On August 28, 2003, Mars will make a closer approach to the Earth than at any time in the last several thousand years--it will come within 34,645,500 miles (55,756,600 km).
Broadly speaking, the Martian surface is divided into two main regions. This division has been referred to as the great "crustal dichotomy". South of a circle inclined by roughly 35' to the planet's equator are ancient, heavily cratered highlands; north of this circle are younger, relatively smooth plains and volcanic features. The boundary between the two regions is formed by a gentle, irregular scarp and low, knobby hills. On average, the southern highlands are some 2.1 kilometers higher in elevation than the northern lowlands.
The southern highlands, which include most of the subequatorial parts of the planet as well as a rather wide tongue extending northward beyond Sinus Sabaeus-Sinus Meridiani, are profuse with craters, so much so that the view looks superficially very much like that of the highlands on the Moon-thus,the discouraging results of the flyby Mariners. In the case of the Moon, there was a long and heated debate about the origin of these craters; on one side were those who believed them to be impact features, on the other were those who maintained that they had been formed by internal processes of some kind-broadly speaking, by volcanism. During the 1960s and early 1970s, the question was finally settled decisively in favor of the impact theory, and there can be no doubt that the Martian craters were formed in the same way.
The impact process has now been worked out in considerable detail. When an object - say, a small asteroid - plunges into the surface of a planet, it produces two interacting shock waves. The first shock wave engulfs the asteroid, vaporizes it, and melts rock at the immediate point of impact. This part of the process absorbs a rela- tively small fraction of the energy of impact; the much greater share goes into producing a second shock wave, which travels radially away from the point of impact, excavates the crater, and throws a rim of disintegrated material around it (the ejecta blanket).
Craters of different diameters have different forms. Very small craters are simply bowl-shaped pits that have a fairly constant depth- to-diameter ratio of about 0.20. Larger craters are more complex.
Violent rebound of the floor from the shock of impact gives rise to a central peak or peaks. In addition, many of the larger craters have terraced walls caused by landslip- the slumping in of rim materials toward the center of the crater. This partial filling in with wall material explains why the more complex craters become shallower with increasing diameter.
Even after the flyby Mariners it was obvious that in general the Martian craters are flatter and more subdued than their lunar counterparts, and usually lack the latters' hummocky surrounds. There is also a relative paucity of smaller pits on Mars (less than about 20 km across). These findings are readily explained as being due to the fact that on Mars, unlike the Moon, there has been con- siderable weathering over time by water, air, and (possibly) glacial erosion. There are other important differences between Martian and lunar craters as well. Summit pits on the central peaks of Martian craters are much more common than on the Moon, and though smaller pits have lunarlike ejecta blankets that have been laid down with ballistic trajectories, those with diameters greater than about 5 kilometers tend to have a dfferent pattern of overlapping sheets of ejecta with lobate margins-the latter being telltale signs of formation by flow across the surface. It has been suggested that the 5 -kilometer diameter transition between ballistic and flow patterns may correspond to the minimum depth that needs to be excavated in order to release subsurface water ice.5
Martian impact craters with diameters greater than about 50 to 70 kilometers are referred to as basins. The very largest basins, greater than 300 kilometers or so across, are multiringed features similar to those already known on the Moon, though again, because of erosive forces the condition of the Martian features is much less pristine.
The Hellas basin, which measures 2,300 kilometers from rim to rim, is the most imposing topographical feature of the Martian southern hemisphere. Its floor is the lowest point on Mars, located some 5 kilometers below the Martian datum. (On Mars, of course, there is no true "sea level"; instead, the reference datum is defined as the level at which the partial carbon dioxide pressure is 6.1 millibars, which marks the triple point of water. For partial pressures of water greater than 6.1 millibars, liquid water can exist under some temperature conditionsns; for partial pressures lower than this, it is always unstable.) 6 In the winter Hellas is often covered with carbon dioxide frost, and at such times can appear brilliant white. Other basins clearly visible in the Mariner 9 images are Isidis and Argyre-the latter measures 1,900 kilometers across and still shows its rim mountains, the Nereidum and Charitum Montes. This basin, too, is often covered with frost. Another major mountain range, the Phlegra Montes, is located in Elysium; these mountains have been identified as remnants of another basin rim that was later inundated by volcanoes. In recent years a number of other large basins have been identified, most of them well-nigh obliterated by Mar- tian weathering.
