Pale Blue Dot: A Vision of the Human Future in Space - Carl Sagan, Ann Druyan (1997)


Midway between Thera and Therasia, fires broke forth from the sea and
continued for four days, so that the whole sea boiled and blazed, and the
fires cast up an island which was gradually elevated as though by levers …
After the cessation of the eruption, the Rhodians, at the time
of their maritime supremacy, were first to venture upon
the scene and to erect on the island a temple.


All over the Earth, you can find a kind of mountain with one striking and unusual feature. Any child can recognize it: The top seems sheared or squared off. If you climb to the summit or fly over it, you discover that the mountain has a hole or crater at its peak. In some mountains of this sort, the craters are small; in others, they are almost as big as the mountain itself. Occasionally, the craters are filled with water. Sometimes they’re filled with a more amazing liquid: You tiptoe to the edge, and see vast, glowing lakes of yellow-red liquid and fountains of fire. These holes in the tops of mountains are called calderas, after the word “caldron,” and the mountains on which they sit are known, of course, as volcanos—after Vulcan, the Roman god of fire. There are perhaps 600 active volcanos discovered on Earth. Some, beneath the oceans, are yet to be found.

A typical volcanic mountain looks safe enough. Natural vegetation runs up its sides. Terraced fields decorate its flanks. Hamlets and shrines nestle at its base. And yet, without warning, after centuries of lassitude, the mountain may explode. Barrages of boulders, torrents of ash drop out of the sky. Rivers of molten rock come pouring down its sides. All over the Earth people imagined that an active volcano was an imprisoned giant or demon struggling to get out.

The eruptions of Mt. St. Helens and Mt. Pinatubo are recent reminders, but examples can be found throughout history. In 1902 a hot, glowing volcanic cloud swept down the slopes of Mt. Pelée and killed 35,000 people in the city of St. Pierre on the Caribbean island of Martinique. Massive mudflows from the eruption of the Nevado del Ruiz volcano in 1985 killed more than 25,000 Colombians. The eruption of Mt. Vesuvius in the first century buried in ash the hapless inhabitants of Pompeii and Herculaneum and killed the intrepid naturalist Pliny the Elder as he made his way up the side of the volcano, intent on arriving at a better understanding of its workings. (Pliny was hardly the last: Fifteen volcanologists have been killed in sundry volcanic eruptions between 1979 and 1993.) The Mediterranean island of Santorin (also called Thera) is in reality the only part above water of the rim of a volcano now inundated by the sea.* The explosion of the Santorin volcano in 1623 B.C. may, some historians think, have helped destroy the great Minoan civilization on the nearby island of Crete and changed the balance of power in early classical civilization. This disaster may be the origin of the Atlantis legend as related by Plato, in which a civilization was destroyed “in a single day and night of misfortune.” It must have been easy back then to think that a god was angry.

Volcanos have naturally been regarded with fear and awe. When medieval Christians viewed the eruption of Mt. Hekla in Iceland and saw churning fragments of soft lava suspended over the summit, they imagined they were seeing the souls of the damned awaiting entrance to Hell. “Fearful howlings, weeping and gnashing of teeth,” “melancholy cries and loud wailings” were dutifully reported. The glowing red lakes and sulfurous gases within the Hekla caldera were thought to be a real glimpse into the underworld and a confirmation of folk beliefs in Hell (and, by symmetry, in its partner, Heaven).

A volcano is, in fact, an aperture to an underground realm much vaster than the thin surface layer that humans inhabit, and far more hostile. The lava that erupts from a volcano is liquid rock—rock raised to its melting point, generally around 1000°C. The lava emerges from a hole in the Earth; as it cools and solidifies, it generates and later remakes the flanks of a volcanic mountain.

The most volcanically active locales on Earth tend to be along ridges on the ocean floor and island arcs—at the junction of two great plates of oceanic crust—either separating from each other, or one slipping under the other. On the seafloor there are long zones of volcanic eruptions—accompanied by swarms of earthquakes and plumes of abyssal smoke and hot water—that we are just beginning to observe with robot and manned submersible vehicles.

Eruptions of lava must mean that the Earth’s interior is extremely hot. Indeed, seismic evidence shows that, only a few hundred kilometers beneath the surface, nearly the entire body of the Earth is at least slightly molten. The interior of the Earth is hot, in part, because radioactive elements there, such as uranium, give off heat as they decay; and in part because the Earth retains some of the original heat released in its formation, when many small worlds fell together by their mutual gravity to make the Earth, and when iron drifted down to form our planet’s core.

