The number of known exoplanets rose from zero to one in the mid-1990s, and has been doubling approximately every two years ever since. Although this can justifiably be called the beginning of an era, an earlier era began in the 1960s when humankind began exploring the Solar System with spacecraft. Even earlier than that, the era of modern scientific study of the Solar System began with Copernicus, Galileo, Brahe, Kepler and Newton. These eras overlap in time, and many individuals have worked across all three. This Review explores what the past can tell us about the future and what the exploration of the Solar System can teach us about exoplanets, and vice versa. We consider two primary examples: the history of water on Venus and Mars; and the study of Jupiter, including its water, with the Juno spacecraft.
The telescope allowed Galileo1 to observe the phases of Venus as it went around the Sun and the orbits of the moons as they went around Jupiter. His observations helped to establish the Copernican heliocentric view2. The telescope was an advance in the technology of its day. Further knowledge of the orbits came from the careful naked-eye (non-telescopic) observations of Brahe and their distillation by Kepler3 into his three laws of planetary motion. Newton4 provided the mathematical model, but he couldn't calculate the planetary masses. All he could say was that Jupiter's density was several times smaller than Earth's — a conclusion that followed from the sizes of the planets relative to the sizes of their moons' orbits. Revealing the actual masses and densities had to wait until Cavendish5 measured Newton's gravitational constant in the laboratory. The astronomical unit — the dimensions of the Solar System — came from timing the transits of Venus from widely separated points on Earth. The first observations of exoplanets6,7 used the radial velocity method, which gives the period of the orbit and a lower bound on the mass. The masses and densities of exoplanets came from the depths of transits8 combined with radial velocity data. Both methods were made possible by advances in the technology of the day.
The ability to measure planetary composition, temperature and cloud coverage took longer, both for Solar System planets and exoplanets. Early infrared observations9 suggested that Jupiter radiated more power than it absorbed from the Sun. Methane and ammonia were detected on Jupiter in 193410, but the main constituent, molecular hydrogen, was not detected until 196011. Meanwhile, the discovery of Jupiter as a radio source12 suggested that Jupiter had radiation belts and a magnetic field. Three years before Pioneer 10 made the first flyby of Jupiter, a three-layer cloud structure for Jupiter had been proposed13 — an ammonia ice cloud at the top, a cloud of ammonium hydrosulfide (NH4SH) in the middle, and a water ice cloud and possibly a liquid water cloud below. That model is still being tested. Transits allow exoplanet spectroscopy, but clouds in the planets' atmospheres often interfere with the detection of gaseous species. Atomic sodium has been detected12, and so has water13,14, but the generally flat spectra imply obscuration by clouds15. Transits allow one to measure the albedo of the planet and its thermal emission. Asymmetries in the phase curves16 provide evidence of winds, but estimates of the wind speeds depend on dynamical models17.
The 1960s marked the end of the first era and the beginning of the second. Our knowledge of Solar System planets then was comparable to our knowledge of exoplanets now. The first era took 400 years and the second about 25. Advances in technology set the pace. Observations and theory are both important, but observations lead. Our Earth-based experience did not prepare us for the richness of the Solar System, and our Solar System experience is already proving inadequate as we discover the richness of exoplanet systems. Without observations to lead us, we are unlikely to ask the right questions or get the right answers from theory alone.
In this Review we use Venus, Mars and Jupiter for examples of long-standing questions that have only partially been answered. Water is the common theme. Because oxygen is the third most abundant element after hydrogen and helium, water is present in many processes ranging from meteorology to geology, and from planet formation to life. Water also defines the habitable zone for exoplanets — the place where water is stable as a liquid. As stated earlier, the objective of this Review is to see how Solar System exploration and exoplanet exploration can benefit each other.
The runaway greenhouse on Venus
Venus is similar to Earth in bulk properties such as mass and density. From their densities one infers that they have similar proportions of rocky mantles and metallic cores. Also, their mean orbital radii are more nearly equal than those of any other pair of planets in the Solar System. The incident sunlight at Venus is 1.91 times that at Earth, but Venus absorbs less power per unit area because the Bond albedo of its sulfuric acid clouds and massive CO2 atmosphere is so high (∼90%)18.
