2005), j 107 m2/s (Abramson et al., 2001), and g 102 m2/s (Cavazzoni et al., 1999; Table 3 Summary of mean properties of the planets’ dynamo regions used to calculate the non-dimensional parameters given in Table 4. RP is planet radius given in Table 1. Molecular transport properties are assumed. References are given in the text. RDI is particularly uncertain for all bodies other than Earth. Dynamo Mercury Earth Jupiter Ganymede Saturn Uranus Neptune X (s1) RD (RP) RDI (RP) q m j g (kg/m3) (m2/s) (m2/s) (m2/s) jBr(r = RD)j (lT) 1.2 106 7.3 105 1.8 104 1.0 105 1.6 104 1.0 104 1.1 104 0.75 0.55 0.95 0.3 0.5 0.8 0.8 0.5 ? 0.19 0.2 ? 0? 0.25 ? 0.6 ? 0.3 ? 8000 11,000 2000 6000 2000 2000 2000 106 106 107 106 107 106 106 105 105 106 105 106 107 107 1 1 1 1 1 102 102 0.6 270 625 32 45 75 73 100 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 Table 4 Order of magnitude estimates of key dimensionless parameters for the planets’ dynamo regions. Rm = 102 is assumed to estimate lower bounds for Re = Rm/Pm and Ro = ReE. The K values neglect contributions from the toroidal field and the unresolved poloidal components, which likely increase the estimate by an order of magnitude. Dynamo E Pr Pm Rm Ro Re K RDI/ RD RD/ RP Mercury Earth Jupiter Ganymede Saturn Uranus Neptune 1012 1015 1019 1013 1018 1016 1016 0.1 0.1 0.1 0.1 0.1 10 10 106 106 107 106 107 108 108 102 102 102 102 102 102 102 104 107 1010 105 109 106 106 108 108 109 108 109 1010 1010 105 0.1 1 103 0.01 104 104 0.6 0.35 0.2 0 0.5 0.6 0.4 0.75 0.55 0.95 0.2 0.5 0.8 0.8 in dynamo models (e.g., Christensen and Aubert, 2006; Olson and Christensen, 2006). As indicated by the Elsasser number, the Lorentz force is significant for the Earth and the gas giants, but may be less so for the other planets. It is often assumed that the planets are in magnetostrophic balance where the Lorentz and Coriolis forces are comparable (e.g., Stevenson, 2003). The low K estimates for Mercury, Ganymede, and the ice giants, however, do not support this hypothesis. Recent dynamo modeling results further dispute the prevalence of magnetostrophic balance (e.g., Christensen and Aubert, 2006; Soderlund et al., in review). The magnetic Prandtl numbers much less than unity suggest a separation of scales between the magnetic and velocity fields where the magnetic flux patches are large compared to the flow eddies. However, the dynamical implications of low viscosity and rapid rotation on convection and dynamo action are not well understood since it is difficult to study low Pm, low E systems. 3.3. Methods Numerical models and laboratory experiments are valuable tools to better understand the dynamics of planetary interiors. Both approaches, however, are unable to reach the extreme planetary parameter values due to computational and technological limitations. Fortunately, many systems exhibit asymptotic behavior at large or small values of the control parameters (e.g., Christensen and Aubert, 2006; Aurnou, 2007; King et al., 2010). Large surveys are carried out to map the behavioral regimes and to determine how the control parameters influence the dynamics. With these surveys, it is possible to develop scaling laws to quantify how the dynamics (e.g., magnetic field strength, flow speeds, heat transfer efficiency) depend on the control parameters. Researchers then attempt to develop asymptotic scaling laws which can be extrapolated to planetary conditions. For a recent review of scaling laws, see Christensen (2010). Moreover, numerical models have shown that convectively-driven dynamos in rotating spherical shells are able to generate some of the key features of the planetary magnetic fields (e.g., Stanley and Bloxham, 2004; Sakuraba and Roberts, 2009; Stanley and Glatzmaier, 2010). Recent planetary dynamo models are summarized in the next section. 4. Planetary dynamo models 4.1. Terrestrial planets and Ganymede In this section we discuss the modeling that has been carried out to understand the dynamos in planets with molten iron-alloy cores. The objects include the terrestrial planets, except for Venus, and the outer planet satellite Ganymede. There is an enormous literature on the modeling of Earth’s dynamo and with our primary focus being the magnetic fields of the other planets, we do not at- tempt to review the geodynamo literature here. Instead we focus on recent modeling attempts to explain the magnetic fields of Mercury and Ganymede. Models of the ancient Martian dynamo will also be discussed. See Kono and Roberts (2002), Glatzmaier (2002), and Wicht et al. (2009) for reviews of geodynamo models. 4.1.1. Mercury Dynamo modeling of Mercury’s magnetic field has focused on explaining why it is so weak. Mercury’s magnetic dipole moment is about 3.3 4.2 1019 Am2 (Anderson et al., 2008). Compared with Earth’s magnetic dipole moment of 7.8 1022 Am2, Mercury’s moment is only about 5 104 that of Earth. Further, Mercury’s magnetic dipole moment is about half that of Ganymede, which is 1.32 1020 Am2, even though Ganymede, stripped of its ice shell, has a radius only about 75% of Mercury’s radius. It has been proposed that a stably-stratified layer at the top of the liquid part of Mercury’s core might be responsible for weakening the external dipole field (Christensen, 2006; Stanley and Mohammadi, 2008). In the Christensen (2006) model, this weakening is due to attenuation of the internally generated field by the electrically conducting stable layer. The Stanley and Mohammadi (2008) model emphasizes the importance of thermal winds in the stable layer that induce unfavorable zonal flows throughout the liquid part of the core. In their models, the magnetic field is strongly time variable and the external dipole can be either weak or strong. Another possible explanation is the thickness of the liquid part of Mercury’s core within which the dynamo operates. Dynamo action in a thin shell can produce weak external dipole fields (Stanley et al., 2005; Takahashi and Matsushima, 2006), while Heimpel et al. (2005b) have produced single plume dynamos in a thick shell that are also consistent with Mercury’s weak magnetic field. Yet another explanation for Mercury’s weak magnetic field involves the self-interaction of its internal dynamo with the magnetosphere created around the planet by the dynamo itself, a type of feedback effect (Glassmeier et al., 2007; Gomez-Perez and Solomon, 2010; Gomez-Perez and Wicht, 2010). Mercury is effectively embedded in an external magnetic field generated by Chapman-Ferraro currents in its magnetopause. This self-generated ambient magnetic field influences the dynamics of the internal dynamo, similar to the way the Jovian magnetic field influences dynamo action in Jupiter’s moon Ganymede (Sarson et al., 1997). Chapman-Ferraro currents generate a magnetic field that enhances the magnetospheric field and tends to cancel the field outside the magnetosphere. In the dynamo region, the Chapman-Ferraro field opposes the dynamo-generated field so that the dynamo is embedded in an ambient field of opposite polarity. Glassmeier et al. (2007) use a kinematic a–X-dynamo model to show that the feedback dynamo indeed has a Mercury-type solution with a weak magnetic field. Recent self-consistent dynamo models with imposed external magnetic fields also support strong magnetospheric feedback on weak internal fields (Gomez-Perez and Solomon, 2010; Gomez-Perez and Wicht, 2010). The possibility that solid iron precipitation in Mercury’s core occurs differently than it does in Earth’s core provides another feasible explanation for Mercury’s weak magnetic field (Vilim et al., 2010). Laboratory experiments have shown that the Fe-S system behaves in a non-ideal way at temperatures and pressures likely encountered in Mercury’s core for sulfur concentrations between about 7 and 12 wt% (Chen et al., 2008). As a consequence, iron could precipitate at different places in the core depending on the sulfur concentration; an Fe snowfall could originate from the top or the mid-point of the core or from both locations. When Fe freezes out at both the top and midway through the core, there are two sources of compositional buoyancy, the heavy iron sinking from the top of the core and from below mid-depth and the light Srich fluid rising above mid-depth. Vilim et al. (2010) modeled G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 dynamo action driven by this unusual source of compositional buoyancy, a double snow layer, and found separate dynamo regions above and below the mid-depth freezing layer that produce magnetic fields of opposite sign with a consequent weakened surface field. Conditions in Ganymede’s core, as discussed below, might also lead to a distinctly non-earthlike form of Fe precipitation (Hauck et al., 2006). The generation of Mercury’s weak magnetic field is not yet understood, although there is no lack of ideas to explain it. The MESSENGER spacecraft is now in orbit around Mercury and data from three flybys are published to date (Uno et al., 2009). With careful analysis of these observations, particularly with respect to separating the externally driven effects from the internally produced magnetic field, we may be able to distinguish among the above explanations. For example, Fig. 4 compares the observed surface magnetic spectra against the spectra predicted by dynamo models. This comparison shows that the model of Christensen (2006) is in best agreement, supporting his conclusion that an electrically conducting stable layer at the top of the dynamo region may be responsible for the weak, dipole-dominated field at the planet surface. 