Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Status of Planetary Science in 1995 OVERVIEW This chapter begins with two sections summarizing the ac- complishments of solar system exploration over the three decades from 1965 to 1995, and the expected scientific questions as of the end of that period. The remaining sections constitute a much more detailed status report for individual classes of objects, with further discussions of open questions. All this material supports, and leads to, the program of missions for the period 1995 to 2015 presented in Chapter 4. State of Planetary Exploration as of 1995 Among the high points already attained or anticipated for the first three decades of planetary study ending in 1995 are: . Mercury: Characterization of physiographic provinces for half the surface; discovery of a planetary magnetic field. 19
20 . Venus: Establishment of atmospheric and cloud composi- tion; characterization of the high-temperature surface environ- ment; preliminary elemental analysis of surface material from landers; study of solar wind interaction; determination of global to- pography and gravitational field; characterization of physiographic provinces from radar images. . Moon: Determination of detailed geological history, chro- nology, and geochemistry of major geological provinces; detailed study of selected samples of surface material; investigation of cra- tering, regolith formation, and interaction of the surface with the solar wind for an airless body; discovery of remanent magnetic fields; seismic characterization; measurement of heat flow; de- termination of composition of the solar wind, both present and ancient. (By 1995, global surface mapping should be achieved or under way.3 . Mars: Near-global mapping of topography, gravity field, and thermal properties; establishment of geological diversity (vol- canoes, canyon lands, polar terrains, etch; discovery of evidence for former extensive surface water (e.g., valley and channel net- works); preliminary surface chemical analysis from landers; estate lishment of structure and chemical and isotopic composition of the atmosphere; determination of geological processes and a relative chronology; study of local and global meteorology over three mar- tian years from landers and orbiters; search for microbial life and organic compounds (yielding negative results). (By 1995, global characterization- morphology, elemental distributions, and some mineralogyof surface units is expected.) . Jupiter system: Study of atmospheric composition and circulation; detailed composition and structure of atmosphere and clouds from direct entry probe; discovery of atmospheric lightning and auroras; detailed characterization of the magnetic field and the magnetosphere (sources and sinks, plasma processes); study of the lo plasma torus and of the interactions between this satellite and the magnetosphere; discovery and characterization of the To volcanoes and interior heat flow; discovery of the ring and several small satellites; comparative studies of icy and rocky planetary oh jects. (The Galileo orbiter will carry out detailed global mapping of the large Galilean satellites and continue efforts in many of the other areas mentioned above, especially the torus and magneto- sphere. The probe will carry out a detailed sounding of Jupiter's atmosphere and clouds.)
21 ~ Saturn system: Initial global study of Saturn and its mag- netosphere; establishment of atmospheric composition differences between Jupiter and Saturn; detailed study of the ring system and investigation of new dynamical phenomena; discovery of several new satellites, including previously unknown orbital configura- tions; measurements of the composition and structure of the at- mosphere and clouds of Titan; low-resolution mapping of satellite surfaces, except Titan. ~ Uranus: Results of Voyager flyby (1986~. (Initial discov- eries include a strong magnetic field with a large inclination and remarkably diverse geology on several of the satellites.) Neptune: Results of Voyager flyby (1989~. Comets: Results of Halley flybys (1986), including imaging of the nucleus, and exploration of the proximate environment. Also, deployment of planned comet rendezvous missions. ~ Asteroids: Results of Galileo flyby of a selected asteroid and of planned flybys by the Comet Rendezvous mission. . Meteorites: Evidence for early magnetic field, late addi- tions of material with differing nucleosynthetic histories, wide- spread high-temperature events In the solar nebula; many exam- ples of core formation In small bodies, basaltic volcanism, extrater- restrial synthesis of arn~no acids; discovery of meteorites from the Moon and possibly Mars. Other Planetary Systems: Discovery that many stars are surrounded by dust clouds or disks, and imagery of one such disk; discovery of a star with a planet-like companion. (Many follow-up studies are expected by 1995.) Applications of these results to the study of planetary origin and evolution include: ~ Establishment of the age of the solar system as 4.6 bil- lion years by analysis of radioactive decay products in the Earth, meteorites, and lunar samples. ~ Dating of the late stages of accretion of the Moon (and presumably the other terrestrial planets) as 3.7 billion years ago, although most of the mass was probably accumulated within the first 107 or 108 years. ~ Determination of a geological chronology for the Moon, with the final major stages of lunar volcanism measured at 3 billion years ago; establishment of the current rate for impact cratering in the Earth-Moon system.
22 ~ Comparative studies of geological processes on the terres- trial planets and the icy satellites of the outer solar system, includ- ing impact cratering, volcanic and tectonic activity, and erosional and depositional processes. ~ Preliminary study of the development and evolution of planetary crusts in planets of different compositions and internal structures, with insight into the role of tectonics and magmat~m in the formation of the crust and interior of the Earth and other planets. ~ Inference that the great bulk of the atmospheres of Earth, Mars, and Venus are all secondary, that is, degassed from the interior or acquired late in accretion, and not remnants of the gas from the solar nebula. . Discovery of unique and as yet unexplained abundances of noble gases (total amounts, relative amounts, and isotopic ratios) on Earth, Mars, and Venus. Discovery of a large (100 times) enrichment of deuterium on Venus compared with Earth. Venus must have started out with much more water (or vapor) than it has now, and a Runaway greenhouses may have caused most of it to be lost. Discovery that all terrestrial bodies have experienced dif- ferentiation, with accompanying volcanism and tectonics, but wit differences in history from one planet to another. ~ Discovery of the uniquely high levels of volcanic activity on To, and preliminary characterization of volcanism based on different physical-chemical systems than had been encountered in the terrestrial planets. In the Saturn system, resurfacing on Enceladus represents yet another example of such volcanic activity. . Discovery of unexpected complexity in the rings of Saturn and Uranus (e.g., the presence of shepherd satellites, of spiral density waves, and of bending waves), providing important insights into the dynamics of self-gravitating spinning disks. ~ In situ investigation of plasma processes of wide astrophy~ ical application in the huge magnetospheres of Jupiter and Saturn. . The determination of the composition of Jupiter's atmo- sphere, which is expected to be representative of the composition of the solar nebula, especially for hydrogen and the noble gases. The abundances that will be determined by the instruments on the Galileo probe will probably become the standard for solar composition.
23 In supporting future investigations, an essential contribution will be made by theorists who endeavor to mode! the natural evm lution of gas-dust disks into stars and their associated planetary bodies. Theoretical investigations of the early stages of this evolu- tion begin with numerical and analytic modeling of star formation, in particular, the conditions under which single stars like the Sun can form. Study of the later stages of this evolution emphasizes modeling the manner and time scale for the accumulation of dust into planetesimals, and the subsequent accumulation of these plan- etesimals into planetary cores of silicates, metal, and ices. In the case of at least Jupiter and Saturn, the final stage of formation involved the gravitational capture of massive envelopes from the gas of the disk. Between now and 1995 we can expect that continuing progress will be made in this field, most likely without the help of crucial observations or sudden theoretical breakthroughs. However, in the absence of a new generation of observational facilities that permit higher resolution imaging of other protostelIar systems, it is quite possible that in the next decade we will not address the first-order questions required to make substantive progress. On the other hand, we can look forward to a significant refinement ant] enhancement of theoretical understanding concerning many aspects of nebular evolution. Much of this progress in theoretical understanding is contingent upon the availability of computational resources of continually greater power. If the first asteroidal flybys occur during the next decade, we can expect to begin to be able to place the great wealth of me- teoritical data into a planetological context. We can also expect that basic information regarding early solar system history will continue to flow from laboratory study of meteorites and strato- spheric collection of interstellar particles. In this connection, it should be pointed out that, to a large extent, the current laW oratory instrumentation used in this work was obtained during lunar sample analysis during the 1960s and early 1970s, and that attention must be given to modernizing the laboratories in which this work is done. Scientific Questions as of 1995 Fundamental questions in planetary science will remain much the same in 1995 as they are today, but new knowledge and new
24 capabilities will alter our view of how to approach them. First, the reconnaissance and exploration of the solar system will by no means be completed. Saturn and Titan are already ripe for in situ investigation and study of interactions among the magnetosphere, rings, and other satellites. Investigation of comets and asteroids will have begun, but intensive study and exploration of the wide diversity of asteroids will remain. In this area we will want to know the following: the overall structure of the asteroid belt and its radial variations of composition and physical characteristics, which are expected to reveal clues about the structure of the pro- toplanetary nebula; the mechanisms that powered the evolution of differentiated asteroids; and the chemical composition and phys- ical character of comet nuclei, in order to determine under what conditions these most primitive planetesimals formed. Internal structure of terrestrial bodies is a broad field for which, apart from the Earth, we still will have only the limited data for the Moon from Apollo, and the even more limited data for Mars from Viking. Even such basic information as crustal thickness will still be lacking. The absolute history of planetary bodies will not be understood without an unambiguous chronology based on radioactive clocks. For example, it is suspected that the martian channels and volcanoes were formed over a protracted period, even though the time scale is based only on crater counts and is very uncertain. There is little prospect of obtaining dates by other means than laboratory analysis of returned samples. Such samples remain valuable long after their acquisition and return to Earth: improved techniques can (and do for the Moon) continue to be applied to the original samples. Only one side of Mercury will have been imaged from space- craft, but all the other terrestrial planets are known to be asym- metric In the distribution of geological provinces. While the Galilean satellites of Jupiter will have been studied in some detail, only the most rudimentary reconnaissance will have been made of the other outer planet satellites. Only single flybys of Saturn, Uranus, and Neptune will have taken place, and the Pluto system will remain unvisited. Our ideas about the origin of this solar system lead us to believe that planet-forming processes occur commonly during star formation. We will want to determine the prevalence and the prop- erties of planetary systems around other stars accurately enough to compare them with one another, as well as with our own system.
25 We will want to carry on detailed studies of protostars in order to ascertain the physical character of their accretion disks, thought to be the sites of planet formation. It seems likely that Earth is the only site of organic life in the solar system, but there is no dearth of organic molecules on or in such objects as meteorites, Titan, the jovian planets, and giant molecular clouds located in other parts of the galaxy. Mars, formerly the object of greatest interest, is now seen to be the site of destruction of organic compounds by an intensely oxidizing atmosphere and soil. Conditions, however, may have been more benign in the remote past. There is still much to be learned about the origin of life by study of the objects mentioned above, and perhaps others such as comets. If other planetary systems exist, they may be seats of organic evolution. PLANETARY GEOSCIENCES During a relatively short period of time, studies of planets made by earth-based telescopes have advanced to detailed in situ measurements from spacecraft of the planets' surfaces and at- mospheres. A complex view of the planets and their satellites continues to emerge. In late 1962, Mariner 2 the first interplanetary spacecraft- flew by Venus: the journey of Voyager 2 is still in progress. The 203-kg Mariner 2 had only six instruments, whereas the 81~kg Voyager 2 has two color TV cameras and ten other advanced instruments. These two spacecraft represent the simple beginning and the sophisticated continuation of solar system exploration. In the early years of exploration, missions were selected more by technical feasibility than by scientific priority. So little was known that any mission greatly increased our knowledge. Now, comparative study of the planets is a significant scientific en- deavor. Great advances in understanding the origin and evolution of the planets and properties of the solar system will come from comparisons of all planetary objects. Common features such as atmospheres, magnetic fields, and geologic processes can be un- derstood best by such comparison. In turn, these comparative planetary studies provide insight about the history and evolution of the Earth. Nevertheless, exploration has shown that each planet is unique and interesting in its own right.
26 Scientific Objectives for Planetary Geosciences The following topics in planetary geosciences contribute to an understanding of the solar system: formation; interior structure, dynamics, and physical state; crustal evolution; and planet mor- phology and surface processes. These topics, and the measurement objectives for them, are discussed below. Formation One key to understanding the formation of the planets is the determination of their chemical and isotopic compositions and the timing of their accretion. The results can be compared for all the planets, satellit es, and meteorites in order to place constraints on models of chemical difl.erentiation as a function of heliocentric or planetocentric distance. The results also shed light on the potential for heat sources important for considerations of internal activity and to assess models of planetary accretion. Interior Structure, Dynamics, and Physical State Measurements of the seismic behavior of planets, the strength and nature of their magnetic and gravity fields, and the heat flow from their interior are critical for determining the characteristics of planetary interiors. When combined with knowledge of mass and composition, the results permit assessment of the nature of possible interior differentiation (core/mantIe/crust) and the pos- sibilities for an internal dynamo. Crustal Evolution A principal objective in planetary exploration is the determi- nation of the age, composition, and distribution of crustal mate- rials, including volatiles. The results allow refinement of models relating to planetary accretion, differentiation, and degassing. In addition, such determinations allow assessment of the style and timing of volcanism and tectonism and their relation to other geo- logical events, as well as the role of volcanism in the evolution of possible atmospheres.
27 Planet Morphology and Surface Processes The types and distributions of landforms and other geological units on planetary surfaces can be determined through geological mapping using remote-sensing data. The results allow assessment of the processes, such as volcanism and tectonism, that have led to the formation and modification of planetary surfaces. Some landforrns, such as dunes and valleys, are indicative of processes associated with wind and water, and thus contribute to models of atmospheric evolution. Assessments must therefore be made of the distribution and exchange of volatiles among the crust, regolith, poles, and atmosphere. Knowledge of the geological processes- volcanism, tectonism, impact cratering, and surficial modifications can be combined with relative and radiometric age determinations of the features associated with those processes to derive geological histories of the planetary surfaces. An important aspect of comparative planetology relates to the origin and evolution of life. Knowledge of the geological environ- ments permits assessment of the likelihood for the evolution and sustenance of organic life, at least in comparison to Earth. The images of Earth taken from space with its thin skin of oceans and clouds help us begin to appreciate the uniqueness of our planet and the fragile balance that makes life here possible. An additional impetus for planetary exploration is the poten- tial for using space resources. In a period when natural resources are being depleted rapidly on Earth, no detailed assessment has been made of the resources that exist in space. The Moon and asteroids may hold significant potential as sources of metals and minerals for utilization in space. The initial utilization of such re- sources may be to support space missions that would travel farther into space, or permanent bases on the Moon or Mars. Measurement Objectives The goals outlined above guide the definition of a set of general scientific objectives as follows: ~ Characterize the internal structure, dynamics, physical state, and bulk composition of the planet of interest; . Characterize the planet's chemical composition and min- eralogy of surface materials on a regional and global scale;
28 ~ Determine the planet's chemical composition, ~runeralogy, and absolute ages of rocks and soil for the principal geologic provinces; ~ Characterize the processes that have produced the land- forms of the planet; ~ Determine the chemical and isotopic composition, distri- bution, and transport of probative compounds that relate to the formation and chemical evolution of the planet's atmosphere, and their incorporation in surface and crustal rocks and polar ice; ~ Characterize the planetary magnetic field and its interac- tion with the upper atmosphere, solar radiation, and the solar wind; ~ Determine the extent of organic chemical and possible bio- logical evolution on Mars and Titan, and explain how the history of the planet constrains these evolutionary processes. The Inner Solar System The inner planets- Mercury, Venus, Earth and its Moon, and Mars range from 0.4 to 1.5 AU in distance from the Sun and are smaller and denser than the outer planets. These terrestrial plan- ets are composed chiefly of rock and metal, are poor in volatiles, and have few satellites. Their densities range from S.4 g/cm3 for Mercury to 3.9 g/cm3 for Mars. The variation in density with so- lar distance has been discussed in the context of a thermodynamic mode} for the proto-solar nebula in which temperature and pres- sure decrease with distance from the nebular center and control the chern~stry of condensed material. However, it may be that primary differences in planetary density are due to accidental variations in fractionation and reaggregation from collisions. After their for- mation, all inner planet surfaces were significantly modified by a wide variety of internal and external processes. Nevertheless, each planet has followed its own evolutionary path. By exploring this diverse family of planets and by comparing their features with those of the Earth, we seek to characterize the evolution of the inner solar system and the causes of the unique aspects of each planet. We also seek to gain insights into the history, as well as the future, of Earth and the life that has evolved on it. Further insights into the terrestrial planets will come from study of the large satellites of the outer solar system as well.
