by Carl Sagan
Like organisms, machines also have their evolutions. The rocket began, like the gunpowder that first powered it, in China where it was used for ceremonial and aesthetic purposes. Imported to Europe around the fourteenth century, it was applied to warfare, discussed in the late nineteenth century as a means of transportation to the planets by the Russian schoolteacher Konstantin Tsiolkovsky, and first developed seriously for high altitude flight by the American scientist Robert Goddard. The German V-2 military rocket of World War II employed virtually all of Goddard’s innovations and culminated in 1948 in the two-stage launching of the V-2/WAC Corporal combination to the then-unprecedented altitude of 400 kilometers. In the 1950’s, engineering advances organized by Sergei Korolov in the Soviet Union and Wernher von Braun in the United States, funded as delivery systems for weapons of mass destruction, led to the first artificial satellites. The pace of progress has continued to be brisk: manned orbital flight; humans orbiting, then landing on the moon; and unmanned spacecraft outward bound throughout the solar system. Many other nations have now launched spacecraft, including Britain, France, Canada, Japan and China, the society that invented the rocket in the first place.
Among the early applications of the space rocket, as Tsiolkovsky and Goddard (who as a young man had read Wells and had been stimulated by the lectures of Percival Lowell) delighted in imagining, were an orbiting scientific station to monitor the Earth from a great height and a probe to search for life on Mars. Both these dreams have now been fulfilled.
Imagine yourself a visitor from some other and quite alien planet, approaching Earth with no preconceptions. Your view of the planet improves as you come closer and more and more fine detail stands out. Is the planet inhabited? At what point can you decide? If there are intelligent beings, perhaps they have created engineering structures that have high-contrast components on a scale of a few kilometers, structures detectable when our optical systems and distance from the Earth provide kilometer resolution. Yet at this level of detail, the Earth seems utterly barren. There is no sign of life, intelligent or otherwise, in places we call Washington, New York, Boston, Moscow, London, Paris, Berlin, Tokyo and Peking. If there are intelligent beings on Earth, they have not much modified the landscape into regular geometrical patterns at kilometer resolution.
But when we improve the resolution tenfold, when we begin to see detail as small as a hundred meters across, the situation changes. Many places on Earth seem suddenly to crystallize out, revealing an intricate pattern of squares and rectangles, straight lines and circles. These are, in fact, the engineering artifacts of intelligent beings: roads, highways, canals, farmland, city streets—a pattern disclosing the twin human passions for Euclidean geometry and territoriality. On this scale, intelligent life can be discerned in Boston and Washington and New York. And at ten-meter resolution, the degree to which the landscape has been reworked first really becomes evident. Humans have been very busy. These photos have been taken in daylight. But at twilight or during the night, other things are visible: oil-well fires in Libya and the Persian Gulf; deepwater illumination by the Japanese squid fishing fleet; the bright lights of large cities. And if, in daylight, we improve our resolution so we can make out things that are a meter across, then we begin to detect for the first time individual organisms—whales, cows, flamingos, people.
Intelligent life on Earth first reveals itself through the geometric regularity of its constructions. If Lowell’s canal network really existed, the conclusion that intelligent beings inhabit Mars might be similarly compelling. For life to be detected on Mars photographically, even from Mars orbit, it must likewise have accomplished a major reworking of the surface. Technical civilizations, canal builders, might be easy to detect. But except for one or two enigmatic features, nothing of the sort is apparent in the exquisite profusion of Martian surface detail uncovered by unmanned spacecraft. However, there are many other possibilities, ranging from large plants and animals to microorganisms, to extinct forms, to a planet that is now and was always lifeless. Because Mars is farther from the Sun than is the Earth, its temperatures are considerably lower. Its air is thin, containing mostly carbon dioxide but also some molecular nitrogen and argon and very small quantities of water vapor, oxygen and ozone. Open bodies of liquid water are impossible today because the atmospheric pressure on Mars is too low to keep even cold water from rapidly boiling. There may be minute quantities of liquid water in pores and capillaries in the soil. The amount of oxygen is far too little for a human being to breathe. The ozone abundance is so small that germicidal ultraviolet radiation from the Sun strikes the Martian surface unimpeded. Could any organism survive in such an environment?
