by Timothy E. Dowling
Jupiter is a beautiful and inspiring planet and has returned major scientific dividends on every investment in exploring it. This largest planet in the solar system has more mass than all of the other planets combined, raging storms that last for decades and longer, 16 moons, the largest bigger than Mercury, and a magnetosphere so vast it would appear as big as the full Moon in the sky— Saturn routinely passes through Jupiter’s magnetotail. There are several reasons why Jupiter gets the lion’s share of attention in planetary exploration. First, it is an extremely common planet type. For example, it is likely that many solar systems have a dominant gas giant planet whose orbital radius is in the neighborhood of the ”snow line,” the approximately 5AU distance from a protosun where the temperature in the protoplanetary nebula drops to the point that ice becomes stable. Its position makes Jupiter the protector of the inner solar system. Because Jupiter contains most of the solar system’s planetary mass, accurate observations of its composition, in particular, the hydrogen-to-helium ratio and the precise abundances of elements heavier than helium, are crucial to the success of solar-system formation models and serve as a springboard for cosmo-chemistry. Jupiter is the case study for all gas giants, including Saturn, Uranus, Neptune, and the nearly 100 extrasolar gas giants discovered to date. Second, the meteorology that shapes the planet’s appearance is more predictable than Earth’s weather. Studying the differences and similarities has led to a better understanding of atmospheric dynamics and chemistry in general. Third, Jupiter’s magnetosphere is second only to the Sun’s; it has an active aurora and strong interactions with its satellites and the solar wind. We can expect that someday a high-order spherical harmonic mapping of the magnetosphere will be used to probe the planet’s interior structure. Finally, because Jupiter is the closest gas giant to Earth, it is the easiest to observe, and because every probe headed for the outer solar system uses Jupiter for a gravity assist, it has had the largest number of close encounters with spacecraft, such that every investigation of the Jovian system is leveraged by its many predecessors (1).
Atmosphere
Jupiter is instantly recognizable by its bands of clouds and its large, ruddy eye, the Great Red Spot. In 1979, the twin Voyager spacecraft returned the first crystal-clear images of the planet, revealing huge stable ovals standing alongside filamentary swirls of white, red, and brown that are as striking to the first-time viewer as any masterpiece of artwork. Jupiter has six times as many jet streams as Earth and storm systems that last a lifetime or longer. In all respects it is a giant planet; it has 71% of the solar system’s total planetary mass and more than 120 times the surface area of Earth.
Studies of the dynamics of Jupiter’s atmosphere show that it has just as much in common with Earth’s oceans as it does with Earth’s atmosphere. Dynamically, Earth’s oceans are giant in size compared to the size of ocean eddies—an ocean basin spans hundreds of eddies—just as Jupiter’s clouds play host to hundreds of storms. In contrast, Earth’s atmosphere is a crowded place and has room for only about a half-dozen large eddies at once. Earth’s atmosphere receives more energy per area than any other planetary atmosphere in the solar system, even more than Venus, which is closer to the Sun but more reflective; so perhaps it is no coincidence that our home planet turns out to have the most unpredictable weather in the solar system. But, Earth also has the weakest winds of all of the planets, and so we must guard against making overly simplistic statements about the role of sunlight. As a case in point, the jet streams on Jupiter are more than twice as fast as those on Earth, even though the sunlight received there is 25 times weaker. And, the winds on the other gas giants are even faster than on Jupiter, even though the Sun is just a bright star in the sky as seen from Saturn, Uranus, and Neptune.
The lack of connection between sunlight and wind speed is a major result of the space-craft era, and to it we can add the following two lessons from Earth’s air-sea system when considering Jupiter’s winds. First, the currents in Earths oceans are mostly caused by wind action—they would run down if the wind stopped blowing. Likewise, Jupiter may have extremely deep, even powerful circulations that are spun up from the alternating jets overhead, rather than the other way around, a tail-wags-the-dog viewpoint that, nevertheless, comes naturally to oceanographers. Second, the process of creating a wind-driven ocean current does not involve a one-dimensional push of the water by the wind, like a child pushing a toy car. Rather, it is a three-dimensional Coriolis reaction to tilts of the constant-pressure surfaces in the top layers of the ocean (caused by Ekman pumping, a boundary-layer effect). The main lesson is that it is not hard to generate strong, alternating jet streams on a rapidly rotating planet, so long as it is not hard to cause and maintain tilts of the constant-pressure surfaces. Considering that planetary rotation is a vast reservoir of momentum and sunlight is not required to keep the planets rotating, perhaps we should not be surprised after all to see that sunlight is not directly involved in determining wind speed. Instead, we should focus on which processes, including moist convection, internal heat, and sunlight, act to warp constant-pressure surfaces in an atmosphere and let the Coriolis effect take care of the rest.
Observations. Jupiter appears to the naked eye as a large, bright star; an ordinary pair of binoculars reveals its famous banded clouds, its Great Red Spot, and its four largest moons, called the Galilean satellites, which Galileo discovered 4 centuries ago using one of the first telescopes. The dark bands on Jupiter are traditionally called belts (dark like a belt worn around a waist), and the light bands are called zones. For a quarter-century, high-resolution images from the spacecrafts Voyager, Galileo, Cassini, and the Hubble Space Telescope, have revealed fine details in Jupiter’s clouds (Fig. 1) that allow determining the winds reliably and repeatedly. Jupiter is particularly photogenic; in contrast, cloud tracking is not reliable for deducing wind speeds on Earth, Uranus, or Neptune, because their clouds are more ephemeral than on Jupiter; Saturn is better, except that its cloud features are muted by an overlying haze. The peaks of Jupiter’s jets are centered precisely on the abrupt color boundaries between the belts and zones, and the speed and location of the jet streams have shown virtually no change during the spacecraft era, even though the color contrast of the jets often changes. The exception that proves the rule is the strong eastward jet at 23°N, which has shown a 20% variation in speed during the spacecraft era; most of the other jets have shown no variation at all.