29 The Moon Apollo yielded an enormous advance in the understanding of planets by providing samples of the Moon and a wealth of other information. The tune scale of the Moon's evolution has been established, and several first-order questions have been answered. Equally important, a basis has been established for interpreting the evolution of other planetary bodies, including the Earth. Dur- ing the accretionary phase of continuous planetesimal in-fall, the Moon appears to have melted to depths of at least a few hundred kilometers. The ancient crust developed during tints maelstrom, with segments repeatedly fragmented and reincorporated into the evolving magmas until a thickness was established that could with- stand the waning bombardment. The larger craters on the Moon record a period of intense bom- bardment that ended about 3.7 billion years ago, a phenomenon that presumably affected all of the inner planets at about the same fume. This bombardment provides a chronological reference, ac- curately measured in the case of the Moon by radioisotope dating techniques, that ~ the basis for constructing the geologic history of Mars, Mercury, and (presumably) Venus. The evolution of the crust of the Moon is known from remote sensing, from instrument data provided by landers, and from study of returned samples. Remote-sensing data show that two major provinces constitute the lunar crust: young (sparsely cratered), low-albedo mare terrains, and old (heavily cratered), high-albedo highlands. The oldest reliably dated rocks on the Moon (from the highlands) are radiometrically dated at about 4.5 billion years old. The youngest mare basaltic lava flows are estimated to be about 2.3 billion years old. The lunar highlands appear to be the result of differentiation at 4.5 billion years and consist of minerals that floated in the melt. However, there is controversy as to whether the upper crust of the Moon was generated in a "magma ocean," whether the whole planet was molten, or whether local areas were successively molten over a long period of time. Geophysical data show that the Moon had a strong magnetic field early in its history, but the field has since disappeared. Al- though most scientists consider the Moon to have a small, partly molten core, its presence is the subject of intense debate due to differences in interpretation of the seismic record. Because all the data provided by Surveyor, Apollo, Soviet
30 landers, and sample returns are gathered from the Earth-facing hemisphere, there is a need to obtain data from other areas of the Moon in order to establish a better understanding of the geochem- ical, geophysical, and geologic history. Since the Soviet Union's Luna 24 sample return in 1976, there have been no missions to the Moon. However, a lunar geoscience orbiter (LOO) may fly before 1995. During its one-year mission, EGO would map the global elemental and mineralogical surface composition, measure surface topography, and map the global gravity field. Although no specific date has been set for this mission, the United States, the Soviet Union, Japan, and the European Space Agency are considering EGO-type missions for the 1991 to l99S period. Mercury The geology of Mercury is known primarily from data returned by the flybys of Mariner 10, in 1974. From photogeological studies and remote-sensing data we have determined that the evolution of the crust of Mercury resembles in many ways the crustal evolution of the Moon. Less than half of the surface was imaged by Mariner 10, however, and there may well be other terrains and processes not yet known. Moreover, most of the images are of moderate resolution, on the order of hundreds of meters, and details of the surface are very poorly known. Most of the known surface of Mercury is heavily cratered and may consist of rocks similar to those of the lunar highlands- differentiated rocks high in silica and alumina. The dominant feature on Mercury is Caloris, an enormous multiringed Trip act basin. The Caloris basin is partly filled with smooth plains ma- terials interpreted as flood lavas similar to the mare basalts on the Moon. However, the characteristic vents and lava flow fronts seen so clearly on the Moon and Mars are less well displayed on Mercury, possibly due to poor image resolution, and so this interpretation is more controversial. Apparently, the silicate crust formed early on Mercury as on the Moon and heavy borr bardment continued. Later flooding of mafic lava flows also occurred, but this may have declined around 3 billion years ago as it did on the Moon. Imaging at higher resolution and coverage for the other half of the planet may discover features indicative of other processes and much younger events. Mapping compositional distributions globally could reveal
31 distinctive provinces and help to explain the high density of the planet. Venus Of the inner planets, we know the least about the geological evolution of Venus. Most of the data for this "sisters planet of Earth have come from the U.S. Pioneer-Venus mission, along with Earth-based radar, and Soviet landers (Veneras ~ through 14, Vegas 1 and 2) and orbiters (Veneras 15 and 16~. The surface of Venus is hidden from view except by radar imaging systems and imaging from the surface. Earth-based radar and data from Pioneer-Venus provide an assessment of the gross topography. Venera-15 and -16 radar images have spatial resolu- tion of about 1 to 2 km and cover part of the northern hemisphere, approximately one-quarter of the planet. Two-thirds of Venus is highland terrain, which includes plateaus higher and more exten- sive than those on Earth. In contrast, the lowland plains ("ocean basins") are only one-third as extensive and one-fifth as deep as the ocean basins on Earth. There are linear mountain belts around Ishtar Terra, the northern "continent," which include the highest mountains, known as Maxwell Mantes. These mountains surmount I,akshmi Planum a plateau that is twice as large and 1000 m higher than the Tibetan Plateau, the largest on Earth. The Beta and AtIa regions, first identified on earth-based radar images, are large shield volcanoes. Beta is composed of high- potassium basalt, while the lowland plains east of it are covered by tholeitic basalts similar to the ocean floor on the Earth and the mare lavas on the Moon. Feldspathic (high silica and alumina) rocks occur in the upland rolling plains that occupy most of the Venus surface. Both Beta and Atia have large gravity anomalies of approximately 135 mgals, corresponding to a compensation depth of more than 100 km, and thus requiring support by an upward flow of mantle material. The indirect discovery of abundant lightning in these areas has been interpreted as indicating the presence of active volcanoes. In general, there is a strong positive correlation between gravity and topography, suggesting that the dominant source is interaction of mantle convection with a surface layer- perhaps a global crust of tens of kilometers thickness. The Tong- wavelength gravity field of Venus thus contrasts sharply with the
32 Earth, which ~ a mix of deep sources (hot spots), plate tectonic, and other effects. The linear mountain belts may be due to tectonic compres- sion; other belts of fractures may be due to crustal extension. Some circular fracture patterns, called coronae, are more than 600 Em across. They may be volcanic-tectonic features, some form of impact-generated structure, or a result of both internal and ex- ternal processes. Many small craters have been observed, which may originate in impacts or volcanism, although it is difficult to determine their origin with the resolution of present images. The high deuterium/hydrogen ratio observed by Pioneer- Venus may indicate the loss of substantial water by photodissoci- ation and hydrogen escape. The yellow-gray color of the surface rocks may indicate that the iron minerals in those rocks have been oxidized like those on the surface of Mars, and fine-grained dark deposits may be windblown fine material that partly covers the surface. There are also extensive rift zones that segment the crust. However, these tectonic features do not form an integrated planet- wide system of rifts and ridges, nor are there obvious continent marginal trenches. Apparently, Venus lacks these indicators of integrated plate tectonic motion; thus Venus must get rid of its internally generated heat largely through conduction, aided pos- sibly by shot spot" volcanism, rift tectonics, and rift volcanism. Venus seems to be closer to the "one plater planets like the Moon and Mars rather than the multiplate Earth, but in volcanic and tectonic behavior it is more complex than Mars. The Magelian (Venus Radar Mapper) mission scheduled to be- gin operation around 1990, will provide near-gIobal images of the surface at 1-km or better resolution. This will enable assessment of the geological processes that have shaped the surface of Venus, and estimates of the ages and sequences of surface units, and in- ternal processes. These images will also allow us to address ques- tions regarding the former existence of liquid water (e.g., ocean shorelines, river channels), possible correlations of topography and gravity, and styles of tectonism and volcanism. Mars The martian surface is divided into two roughly equal hemi- spheres, the southern highlands and the northern lowlands. The southern highlands have elevations from 1 to 10 km above the
33 planetary reference elevation. This region is dominated by large impact craters and several enormous impact basins, of which Helias and Argyre are the largest. This ancient, heavily cratered terrain may be similar in some respects to the lunar terrae, formed 3.9 to 4.2 billion years ago. However, the martian highlands are partly covered by younger lavas and mantling deposits possibly of sedi- mentary origin. These highlands also show extensive reworking by wind and water. Although the composition of the highland rocks is not known, they probably include felsic rocks. The northern lowlands consist of young lava flows similar to the more mafic lunar mare lavas, and have been modified by aeolian, fluvial, and periglacial processes. Two impressive volcanic provinces dominate Mars: the Thar- sis region and the Elysium region. The Tharsis region includes Olympus Mons and three other very large (550 km across) voica- noes surmounting the Tharsis Risen an elevated area standing about 10 km high plus many smaller volcanoes. Olympus Mans, 600 km in diameter and more than 26 km high, is one of the larger volcanoes in the solar system. The Elysium region also includes several large volcanoes, but appears to be older than the Tharsis region. Moreover, the morphologies of the Elysium volcanoes and the lava flows are different from Tharsis and may indicate differ- ences in the style of volcanism or differences in the composition of the magma at the time of eruption. Although the Tharsis and Elysium regions are impressive, by far the greatest extent of volcanic rocks occur as various flood lavas and lava plains in the northern lowlands and the southern cratered terrain. Still other large volcanic deposits may represent high-silica eruptions and may include ash flow tufts that have been heavily eroded by the wind. The polar areas are covered by extensive, thinly layered de- posits, probably of sedimentary origin, that are associated with the permanent and ephemeral ice deposits. The perennial ice caps may be water ice in the north and carbon dioxide in the south. In addition, a vast dune field surrounds the north polar region. Dunes also occur In isolated fields elsewhere on the planet, where they partly fill many of the craters. Many parts of Mars have been extensively modified by fault- ing. Valles Marineris is a canyon system more than 4000 km long that resulted from rifting and other tectonic processes. Extensive layered deposits are visible in the sides of the canyons and in mesas
34 within the canyon system. Some of these deposits may have arisen in vast lakes that once filled the canyons. Other theories suggest that some of the layered deposits are lava flows, volcanic explosive deposits, or wind-laid sediments. Regardless of origin, the deposits may represent a large part of martian geologic history, just as the deposits in the Grand Canyon of Arizona represent a substantial fraction of Earth's history. Some regions of Mars are dissected by large and small channels cut into young rocks, indicating that liquid water existed late in martian history. Other more degraded channels dissect ancient terrain and appear to have formed earlier in martian history. Thus, there may have been many periods in Mars' past when liquid water could exist on the surface. Despite several successful missions to Mars, including the two Viking landers, many fundamental questions regarding its present state and geological history remain unanswered. The Mars Ob- server mnssion will return data in the early 1990s, providing global maps of the surface composition, details of the topography, and data on the lower atmosphere. However, questions about the inte- rior (presence of a core, nature of the mantle, etc.) and of active surface processes will remain unanswered. Internal Characteristics of the Inner Planets All the inner planets, including the Moon, underwent signif- icant early heating, melting, and differentiation, but the evolu- tion of the Moon and Mercury terminated early as heat was lost rapidly due to their small size. Like Earth, Mars and Venus are presumably sufficiently large that they are losing internal heat slowly. Their heat sources have continued to operate over billions of years, manifested at their surface in the form of volcanic and tectonic features. Earth's surface continues to evolve dynamically. Crustal ma- terial Is continually created at mid-ocean ridges and destroyed beneath deep-sea trenches, as the plates that make up the Earth's crust move in more or less steady relative motion. The formation of mountain belts, the development of volcanic chains, and the driving force behind many large earthquakes are linked to these plate motions. Neither the Moon, Mercury, nor Mars shows global tectonics of such vigor; the surfaces of the Moon and Mercury are
35 old and preserve a record of early heavy cratering and ancient vol- canism. The surface of Mars also shows heavy cratering modified by wind and water erosion, but also demonstrates an extensive history of volcanism and tecton~sm. Data for Venus, although lim- ited, show that this planet also underwent differentiation and has experienced tectonism and voican~m. The collective study of the inner planets implies that all have been melted and internally differentiated, leading to core, mantle, and lithosphere. Earth's interior is known from seismic measure- ments to be layered, a product of global differentiation. At Earth's center ~ a metallic core, largely fluid and in convective motion, but with a small solid inner core. The core is surrounded by a mantle of ferro-magnesian silicates, mostly solid and in very slow convective motion. At the surface, the mantle is capped by a thin, rigid lithosphere of mostly igneous and metamorphic rocks, over- la~n by a veneer of volcanic rocks and sedimentary material. Each of the other terrestrial planets is thought to be similarly layered, but the evidence is lim~tecI. Fundamental unresolved issues in inner planet studies are: (1) the nature of the convective motions that drive plate tectonics on Earth and the importance of such tectonic processes early in Earth's history, (2) the character of tectonism on other planets and satellites, and, (3) the causes of the major differences in evolutionary style. Magnetic Fields of the Inner Planets Earth has a substantial dipolar magnetic field of internal ori- gin, evidently produced by the action of a hydro-magnetic dynamo sustained by motions in the fluid core of the rotating planet. Of the other inner planets, only Mercury has a magnetosphere com- parable in character with that of Earth, though much smaller in size. The Moon shows evidence for an early complex magnetic history, now recorded in the remnant magnetism of lunar rocks and the lunar crust, but the origin of the field is not known. Slowly rotating Venus apparently has no internal magnetic field. The existence of a martian magnetic field ~ debatable because of scanty measurements; however, if a field exists, it is small. The wide differences in the nature of planetary magnetic fields are not understood but may be related to rotation rates and the nature of the core.