The molten rock, or magma, rises through fissures in the surrounding heavier solid rocks. We can imagine vast subterranean caverns filled with glowing, red, bubbling, viscous liquids that shoot up toward the surface if a suitable channel is by chance provided. The magma, called lava as it pours out of the summit caldera, does indeed arise from the underworld. The souls of the damned have so far eluded detection.

Once the volcano is fully built from successive outpourings, and the lava is no longer spewing up into the caldera, then it becomes just like any other mountain—slowly eroding because of rainfall and windblown debris and, eventually, the movement of continental plates across the Earth’s surface. “How many years can a mountain exist before it is washed to the sea?” asked Bob Dylan in the ballad “Blowing in the Wind.” The answer depends on which planet we’re talking about. For the Earth, it’s typically about ten million years. So mountains, volcanic and otherwise, must be built on the same timescale; otherwise the Earth would be everywhere smooth as Kansas.*

Volcanic explosions can punch vast quantities of matter—mainly fine droplets of sulfuric acid—into the stratosphere. There, for a year or two, they reflect sunlight back to space and cool the Earth. This happened recently with the Philippine volcano, Mt. Pinatubo, and disastrously in 1815–16 after the eruption of the Indonesian volcano Mt. Tambora, which resulted in the famine-ridden “year without a summer.” A volcanic eruption in Taupo, New Zealand, in the year 177 cooled the climate of the Mediterranean, half a world away, and dropped fine particles onto the Greenland ice cap. The explosion of Mt. Mazama in Oregon (which left the caldera now called Crater Lake) in 4803 B.C. had climatic consequences throughout the northern hemisphere. Studies of volcanic effects on the climate were on the investigative path that eventually led to the discovery of nuclear winter. They provide important tests of our use of computer models to predict future climate change. Volcanic particles injected into the upper air are also an additional cause of thinning of the ozone layer.

So a large volcanic explosion in some unfrequented and obscure part of the world can alter the environment on a global scale. Both in their origins and in their effects, volcanos remind us of how vulnerable we are to minor burps and sneezes in the Earth’s internal metabolism, and how important it is for us to understand how this subterranean heat engine works.

IN THE FINAL STAGES of formation of the Earth—as well as the Moon, Mars, and Venus—impacts by small worlds are thought to have generated global magma oceans. Molten rock flooded the pre-existing topography. Great floods, tidal waves kilometers high, of flowing, red-hot liquid magma welled up from the interior and poured over the surface of the planet, burying everything in their path: mountains, channels, craters, perhaps even the last evidence of much earlier, more clement times. The geological odometer was reset. All accessible records of surface geology begin with the last global magma flood. Before they cool and solidify, oceans of lava may be hundreds or even thousands of kilometers thick. In our time, billions of years later, the surface of such a world may be quiet, inactive, with no hint of current vulcanism. Or there may be—as on Earth—a few small-scale but active reminders of an epoch when the entire surface was flooded with liquid rock.

In the early days of planetary geology, ground-based telescopic observations were all the data we had. A fervent debate had been running for half a century on whether the craters of the Moon were due to impacts or volcanos. A few low mounds with summit calderas were found—almost certainly lunar volcanos. But the big craters—bowl- or pan-shaped and sitting on the flat ground and not the tops of mountains—were a different story. Some geologists saw in them similarities with certain highly eroded volcanos on Earth. Others did not. The best counterargument was that we know there are asteroids and comets that fly past the Moon; they must hit it sometimes; and the collisions must make craters. Over the history of the Moon a large number of such craters should have been punched out. So if the craters we see are not due to impacts, where then are the impact craters? We now know from direct laboratory examination of lunar craters that they are almost entirely of impact origin. But 4 billion years ago this little world, nearly dead today, was bubbling and churning away, driven by primeval vulcanism from sources of internal heat now long gone.

In November 1971, NASA’s Mariner 9 spacecraft arrived at Mars to find the planet completely obscured by a global dust storm. Almost the only features to be seen were four circular spots rising out of the reddish murk. But there was something peculiar about them: They had holes in their tops. As the storm cleared, we were able to see unmistakably that we had been viewing four huge volcanic mountains penetrating through the dust cloud, each capped by a great summit caldera.

After the storm dissipated, the true scale of these volcanos became clear. The largest—appropriately named Olympus Mons, or Mt. Olympus, after the home of the Greek gods—is more than 25 kilometers (roughly 15 miles) high, dwarfing not only the largest volcano on Earth but also the largest mountain of any sort, Mt. Everest, which stands 9 kilometers above the Tibetan plateau. There are some 20 large volcanos on Mars, but none so massive as Olympus Mons, which has a volume about 100 times that of the largest volcano on Earth, Mauna Loa in Hawaii.

