One of the major goals of planetary exploration is to determine the surface histories of the solid planets and satellites. Surface histories tell us how these bodies evolved through time and provide information on the probable causes for observed differences. To establish a surface history, it is necessary to determine the sequence of various geologic events and, if possible, their duration. Two basic types of dating are possible: absolute and relative. Absolute age dating determines the "calendar" time at which a rock, surface, or feature formed; relative age dating determines the order-but not the time-of formation.
3.1. Absolute Dating
The traditional and most reliable method of absolute age dating requires laboratory analysis of samples. Most rocks contain small amounts of radioactive isotopes, such as 238U, 235U, 40K, and 87Rb, which decay at known rates. If the rocks have remained as closed isotopic systems, it is possible to calculate their age by measuring the amount of radiogenic isotopes relative to the amount of stable isotopes now present. In practice, this procedure requires an accurate assessment of the initial abundances of the isotopes produced in the radioactive decay. The problem becomes intricate if more than one event that affected the radiogenic isotope systems has occurred during the evolution of the rock. Many rocks have complex histories, and the challenge in isotopic age determination is to unravel and date not one, but each of the events that affected their evolution.
 To date, only terrestrial, lunar, and meteoritic samples have been dated by isotopic methods. The oldest terrestrial rocks, found in the Precambrian shield of Greenland, are about 3.8 billion years old. Most of Earth's surface (the ocean basins) was formed by seafloor spreading during the last 200 million years (about the last 5 percent of geologic history). Hence Earth's surface is geologically very young. In contrast, the Moon's surface is very old. The youngest extensive stratigraphic units dated by isotopic methods are the mare basalts, which range in age from about 3.3 to 3.8 billion years. Rocks recovered from the lunar highlands are even older, and ages in excess of 4.3 billion years have been measured.
The isotopic method of determining absolute age is the most accurate and desirable way of dating planetary surfaces, but collecting and returning rock samples from distant planets and satellites is a difficult and expensive endeavor. Furthermore, some surfaces, such as those of the icy satellites of Jupiter and Saturn, may not yield rocks that are datable by current isotopic techniques. On the other hand, most solid bodies in the solar system display a record of accumulated impact cratering on their surfaces. The total number of craters recorded by a surface is a measure of its age. If the rate at which craters are formed is known, then it is possible to estimate the absolute age of the surface.
The present rate of crater formation can be estimated from telescopic observations of various planet-crossing objects. These objects include small bodies of asteroidal appearance and the nuclei of comets. The population of these objects can be estimated by statistical methods from the rates at which they are discovered by systematic searches of the sky (Shoemaker et al., 1979). Using the techniques of statistical celestial mechanics, first developed by E. J. Opik (1951), collision rates with a given planet or satellite can be derived from a knowledge of the orbits of these small planet-crossing bodies. Sizes and the size distribution can be estimated using various remote sensing techniques. An assessment of mineral composition can be made from spectrophotometric observations, and plausible densities and masses can then be assigned to well-observed small bodies (chapter 7). Cratering rates are estimated from the collision rates and from the masses and impact velocities of the colliding bodies by means of either empirical crater scaling laws or by more elaborate computer calculation of crater formation (Shoemaker, 1977).
 Significant uncertainties are associated with each of these steps, particularly with the assignment of masses and with the calculation of crater sizes. One vexing problem is that, although comets have been and remain important impacting bodies, our knowledge of the sizes and densities of their nuclei remains especially poor (Wilkening, 1982). Therefore, it is extremely important to obtain independent information on the present cratering rate on different planets or satellites to check and calibrate the cratering rate calculations. Fortunately, Earth provides one such check for the inner solar system. Accurate determinations of recent cratering rates on Earth are vital to the estimation of accurate absolute ages from crater densities on the terrestrial planets. The Earth-Moon system also provides the essential record needed to determine the past variation of this cratering rate (Hartmann, 1972a).
If the cratering history is known for one planet or planet-satellite system, then, in principle, it can be derived for other planets and satellites, provided that the bodies impacting the various planets and satellites are dynamically related. In the case of asteroidal bodies that collide with Earth, it has been shown that these bodies are closely related to asteroidal objects that impact the other terrestrial planets. It has also been demonstrated that the flux of comets in the neighborhood of the terrestrial planets is closely linked to their flux in the neighborhood of Jupiter (Shoemaker and Helin, 1977).
Crater densities indicate that the present rates of formation of large craters on each of the terrestrial planets and on the Moon are approximately within a factor of 2 of the present cratering rate on Earth. Ten kilometer diameter craters are produced on Earth at the rate of ~2 X 10-14 km-2 yr- 1 (Shoemaker et al., 1979). From the calculated present cratering rates and the observed history of cratering in the Earth-Moon system, it can be shown that the period of early heavy bombardment probably ended 3.5 billion years ago on each of the terrestrial planets (Hartmann, 1972; Soderblom et al., 1974). On Mars, the analysis indicates that volcanism and plains formation extended through much of the post-heavy bombardment period. Much work remains to be done, however, to refine the accuracy of these age estimates based on crater densities. More extensive telescopic observations are needed to improve our knowledge of the physical properties and collision rates of the planet-crossing bodies, and computer models must be refined to estimate  more accurately the crater sizes produced on various planetary surfaces.