36 Characterizing the particle and field environment, including internal magnetic fields, is unportant for understanding the solar wind's interaction with a planet. Earth's field extends through a volume of space many times larger than the planetary volume, forming an umbrella that shields Earth from the interplanetary plasma; in contrast, the solar wind blasts the Moon directly. Al- though Venus does not have an internal field, there is a complex interaction with the solar wind that may be similar to that of comets. Atmospheric-Climatic Connections of the Inner Planets The geological record and measurements of planetary atmo- spheres provide clues to the evolution of surfaces and climate history. The Viking entry and lander measurements of nitrogen isotopes in the martian atmosphere indicate that large quantities of volatiles have been lost from Mars. It is inferred from this and other measurements that Mars has outgassed and subsequently lost to space or to permafrost cold traps the global equivalent of a depth of some tens of meters of water. Because of the planet's low atmospheric pressure, liquid water cannot exist on Mars at present. However, there is strong geologic evidence for the earlier presence of liquid water on Mars and the suggestion of a hydrologic cycle and thus a long, perhaps episodic history of free water on the martian surface. NASA's Mars Observer, planned to gather data in the early 1990s, will shed light on key questions regarding the present and past interactions between the atmosphere and lithosphere. This mission is proposed to provide: (1) a global view of the distribution of elements and minerals on the martian surface, (2) the water vapor, clouds, and dust distribution in the atmosphere and the vertical temperature profile, and (3) the distribution of ice on or near the surface. In addition, the Mars Observer will obtain global radar altunetry data that are crucial to the integration of the image, gravity, thermal, and chemical information. A Mars aeronomy mission, possibly flying at the same time as the Mars Observer, could explore the planet's upper atmosphere and interaction with the solar wind and answer long-standing questions about Mars' internal and external magnetic fields. In addition, such a mission could provide data on the net mass ex- change between the atmosphere and the solar wind, and provide
37 important clues regarding the history of Mars' present and past at- mospheres by determination of the rates at which various volatile elements escape from the martian ionosphere. RoLky Satellites The Voyager results made possible the geological study of a host of new objects, ranging from relatively large, silicate bodies to small satellites composed predominantly of ice. Collectively, these objects show surfaces that have experienced impact, volcanic, and tectonic processes similar to the inner planets. In addition, they show processes not seen, or at least not fully appreciated, on the inner planets, including volcanism induced by tidal heating, sulfur- driven volcanic eruptions, deformation of surface features through slow flow of ice-rich crusts, and resurfacing through the eruptions of ice-rich materials. Voyager results also caused a reassessment of any notion that internal planetary activity is simply correlated with planetary size. Larger objects contain more radioactive constituents rela- tive to their surface area, and hence generate more heat to drive planetary tectonic activity. Thus, it was thought that small bod- ies would coo! and freeze quickly and have short, simple histories of internal activity In comparison to large bodies. Although this seemed to be the case with the inner planets, the concept was drastically modified by Voyager observations. To, a body the size of Earth's Moon, had nine volcanic eruptions in progress dur- ing the encounter, making it the most internally active body in the solar system. The heat to drive these volcanoes is likely de- rived predominantly from tidal stresses created by Jupiter and the nearby satellite Europa, rather than from radioactive decay, as on larger planetary bodies. Some small satellites of Saturn (notably Enceladus) and of Uranus (such as Miranda) show evidence of resurfacing and extensive tectonism. This is indicative of inter- nal activity. On the other hand, some other larger and smaller satellites appear to have less vigorous or no internal activity. Many of the questions raised by Voyagers 1 and 2 during their brief encounters with the Jupiter system will be addressed by the Galileo spacecraft scheduled to arrive at its destination in the m~d-1990s. During its Month mission, Galileo will obtain better estimates of the chemical composition and physical state of the satellites, along with data on magnetic field and particle
38 fluxes in the Jupiter system. The cratering record, the nature of volcanic processes on Io and possibly Europa, and the styles of resurfacing and tectonic processes on Europa and Ganymede will all be substantially better known after Galileo. Studies of the outer planet icy satellites with their solid crusts and possibly mobile mantles represent an opportunity to address the fundamental problems of the physics, chemistry, and geology of deformed crusts. They will also allow us to study the internal con- stitution of bodies that differ radically from the inner planets. By 1995, data will be available for a range of bodies, from those with thoroughly deformed crusts to those that are minimally disturbed. With additional geophysical and geochem~cal data we may be able to verify the causes of the formation of the core, mantle, and lithospheres of these bodies. We may also be able to characterize the lithosphere, asthenosphere, and mantle, and their interactions to produce tectonism and volcanism in terrestrial bodies. Subtle differences in chemistry and phase are necessary to make the ter- restrial systems function. Study of systems that involve water-ice crusts and liquid-water mantles may broaden our appreciation of these fundamental problems. Satellites of Jupiter The most important advance in understanding satellite evolu- tion arose from the observations of the Jupiter system by Voyager. This understanding will be further advanced by the Galileo m~s- sion. The Galilean satellites of Jupiter form a well-ordered family of bodies that cover a wide range of internal dynamism. The inner two, lo and Europa, are about the size of Earth's Moon. Their densities, 3.5 and 3.2 g/cm3 respectively, indicate that they are probably composed primarily of silicate materials. The outer two, Ganymede and Callisto, are about the size of Mercury and appear to be mixtures of silicates and water, indicated by their densities of 2.0 and 1.8 g/cm3, respectively. As discussed above, the discovery of active volcanoes on To by Voyager makes it the most volcanically dynamic body known in the solar system. Internal heat appears to be generated by the tidal stresses in To, and the eruptions, apparently driven by sulfur dioxide, reach heights of 250 km. In addition to the products of active volcanic explosions, the surface of To is dominated by other volcanic features, including a variety of domes, caIderas,
39 collapsed depressions, and digitate lava flows that radiate from volcanic centers. The absence of impact craters at the resolution of Voyager images indicates that the surface of Io is very young. Evidently, resurfacing by lava flows and deposits from volcanic emissions is taking place very rapidly. The composition of the surface material on Io is enigmatic. The presence of sulfur is indicated by spectral reflectance data and various shot spots" having temperatures consistent with molten sulfur. However, pure sulfur has insufficient strength to form large, steep landforms. The presence of mountains as high as 8 km and scarps up to 1.5 km indicates the probable presence of silicate rocks. The surface of Europa has very high albedo. This probably indicates an ice or ice-rock crust, which Is often densely fractured. Europa is also thought to have a water substrate and a rocky interior. Only a few impact craters have been identified, and, like To, the surface is considered to be very young. Ganymede displays two fundamental surface units: an older, heavily cratered, clerk terrain and a younger, brighter unit that has been extensively modified by fracturing and other tectonic processes. Impact craters exhibit both bright and dark ejecta; many of the craters have been deformed by viscous flow of the icy crust moving under its own weight. However, some craters retain their topographic expression, and it has been proposed that there was higher heat flow early in the history of Ganymede, which allowed viscous creep to occur at a high rate. Later cooling led to a more rigid crust. The outermost Galilean satellite is Callisto. Its surface is very heavily cratered and includes several multiringed structures. Cal- listo does not appear to have experienced any tectonic deformation or volcanism. Satellites of Saturn The satellites of Saturn are more diverse and irregular in their crustal evolution than those of Jupiter. The most interesting is Titan. It has Tow density (1.9 g/cm3) and a thick atmosphere (1.5 bars) of predorn~nantly nitrogen, plus methane and minor constituents. Clouds masked views of the surface from Voyager; consequently, little is known about the crustal evolution of Titan. Enceladus is a particularly intriguing satellite of Saturn. It
40 displays ancient, heavily cratered terrain, a crust that is broken by faults, and areas that have been resurfaced. Its density (1.2 g/cm3) is consistent with water ice, but substantial amounts of methane and other ices may also be present. Enceladus, like To, may have been volcanically active as a result of tidal stresses, giving it a crust and mantle that may still be active. The other Saturnian satellites are heavily cratered and are so cold and rigid that the craters retain their original topo- graphic form and are not viscously deformed, as are the craters on Ganymede and Callisto in the Jupiter system. They exhibit different degrees of internal activity, varying from rift valleys, as on Tethys, to faulted and partially resurfaced crusts, as on Dione. Bright wispy zones, some of which are associated with faults, may reflect water frosts erupted from fissures. lapetus has a larger variation in albedo (dark to bright) than any other satellite in the solar system. It may be affected by dark material swept up from orbit, emplaced by impact, or deposited via internal activity. The current data are of highly inadequate resolution to resolve this question. Satellites of Uranus and Neptune The Voyager 2 encounters of the Uranus system in early 1986 returnee] the first images and other data on the nature of the satellites. Ground-based studies have already demonstrated that the five known satellites of Uranus are icy objects, perhaps similar in bulk composition to the satellites of Saturn, but with darker (dirtier) surfaces. All five satellites were imaged by Voyager 2 at resolutions of a few kilometers or better, but the mission emphasis was on the innermost known satellite, Miranda, which was encoun- tered at close range, yielding subkilometer resolutions, as good as any obtained by Voyager of the satellites of Jupiter and Saturn. The 1989 Voyager encounter with Neptune ~ being planned with that planet's largest satellite, Triton, as a prime target. The spacecraft will fly close to Triton and will provide an occultation as seen from Earth. The occultation is particularly important since Triton is known to have an atmosphere. Ground-based telescopic studies have also revealed evidence of liquid nitrogen and perhaps hydrocarbons on the surface, as well as frozen methane, making this nearly lunar-sized object potentially one of the most interesting members of the satellite family.
41 ATMOSPHERES Atmospheres of planets and moons exist in an enormous va- riety. The terrestrial objects Venus, Mars, and Titan have at- mospheres somewhat similar to Earth's, even though they span several orders of magnitude in surface pressure. The Jovian plan- ets, including Jupiter, Saturn, Uranus, and Neptune, have ex- tremely deep atmospheres of which we can expect to explore only the outermost skin. They are dominated by hydrogen and helium rather than the oxygen, nitrogen, and carbon dioxide of the ter- restrial planets. Very tenuous atmospheres are found on Mercury, Io, comets, and probably a few of the icy satellites of the jovian planets; in many ways they resemble the outermost parts of the denser atmospheres, exhibiting phenomena such as escape and the presence of ionization. Atmospheres are studied to determine their present state- their composition, structure, meteorologyand also to find clues to their origin and evolution. To first order, the gases in the jovian planets resemble those in the Sun, and are therefore taken to be of primary origin, with little change since their formation. Such gases are almost, but not quite, absent on the terrestrial planets; instead we find "volatiles~ that could plausibly have accreted as components of solids. Later degassing and chemical alteration would then produce what is found today. For example, photolysis of ammonia and loss of hydrogen to space give nitrogen. Likewise, life on the Earth has converted carbon dioxide to oxygen, organic molecules, and buried carbon in the form of carbonate rocks. The traces of ~primary" gases, neon, and heavier noble gases can only be measured from within the atmosphere or on a sample. In this area our present information is limited to Earth, Mars, Venus, and the parent bodies of certain meteorites. The most remarkable thing about these results is their diversity, which has so far resisted any attempts at an overall explanation. Nevertheless, it is important to know that primary atmospheres either never existed on bodies as large as the Earth or were almost entirely lost. Earth, Mars, and Venus The atmospheres of Earth, Mars, and Venus have all evolved markedly from their initial states. The other planets show that
42 biological activity, which seems to have dominated on Earth, is not the only agent that can have a profound effect on a planetary atmosphere or surface. Viking measurements of the Mars atmosphere showed that a large fraction of the original nitrogen has been lost. The abundance of water must have been enough at one time to cut large numbers of fluvial channels; current surface temperatures and pressures do not allow water in the liquid state. Mars may have lost an amount of water equivalent to a layer 10 m deep to space and to sinks below the surface, although some resides in the polar caps. Even larger amounts are missing from Venus, as shown by the huge enrichment of deuterium in the atmosphere measured by Pioneer Venus. Depending on the exact style of loss, as much as an earth ocean of water could have been lost over the life of the planet. This could explain the immense underabundance of water on Venus relative to the Earth, since the planets are otherwise quite similar. However, the original amount is very poorly determined, and the site of the oxygen left behind by the escaping hydrogen has not been established. It is usual to assume that the loss took place early in Venus' history. However, this timing should be determined and any possible relation to a change in the tectonic style of the planet should be explored. The history of Earth is similar to the histories of other terres- trial planets. Common aspects include early global differentiation of crust and core, outgassing and evolution of an atmosphere, and early bombardment of the surface by a heavy flux of meteoroids. Of course, Earth has many attributes not shared by any other planet. These include its oceans, its high oxygen abundance, its tectonic motions and the consequent complex history of crustal de- formation, its life form, and its development of a global magnetic field and magnetosphere. Earth is the only planet that has large quantities of free water on its surface and in its atmosphere. The dynamics of Earth's oceans play a large, incompletely understood role in the regula- tion of the terrestrial climate. Earth is also unique in the large quantities of molecular oxygen in its atmosphere as the result of biological activity. Venus makes a startling contrast: it is cov- ered by a dense global blanket of clouds composed of sulfuric acid droplets and has a thick, hot atmosphere of carbon dioxide. Cloud motions and tracking of descent probes and balloons indicate a global wind pattern with substantial dependence on height for the
43 mean wind speed. Surface winds are mild, but 100 m/s winds blow at the cloud tops. Martian winds are variable, as on Earth, with annual episodes of high-velocity winds that often give rise to global dust storm. Mars also has marked seasons, with carbon dioxide cycling between the polar caps driving a major component of atmospheric circulation. Layered sedimentary deposits at the martian poles are evidence of tong-term climatic changes, whose origins are poorly understood. On Earth, such climatic changes have given rise to the periodic ice ages. Earth is the only planet whose surface, atmosphere, and hydrosphere have provided an en- vironment conducive to the development of life and the evolution of complex living organisms. These life forms have substantially influenced the chemistry of the atmosphere and the oceans and the major sedimentary rocks on Earth's surface. The Viking mission showed the absence of detectable organic molecules on Mars. It also revealed that an intense ultraviolet flux from the Sun reaches the surface. This suggests that living organisms are not present on Mars now. Whether Mars was less hostile to the development of life during earlier times, when it may have had a denser atmosphere and flowing surface water, Is still an open question. Other basic questions about Mars are the fate of all the missing water and the nature of the current hydrological cycle linking the polar caps, ground water and ice, and the atmosphere. A rare group of meteorites, called SNC for the initials of places where the first examples were found, are widely believed to be samples from a martian lava flow. Most remarkably, they contain gases whose elemental and isotopic composition is substantially identical to that measured by the Viking landers, within the errors of the latter. Study of the present state of an atmosphere is carried out both in situ and remotely by flybys and orbiters. Except for Mars, all the terrestrial bodies with substantial atmospheres exhibit a greenhouse effect: the surface temperatures are higher than they would be in the absence of an atmosphere. There is a great deal of interest in this phenomenon from the standpoint of past and future climates on the Earth, where we are in the midst of a grand experiment of the effect of massive injections of carbon dioxide from human activity. Clouds, smogs, and dust are interesting both in their own right and as tracers of atmospheric motions. Observed clouds include water and ice on Earth, ice and solid carbon dioxide on Mars, ammonia, methane, and perhaps liquid water or ice on
44 the jovian planets. Smogs (produced photochemically rather than by condensation) are found on Earth, Venus, Titan, and at least some of the jovian planets. Dust is very important on Mars and is significant on the Earth as weli. Theoretical meteorology of the Earth is finely tuned to our oh servational knowledge. It has been extended to Mars with consid- erable success, but is very unsatisfactory with Venus and Jupiter. On Venus we observe winds near the cloud tops (about 65-km altitude) blowing from the east at 100 m/s, and similar winds are thought to exist in the region of the ionosphere. Not only were these motions unpredicted, they still cannot be explained in any fundamental way now that they have been observed. On Jupiter and Saturn, similar velocities are seen, but with large shears be- tween zones. These are just as far from being explained. Progress in meteorological theory will rely on a combination of further theoretical and computational work, and more and dif- ferent observations. A theory with true predictive power would be especially important in the area of climatic change. Improved short-term forecasting, however, is more likely to be obtained by still-more-detailed observations of the Earth itself. The study of photochemistry and photoionization is much more readily transferable between planets. Mars and Venus are excellent Earth analogs. The dayside ionosphere of Venus is well understood, and the stratospheres exhibit similar phenomena. Limited ciata on the Mars ionosphere suggest that it is similar to that of Venus, but measurements from a long-lived orbiter are needed. Catalytic ozone destruction is much more important in the stratospheres of both Mars and Venus than on Earth. Both Venus and Earth have sulfate layers, but again this layer is much denser on Venus. Further study of these planets should continue to shed light on terrestrial pollution problems. The nightside upper atmosphere and ionosphere of Venus are unexpectedly cold. The corresponding region of Mars is unex- plored and may or may not be similar. Comparative study should cast light on this mysterious phenomenon. Titan Titan, the largest satellite of Saturn, is unique among satellites in having a dense atmosphere: predominantly nitrogen, but also containing a small amount of methane and possibly argon. The
45 surface pressure is 1.5 bars. Since the surface gravity is only 135 cm/s2, the amount of gas per unit area is nearly 11 times that of the Earth. A nearly uniform orange haze hides the surface and any con- densation clouds that might exist. This orange haze ~ likely composed of condensed hydrocarbons, nitrite compounds, poly- acetylenes, and HCN. The precipitation of these compounds may form a deep layer lying on the surface, or they may be dissolved in a liquid ocean. Dense methane clouds probably lie some distance above the surface. This rich inventory of organics in a nitrogen medium provides a natural laboratory for the study of prebiotic organic chemistry, which may be relevant to the question of the origins of life on Earth. The origin of nitrogen on Titan ~ a fundamental, unsolved question. One theory is that the nitrogen could be primordial, in- corporated in a ciathrate ice during accretion. Two other sources have also been suggested: photolysis of ammonia and high-temper- ature formation from ammonia during impacts. The other atmo- spheric constituents can be derived by photochemistry from a nitrogen-methane rn~xture, except for the traces of carbon monox- ide and carbon dioxide. The carbon monoxide could be outgassing from the interior, or could be derived from the ice in incoming meteoric material, with the carbon coming from the atmospheric methane. Titan is embedded in a torus of escaped gases, which includes atomic hydrogen, observed by Voyagers 1 and 2, and probably molecular hydrogen and nitrogen. ionized torus material con- tributes to the plasma in Saturn's magnetosphere: impact by magnetospheric particles is an important loss process for the neu- tral torus. Voyager 1 passed through Titan's magnetospheric wake and observed a number of changes in the plasma and magnetic en- vironment. The existence and nature of the predicted ocean of ethane and nitrogen (or alternatives, such as lakes or puddles) should be studied, both remotely and in situ. There are important interac- tions of such an ocean with the atmosphere, for which it would be a source of volatiles and a sink of photochemical products. Of the volatiles, methane is a likely candidate to produce the dense cloud layers, whose presence is strongly suspected, but which have not been confirmed because they are hidden by the photochemical smog.