To test this question, many years ago my colleagues and I prepared chambers that simulated the Martian environment as it was then known, inoculated them with terrestrial microorganisms and waited to see if anybody survived. Such chambers are called, of course, Mars Jars. The Mars Jars cycled the temperatures within a typical Martian range from a little above the freezing point around noon to about – 80°C just before dawn, in an anoxic atmosphere composed chiefly of CO2 and N2. Ultraviolet lamps reproduced the fierce solar flux. No liquid water was present except for very thin films wetting individual sand grains. Some microbes froze to death after the first night and were never heard from again. Others gasped and perished from lack of oxygen. Some died of thirst, and some were fried by the ultraviolet light. But there were always a fair number of varieties of terrestrial microbes that did not need oxygen; that temporarily closed up shop when the temperatures dropped too low; that hid from the ultraviolet light under pebbles or thin layers of sand. In other experiments, when small quantities of liquid water were present, the microbes actually grew. If terrestrial microbes can survive the Martian environment, how much better Martian microbes, if they exist, must do on Mars. But first we must get there.
The Soviet Union maintains an active program of unmanned planetary exploration. Every year or two the relative positions of the planets and the physics of Kepler and Newton permit the launch of a spacecraft to Mars or Venus with a minimum expenditure of energy. Since the early 1960’s the U.S.S.R. has missed few such opportunities. Soviet persistence and engineering skills have eventually paid off handsomely. Five Soviet spacecraft—Veneras 8 through 12—have landed on Venus and successfully returned data from the surface, no insignificant feat in so hot, dense and corrosive a planetary atmosphere. Yet despite many attempts, the Soviet Union has never landed successfully on Mars—a place that, at least at first sight, seems more hospitable, with chilly temperatures, a much thinner atmosphere and more benign gases; with polar ice caps, clear pink skies, great sand dunes, ancient river beds, a vast rift valley, the largest volcanic construct, so far as we know, in the solar system, and balmy equatorial summer afternoons. It is a far more Earth-like world than Venus.
In 1971, the Soviet Mars 3 spacecraft entered the Martian atmosphere. According to the information automatically radioed back, it successfully deployed its landing systems during entry, correctly oriented its ablation shield downward, properly unfurled its great parachute and fired its retro-rockets near the end of its descent path. According to the data returned by Mars 3, it should have landed successfully on the red planet. But after landing, the spacecraft returned a twenty-second fragment of a featureless television picture to Earth and then mysteriously failed. In 1973, a quite similar sequence of events occurred with the Mars 6 lander, in that case the failure occurring within one second of touchdown. What went wrong?
The first illustration I ever saw of Mars 3 was on a Soviet postage stamp (denomination, 16 kopecks), which depicted the spacecraft descending through a kind of purple muck. The artist was trying, I think, to illustrate dust and high winds: Mars 3 had entered the Martian atmosphere during an enormous global dust storm. We have evidence from the U.S. Mariner 9 mission that near-surface winds of more than 140 meters per second—faster than half the speed of sound on Mars—arose in that storm. Both our Soviet colleag
ues and we think it likely that these high winds caught the Mars 3 spacecraft with parachute unfurled, so that it landed gently in the vertical direction but with breakneck speed in the horizontal direction. A spacecraft descending on the shrouds of a large parachute is particularly vulnerable to horizontal winds. After landing, Mars 3 may have made a few bounces, hit a boulder or other example of Martian relief, tipped over, lost the radio link with its carrier “bus” and failed.