By counting the accumulated impact craters (made by small impacting asteroids, and readily distinguished from summit calderas) on the flanks of the volcanos, estimates of their ages can be derived. Some Martian volcanos turn out to be a few billion years old, although none dates back to the very origin of Mars, about 4.5 billion years ago. Some, including Olympus Mons, are comparatively new—perhaps only a few hundred million years old. It is clear that enormous volcanic explosions occurred early in Martian history, perhaps providing an atmosphere much denser than the one Mars holds today. What would the place have looked like if we had visited it then?

Some volcanic flows on Mars (for example, in Cerberus) formed as recently as 200 million years ago. It is, I suppose, even possible—although there is no evidence either way—that Olympus Mons, the largest volcano we know about for certain in the Solar System, will be active again. Volcanologists, a patient sort, would doubtless welcome the event.

In 1990–93 the Magellan spacecraft returned surprising radar data about the landforms of Venus. Cartographers prepared maps of almost the entire planet, with fine detail down to about 100 meters, the goal-line-to-goal-line distance in an American football stadium. More data were radioed home by Magellan than by all other planetary missions combined. Since much of the ocean floor remains unexplored (except perhaps for still-classified data acquired by the U.S. and Soviet navies), we may know more about the surface topography of Venus than about any other planet, Earth included. Much of the geology of Venus is unlike anything seen on Earth or anywhere else. Planetary geologists have given these landforms names, but that doesn’t mean we fully understand how they’re formed.

Because the surface temperature of Venus is almost 470°C (900°F), the rocks there are much closer to their melting points than are those at the surface of the Earth. Rocks begin to soften and flow at much shallower depths on Venus than on Earth. This is very likely the reason that many geological features on Venus seem to be plastic and deformed.

The planet is covered by volcanic plains and highland plateaus. The geological constructs include volcanic cones, probable shield volcanos, and calderas. There are many places where we can see that lava has erupted in vast floods. Some plains features ranging to over 200 kilometers in size are playfully called “ticks” and “arachnoids” (which translates roughly as “spiderlike things”)—because they are circular depressions surrounded by concentric rings, while long, spindly surface cracks extend radially out from the center. Odd, flat “pancake domes”—a geological feature unknown on Earth, but probably a kind of volcano—are perhaps formed by thick, viscous lava slowly flowing uniformly in all directions. There are many examples of more irregular lava flows. Curious ring structures called “coronae” range up to some 2,000 kilometers across. The distinctive lava flows on stifling hot Venus offer up a rich menu of geological mysteries.

The most unexpected and peculiar features are the sinuous channels—with meanders and oxbows, looking just like river valleys on Earth. The longest are longer than the greatest rivers on Earth. But it is much too hot for liquid water on Venus. And we can tell from the absence of small impact craters that the atmosphere has been this thick, driving as great a greenhouse effect, for as long as the present surface has been in existence. (If it had been much thinner, intermediate-sized asteroids would not have burned up on entry into the atmosphere, but would have survived to excavate craters as they impact this planet’s surface.) Lava flowing downhill does make sinuous channels (sometimes under the ground, followed by collapse of the roof of the channel). But even at the temperatures of Venus, the lavas radiate heat, cool, slow, congeal, and stop. The magma freezes solid. Lava channels cannot go even 10 percent of the length of the long Venus channels before they solidify. Some planetary geologists think there must be a special thin, watery, inviscid lava generated on Venus. But this is a speculation supported by no other data, and a confession of our ignorance.

The thick atmosphere moves sluggishly; because it’s so dense, though, it’s very good at lifting and moving fine particles. There are wind streaks on Venus, largely emanating from impact craters, in which the prevailing winds have scoured piles of sand and dust and provided a sort of weather vane imprinted on the surface. Here and there we seem to see fields of sand dunes, and provinces where wind erosion has sculpted volcanic landforms. These aeolian processes take place in slow motion, as if at the bottom of the sea. The winds are feeble at the surface of Venus. It may take only a soft gust to raise a cloud of fine particles, but in that stifling inferno a gust is hard to come by.

There are many impact craters on Venus, but nothing like the number on the Moon or Mars. Craters smaller than a few kilometers across are oddly missing. The reason is understood: Small asteroids and comets are broken up on entry into the dense Venus atmosphere before they can hit the surface. The observed cutoff in crater size corresponds very well to the present density of the atmosphere of Venus. Certain irregular splotches seen on the Magellan images are thought to be the remains of impactors that broke up in the thick air before they could gouge out a crater.