With these similarities, one might expect that Venus and Earth would have similar amounts of volatile elements like oxygen, carbon, nitrogen, hydrogen, sulfur and the noble gases. By analogy with Earth, one would expect these elements to be tied up in compounds like water, carbon dioxide, molecular nitrogen, sulfates and sulfur dioxide. One might expect a warmer, wetter climate on Venus because of the higher incident sunlight, although perhaps the higher albedo would cancel this effect. There are many similarities, but there are also profound differences. First, there are no oceans and very little water vapour in the Venusian atmosphere. Second, its surface temperature is 730 K, which is hot enough to melt lead and warmer than the critical temperature of water.
The volatile inventories for Venus, Earth, and Mars are given in Table 1. Here ‘volatile' means capable of having a substantial vapour pressure at planetary temperatures. The units are mass of volatile per unit mass of planet. The total amount includes the atmosphere, oceans, frost and minerals such as carbonates in the crust and mantle. On Earth carbonates have formed from CO2 gas dissolving in the ocean and weathering the calcium- and magnesium-bearing igneous rocks to form calcium and magnesium carbonates. The CO2 in carbonate rocks on Earth is comparable to the massive CO2 atmosphere of Venus, where the surface pressure is 90 bars. Although the Venusian atmosphere is only 3.5% N2, the total amount of N2 is around three times that found in Earth's atmosphere. For the purpose of this discussion, the N2 amounts are comparable. Finally, the amounts of 40Ar are comparable. This indicates that the amounts of outgassing are comparable because 40Ar is a product of radioactive decay of 40K in crustal rocks.
The big discrepancy in Table 1 is water. If Earth's oceans were vapourized but did not escape, the atmosphere would be almost entirely water vapour and would have a surface pressure of 260 bars. Venus has 10,−5 times less water than Earth. Was Venus born dry? The volatiles in the inner Solar System are thought to have condensed in the outer Solar System and come in later, after the Earth had cooled down. The original theory involved icy planetesimals, which are assumed to be larger versions of today's comets19,20. Recent geochemical evidence, particularly the high deuterium-to-hydrogen (D/H) ratio in comets compared with the Earth, favours an asteroidal origin in the form of carbonaceous chondrites21,
Water vapour is a potent greenhouse gas, and its abundance in Earth's atmosphere is controlled by an equilibrium with the oceans. A warmer ocean means more water vapour and a stronger greenhouse effect — an example of positive feedback. As a result, there is a maximum emitted flux that an atmosphere with a fixed relative humidity can radiate26. This flux is reached when all the spectral windows are closed and the surface is no longer able to radiate directly into space. Further warming of the ocean merely raises the altitude of the emission, but it does not change the emitting temperature. If the absorbed sunlight was greater than the maximum emitted flux, the planet could not be in equilibrium. Spectral windows — where the infrared opacity is low — complicate the picture27,28, but for most gaseous mixtures characteristic of terrestrial planet atmospheres, the critical orbital distance is between Earth and Venus. Instability means that a steady state with an ocean is impossible. If oceans were present initially, they would evaporate and the pressure would be 260 bars. The feedback stops only when the reservoir runs dry.
A water vapour atmosphere is vulnerable to photodissociation25,29, in which ultraviolet radiation from the Sun breaks water molecules into H and OH. Hydrogen, the lighter gas, escapes into space, leaving behind the O and OH to combine with the iron in surface rocks. Or perhaps they combine with outgassed CO to make CO2. It is possible that Venus lost an ocean's worth of water but Earth did not because Earth was too far from the Sun for the instability to develop.
Evidence for this is found in the deuterium to hydrogen ratios shown in Fig. 1. The D/H ratio on Earth is less than that of comets, close to that of meteorites, and greater than that of the primordial Sun and giant planets. The D/H of Mars is seven times that of Earth; for Venus it is around 150 times larger, which is a huge discrepancy. High D/H could come about during escape to space because the lighter H escapes faster, leaving more of the heavier D behind. The high D/H of Venus suggests that Venus lost a large amount of water, but estimating how much is difficult. The fractionation factor — the ratio of the escape rate for D to the escape rate for H — depends on many unknown factors, so we can't say for sure that Venus once had an ocean's worth of water. The history of water on Venus is one of the enduring questions about climate history and the evolution of terrestrial planets.