But why did Mars 3 enter in the midst of a great dust storm? The Mars 3 mission was rigidly organized before launch. Every step it was to perform was loaded into the on-board computer before it left Earth. There was no opportunity to change the computer program, even as the extent of the great 1971 dust storm became clear. In the jargon of space exploration, the Mars 3 mission was preprogrammed, not adaptive. The failure of Mars 6 is more mysterious. There was no planet-wide storm when this spacecraft entered the Martian atmosphere, and no reason to suspect a local storm, as sometimes happens, at the landing site. Perhaps there was an engineering failure just at the moment of touchdown. Or perhaps there is something particularly dangerous about the Martian surface.
The combination of Soviet successes in landing on Venus and Soviet failures in landing on Mars naturally caused us some concern about the U.S. Viking mission, which had been informally scheduled to set one of its two descent craft gently down on the Martian surface on the Bicentennial of the United States, July 4, 1976. Like its Soviet predecessors, the Viking landing maneuver involved an ablation shield, a parachute and retro-rockets. Because the Martian atmosphere is only 1 percent as dense as the Earth’s, a very large parachute, eighteen meters in diameter, was deployed to slow the spacecraft as it entered the thin air of Mars. The atmosphere is so thin that if Viking had landed at a high elevation there would not have been enough atmosphere to brake the descent adequately: it would have crashed. One requirement, therefore, was for a landing site in a low-lying region. From Mariner 9 results and ground-based radar studies, we knew many such areas.
To avoid the probable fate of Mars 3, we wanted Viking to land in a place and time at which the winds were low. Winds that would make the lander crash were probably strong enough to lift dust off the surface. If we could check that the candidate landing site was not covered with sifting, drifting dust, we would have at least a fair chance of guaranteeing that the winds were not intolerably high. This was one reason that each Viking lander was carried into Mars orbit with its orbiter, and descent delayed until the orbiter surveyed the landing site. We had discovered with Mariner 9 that characteristic changes in the bright and dark patterns on the Martian surface occur during times of high winds. We certainly would not have certified a Viking landing site as safe if orbital photographs had shown such shifting patterns. But our guarantees could not be 100 percent reliable. For example, we could imagine a landing site at which the winds were so strong that all mobile dust had already been blown away. We would then have had no indication of the high winds that might have been there. Detailed weather predictions for Mars were, of course, much less reliable than for Earth. (Indeed one of the many objectives of the Viking mission was to improve our understanding of the weather on both planets.)
Because of communication and temperature constraints, Viking could not land at high Martian latitudes. Farther poleward than about 45 or 50 degrees in both hemispheres, either the time of useful communication of the spacecraft with the Earth or the period during which the spacecraft would avoid dangerously low temperatures would have been awkwardly short.
We did not wish to land in too rough a place. The spacecraft might have tipped over and crashed, or at the least its mechanical arm, intended to acquire Martian soil samples, might have become wedged or been left waving helplessly a meter too high above the surface. Likewise, we did not want to land in places that were too soft. If the spacecraft’s three landing pods had sunk deeply into a loosely packed soil, various undesirable consequences would have followed, including immobilization of the sample arm. But we did not want to land in a place that was too hard either—had we landed in a vitreous lava field, for example, with no powdery surface material, the mechanical arm would have been unable to acquire the samples vital to the projected chemistry and biology experiments.
The best photographs then available of Mars—from the Mariner 9 orbiter—showed features no smaller than 90 meters (100 yards) across. The Viking orbiter pictures improved this figure only slightly. Boulders one meter (three feet) in size were entirely invisible in such photographs, and could have had disastrous consequences for the Viking lander. Likewise, a deep, soft powder might have been indetectable photographically. Fortunately, there was a technique that enabled us to determine the roughness or softness of a candidate landing site: radar. A very rough place would scatter radar from Earth off to the sides of the beam and therefore appear poorly reflective, or radar-dark. A very soft place would also appear poorly reflective because of the many interstices between individual sand grains. While we were unable to distinguish between rough places and soft places, we did not need to make such distinctions for landing-site selection. Both, we knew, were dangerous. Preliminary radar surveys suggested that as much as a quarter to a third of the surface area of Mars might be radar-dark, and therefore dangerous for Viking. But not all of Mars can be viewed by Earth-based radar—only a swath between about 25° N and about 25° S. The Viking orbiter carried no radar system of its own to map the surface.
