The Cosmic Connection

by Carl Sagan

  Lesser mountains were produced by volcanic events. Hot molten rock, called lava, upwells through tubes in the upper layers of the Earth – tubes of structural weakness through which the underlying pressure is relieved – and produces large surface piles of cooling volcanic slag. The resulting hole in the top of the volcanic mountain – the geologists call it a summit caldera – is the channel through which successive episodes of lava-upwelling occur. In the summit caldera of an active volcano, as, for example, in Hawaii, we can actually see molten lava. These individual volcanic mountains and mountain ranges, which are not really separate entities, are signs of a geologically vigorous and dynamic Earth.

  What about Mars? It is a smaller planet than Earth; its central pressures and temperatures are less; it has a lower average density than Earth. These circumstances combine to suggest that Mars should be geologically less active than Earth, perhaps like the Moon. But even on the Moon, a much smaller object than Mars, with even lower anticipated interior temperatures than Mars, recent signs of volcanic activity have been uncovered by the Apollo missions. We do not even today understand the connection between the size and structure of a planet and the presence of volcanism and mountains, although we do know that there are no significant folded mountain ranges on the Moon.

  Our present ignorance on this subject is exceeded by the ignorance of the early planetary astronomers, less than a century ago, as they peered through small telescopes and tried to guess what distant Mars was like. One of the earliest astronomers to commit himself on the question of mountains on Mars was Percival Lowell. Lowell believed (see Chapter 18) that he had found evidence of an extensive network of straight lines, crisscrossing the Martian surface with remarkable regularity and straightness, and that could only have been produced by a race of intelligent beings on that planet. He believed that these "canals" were truly canals carrying water. We now know that the problem was not so much with his logic as with his observations; none of the Mariner or other recent quantitative observations of Mars have shown any sign of the Lowellian canals.

  In the 1890s Lowell argued that Mars must have no mountains, because mountains would be a severe impediment to the construction of a comprehensive network of canals. But surely a race that could construct a planet-wide network of canals should be able now and again to mow down an awkwardly placed mountain.

  Nevertheless, Lowell was among the very first astronomers to apply an actual observational test to the question of mountains on Mars. He looked beyond the terminator. The terminator is the line – sharp or fuzzy, depending on the absence or presence of a planetary atmosphere – that separates the day from the night side of a planet. The terminator moves around the planet once a day – the local planetary day. But if there are mountains just on the night side of the terminator, the mountains will receive the rays of the setting Sun when their adjacent valleys are in darkness. Galileo first used this technique to discover what he called the mountains of the Moon – although the lunar mountains are mainly enormous pieces of rubble that fell out of the sky in the final phases of the formation of the Moon, rather than mountains of the terrestrial type, produced by a geologically active interior.

  Lowell and his collaborators found cases of bright projections beyond the Martian terminator, illuminated by the rays of the setting Sun. But when they calculated their altitudes – an easy task for anyone grounded in high school geometry – the mountains were found to be many tens of miles high. Such elevations on Mars seemed to him absurd because of his canal argument. Moreover, the next day – the day on Mars is almost exactly the same length as a day on Earth – when the feature was seen again, its position had changed. This behavior is quite uncharacteristic of mountains of whatever origin, and Lowell correctly concluded that he had been seeing dust storms, in which fine particles from the Martian surface had been carried some tens of miles into the Martian atmosphere.

  Such dust storms are also observed when we examine through the telescope the day side of Mars. We sometimes see that the characteristic configuration of bright and dark markings on the planet is temporarily obscured. There is an intrusion of bright-area material into the dark area, followed by a reappearance of the former configuration. These changes were interpreted in Lowell's time as dust storms arising in the bright areas and obscuring the adjacent dark areas. The present interpretation, based on the full range of Mariner 9 close-up observations, confirms this view (see Chapter 19).