4.1.2. Ganymede Dynamo action in Ganymede’s core was first studied by Sarson et al. (1997) who investigated how Jupiter’s magnetic field might influence magnetic field generation in the satellite. The Jovian magnetic field at Ganymede’s orbit is much weaker than the satellite’s intrinsic magnetic field and, therefore, is not expected to play a significant role in the operation of Ganymede’s dynamo. The dynamo model of Sarson et al. (1997) confirmed this, but left open the possibility that in the initial stages of dynamo activity, the Jovian magnetic field could provide a seed field that would explain the anti-alignment of Jupiter’s and Ganymede’s dipole moments. Similar to Mercury, Ganymede’s dynamo acts in a relatively low pressure, low temperature environment compared with the Earth. Solidification of Ganymede’s core and the associated compositional driving of its dynamo might then be completely different from the freezing out of Earth’s core and the compositional buoyancy driving the geodynamo. According to Hauck et al. (2006), Ganymede’s dynamo can be driven in part by compositional buoyancy released through the sinking of Fe snow formed below the core-mantle boundary or by the upward flotation of solid FeS formed in the Power / Dipole power 100 10-1 10-2 Uno et al., 2009 Heimpel et al., 2005a Takahashi and Matsushima, 2006 Christensen, 2006 10-3 Vilim et al., 2010 10-4 1 2 3 4 5 6 Spherical Harmonic Degree, l Fig. 4. Normalized surface magnetic power spectra for Mercury from Uno et al. (2009) and select dynamo models: the v = 0.15 case of Heimpel et al. (2005b), the Ra = 2 105 case of Takahashi and Matsushima (2006), case II of Christensen (2006), and model 1 of Vilim et al. (2010). 101 deep core. These alternative scenarios for the thermochemical state of the core arise because the eutectic melting temperatures and eutectic sulfur contents in the binary Fe–FeS system decrease with increasing pressure in the range of core pressures in Ganymede (14 GPa). Iron snow or flotation of solid FeS occur upon cooling of Ganymede’s core depending on whether the core composition is more or less iron-rich than eutectic. Hauck et al. (2006) did scaling calculations and concluded that for iron snow cooling scenarios, typical Rossby and magnetic Reynolds numbers are consistent with dynamo action occurring in the core and estimates of the generated magnetic field agree with the strength of Ganymede’s field. Though Ganymede is larger than Mercury, the silicate-metal part of the satellite is only about the size of Io and the Moon, and Ganymede’s core is at most about half that large (Schubert et al., 2004). It is perhaps surprising that such a small core could support dynamo action over a geologically long time. What supplies the energy to drive the dynamo? Why is the core not almost completely solidified? Bland et al. (2008) and Kimura et al. (2009) have investigated thermal history models for Ganymede with the aim of identifying scenarios that allow for present day dynamo activity. Bland et al. (2008) conclude that a present day dynamo could exist only if the sulfur mass fraction in Ganymede’s core is very low (<3%) or very high (>21%), and only if the silicate mantle could cool rapidly (e.g., a mantle with a wet olivine rheology). Kimura et al. (2009) conclude that a Ganymede dynamo could currently be operative for a sufficiently large sulfur-rich metallic core in which dynamo activity started late in the satellite’s evolution. In contrast, Hauck et al. (2006) inferred that a dynamo could exist at present for a broad range of core sulfur concentrations. Bland et al. (2008) also studied whether tidal heating during passage through an eccentricity pumping Laplace-like resonance in Ganymede’s past could enable present day dynamo action in the metallic core. They found that such an event would not facilitate the operation of a present day dynamo. Of course, Ganymede does have a dynamo at present which may indicate that its core formed relatively late in the satellite’s thermal evolution. To date, there are no published quantitative models of Ganymede’s dynamo. The studies of Hauck et al. (2006), Bland et al. (2008), and Kimura et al. (2009) use scaling arguments and qualitative criteria to determine whether dynamo action could occur in their Ganymede models. Progress in understanding how Ganymede generates its magnetic field requires at least the numerical modeling of dynamo activity in Ganymede models derived consistently from thermal-orbital evolution models of the satellite. 4.1.3. Mars Mars crustal magnetization preserves a record of its ancient dynamo. The record has been exploited to infer the nature of the Martian dynamo, but only a few attempts have been made to model dynamo action in the special circumstances of ancient Mars. As discussed in Section 2.2, one of the main questions about the Martian dynamo is the period in which it was active. When did dynamo action begin and when did it end? Why did the Martian dynamo stop working? There is an observed absence of magnetic anomalies over the major impact basins, perhaps due to impact demagnetization (Hood et al., 2003; Mohit and Arkani-Hamed, 2004). This has led many investigators to the conclusion that the Martian dynamo had an early onset and ceased before the formation of the major impact basins (Acuña et al., 1999; Frey, 2003; Arkani-Hamed, 2004). The only constraint on the timing of the dynamo comes from the magnetized Allan Hills meteorite ALH84001. This meteorite is the oldest known Martian rock, dated between 3.9 and 4.1 Ga (Weiss et al., 2002). The onset of dynamo activity in Mars, therefore, occurred around 4 Ga or earlier. Bibring et al. (2006) presented a global mineralogical map of Mars and 102 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 concluded that a rapid drop in atmospheric pressure near the end of the Noachian (first Martian geologic period) could have been triggered by a rapid decrease in dynamo field strength. It has also been suggested that the major impacts could have altered the internal temperature and convective character of Mars and may have led to the early demise of the dynamo (Frey et al., 2007; Lillis et al., 2008b). Early dynamo cessation has been challenged by Schubert et al. (2000) (see Section 2.2). More recently it has been recognized that some volcanic edifices which postdate the formation of the major impact basins on Mars are magnetized. Hood et al. (2010) demonstrated that the Apollinaris Mons construct, dated at 3.7 Ga (Werner, 2009), is the most likely source of the magnetic field anomaly centered on the volcano. Milbury and Schubert (2010) did a statistical analysis of the global magnetic field and found very small differences between Noachian and Hesperian geologic units, suggesting that the dynamo persisted into the Hesperian. Further, volcanic edifices at both Hadriaca Patera and Tyrrhenus Mons were built in the period 3.7–4.0 Ga and show signs of magnetization (Greeley and Crown, 1990; Williams et al., 2008; Werner, 2009; Robbins et al., 2011). Lillis et al. (2006) used the electron reflectometry data from Mars Global Surveyor and found a magnetic signature associated with Hadriaca Patera, an edifice located on the rim of the Hellas impact basin that postdates basin formation. Milbury et al. (2011) modeled a large number of magnetic anomalies collocated with gravity anomalies within the region surrounding Tyrrhenus Mons. They found a correlation of paleopole locations with units of different geologic age showing a progression of pole positions from initial magnetization to pole reversal through polar wander. The numerical dynamo simulations of Kuang et al. (2008) suggest possible behaviors of the Martian dynamo toward the end of its activity. Near its demise, the Martian dynamo could have reversed frequently, and it could have operated as a subcritical dynamo since the energy to sustain a dynamo is significantly less than what is required to excite it. Subcritical dynamo activity would lengthen the life of the magnetic field. In addition, it would enable an abrupt termination of dynamo activity. Once turned off, it would be difficult to reactivate the dynamo without a substantial increase of the buoyancy force in the Martian core. Glatzmaier et al. (1999) and Stanley and Glatzmaier (2010) suggest another possible termination scenario for the Martian dynamo involving a critical role of reversals. Mars’ dynamo could have ended by going through a series of dipole reversals and not completely recovering after each one. Evidence for dipole reversal in the Tyrrhenus Mons region of Mars (Milbury et al., 2011) might indicate the