Many studies use the runaway greenhouse idea to define the inner boundary of the habitable zones of exoplanet host stars. The Solar System provides a fairly imprecise boundary — probably between Venus and Earth — and only for a single star. Perhaps someday scientists will have so many well-documented exoplanet pairs spanning this boundary that they will tell us precisely where the Solar System's habitable zone lies. That would be an example of exoplanet research informing human beings about their own Solar System and its history.
Liquid water on Mars and the faint young Sun
Starting in the late 1960s, spacecraft have observed large-scale fluvial flow features dotted with impact craters on Mars, suggesting that water flowed there in ancient times. Valley networks with braided tributaries indicate precipitation falling over large areas30. Deltaic sediments at the termination of channels suggest standing bodies of liquid water31. The oldest terrain, the Noachian — which indicates the first billion years of Mars history based on the cratering record — shows many signs of aqueous alteration32,33. Phyllosilicates (clays), which require liquid water and near-neutral pH conditions to form, are the best example34. Whether these conditions existed on the surface or in the subsurface is a debated question35. The next-oldest terrain, the Hesperian, was characterized by drier, more acidic conditions and minerals like evaporates and sulfates, which also require liquid water to form32. Volcanoes from this period left their mark on the surface. The current period, the Amazonian, is characterized by cold, hyper-acid, oxidizing surface conditions and little chemical weathering. Whether these changes were driven from within the planet, at the surface, or outside, is uncertain.
Although today the surface temperature of Mars at noon on the equator can rise to a comfortable 20 °C, there is no ice or liquid water there. The dew point — the temperature of the atmosphere where water vapour would condense — is never more than about 205 K. This value is set by the polar ice caps during their warmest season, which is early summer36. Any water, ice or liquid, cannot be in equilibrium with the atmosphere at a temperature higher than the dew point. Liquid water might exist beneath the ice or seeping out of the ground. Transient dark streaks and gullies37,38 running down hillsides may involve liquid water, especially if the water is salty enough to remain liquid at Martian temperatures39,40. There is now spectral evidence for hydrated salts in areas of recurring slope streaks41.
The evidence for a warm, wet early Mars has problems, however, because stellar models42 and observation of Sun-like stars43 suggest that the Sun's luminosity during the first billion years was 20–30% lower than today's value. This is also known as the faint young Sun paradox44. Theories to ‘solve' this paradox have focused on massive amounts of greenhouse gases in the early atmosphere. Scientists have tried putting several bars of CO2 into their atmospheric models44,45 — much more than the amounts given in Table 1 — and have artificially heated the planet enough to vapourize the permafrost, thereby creating a substantial greenhouse effect due to the combined effects of water and CO2. But at Mars's distance the faint young Sun is not powerful enough to maintain this greenhouse. The water rains out, and the greenhouse effect of CO2 is not enough to keep the planet warm. In other words, Mars reverts to its cold, dry self32. Models investigating the cold, dry early Mars46 hold that transient warming caused the melting of snow and ice deposits and thus a temporarily active hydrological cycle, leading to erosion of the valley networks and other fluvial features. Precise details of the warming mechanisms remain unclear, but impacts, volcanism and orbital forcing are all likely to have played important roles.
A problem with greenhouse theories, whether steady-state or transient, is locating the CO2 and water today. The masses of CO2 and H2O ices at the surface of Mars (Table 1) have been measured with good precision from the topography of the polar caps and from ice-penetrating radar, which can distinguish water ice from CO2 ice. Uncertainty in the mass arises from the other ice reservoirs. The large areas down to 45° latitude in both hemispheres are soils with 35–50% permafrost by mass47. The thickness of the permafrost soils is unknown, but it is at least 1 m. A similar problem exists for CO2. The CO2 in the atmosphere is about equal to the CO2 in the ice caps, and together they make up the CO2 total shown in Table 1. There may be buried carbonates on Mars, although they are not plentiful at the surface34. And there may be buried CO2 frost, although it is much more volatile than water and is more likely to sublimate into the atmosphere. As a fraction of the planet's mass, the known amount of CO2 is much less than that of Venus and Earth.
Escape to space is a possibility. Mars is a small planet with a relatively low gravitational binding energy. It lacks a global magnetic field to protect its atmosphere from solar wind. A range of escape processes — Jeans (thermal) escape, hydrodynamic escape, photodissociative escape and stripping by solar wind — might have combined to drive off the massive amounts of CO2 and H2O postulated for the early Martian climate. The MAVEN spacecraft, which is currently in orbit around Mars, is dedicated to studying these processes as they operate today48,49. Observations of Mars-sized exoplanets are still very rare50 and the characterization of their atmospheres and surfaces is not yet possible, but future developments in exoplanet research will perhaps give us an ensemble of Mars-like planets in various stages of evolution, which may indicate what happened on our Mars long ago.