  Lowell and his contemporaries called the bright areas "deserts," and this, too, seems to be an appropriate name. The Lowellians concerned themselves with the problem of whether bright areas tended to be higher or lower than dark areas, even though the elevation difference was expected to be extremely small. A dark area seen at the illuminated limb, or edge, of the planet seemed to be a notch or depression. But this could be understood merely in terms of the darkness of the dark area: If it were dark against a dark sky, we would not see it at all. We might gain the mistaken impression of a notch or depression. The prevailing opinion of most astronomers seems to have been that the dark areas were slightly lower than the bright, but the difference was estimated by Lowell as only half a mile or less.

  In 1966, I re-examined this problem with Dr. James Pollack. We used two main arguments. Mars has in its winter hemisphere a large polar cap which, at various times, has been ascribed to frozen water or frozen carbon dioxide. Even at the present time its composition is unsettled; both substances are probably present. As the polar cap retreats in each hemisphere, once each year, there are regions where frost is left behind. Later, when the frost leaves these regions, they are found to be brighter than their surroundings. By analogy with the Earth, we might expect them to be high mountainous regions that remain frosted after the snows of the valleys have melted or evaporated. Indeed, one Martian polar region – the so-called Mountains of Mitchel – was identified as mountainous by this argument alone.

  But why are terrestrial mountains the last places to be frost-free? Because it is colder as we walk uphill, as every mountaineer knows. But why does it get colder as we walk uphill? Do the reasons that make terrestrial mountaintops colder than their bases apply to Mars?

  We concluded that all factors that make it colder while walking uphill on the Earth are inoperative on Mars, mainly because of the very thin Martian atmosphere. But the winds on Mars should be higher at mountaintops than in valleys, as on Earth. This is not a conclusion from analogy, but is based on the appropriate physics. Therefore, we imagined that snows are removed by high winds preferentially from the mountains of Mars, and that the bright areas that retain frost on Mars are, therefore, low.

  Our second line of attack was based on the radar observations of Mars, which began in the middle 1960s. There was one piece of evidence that immediately caught our attention. When the small central part of the radar beam was positioned directly on a dark area of Mars, only a very small fraction of the radar signal was returned to Earth. But when an adjacent bright area, on one side or the other of this dark region, was under the center of the radar beam, the reflection was much stronger. This could be understood if the dark area were either much higher or much lower than the adjacent bright area. From the preliminary radar evidence then available, we concluded that if it had to be one or the other of these two alternatives, the dark areas had to be systematically high on Mars. We concluded that major elevation differences existed on Mars, in some cases as much as ten miles between adjacent bright and dark areas. The large-scale slopes were at most only a few degrees – not a very steep grade – and both the elevation differences and slopes were comparable to those on Earth, although the elevations seemed to be greater than here. The notion that the deserts generally were lowlands seemed consistent with the notion of fine sand and dust being trapped in low valleys, with the tops of mountains – where the winds are higher – being scoured of small, bright, fine particles.

  In the few years following our analysis many more detailed radar studies were done – principally by a group
at the Haystack Observatory of the Massachusetts Institute of Technology, headed by Professor Gordon Pettengill. For the first time it was possible to do direct radar altitude measurements. Instead of using our indirect arguments, the technology had reached a point where it was possible to measure how long it took the radar signal to reach Mars and be returned from it. Those places on Mars from which the radar signal took longest to return were farthest from us, and, therefore, deepest. Those regions from which the radar signal took the least time to return were closer to us, and, therefore, highest. In this way the first topographic maps of selected regions of the Martian surface were constructed. The maximum elevation differences and slopes were just about what we had concluded by much more indirect means.

  But dark areas did not appear to be systematically higher than bright areas. Pettengill and his colleagues found that a bright region of Mars called Tharsis appeared to be very high – perhaps the highest region sampled on the planet. A major Martian bright circular area called Hellas – Greek for "Greece" – indeed turned out to be very low from later nonradar observations. A somewhat similar feature called Elysium, also large and bright and roughly circular, turned out to be high. The darkest big Martian area, Syrtis Major, turned out to be a steep slope.