Another method that has been used successfully on the Moon to estimate absolute ages involves the correlation of the morphology of small craters (1 km in diameter) with the absolute age of a surface determined from isotopic measurements (Shoemaker, 1966). The technique depends on an erosion model that relates the shape of a crater to the integrated flux of meteoroids and secondary debris that have impacted the surface since the crater was fresh. The method provides a means of estimating absolute surface ages in areas not sampled by the Apollo missions and suggests that some mare regions may be as young as about two billion years.
3.2. Relative Dating
Although it is not always possible to date a geologic event or surface on an absolute time scale, it may be possible to establish the order in which events occurred by the traditional methods of superposition and cross-cutting relationship among various geologic units. Material units that were deposited on other units clearly postdate the units on which they lie. For example, lava flows that are observed to embay crater rims or associated ejecta deposits were emplaced after the formation of the craters. Craters and their associated ejecta blankets that overlie these flows postdate the emplacement of the flows. Tectonic structures such as faults can be dated relative to other events by their cross-cutting relationships. For example, faults that cut the cratered highlands on the Moon but terminate at the boundary of a unit that embays the highlands indicate that a period of crustal deformation occurred after the formation of the highland unit but before the emplacement of the unit that embays it.
Another method of relative age dating involves the relative abundance of impact craters on different geologic units. This method makes no assumptions regarding the flux history of impacting objects and relies only on the principle that an older surface will have accumulated a greater number of craters than a younger one. An example is provided by the difference in the abundance of large craters between the lunar maria and highlands. The relatively high number of craters in the highlands indicates that the highland rocks are older than the mare lava flows, an observation that was confirmed by the Apollo missions. Although this method of relative age.....
 ....dating is rather straightforward, there are certain complications that must be taken into account. First, primary impacts produce large numbers of secondary craters that may be widely dispersed over the surface. The size of secondary craters depends on the energy of the primary impact and can attain diameters up to about 20 to 25 km for large basin-forming collisions. Should a significant number of secondary craters be included in a crater count, a spurious relative age will result. This problem can be minimized by counting only the larger craters and excluding obvious secondaries (those which occur in clusters or display irregular rim structure). Another potential problem is that certain geologic processes such as lava flows may leave the rims of large craters exposed. If the imaging resolution is not sufficient to determine whether or not a given crater is superimposed on a specific geologic unit, it may be erroneously included in the crater counts used to determine the relative age of this unit. Obviously, relative age dating between surfaces on different planets and satellites based on comparisons between crater abundances is not reliable without a knowledge of the orbits and the flux histories of the impacting objects and of the relative efficacy with which craters are produced on the various objects.
In summary, relative age dating based on crater abundances and on traditional superposition and transection relationships provides a powerful means of determining the sequence of events that have shaped a planet's or satellite's surface. The geologic history of a body can be reconstructed, and time constraints can be placed on its thermal history and internal dynamics.
3.3. Potential New Information on Solid Bodies
In the next decade improved calibrations of the cratering time scales and new observations by spacecraft can be expected to yield major advances in our understanding of the evolution of solid bodies of the solar system. Some of the advances that can be anticipated are outlined briefly below.
No new spacecraft observations of Mercury are expected in the next decade, but significant improvements in the calibration of the cratering time scale on Mercury can be achieved. The present cratering rate on the planet is close to the current cratering rate on the Moon. Collision of long-period comet nuclei may have produced  somewhat more than half of the recent large craters and planet-crossing asteroids somewhat less than half. This proportion is slightly different than that for the Moon, but, to a close approximation, the post-heavy bombardment cratering history of Mercury is believed to have followed the same pattern as that on the Moon. Studies of the highland crater size distributions suggest that the same family of projectiles struck Mercury and the Moon during late heavy bombardment (Strom, 1979). If the history of decay of late heavy bombardment on Mercury also followed the history of late heavy bombardment on the Moon, the youngest large basin discovered, Caloris, is comparable in age to the youngest large lunar basins, Orientale and Imbrium, and all of the plains units are older than 3 billion years. Significant improvements in the calibration of cratering time scales on Mercury can be expected as our general understanding of cratering rates in the inner solar system improves through continued observational and theoretical work.
Existing radar observations suggest that ancient cratered surfaces may be present on Venus and that the crater populations may be roughly comparable to those of the lunar highlands (Masursky et al., 1981), although a volcanic or tectonic origin for many of the crater-like features is also possible at this stage. Ages of major features of probable volcanic origin and the broad outline of the history of volcanism on Venus could be ascertained from more detailed radar imaging data. The great Ishtar Terra plateau is evidently underlaid by a thick slab of differentiated crust. Crater densities could determine the age of this feature and thus place bounds on the time at which global differentiation of Venus' crust occurred.