termination of the Martian dynamo in the Hesperian. The spatial distribution of Mars crustal magnetic field confronts us with another major puzzle (Fig. 3). Is the southern hemisphere confinement of the magnetic anomalies a consequence of post dynamo modification of the magnetized crust by endogenic processes or impacts, or did the Martian dynamo produce a field that reflects the current magnetization distribution? Since the asymmetry of the crustal magnetic field coincides with the crustal dichotomy, it is possible that whatever process created the dichotomy also destroyed the crustal magnetization in the northern hemisphere. Plausible causes of the dichotomy include endogenic removal of crust in the northern hemisphere or exogenic disruption of the northern hemisphere crust by a giant impact (Wilhelms and Squyres, 1984; Frey and Schultz, 1988; Andrews-Hanna et al., 2008; Marinova et al., 2008; Nimmo et al., 2008). An endogenic process that could have removed crust from the northern hemisphere is degree-1 mantle convection or overturn (Schubert and Lingenfelter, 1973; Lingenfelter and Schubert, 1973; Weinstein, 1995; Harder, 1998; Breuer et al., 1998; Zhong and Zuber, 2001; ElkinsTanton et al., 2003; Elkins-Tanton et al., 2005; Roberts and Zhong, 2006; Ke and Solomatov, 2006). The endogenic and exogenic mechanisms can both not only remove magnetized crust from the northern hemisphere, but can also create conditions deep in the Martian mantle that influence convection and dynamo generation in the Martian core (Watters et al., 2009). Both mechanisms, upwelling in the northern hemisphere and energy deposition by impact, imply a hotter deep mantle in the northern hemisphere of Mars and, therefore, a reduced heat flux from the core into the northern hemisphere mantle. Stanley et al. (2008) investigate the effect of this core-mantle boundary heat flux dichotomy on magnetic field generation in the core by imposing a large outward core-mantle boundary heat flux in the southern hemisphere of a dynamo model and an inward core-mantle boundary heat flux in the northern hemisphere. The heat flux pattern imposed on the outer shell boundary of the model and the resulting magnetic field upward continued to the planet surface are shown in Fig. 5. Their dynamo model generates a magnetic field concentrated in and strongest in the southern hemisphere. This largely occurs because the core-mantle heat flux boundary condition made the northern hemisphere of the core subadiabatic and focused core convection in the southern hemisphere. Thus, it is possible that the concentration of Martian magnetization to its southern hemisphere could have been the result of a peculiar southern hemisphere Martian core dynamo or the destruction of magnetized northern hemisphere crust by post-magnetization processes. While the hemispheric dynamo of Stanley et al. (2008) shows that it is possible to confine the Martian magnetic field to the southern hemisphere of the planet, it does not explain the further concentration of the observed magnetization to the Terra Cimmeria-Terra Sirenum region of the southern highlands. It is not apparent that impact demagnetization of the southern hemisphere crust can account for this localization of magnetization either. Perhaps an even more restrictive condition on heat flow at the core-mantle boundary or some azimuthal dependence of the condition could explain the observation. 4.2. Giant planets The magnetic fields of the giant planets are likely coupled to their surface zonal flows and convective heat transfer patterns. These planets have strong winds that are organized into zonal jets. The wind velocities are measured with respect to the planet’s rotation rate, assumed to correspond with the rotation rate of the deep interior and magnetic field (e.g., Russell et al., 2001; Giampieri and Dougherty, 2004). Jupiter and Saturn have strong prograde (eastward) equatorial jets and multiple smaller-scale jets at higher latitudes (e.g., Porco et al., 2003; Sanchez-Lavega et al., 2000). The high latitude jets of Jupiter alternate in direction, while those of Saturn are exclusively prograde assuming the Voyager-derived rotation rate (Desch and Kaiser, 1981) and alternate in direction assuming the planetary rotation rate estimates of Anderson and Schubert (2007) and Read et al. (2009). In contrast, the zonal winds of Uranus and Neptune are retrograde (westward) at low latitudes and prograde at high latitudes (e.g., Sromovsky et al., 2001; Hammel et al., 2005). The wind speeds, however, can vary up to 50% due to uncertainties in the rotation rates (Helled et al., 2010). It is not presently known to what depths the winds of the giant planets extend, although the Galileo probe suggested that the Jovian winds penetrate to at least the 20 bar pressure level (Atkinson et al., 1998). The latitudinal internal heat flux profiles peak in the polar regions with a minimum near the equator on the gas giants (Pirraglia, 1984) and peak in the equatorial and polar regions with minima at mid-latitudes on Neptune (Pearl and Conrath, 1991). The heat flux pattern of Uranus is not well constrained (Pearl et al., 1990). Modelers attempt to self-consistently generate these observations, to understand the underlying mechanisms, and to explain 103 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 (a) Imposed Heat Flux (b) Radial Magnetic Field -15 µT 0 µT 15 µT Fig. 5. (a) Imposed outer boundary heat flux pattern in the Stanley et al. (2008) model. Red (blue) indicates inward (outward) directed heat flux. (b) Simulated radial magnetic field upward continued to the surface of Mars. Red (blue) indicates outward (inward) directed field. the fundamental differences in behavior between the gas and ice giants. Deep convection modeling efforts are typically channelled into non-magnetic rotating convection models and into planetary dynamo models. Non-magnetic models are also important to consider because the electrically insulating regions are likely coupled to the dynamo region and the mechanisms responsible for driving zonal flows and controlling the heat transfer are more easily understood without the added complexity of the Lorentz force. The results of recent gas giant-like and ice giant-like models are summarized below. 4.2.1. Gas giants Busse (1976) first proposed that deep convection in the molecular envelopes may be responsible for the zonal winds of Jupiter. He envisioned that the alternating jets are the surface expressions of differentially-rotating cylinders aligned with the rotation axis. In this hypothesis, the fluid motions are quasigeostrophic and do not strongly vary along the direction of the rotation axis, in agreement with the Taylor–Proudman theorem (Taylor, 1917; Proudman, 1916). Gas giant-like zonal flows with multiple jets were first obtained in rapidly-rotating Boussinesq convection models by Christensen (2001, 2002). More recently, Heimpel et al. (2005a), Heimpel and Aurnou (2007), and Aurnou et al. (2008) have shown that turbulent, rapidly-rotating convection in thin spherical shells produces zonal flow and heat flux patterns that are consistent with those observed on Jupiter. The simulated zonal winds of the Heimpel et al. (2005a) model are shown in Fig. 6. The equatorial jets in these models result from Reynolds stresses due to correlations between the fluctuating cylindrical and azimuthal velocity components (e.g., Zhang, 1992). Outside the tangent cylinder, the imaginary right cylinder tangent to the inner shell at the equator, the height of the convection columns decreases with increasing cylindrical radius. As a result, fluid moving away from the rotation axis must speed up and fluid moving toward the rotation axis must slow down in order to conserve potential vorticity. This leads to spiraling convection cells in which cylindrically outward flows correlate with prograde azimuthal flows, and vice versa. These Reynolds stresses transport positive angular momentum outward and negative angular momentum inward, driving zonal flows that are prograde near the outer boundary and retrograde near the inner boundary. Inside the tangent cylinder, quasigeostrophic turbulence leads to multiple jets whose widths are determined by the Rhines scale (Rhines, 1975; Heimpel and Aurnou, 2007). This style of zonal flow inhibits equatorial heat transfer and promotes polar Fig. 6. Instantaneous azimuthal winds on the inner and outer boundaries of the Heimpel et al. (2005a) model. Red (blue) indicates prograde (retrograde) flow. heat transfer (Aurnou et al., 2008). In the equatorial plane, the convective flows are less efficient at transferring heat outward since they must compete with the strong azimuthal forcing. The convective flows in the polar regions are most efficient because the zonal flows do not inhibit radial motions. Density stratification must also be considered to accurately simulate the dynamics of the gas giants. Flow speeds will tend to decrease with increasing density and the convection will not tend to manifest as axially-invariant columns (Glatzmaier et al., 2009). Thus, the vortex-stretching mechanism