Jupiter’s surface clouds top out at a pressure level of about 0.7 bar (1bar = 105Pa, approximately sea-level pressure on Earth); the tops in the zones are somewhat higher than in the belts. The clouds are predominantly white ammonia ice that is colored by unknown chromophores, probably sulfur or phosphorus compounds in trace amounts; the chemical makeup of the chromophores has been a long-standing open question that will probably require in situ measurements to answer. Occasionally, a little red spot appears in the Northern Hemisphere that has spectral properties identical to those of the Southern Hemisphere’s Great Red Spot and at about the same latitude reflected across the equator. Why this particular latitude range, 19-23°, consistently harbors the reddest pigment found in the atmosphere is an intriguing question. The Galileo Orbiter has provided evidence that not all of the surface clouds are ammonia. Specifically, the occasional, sudden appearance of a bright-white cloud seen in the filamentary-cloud areas with cyclonic shear (”cyclonic” meaning in the same direction as the planet’s rotation, implying low pressure), is the top of a giant cumulus water cloud that has punched up through the ammonia cloud from the 5-6 bar region below. These water clouds stand about three times taller than cumulus towers on Earth and would be breathtaking to view from the side; the largest ones erupt on the northwest side of the Great Red Spot and cover an area the size of Alaska before they are sheared apart. Such energetic moist convection may hold the key to the transfer of energy from Jupiter’s interior to its atmosphere.
Figure 1. Hubble Space Telescope (HST) image of Jupiter bearing the scars of the Comet Shoemaker-Levy 9 impacts. From left to right, the impact sites are the E/F complex on the edge of the planet, the star-shaped H site, the sites for N, Q1, Q2, and R, and the D/G complex on the far right limb. Also visible is the Great Red Spot and several smaller storms (Hubble Space Telescope/NASA).
Jupiter has hundreds of stable oval-shaped storms (Fig. 2) that populate its anticyclonic (high-pressure) shear zones. They tend to get more numerous and smaller as latitude increases toward the poles. The Cassini flyby data show that the alternating jets extend poleward all the way to at least 7 80°, even though the belt-zone banding gives way to a leopard-skin appearance poleward of 60°. Ultraviolet-filter images that are sensitive to aerosols in Jupiter’s stratosphere show occasional large vortices and wavy boundaries in Jupiter’s stratosphere that are unrelated to the eddies in the troposphere below. Eddies in Jupiter’s atmosphere last much longer than their cousins on Earth, but they do have a life cycle, albeit with a life span closer to three score and ten than the couple of weeks for which major storms remain coherent on Earth. For example, at 33°S, the three White Ovals BC, DE, and FA that emerged from a wavy disturbance in 1939 (which had six pinched areas labeled A through F) existed virtually unchanged for 60 years, until they merged together sequentially in the late 1990s to form a single White Oval, called BA. Like Old Faithful, Jupiter’s Great Red Spot (GRS) seems to have been a part of the planet’s landscape forever—Robert Hooke reported sighting a giant spot on Jupiter more than 3 centuries ago. We can see firsthand how this giant storm persists in the face of dissipation by watching movies of Jupiter’s cloud-top dynamics. They show that the GRS ingests a steady diet of small anticyclones that are born in the cyclonic, filamentary region to its northwest and then travel at latitude 19°S all the way around the planet until they are swept into the GRS from its eastern side. This means that the GRS should not be treated as a local storm but as the most obvious component of a global system in dynamic equilibrium, on which the Sun never sets. It would also be a mistake to treat the GRS as eternal; it has been shrinking in longitudinal extent during the past century. The trend has been precisely measured during the last quarter-century and raises the possibility that the GRS may lose its stable elliptical shape and cease to exist at some point, perhaps within the reader’s lifetime.
Figure 2. Galileo spacecraft image of Jupiter’s swirling storms and stratospheric haze in the vicinity of latitude 50°N. This is a false-color image in which red indicates high-altitude features and blue indicates low-altitude features. North is up and the line of sight looking toward the limb emphasizes the high-altitude haze.
Above the Clouds. Planetary scientists are always hunting for ways to infer vertical information from horizontal (map-view) information that is gathered remotely. Given a rapidly rotating planet like Jupiter, one such important technique allows for the determination of wind shear with respect to height above the cloud tops, where, by definition, there are no visible tracers of the motion. The thermal-wind equation makes this possible; it is a three-dimensional relationship that holds when the Coriolis acceleration strongly couples the atmospheric momentum (wind) to the mass (pressure), such that the horizontal wind runs parallel to the axis of tilt of constant-pressure surfaces and the speed is proportional to the angle of the tilt; atmospheres governed by this coupling are said to be in geostrophic balance. The trick is to make horizontal temperature maps using an infrared camera, because horizontal temperature gradients regions where constant-pressure surfaces are not parallel, and hence regions where the horizontal wind speed changes with height. In Jupiter’s upper troposphere, the anticyclonic structures (high-pressure regions), including both the long-lived vortices like the Great Red Spot as well as the zones themselves, are cooler than their surroundings, about 8K cooler in the case of the Great Red Spot, whereas the cyclonic vortices and belts are correspondingly warmer than average. Both of these trends signify that the winds on Jupiter are decaying with height in the upper troposphere. For example, it is inferred that the peak cloud-top winds in the Great Red Spot, which were about 120 m/s during the Voyager encounters, drop to zero in the stratosphere at around 50 mbar. Numerical models have demonstrated that this trend helps to stabilize the winds against shear instability.