46 An atmospheric circulation, especially strong in the strato- sphere, has been inferred from the temperature field observed by Voyager 1. Testing by direct wind measurements could tell us whether our understanding, based on Earth and Venus data, is good enough to encompass Titan as well. Finally, the hydrogen torus surrounding Titan's orbit has provoked speculation about the presence of other atoms and molecules in the torus, and the torus' interactions with the magnetospheric plasma. Also of in- terest is the possibility of sources for the torus other than Titan itself; Saturn is a possible candidate. lo and the Plasma Torus To, Jupiter's innermost large satellite, exhibits such remark- able phenomena as volcanoes believed to be driven by sulfur diox- ide, a tenuous atmosphere of sulfur dioxide, a persistent extended cloud of sodium atoms, and a plasma torus containing ions of sul- fur and oxygen enveloping the orbit. In turn, many of these ions become energized and populate the magnetosphere, and probably drive the large escape rates that populate the torus itself. The energy source for the vuicanism is tidal heating by Jupiter and nearby satellites. Most of the other aspects are poorly understood at best, and some are totally mysterious. For example, although it seems clear that sulfur dioxide is passing through the atmo- sphere to the torus, there is no agreement on the density of the atmosphere. Study of all these phenomena would be near the top of any priority list if it were not for the unbearable environment of the en- veloping magnetosphere. Pioneer and Voyager instruments mak- ing a single pass have been damaged, and radiation damage is a primary concern for the Galileo spacecraft and its instruments. A prolonged stay near To does not seem possible with current or easily foreseen technology. The most burning question about To's atmosphere is the ac- tual quantity of sulfur dioxide and its variations with latitude and time of day. The amounts of other gases, such as oxygen, are also of great interest. A large pressure bulge is expected near the sum solar point, and some models predict strong winds blowing away from this region. Probably the most remarkable thing about this atmosphere is the huge escape fluxes of sulfur, oxygen, and sulfur dioxide that populate the torus, and that require replacement in
47 a very short time. The mechanism itself is obscure, and there is the possibility of large variations with time. Among the atoms involved is sodium, which gives a bright glow easily observed from Earth. It must be a substantial component of the crust, but the chemical form is unknown. Although the ionic composition of the torus is fairly well known, the energy sources are not fully established. Since this object drives much of the rest of the magnetosphere, it needs to be understood as well as possible. Another energetic phenomenon related to lo is the intense radio (decametric) bursts, which have been studied from Earth for decades but whose origin is still being discussed. Jovian Planets From their very low mean densities we know that Jupiter, Saturn, Uranus, and Neptune have extremely deep atmospheres, extending perhaps more than halfway to the centers. They also possess cores of 10 to 20 earth masses similar to large terrestrial planets, but possibly much richer in the ice-form~ng compounds water, methane, and ammonia. A large percentage of the atmo- spheres appears to be made up of these compounds; there is strong evidence for methane and ammonia, but much weaker indications of water. They also form cloud layers, perhaps more than one per planet, but the highest layer (ammonia on Jupiter and Saturn, methane on Uranus and Neptune) tends to obscure the deeper ones. Particularly for Jupiter, the cloud patterns are arranged in bands parallel to the equator, instead of the cyclonic whirls found on Earth. Several simple organic molecules and a dark strato- spheric smog, also seen on Titan, are believed to be produced photochemically from methane. Jupiter, Saturn, and Neptune all have internal heat sources that are very large by terrestrial standards, and comparable to the heat they receive from the Sun. Uranus' source may be almost as large as Neptune's, but it cannot be resolved from the reradiated solar heat. Such a flux from the interior is expected to have profound effects on the structure and circulation of an atmosphere, but the details are not understood. The heat source itself is of major interest; it is probably a combination of residual heat from planetary formation and gravitational energy from continued differentiation and rainout of heavier elements, such as helium, due
48 to immiscibility with hydrogen. The former heat source dominates on Jupiter and the latter on Saturn. Precise in situ measurements of the helium-to-hydrogen ratio in the atmospheres of Saturn, Uranus, and Neptune will contribute to an understanding of these processes. The dynamical properties of the interior are closely coupled to the composition, structure, heat flow, and rotation of the planet. Convective flow of conducting matter generates an external mag- netic field that reflects some of the properties of the internal flows, with changes in the Sow producing secular changes in the magnetic field. These properties of planetary magnetic fields can be best determined from orbiting spacecraft with a small periapsis. Differences in the bulk composition of the outer planets are related to differences in the temperature, pressure, and chemistry occurring at different radial locations in the solar nebula at the time of planetary formation and to the nature of the accretion and collapse process. Because the major species such as carbon, nitrogen, and oxygen are present mainly as methane, ammonia, and water, which are deficient in the upper atmosphere due to condensation into clouds, their abundances must be determined by probes below the cloud decks. On Uranus and Neptune, it is thought that the water cloud base will lie below 100 bars. It is often presumed that the different-colored clouds of Jupiter are at different altitudes and indicate regions of upwelling and downwelling. However, there is little direct evidence that this is the case and there are other indications from observations of atmospheric scattering properties that are difficult to reconcile with such a model. Synoptic multiband observation with moderate spatial resolution would be valuable, as would in situ observations of cloud depths in different latitudinal regions. Internal structure is related to the origin and evolution of the planets, since it depends on bulk composition, on the accre- tion process, and on subsequent evolution. It ~ also of interest for the insight it provides into the properties of matter at high temperatures and pressures. Knowledge of the bulk composition will be essential, and improved values for the higher gravitational moments, as can be derived from an orbiting spacecraft, would be useful. Laboratory and computer studies of high-pressure proper- ties are also essential. It is possible that the observed zonal wind flows are an atmo- spheric skin effect extending only as deep as sunlight penetrates
49 (several bars). Alternatively, the zonal winds may reflect an in- ternal flow pattern extending deep into the neutral atmosphere, ciriven by the internal heat source. In situ wind measurements with a simple atmospheric probe, and synoptic optical, infrared, and microwave observations from an orbiting spacecraft, may provide relevant information. RINGS ~ ~ ~ ~ , Planetary rings, once thought to be unique to Saturn, have been observed around all the giant planets except Neptune. Even for Neptune, there is evidence for the existence of partial rings. The ring of Jupiter Is optically thin and composed of dustlike small particles. Saturn's rings are broad, bright, and opaque, whereas the rings of Uranus are narrow and dark. They all lie predominantly within the Roche limit, where tide] forces would destroy a self-gravitating body, and also within the planetary magnetosphere. The goals of the study of planetary rings include three major objectives. The first is to understand their composition, active processes, and origin. A second objective Is to study their active processes as analogs of those that operate in other flattened, rotat- ing, dynamic systems like galaxies, accretion disks, and our own solar system at an earlier stage. A third objective is to study the particles as remnants of an earlier stage of solar system evolution; they are not as primitive as the comets, but less processed than the larger planets and satellites in the outer solar system. The best-studied ring system is Saturn's, which has been oh served from the ground for centuries. The most detailed informa- tion on planetary rings is from Voyager spacecraft observations. Ground-based radar, photometry, and infrared spectroscopy have been complemented by spacecraft imaging, spectroscopy, and oc- cultation. We now have a reasonably comprehensive inventory of the ring material surrounding Jupiter, Saturn, and Uranus, and a preliminary understanding of some important dynarn~c processes in each of these systems. Continuing theoretical modeling using existing data sets will focus questions about the physical processes that govern the morphology and stability of planetary rings. The occurrence of rings around the massive planets is evidence of an evolutionary path parallel to planetary aggregation. The
so . nonlinear, dynamical interactions responsible for forming and flat- tening planetary rings also operate in galaxies, planetary systems, and stellar accretion disks. Generally, rings consist of planetary material that accreted originally with the planet. This mate- rial was either never incorporated into larger bodies, or may have formed satellite bodies that were later broken up. In both Saturn's E-ring and the Jovian ring, there is Voyager evidence for short par- ticle lifetimes, arguing for continuous replenishment from satellite surfaces. A fundamental open question is the age of the Saturn ring system: the proximity of the inner satellites is inconsistent with the action throughout the age of the solar system of current dynamic processes, which push them away from the rings. Saturn's rings are mostly water ice and emit thermal radiation in energy balance with incident sunlight. Radar, radio occultation, and spectrophotometric studies indicate that particles larger than 1 cm are responsible for most of the Saturn ring opacity. The composition and thermal state of Jupiter's and Uranus' rings are unknown, but interaction with magnetospheric ions at Jupiter may both heat and differentiate ring material. Remote sensing has a limited capability to characterize ring particle composition and size: by 1995, any spectroscopically active constituents should have been identified in both the solid and gas phases. In the subsequent period, a ring rendezvous mission could provide in situ analysis of particle and ring-atmosphere composition. Ring particles display complex collective interactions. Voy- ager Saturn data show mass clustering into thousands of ring features. Density and bending waves, spokes, and even multiple strands have also been observed. Nine distinct, narrow rings have been identified at Uranus. Many of the Saturn structures are gravitationally induced by satellite orbital resonance; for exam- ple, the outer ring edges of Saturn's brighter ring occur at radial distances where particle orbital periods are commensurate with those of satellites. Small, close satellites "shepherd" the- F-ring. A satisfactory explanation does not exist for the multitude of smaller- scale Saturn ring features, and explanations of the broad structure of Uranus' rings and Jupiter's ring are unconfirmed. Since ring particles may acquire net charge by either ultraviolet or charged particle irradiation, electrodynamic forces influence the motions of the smallest particles, probably producing the Saturn ring spokes and likely limiting the lifetime of the small particles in Jupiter's
51 ring. A ring rendezvous could measure the plasma environment of Saturn's rings in the vicinity of a spoke feature. The orbit pole of an inclined elliptical ring processes in re- sponse to the higher moments of the planetary mass distribution. This effect has been observed for Uranus' epsilon ring, and has provided a value of ]2, the second gravitation moment, with a small uncertainty. By 1995, improved values for the even the mul- tipole moments of giant planet gravity should be available from improved ring and satellite astrometry. In the mid-199Os, we will have new observations of the jovian ring from the Galileo orbiter and observations of Saturn's rings (including stellar occultations), from both earth-based and earth- orbiting instruments. We will also have data from Voyager studies of the uranian rings, and will have determined at least upper limits on any Neptune ring material. In addition, we expect that a Saturn orbiter mission will be en route to Saturn to arrive at the turn of the century. For the Jovian and saturnian systems we can hope to have some long-term, detailed information on dynamics and secular changes. We also expect to know about the spatial variation of important observables like composition within those ring systems. However, the details of individual particles and their mutual interactions still may be hidden, and we will still lack any in situ measurement of their composition and local environment. The origin and nature of both the large-scale and the small-scale structure in Saturn's rings are likely to be resolved only when the composition, physical properties, and size distributions of particles in the A-, B-, C-, and D-rings are adequately characterized. Among the dynamical processes believed to be important in Saturn's rings are direct collisions, gravitational scattering, dif- fusion and angular momentum flow, resonance interactions with various satellites and Saturn, electrodynamic processes, and possi- bly diffusional instabilities. The exact nature of these processes is not understood, however. For example, although the outer edges of the A- and B-rings are known to result from satellite resonances, the edges are much more abrupt than expected. A more complete understanding of the nature of the processes and their importance depends upon knowledge of inelastic collisions between ring parti- cles and of the particles' relative velocities. There is currently no direct information on either of these. Although it is generally agreed that the outer edge of the A- ring is maintained by the outward transfer of angular momentum
52 to the small co-orbital satellites, the resulting dynamical time scale for the expansion of the orbit of these satellites is short, suggesting either that they are transferring angular momentum to another satellite through some unrecognized mechanisms or that the outward flow of angular momentum from the A-ring is smaller than thought. The dynamical time scale for the inner edges of the rings also poses a problem, since there is no known mechanism for transferring angular momentum to the ring particles to halt their inward spiral toward Saturn. The thickness of the rings is dynamically determined by the balance between energy lost and the random energy supplied dur- ing collisions. This clearly depends upon the relative importance of the various interparticle interactions, including direct collisions and gravitational scattering. These two physical processes are dis- tinct and result in different dependences of the vertical distribution of particles on their relative velocities, size distribution, and phys- ical properties. The vertical spread of the largest particles could be quite different from that of the smaller particles. The overall composition of the ring material is related to the origin of the ring, while any segregation of the material according to particle size or location involves dynamical processes that have either produced or maintained the segregation. Closely related to the composition Is the physical nature of the individual particles. They may be fluffy or solid, cohesive or merely short-lived aggre- gations of smaller pieces. Observations of individual particles from a ring rendezvous could provide this information. Models for the formation of the observed radial spokes involve unobserved aspects of the plasma and neutral particle environment of the rings. Although synoptic studies from a Saturn orbiter will considerably improve our knowledge of the kinematics of spoke creation and dissipation, measurement of the ring environment by a ring rendezvous would directly address the nature of the processes involved. For the uranian rings, answering current questions will involve providing an inventory of the ring material and associated satel- lites, determining the relative role of self-gravity and collisions in creating the morphology of the elliptical rings, and discovering waves and other small-scale structures not visible from earth-based stellar occultations. All of these were addressed by the Voyager Uranus encounter in 1986. This, however, will be the last look at the uranian rings before 1995.