of driving the zonal flow is no longer appropriate. Rather, compressional torques produce negative vorticity in the expanding, uprising plumes and positive vorticity in the contracting, sinking plumes. Since fluid moving radially outward will tend to become faster and fluid moving radially inward will tend to become slower due to density effects, the plumes tilt and correlations again develop between the fluctuating cylindrical and azimuthal motions (Glatzmaier and Gilman, 1981; Evonuk, 2008; Glatzmaier et al., 2009). Reynolds stresses drive the zonal flow by transporting positive angular momentum outward and negative angular momentum inward, resulting in equatorial zonal flows that have the same organization as models without density stratification. Jones and Kuzanyan (2009) contrast anelastic and Boussinesq convection in rotating spherical shell 104 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 models and find that the zonal flow pattern is not strongly affected by the density stratification, but the convective flow speeds are suppressed at depth in the anelastic models. Electrical conductivity stratification makes the models even more realistic since it allows coupling between the lower conductivity molecular envelope and the metallic mantle. Stanley and Glatzmaier (2010) have carried out dynamo simulations with both density stratification and radially-varying electrical conductivity. Rapidly-rotating, turbulent convection in a thin shell geometry produces zonal flows and a radial magnetic field that agree to first order with the observations of the gas giants (Fig. 7). The zonal flows are maintained by the compressional torque mechanism. The magnetic field is dipole-dominated, but has significant power in the higher order modes since the axisymmetric radial field exhibits a banded structure with latitude. Additional measurements are needed to improve the spatial resolution of the planets’ magnetic fields in order to compare against the model predictions. For Saturn, modelers must also attempt to explain the strong axisymmetry of the magnetic field. Stevenson (1980, 1982) proposed that differential rotation in a stably-stratified layer created by helium rain-out near the molecular-metallic hydrogen transition could diminish the non-axisymmetric components of the magnetic field. Recent dynamo models support this hypothesis. The kinematic dynamo models of Schubert et al. (2004) show that the morphology of the dynamo is determined by the characteristics of the zonal thermal winds in the overlying stable layer. Christensen and Wicht (2008) and Stanley (2010), respectively, show that thick and thin stably-stratified layers overlying the dynamo region develop differential rotation through thermal winds; 4.2.2. Ice giants Convection in the molecular envelopes of Uranus and Neptune has also been proposed to drive their observed zonal flows and to be dynamically coupled to their dynamo regions (Suomi et al., 1991; Aurnou et al., 2007). Further, since the ice giants are estimated to have the most turbulent fluid flows in the solar system (compare Re in Table 4), inertial effects may be important in these planets. The non-magnetic Boussinesq models of Aurnou et al. (2007), non-magnetic anelastic models of Brun and Palacios (2009), and Boussinesq dynamo models of Soderlund and Aurnou (2010) show that strongly-forced convection in rotating spherical shells, regardless of shell thickness, can drive zonal flows that qualitatively agree with the observations of Uranus and Neptune. The simulated zonal winds and radial magnetic field of the Soderlund and Aurnou (2010) model are shown in Fig. 8. In these models, the Taylor–Proudman constraint is not valid due to strong inertial effects, leading to three-dimensional convection. This convective turbulence strongly mixes the fluid and homogenizes the temperature and absolute angular momentum (Scorer, 1966; Gough and Lynden-Bell, 1968; Bretherton and Turner, 1968). Aurnou et al. (2007) find that retrograde equatorial jets and strong prograde jets at high latitudes develop in order to conserve angular momentum. Soderlund and Aurnou (2010) show that this convective regime also generates heat flux patterns and magnetic field morphologies that are consistent with observations of the ice giants. The importance of inertia for the transition from dipolar to multipolar (b) Radial Magnetic Field 90 90 60 60 30 30 Latitude Latitude (a) Zonal Flows this flow can suppress rapidly-varying, higher-order magnetic field components through a skin effect mechanism. 0 0 -30 -30 -60 -60 -90 0 200 400 -90 -3000 0 3000 Intensity, µT Velocity, m/s Fig. 7. (a) Zonal flow and (b) radial magnetic field snapshots on the outer boundary of the Stanley and Glatzmaier (2010) model. In (a), red (blue) indicates prograde (retrograde) flow. In (b), red (blue) indicates outward (inward) directed fields. The latitudinal profiles are averaged over all longitudes. (b) Radial Magnetic Field 90 90 60 60 30 30 Latitude Latitude (a) Zonal Flows 0 0 -30 -30 -60 -60 -90 -90 -500 0 Velocity, m/s 500 -103 0 Intensity, µT 103 Fig. 8. Snapshots of (a) zonal flow and (b) radial magnetic field up to harmonic three on the outer boundary of the Soderlund and Aurnou (2010) model. In (a), red (blue) indicates prograde (retrograde) flow. In (b), purple (green) indicates outward (inward) directed fields. The latitudinal profiles are averaged over all longitudes. G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 magnetic field generation has also been shown in the geodynamo models of Grote et al. (1999, 2000), Kutzner and Christensen (2000), Kutzner and Christensen (2002), Simitev and Busse (2005), Aubert (2005), Sreenivasan and Jones (2006); and Soderlund et al. (in review). Other groups have also produced ice giant-like magnetic fields. Gomez-Perez and Heimpel (2007) find that dynamo models tend to become multipolar as the electrical conductivity is decreased and as strong zonal flows develop. In addition, Stanley and Bloxham (2004), Stanley and Bloxham (2006) find multipolar dynamos when moderately strong convection occurs in a thin spherical shell overlying a stably-stratified fluid. They suggest that solid, electrically conducting inner cores act to stabilize the magnetic field and promote dipolar dynamos, while stably-stratified fluid cores do not have this anchoring effect. While this model may be appropriate for Uranus, it is inconsistent with recent internal structure models of Neptune (Fortney et al., 2011). Further, the zonal flows in each group’s models have strong prograde equatorial jets, opposite to those observed. 5. Conclusions and outlook Dynamo models yield a diverse spectrum of dynamical behaviors and are able to reproduce aspects of the planets’ magnetic fields, fluid flows, and thermal emissions. Thus, through these models, we learn about the dynamics and underlying mechanisms that may be occurring in planetary interiors. Further modeling progress and additional constraints on the magnetic fields and the internal structures and dynamics derived from upcoming spacecraft missions will likely allow us to develop sophisticated models that capture the relevant physics and to make predictions about the evolution of planetary magnetic fields. For example, technological advances will enable models to reach lower Ekman and magnetic Prandlt numbers, to develop asymptotic scaling laws, and to incorporate significant stratification in density and electrical conductivity. By understanding the dynamos of our solar system, we may also use these models and associated scaling laws to predict the magnetic field strengths and morphologies of exoplanetary dynamos. As for observations, the MESSENGER and Bepi-Columbo missions will help constrain the structure of Mercury’s dynamo region and the behavior of the body’s magnetic field. In addition, a possible future mission to the Jupiter system would help to characterize Ganymede’s dynamo and internal structure. The upcoming Juno mission will critically improve our understanding of Jupiter by resolving the magnetic field up to spherical harmonic degree 14, detecting short timescale secular variation, and constraining the depth of the zonal winds (Connerney, private communication). The Cassini spacecraft may closely orbit, and eventually impact, Saturn in the possible extended-extended phase of the mission, allowing the magnetic and gravitational fields to be mapped in further detail (Cassini Mission Plan, Revision N). Missions to Uranus and/or Neptune are necessary to better constrain the internal structures and dynamics of the ice giants and to map the magnetic fields globally and with higher resolution. Improved understanding of planetary dynamos will come not only from additional observations of planetary magnetic fields by future spacecraft missions, but also from new modeling efforts to address the diversity of physical settings in which planetary dynamos operate. The release of compositional buoyancy in the cores of Ganymede and Mercury can be fundamentally different from the source of compositional buoyancy in the Earth’s core, and numerical