Water on Jupiter and the Juno spacecraft
Measuring the global abundance of water and its spatial distribution is a prime objective of the Juno mission, which is currently orbiting Jupiter. The first results of the Juno mission are being prepared for publication. Here we review some of the basic questions that the mission is addressing. Water is important for understanding Jupiter's internal structure, it is vital for meteorology as a source of latent heat and power for storms, and it may have played a role in delivering other volatiles to Jupiter and spreading them around the Solar System. The Galileo probe found that Jupiter is enriched in the volatile elements C, N, S, Ar, Kr and Xe, relative to solar element-to-H ratios51,
First, the helium is somewhat depleted; that is, its enrichment factor is less than 1. This has been explained54 as a result of helium rain in the metallic hydrogen interior of Jupiter. As the planet cools, the helium condenses and falls towards the centre. Since the atmosphere is well-mixed by convection, helium is also depleted in the atmosphere. The helium raindrops are thought to remove neon, which would account for its very low enrichment factor in the atmosphere. The second surprise is the similarity of the enrichment factors for C, N, S, Ar, Kr and Xe, which fall in the range of 2.1–4.9. How all of these elements, with their different chemistries and different degrees of volatility, could be enriched by similar factors is a major puzzle.
Another puzzle for Jupiter is that the enrichment factor for O is so low, although this could be just bad luck. The Galileo probe went into a ‘5 μm hot spot', a dry spot in which the absence of clouds allows thermal radiation at 5 μm to escape from deeper and warmer levels. Since water condenses deeper than NH3 and H2S, its depletion down to the deepest levels probed by Galileo, despite the enrichment of the other gases, could be explained. This explanation is plausible but has not been tested. The global depletion of water would be a major surprise because one would expect oxygen to behave like the other reactive elements C, N and S during Solar System formation.
If O had an enrichment factor in the 3–5 range, one could explain the enrichment factors of C, N, S and O by forming them into solid compounds during Solar System formation and blowing 60–80% of the original H and He away. The early solar wind seems capable of doing this, but it doesn't work as well for the noble gases.
The problem is that Ar, Kr and Xe remain gases down to very low temperature (for example, 35 K for Ar). At higher temperatures, Ar would get blown away with the main gases H and He, and the Ar enrichment factor would therefore be 1. The observed enrichment factor of 2.5 is a mystery, but the solution may hinge on water abundance. Ar and the other noble gases could have ridden in on icy planetesimals from the outer Solar System55, where temperatures are 35 K and below. Or they could have become trapped at low temperature in a hydrate clathrate56, which is a water lattice with molecules of other gases in the interstices. The clathrate hypothesis requires a very high global water abundance, with an O/H enrichment factor greater than 10. Either way, in this hypothesis the water comes in with its load of noble gases from the cold outer Solar System after the giant planets have formed. Again, the global depletion of water would be a puzzle because it would leave the noble gas enrichment factors unexplained.
This is where Juno comes in. It is equipped with a microwave radiometer (MWR)57 that measures thermal emission arising from depths up to 100 bars and temperatures up to 800 K. This is the region well below cloud base, which is probably at 6–10 bars. This means Juno's MWR is likely to provide a representative sample of the global water abundance. Water shows up mainly in its effect on the lapse rate — the temperature decrease with altitude when water is condensing, compared with that for a dry atmosphere. The direct effect of water on the opacity of Jupiter's atmosphere is small compared with that of ammonia. Finding places where water is condensing is therefore an important part of the MWR plan. Another instrument on Juno, the Jovian Infrared Auroral Mapper (JIRAM), can detect water in thermal radiation emitted at 5 μm58. However, JIRAM only sees down to pressures of ∼5 bars, where water is condensing, so it cannot measure the global water abundance.