Just as in Earth’s atmosphere, ultraviolet light from the Sun heats up Jupiter’s upper atmosphere to form a stratosphere; the difference is that ozone is the primary UV absorber on Earth whereas methane and a long list of hydrocarbons are the UV absorbers on Jupiter. The abundance of photochemical products in Jupiter’s stratosphere makes its chemistry easier to probe spec-troscopically than either the troposphere below or the thermosphere above. An entire array of hydrocarbons has been identified spectroscopically: ethane, acetylene, methylradicals, ethylene, methylacetylene, allene, propane, diacetylene, benzene—the list continues—plus stable isotopes of the above made with 13C and D (deuterium). There are also trace oxygen compounds in the stratosphere, probably supplied by cometary impacts; the most prominent are carbon monoxide, carbon dioxide, and water vapor; and there are other compounds involving nitrogen, phosphorus, and sulfur. A hotbed of disequilibrium chemistry exists near each pole in the auroral regions that forms polar haze layers. Galileo Probe. The Galileo Probe mission lasted an hour on 7 December 1995 and returned 3.5 Mbits of data to a depth of 200 km below Jupiter’s cloud tops. Probes have been successfully deployed in planetary atmospheres before, including Venus, Earth, and Mars, but the success of this experiment was a milestone in spacecraft engineering because the Probe entered Jupiter’s upper atmosphere at a speed of 47 km/s, the fastest atmospheric entry ever attempted. The only serious problem to occur was that the Probe’s heat insulation failed to maintain the temperature inside it within the design-specification range of — 20° to + 50°C and the temperature varied at a rate of — 5° to + 4°C/min, which was three times more rapid than expected; so the instrument readouts had to be recalibrated using identical twins on the ground (2).
A monitoring campaign at NASA’s Infrared Telescope Facility (IRTF) in Mauna Kea, Hawaii, showed that the Probe hit the southern rim of one of the dozen or so 5-um wavelength hot spots that drift along latitude 7°N at any given time. Hot spots are so-called because they appear bright in infrared images, signaling that they are cloud-free skylights that allow the radiation from the interior to escape to space (”hot” refers to the IR brightness temperature). The Probe’s atmospheric structure instrument, cloud-particle detector (nephelome-ter), and other instruments show that it did not encounter the expected three-tiered layers of ammonia, ammonium hydrosulfide, and water clouds, but instead saw only the barest hint of any clouds at all. Interestingly, the locations of con-densibles were pushed down into the planet in an organized fashion without being mixed together. One theory is that the hot spots are the troughs of large-amplitude waves that encircle Jupiter just north of the equator and that this wave system includes the white equatorial-plume structures.
Analyzing the Doppler shifts in the Probe’s telemetry allowed determining its precise line-of-sight motions. The largest contributions came from the Probe’s vertical descent and the rotation of Jupiter relative to the Galileo Or-biter overhead, but once these effects were accurately removed, a clear signal of the horizontal winds emerged. The Probe encountered an eastward (prograde) wind of 100 m/s at the cloud-top level, as expected, and then the wind speed increased rapidly to approximately 170 m/s at 4 bars; after that, it maintained a similar speed down to the 21 bar level, where telemetry was lost. Modest vertical winds of 1-2 m/s were encountered that are consistent with buoyancy waves (internal gravity waves) in the atmosphere similar to those in Earth’s atmosphere.
What happened to the Galileo probe after it accomplished its 1-hour mission? Because the Probe was vented, it is thought that only a few sealed boxes inside it were crushed once high pressures were encountered, but the rest of the aluminum/titanium structure was probably able to equilibrate to the rapidly increasing pressures during descent. Thus, the ultimate fate was most likely vaporization by the intense temperatures of Jupiter’s interior. First to melt was probably the parachute, made of Dacron, when the probe encountered temperatures exceeding 260°C, estimated at less than an hour after its last received transmission. Once the parachute failed, the probe’s descent hastened, and before 2 hours had elapsed after the end of the telemetry, in an ambient pressure exceeding 280 bars, the Probe’s aluminum components reached 660°C and melted. The melting temperature of titanium is 1,680 °C, so the housing survived for about another 6 hours before melting and breaking up into droplets at about 2000 bars. The droplets then rained downward for another hour or so until they evaporated. From the time it started returning its telemetry until it was completely vaporized, the Probe probably lasted only about 10 hours, or about 1 Jupiter day (3).
Stability of the Jets. Part of the mystery of why Jupiter has so many alternating jet streams, and why they are rock steady, comes from not being able to see the abyssal circulations hidden beneath the cloud tops; this leaves open many possibilities. For example, Jupiter’s jets could be seated in the visible atmosphere, with the internal circulations spun up to a greater or lesser degree as a consequence. Or they could be rooted in the convecting interior where the observable patterns in the cloud tops are shaped by conditions below. Or they could be differentially accelerated at all levels by external satellite tides. Each of these possibilities is competently represented in the scientific literature, and even a mix of all three is possible. The first possibility is that the jets are shaped in the shallow atmosphere because there are intrinsic length scales that bear on the undulation of Jupiter’s constant-pressure surfaces with respect to latitude and hence, on the spacing and amplitude of its jet streams. One is called the deformation radius, Ld = c/\f \ (valid away from the equator), the ratio of the gravity-wave speed, c, to the Coriolis parameter, f = 2Osin(lat), where O is the planets angular velocity. This length scale is attributed to Rossby, when used in the meteorological context and is mathematically analogous to the Debye shielding length in the theory of magnetized plasmas. For vortices, it is the distance beyond which they do not strongly feel each other’s presence. The deformation radius, it is estimated, is of the order of 2000 km in Jupiter’s troposphere, but probably has natural variability of the order of 50% because it is sensitive to the distribution of water vapor, which appears to be quite heterogeneous on Jupiter. Several studies suggest that it is not Ld, but rather a second length scale attributed to Rhines that sets the spacing of jets, which may be written Lp = (Ua/O)1/2, where U is the scale of the horizontal velocity and a is the radius of the planet. In the 1990s, it was shown numerically that persistent jets form in both a wide channel and a full spherical shell whose spacing is controlled by Lp and furthermore, that the number of jets that emerge qualitatively matches the different numbers observed for Jupiter, Saturn, Uranus, and Neptune.