dynamo models that focus on these different buoyancy distributions are needed. Strong density and electrical conductivity stratifications occur in the interiors of the giant planets and dyna- 105 mo models that can account for these extreme property variations need to be developed. To understand dynamos that operated in the past, and even some presently active dynamos, connections with thermal evolution models of the planets and satellites are required. The internal structures of the outer planets are especially uncertain and need to be better known if the dynamos in these planets are to be modeled and understood. Even the rotation rates of the giant planet interiors need to be determined with more confidence before we can really know the significance of the observed zonal circulations. This is especially true of Saturn and possibly the ice giants. Gaps in our knowledge of planetary properties place serious limitations on our ability to understand the variety of planetary dynamos found in our solar system. Despite these obstacles, we are optimistic about future discoveries and modeling efforts that will progress the study of planetary magnetic fields. Acknowledgments G.S. was supported by NASA grant NNX09AB57G from the Planetary Atmospheres Program and by NSF Grant AST-0909206 from the Planetary Astronomy Program. K.M.S. was supported by NASA Grant NNX09AB61G from the Planetary Atmospheres program. References Abramson, E.H., 2005. Viscosity of water measured to pressures of 6 GPa and temperatures of 300 °C. Phys. Rev. E 76, 051203. Abramson, E.H., Brown, J.M., Slutsky, L.J., 2001. The thermal diffusivity of water at high pressures and temperatures. J. Chem. Phys. 115, 10461–10463. Acuña, M.H. et al., 1998. Magnetic field and plasma observations at Mars: initial results of the Mars Global Surveyor Mission. Science 279, 1676–1680. Acuña, M.H. et al., 1999. Global distribution of crustal magnetism discovered by the Mars Global Surveyor MAG/ER experiment. Science 284, 790–793. Acuña, M.H. et al., 2001. The magnetic field of Mars: summary of results from the aerobraking and mapping orbits. J. Geophys. Res. 106, 23403–23417. Aharonson, O., Auber, M., Solomon, S., 2005. Crustal remnance in an internally magnetized non-uniform shell: a possible source for Mercury’s magnetic field? Earth Planet. Sci. Lett. 218, 261–268. Anderson, B.J., Acuña, M.H., Korth, H., Purucker, M.E., Johnson, C.L., Slavin, J.A., Solomon, S.C., McNutt, R.L., 2008. The structure of Mercury’s magnetic field from MESSENGER’s first flyby. Science 321, 82–85. Anderson, B.J., Acuña, M.H., Korth, H., Slavin, J.A., Uno, H., Johnson, C.L., Purucker, M.E., Solomon, S.C., Raines, J.M., Zurbuchen, T.H., Gloeckler, G., McNutt, R.L., 2010. The magnetic field of Mercury. Space Sci. Rev. 152, 307–339. Anderson, J.D., Schubert, G., 2007. Saturn’s gravitational field, internal rotation, and interior structure. Science 371, 1384–1387. Andrews-Hanna, J.C., Zuber, M.T., Banerdt, W.B., 2008. The Borealis basin and the origin of the martian crustal dichotomy. Nature 453, 1212–1215. Arkani-Hamed, J., 2004. Timing of the Martian core dynamo. J. Geophys. Res. 109, E03006. Atkinson, D.H., Pollack, J.B., Seiff, A., 1998. The Galileo probe doppler wind experiment: measurement of the deep zonal winds on Jupiter. J. Geophys. Res. 103, 22911–22928. Aubert, J., 2005. Steady zonal flows in spherical shell dynamos. J. Fluid Mech. 542, 53–67. Aurnou, J.M., 2007. Planetary core dynamics and convective heat transfer scaling. Geophys. Astro. Fluid Dyn. 101, 327–345. Aurnou, J.M., Heimpel, M.H., Wicht, J., 2007. The effects of vigorous mixing in a convective model of zonal flow on the ice giants. Icarus 190, 110–126. Aurnou, J.M., Heimpel, M., Allen, L., King, E.M., Wicht, J., 2008. Convective heat transfer and the pattern of thermal emission on the gas giants. Geophys. J. Int. 173, 793–801. Balogh, A., Dougherty, M.K., Forsyth, R.J., Southwood, D.J., Tsurutani, B.T., Murphy, N., Burton, M.E., 1992. Magnetic field observations in the vicinity of Jupiter during the Ulysses flyby. Science 257, 1515–1518. Bibring, J.P. et al., 2006. Global mineralogical and aqueous Mars history derived from OMEGA/Mars express data. Science 312, 400–404. Bland, M.T., Showman, A.P., Tobie, G., 2008. The production of Ganymede’s magnetic field. Icarus 198, 384–399. Bloxham, J., Gubbins, D., Jackson, A., 1989. Geomagnetic secular variation. Philos. Trans. R. Soc. Lond. Ser. A 329, 415–502. Bloxham, J., Jackson, A., 1991. Fluid flow near the surface of Earth’s outer core. Rev. Geophys. 29, 97–120. Binder, A.B., 1998. Lunar prospector: overview. Science 281, 1475–1476. Boehler, R., 1996. Melting temperature of the Earth’s mantle and core: Earth’s thermal structure. Annu. Rev. Earth Planet. Sci. 24, 15–40. Bogue, S.W., Merrill, R.T., 1992. The character of the field during geomagnetic reversals. Ann. Rev. Earth Planet. Sci. 20, 181–219. 106 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 Bretherton, F.P., Turner, J.S., 1968. On the mixing of angular momentum in a stirred rotating fluid. J. Fluid Mech. 32, 449–464. Breuer, D., Yuen, D.A., Spohn, T., Zhang, S., 1998. Three dimensional models of Martian mantle convection with phase transitions. Geophys. Res. Lett. 25, 229–232. Bridge, H.S., Lazarus, A.J., Snyder, C.W., Smith, E.J., Davis Jr., L., Coleman Jr., P.J., Jones, D.E., 1969. The Venus Atmosphere, 533. Brun, A.S., Palacios, A., 2009. Numerical simulations of a rotating red giant star. I. Three-dimensional models of turbulent convection and associated mean flows. Astrophys. J. 702, 1078–1097. Bullard, E.C., 1949. The magnetic field within the Earth. Proc. R. Soc. Lond. A 197, 433–453. Burke, B.F., Franklin, K.L., 1955. Observations of a variable radio source associated with the planet Jupiter. J. Geophys. Res. 60, 213–217. Burton, M.E., Dougherty, M.K., Russell, C.T., 2009. Model of Saturn’s internal planetary magnetic field based on Cassini observations. Planet. Space Sci. 57, 1706–1713. Busse, F.H., 1976. A simple model of convection in the Jovian atmosphere. Icarus 20, 255–260. Cande, S.C., Kent, D.V., 1995. A new geomagnetic polarity time scale for the Late Cretaceous and Cenozoic. J. Geophys. Res. 97, 13917–13951. Carr, T.D., Schauble, J.J., Schauble, C.C., 1981. Pre-encounter distributions of Saturn’s low frequency radio emission. Nature 292, 745–747. Carr, T.D., Desch, M.D., Alexander, J.K., 1983. Phenomenology of magnetospheric radio emissions. In: Dessler, A.J. (Ed.), Physics of the Jovian Magnetosphere. Cambridge Univ. Press, Cambridge, UK. Cavazzoni, C., Chiarotti, G.L., Scandolo, S., Tosatti, E., Bernasconi, M., Parrinello, M., 1999. Superionic and metallic states of water and ammonia at giant planet conditions. Science 283, 44–46. Chen, B., Li, J., Hauck II, S.A., 2008. Non-ideal liquidus curve in the Fe-S system and Mercury’s snowing core. Geophys. Res. Lett. 35, L07201. Christensen, U.R., 2001. Zonal flow driven by deep convection on the major planets. Geophys. Res. Lett. 28, 2553–2556. Christensen, U.R., 2002. Zonal flow driven by strongly supercritical convection in rotating spherical shells. J. Fluid Mech. 470, 115–133. Christensen, U.R., 2006. A deep dynamo generating Mercury’s magnetic field. Nature 444, 1056–1058. Christensen, U.R., 2010. Dynamo scaling laws and applications to the planets. Space Sci. Rev. 152, 565–590. Christensen, U.R., Aubert, J., 2006. Scaling properties of convection-driven dynamos in rotating spherical shells and application to planetary magnetic fields. Geophys. J. Int. 166, 97–114. Christensen, U.R., Wicht, J., 2008. Models of magnetic field generation in partly stable planetary cores: applications to Mercury and Saturn. Icarus 196, 16–34. Clune, T.C., Elliot, J.R., Miesch, M., Toomre, J., Glatzmaier, G.A., 1999. Computational aspects of a code to study rotating turbulent convection in spherical shells. Parallel Computing 25, 361–380. Coleman Jr., P.J., Schubert, G., Russell, C.T., Sharp, L.R., 1972. Satellite measurements of the Moon’s magnetic field: a preliminary report. Moon 4, 419–429. Connerney, J.E.P., 1993. Magnetic fields of the outer planets. J. Geophys. Res. 98, 18659–18679. Connerney, J.E.P. et al., 1999. Magnetic lineations in the ancient crust of Mars. Science 284, 794–798. Connerney, J.E.P. et al., 2001. The global magnetic field of Mars and implications for crustal evolution. Geophys. Res. Lett. 28, 4015–4018. Desch, M.D., Kaiser, M.L., 1981. Voyager measurement of the rotation period of Saturn’s magnetic field. Geophys. Res. Lett. 8, 253–256. de Pater, I., Lissauer, J.J., 2001. Planetary science. Cambridge Univ. Press, Cambridge, UK. Dougherty, M.K. et al., 2005. Cassini magnetometer observations during Saturn orbit insertion. Science 307, 1266–1270. Dwyer, C.A., Stevenson, D.J., Nimmo, F., 2010. An early nutation-driven lunar dynamo. American Geophysical Union, Fall Meeting, Abstract #GP31A-02. Dziewonski, A.M., Anderson, D.L., 1981. Preliminary reference earth model. Phys. Earth Planet. Int. 25, 297–356. Elkins-Tanton, L.T., Parmentier, E.M., Hess, P.C., 2003. A model for martian magma ocean crystallization and