Another goal of the MWR is to find evidence of large-scale circulation below the clouds. The MWR will be able to see ammonia varying with altitude and latitude, perhaps in patterns related to those in the visible cloud bands, down to the 100-bar level. This is relevant to another long-standing question about the giant planets: whether the winds are shallow or deep. A rotating fluid with a well-mixed (that is, isentropic) interior could have its zonal jets — the east–west winds in the visible atmosphere — extending deep down on concentric cylinders aligned with the planetary rotation axis59,60. Weak mixing and a stable stratification destroy the concentric cylinders, which means the zonal winds could be shallow. In contrast, finding variations in ammonia and water down to 100 bars — well below the depths heated by sunlight — would imply deep winds.
Comparative planetology of the gas giants will be possible by comparing Juno's results at Jupiter with Cassini's results at Saturn. Cassini has been in orbit around Saturn since 2004, and it would be impossible to cover all of its discoveries here. Nevertheless we mention a few. For Saturn, there was no entry vehicle like the Galileo probe — the Cassini spacecraft will not survive its final plunge into Saturn on 15 September 2017 — and no six-channel microwave radiometer like Juno's MWR. Thus we must rely on infrared and visible spectroscopy, and only methane (CH4) and phosphine (PH3) have been measured with enough confidence to infer global abundances61,
Viewed from Earth, both Jupiter and Saturn exhibit large-scale cloud bands and associated jet streams that circle the planet at constant latitude. Viewed from above the poles (Figs 2,3), Jupiter shows a more chaotic structure, which is different from that on Saturn (Fig. 4). Since Juno is the first spacecraft to fly along the terminator directly over the poles of Jupiter, these images are unprecedented. Instead of parallel bands, Jupiter's poles have irregular compact structures with diameters of up to 8,000 km. One such feature straddles the terminator in the north; its ring of clouds catches the sunlight beyond the terminator, which implies its clouds protrude 50–100 km above the average. Unlike Saturn's north pole, there is no hexagon-shaped feature — a meandering jet stream near 75° latitude — and no compact circular structure occupying the region from 89° to the pole. Because of the small tilt of Jupiter's rotation axis relative to its orbit axis, the poles are close to the terminator, although the south pole is in sunlight and the north pole is in darkness.
Interiors and magnetic fields on Jupiter and Saturn
The mass of Jupiter's core and the distribution of heavy elements are relevant to its formation and cooling history. The basic data are the non-spherical components of the planet's gravitational field. To interpret the data one would like Jupiter's temperature and composition, together with a reliable equation of state. The MWR will measure temperature and the abundance ratios of water and ammonia at 100 bars. Since Jupiter is likely to be well-mixed and adiabatic below 100 bars, the measurements provide a useful constraint on interior models.
The planet's gravitational field affects the orbit, which is determined by analysing the Doppler shift of the radio signal between the spacecraft and Earth. Juno's highly inclined orbit and low altitude at the equator are critical for revealing the non-spherical components of the field. Cassini follows a similar path, with similar objectives, leading up to its final plunge into Saturn. Rotation makes the planet oblate; it produces a bulge at the planet's equator that shows up in the low-order axisymmetric components of the gravitational field. A planet with a large fraction of its mass in a dense core will have smaller values of these components than a planet with a more uniform mass distribution. This is because the core is less affected by rotation — it stays nearly spherical despite the planet's spin. Theories of giant planet formation64 say that a core 10–20 times larger than that of the Earth is needed to start the runaway accretion of gas onto the planet while it is forming. Saturn has a core of about this size65. A core this size inside Jupiter would be harder to detect because it would be a smaller fraction of the planet's mass. In fact, Jupiter's core has not yet been measured, which makes it an important objective of the Juno mission.
Differential rotation — different parts of the planet rotating at different rates — shows up in the higher-order components of the gravitational field. If the interior were chemically homogeneous and isentropic, winds at the cloud-top level would be the surface manifestation of differential rotation on concentric cylinders extending deep into the planet. Balancing the centrifugal force of the winds would require a redistribution of mass inside the planet. Since Jupiter and Saturn have 4–6 zonal jets in each hemisphere, the winds would show up as higher-order terms in the gravitational field — a ring of extra mass associated with each jet66. With their close-in orbits, Juno and Cassini are capable of measuring these terms.