When the frequencies of craters of different sizes are plotted for the highlands of the Moon, Mercury, and Mars, the resulting distri- butions, although not identical, demonstrate the existence of two distinct cratering populations: an older population, made during a period of heavy bombardment in which the great basins and the majority of the craters of the rugged highlands were formed; and a younger population that may be traced in the more recent plains and was formed by post-late heavy bombardment.
When the solar system began to form 4.6 billion years ago, a rotating disk of gas and dust began, grain by grain, to accrete into larger objects known as planetesimals; these in turn accreted into the planets. Some of the planetesimals became rather large in their own right and collided with the planets, with fateful consequences. For instance, a smash-up involving a Mars-sized body and the Earth is believed to have given rise to the Moon. Another collision between a large object and Mars about 4.2 billion years ago may well account for the Martian crustal dichotomy. In the low-lying plains of the northern hemisphere, which are now largely covered with sedimentary debris, a gigantic impact basin has been tentatively identified. The Borealis basin, as it is called, is 7,700 kilometers across and is centered in Vastitas Borealis (50' N, 190' w); it is in- deed vast-altogether it covers some 8o percent of the northern hemisphere plains.7
Obviously space was much more crowded in the early history of the solar system than it is now, and impacts were more frequent. The residue of accretional material subjected the Moon and planets to a massive late bombardment in which the impacts occurred at such high rates that their surfaces became saturated with craters -in other words, the formation of new craters could take place only by obliterating preexisting ones. In the case of the Moon, this violent period came to an end about 3.8 billion years ago. It may have ended at about the same time on Mars, although there is evidence to suggest that it continued until somewhat later. In any case, the objects derived from accretional debris became extinct. Henceforth, craters were added at a much lower rate and were caused by the occasional impacts of asteroids and comets (for Mars, which lies very near the asteroid belt, it has been estimated that asteroids create seven times more craters than comets do). It is these comets and asteroids that are responsible for the post-late heavy bombardment cratering.
Various areas on the Martian surface show widely different cratering densities; the most heavily cratered areas are the oldest. Thus the relative ages of surface units (stratigraphic relationships) can be worked out, and by making certain assumptions about when the late heavy bombardment phase ended and the rates of cratering since, one can even go so far as to estimate their absolute ages (though really reliable values must await the return of actual surface materials from Mars). The oldest units make up the so-called Noachian system, which is represented by the ancient cratered terrain centered on Noachis Terra (approximate ages 4.60 to 3.80 billion years). Overlying the rocks of Noachian age are those of the Hesperian system, whose units are characterized by ridged plains material, of which examples are found in Hesperia Planum and Vastitas Borealis (approximate ages 3.80-3-55 billion years). Finally, the Amazonian system consists of largely smooth plains material such as that which covers Acidalia, Amazonis, and Elysium Planitia (ages less than 3.55 billion years).
Aeologists once believed that Mars was cool early in its history, and that it formed a hot core at a relatively advanced stage, after heating due to decay of radioactive materials warmed it sufficiently to initiate melting of rock. This would have implied a late onset of volcanism. We now know otherwise, and from an unimpeachable source: through direct analysis of material from Mars itself.
This material exists in the form of several unusual meteorites - one fell at Chassigny, France, in 1815; others fell at Shergotty, India, in 1865 and at Nakhla, Egypt, in 1911; and several others have been identified as well. Their Martian origins have, however, been suspected only since 1981. The meteorites are classified into three groups: shergottites, nakhhtes, and chassignites (collectively known as SNCS). All are of the common stony type, but they are very young compared with the 4.5-bilhon-year age of most meteorites. Gases captured in shock-generated glassy nodules within them were analyzed and proved to have the exact composition of the same gases in the Martian atmosphere; they also contain small amounts of water and water-altered minerals. There can be little doubt that they are Martian. They were blasted into space in one or (more probably) several impacts, with such force that they reached escape velocity and eventually reached Earth.