53 It ~ not possible to predict what will be the most important questions arising after the new Voyager results, since it Is likely that we will find that some of our current concepts are incorrect. However, a number of major questions remain open after the Voyager Uranus flyby. These are the detailed dynamics, long- term changes, composition, and spatial variation of the rings. Further, we will need to know their neutral and charged-particle environment. One discovery is that the fringes of the atmosphere actually envelop the rings and cause a very substantial drag yet another indication of a short time scale. Clearly, a Tong-lived Uranus orbiter is needed to address these questions. Jupiter's ring wall be extensively imaged by the Galileo orbiter before 1995. The Galileo observations will help answer important questions for this system. The COD (charge-coupled device) cam- era on Galileo Is more sensitive than Voyager's and will make mul- tiple observations of the jovian ring and small satellites nearby. This will refine our knowledge of the morphology of the ring and better define the particle size distribution. We expect increased understanding of the evolution of individual particles under the combined effects of plasma and gravity. We hope to clarify the relation between the ring particles and the small moons embedded in it, which appear to be supplying ring material. Since the life history and lifetime of the small jovian ring particles are strongly affected by the plasma environment, deeper understanding requires directly measuring that environment. This requires in situ measurement of the inner jovian magnetosphere, most plausibly using a Jupiter orbiter with a low perijove. INTERIORS OF THE GIANT PLANETS Our knowledge of the giant planet interiors is indirect and relies heavily on theoretical plausibility arguments. The proce- dure involves the computation of theoretical models with several adjustable parameters that are varied to match observed quan- tities. These quantities include the mass, radius (and hence the mean density), and often several of the higher gravitational mo- ments that depend both on the rotation period and the internal distribution of the mass. Prior to spacecraft flybys of many of the giant planets, the radii of Jupiter and Saturn were known only from remote mea- surements, the masses were fairly accurately known from planetary
54 perturbations, and the first two gravitational moments were ap- proximately known from satellite perturbation; the observational situation for Uranus and Neptune was much worse. The flybys that have taken place through the Jupiter, Saturn, and Uranus systems have much improved the accuracy of all these quantities, and meanwhile there has been an extensive effort-to determine these quantities better for Neptune by remote observations. By 1990 -he should also have spacecraft flyby observations for Nep- tune. The construction of interior models depends upon having a good equation of state, which, in turn, depends upon a knowledge of the elemental constitution of the interior. It has been customary to assume that the major elements, hydrogen and helium, should be present in solar proportions, since there is no known way in which the planets can be formed that would result in the chemical fractionation of these two gases. Models composed of hydrogen and helium alone are deficient, and do not fit the observations; the discrepancies indicate that there is an excess of material of higher mean molecular weight near the center. The mass excess indicated is approximately 10 earth masses for each of the giant planets. This is a large excess relative to the planetary mass for Uranus and Neptune, and so can be obtained from even crude knowledge of the planetary parameters, but it is a much smaller fraction of the planetary mass for Jupiter and Saturn, so that the higher precision of the knowledge of the planetary parameters has been essential in those two cases. We can say very little about the distribution of this excess mass, other than that it is core-concentrated, and we cannot determine directly whether it is entirely rocky in composition or whether there may be a major contribution to it from "ices" such as water or ammonia. One major theoretical question is the solubility of helium in hydrogen in the interior. Remote sensing measurements from the Saturn flybys have indicated a reduced helium/hydrogen ratio in the Saturn atmosphere compared to Jupiter. This has been interpreted as indicating insolubility of helium in hydrogen on Saturn, where the interior temperature is lower than in Jupiter. Thus it is expected that helium has trainees out of the Saturn envelope to lower depths. A consequence of this is an additional gravitational energy release on Saturn, which may account for its observed infrared luminosity, much higher than would be expected from a straightforward cooling history since the planet was formed
55 and contracted to its present size. By the mid-199Os, we expect to have direct helium/hydrogen abundance measurements for the Jupiter atmosphere from the Galileo entry probe, and can be looking forward to sirn~lar measurements for Saturn, which will improve the accuracy of this import ant measurement. A major question of the interior physics of Jupiter and Saturn is the nature of the transition interface from molecular hydrogen to met alRic hydrogen, which is expected to occur at a pressure in excess of one megabar. All of the giant planet interiors are ex- pected to be connectively unstable, which would normally keep the envelopes well cruxes. However, this interface to metallic hydrogen may locally suppress convection, and this may help to maintain a disparity in the helium/hydrogen ratio in the upper and lower envelope of Saturn. Related to this may be the puzzle that an internal infrared luminosity has not been detected for Uranus, but one is readily found for the very similar planet Neptune. These are among the important questions that can be addressed through improved modeling, which, in turn, depends upon better measure- ments of envelope composition and gravitational moments. Another theoretical expectation is that the cores of the giant planets, whether of rock or ice composition, would dissolve into the overlying hydrogen at the central pressures of the planets, if given an opportunity to do so (unless the rocky material arrives in very large chunks). The implication is that they have not been given a chance to do so, and hence the mass excesses in the cores have not been produced by infall of material into the assembled planets and settling through their envelopes. This points toward a process of formation of the giant planets in which the cores formed first and then captured the hydrogen and helium. Thm, in turn, places a number of important constraints on the history of the solar system, and it requires confirmation, not only through the construction of better interior models, but also through better laboratory measurements of the relevant physics. In the mid-199Os, we expect to know the relative abundances of hydrogen, helium, carbon, nitrogen, and hopefully oxygen in the Jovian atmosphere as a result of the probe entry in the Galileo mission. It should be possible to make better measurements of the higher gravitational moments because of the Galileo mission. The heat flow from the interior of the planet and the configuration of the Jovian magnetic field should be measured somewhat better.
56 These will allow improved models of the planet to be constructed, but our knowledge of the planetary core will remain very crude. For Jupiter it will be desirable to have a follow-up mission that includes as a component a spacecraft in polar orbit with at least part of the orbit lying close to the planet. This will make possible higher-precision measurements of the gravitational moments and will allow a complete mapping of the magnetic field, including the determination of many of the higher magnetic moments. If the abundance of oxygen is inadequately determined in the Galileo mission, then a follow-up probe designed to obtain this highly important parameter will also be desirable. Laboratory studies of the equations of state of the relevant materials and of the solubility of heavier elements in hydrogen at higher pressures will contribute to an understanding of the structure of the interior. Both laboratory and theoretical studies of the metallic-molecular transition in hydrogen will be critical to the interpretation of the jovian interior. A repeat of the polar orbital measurements after an interval of several decades will be important for studying changes in the higher magnetic moments of Jupiter, and hence possibly deducing some aspects of the internal motions of the gas in the envelope at the depth at which the jovian magnetic dynamo operates. All of these measurements are also desirable at Saturn. A planetary probe is essential to measure the helium/hydrogen ratio accurately and to obtain the abundances of carbon, nitrogen, and oxygen relative to hydrogen. Polar orbiters are needed to measure gravitational and magnetic moments of the planet and to mea- sure the time variations of the latter. The interpretation of the interior structure will be assisted by the laboratory measurements mentioned above. For Uranus and Neptune only a preliminary flyby of the plan- ets will have been accomplished by 1995; this occurred quite far from the planetary surface in the case of Uranus, but will be much closer in the case of Neptune. That will give us an improved un- derstanding of the structure of these planets. However, hydrogen and helium make up only a relatively small part of the mass of these two planets, and so it is still highly desirable to obtain the best possible measurements of the higher gravitational moments to obtain the interior mass distribution, and to measure the envelope composition, the heat flow from the interior, and the magnetic mo- ments and their time variations with considerable accuracy. This
57 will require entry probes and orbiting spacecraft for these planets beyond 1995. PLANETARY MAGNETISM Although the existence of Earth's magnetic field has been known for hundreds of years, it ~ only during the twentieth cen- tury that magnetic fields have come to be known as a phenomenon widespread throughout the universe. In the early part of this cen- tury, sunspots were discovered to contain intense magnetic fields. During the past several decades magnetic fields have been found associated with the rest of the Sun, the interplanetary and inter- steliar media, many of the planets, galaxies, and a large number of stars. There is indication that the Moon was strongly magnetized early in its history and that the protosolar nebula from which the planets formed also possessed a strong magnetic field. Indeed, it is now clear that the absence of a magnetic field in a large cosmical object is the exception rather than the commonplace. Of the planets investigated so far, only Venus ant! Mars have no, or very weak, magnetic fields. Mercury, Earth, Jupiter, Sat- urn, and Uranus all have strong fields. Most of these fields are predominantly dipolar, approximately centered within the plan- ets, and nearly aligned with the planets' rotation axes. The most startling exception is Uranus, having a dipole highly inclined with respect to its rotation, and displaced far from the planet's cen- ter. Although such a large departure from symmetry is unusual among known planetary fields, astronomical studies indicate that many stars have magnetic fields with marked departures from symmetry. The surface intensities of known planetary magnetic fields typically fall in the range of a few tenths of a gauss to 10 G at the planetary surfaces, although Mercury's magnetic field is weak in comparison to those of most of the other planetsa few thousandths of a gauss. A distorted magnetic field stores energy and exerts forces on the medium in which it is embedded. Thus strong magnetic fields can influence the dynamical behavior and evolution of cosmical systems. The rapid release of energy stored in magnetic fields causes explosive, flaring outbursts in many systems, ranging from planetary magnetospheres to the solar corona and high-energy astrophysical objects. The persistence and behavior of cosrnical magnetic fields are
58 now understood to result from similar physical phenomena occur- ring in a variety of different objects. The general characteristics of objects possessing magnetic fields are that they are large, elec- trically conducting, and rotating fluid bodies. The ability of an object to retain a magnetic field depends on a combination of large physical scale and high electrical conductivity. Most cosrrucal oh jects are large by their very nature, and they conduct electricity in extensive metallic or gaseous ionized regions. The generation of a magnetic field results from the organizing influence of rotation on the convective motions of an electrical conductor. Magnetic fields are ubiquitous in the universe because most natural objects have these attributes, and because large quantities of free mag- netic charge, capable of rapidly short-circuiting magnetic fields, are absent. In addition to being important targets of scientific scrutiny in their own right, planetary magnetic fields provide significant clues to the interior states of planets. In view of the fact that magnetic fields seem to be very easily generated when the require- ments described above are satisfied, the absence of a magnetic field in a planet imposes a severe constraint on the state and the motions of its interior fluid. Moreover, detailed observations of a planetary magnetic field can yield important information about the characteristics of the generating fluid motions. Indeed, it was detailed studies of the temporal behavior of the geomagnetic field that provoked early ideas about the motion of fluids in Earth's core and that provided the earliest insights into the regeneration of magnetic fields in natural objects through the phenomenon of the hydromagnetic dynamo. More recently, the discovery of a magnetic field in Mercury was one of the unexpected discover- ies that resulted from detailed spacecraft investigations of that planet. The fact that Mercury possesses a magnetic field poses still unresolved questions about its interior structure and thermal evolution. Generation of Planetary Magnetic Fields Our present theoretical understanding of hydromagnetic dy- namos gives us a strong foundation for understanding the genera- tion and dynamical behavior of magnetic fields in the wide variety of cosmical objects that possess them. However, a great dead re- mains to be learned. Most of our current understanding is based
59 on linear kinematical theories, which are strictly confined to those situations in which the magnetic field does not itself strongly influ- ence the state of the system. But fundamental considerations and observations indicate that many natural magnetic fields do not operate in such a simple regime. Rather, many systems develop to a fully nonlinear state in which the magnetic forces are as large as any of the others. Advances in our understanding of magnetic field generation and behavior rely on a combination of theoretical and observational studies. Because of the extreme nonlinearity in their behavior, a variety of states may be available to hyd-romagnetic systems even with sunilar geometries and boundary conditions. In this respect, the behavior of hydromagnetic dynamos is analogous to that of most complicated cosmic systems. In a hydromagnetic dynamo, the magnetic field is produced by large-scale electrical currents flowing in a fluid conductor. The electrical currents are, in turn, generated by the fluid's motion across the magnetic field lines. A self-sustaining hydromagnetic dynamo is said to occur when the fluid motion is sufficiently vig- orous that new electrical current ~ generated at a rate that com- pensates or overcomes the dissipation of the current by electrical resistance. The ultimate energy source for dynamo magnetic fields derives from the forces that drive the fluid convection. This magnetic field generation process is essentially similar to the mechanism whereby electricity is produced in electrical generators. However, because the fluid systems that occur in natural bodies have so many degrees of freedom, the complexity of behavior of natural dynamos and the variety of states in which they can exist are vast in comparison with the simplicity of electrical generators. Strictly speaking, the fluid motions in dynamos cannot gen- erate magnetic fields from scratch. The dynamo process is one of field maintenance through regeneration and amplification. In order to get the process started, an initial magnetic field is re- quired to be present, although it may be arbitrarily weak. Such weak magnetic fields are omnipresent, generated by thermoelectric currents and by a variety of other astrophysical and cosmological phenomena. Not all fluid flows are capable of maintaining magnetic fields. However, a large variety of bows are capable, and most natural objects satisfy the criteria for producing such flows. As already noted, regenerative fluid motions occur in objects that rotate and
60 convect. Most cosmical objects of planetary size or larger satisfy both criteria. The motions need not be very vigorous if the fluid is highly electrically conducting. Earth's magnetic field is generated in the planet's liquid iron core, which has an electrical conductivity not atypical of cosmical objects. The regenerative fluid motions in Earth's core are estimated to be less than a millimeter per second, corresponding approximately to the velocity amplitude that theoretical calculations show to be necessary. Because of the essential organizing influence of rotation on dy- namo fluid motion in rotating objects, and because of the impor- tant role played by differential rotation, the close correspondence between dynamo magnetic fields and planetary rotation poles is not mysterious. Because large-scale planetary magnetic fields re- sult from correlated regenerative action at smaller scales, the fact that most fields are nearly centered in their respective planets is also not hard to understand. Moreover, deviations of plane- tary magnetic fields from the ~ideal" centered and axially aligned structure can be understood on the basis of random variations in the fluid motions, although other effects may also occur. In fact, the geomagnetic field the only planetary magnetic field for which detailed, long-term measurements are availabl~deviates in a ran- dom fashion from the centered, axial ideal, with the time scales of variability being some thousands of years. Averaged over very long times, the geomagnetic field is highly centered and axially aligned with the pole of rotation. Uranus' highly tilted and eccentric magnetic field challenges the simplest theoretical models of field generation. However, the significance of this challenge is difficult to assess without measure- ments of the field's possible variation with time. Mathematical analysm of the magnetohydrodynamic dynamo process shows that magnetic field states of several types can occur. The fields can be generated with a variety of spatial structures and time variations, depending on the geometrical structure of the fluid motion and its amplitude. Some dynamo modes are stationary with time, while others grow or decay monotonically; another class of dynamo modes consists of fields that oscillate periodically and migrate through the generation regions. The polarity of a cosmical magnetic field is not dictated by the dynamo process; if a field can be maintained by dynamo action, then the same field, but with opposite polarity, can be maintained equally well. Magnetic fields that are observed in natural objects clearly
