Craters can be used to determine how old one surface is relative to another because the rate of impacts over time is thought to have declined slowly over the past three-quarters of solar system history, having decreased quickly prior to that from a much larger initial rate. Surfaces that are young, that is, which have been renewed through lava flows, mountain building, erosion by water, and other geologic processes, will show fewer craters than surfaces that are much less active, or older. Because of this we can use the abundance of craters on various surfaces of a planetary body to determine, in a relative sense, when certain kinds of geologic processes occurred relative to others. The
excavation and beginning of
uplift
central uplift and rim collapse
final crater
excavation and beginning of
uplift
central uplift forms
peak ring crater
(a) (b)
Figure 7.2Stages in the formation of (a) small and (b) large craters. Modified from Melosh (1989, p. 142) by permission of Oxford University Press.
freshness of craters, that is, how bright their debris or ejecta blankets appear and how sharp their features are, provides an additional refinement to the relative dating process. It is also possible to compare relative ages from one planet or moon to another, provided one can calibrate the rate of impacts in one part of the solar system relative to another.
An example from the Moon provides a classic illustration. The bright areas of the surface of the Moon are very heavily cratered regions called thelunar highlands, as revealed by telescopes and images from lunar orbiting spacecraft (Figure 7.4a). In these regions the density of craters is so great that craters overlap with and are superimposed on each other; down to the limit of resolution on the image, one sees a scene filled with craters.
The dark portions of the Moon, on the other hand, consist of areas that are smooth and relatively devoid of craters. These mare (Latin for seas, the seventeenth and eighteenth century interpretion from telescopic views) show ample evidence of craters that have been partly covered or obliterated by the mate- rial that makes up the smooth, dark surfaces (Figure 7.4b). The simplest interpretation is that the mare are lowland basins that were flooded by lavas sometime after an early, heavy bombard- ment of the Moon occurred. The flooding obliterated most of the craters, leaving a fresh surface on which some remains of old craters can be seen, and a few small, new craters were formed by impacts after the lava solidified.
Explorations by the Apollo astronauts from 1969 to 1972 returned nearly 400 kg (900 lbs) of moon rocks from mare and highland regions. Radioisotopic techniques, described in Chapter 5, were used to provide absolute dates for the solidifi- cation of these rocks from original molten materials. The lunar highlands are old, with rocks dating as old as 4.5 billion years.
The mare deposits typically are 4.2 billion to 4.3 billion years old, significantly younger than the highlands.
The age estimates based on the cratering density on the lunar surface are confirmed by the absolute dating of mare and high- lands provided by rock samples. It then would appear possible to use crater densities on other worlds, not accessible for sample collection at present, to construct chronologies as well. The most
straightforward chronology involves determining relative ages of events on a surface, that is, which event preceded another.
This simply requires counting craters as well as looking for evi- dence of craters partly obliterated by geologic processes. More difficult is to try to assign actual dates, which requires assuming that the lunar crater density and ages based on Moon rocks can be transferred directly to other bodies in the solar system.
7.2.1 Relative ages of events on a planetary surface Impact craters can be used to determine the relative ages of geo- logic features on a planetary surface. Two examples of this are shown, one from Mars and one from Jupiter’s moon Ganymede, using images fromVikingandGalileomissions in Figures 7.5a and 7.5b, respectively. In the case of Mars, the geologic features of interest are channels that clearly were cut by water, but today are dry along with the rest of the planet. Are the features young or ancient? Was the climate wet up through recent times, such that life might have evolved to an advanced stage?
Examination of theVikingimages such as that in figure 7.5a reveals that the Martian channels typically are overlain by impact craters, some fairly substantial in size. Other regions of the Mar- tian surface have far fewer craters, and hence we can say that the channels are, relatively speaking, ancient. Determining a more exact age requires tying the cratering rate to some abso- lute timescale. At the same time, we know that the channels are not among the oldest features, either, because many cut through craters that must therefore be older than the channels. A chronol- ogy can be assembled in which channel formation occurs after formation of the oldest Martian terrains but before a number of other geologic events that are recorded in the surface.