Both Juno and Cassini have magnetometers for measuring the magnetic field at the spacecraft, which enables one to separate the dynamo field that is generated inside the planet from that generated by currents in the magnetosphere67. The roughness of the magnetic field at the surface — the departure from a planet-centred rotationally aligned dipole — is greater when the field is generated at shallow depths. Measuring the roughness allows us to learn where and how the field is generated. Dynamo fields occur in objects of various sizes, and all the giant planets have them. Earth and Mercury also have them, but Venus and Mars do not. Even Ganymede, a moon of Jupiter, has a dynamo field. Saturn's field is unique68,69 because it is axisymmetric and aligned with the rotation axis to within 0.1°. By narrowing the upper limit on this angle, Cassini will be investigating the uniqueness of Saturn's field. A misalignment would appear as a small daily wobble in the field, which would indicate the rotation rate of Saturn's interior and aid in the interpretation of the low-order axisymmetric components of the gravitational field, as described above for Jupiter.
Planetary magnetospheres are low-density regions populated by charged particles that are protected from the solar wind by the planet's magnetic field. For Jupiter and Saturn the bulk of the plasma comes from the geologically active moons Io and Enceladus, respectively70,71. The volcanoes of Io spew out plumes of SO2, some of which escapes into orbit around Jupiter. The gas gets ionized and picked up by the Jovian magnetic field as it sweeps past the slower-moving moon. The heavy ions of the Jovian plasma are therefore mostly S+, O+, S2+ and O2+. Similarly, the cracks at the south pole of Enceladus spew out plumes of water vapour and salty ice, with a few per cent of other gases such as CO2, CH4, NH3 and higher hydrocarbons up to benzene. The heavy ions of the Saturnian plasma are therefore mostly O+, H2O+ and OH+.
The magnetic field tries to hold on to the plasma and maintain its co-rotation with the planet. The plasma flows outwards, since it is constantly being resupplied by the moons. To maintain co-rotation the outward-flowing plasma must gain angular momentum, which requires a j × B force directed east by a radially outward current72. Here j is the current and B is the magnetic field, which points downward at the equator. At the outer edge of the co-rotating plasma disk the current flows out of the equatorial plane, poleward along the field lines, whose electrical resistance is essentially zero. The field-aligned currents plunge into the planet's ionosphere at high latitudes to produce an aurora. Collisions with the neutral atmosphere try to keep the ionosphere and magnetosphere rotating to replace the angular momentum lost to the plasma disk. This drives a current in the ionosphere towards the equator and closes the circuit with the plasma disk. The currents at Jupiter have not been fully measured, because previous spacecraft have not flown repeatedly through the field lines in the polar regions, as Juno will. Cassini, during its grand finale, will fly over the poles of Saturn closer than it ever has before. Measuring the field-aligned currents is important for understanding how co-rotation is maintained and where it breaks down. In addition, other sources of auroral electrons may exist, and by flying through the auroral field lines Juno and Cassini will identify those sources.
All these measurements will provide a greater global understanding of gas giant planets. In this field, the potential for planetary–exoplanetary cross-studies is huge, as gas giants constitute a sizable part of the easily observable exoplanets. Again, close-ups like those given by Juno at Jupiter and Cassini at Saturn will permit detailed studies of their physics, formation and evolution, whereas exoplanet monitoring will give statistics and the possibility of studying regimes that do not exist in our Solar System.
We have mentioned some of the ways that exoplanet research and Solar System research can benefit each other. The analogy with Earth science and planetary science is appropriate. Earth science provides the knowledge base on which planetary scientists draw. Earth science has more data, but planetary science addresses the larger questions. Planetary science takes familiar processes and provides extreme examples, which sometimes are the key to understanding. Exoplanet research does the same thing for both planetary science and Earth science, but it has a statistical advantage: there are many planets out there, and they all have something to tell us. The Solar System, for all its diversity and many small bodies, has only eight planets and one star.
Our examples have shown that both theory and observation are important. But given the need for technological advances — rovers, spacecraft, telescopes and optical devices — it would be wrong to limit our scientific objectives to well-posed questions with overly concise answers. Wide bandwidth is appropriate, both for planetary science and exoplanet science. If a new technology appears, use it. Let it lead to new discoveries. Planets are full of surprises at all scales of time and space. If we approach them at closer range, or at higher spectral and spatial resolution, or with longer time coverage, we will be surprised, and we will be learning something new.
How to cite this article: Ingersoll, A. P. Three eras of planetary exploration. Nat. Astron. 1, 0010 (2017).
This research was supported by NASA through the Juno and Cassini Projects.