There were many constraints—perhaps, we feared, too many. Our landing sites had to be not too high, too windy, too hard, too soft, too rough or too close to the pole. It was remarkable that there were any places at all on Mars that simultaneously satisfied all our safety criteria. But it was also clear that our search for safe harbors had led us to landing sites that were, by and large, dull.
When each of the two Viking orbiter-lander combinations was inserted into Martian orbit, it was unalterably committed to landing at a certain latitude on Mars. If the low point in the orbit was at 21° Martian north latitude, the lander would touch down at 21° N, although, by waiting for the planet to turn beneath it, it could land at any longitude whatever. Thus the Viking science teams selected candidate latitudes for which there was more than one promising site. Viking 1 was targeted for 21° N. The prime site was in a region called Chryse (Greek for “the land of gold”), near the confluence of four sinuous channels thought to have been carved in previous epochs of Martian history by running water. The Chryse site seemed to satisfy all safety criteria. But the radar observations had been made nearby, not in the Chryse landing site itself. Radar observations of Chryse were made for the first time—because of the geometry of Earth and Mars—only a few weeks before the nominal landing date.
The candidate landing latitude for Viking 2 was 44° N; the prime site, a locale called Cydonia, chosen because, according to some theoretical arguments, there was a significant chance of small quantities of liquid water there, at least at some time during the Martian year. Since the Viking biology experiments were strongly oriented toward organisms that are comfortable in liquid water, some scientists held that the chance of Viking finding life would be substantially improved in Cydonia. On the other hand, it was argued that, on so windy a planet as Mars, microorganisms should be everywhere if they are anywhere. There seemed to be merit to both positions, and it was difficult to decide between them. What was quite clear, however, was that 44° N was completely inaccessible to radar site-certification; we had to accept a significant risk of failure with Viking 2 if it was committed to high northern latitudes. It was sometimes argued that if Viking 1 was down and working well we could afford to accept a greater risk with Viking 2. I found myself making very conservative recommendations on the fate of a billion-dollar mission. I could imagine, for example, a key instrument failure in Chryse just after an unfortunate crash landing in Cydonia. To improve the Viking options, additional landing sites, geologically very different from Chryse and Cydonia, were selected in the radar-certified r
egion near 4° S latitude. A decision on whether Viking 2 would set down at high or at low latitude was not made until virtually the last minute, when a place with the hopeful name of Utopia, at the same latitude as Cydonia, was chosen.
For Viking 1, the original landing site seemed, after we examined orbiter photographs and late-breaking Earth-based radar data, unacceptably risky. For a while I worried that Viking 1 had been condemned, like the legendary Flying Dutchman, to wander the skies of Mars forever, never to find safe haven. Eventually we found a suitable spot, still in Chryse but far from the confluence of the four ancient channels. The delay prevented us from setting down on July 4, 1976, but it was generally agreed that a crash landing on that date would have been an unsatisfactory two hundredth birthday present for the United States. We deboosted from orbit and entered the Martian atmosphere sixteen days later.
After an interplanetary voyage of a year and a half, covering a hundred million kilometers the long way round the Sun, each orbiter/lander combination was inserted into its proper orbit about Mars; the orbiters surveyed candidate landing sites; the landers entered the Martian atmosphere on radio command and correctly oriented ablation shields, deployed parachutes, divested coverings, and fired retro-rockets. In Chryse and Utopia, for the first time in human history, spacecraft had touched down, gently and safely, on the red planet. These triumphant landings were due in considerable part to the great skill invested in their design, fabrication and testing, and to the abilities of the spacecraft controllers. But for so dangerous and mysterious a planet as Mars, there was also at least an element of luck.