Ganymede, the third of Jupiter’s four giant moons (Io is closest to Jupiter, and then Europa, Ganymede, and Callisto), shows lines in its spectrum typical of water ice. However, the mass of the planet is too heavy given its volume (mass over volume is density) to be pure water ice. The best guess, based on models of solar system formation, is that the heavier component is a silicate, or common rocky material containing silicon, oxygen,
(a) (b)
(d) (c)
(e) (f)
Figure 7.3Varieties of impact craters: (a) large multiring basin, Mare Oriental, on the Moon; (b) classic large crater, Copernicus, with central peak, on the Moon; (c) smaller lunar craters without peaks; (d) pedestal crater on Mars, formed by melting of ground ice during impact. (e) relaxed craters, orpalimpsestson Ganymede (Voyagerimage); (f) eroded crater on Earth, now comprising Lake Manicouagan, Quebec, Canada (see color version in plates section). Photos (a) through (f) are courtesy of NASA.
64
(a)
(b)
Figure 7.4Two very different terrains on the Moon: (a) the lunar highlands show craters of all sizes filling all available surface space; (b) the lunar mare regions are smooth, dark plains with a few fresh craters and remains of large craters, in various states of preservation, which were present when lavas flooded the lunar surface.
(a) (b)
Figure 7.5(a) NASA/Vikingimage of Martian surface cut by channels; (b) NASA/Galileoimage of dark and light terrains on Jupiter’s moon, Ganymede.
magnesium, some iron, and other elements. Therefore, unlike our own Moon, Earth, Venus, Mercury, and Mars, which are made up mostly of silicon-bearing rock and metal, Ganymede is half-rock, half-ice. This is true for Callisto, as well, but not Europa and Io: they are both mostly rock, although Europa has an outer veneer of ice and possibly liquid water.
One might expect, from terrestrial experience, that the ice might behave differently in an impact than rock. In fact, the pressures and temperatures in the hypervelocity impacts we have been describing are so large that there is little difference. Fur- thermore, temperatures in the distant outer solar system, where Jupiter and its moons reside, are so low that ice behaves much like rock as a material making up the solidcrusts, or outer lay- ers, of Ganymede and Callisto. The surface temperature near the equator of these moons is typically 165 K, very far below the ice melting point of 273 K.
Images of Ganymede reveal two types of surfaces: a dark, heavily cratered terrain, and a bright, lightly cratered terrain.
The paucity of craters on the latter surface immediately suggests that it is a younger feature, perhaps ice that has been extruded from the interior along cracks and flowed outward. In places, it is possible to see where a crater on the older dark terrain has been partially obliterated by the new material. It is also possible to tell something about how the cracks and new material formed by looking at distortions in partially preserved craters along the edges of the bright terrain. The difference in brightness between the dark and light terrains remains a matter of speculation; sili- cates and perhaps some carbon-bearing materials are well mixed with the water ice, perhaps dating back to the original formation of Ganymede.
The ability to learn something about the sequence of events on a surface by looking at crater densities is a tool of primary importance in solar system studies. It is a new development of a much older technique applied to Earth geology to look at superpositionof layers to assemble a history of a given region.
On Earth, water and geologic activity have effectively erased the cratering record, so that the use of craters as a geologic tool was a novel idea that did not come into its own until planetary exploration began some three decades ago.
7.2.2 Absolute chronology of solar system events Relative age dating is limited in the amount of information derived. Ideally, one wants to assign ages to events on the sur- faces of planets and moons so as to understand their history and ultimately that of the solar system. Imagine how limited our own understanding of the history of human cultures would be if we only knew the order of events, but not their antiquity or duration.
In the case of Earth’sgeologichistory, even before radioiso- topic dating provided reliable dates, estimates of ages could be made on the basis of notions of the accumulation rate ofsedi- ments, debris brought from high to low places by the action of water. Early work tended to overestimate the rates of sedimenta- tion and hence produced a compressed timescale relative to what is accepted today based on radioisotopic determinations. With the help of radioisotopic dating, the rates of geologic processes
are now better understood and calibrated, such that indirect dating techniques such as sedimentation are enhanced as tools in assembling the history of Earth.