overturn. Meteoritics and Planetary 38, 1753–1772. Elkins-Tanton, L.T., Hess, P.C., Parmentier, E.M., 2005. Possible formation of ancient crust on Mars through magma ocean processes. J. Geophys. Res. 110, E12S01. Elsasser, W.M., 1939. On the origin of the Earth’s magnetic field. Phys. Rev. 55, 489– 498. Evonuk, M., 2008. The role of density stratification in generating zonal flow structures in a rotating fluid. Astrophys. J. 673, 1154–1159. Finlay, C.C., Dumberry, M., Chulliat, A., Pais, M.A., 2010. Short timescale core dynamics: theory and observations. Space Sci. Rev. 152, 177–218. Folkner, W.M., Yoder, C.F., Yuan, D.N., Standish, E.M., Preston, R.A., 1997. Interior structure and seasonal mass redistribution of Mars from radio tracking of Mars Pathfinder. Science 278, 1749–1752. Fortney, J.J., Ikoma, M., Nettelmann, N., Guillot, T., Marley, M.S., 2011. Selfconsistent model atmospheres and the cooling of the solar system’s giant planets. Astrophys. J. 729, 32–46. Frey, H.V., Schultz, R.A., 1988. Large impact basins and the mega-impact origin for the crustal dichotomy on Mars. Geophys. Res. Lett. 15, 229–232. Frey, H.V., 2003. Buried impact basins and the earliest history of Mars. Sixth Intl. Conf. on Mars, Calif. Inst. of Technol., Pasadena, CA, 20–25 July. Frey, H.V., Edgar, L., Lillis, R., 2007. Very large visible and buried impact basins on Mars: implications for internal and crustal evolution and the late heavy bombardment in the inner solar system. Seventh Intl. Conf. on Mars, Calif. Inst. of Technol., Pasadena, CA, 9–13 July. Garrick-Bethell, I., Weiss, B.P., Shuster, D.L., Buz, J., 2009. Early lunar magnetism. Science 323, 356–359. Giampieri, G., Balogh, A., 2002. Mercury’s thermoelectric dynamo model revisited. Planet. Space Sci. 50, 757–762. Giampieri, G., Dougherty, M., 2004. Rotation rate of Saturn’s interior from magnetic field observations. Geophys. Res. Lett. 31, L16701. Glassmeier, K.H., Auster, H.U., Motschmann, U., 2007. A feedback dynamo generating Mercury’s magnetic field. Geophys. Res. Lett. 34, L22201. Glatzmaier, G.A., 2002. Geodynamo simulations – How realistic are they? Annu. Rev. Earth Planet. Sci. 30, 237–257. Glatzmaier, G.A., Gilman, P.A., 1981. Compressible convection in a rotating spherical shell. III. Analytic model for compressible vorticity waves. Astrophys. J. Suppl. 45, 381–388. Glatzmaier, G.A., Coe, R.S., Hongre, L., Roberts, P.H., 1999. The role of the Earth’s mantle in controlling the frequency of geomagnetic reversals. Nature 401, 885–890. Glatzmaier, G.A., Evonuk, M., Rogers, T.M., 2009. Differential rotation in giant planets maintained by density-stratified turbulent convection. Geophys. Astrophys. Fluid Dyn. 103, 31–51. Gomez-Perez, N., Heimpel, M.H., 2007. Numerical models of zonal flow dynamos: an application to the ice giants. Geophys. Astrophys. Fluid Dyn. 101, 371–388. Gomez-Perez, N., Solomon, S., 2010. Mercury’s weak magnetic field: a result of magnetospheric feedback? Geophys. Res. Lett. 37, L20204. Gomez-Perez, N., Wicht, J., 2010. Behavior of planetary dynamos under the influence of external magnetic fields: application to Mercury and Ganymede. Icarus 1, 53–62. Gough, D.O., Lynden-Bell, D., 1968. Vorticity expulsion by turbulence: astrophysical implications of an AlkaSeltzer experiment. J. Fluid Mech. 32, 437–447. Greeley, R., Crown, D.A., 1990. Volcanic geology of tyrrhena patera. Mars. J. Geophys. Res. 95 (B5), 7133–7149. Grote, E., Busse, F.H., Tilgner, A., 1999. Convection-driven quadrupolar dynamos in rotating spherical shells. Phys. Rev. E 60, 5025–5028. Grote, E., Busse, F.H., Tilgner, A., 2000. Regular and chaotic spherical shell dynamos. Phys. Earth Planet. Int. 117, 259–272. Guillot, T., 1999a. Interiors of giant planets inside and outside the solar system. Science 286, 72–77. Guillot, T., 1999b. A comparison of the interiors of Jupiter and Saturn. Planet. Space Sci. 47, 1183–1200. Guillot, T., 2005. The interiors of giant planets: models and outstanding questions. Annu. Rev. Earth Planet. Sci. 33, 493–530. Gubbins, D., 2001. The Rayleigh number for convection in the Earth’s core. Phys. Earth Planet. Int. 128, 3–12. Gurnett, D.A. et al., 2007. The variable rotation period of the inner region of Saturn’s plasma disk. Science 316, 442–445. Hale, C., Dunlop, D., 1984. Evidence for an early archean geomagnetic field: a paleomagnetic study of the Komati Formation, Barberton Greenstone Belt. South Africa Geophys. Res. Lett. 11, 97–100. Hammel, H.B., de Pater, I., Gibbard, S., Lockwood, G.W., Rages, K., 2005. Uranus in 2003: zonal winds, banded structure, and discrete features. Icarus 175, 534– 545. Harder, H., 1998. Phase transitions and the three-dimensional planform of thermal convection in the Martian mantle. J. Geophys. Res. 103, 16775–16797. Hauck II, S.A., Aurnou, J.M., Dombard, A.J., 2006. Sulfur’s impact on core evolution and magnetic field generation on Ganymede. J. Geophys. Res. 111, E09008. Heimpel, M.H., Aurnou, J.M., Wicht, J., 2005a. Simulation of equatorial and highlatitude jets on Jupiter in a deep convection model. Nature 438, 193–196. Heimpel, M.H., Aurnou, J.M., Al-Shamali, F., Gomez-Perez, N., 2005b. A numerical study of dynamo action as a function of spherical shell geometry. Earth Planet. Sci. Lett. 236, 542–557. Heimpel, M.H., Aurnou, J.M., 2007. Turbulent convection in rapidly rotating spherical shells: a model for equatorial and high-latitude jets on Jupiter and Saturn. Icarus 187, 540–557. Helled, R., Anderson, J., Schubert, G., 2010. Uranus and Neptune: shape and rotation. Icarus 210, 446–454. Hewitt, J.M., McKenzie, D.P., Weiss, N.O., 1975. Dissipative heating in convective flows. J. Fluid Mech. 68, 721–738. Holme, R., Bloxham, J., 1996. The magnetic fields of Uranus and Neptune: methods and models. J. Geophys. Res. 101, 2177–2200. Hood, L.L., Richmond, N.C., Pierazzo, E., Rochette, P., 2003. Distribution of crustal magnetic fields on Mars: shock effects of basin-forming impacts. Geophys. Res. Lett. 30, 1281. Hood, L.L., Harrison, K.P., Langlais, B., Lillis, R.J., Poulet, F., Williams, D.A., 2010. Magnetic anomalies near Apollinaris Patera and the Medusae Fossae Formation in Lucus Planum, Mars. Icarus 208, 118–131. Hubbard, W., Podolak, M., Stevenson, D., 1995. Interior of Neptune. In: Cruickshank, D. (Ed.), Neptune and Triton. Univ. Arizona Press, Tucson. Hulot, G., Finlay, C.C., Constable, C.G., Olsen, N., Mandea, M., 2010. The magnetic field of planet Earth. Space Sci. Rev. 152, 159–222. Iess, L., Rappaport, N.J., Jacobson, R.A., Racioppa, P., Stevenson, D.J., Tortora, P., Armstrong, J.W., Asmar, S.W., 2010. Gravity field, shape, and moment of inertia of Titan. Science 327, 1367–1369. Jia, X., Kivelson, M.G., Khurana, K.K., Walker, R.J., 2010. Magnetic fields of the satellites of Jupiter and Saturn. Space Sci. Rev. 152, 271–305. Johnson, C., Constable, C., Tauxe, L., 2003. Mapping long-term changes in Earth’s magnetic field. Science 300, 2044–2045. G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 Jones, C.A., 2000. Convection-driven geodynamo models. Philos. Trans. R. Soc. Lond. Ser. A 204, 227–238. Jones, C.A., Kuzanyan, K.M., 2009. Compressible convection in the deep atmospheres of giant planets. Icarus 358, 873–897. Kabin, K., Heimpel, M.H., Rankin, R., Aurnou, J.M., Gomez-Perez, N., Paral, J., Gombosi, T.I., Zurbuchen, T.H., Koehn, P.L., DeZeeuw, D.L., 2008. Global MHD modeling of Mercury’s magnetosphere with applications to the MESSENGER mission and dynamo theory. Icarus 195, 1–15. Ke, Y., Solomatov, V.S., 2006. Early transient superplumes and the origin of the Martian crustal dichotomy. J. Geophys. Res. 111, E10001. Khan, A., Mosegaard, K., Williams, J.G., Lognonné, P., 2004. Does the Moon possess a molten core? Probing the deep lunar interior using results from LLR and lunar prospector. J. Geophys. Res. 109, E09007. Khurana, K.K., Jia, X., Kivelson, M.G., Nimmo, F., Schubert, G., Russell, C.T., 2011. Evidence of a global magma ocean in Io’s interior. Science. doi:10.1126/ science.1201425. Kimura, J., Nakagawa, T., Kurita, K., 2009. Size and compositional constraints of Ganymede’s metallic core for driving an active dynamo. Icarus 202, 216–224. King, E.M., Soderlund, K.M., Christensen, U.R., Wicht, J., Aurnou, J.M., 2010. Convective heat transfer in planetary dynamo models. Geochem. Geophys. Geosyst. 11, Q06016. Kivelson, M.G., Khurana, K.K., Russell, C.T., Walker, R.J., Warnecke, J., Coroniti, F.V., Polanskey, C., Southwood, D.J., Schubert, G., 1996. Discovery of Ganymede’s magnetic field by the Galileo spacecraft. Nature 384, 537–541. Kivelson, M.G., Khurana, K.K., Volwerk, M., 2002. The permanent and inductive magnetic moments of Ganymede. Icarus 157, 507–522. Kono, M., Roberts, P.H., 2002. Recent geodynamo simulations and observations of the geomagnetic field. Rev. Geophys. 