Detailed analyses of the SNC meteorites made it clear that not long after Mars accreted into a world, its interior was already hot and had differentiated into a nickel-iron core, mantle, and crust. Much of the heat at this stage must have come from the energy of the impacts themselves. Thus, Martian volcanism began early. Dur- ing late Noachian and early Hesperian times, melted rock (magma) began to reach the surface. Extensive ridged plains were laid down in areas of the southern hemisphere; the so-called highland paterae also emerged, of which four are located near the Hellas basin and probably were formed in relation to deep-seated fractures produced during the impact that formed it. The best-known example is Patera Tyrrhena (at 23' S, 255' w), which seems to have been a volcano of the explosive type, as its irregular summit caldera (40 km long by 12 km wide) is surrounded by large quantities of ash. The plateau of Syrtis Major Planitia (10' N, longitude 2900 W) is another early volcanic region; its activity seems to have begun during postimpact adjustments of the crust around the Isidis basin. The dark materials that cover it originated from a low-relief volcanic shield.
This early volcanic period was characterized by the rapid escape of heat from the interior to the surface. Inevitably, the planet began to cool; as it did, convection of the mantle decreased, the over- lying crust steadily thickened, and surface volcanism came to be concentrated in ever more limited regions. For some reason not yet entirely understood, the main volcanic activity came to center on two areas: Elysium and Tharsis. These areas marked the locations of "hot spots" or mantle plumes -places where a column of heated material rose from the mantle. Why there should have been only two such plumes on Mars is not known, but the consequences are obvious enough. Lava flooding occurred in these regions on an enormous scale, producing a domical buildup of material which stretched the overlying crust and produced belts of intense fracturing. Along the equator, Vafles Marineris began to open when a series of deep troughs formed, oriented radially to the Tharsis rise; later, these troughs were eroded into spurs and gullies. In Elysium, the first volcano to appear was Hecates Tholus, a shield structure some 180 kilometers across and 6 kilometers high. The center of the eruptions then shifted some 850 kilometers to the south to form the dome-shaped Albor Tholus. Still later eruptions gave rise to Elysium Mons, a shield volcano some 500 kilometers across at its base and standing some 9 kilometers above the surrounding plain; its summit is marked by a 14-kilometer-wide caldera.
By far the greatest activity took place in Tharsis, to the west. The Tharsis bulge stands out like an enormous hump relative to the ellipsoidal shape that the planet would have were it in equilibrium with itself; this hump extends 4,000 kilometers north-south from the plains bordering Mareotis Fossae to Solis Planum, and 3,000 kilometers east-west from Lunae Planum to Amazonis and Arcadia Planitia. Its average level stands some 8-10 kilometers above the Martian datum. In late Hesperian time, the volcanic activity in the region centered on Alba Patera, which lies on the bulge's north flank in an extensively fractured landscape; the caldera itself shows little vertical relief, but measures 1,500 kilometers across. By early Amazonian time, a fault line running on the northwest flank of the Tharsis bulge had become active, giving rise to the three great shield volcanoes of the Tharsis Montes. Several smaller shields and vol- canic domes lie close to the same line - Biblis, Ulysses, and Uranius Paterae, and Uranius, Ceraunius, and Tharsis Tholi. Finally, on the southeastern flank of the Tharsis bulge, Some 1,200 kilometers northeast of the Tharsis Montes, arose Olympus Mons, whose slopes make up what is probably the youngest surface on Mars. The eruptions here continued long after cooling of the planet's interior had extinguished active volcanism elsewhere -the most recent may have occurred as little as 300 million years ago.
Of the many discoveries of Mariner 9, by all odds the most exciting was the recognition of valley networks and outflow channels, which can hardly have formed otherwise than by the action of running water at or near the surface. Their existence has provided the strongest evidence that Mars may have undergone major climatic changes over time.