The situation for an absolute chronology from planetary cra- tering is similar in that radioisotopic dating has been used to construct a chronology for Earth’s Moon, which then has been applied, with caveats, to other solar system bodies. The Moon is the only body for which radioisotopic dating of terrains of vary- ing crater density can be performed; Earth’s cratering record is too sparse. (It is not possible to determine unequivocally from what part of the Martian surface the SNC meteorites were derived; hence they cannot help calibrate the cratering record on Mars.)
The oldest parts of the Moon, the highlands, have by far the largest number of craters; the younger mare possess the least.
This is consistent with the decreasing population over time of debris in orbits around the Sun. Theories of planet formation, which we discuss in Chapter 10, hold that the planets were assembled from smaller pieces of rock and ice through relatively low-speed collisions that allowed the pieces to stick together. In the final phases of this process, most of this protoplanetary material was perturbed by close encounters with the planets into highly elliptical orbits, guaranteeing that any subsequent collisions with the planets would be at high speeds, producing craters. Over time this remnant debris of planet formation was swept up by the planets, so that the available impactor population has decreased dramatically from the beginning to the present day.
A simple law governing the rate of impacts over time, con- sistent with the sweep-up picture described above, and with the lunar cratering record, has the inverse exponential form shown in Figure 7.6. The curve is characterized by a very steep decrease initially, as large amounts of material are swept up by the nearly fully grown planets, followed by a transition to a slowly decreasing rate of impacts. The cratering record on the Moon tells us when the transition occurs between these two regimes. Further, it provides information about the tail-off in impacts at later times, though with limited capability because of the paucity of new craters. More difficult to discern is the precise steepness of the early curve, because the cratering rate was so high that lunar highland surfaces are completely cov- ered with craters: new impacts simply obliterate all or part of old ones and only a lower limit on the ancient cratering rate is accessible.
The dating of Moon rocks fixes the transition in the cratering curve at roughly 3.8 billion to 4.0 billion years before present;
the period of intense cratering before that is called theLate Heavy Bombardment, referring to the tail end of the planet- formation (accretion) process. Interestingly, the oldest whole rock samples on Earth date back to roughly the same time. We know that this does not represent the age of Earth because the rocks are rather evolved, showing the action of liquid water on their chemistry and texture; additionally, meteorites record much earlier dates back to 4.56 billion years before present.
Instead, Earth was simply too active geologically at earlier times to preserve older rocks and, as we see in later chapters, had little or no continental land mass on which such rocks could be preserved.
500
400
300
Crater rate (arbitrary units) 200
100
0 4 3
Billion years before present
2 1 0
Copernicus lava flow
Oceanus Procellarum lava flow
Mare Tranquillitatis Imbrium
Basin decaying cratering component Fra
Mauro
steady cratering component
Age of moon
Figure 7.6Number of impacts versus time on the surface of the Moon; the curve is labeled with ages of rocks collected at the Apollolanding sites, and an estimate for the age of the large crater Copernicus. The dashed line shows what the cratering rate would look like if the recent cratering rate were extrapolated along a straight line back to the beginning. After Lang and Whitney (1991).
An additional piece of information on the bombardment his- tory of the Moon, one crucial for calibrating the impact history of other solar system bodies, is the distribution of crater sizes.
Crater sizes are related fairly directly to those of the original
impactors, and hence a model of crater formation can yield the original impactor size distribution.
A typical crater population will exhibit the evolution sketched in Figure 7.7. On a plot showing the number of craters below a given size, versus size, the data tend to fall on a broken line with two different slopes. The steeper slope, occurring for larger crater diameters, directly reflects the size distribution of the orig- inal impacting population that produced the craters. At smaller crater sizes, saturation effects tend to reduce the number of observed craters: Smaller craters are so numerous that they read- ily fill a surface until new craters simply obliterate the old ones.
Over time, as shown in the figure, the breakpoint at which sat- uration takes over moves to larger crater sizes as the surface is increasingly filled. Determination of the breakpoint on such a plot for a cratered region of a planetary surface provides a measure of its age when correlated against surfaces that are absolutely dated, such as those of the Moon.
Mercury and Mars show heavily cratered terrains with dis- tributions similar to those on the Moon. Not only does this allow us to determine how ancient the various terrains are, it also leads us to conclude that the impactor populations on the Moon, Mars, and Mercury are similar. This strongly suggests that the population of impactors that have struck the Moon over time originate from beyond Earth orbit, and in fact are in orbits around the Sun that take them well beyond Mars into the outer solar system. Some of these impactors may have been icy bod- ies derived from reservoirs of debris beyond the orbit of Jupiter, and left over from planetary formation. Additional impact debris likely is derived from the asteroid belt between Mars and Jupiter.