Extensive high-quality spacecraft observations are already available. Important advances in our knowledge of the evolution of Mars will come largely from further study of these data and from improvements in the martian cratering time scale. The present production of craters is dominated by shallow Mars-crossing asteroids. An intensive telescopic study of these asteroids may be expected to provide a fairly accurate cratering time scale. Present evidence suggests that the current rate of production of small (I km in diameter) craters is about twice as high as the rate on Earth (Shoemaker et al.,  1979). Ages for the youngest lavas on Olympus Mons and on the giant Tharsis volcanoes indicated by this cratering rate are less than 10 million years. As the eruptive histories of these volcanoes span at least several hundred million years, the young ages of the lavas suggest that these features are merely dormant and that Mars is a volcanically active planet (Carr, 1981). A better calibration of the cratering time scale on Mars would not only sharpen our understanding of the time development of martian volcanism, but could also yield reliable age estimates for the various fluvial channels. However, no such calibration will be secure until samples of the Martian surfaces have been dated by isotopic techniques.
Abundant isotopic age determinations are available for presumed asteroid fragments available to us as meteorites. The difficulty is that few (if any) meteorites can be identified confidently with a specific asteroidal parent body. Progress can be expected from theoretical studies of the dynamical mechanisms of delivery of meteorites to Earth and of the collisional evolution of the asteroids (Wetherill, 1977). Significant progress may also be made in determining the minerologic compositions of asteroid surfaces from remote sensing observations (chapter 7), thus aiding in the ultimate identifications of some meteorite parent bodies (Chapman and Gaffoy, 1979; Gaffey and McCord, 1979; Larson and Veeder, 1979). However, spacecraft missions to one or more asteroids represent the next major step in unraveling asteroid histories; it may even be feasible to obtain samples from some near-Earth objects.
3.3.5. Satellites of Jupiter and Saturn
Many of the large satellites orbiting Jupiter and Saturn are now known to have experienced complex surface histories (Smith et al., 1979b, 1981). Io, the innermost large satellite of Jupiter, has the highest rate of volcanic activity of any body in the solar system. Europa and Ganymede have been largely resurfaced by mechanisms that probably involved flooding of the surface by water. A preliminary evaluation of the rate of cratering by comet impact suggests that the entire surface of Io may be no older than a million years and that the surface of Europa is less than 100 million years old. Partial resurfacing of Ganymede probably occurred near the end of a period of heavy bombardment, probably between 3.5 and 4.0  billion years ago. The largest uncertainty in the chronology of events in the Jovian satellite system resides in our estimates of the sizes and masses of impacting comet nuclei. An intensive program of photometric and radiometric observations of comet nuclei at large heliocentric distances (where comets are least active) is needed to improve our knowledge of their dimensions; the calculated cratering rates depend critically on this parameter.
The geologic histories of Saturn's small icy satellites appear to be as complex as those of the much larger icy satellites of Jupiter (Smith et al., 1981). Successive surface layers were formed on Enceladus over the span of geologic time. Tethys, Dione, Rhea, and Iapetus show evidence of internal activity, and plains-forming eruptions occurred comparatively early in the history of each body. A complex cratering record is preserved on each satellite, but two factors complicate the interpretation. First, the present cratering rate probably is dominated by Saturn-family periodic comets, the existence of which is inferred largely from dynamical theory, rather than from actual observations. These comets exist far from the Sun and Earth, and their sizes and numbers cannot be determined by telescopic studies. Second, the early cratering history may include numerous craters produced by collisions with debris in orbit around Saturn. The rates and history of cratering by this debris are not readily determined from theory. Thus, the cratering history of Saturn's satellites, a key to understanding the evolution of their surfaces, will continue to be a challenging problem during the coming decade.
Dating techniques are essential to geologic studies of planets and satellites in that they provide the timing and duration of events that affected the evolution of the surfaces we see today. The most secure dating techniques remain those based on isotopic measurements, and great efforts should be made to extend these to other bodies (especially Mars). The direct dating of lunar samples refined our understanding of not only the Moon's evolution, but also of the evolution of terrestrial planets as a whole, to a degree that is impossible to overemphasize. There can be little doubt that the absolute dating of additional planetary materials would prove equally informative.
 Important dating information can also be obtained from careful crater counts. To interpret such crater counts fully, however, requires information on the current and past populations of impacting bodies. Although much progress has been made in recent years in evaluating the flux of impacting objects in Earth's vicinity, too few reliable data exist for the solar system beyond the asteroid belt. Specifically, our knowledge of the numbers, and especially of the dimensions, of comets is too poor to predict accurate cratering rates m the outer solar system. This situation could be improved significantly by a concentrated program of Earth-based observations of comets. Ample work also remains to be done in refining theoretical models that seek to tie in the current flux of cratering objects to its history over the past 4.6 billion years.
The importance of reliably identifying craters on the surfaces of Venus and Titan, and thereby estimating ages, cannot be overemphasized.