The valley networks are in general the older features. They have tributaries, so that they look very much like dry riverbeds, and they lie almost entirely (about 98 percent) in the heavily cratered highlands of the southern hemisphere - and so must be as ancient as they. They are typically 1 to 2 kilometers wide, and not very long; even including their tributary systems, networks seldom go on for more than a few hundred kilometers. The longest-Ma'adim Vallis (centered at 20' s, 182' w), Al Quahira (at 18' S, 196' w), and Nir- gal Valis (north of Argyre at 28' S, 40' w) -range in length from 400 to 800 kilometers. At first it was hoped that the valley networks might be proof that precipitation took place on early Mars -rain, in other words -though more recent studies have shown that they give every indication of having been formed by groundwater sap- ping due to melting of an ice-rich permafrost.8
Some valleys formed after the end of the period of heavy bombardment; for example, the system located on the northern flanks of Alba Patera. It is certainly young, and may well be of early Amazonian age. In every way it looks similar to the fluvial valleys produced by runoff on the flanks of the Hawaiian volcanoes, and like the latter is believed to have been formed by surface runoff (though again produced by a sapping process rather than by rainwaters Compared with the comparatively tiny valley networks, the outflow channels are features on a grand scale; typically they measure hundreds of kilometers long and tens of kilometers wide. They generally emerge in areas of the surface that have undergone collapse, such as chaotic terrain or canyons. Several outstanding examples - the Ares, Tiu, and Simud Valles -originate in the chaotic terrain in Aurorae Sinus at the eastern end of Vafles Marineris; Kasei Vallis extends from Echus Chasma, a canyon just west of Lunae Planum; and the Maja, Vedra, and Bahram Valles all arise from Juventae Chasma, on the opposite flank of Lunae Planum. All of these channels then converge on and disappear into the southern floor of Chryse Planitia on the eastern margin of the Tharsis bulge. There are also many outflow channels in Elysium, northwest of the volcanic province, whence they debouch into the low-lying northern plains; still others are found in Memnonia, Amazonis Planitia, and on the rim of Hellas.
These enormous channels lack tributaries; some, such as Mangala Vallis in the upland region of Memnonia on the border of Amazonis Planitia, are characterized by sculpted landforms such as teardrop or lemniscate islands. The closest terrestrial analogy to these features is the Channeled Scabland of eastern Washington, in the United States, which was formed at the end of the last Ice Age when much of western Montana was covered by a glacial meltwater lake (Lake Missoula). The lake was held in check by an ice dam lying across northern Idaho; when the dam suddenly broke, it released the captured water in a flash flood that drained the whole Columbia Plateau as far as the Pacific Ocean over a period of several days.10 There is every reason to believe that the Martian channels were also formed by catastrophic flooding, but on an even more monumental scale. The outflow channels are generally younger than the valley networks, and all were formed after the period of heavy bombardment; the circum-Chryse channels have been dated to Hesperian times, and Mangala Vallis is of Amazonian age. Many of the channels seem to record multiple flooding events over a long period, and apparently they could even form under present conditions; although liquid water is unstable on the surface of Mars because of the low atmospheric pressure (any that formed there would rapidly boil away), this poses no obstacle to massive floods of brief duration. What was the source of all the water? It is probable that early in its history Mars had a much more substantial atmosphere than it does now. Over time, the oxygen present bonded with rock, turning it red, and the water seeped down through the meteorite- fractured regolith. In the upper parts of the regolith the water froze, but farther down the temperature was still warm enough for the water to remain liquid, and seas formed deep within Mars, pooling beneath an ice-rich permafrost layer perhaps several kilometers thick. Volcanism such as that in Tharsis caused extensive melting of the permafrost layer, releasing the water through permeable volcanic rocks, out along the great fracture systems such as Memnonia Fossae and VaUes Marineris, and finally onto the surface in vast catastrophic floods.
THE IMPLICATIONS of the valley networks and outflow channels are still being debated, and I shall have more to say about them in the next chapter, but there can be little doubt that their discovery was the single most important revelation of Matiner 9. To astronomers, geologists, and laypersons alike they suggested the distinct possibility that Mars, in having once allowed running water on its surface, may not always have been so forbiddingly severe as it is now. This in turn gave a tremendous impetus to the next American mission, Viking, whose lofty goal was nothing less than to commence the in situ search for evidence of life on Mars.
The odds of finding living organisms on Mars were obviously very slim, but it seemed that they might not be nil. Such a search was not without its chimerical aspects, but it had been all but inevitable ever since Percival Lowell had sat at the eyepiece of his telescope and, scanning the little globe of orange-yellow spotted by transverse stripes of color, had savored the thought: Here there be life!