But, the devil is in the details, and there is evidence to suggest that these shallow-atmosphere models all miss one key feature of Jupiters jet streams, namely, they all fail to reproduce the sharpness of Jupiter’s westward jets. Jupiter exhibits several westward jets with curvatures that are a factor of 2 or more greater than the beta parameter, b = df/dy, where y is latitude (expressed in the local-Cartesian sense in units of length). Such sharp westward jets appear impossible to achieve in shallow-atmosphere models where there is little or no deep circulation, because the shear flow tends to become unstable through the process of barotropic shear instability. The dynamics is anisotropic because of the planetary rotation, and there not a similar restriction on the shape of shallow eastward jets. However, obtaining an equatorial jet that is eastward, as on Jupiter and Saturn, also seems to be difficult to obtain from shallow-atmosphere models, whereas the westward equatorial jets of Uranus and Neptune are not a problem.
If instead, the jets extend into Jupiter’s interior, then important new possibilities arise. The Galileo Probe results show that the winds at 7°N do extend into the planet. Furthermore, an analysis of Voyager wind data in and around the Great Red Spot and White Oval BC has produced a family of abyssal circulations covering the latitude range 10°Sto40°S that have the following testable property: a model GRS placed over one of these deep-wind profiles reproduces the observed variation along streamlines of absolute vorticity, q+f, where q is the relative vorticity (the vertical component of the curl of the velocity). These empirically determined circulations happen to correspond to the special case Ld=Lp, when the latter is written Lp = (u/Qy)1/2, where Qy is the gradient of the potential vorticity (quasigeostrophic), which includes the plan-etary-vorticity gradient or beta parameter, b = df/dy = 2(O/a) cos (lat) as above; the relative-vorticity gradient, dq/dy e — d2u/dy2; and the vertical-shear (baro-clinic) term, which is needed when Ld is much smaller than the planet’s radius, as is the case for Jupiter. This special condition may alternatively be stated u = QyLd, and it has significance for two reasons. First, it corresponds to the case of marginal shear stability with respect to a stability criterion that traces back to Kelvin, but is notably absent from most meteorology text topics, and second, it implies significant abyssal circulations that differ from the cloud-top winds. Cross-Disciplinary Physics. Considering that, at some level, there is perfect correspondence between the mathematics that describes the dynamics of rapidly rotating atmospheres and that which describes magnetized plasmas, the type that arises in fusion power-generation studies and have economic significance, it is particularly interesting that the same shear-stability theorem just mentioned also arises in fusion-related experiments. For example, O’Neil and Smith (4) discuss the two branches of shear stability theorems pioneered by Kelvin and proved for nonlinear (large amplitude) perturbations by Arnol’d in the mid-1960s. They point out that one branch is much easier to establish than the other and hence, is much better represented in the plasma literature, but that the harder one is needed to understand shears in their nonneutral plasma column. Completely parallel to this situation, most meteorological text topics establish to one degree or another the stability theorems attributed to Fj0rtoft, Charney, and Stern, and Rayleigh and Kuo (the barotropic stability criterion), which we have listed in order of decreasing generality and which are all related members of the easier branch, while leaving out any word about the second branch, referred to in the literature as Arnol’d's second stability theorem. It is tantalizing that when fusion researchers need a long-lived vortex and stable shears in their plasma column, they have the option of copying how Jupiter does it. Deep Winds. The other reason for the significance of marginal stability of Jupiter’s winds to Arnol’d's second stability criterion is that this condition implies strong, alternating jet streams in Jupiter’s interior that differ from those seen in the cloud tops. By assuming that the 454 m/s speed of the dark ring seen propagating outward from each of the Comet Shoemaker c Levy 9 impact sites is the gravity-wave speed in Jupiter’s atmosphere—not a firmly established fact—a unique member of the family of abyssal circulations mentioned before is singled out that leads to the pre-Probe prediction that Jupiter’s westward jets change little with depth but that its eastward jets increase in strength by 50-100% with depth. This prediction for the eastward jets closely matches the results of the Probe’s Doppler wind experiment, with the caveat that the Probe’s latitude of 7°N was too close to the equator for the strong Coriolis effect assumed by this (quasigeostrophic) theory.
If Jupiter’s jets are deep, then they must be stable simultaneously in the two different geometries that they span, the atmosphere and the interior. There is a promising lead regarding the atmosphere in the form of Arnol’d's second stability criterion, sharp westward jets and all. But the necessary deep winds for this would be prone to barotropic instability themselves, so it seems we have jumped out of the frying pan and into the fire. That would be the case were the deep winds subject to the thin spherical-shell geometry of the atmosphere, but they are not, as illustrated by work on the stability of deep-seated jet streams. Several groups have considered the possibility that Jupiter’s jet streams are rooted deeply where the planet’s internal heat source drives convection and where there is no confinement of motions inside a thin spherical shell. The problem is made all the more intriguing by the Taylor-Proudman effect, which inhibits motions in the direction parallel to the planet’s rotational axis. A series of studies have shown that deep-seated convection can generate alternating jets at the top ofa convecting sphere. Numerical simulations in the 1990s of a rapidly rotating, deep fluid shell achieved a broad eastward flow at the equator and alternating jets at higher latitudes. However, the jets are barely discernible through the large noise of the convection in the work to date. It now seems probable that both a deep-spherical interior and a thin, stable atmosphere coupled together are needed to model Jupiter properly. One recent coupled model has thermally driven convection in the interior that evolves to concentrate motions via “teleconvection” in the stable outer layer. More coupled atmosphere-interior experiments are needed to determine which regime ultimately picks the scale of the jets on the gas giants, or whether it is a negotiated deal.