  Why were Pollack and I only partly right? Because of Occam's Razor, a convenient and frequently used principle in science, but one that is not infallible. Occam's Razor recommends that, when faced with two equally good hypotheses, we choose the simpler. We had assumed that dark areas were either systematically high or systematically low. If that were the case, dark areas would have to be systematically high. But that is not the case; dark areas can be either high or low. Our conclusions only reflected our assumptions.

  But I am very pleased that we were able, through logic and physics, to get the story at least partly right, and to demonstrate that there are enormous elevation differences on Mars, elevations much vaster than Lowell had expected. I find it more difficult, but also much more fun, to get the right answer by indirect reasoning and before all the evidence is in. It's what a theoretician does in science. But the conclusions drawn in this way are obviously more risky than those drawn by direct measurement, and most scientists withhold judgment until there is more direct evidence available. The principal function of such detective work – apart from entertaining the theoretician – is probably to so annoy and enrage the observationalists that they are forced, in a fury of disbelief, to perform the critical measurements.

  17. The Mountains of Mars

  II. Observations From Space

  The epic flight of Mariner 9 to Mars in 1971 produced a new set of definitive and direct measurements concerning the mountains and elevations of Mars. Moderately complete elevation terrain maps of Mars have been developed as a result of the ultraviolet spectrometer, the infrared interferometric spectrometer, and the S-band occultation experiments aboard Mariner 9. But the most striking information on the mountains of Mars came from the television experiment.

  The first pictures that Mariner 9 returned from Mars, obtained even before orbital insertion on November 14, 1971, showed an almost completely featureless planet. The south polar cap could be discerned dimly, but the bright and dark markings, which had been seen and debated for over a century, were nowhere to be found. This was not a failure of the television camera, but rather the result of a spectacular planet-wide dust storm, which had begun in late September and would not significantly subside until early January.

  The earliest pre-orbital pictures and the first few days of orbital pictures showed no significant nonpolar detail – except in the region of Tharsis. Here, there were four dark, somewhat irregular spots to be seen, three of them in an approximate straight line running northeast to southwest; the fourth was isolated away from them and to the west. Since there was otherwise nothing much visible on the planet, I devoted some attention to these spots in the early phases of the mission – so much attention that for a while they were known as "Carl's Marks" by several of my wittier co-investigators. I, in turn, proposed naming them Harpo, Groucho, Chico, and Zeppo, but this was all before their significance was established.

  The isolated spot corresponded in position quite well with the classical Martian feature named Nix Olympica – Greek for the Snows of Olympus, the home of the gods. The other three spots seemed to correspond to no familiar Martian surface features. But Bradford Smith, astronomer at New Mexico State University, pointed out that they corresponded quite well (as did Nix Olympica) to places on Mars that exhibited local afternoon brightening as observed from Earth. In some of Smith's ground-based telescopic photographs, obtained with a blue or violet filter and when there was no dust storm on Mars, these four places appeared as brilliant white spots, even though the contrast between the usual bright and dark areas was very small and the usual markings of Mars were indiscernible (the usual situation when Mars is viewed in blue or violet light rather than in orange or red light). Were we observing some sort of dark clouds in the midst of the dust storm at sites where bright clouds were usually found?

  Another Mariner 9 experimenter, William Hartmann of Science Applications, Inc., Tucson, Arizona, performed a computer contrast-enhancement of the original photographs of the four spots and found some faint indication of circular central regions in at least two of them. Indeed, the Mariner 6 and 7 photographs of Nix Olympica, taken in 1969, showed a similar indication there.