40, 1013. Konopliv, A.S., Yoder, C.F., 1996. Venusian k(2) tidal Love number from Magellan and PVO tracking data. Geophys. Res. Lett. 23, 1857–1860. Konopliv, A.S., Binder, A.B., Hood, L.L., Kucinskas, A.B., Sjogren, W.L., Williams, J.G., 1998. Improved gravity field of the Moon from lunar prospector. Science 281, 1476–1480. Kuang, W., Jiang, W., Wang, T., 2008. Sudden termination of Martian dynamo? Implications from subcritical dynamo simulations. Geophys. Res. Lett. 35, L14204. Kutzner, C., Christensen, U.R., 2000. Effects of driving mechanisms in geodynamo models. Geophys. Res. Lett. 27, 29–32. Kutzner, C., Christensen, U.R., 2002. From stable dipolar to reversing numerical dynamos. Phys. Earth Planet. Int. 131, 29–45. Langlais, B., Purucker, M.E., Mandea, M., 2004. Crustal magnetic field of Mars. J. Geophys. Res. 109, E02008. Langlais, B., Lesur, V., Purucker, M.E., Connerney, J.E.P., Mandea, M., 2010. Crustal magnetic fields of terrestrial planets. Space Sci. Rev. 152, 223–249. Lee, K.K.M. et al., 2006. Laser-driven shock experiments on precompressed water: implications for ‘‘icy’’ giant planets. J. Chem. Phys. 125, 014701. Lillis, R.J., Mitchell, D.L., Lin, R.P., Connerney, J.E.P., Acuña, M.H., 2004. Mapping crustal magnetic fields at Mars using electron reflectometry. Geophys. Res. Lett. 31, L15702. Lillis, R.J., Manga, M., Mitchell, D.L., Lin, R.P., Acuña, M.H., 2005. Evidence for a second Martian dynamo from electron reflection magnetometry. Lunar Planet. Sci. XXXVI 1578. Lillis, R.J., Manga, M., Mitchell, D.L., Lin, R.P., Acuña, M.H., 2006. Unusual magnetic signature of the Hadriaca Patera Volcano: implications for early Mars. Geophys. Res. Lett. 33, L03202. Lillis, R.J., Frey, H.V., Manga, M., 2008a. An improved crustal magnetic field map of Mars from electron reflectometry: highland volcano magmatic history and the end of the martian dynamo. Icarus 194, 575–596. Lillis, R.J., Frey, H.V., Manga, M., 2008b. Rapid decrease in Martian crustal magnetization in the Noachian era: implications for the dynamo and climate of early Mars. Geophys. Res. Lett. 35, L14203. Lingenfelter, R.E., Schubert, G., 1973. Evidence for convection in planetary interiors from first-order topography. Earth Moon Planets 7, 172–180. Liu, J., Goldreich, P.M., Stevenson, D.J., 2008. Constraints on deep-seated zonal winds inside Jupiter and Saturn. Icarus 196, 653–664. Lodders, K., Fegley Jr., B., 1998. The Planetary Scientist’s Companion. Oxford Univ. Press, Oxford. Longhi, J., Knittle, E., Holloway, J.R., Waenke, H., 1992. The bulk composition, mineralogy and internal structure of Mars. In: Kieffer, H.H. et al. (Eds.), Mars. Univ. of Ariz. Press, Tucson. Malavergne, V., Toplis, M.J., Berthet, S., Jones, J., 2010. Highly reducing conditions during core formation on Mercury: implications for internal structure and the origin of a magnetic field. Icarus 206, 199–209. Malkus, W.V.R., 1994. Energy sources for planetary dynamos. In: Proctor, M.R.E., Gilbert, A.D. (Eds.), Lectures on Solar and Planetary Dynamos. Cambridge Univ. Press, Cambridge, UK. Margot, J., Peale, S., Jurgens, R., Slade, M., Holin, I., 2007. Large longitude libration of Mercury reveals a molten core. Science 316, 710–714. Marinova, M.M., Aharonson, O., Asphaug, E., 2008. Mega-impact formation of the Mars hemispheric dichotomy. Nature 453, 1216–1219. McComas, D.J., Allegrini, F., Bagenal, F., Crary, F., Ebert, R.W., Elliott, H., Stern, A., Valek, P., 2007. Diverse plasma populations and structures in Jupiter’s magnetotail. Science 318, 217–220. McEwen, A.S., Lunine, J.I., Carr, M.H., 1989. Dynamic Geophysics of Io. In Time Variable Phenomena in the Jovian System, NASA-SP-494, pp. 11–46. McNutt Jr., R.L. et al., 2007. Energetic particles in the Jovian magnetotail. Science 318, 220–222. 107 McNutt Jr., R.L. et al., 2010. The MESSENGER mission: results from the first two Mercury flybys. Acta Astronautica 67, 681–687. Merrill, R.T., McElhinny, M.W., McFadden, P.L., 1996. The magnetic field of the Earth: paleomagnetism, the core, and the deep mantle. Academic Press, London. Milbury, C., Schubert, G., 2010. Search for the global signature of the Martian dynamo. J. Geophys. Res. 115, E10010. Milbury, C., Schubert, G., Raymond, C., Smrekar, S., 2011. Modeling gravity and magnetic anomalies at Tyrrhena Patera, Mars: implications for early polar wander and a late dynamo. In review. J. Geophys. Res.. Mitchell, D.L., Lillis, R.J., Lin, R.P., Connerney, J.E.P., Acuña, M.H., 2007. A global map of Mars’ crustal magnetic field based on electron reflectometry. J. Geophys. Res. 112, E01002. Mohit, P.S., Arkani-Hamed, J., 2004. Impact demagnetization of the Martian crust. Icarus 168, 305–317. Nellis, W.J., 2000. Metallization of fluid hydrogen at 140 GPa (1.4 Mbar): implications for Jupiter. Planet. Space Sci. 48, 671–677. Ness, N.F., Behannon, K., Lepping, R., Whang, Y., 1975. Magnetic field of Mercury confirmed. Science 255, 204–206. Ness, N.F., Behannon, K., Lepping, R., Whang, Y., 1976. Observations of Mercury’s magnetic field. Icarus 28, 479–488. Ness, N.F., Acuña, M.H., Lepping, R.P., Burlaga, L.F., Behannon, K.W., Neubauer, F.M., 1979. Magnetic field studies at Jupiter by Voyager 1: preliminary results. Science 204, 982–987. Ness, N.F. et al., 1981. Magnetic field studies by Voyager 1 – preliminary results at Saturn. Science 212, 211–217. Ness, N.F., Acuña, M.H., Behannon, K.W., Burlaga, L.F., Connerney, J.E.P., Lepping, R.P., Neubauer, F.M., 1982. Magnetic field studies by Voyager 2: preliminary results at Saturn. Science 215, 558–563. Ness, N.F., Acuña, M.H., Behannon, K.W., Burlaga, L.F., Connerney, J.E.P., Lepping, R.P., Neubauer, F.M., 1986. Magnetic fields at Uranus. Science 233, 85–89. Ness, N.F., Acuña, M.H., Burlaga, L.F., Connerney, J.E.P., Lepping, R.P., Neubauer, F.M., 1989. Magnetic fields at Neptune. Science 246, 1473–1478. Ness, N.F. et al., 1999. MGS magnetic fields and electron reflectometer investigation: discovery of paleomagnetic fields due to crustal remanence. Adv. Space Res. 23, 1879–1886. Nimmo, F., Stevenson, D.J., 2000. Influence of early plate tectonics on the thermal evolution and magnetic field of Mars. J. Geophys. Res. 105, 11969–11979. Nimmo, F., Hart, S.D., Korycansky, D., Agnor, C.B., 2008. Implications of an impact origin for the martian hemispheric dichotomy. Nature 453, 1220–1223. Olsen, N., Mandea, M., 2008. Rapidly changing flows in the Earth’s core. Nature 1, 390–394. Olson, P., Christensen, U.R., 2006. Dipole moment scaling for convection-driven planetary dynamos. Earth Planet. Sci. Lett. 250, 561–571. Pearl, J.C., Conrath, B.J., 1991. The albedo, effective temperature, and energy balance of Neptune as determined from Voyager data. J. Geophys. Res. Suppl. 96, 18921– 18930. Pearl, J.C., Conrath, B.J., Hanel, R.A., Pirraglia, J.A., 1990. The albedo, effective temperature, and energy balance of Uranus as determined from Voyager IRIS data. Icarus 84, 12–28. Phillips, J.L., Russell, C.T., 1987. Upper limit on the intrinsic magnetic field of Venus. J. Geophys. Res. 92, 2253–2263. Pirraglia, J.A., 1984. Meridional energy balance of Jupiter. Icarus 59, 169–176. Podolak, M., Hubbard, W.B., Stevenson, D.J., 1991. Models of Uranus’ interior and magnetic field. In: Bergstralh, J.T., Miner, E.D. (Eds.), Uranus, Matthews, M.S., Univ. of Ariz. Press, Tucson. Porco, C.C. et al., 2003. Cassini imaging of Jupiter’s atmosphere, satellites and rings. Science 299, 1541–1547. Proudman, J., 1916. On the motion of solids in a liquid possessing vorticity. Proc. R. Soc. Lond. A 92, 408–424. Purucker, M.E., Ravat, D., Frey, H., Voorhies, C., Sabaka, T., Acuña, M., 2000. An altitude-normalized magnetic map of Mars and its interpretation. Geophys. Res. Lett. 27, 2449–2452. Purucker, M.E., 2008. A global model of the internal magnetic field of the Moon based on lunar prospector magnetometer observations. Icarus 197, 19–23. Read, P.L., Dowling, T.E., Schubert, G., 2009. Saturn’s rotation period from its atmospheric planetary-wave configuration. Nature 460, 608–610. Rhines, P.B., 1975. Waves and turbulence on a beta-plane. J. Fluid Mech. 69, 417– 443. Riner, M.A., Bina, C.R., Robinson, M.S., Desch, S.J., 2008. Internal structure of Mercury: implications of a molten core. J. Geophys. Res. 113, E08013. Robbins, S.J., Achille, G.D., Hynek, B.M., 2011. The volcanic history of Mars: highresolution crater-based studies of the calderas of 20 volcanoes. Icarus 211, 1179–1203. Roberts, J.H., Zhong, S., 2006. Degree-1 convection in the Martian mantle and the origin of the hemispheric dichotomy. J. Geophys. Res. 111, E06013. Russell, C.T., Weiss, H., Coleman, P.J., Jr., Soderblom, L.A., Stuart-Alexander, D.E., Wilhelms, D.E., 1977. In: Proc. 8th Lunar Sci. Conf., Pergamon, New York, pp. 1171–1185. Russell, C.T., Elphic, R.C., Slavin, J.A., 1980. Limits on the possible intrinsic magnetic field of Venus. J. Geophys. Res. 85 (A13), 8319–8332. Russell, C.T., Yu, Z.J., Kivelson, M.G., 2001. The rotation period of Jupiter. Geophys. Res. Lett. 28, 1911–1912. Russell, C.T., Dougherty, M.K., 2010. Magnetic fields of the outer planets. Space Sci. Rev. 152, 251–269. Sakuraba, M., Roberts, P.H., 2009. Generation of a strong magnetic field using uniform heat flux at the surface of the core. Nature Geol. 2, 802–805. 