The third class of hypotheses for Jupiter’s jet streams involves the intriguing possibility that the winds are shaped and accelerated not by internal forces as before, but by satellite tides. If the interior of Jupiter on average is modestly statically stable to convection, tides can couple to it that are dominated by higher order Hough modes. These tend to produce banding that has alternating accelerations of the order of 1 cms ~1 day ~1. The dominant tides come from Io, Titan, Ariel, and Triton, respectively, for Jupiter, Saturn, Uranus, and Neptune. This idea is an area of active research, and it is motivating the search for observational evidence of a tidal response at Jupiter’s cloud level.
Interior
The inside of Jupiter has to be a fascinating place. Whatever the abyssal circulations are that fill the interior, they must certainly be laced into intricate convective patterns that are shaped by the planet’s rapid rotation and strong magnetic field. Most of the interior is metallic hydrogen, which is conductive and the root of Jupiter’s strong magnetic field. This is magnetohydrodynamics (MHD) at its best, and although we have the barest inkling of what is going on, there is great hope in uncovering more, the closer we scrutinize Jupiter.
Jupiter is an order of magnitude too small to fuse hydrogen in its core and thus be classified as a star, but it does emit more radiation than it absorbs. Only 60% of Jupiter’s emitted infrared flux is reradiated solar energy; the rest percolates upward from deep inside the planet, the result of settling and the release of gravitational potential energy that continues 4.6 billion years after the formation of the planet. The temperature reaches about 10,000 K in the interior; the exact maximum value and its profile with depth are not precisely known because they are sensitive to the detailed composition. For example, if Jupiter’s interior contains reduced amounts of the alkali metals, sodium and potassium, then the heat-transfer mechanism could change from convection to radiation where the temperature is in the range 1200-1500 K and the effective opacity of hydrogen and helium dips; such a radiative zone would have implications for the overlying dynamic meteorology and shear stability. But more likely, the alkali metals are not depleted, and they maintain opacity across this temperature range such that the heat-transfer mode is convection throughout Jupiter’s interior.
A low-flying spacecraft that orbited Jupiter a few thousand km above its cloud tops could measure gravitational anomalies on the order of 1 mgal (10 ~ 5ms ~2) and determine the high-order, spherical-harmonic coefficients of the gravity field, thereby illuminating the interior structure of Jupiter; such a mission is on the drawing board. Two other classes of observations that might open up the study of Jupiter’s interior, in the same fundamental way that he-lioseismology revolutionized the study of solar interiors, are the detection offree oscillations and of tidal responses to satellites. Both are being actively pursued by observers, but the signals are weak and to date there have not been clear results. On the other hand, the entire Jovian system is affected by the makeup of Jupiter’s interior, whether the influence be gravitational, magnetic, chemical, thermal, or a combination, and we can turn this fact around to infer properties of the interior indirectly.
Hydrogen: An Alkali Metal. The elements that occupy the leftmost column on the periodic table are the alkali metals, and hydrogen is a member of this group. Converting molecular hydrogen, H2, into monatomic metallic hydrogen in the laboratory is a modern feat that requires extremely high pressure. Most of the interior of Jupiter exceeds this pressure threshold, and hence the behavior of metallic hydrogen has important implications for our understanding of Jupiter’s interior structure and the generation of its magnetic field. By briefly creating shock pressures up to 180 GPa (1.8Mbar) and temperatures up to 4000 K in the laboratory, researchers have been able to measure the electrical conductivity of fluid metallic hydrogen. Currently, there are two differing sets of results that disagree to a level that impacts Jupiter-interior modeling, but it is likely that the picture will clear up soon because more experiments are being performed. Present indications are that the change from molecular to metallic hydrogen is not a first-order phase transition, which would imply an abrupt boundary in Jupiter’s interior and be significant for both atmospheric dynamics and chemistry, but rather is continuous and complex. Some caution must be exercised because the laser-shocks created in the laboratory may yet turn out to be supercritical. The current picture (5) is that hydrogen begins to dissociate around 40 GPa, to form significant metal-like electrical conductivity around 140 GPa; significantly it has the same value of conductivity as the fluid alkali metals Cs and Rb undergoing the same transition, and becomes completely metallic at around 300 GPa. The pressure 140 GPa (1.4Mbar) corresponds to a depth below Jupiter’s surface clouds of only about 10% of the planet’s 71,492-km equatorial radius (earlier it was thought that the depth of the metallic transition was closer to 25% of the radius). Thus, approximately (0.90)3 = 73% of the planet’s volume contains metallic hydrogen. Mapping Jupiter’s higher order magnetic field components using a magnetometer carried by either a satellite in low orbit or a ramjet flying in the atmosphere will provide new information about the interior structure.
Jupiter’s magnetic field is similar to Earth’s in many ways. The dipole tilt for Jupiter is 9.6°, which is only 2° less than the current value for Earth, and if the low-order moments for both Jupiter and Earth are extrapolated inward to their point of origin, the relative strengths of the moments are the same for both planets. Thus, a similar dipole-style dynamo process that involves convection ofa conductive fluid is probably producing the magnetic fields in both planets. One major difference is that Earth’s core-mantle boundary, whose position and character are precisely known from seismology, is abrupt, whereas the outer envelope of Jupiter’s dynamo may not be so. Uranus and Neptune are much different; they have larger quadrupole moments, larger dipole offsets, and larger tilts, so they may have quadrupole-style dynamos. The tilt of Saturn’s field is enigmatic; it is less than 1°; the Cassini orbiter will revisit the question of why Saturn is unique in this regard.
Helium is the most important constituent in Jupiter’s atmosphere after hydrogen, just as in the Sun. A primary result of the Galileo Probe was to determine accurately the helium mole fraction, which is 13.59 + 0.27%. This value is less than that inferred as the original fraction based on models ofthe Sun’s history; so some of the helium has probably rained toward the center of Jupiter. Because neon tends to dissolve in helium drops, it is consistent that the neon mixing ratio is lower on Jupiter than in the Sun. As a rule of thumb, the Galileo Probe found that, except for oxygen, the most common elements heavier than hydrogen and helium on Jupiter are about three-times enriched compared to the Sun. The oxygen concentration on Jupiter appears to be at least solar but is complicated by the fact that it occurs as water, which is a major player in the atmosphere’s dynamic meteorology, and consequently its distribution is heterogeneous.