61 correspond to these calculated behaviors. The Sun's magnetic field seem to correspond to an oscillating and migrating dynamo mode. Earth's magnetic field seems to correspond to a low-order stationary solution of the dynamo equations. For the other plan- ets, too little is known about the temporal behavior to guess whether any of their magnetic fields are oscillating. The oscilIa- tion periods "e probably too long to observe directly. Uranus' magnetic field ~ most provocative in this respect. The correct interpretation of its unusual spatial structure depends strongly on whether the field is in a stationary state, an oscillating state, or a transient configuration deviating markedly from a more regular average structure. ~ Earth's Magnetic Field For obvious reasons, Earth's magnetic field is the best studied of all cosmical magnetic fields. Along with the magnetic field of the Sun, it constitutes the most unportant paradigm of cosmi- cal magnetic field generation. Detailed global measurements of the contemporary geomagnetic field have been made for over 150 years- a period of time adequate to discern the dominant fea- tures of the field's secular variations. A substantial record of the geomagnetic field's long-term behavior has been extracted from paleomagnetic studies of magnetized rocks and from the magne- tization patterns preserved In the sea floor as new oceanic crust cools and spreads away from the mid-oceanic ridges. This paleo- magnetic record yields a fairly complete history extending several hundred million years into the past, although, because of sam- pling problems, the record is probably not reliably complete to time scales less than several tens of thousands of years. Perhaps the most startling behavior recorded in the geomag- netic record is the phenomenon of geomagnetic reversal. From time to time, the geomagnetic field spontaneously and suddenly changes its polarity. Intervals between polarity reversals are of ran- dom durationaveraging some 200,000 years, but with a nearly Poisson distribution. In the intervals between reversab the field exists in a quasi-static, but noisy, state. The actual reversal events proceed very quickly, taking only some 5,000 years. This time is far shorter than the free decay time of the dipole mode, thus requiring that reversals involve an active dynamical mechanism rather than
62 a passive shutting down and restarting of dynamo regeneration within Earth. The geomagnetic field aLso exhibits occasional large, but ap- parently short-lived, excursions from its normal structure. Be- cause of the difficulty of extracting from the paleomagnetic record detailed structural information about short-lived events, these transients are not well mapped. However, it is possible that the geomagnetic field might occasionally resemble what we have re- cently measured at Uranus. Studying Planetary Magnetic Field It is probably safe to say that the fundamental basis of mag- netic field generation in planets is understood. The physical and mathematical machinery of Maxwell's equations and Newton's laws has led to a dynamo theory of magnetic field generation that accounts for basic aspects of the existence and behavior of known magnetic fields. However, beyond that, our understanding quickly fades to a frontier of unanswered questions. A hydromagnetic dynamoconsisting of a coupled system of fluid and magnetic field has many degrees of freedom, leading to a highly complex set of possible behaviors. It is not feasible to explore theoretically the full range of mathematically possi- ble behaviors. An effective approach to understanding requires both theoretical study and detailed observations of actually oc- curring magnetic fields. Theoretical investigations create our un- derstanding of the possible varieties of dynamo magnetic fields, their behaviors, and the conditions under which these occur. De- tailed measurements of magnetic fields guide our knowledge about which among these theoretical possibilities are actually realized in physical systems and provoke a more penetrating theoretical understanding by revealing unanticipated phenomena. Because magnetic fields are generated by fluid flows, a major deficit in our understanding of any natural magnetic field results from continuing uncertainty about the detailed physical character of the fluid flows themselves. In the specific case of planets, ig- norance of the state of interior fluid and its motion inhibits the predictive power of even the incomplete theories now in hand. The sources of energy and buoyancy that drive fluid convection in planetary interiors are not well understood. Even for Earth,
63 it is not yet known what is the motive force that drives convec- tion in the core fluid; the most popular candidates are radioactive heat generation and differentiation of heavier materials toward the center. Most of our present theoretical understanding of magnetic field generation is confined to the so-called kinematical theory- or, equivalently, the weak-magnetic-field limit of the theory in which the fluid motion is negligibly perturbed by the magnetic field's Lorentz force. It is possible that many peculiar behaviors of natural magnetic fields will only be understood on the basis of more complete theories that can take into account the effects of the magnetic forces. So far, the mathematical tools with which to accomplish this are in a rudimentary state. This problem is likely to be overcome only by application of the largest computers now · coming Into use. Observationally, it is important to extend our knowledge of the natural magnetic fields existing in the solar system and the variety of their behaviors. Up to now we have been surprised at every turn, for the most part by the startlingly pervasive occur- rence of cosmical magnetism and its astounding variety of forms. Important advances in understanding will come from the combina- tion of information obtained about the magnetic fields of planets with the much greater detail of knowledge that can be extracted from Earth and the Sun. What understanding has been achieved for the Earth and the Sun has largely resulted from our ability to make measurements of the temporal variations, as well as the spatial structures, of those magnetic fields. Because of its accessi- bility, Earth is a crucial target for detailed study of its magnetic field behavior; what is learned from studying Earth will influence our understanding of all cosmical magnetic fields. It is especially important to ascertain the spatial structures and time dependence of the geomagnetic field on both large and intermediate scales, and over both short and geological times. Measurements of the tempo- ral variations of other planetary magnetic fields are also of great importance. Because of the commonality of physical processes in- volved in cos~riical magnetic field generation, deep understanding will be achieved only from an integrated approach. It should be noted that Jupiter, Saturn, and Uranus have large and dynamic magnetospheres. Exploration of these objects should go hand in hand with intensive study of the Earth's magnetosphere
64 and tail. This area is discussed in the report of the Task Group on Solar and Space Physics. PRIMITIVE BODIES AND THE ORIGIN OF THE SOLAR SYSTEM A major goal of space science is understanding how the Sun and planets were formed. Progress toward achieving this goal in- volves synthesis of knowledge obtained from many sources, includ- ing missions to planetary bodies, astronomical observations from earth-based and earth-orbiting observatories, and geological, geo- physical, and geochem~cal studies of the Earth itself. Knowledge from these sources is then combined with theoretical investigations that link observational data to fundamental physical and chemical laws and processes. A special role is played by observations of bodies that have been relatively unaltered since the formation of the solar system: comets, asteroids, and the meteoroidal and meteoritic fragments derived from these bodies. Subsequent sections in this report address our anticipated state of knowledge of these bodies in 1995, the principal questions that are likely to be unanswered at that time regarding these bodies, and the programs and technical needs required to address these questions. It is not appropriate to discuss the broader questions of origin in this report. However, the task group has provided a discussion of how work on this large problem may be expected to proceed. There emerges from this discussion a proposal that does involve specific technical requirements related to the planned space station. This is the search for planetary systems other than our own. This exciting opportunity represents a unique advance in our approach to planetary studies, and is fundamental to future progress in understanding how our own planetary system and life on Earth began. The Origin of the Sun and Planets How the solar system was formed is an unresolved question fundamental to planetary science. Answering this question is a prime objective of the NASA planetary program. Serious attempts to devise a theory for the origin of the solar system go back more than three centuries.
65 During the last decade 8 qualitative change has developed in our approach to this problem. This has largely been stimu- lated by the great wealth of information returned from lunar and planetary missions, by astronomical observations of young stel- lar objects, and from laboratory investigations of extraterrestrial material: meteorites, lunar samples, and interplanetary dust par- ticles. What once was a field of science populated by a few lonely thinkers defending complete but idiosyncratic theories of solar system origin is now a field characterized by a healthy interplay among theory, observation, and experiment. In association with this change, there has been significant evolution in thinking concerning the origin of the solar system. The picture that ~ emerging is one in which formation of disks of dust and gas ~ an essential component of the star formation process. Whether stars form out of such disks, or whether the stars are the principal source of the disks is not yet clear. It is now generally accepted that these disks provide the birthsite for planetary companions to stars. While this general picture Is similar in a general way to previous "nebular" models, there is now a growing body of observational evidence that supports it. Another principal characteristic of current work on solar sys- tem origin Is its strongly interdisciplinary nature. The formation of circumstelIar disks ~ inferred from infrared and radio astro- nomical observations of cool, dense interstellar clouds, and den- sity irregularities (called ~cores") contained therein. Aspects of stellar astrophysics are also relevant because meteoritic studies show that the early solar system contained significant quantities of now-extinct, short-lived radioactive isotopes produced in stellar nucleosynthesis, such as 26Al, that were rapidly mixed into the proto-solar nebula, and may have been an important heat source in the earliest solar system. In addition, other products of stellar nucleosynthesis were incompletely mixed into the material from which the Sun and planets were formed, as evidenced by other "isotopic anomalies in meteorites. Knowledge obtained from lunar and planetary missions is also central to building our understanding of earliest solar sys- tem history. The detailed chemical and isotopic compositions of the various planetary bodies including satellites, comets, and asteroids are particularly relevant because of their bearing on our understanding of processes and thermodynamics within the solar nebula. Sometimes these planetary observations match in
66 a natural way current thinking of early solar system history. An example of this is the evidence for extensive early cratering found on the less geologically active planetary bodies. This record fits with that expected from the sweep-up of residual planetesimals following the principal stage of formation of these planets. How- ever, unsettling surprises also occur. Perhaps the best example of this was the discovery by Pioneer and Venera spacecraft that Venus' atmosphere contains 50 to 100 times as much 20Ne and 36Ar as Earth's atmosphere. Not only was this unpredicted by all present theories of planet formation, but even now reconciliation of these observations with theory ~ in an unsatisfactory state. Meteorites represent another highly rewarding source of data relevant to solar system formation. Most meteorites are fragments of rock and metal, broken off during collsions among asteroids. Detailed laboratory chemical, petrographic, and isotopic investi- gations of this material continue to provide a wealth of information regarding conditions that prevailed in the primordial solar nebula, as well as of processes that occurred during the formation and subsequent history of these small planets. Finally, theorists endeavor to model the natural evolution clef gas-dust disks into stars and their associated planetary bodies. At present we are far from achieving anything that might be called a definitive understanding of this evolution of a gas-dust disk into a planetary system. C)n the other hand, on a number of aspects of the problem there are developing "shared understandings" that are likely to evolve into a consensus. An example is the general agreement that the Sun and planets evolved from a disk of dust and gas on a time scale of ~108 years. It is also generally recog- nized that achieving an understanding of the mechanisms by which angular momentum, energy, and mass are transported during the evolution of this disk is essential. A number of investigators are engaged in theoretical studies directed toward evaluating alterna- tive physical processes (e.g., turbulent viscosity or gravitational torques) by which this transport is effected. At the opposite end of the sequence of events leading from a gait disk to planets rs the final accumulation of the Earth and other terrestrial planets from planetesimals. A considerable quan- tity of analytic and numerical work has been carried out on the dynamics of this stage of planetary formation. Although many unanswered questions remain, the answers to several important
67 questions appear to be quite model-independent (i.e., not depen- dent upon whether the accumulation took place in the presence of nebular gas or after the dissipation of this gas). If the terrestrial planets did indeed form from the accumulation of planetesunals, it is found that these planetesimals must have included quite large objects, greater than 100 km in diameter, and possibly as large as the Moon or even Mars. Accumulation of such large bodies onto a planet will result in high initial internal temperatures, inasmuch as the mechanisms for radiating into space the large quantities of energy resulting from the impacts of bodies this large are in- efficient. Geologically important corollaries are the prediction of extensive partial melting during the formation of the Earth, pri- mordial chemical differentiation, and formation of the Earth's iron core simultaneously with the formation of the Earth. With regard to the formation of the outer planets, serious fundamental problems exist. It seems most likely that solid cores with masses of about 15 times the mass of Earth formed first, which then accumulated gas from the nebula. There are very good, if not compelling, reasons for believing that these cores formed before the final accumulation of the terrestrial planets. However, at least in their most simple form, present theories of planetary accumulation lead to the opposite conclusion, that the accumulation of the outer planets required much more time than was required for the growth of the terrestrial planets. Promising suggestions that could resolve this paradox have been made and are now awaiting serious theoretical evaluation. If the first asteroidal flybys occur in the next decade, we can expect them to begin to help us place the great wealth of mete- oritical data into a planetological context. This should make a significant contribution to our ability to evaluate such questions as the extent to which the available meteoritic sample is represen- tative of the asteroidal region, and to address puzzling questions regarding the apparent limited variety of meteorite sources. These exploratory encounters can also provide basic information that will facilitate the intelligent design of subsequent, more sophisticated asteroidal missions. During the next decade it may also be expected that basic in- formation regarding early solar system history will continue to flow from laboratory study of meteorites and stratospheric collection of interstellar particles. In this connection, it should be pointed out that to a large extent the current laboratory instrumentation
68 used in this work was obtained in the course of lunar sample anal- ysis during the 1960s and early 1970s. Attention must be given to modernizing the laboratories in which this work ~ done if this potentially major source of information is to be fully utilized. Among the fundamental questions in the area of solar system · origin are: Formation of Mingle stars. Recent observations have shown that perhaps as many as 90 percent of Al stars are mem- bers of binary systems; the Sun appears somewhat special in this regard. In order to specify the physical conditions in the earliest phases of the solar nebula, it is necessary to understand the way in which single stars are formed, how the process differs from that for formation of binary systems, and ways in which the formation of a planetary system is related to the way in which binary stellar systems are formed. . Nonstellar companions of the Sun. The present paradigm for formation of the solar system involves a nebular component out of which the Sun forms. A fundamental question is whether other large, but nonstelIar objects can form in the nebula at the same time that the Sun is forming. This question is of dual significance; it relates to whether there is a generic relationship between the formation of planetary and stellar binary systems, and it relates to the question of how protoplanetary objects may be formed. In this latter sense this question is a basic part of the question of how planets form. . Residual presolar disks. In order to develop a clear un- derstanding of the evolution of the solar nebula as that evolution affects the formation and evolution of individual components, it is important that we characterize the nebula and its properties as much as is possible. Particular emphasis should be placed on the mass and extent of the nebula, and its variation in time. Also of interest are the spatial and temporal variations of temperature and composition, both of which are essential to understanding the dovetails of planet formation. Formation of planelesimals. One of the possible models for formation of the terrestrial planets, and perhaps for the cores of the outer planets as well, involves the formation of small solid bodies called planetesimals. Progress in understanding this pivotal phase, especially in the context of specific nebular models, is central to gaining an overall understanding of the planet formation problem.