108 G. Schubert, K.M. Soderlund / Physics of the Earth and Planetary Interiors 187 (2011) 92–108 Sanchez-Lavega, A., Rojas, J.F., Sada, P.V., 2000. Saturn’s zonal winds at cloud level. Icarus 147, 405–420. Sanchez-Lavega, A., 2004. The magnetic field in giant extrasolar planets. Astrophys. J. 609, L87–L90. Sarson, G.R., Jones, C.A., Zhang, K., Schubert, G., 1997. Magnetoconvection dynamos and the magnetic fields of Io and Ganymede. Science 276, 1106–1108. Schubert, G., Lingenfelter, R.E., 1973. Martian centre of mass – centre of figure offset. Nature 242, 251–252. Schubert, G., Ross, M.N., Stevenson, D.J., Spohn, T., 1988. Mercury’s thermal history and the generation of its magnetic field. In: Vilas, F., Chapman, C.R., Matthews, M.S. (Eds.), Mercury. Univ. of Ariz. Press, Tucson. Schubert, G., Zhang, K., Kivelson, M.G., Anderson, J.D., 1996. The magnetic field and internal structure of Ganymede. Nature 384, 544–545. Schubert, G., Russell, C.T., Moore, W.B., 2000. Geophysics – timing of the Martian dynamo. Nature 408, 666–667. Schubert, G., Turcotte, D.L., Olson, P., 2001. Mantle Convection in the Earth and Planets. Cambridge Univ. Press, Cambridge, UK. Schubert, G., Chan, K., Liao, X., Zhang, K., 2004. Planetary dynamos: effects of electrically conducting flows overlying turbulent regions of magnetic field generation. Icarus 172, 305–315. Schubert, G., Anderson, J.D., Spohn, T., McKinnon, W.B., 2007. Interior composition, structure and dynamics of the Galilean satellites. In: Bagenal, F., Dowling, T.E., McKinnon, W.B. (Eds.), Jupiter. Cambridge Univ. Press, Cambridge, UK. Scorer, R.S., 1966. Origin of cylcones. Sci. J. 2, 46–52. Shkolnik, E., Walker, G.A.H., Bohlender, D.A., Gu, P.-G., Kuerster, M., 2005. Hot Jupiters and hot spots: the short- and long-term chromospheric activity on stars with giant planets. Astrophys. J. 622, 1075–1090. Shkolnik, E., Bohlender, D.A., Walker, G.A.H., Collier Cameron, A., 2008. The on/off nature of star–planet interactions. Astrophys. J. 676, 628–638. Simitev, R., Busse, F.H., 2005. Prandtl-number dependence of convection-driven dynamos in rotating spherical fluid shells. J. Fluid Mech. 532, 365–388. Smith, E.J., Davis, L., Jones, D.E., Coleman, P.J., Colburn, D.S., Pyal, P., Sonett, C.P., Frandsen, A.M.A., 1974. The planetary magnetic field and magnetosphere of Jupiter: Pioneer 10. J. Geophys. Res. 79, 3501–3513. Smith, E.J., Davis, L., Jones, D.E., Coleman, P.J., Colburn, D.S., Pyal, P., Sonett, C.P., 1975. Jupiter’s magnetic field, magnetosphere, and interaction with the solar wind: Pioneer II. Science 188, 451–454. Smith, E.J., Davis, L., Jones, D.E., Coleman, P.J., Colburn, D.S., Sonett, C.P., 1980. Saturn’s magnetic field and magnetosphere. Science 207, 407–410. Smith, E.J. et al., 2010. The equatorial shape and gravity field of Mercury from MESSENGER flybys 1 and 2. Science 209, 88–100. Smylie, D.E., Rochester, M.G., 1980. Compressibility, core dynamics and the subseismic wave equation. Phys. Earth Planet. Int. 24, 308–319. Solomon, S., 2011. A new look at the planet Mercury. Phys. Today 64, 50–55. Soderlund, K.M., Aurnou, J.M., 2010. Simulation of an ice giant-style dynamo. American Geophysical Union Fall Meeting 2010. San Francisco. Soderlund, K.M., King, E.M., Aurnou, J.M., in review. The weak influence of magnetic fields in planetary dynamo models. Earth Planet. Sci. Lett. Sreenivasan, B., Jones, C.A., 2006. The role of inertia in the evolution of spherical dynamos. Geophys. J. Int. 164, 467–476. Sromovsky, L.A., Fry, P.M., Dowling, T.E., Baines, K.H., Limaye, S.S., 2001. Neptune’s atmospheric circulation and cloud morphology: Changes revealed by 1998 HST imaging. Icarus 150, 244–260. Stanley, S., 2010. A dynamo model for axisymmetrizing Saturn’s magnetic field. Geophys. Res. Lett. 37, L05201. Stanley, S., Bloxham, J., 2004. Convective-region geometry as the cause of Uranus’ and Neptune’s magnetic fields. Nature 428, 151–153. Stanley, S., Bloxham, J., 2006. Numerical dynamo models of Uranus’ and Neptune’s unusual magnetic fields. Icarus 184, 556–572. Stanley, S., Bloxham, J., Hutchison, W.E., Zuber, M.T., 2005. Thin shell dynamo models consistent with Mercury’s weak observed magnetic field. Earth Planet. Sci. Lett. 234, 27–38. Stanley, S., Mohammadi, A., 2008. Effects of an outer thin stably stratified layer on planetary dynamos. Phys. Earth Planet. Sci. 168, 179–190. Stanley, S., Elkins-Tanton, L., Zuber, M., Parmentier, E.M., 2008. Mars’ paleomagnetic field as the result of a single-hemisphere dynamo. Science 321, 1822–1825. Stanley, S., Glatzmaier, G.A., 2010. Dynamo models for planets other than Earth. Space Sci. Rev. 152, 617–649. Stegman, D.R., Jellinek, A.M., Zatman, S.A., Baumgardner, J.R., Richards, M.A., 2003. An early lunar core dynamo driven by thermochemical mantle convection. Nature 421, 143–146. Stevenson, D.J., 1980. Saturn’s luminosity and magnetism. Science 208, 746–748. Stevenson, D.J., 1982. Reducing the non-axisymmetry of a planetary dynamo and an application to Saturn. Geophys. Astrophys. Fluid Dyn. 21, 113–127. Stevenson, D.J., 1983. Planetary magnetic fields. Rep. Prog. Phys. 46, 555–620. Stevenson, D.J., 1987. Mercury’s magnetic field: a thermoelectric dynamo? Earth Planet. Sci. Lett. 82, 114–120. Stevenson, D.J., 2003. Planetary magnetic fields. Earth Planet. Sci. Lett. 208, 1–11. Stevenson, D.J., 2010. Planetary magnetic fields: achievements and prospects. Space Sci. Rev. 152, 651–664. Stevenson, D.J., Salpeter, E.E., 1977. The phase diagram and transport properties for hydrogen–helium fluid planets. Astrophys. J. Suppl. 35, 221–237. Stevenson, D.J., Spohn, T., Schubert, G., 1983. Magnetism and thermal evolution of the terrestrial planets. Icarus 54, 466–489. Stacey, F., 2007. Core properties, physical. In: Gubbins, D., Herrero-Bervera, E. (Eds.), Encyclopedia of Geomagnetism and Paleomagnetism. Springer-Verlag, Netherlands. Suomi, V.E., Limaye, S.S., Johnson, D.R., 1991. High winds of Neptune: a possible mechanism. Science 251, 929–932. Takahashi, F., Matsushima, M., 2006. Dipolar and non-dipolar dynamos in a thin shell geometry with implications for the magnetic field of Mercury. Geophys. Res. Lett. 33, L10202. Takahashi, F., Tsunakawa, H., 2009. Thermal core-mantle coupling in an early lunar dynamo: implications for a global magnetic field and magnetosphere of the early Moon. Geophys. Res. Lett. 36, L24202. Taylor, G., 1917. Motion of solids in fluids when the flow is not irrotational. Proc. R. Soc. Lond. A 93, 99–113. Tilgner, A., 2007. Kinematic dynamos with precession driven flow in a sphere. Geophys. Astrophys. Fluid Dyn. 101, 1–9. Uno, H., Johnson, C.L., Anderson, B.J., Korth, H., Solomon, S.C., 2009. Modeling Mercury’s internal magnetic field with smooth inversions. Earth Planet. Sci. Lett. 285, 328–339. Vilim, R., Stanley, S., Hauck II, S.A., 2010. Iron snow zones as a mechanism for generating Mercury’s weak observed magnetic field. J. Geophys. Res. 115, E11003. Watters, W., Zuber, M., Hager, B., 2009. Thermal perturbations caused by large impacts and consequences for mantle convection. J. Geophys. Res. 114, E02001. Weinstein, S.A., 1995. The effects of a deep mantle endothermic phase change on the structure of thermal convection in silicate planets. J. Geophys. Res. 100 (E6), 11719–11728. Weiss, B.P., Vali, H., Baudenbacher, F.J., Kirschvink, J.L., Stewart, S.T., Shuster, D.L., 2002. Records of an ancient Martian magnetic field in ALH84001. Earth Planet. Sci. Lett. 201, 449–463. Weiss, B.P., Berdahl, J.S., Elkins-Tanton, L., Stanley, S., Lima, E.A., Carporzen, L., 2008. Magnetism on the angrite parent body and the early differentiation of planetesimals. Science 322, 713–716. Werner, S.C., 2009. The global martian volcanic evolutionary history. Icarus 201, 44–68. Wicht, J., Stellmach, S., Harder, H., 2009. Numerical models of the geodynamo: from fundamental cartesian models to 3D simulations of field reversals. In: Geomagnetic Field Variations, Springer, p. 107. Wienbruch, U., Spohn, T., 1995. A self-sustained magnetic field on Io? Planet. Space Sci. 43, 1045–1057. Wilhelms, D.E., Squyres, S.W., 1984. The martian hemispheric dichotomy may be due to a giant impact basin. Nature 309, 138–140. Williams, J.-P., Aharonson, O., Nimmo, F., 2007. Powering Mercury’s dynamo. Geophys. Res. Lett. 34, L21201. Williams, D.A., Greeley, R., Werner, S.C., Michael, G., Crown, D.A., Neukum, G., Raitala, J., 2008. Tyrrhena Patera: geologic history derived from Mars express high resolution stereo camera. J. Geophys. Res. 113, E11005. Yoder, C.F., Konopliv, A.S., Yuan, D.N., Standish, E.M., Folkner, W.M., 2003. Fluid core size of Mars from detection of the solar tide. Science 300, 299–303. Yu, Z.J., Leinweber, H., Russell, C.T., 2010. Galileo constraints on the secular variation of the Jovian magnetic field. J. Geophys. Res. 115, E03002. Zhang, K., 1992. Spiraling columnar convection in rapidly rotating spherical fluid shells. J. Fluid Mech. 236, 535–554. Zhang, C.Z., Zhang, K., 1995. On the internal structure and magnetic fields of Venus. Earth Moon Planets 69, 237–247. Zharkov, V.N., 1983. Models of the internal structure of Venus. Earth Moon Planets 29, 139–175. Zhong, S., Zuber, M.T., 2001. Degree-1 mantle convection and the crustal dichotomy on Mars. Earth Planet. Sci. Lett. 189, 75–84.