Satellites
Starting on the inside and working out, Jupiter’s largest sixteen moons are Metis, Adrastea, Amalthea, and Thebe; the four Galilean satellites Io, Europa,Ganymede, and Callisto; then Leda, Himalia, Lysithea, Elara, Ananke, Carme, Pasipha, and Sinope. The Galileo Orbiter has vastly enhanced our understanding and appreciation of these worlds, and there is amazing diversity among the siblings. Io has the most colorful coat and the most influence on the physics of the Jovian system. But, when the topic turns to liquid water and the possibility of extraterrestrial life inside the solar system, Europa is uppermost on everyone’s mind. And then there is the largest, Ganymede, which outclasses two planets and has its own magnetosphere. The smaller satellites are active too; the innermost ones are involved in shaping Jupiter’s ring system. Each satellite has a different story to tell (6), but we begin with the technicolor dreamcoat. Io. The innermost Galilean satellite has stood out for almost a century as an enigmatic moon. Laplace (Mecanique Celeste, Vol. 4, 1805) studied the intriguing 4:2:1 resonance among the orbital periods of Io, Europa, and Ganymede that now bear his name. In 1927, it was noted that Io (Fig. 3) has a pronounced variation in brightness with orbital phase angle. In 1964, there was a report of an anomalous brightening of Io’s surface as it emerged from behind Jupiter. Also in 1964, the Australian meteorologist Bigg discovered that radio waves in the decametric wavelength range emitted by Jupiter showed a modulation that was related to Io’s orbital position relative to the observer. As we now know, the interactions between Io, the plasma torus that encircles Jupiter and is supplied by Io’s vol-canism, and Jupiter’s magnetosphere are complex and contain many feedback mechanisms. In 1971, Io occulted a bright star, providing an accurate determination of its radius and density, both of which are about 5% larger than Earth’s Moon—they could be twins. However, whereas the Moon has a global heat flow of about 0.02 Wm”2, Io’s heat flow is 1 to 3Wm”2. At the start of the 1970s, observers were looking for evidence of a Moon-like surface, perhaps covered with some water or ammonia frost. What was found was discordant photometry at visible and infrared wavelengths, sodium D line emission, and spectral evidence for sulfur, none of which is Moon-like. The Pioneer spacecraft encounters revealed that Io has an ionosphere and a thin atmosphere. In 1975, 2 years before the Voyager encounters, strong absorption near 4 mm was detected that later proved to be sulfur dioxide. In 1979, in support for Voyager, but prior to the encounters, observers discovered intense temporary brightening in the infrared from 2 to 5 mm and evidence that some of Io’s surface is at 600 K compared to the daytime average of 130 K; this result was met with skepticism.
Figure 3. Hubble Space Telescope UV image of Jovian northern aurora: The polar cap. Auroral footprints can be seen in this image from Io (along the left-hand limb), Ganymede (near the center), and Europa (just below and to the right of Ganymede’s auroral footprint). These emissions, produced by electric currents generated by the satellites, flow along Jupiter’s magnetic field and bounce in and out of the upper atmosphere (Hubble Space Telescope/NASA).
And then, just 1 week before the 5 March 1979 Voyager 1 encounter, S. Peale, P. Cassen, and R. Reynolds (7) published a paper entitled ”Melting of Io by tidal dissipation,” in which they predicted, ”Consequences of a largely molten interior may be evident in pictures of Io’s surface returned by Voyager 1,” and ”widespread and recurrent surface volcanism would occur” as a result of Io’s role in the Laplace resonance (7). The general mechanism is as follows. Were it not for the resonant forcing of the Laplace resonance, Io’s orbit would have an eccentricity of only e = 0.00001, which would produce negligible tidal heating, but the value is elevated to e = 0.0043. To the eye, this is still a small eccentricity, and the corresponding elliptical orbit still looks like a circle, but one that has Jupiter not quite in the middle because it is centered on one of the two foci. The main heating comes from the fact that Io turns on its axis with clockwork precision at the average rate of its orbit—it keeps the same face pointed toward Jupiter just as the Moon does to Earth—but slides faster through its orbit when closer to Jupiter (perijove) and slower when farther away (apojove), so that it nods back and forth, causing the tidal stresses and driving the most active volcanism in the solar system.
Voyager 1 found widespread volcanism on Io and no impact craters. The dramatic prediction and swift confirmation of Io’s volcanism is one of the major success stories in theoretical planetary science. Today, there exists a large amount of ground-based and spacecraft observations, including more than a half-dozen close encounters of Io by the Galileo Orbiter that provide a tantalizing view of the plasma physics and magnetohydrodynamics in the Jovian system. Two major classes of volcanoes have been identified on Io. Pele-type (Fig. 4) eruptions are large, up to 300 km high, and deposit relatively dark red material; they occur in a restricted region from longitude 240° to 360°. Prometheus-type eruptions are 50-120 km high, last months or years, and deposit bright white materials; they occur all around an equatorial band.
Sulfur dioxide frost covers much of Io’s surface, and volcanism and sublimation supports a tenuous SO2 atmosphere. Beautiful UV auroras have been observed in Io’s atmosphere. Some of that atmospheric mass is injected into Jupiter’s magnetosphere at a rate that depends in part on the level of Io’s volcanic activity and supplies electrons and S + ,S2 + ,O +, and O2 + ions into a torus of plasma that orbits Jupiter and is appropriately called the Io plasma torus. The plasma is quickly caught up in Jupiter’s magnetic field and then orbits in the same 9.9-hour period at which the planet rotates; this is much faster than Io’s 42.5-hour orbital period. The result is that the plasma streams past Io at a relative speed of about 60 km/s and as a consequence, generates an enormous, 10-million-ampere electrical circuit. A giant, arching structure called the Io Flux Tube connects Io and Jupiter along Jupiter’s magnetic field lines. The location where this tube enters Jupiter’s atmosphere near each of Jupiter’s poles is called the Io footprint and is the site of intense, localized auroral emissions from Jupiter’s ionosphere. The most spectacular images are observed in the infrared at the 3.4-mm wavelength, because the methane in Jupiter’s atmosphere strongly absorbs, and hence eliminates, any light coming up from below the ionosphere at this wavelength, whereas the main ion present, H^, produces several strong emission lines in the same wavelength region. Io’s footprint has been located in UV and visual-wavelength images as well as in IR images. The physical mechanisms involved in the interaction between the ionospheres of Io and Jupiter are not fully understood.