69 ~ Formation of the three major classes of planets. Estate fishing the manner in which each of the three major classes of planets- terrestrial, Jovian, and uranian formed is a central is- sue in understanding the origin of the solar system. The temporal aspects of this topic are of particular importance; for example, if Jupiter formed before the other planets it may have controlled to some extent the formation of the other planets. If, on the other hand, the terrestrial planets formed before Jupiter, the details of their formation are likely to differ substantially from those as- sociated with formation in the presence of a massive object like Jupiter. ~ Formation of satellites and small primitive bodies. Much of the information that we have concerning the very early history of the solar system comes from the study of smaller bodies. The reason that the small bodies are significant in this regard is that they do not undergo much metamorphic evolution and therefore retain a reasonably unaltered record of conditions at the time of their formation. It is important then to place the formation of these objects in the context of broader nebular models. ~ Loss of the nebular gas. One of the challenging problems in understanding the early evolution of the solar system is to identify when and how the bulk of the nebular gas (a hundredth of a solar mass or more) left the solar system. The timing is particularly relevant as it relates to whether accretion of solid material occurred in the presence or absence of significant amounts of gas. Probably at least Uranus and Neptune formed either after or during removal of the gas. In addition to understanding when removal occurred, it ~ important to understand the physical mechanisms involved in removing this amount of gas. There is growing evidence from astronomical studies that young stars that are thought to be similar to the Sun in its youth are associated with major bipolar mass loss. The relationship between this bipolar flow and dispersal of a circumstelIar nebula is unknown. ~ Initial states of the planets. A comprehensive study of the planets involves a characterization not only of their current state, but also of their evolution. An essential ingredient in this analysis is specification of the initial state of planets. Such specification includes temperature and compositional gradients or profiles. This type of information is provided through models of the formation of the planets and can be used in conjunction with observations
70 to assess, for example, models for formation and evolution of planetary atmospheres. Search for and Study of Other Planetary Systems Underlying all modern efforts to understand! the origin of the solar system is the assumption that the system is a natural re- sult of star formation and that similar events take place when other stars are formed. A consequence of this view is that most stars should be accompanied by retinues of nonstelIar compan- ions. Empirical evidence about this relatively simple assumption and its broader consequences form the only viable foundation for significant advances in our understanding of the origin of the solar system. A major challenge is to identify what features, if any, of the so- lar system are prototypical or representative of the general process of planetary system formation. For example, theoretical models are being developed that lead naturally to the expectation of the formation of jovian-mass planetary companions to all solar-type stars. Observational evidence either to validate this consequence or to refute it, and thereby provide an empirical foundation for a next generation of theoretical models, is thus essential to our knowledge of the origin of the solar system. Another fundamental challenge is to detect, study, and relate to theories of planetary for- mation the broad range of circumsteliar material and phenomena that may represent precursor or post-formation phases of other planetary systems. Anticipated State of Knowledge ire 1995. At the present time there is no unambiguous evidence for the existence of another plane- tary system, let alone detailed information concerning statistical properties of planetary systems in general, or of structural details of specific planetary systems. IRAS and some ground-based in- struments have detected discrete dusty structures associated with some nearby main-sequence stars, where the dust particles are some 2 orders of magnitude larger than typical interstellar dust particles (~100 Em versus ~0.2 ,um). However, these observa- tions do not of thernseIves signify detection of another planetary system. SubstelIar companions to nearby stars have also been de- tected with the use of ground-based instrumentation. As yet, these observations are unable to distinguish between coplanar planetary
71 systems similar to our own, and low mass members of more com- monplace double star systems. Advances are under way to develop instruments that will permit a search of sufficient accuracy to allow a quantum jump in the available data base. By 1995 we can expect to have 5 to 10 years of survey data from ground-based, special-purpose spectroscopic telescopes with detectors capable of measurements accurate to better than 10 m/s. While the spectroscopic technique of searching for other planetary systems does have its limitations, such a search will provide con- straints on our current models. The ability of this technique to detect jovian-mass companions in orbits with periods of a few years is significant, and evidence either positive or negative on such objects would have a useful, perhaps revolutionary, effect on our efforts to understand the origin of the solar system. It is con- ceivable that another planetary system could be discovered prior to 1995 using some of the ground-based instrumentation that is currently being developed (e.g., high-prec~sion spectroscopic and astrometric systems). It is very unlikely, however, that these shy tems will be able to go beyond a simple detection to provide significant data regarding the statistical properties of other plan- etary systems. It would appear that as of 1995, we will have only begun to collect the type of data essential to an understanding of the origin of the solar system. Detection of planetary systems therefore looms as one of the major challenges of any long-term plans in planetary research. There are four major questions associated with this area of planetary research. The first concerns the possible uniqueness of our solar system. This can be answered by detection of at least one other planetary system. If this has not been accomplished prior to 1995, it should be a priority effort thereafter. As exciting as the confirmed discovery of another planetary system would be, the major scientific advances of such a discovery will lie in answering the remaining three questions in this area of study. First, we must seek to characterize planetary systems on a statistical bash. For example: What is the frequency of occurrence of planetary systems as a function of spectral type of their central star? What are the masses of the planetary bodies relative to that of their central star? Answering these questions will require instrumentation capable of surveying a large number (100 to 1000) of stars to a level of accuracy that permits detection of objects comparable in mass to Uranus and Neptune. This
72 type of data will permit us to deduce the general features of planetary systems those features that must be explained by a comprehensive theory from features that are peculiar to a given planetary system. A third major question concerns the detailed nature of specific planetary systems. This type of effort must involve a longer time base as it requires precise determinations of planetary masses and orbits, and as much information on composition as possible. These types of data are required to both test and constrain models of processes within the gas-dust nebulae from whence planetary systems are presumed to form. A fourth question concerns the coo! material around stars in various stages of evolution. A central aspect of most modern theories on the formation of the solar system is the postulation of an accretion disk that evolves in such a way that it produces a central condensation (a star) and a suite of companions (planets). The previous questions focus on detection of the products of the disk evolution, the planetary bodies. No less important ~ an effort to characterize the nature and behavior of the disks associated with preplanetary systems, young planetary systems, and mature ones. Identification and study of this circumstelIar material wait help place the formation of planetary systems in the broader context of the problem of star formation. These four major questions can be addressed provided that certain technologies and instrumentation are developed in a t~rnely manner. The first three questions are most effectively addressed by an astrometric telescope of l~,uarcsec accuracy. Addressing the fourth question requires a low-scattering telescope to image and conduct spectrophotometric analysis of coo} circumsteliar mate- rial. Such systems must be located in space because turbulence in Earth's atmosphere places unacceptable limits on image size and stability at ground-based sites. In this regard it is important to note that the manned base of the Space Station could provide an ideal platform for such a tele- scope. It has the long-term stability and data-handling capability required for a system of this type. A roughly Midyear observing program of all stars within 30 parsecs of the Sun is envisioned for this telescope. The technology is at hand to build such a telescope, and it would be a relatively inexpensive device by today's stan- cards. This telescope and its scientific objectives provide perhaps the most exciting scientific use of the Space Station and should
73 therefore be considered a prime candidate for implementation. It could complete its survey within 15 to 20 years of its initial operation. Asteroid, SmaB Satellites, and Meteorites General Characteristics The asteroids are small bodies that orbit the Sun, for the most part at distances of 2 to 3.5 AU. The total mass of material in the asteroid belt is less than 0.001 earth masses. During the past 15 years there has been a phenomenal increase in our knowledge of the physical, chemical, and dynamical properties of asteroids. This has been achieved by earth-based telescopic observations of their positions, rotation, and spectral reflectance and emission. The largest asteroid, 1 Ceres, has a diameter of 1000 km, and there are more than 30 others larger than 20 km. This population grades down to more numerous smaller bodies; there are 105 to 106 asteroids greater than 1 km in diameter. The size distribution is such that most of the mass is found in the largest bodies, however. The mean densities of asteroids seem to be 2 to 3 g/cm3. Optical properties, including variations of optical magnitude with rotation, optical and near-infrared spectral reflectance, and thermal radiance have been measured for a large number of as- teroids. Most asteroids are distributed bimodally into one of two groups usually designated as C and S. Most abundant are the dark (geometric albedo of 0.03 to 0.04) C asteroids that exhibit rather featureless reflectance spectra. Both of these characteristics probably result from possibly small quantities of opaque minerals. This is consistent with identification of these asteroids as being of similar composition to carbonaceous meteorites, but the identifi- cation is not unique. Although there are marked exceptions (e.g., 313 Chaldaea), the C asteroids are concentrated toward the outer part of the asteroid belt. The S asteroids exhibit larger geometric albedos (~0.16) and absorption features attributed to mixtures of pyroxene, olivine, and metal. Qualitatively, they are thus mineralogically similar to ordinary chondrites, as well as to fairly well-bred igneous assemblages resulting from differentiation of asteroids of overall ordinary chondrite composition. Quantitatively, in most cases the relative proportions of these minerals do not appear to agree with
74 those of ordinary chondrites. It is not clear whether this discrep- ancy is only a surficial phenomenon, or of greater significance. It is quite possible that some S asteroids are differentiated bodies, whereas others are similar or identical to ordinary chondrites in composition. Some asteroidal reflectance spectra clearly indicate differenti- ation, such as E (enstatite) and M (metal). A number of asteroids, including many of the largest, do not fit into any of the standard classes. Some of these, particularly the 500 km-diameter body 4 Vesta, of basaltic composition, are presumably differentiated, whereas others, such as 1 Ceres and 2 PalIas may be undifferenti- ated. The orbits of asteroids intersect one another, and the expected frequency of asteroidal collisions (relative velocity of ~5 km/s) can be calculated in a rather straightforward way. There are signif- icant uncertainties in calculating the effects of these collisions, but it is very probable that almost all bodies less than 30 km in diameter are fragments of larger bodies, whereas asteroids larger than 100 km in diameter may be collisionally altered primordial planetesimals. The physical structure of all asteroids is undoubt- edly influenced by their collisional history. In detail, these effects are not clear, the possibilities ranging from rubble piles of possi- bly bizarre geometrical appearance, to thick, blocky regoliths, to nearly clean rocky fragments. There is ambiguous evidence that some asteroids represent multiple, mutually orbiting bodies. In the present solar system, almost all asteroids appear to be in orbits that are stable (except for the possibility of collision) on time scales that are long compared to the lifetime of the solar system (~10~° years). Exceptions are asteroids produced as colli- sion fragments adjacent to regions in which their orbital period, or other characteristic frequencies, are in resonance with the major planets, particularly Jupiter. The Earth-approaching Apollo and Amor asteroids are of some special interest. These bodies have perihelia within 1.3 AU, many of them pass inside the orbits of Earth and Venus, and a few pass within the orbit of Mercury. Close approaches to these planets cause the orbits of these bodies to be unstable on a time scale that is short compared to the age of the solar system. The observed Apollo-Amor objects thus constitute a steady-state population, and sources are required to sustain the observed population. These sources are probably twofold. Some bodies are fragments of main
75 belt asteroids, transferred into Earth-approaching orbits by the same resonant mechanisms responsible for the delivery of some meteorites from the asteroid belt to Earth. Others are likely to be the devolatilized nuclei of short-period comets, the residue remaining after ~04 episodes of solar heating during perihelion passage. An uncertain estimate of the relative proportions, based on interpretation of observations and theoretical reasoning, is that at least 20 percent of these bodies are derived from the asteroid belt, and that at least 10 percent are of cometary origin. The Earth-approaching bodies of asteroidal origin have a close kinship to meteorites, and many meteorites are probably collision fragments of these bodies. It is plausible to believe that a mission to a body of this kind would provide similar information to that obtained by a mission to a main belt asteroid of the same size but at a lower cost, because of the lower energy requirements characteristic of many of these near-Earth orbits. For example, the large Amor object 433 Eros (~2~km diameter) is likely to be of asteroidal origin. If so, it is undoubtedly a collision fragment rather than a primordial body, and may be expected to display the internal structure of a large asteroid. It is large enough to retain on its surface some of its collisional ejecta, facilitating collection of this material. The Earth-approaching bodies most likely to be of cometary origin tend to be in higher-velocity orbits than those of asteroidal origin. It is not out of the question that some of the low-velocity bodies may also be of cometary origin, but there is no observational evidence supporting this at present. A mission to an extinct comet may be able to study and sample cometary material difficult to access during the active phase of a comet's life. Such a mission may also provide information regarding the possible relationship between comets and some carbonaceous meteorites. Although earth-based observations have resulted in remarkable progress in our understanding of asteroids, qualitatively new approaches will soon be required if this progress Is to continue. Anticipated State of Knowledge in 1995 Space Missions. The next major step in asteroidal science is clearly in situ studies of asteroidal bodies. Missions being planned for encounter prior to 1995, if carried out, provide a first step in this program. They represent two or three fast flybys of asteroids
76 by spacecraft on the way to other prime target planetary oh jects. These include Galileo, CRAP (Comet Rendezvous/Asteroid Flyby), and the Saturn orbiter/Titan probe. These missions will be equipped with imaging that can achieve on the order of 100 m resolution of the asteroid surface. In addition, missions will map the surface in several spectral regions to provide mineralogi- cal information. The observations should be of significant help in the interpretation of data from ground-based spectrophotometric observations limited to integral whole-disk reflectance spectra. These missions will begin to address first-order questions such as large-scale topography, density, and composition of the aster- oids. For the first time we will actually be able to see what an asteroid looks like, and will bring asteroidal observations to a state comparable to lunar studies prior to the space program. The Voyager missions have imaged Amalthea and many of the smaller satellites of the Saturnian system, but at a resolution inadequate to tell much about the nature of the surfaces or the processes that have operated on these bodies. Phobos and Deimos are the only small satellites imaged with adequate resolution to tell much about the nature of the surface (5-m resolution was achieved by the Viking missions). The U.S.S.R. plans to make additional measurements of the martian satellites during the Phobos mission in 1988. Ground-based, Earth-orbital, and Laboratory Studies. Important ground-based and laboratory studies of asteroids, minor satel- lites, and meteorites may be expected to continue during the next decade. Much of this work falls into the category of "small sci- ence,n i.e., work that is best not programmed in advance, but allowed to evolve naturally through short-term, peer-reviewed grants. It is obviously difficult to determine what we will have learned from these studies by 1995, but significant advances are likely. For example, the number of known asteroids, including those in near-Earth and Mars-crossing space, will grow substantially. A significantly increased number of near-Earth asteroids will be available for spectroscopic studies and comparison with the popu- lation of main-belt asteroids. The number of candidate near-Earth asteroids for future missions also will increase significantly. Studies of asteroids by the Space Telescope (ST) also may advance our knowledge of the asteroids. The ST will provide
77 images of about 20 km resolution at 2 AU from the Earth and will determine the shape of many asteroids. In addition, the high spatial resolution will allow reflectance spectra to be taken at many spots to test for surface heterogeneity. The capabilities of ST will also allow reflectance spectra to be taken in the ultraviolet region, and perhaps in the future, in more of the infrared region. These regions may show more diagnostic features than have been available from ground-based studies. Many asteroids have been grouped according to spectral re- fiectance features, and many of these features have been related to known meteorite types. One difficulty in the link between asteroids ant} meteorites is that the reflectance spectra of the most common meteorites, the ordinary chondrites, are not precisely matched by the main-belt asteroid spectral reflectance groups. The closest match, some of the S asteroids, may actuary be the source of the ordinary chondrites if the surficial features of the asteroids have been modified by exposure to the space environment. By 1995 these processes may be better understood as a result of labora- tory studies, facilitating a closer observational link between the meteorites and asteroids. There are good dynarn~cal reasons for believing that most ordinary chondrite meteorites are derived from a limited (~0.05 AU) region of the asteroid belt in the vicinity of the 3:1 Kirkwood gap at 2.5 AU. (Asteroids with values of their sern~major axes in this region wiD have periods in resonance with the orbital period of Jupiter). Except for the questions of spectrophotometric inter- pretation mentioned above, the known larger S asteroids, in the vicinity of 2.5 AU are prone candidate sources for ordinary chon- drites. Most of the meteorite population is clearly derived from the main-belt asteroids, and their study provides our most detailed understanding of the nature and properties of the asteroids. Because of the diversity of studies on meteorites, it is dif- ficult to predict the state of knowledge in 1995. Prior to 1995, we can expect much new information will be obtained through the NSF-sponsored Antarctic meteorite collection program and the de- velopment of new analytical techniques. This collection represents a les~biased sample of the small meteorite end of the terrestrial meteorite flux and a substantial increase in the number of stony meteorites available for study. For this reason it may provide a look at meteorites derived from the asteroid belt at velocities so high that a significant yield of large meteorites ~ not expected.
78 This possibility appears to have already been realized by iden- tification of several meteorites of lunar origin In the Antarctic collection. A large group of meteorites, the achondrites, have clearly undergone igneous differentiation processes dated at 4.5 billion years nearly contemporaneous with the formation of the solar system. Besides giving evidence that at Idast some asteroids were once geologically active, the meteorites allow the study of igneous processes that occurred under a set of conditions unlike those of Earth or Moon, and thus broaden our understanding of the effects of different parameters that cannot otherwise be studied in the laboratory. In addition, most of the iron meteorites formed very early in solar system history in cores of many (up to 50) different asteroids. These data provide strong evidence for the presence of a substantial heat source capable of producing primordial melting of the asteroids (and possibly planetary planetesimals as well). Many of the meteorites are breccias (made up of fragments of other preexisting rock) and provide evidence for the nature of collisional processes at an early epoch. The meteorites, and hence the asteroids, also retain a record of the cosmic-ray intensity in the past, as well as a record of significant (up to a few gauss) magnetic fields that were present before the asteroids accumulated. Several of the carbonaceous chondrites have been found to contain amino acids and other complicated organic molecules, which are clearly of extraterrestrial origin. Carbonaceous chondrites also contain calcium and aluminum-rich inclusions that appear to be among the earliest objects to have formed in our solar system. These inclusions provide evidence of very complicated processes in the early solar nebula involving multiple episodes of high temperature (~1400K). They also contain a record of the nucleosynthesis of elements that were added to our solar system just before the inclusions formed. A problem associated with all of these meteorite studies, how- ever, is the lack of a geological context for the interpretation of the observations. An asteroid sample return mission would go a long way toward providing this context. It would also provide the in situ data that would substantially increase the value of the remote studies of the asteroids. Similarly, the SNC meteorites can answer a number of questions about Mars, but they are clearly derived from a lava flow, whose composition is unlikely to represent that of the surface as a whole.