  By this time, the extent and severity of the dust storm had become evident, and part of our preplanned mission for Mariner 9 to map the planet had to be postponed. This then freed a significant picture-taking ability for high-resolution, close-up photographs of the four spots. These experiments of opportunity were possible only because Mariner 9 had a major adaptive capability. The scan platform, on which the cameras were located, could be aimed at many desired spots on Mars, and the technical staff of the Jet Propulsion Laboratory of the California Institute of Technology was able to change its plans quickly enough to accommodate the changed scientific needs of the mission. Because of the design of the spacecraft and the adaptability of its controllers, the first close-up photographs of the four spots began coming in.

  Each spot had a vaguely circular center. There were parallel arcuate segments. There was a kind of scalloping. All these features were dark against a bright surrounding, corresponding to the dark appearance of the spots as seen initially in low resolution.

  The particular shapes that we had seen in the early pictures held no particular significance for me. But I was struck by the fact that these circular features occurred in Tharsis, the highest region on Mars. These features were craters. Why were we seeing them and virtually no other Martian features? Because they must be the highest regions in Tharsis, a region already enormously elevated. The four spots, therefore, seemed to me to be vast mountains poking through the dust. I proposed that as time went on and the dust storm settled (from experience with other Martian global dust storms over decades of observation, we knew the dust storms would have to settle eventually), we would see more and more of these mountains, clear down to their bases. I even thought it possible that we could produce topographic maps from the sequence of emerging detail as the dust settled. Unfortunately, the settling out of the dust was a very irregular affair, and this suggestion has not yet borne fruit.

  Geologist members of the Mariner 9 television team, such as Harold Masursky and John McCauley, of the U. S. Geological Survey, were taken with the form of the craters, and quite early identified them – by analogy with similar features on Earth – as vast volcanic piles with summit calderas. I have always been mistrustful of arguments from terrestrial analogy. After all, Mars is quite another place. For all we knew – at least, for all I knew – quite different geological processes might operate there, and Earth-like features might be produced by different causes.

  However, by another route I reached the same conclusion as the geologists: There are only two processes we know that produce craters – the impact of interp
lanetary debris (the origin, for example, of most of the craters on the Moon) and vulcanism. It would be asking too much to expect that the large meteorites or small asteroids that carved out four of the largest impact craters in Tharsis knew enough to land on the top of the four highest mountains in Tharsis. Much more plausible is the idea that the mechanism that made the mountain made the crater. That mechanism is called vulcanism.

  As the dust storm cleared, the true magnitude of these four volcanic mountains became clear. The largest of them, Nix Olympica, is five hundred miles across, larger than the largest such feature on the Earth, the Hawaiian Islands. The altitudes of the spots have not yet been determined with precision, but they appear to be ten to twenty miles above the mean level of the planet. (We cannot talk of sea level on Mars because there are not – today, at any rate – any seas there.) Over a dozen smaller volcanoes have since been found in other regions of Mars.

  The infrared radiometer on Mariner 9 showed no sign of hot lava in the summit calderas of the craters. On the other hand, their fresh appearance and the almost total absence of meteorite cratering on their slopes show them to be very young objects, geologically speaking – probably no more than a few hundred million years old, possibly younger.

  The association of clouds with these volcanic mountains could be due to contemporary outgassing from the calderas – steam, for example, being exhaled up volcanic vents. But it seems more likely that the clouds are present at the summits of these mountains precisely because these mountains are so high. An imaginary parcel of Martian air, rising along the slope of the mountain, expands and cools. (The air gets colder as we go upward in the Martian atmosphere. But because the air is so thin on Mars, it cannot exchange heat well with the surface; thus the surface does not get cold as we go uphill on Mars, as we discussed earlier.) When the temperature in the parcel of air drops below the freezing point of water, all of the water vapor in the parcel condenses out into ice crystals. The amount of water vapor we know to exist in the Martian atmosphere, the heights of the mountains, and the amount of small ice crystals necessary to produce a visible cloud together work out correctly for this to be the explanation of mountain clouds on Mars.