Figure 4. Galileo spacecraft image of Jupiter’s moon Io. The giant active volcano Pele is prominent in this image. Much of Io’s surface is covered by sulfur dioxide frost, and many of the colors may represent allotropes of sulfur.
Ganymede. The largest planet in the solar system fittingly has the largest satellite, Ganymede. It is larger than Pluto and Mercury and maintains its own magnetosphere, like a bubble inside of Jupiter’s magnetosphere; it has a field strength an order of magnitude stronger than the typical 120-nT Jovian field strength at Ganymede’s orbit. There is a trace atmosphere of oxygen coming from the 50-90% ice surface. Data from two Galileo close encounters reveal that Ganymede (Fig. 5) has the lowest measured moment of inertia of any solid body in the solar system, meaning that most of its dense material has sunk to the center, probably forming a silicate core with an iron center and leaving a thick icy mantle outside. It is not surprising that Ganymede’s geology is as complicated as any planet’s. About 60% of the surface has been reworked into bright, grooved terrain and shows evidence of strains larger than 30% that have torn its craters apart. The other 40% of the surface is heavily cratered and is darkened by a thin layer of residue or slag that is probably left over from sublimation of the ice. Callisto. A surprise of planetary exploration has been that each of the more than 60 moons in the solar system is different. A case in point is Callisto, which unlike its three Galilean-satellite siblings, has a heavily cratered surface and almost no trace of other geologic processes, except that some of its small craters (<1km) appear to have been broken apart into large blocks by an unknown agent of erosion. One theory for the uniqueness of Callisto is that it is the only Galilean satellite that has never experienced orbital resonance with its siblings. Callisto’s large craters tend to be fatter than craters seen in the inner solar system. Spectroscopic observations show that there is plenty of ice worked into the surface, about 50% on average; this is also not Earth’s Moon, but there are regions on Callisto that are free of ice. Not much can be said about its interior structure, to what extent it is differentiated into a silicate core and an icy mantle, but its moment of inertia is somewhat lower than that of a uniform sphere, implying some differentiation. Given Callisto’s old-man appearance compared to Europa and Io, it was something of a surprise when two Galileo close encounters suggested that Callisto actively responds to the constant changes of Jupiter’s magnetic field (due to Jupiter’s rotation and magnetic-field tilt). An induced dipole that flips with the forcing from Jupiter is evident, which implies a conducting layer somewhere inside the satellite. A trace amount of ammonia would have a strong antifreeze effect and is a distinct possibility for keeping a layer of liquid inside Callisto and could also explain the erosion of the smallest craters. But researchers are not sure how the satellite could avoid greater differentiation if this is the case. A lander might be able to solve the mystery by producing a longer magnetic field record and by sampling the surface.
Figure 5. Galileo spacecraft image of Ganymede showing a line of 13 closely spaced craters. A fragmented comet similar to Comet Shoemaker-Levy 9 probably caused this feature.
Europa. Europa is one of the smoothest spheres ever produced in nature. One has to search hard to find its handful of craters. Europa is tidally heated by the same Laplace resonance that drives Io’s volcanos, and it is clear that its surface has been worked and reworked many times. Some of the surface is mottled and composed of a jumble of blocks, called chaos terrain, that probably indicates local melting of ice. On the large scale, there is a remarkable series of dark lines that crisscross the surface. Many causes may contribute to the stress field that produces these, including nonsynchronous rotation, polar wander (a tendency for the polar ice to thicken such that the ice shell’s global orientation becomes unstable), global contraction or expansion, or tidal stress. Europa gets between 100 and 1000 times the radiative flux that Earth’s Moon receives from the solar wind, causing significant chemical changes in new surface material in less than a decade.
The magnetic signature around Europa suggests that there is an induced dipole from Jupiter’s time-variable magnetic field, which implies a briny subsurface ocean. But the data are enigmatic, and a fixed dipole cannot be ruled out. Taken as an ensemble, essentially all of the features associated with Europa’s geology, magnetic response, and even the presence of salty contaminants on its surface, suggest that liquid water lies beneath the surface. However, each indicator can also be explained individually without liquid water, so one must guard against overinterpretation. Even so, the possibility of an extraterrestrial liquid ocean has policy implications; for example, after its useful lifetime, the Galileo orbiter will be scuttled into Jupiter’s atmosphere to eliminate any possibility that it might one day collide with Europa. A new mission to Europa is high on the NASA’s priority list; the goal is to settle definitively the question of a liquid ocean. Part of the excitement lies in the fact that ecosystems exist on Earth that do not depend on sunlight; the most relevant are the varied life-forms that derive their energy from “black smoker” hydrothermal vents in the deep oceans. It is not hard to conceive of such vents inside of Europa, but there is a problem involving the chemistry needed to harness that energy. The oxidants in Earth’s hydrothermal vents are ultimately tied to atmospheric oxygen through the overturning action of plate tectonics. Without such cultivation, Europa’s mantle would become a reducer rather than an oxidizer, and its vents would emit methane (CH4) rather than carbon dioxide (CO2), which would be poisonous to Earth’s vent ecosystems and presumably also to Europa’s. But, when the subject is life, at least on Earth, there always seems to be a way; for example, rare bacteria exist that use hydrogen to reduce minerals, and similar organisms might not find Europa completely inhospitable.