79 Questions in Asteroid Science. Accessing the information that asteroids contain about the early solar system will require much more detailed understanding of the asteroids' surface composi- tions, their internal structure, the degree to which they are het- erogeneous, and their dynamical and collisional evolution. At present our knowledge of the composition of specific aster- oids is almost entirely founded on earth-based spectrophotometric data. Although we may expect this situation to change some- what between now and 1995, for almost all asteroidal bodies it will still prevail. Earth-based data represent a surficial average over the entire observed disk of the asteroid, and heterogeneity is at present only crudely exhibited as a consequence of aster- oid rotation. When looked at carefully enough, every asteroid appears spectrometrically unique, and different in detail from bad oratory spectra of meteorites. Are these differences fundamental or are they primarily the result of heterogeneity or exposure to the surface environment or regolith phenomena? Making these distinctions requires remote sensing data of higher resolution. In addition, corroborating "ground truth" data are lacking, and will be required to ensure that unanticipated mineralogical differences are not overlooked or misinterpreted. It is likely that asteroidal collisions have exposed interior re- gions of asteroidal bodies. When combined with higher resolution spectral and imaging data, this affords an opportunity to overcome the apparent irritation to surficial composition. This may facil- itate addressing such questions as the fragmentation history and internal structure of differentiated asteroids, and the identification of specific asteroids as sources of particular meteorite classes. A better understanding of this asteroid collisional evolution is also needed In order to learn which asteroids are primitive objects, as opposed to collision products of larger bodies, and whether they are best thought of as "rubble piles," megaregoliths, or simply as solid rocks. In this same connection, present observations of Hi- rayama families serve as a key source of information concerning as- teroidal collision phenomena. Their apparent spectrophotometric heterogeneity calls into question the basic assumptions supporting these inferences. With these more detailed compositional data it will also be possible to address the relationship between the heliocentric dis- tance of asteroids and their chemical composition. Understanding this possible relationship is central to examining the conventional 1
80 assumption that the formative solar system had a marked radial temperature dependence, related to the composition of planetes- imals and planets formed at different locations in the early solar system. COMETS General Characteristics Comets are thought to be small conglomerates of rock, ice, and dust several kilometers in diameter and 10~5 to 10~8 g in mass formed during the early years of the solar system's history. Today most reside in the so-called Oort cloud, in loosely bound orbits at tens of thousands of astronomical units from the Sun. Perturbations induced by the gravity of passing stars and inter- stelIar clouds occasionally alter a comet's orbit and send it near the Sun, where solar heat evaporates the ice. The subsequent outflow of gas and entrained dust, illuminated by sunlight, produces the comets. Some of these comets that venture into the solar system are further influenced by the gravity of planets, becoming trapped in periodic orbits near comets. Some of these comets that ven- ture into the solar system are further influenced by the gravity of planets, becoming trapped in periodic orbits near the Sun. Such periodic comets appear regularly for thousands of years until their volatile material is depleted. Several schools of thought hold that comet-like objects were among the fundamental building blocks of some larger planetary bodies. As a result of their small sizes and their large average distances from the Sun, evolutionary processes that differentiated the planets are thought to have been insignificant for many comets. They may have played a role in later states of planetary evolution, perhaps by providing volatile constituents for some atmospheres. It has been speculated that some of these cometary constituents were essential to the origin of life. A bright comet appears as a roughly spherical coma or at- mosphere composed of comparable quantities of dust and volatile species such as neutral gases and ions. The curved, relatively featureless dust tad! directed almost exactly away from the Sun shows considerable temporal and spatial structure. Depending on a comet's distance from the Sun and the light in which it is oW served, its coma can be quite large, in the range of 104 to 107 km
81 when a comet ~ at 1 AU. The plasma and dust tails are even larger in the case of a bright comet, some 107 to 108 km in extent. All of these phenomena result from the gas and dust that emanate from the nucleus; the nucleus itself is small, of the order of 1 to 10 km in diameter for typical comets. The ice and snow in comet nuclei are composed of condensed gases and other volatile materials, including water and probably carbon monoxide, carbon dioxide, HCN, CH3CN, and uniden- tified complex organic molecules. These species are the parent molecules of the molecules and ions observed in the coma and tail. The nonvolatile material is in the form of grains ranging from subrn~crometer-sized dust to sand grains and perhaps pebbles and boulders, containing silicates and possibly metals, oxides, sulfides, and organic compounds. When solar heat vaporizes the volatile material, the outflowing gas carries smaller solid particles with it. Since the comet's gravity is finite, though weak, any larger peW bles and boulders are likely to remain bound on the surface-of the nucleus possibly leading to the formation of an "extinct comet of asteroidal appearance. Some of the earth-approaching "asteroids" may be highly evolved comets of this kind. Since many comets show evidence of directional emission of gas and dust, it appears that the surfaces of their nuclei are inhomogeneous and may have localized active regions. A tentative estimate of the volatile fractions of four recent comets has been made from their apparent production rates of carbon, oxygen, and nitrogen. Although the parent molecules are uncertain, they seem to be composed mainly of hydrogen, carbon, nitrogen, and oxygen. The mass ratio of the dust to gas liberated from a nucleus has been estimated in two cases to be 0.5 and 1.7 within a factor of 2. By comparison, the ratio of volatile to nonvolatile components is about 100 for solar material. These results imply that hydrogen and helium are depleted in comets. Nevertheless, comets seem to contain 3 to 10 times as much volatile material as the most volatile-rich meteorites. Thus comets appear to have formed from material at temperatures much lower than that characteristic of meteorites, at about 150K as opposed to more than 400K. This suggests that cometary material is the least- differentiated and best-preserved product of the preplanetary solar nebula that is known to remain in existence. It is speculated that some fraction of comet dust may be unaltered interstellar material. It may be possible to illuminate
82 this question by establishing some elemental isotopic ratios. Rel- ative isotopic abundances of the elements reflect their formation processes. Carbon Is a good example. Bodies within the solar system, including the Sun, Moon, terrestrial planets, meteorites, and Jupiter, exhibit a common value of about 90 for the i2C/~3C isotopic ratio. Red giant stars show a range from 12 to 50 for this ratio. In carbon stars the ratio falls in a wide range from 2 to 100. Although observed values in the interstellar medium also span a wide range from 13 to 105, some investigators have argued that a value of 40 Is representative. Many of the so-called "BrownIee particles," which are collected in the stratosphere, are suspected to originate In comets. At best, however, they are necessarily deprived of their most volatile components. State of Enowledge In 1995 Our ideas about comets are constructed from lirn~ted remote observations as well as from our more general notions about solar system bodies. Significant advances in our knowledge of comets are expected during the decade preceding 1995, as a result of ini- tial spacecraft missions carried out by a number of different space agencies. The NASA fly-through of Comet Giacobini-Zinner con- ducted the first in situ measurements of a comet. The instrument suite carried by that spacecraft was designed for solar wind and magnetospheric studies. The primary contribution of the mission was to characterize the particle and field distribution around the comet and to establish the basic features of that comet's ~nter- action with the solar wind. Several important physical problems were addressed by the Giacobini-Zinner mission, including the medium-energy, nonthermal particle distribution, the morphology of the magnetic field, the character of the interaction between cometary gas and the solar wind plasma, and plasma instabil- ities excited by the interaction. Although the Giacobini-Zinner investigation was the first in situ comet investigation, neither the encounter orbit nor the instruments were chosen with the comet in mind. The same is true of the Pioneer Venus orbiter, which obtained images in radiation scattered by atomic hydrogen. The three scientific spacecraft that encountered Halley's Comet were specifically designed for the purpose. The ESA Giotto project aimed for the closest approach to the comet's nucleus- passing as close as 500 km. Giotto passed through all of the outer
83 comet/solar wind interaction layers and into the inner cometary coma. The instruments carried out a first-order characterization of the solar wind interaction, a crude characterization of the den- sity and composition of cometary gases, and a crude analysis of the elemental composition of cometary dust. In addition, a limited number of medium-resolution pictures of the comet's nucleus were obtained. The Soviet Halley project sent two spacecraft through the outer regions of the comet's interaction with the solar wind. Gen- erally speaking, the information from the Soviet mission was sim~- lar to that from Giotto. The Japanese Suisei and Sakigake missions provided coordinated information on the solar wind flow upstream of the comet. Altogether, the investigations of Comets Halley and Giacobini- Zinner achieved a gross characterization of the morphology of large-scale cometary phenomena. As valuable as these investiga- tions were, most of the highest-priority questions that challenge our understanding of comets and that promise to reveal clues about the early solar system remain unanswered. Detailed understanding of comets will require more intimate and extensive measurements than are accessible to flyby investi- gations. Comet rendezvous and comet nucleus sample return will be the means by which the major objectives of cometary science will be realized. The United States will carry out a comet ren- dezvous mission by the middle of the l990s, which will follow a comet through most of its inner-solar-system passage. Analyses of cometary gas and dust, energetic particles, and magnetic fields and plasmas, as well as detailed investigations of the structure and gross composition of the comet nucleus will be carried out by the comet rendezvous mission. Successful completion of the comet rendezvous should an- swer many outstanding questions about the gross characteristics of cometary features and phenomena. The next obvious step in the study of comets will be comet sample return. A crude sample return may be accomplished by flying a collector, at high velocity, through a cometary coma. However, the material returned in this way will retain only the information about its basic elemental and isotopic composition. In some ways, this situation resembles that of the stratospheric "BrownIee particles," which do in fact retain
84 much of their original structure. Valuable as they are, these par- ticles cannot be identified with any particular comet, and in fact may not even arise from comets at all. By 1995, the next major step in cometary science will be to return an intact sample of comet nucleus material. This material can then be analyzed in detail in laboratories in order to carry out the mineralogical, chemical, and isotopic analyses that are needed to unravel the formation processes and evolutionary history of comets. Questions in Cometary Science It is expected that as a result of spacecraft missions and earth- based studies, our knowledge of comets will increase significantly during the next decade. It ~ clear that those results will represent only the beginning of our quest for understanding these most primitive and unevolved aggregates of matter assembled during the birth of the solar system. An important set of questions concerns the description of the present state of cometary nuclei. These include descriptions of the chemical, isotopic, and mineralogical composition of comets, their internal structure, and the range of variation of these characteris- tics between different comets. Questions of another class address cometary processes. As oW served, comets are complex systems of neutral gas plasma, dust, and larger solid bodies. We must understand the physical pro- cesses that determine the loss of material from the cometary nu- cleus and the resulting short-term evolution of its structure, and those that produce the elaborate extended coma and tail of the comets. Some of these same physical processes determine the nongrav- itational evolution of cometary orbits, that is, a Rocket effect" caused by the asymmetrical emission of gases from the nucleus. These altered orbits can have a substantial effect on the probabil- ity that a comet will impact planets and are central to the question of the evolution of active comets into Apollo objects of asteroidal appearance. This knowledge of the present state of comets and the physi- cal processes to which they are subject is required to understand the properties of earth-impacting material derived from comets. It is known that meteors, meteorites, and cosrn~c dust represent
85 the impact of cometary and asteroidal materials, but the relation- ship of this material to their sources is imperfectly understood. Like returned samples, laboratory study of this material provides important information regarding these bodies and their origin. Characterization and identification of possible cometary material are required. The resulting knowledge of present-day cometary composition, structure, and processes is fundamental to understanding where, when, and how the comet nuclei formed. In particular, we need to understand the age and any alteration of the various components of cometary nuclei. To what extent do they represent evolved solar system material, early solar nebula condensates, or solids of interstellar origin? This extrapolation back to the time of the origin of the solar system is also needed to interpret the extent to which comet-like objects contributed to the formation of the giant planets and the volatile inventories of the terrestrial planets. Comet Measurements and Technical Requirements Addressing the scientific questions cited above will require detailed investigations of the comet's nucleus and the emitted dust, as well as the gas, plasmas, and fields in the comet. Comelary Nucleus. Investigations of the comet nucleus should be aimed at establishing its composition and its physical and structural characteristics. Compositional measurements should determine the atomic, molecular, and mineralogical content of the refractory and the volatile solids. Together with measurements of the physical and structural features, this information will help in ascertaining the history of cometary matter and the processes responsible for its formation and for the assembly of cometary nuclei. Accurate isotopic measurements should be carried out on both the refractory and volatile constituents to explore the nucleosyn- thetic history and establish a time scale for major events in the history of cometary material. Both the compositional and struc- tural investigations should be extended over the variety of physical scales that characterize cometary nucleus material; this covers a range from the microscopic grains to the full size of major macro- scopic components of cometary nuclei. It is desirable to identify the major mineral assemblages of the nucleus for those constituents
86 that make up 5 percent or more of the comet's composition, and with a resolution of better than 10 percent of the nuclear diameter. On the small scales, complete determination of the structure and composition of cometary material requires detailed analysis of the dust component to discern the dust's character and origin. Cometary Atmosphere. Measurements of the gaseous component of cometary effluent will reveal information about the most volatile materials in the nucleus. Spacecraft should have the capability to identify and determine the abundances of all molecular species in the mass range 1 to several hundred, and to determine the isotopic ratios for the important species at an accuracy sufficient to identify deviations from solar composition and other anomalous isotopic variations. A major purpose of the cometary atmospheric mea- surements is to ascertain the composition of the so-called parent molecules of cometary effluent and their evolution as they leave the comet. Both the neutral and ionized components of the cometary atmosphere should be analyzed in detail, including the variations in composition with distance from the comet's nucleus. Solar Wind Interaction. The structure of the large-scale cometary phenomena should be deterrn~ned. Measurements of the electro- magnetic fields, plasma, energetic particles, and neutral gas should be made in a volume surrounding the nucleus and encompassing the upstream coma and solar wind interaction region, as well as a significant volume of the tail. Measurements should be made with sufficient spatial and temporal completeness to allow identi- fication of the major dynarn~cal physical processes, including tran- sients, that play important roles in shaping the overall cometary structure and providing its sources of energy. Technical Capabilities. Cometary studies require a sophisticated complement of spaceborne and laboratory instrumentation that will not be discussed in detail here. Scientific investigations of comets need a launch system able to reach a variety of orbits and to maneuver for an extended period of time near the comet. At present, low-thrust electric propulsion seems to provide, by a wide margin, the best propulsion system for such missions. Achieving the scientific objectives of comet exploration will require missions of months' duration designed to carry out ex- tended investigations of a single comet during both its most active
87 and its more quiescent times. Return of high-~ntegrity samples of cometary nucleus material to Earth for detailed analysm in ter- restrial laboratories will be essential to realize the overall goals of comet science, as well as of solar system science in general. However, adequate analyses should be carried out in situ, because these materials may well undergo significant changes once they are removed from their natural environment.