The crater distributions on the moons of the giant planets are similar neither to those of the Moon, Mars, and Mercury nor to each other. Each giant planet seems to have defined a unique population of impactors for its moons, with the only gen- eral resemblance that the population has decreased sharply with time. The cratering size-frequency distribution for the moons of
early
equilibrium diameter
Diameter (logarithmic scale)
Cumulative number of craters (logarithmic scale)
crater production
equilibrium observed
later
larger equilibrium diameter
Diameter (logarithmic scale)
Cumulative number of craters (logarithmic scale)
crater production
equilibrium observed
Figure 7.7How a population of craters evolves over time. Shown in each graph is the number of craters with a diameter less thanD, as a function ofD. The thinner lines are guides to two idealized populations of craters. The steeper line, “production,” is what is produced directly from a particular size distribution of the impacting bodies. The shallower “equilibrium” line is the result of saturation, i.e., obliteration of craters by newer impacts in a very crowded crater field. The right-hand panel represents the situation at a time later than the left-hand panel, showing that, as a surface gets older, the effects of saturation extend to larger and larger crater diameters. Redrawn from Melosh (1989, p. 192).
1
10–2 10–8
10–6 10–4 10–2 102
1
1 102 104
10 100
Diameter (m) Gusev Crater
(Spirit landing site)
Meridiani Planum (Opportunity landing site)
3 Ga 3.5 Ga
4 Ga
1 Ga 100 Ma 10 Ma
Measured craters Measured craters
Model fit Measured hollows
Model fit
Saturation equilibrium Isochrons
No. Craters/km2
1000 10 100 1 10 100
(a)
Diameter (m)
(b)
Figure 7.8Examples of crater size-frequency distributions for the two sites on Mars where the Mars Exploration Rovers ranged over the surface.
Plotted are the cumulative number of craters per square kilometer versus diameter. (a) At theSpiritsite, a large number of hollows exist which may be eroded or partly buried craters; these are indicated with empty circles. The saturation curve defined in Figure 7.7 is shown as a heavy black line, and model fits for different ages in millions of years (Ma) and billions of years (Ga) are shown as lines of constant age (“isochrones”). At theSpirit site a classic fit for an ancient terrain, in which the smaller craters are saturated while the larger ones provide an age, is obtained (dashed line).
(b) For theOpportunitysite, which is suspected to have been more eroded by the action of water, the fit is poorer, presumably because the crater are not uniformly preserved and many are unrecognizable or obliterated. From Smithet al. (2008).
the giant planets probably can be understood best by invoking two populations of impactors: those in solar orbit, perhaps the same as those that had peppered the inner solar system with craters, and a unique population of debris orbiting each of the giant planets. This local debris, likely the leftovers from the for- mation of the giant planets and their moons, has had a different history for each of Jupiter, Saturn, Uranus, and Neptune. In the case of Saturn, there is even evidence in the crater record for the break up of a large moon late in solar system history, per- haps in the orbit now occupied by the irregularly shaped moon Hyperion.
The cratering record in the outer solar system is a use- ful tool for assembling a rough history of this region but, as yet, it is too remote to retrieve samples of icy moons for radioisotopic dating. From studies of craters we now know that Callisto’s heavily cratered surface probably dates back close to the beginning of the solar system; the heavily cratered ter- rain on Ganymede might be slightly younger but certainly pre- dates the bright lightly cratered regions on that moon. Saturn’s small moon, Enceladus, exhibits smooth regions bordered by areas where craters have been partly covered by bright flows;
clearly this moon has been active enough that fresh water ice and other compounds poured out onto the surface recently. Uranus’
satellites differ in their crater density and hence bespeak vary- ing levels of geologic activity the nature of which is otherwise poorly understood. Finally, Jupiter’s Io is devoid of craters and the same is nearly true of Jupiter’s Europa and Neptune’s Tri- ton, satellites that possess a history whose levels of geologic
activity rival or exceed that of our Earth in erasing the cratering record.