Small Satellites. Jupiter plays host to at least an additional 30 small, outer satellites and also to Trojan asteroids that lead or trail the planet by 60° in the same orbit (at the stable L4 and L5 Lagrange points, respectively). The orbits of the outer satellites tend to be large, have high eccentricity and high inclination, and often have retrograde orbits. The gravitational reach of Jupiter, its Hill sphere, is about one-third of an AU or just over 700 Jovian radii. The orbits of the outer satellites average about half this distance, which makes them prone to perturbations. Not much is known yet about their chemical composition, but there is the suggestion in color data that they may form groups of fragments that come from distinct parent bodies. Jupiter also has a tendency to grab comets into an orbit around itself at the rate of a few per century. The most famous is Comet Shoemaker-Levy 9, which was captured by Jupiter and immediately afterwards collided with the planet; every piece of telescopic glass in the solar system was pointed to witness the event. Such collisions may occur a few times per millennium. Jupiter’s Trojans could possibly be snow-line planetesimals, in which case they would be as valuable as comets for uncovering facts about the protoplan-etary nebula.
Rings and Dust. One goal of the Voyager 1 encounter with Jupiter was to discover the existence of any faint ring or rings (Fig. 6). No major rings had ever been seen in the backscattered sunlight that is visible from Earth’s inferior orbit, but this did not rule out a ring made of dust, which is best viewed in forward-scattered sunlight, as anyone will attest who has driven into a sunrise or sunset with a dusty windshield. The goal was met handsomely by a single image of the dark side of Jupiter where the forward-scattered sunlight illuminated a distinct ring. The same method was successfully used by Voyager 2 at Uranus and Neptune to capture images of the dusty components of their narrow-ring systems. The discovery at Jupiter was followed up 4 months later by an imaging sequence, customized on-the-fly for Voyager 2. Today, Jupiter’s ring can be imaged directly from Earth using infrared telescopes.
All four gas giants have ring systems, whereas none of the terrestrial-class planets have them; this suggests that a ready supply of orbiting mass is needed to make a ring. Saturn’s ring system is the most majestic in the solar system, but each shows structure as a function of orbital radius, sometimes gaps and sometimes regions of enhanced density, that are the result of resonances between the orbital elements of the ring particles and a particular satellite, or pair of satellites. Jupiter’s dusty ring system harbors an additional feature; it contains resonances with Jupiter’s inclined and rotating magnetic field, called Lorentz resonances, a prime example of the way the dynamics of orbiting dust is affected by electromagnetic and gravitational forces simultaneously. There are three components of Jupiter’s ring system. The main ring is flat and has a sharp outer edge at an orbital radius of 129-130 km and is about 7000 km wide. The ring’s dust supply comes from Adrastea and Metis. The two other satellites orbiting outside the main ring, Amalthea and Thebe, guard two faint rings called the gossamer rings; Galileo images reveal structure even in these rings. The out-of-plane thickness of each ring matches the inclination of its associated satellite, and Galileo images show dust coming from Amalthea and Thebe that supplies the gossamer rings. The third component is called the halo, which starts at the main ring and flares out as it extends about 20,000 km inward to the midway point with the planet’s cloud tops.
Figure 6. Sequence of Infrared Telescope Facility (IRTF) images of Jupiter, its ring, and the two inner satellites Metis and Amalthea. The images span 2 hours; time increases from upper left to lower right. Metis first appears in the second image, following transit across the face of the planet. The brighter satellite, Amalthea, first appears in the third image before transiting across the planet.
Jupiter’s dust environment is as well sampled as Earth’s—five dust detectors have flown through the Jupiter system to date. The Pioneer 10 and 11 spacecraft flew by Jupiter in 1973 and 1974, respectively, and recorded a thousandfold increase in dust as they entered the Jovian system. Neither Voyager 1 or 2 carried a dust detector, but in 1992, the Ulysses spacecraft passed through the Jupiter system to achieve its primary mission of a high-solar latitude orbit around the Sun and carried a dust detector that had greatly increased sensitivity compared to the Pioneer instruments. It discovered bursts of dust coming from the Jovian system that extend far into interplanetary space, and these were also recorded by a similar instrument carried into Jupiter orbit by the Galileo spacecraft. Time series of the dust hits on the various detectors show a strong 5-hour oscillation that is modulated by the passing of Jupiter’s magnetic equator over the spacecraft. Analysis of the dust streams emanating from Jupiter show that charged dust particles interact with the interplanetary magnetic field, so that the paths of smaller particles are bent more than the paths of larger ones, just as in a mass spectrometer.
Each time the Galileo Orbiter passed close by one of the major satellites, it detected a jump in the dust concentration, and when this happened, the speed of the dust hitting the detector matched the speed of the spacecraft relative to each satellite. This is how the existence of clouds of dust surrounding each moon was discovered. These clouds are probably stirred up by a continuous bombardment of small meteoroids. The active volcanism on Io provides a much greater source of dust, and in addition, the Io plasma torus is a major source of charge for the dust caught up in the Jovian system. Because it is an orbiter instead of a flyby spacecraft, Galileo has been able to spend years mapping in detail the dust distribution inside the Jovian system, including clouds of dust around the Galilean satellites and the interaction of interplanetary dust with Jupiter’s powerful magnetosphere. A unique opportunity came when the Cassini spacecraft, bound for Saturn but making a close encounter with Jupiter in 2000 to gain a gravity assist, carried the most sensitive dust detector yet through the system. The result was that one particular dust stream was observed in situ at two different times and positions by Cassini and Galileo.
We can expect even closer scrutiny of Jupiter, its satellites, and its magnetosphere in the decades to come, as Europa is fully investigated and as the high-order spherical harmonics of the planet’s gravitational and magnetic fields are used to peer into its interior. One day, we may hope to appreciate Jupiter’s role fully as protector of the inner solar system and king of the planets.