Most of the impact craters are remarkably pristine and well preserved; only a few percent of them have been engulfed by subsequent lava flows. The surface of Venus as revealed by Magellan is very young. There are so few impact craters that everything older than about 500 million years* must have been eradicated—on a planet almost certainly 4.5 billion years old. There is only one plausible erosive agent adequate for what we see: vulcanism. All over the planet craters, mountains, and other geological features have been inundated by seas of lava that once welled up from the inside, flowed far, and froze.

After examining so young a surface covered with congealed magma, you might wonder if there are any active volcanos left. None has been found for certain, but there are a few—for example, one called Maat Mons—that appear to be surrounded by fresh lava and which may indeed still be churning and belching. There is some evidence that the abundance of sulfur compounds in the high atmosphere varies with time, as if volcanos at the surface were episodically injecting these materials into the atmosphere. When the volcanos are quiescent, the sulfur compounds simply fall out of the air. There’s also disputed evidence of lightning playing around the mountaintops of Venus, as sometimes happens on active volcanos on Earth. But we do not know for certain whether there is ongoing vulcanism on Venus. That’s a matter for future missions.

Some scientists believe that until about 500 million years ago the Venus surface was almost entirely devoid of landforms. Streams and oceans of molten rock were relentlessly pouring out of the interior, filling in and covering over any relief that had managed to form. Had you plummeted down through the clouds in that long-ago time, the surface would have been nearly uniform and featureless. At night the landscape would have been hellishly glowing from the red heat of molten lava. In this view, the great internal heat engine of Venus, which supplied copious amounts of magma to the surface until about 500 million years ago, has now turned off. The planetary heat engine has finally run down.

In another provocative theoretical model, this one by the geophysicist Donald Turcotte, Venus has plate tectonics like the Earth’s—but it turns off and on. Right now, he proposes, the plate tectonics are off; “continents” do not move along the surface, do not crash into one another, do not thereby raise mountain ranges, and are not later subducted into the deep interior. After hundreds of millions of years of quiescence, though, plate tectonics always breaks out and surface features are flooded by lava, destroyed by mountain building, subducted, and otherwise obliterated. The last such breakout ended about 500 million years ago, Turcotte suggests, and everything has been quiet since. However, the presence of coronae may signify—on timescales that are geologically in the near future—that massive changes on the surface of Venus are about to break out again.

EVEN MORE UNEXPECTED than the great Martian volcanos or the magma-flooded surface of Venus is what awaited us when the Voyager 1 spacecraft encountered Io, the innermost of the four large Galilean moons of Jupiter, in March 1979. There we found a strange, small, multihued world positively awash in volcanos. As we watched in astonishment, eight active plumes poured gas and fine particles up into the sky. The largest, now called Pelé—after the Hawaiian volcano goddess—projected a fountain of material 250 kilometers into space, higher above the surface of Io than some astronauts have ventured above the Earth. By the time Voyager 2 arrived at Io, four months later, Pelé had turned itself off, although six of the other plumes were still active, at least one new plume had been discovered, and another caldera, named Surt, had changed its color dramatically.

The colors of Io, even though exaggerated in NASA’s color-enhanced images, are like none elsewhere in the Solar System. The currently favored explanation is that the Ionian volcanos are driven not by upwelling molten rock, as on the Earth, the Moon, Venus, and Mars, but by upwelling sulfur dioxide and molten sulfur. The surface is covered with volcanic mountains, volcanic calderas, vents, and lakes of molten sulfur. Various forms and compounds of sulfur have been detected on the surface of Io and in nearby space—the volcanos blow some of the sulfur off Io altogether.* These findings have suggested to some an underground sea of liquid sulfur that issues to the surface at points of weakness, generates a shallow volcanic mound, trickles downhill, and freezes, its final color determined by its temperature on eruption.

On the Moon or Mars, you can find many places that have changed little in a billion years. On Io, in a century, much of the surface should be reflooded, filled in or washed away by new volcanic flows. Maps of Io will then quickly become obsolete, and cartography of Io will have become a growth industry.

All this seems to follow readily enough from the Voyager observations. The rate at which the surface is covered over by current volcanic flows implies massive changes in 50 or 100 years, a prediction that luckily can be tested. The Voyager images of Io can be compared with much poorer images taken by ground-based telescopes 50 years earlier, and by the Hubble Space Telescope 13 years later. The surprising conclusion seems to be that the big surface markings on Io have hardly changed at all. Clearly, we’re missing something.

A VOLCANO in one sense represents the insides of a planet gushing out, a wound that eventually heals itself by cooling, only to be replaced by new stigmata. Different worlds have different insides. The discovery of liquid-sulfur vulcanism on Io was a little like finding that an old acquaintance, when cut, bleeds green. You had no idea such differences were possible. He seemed so ordinary.

We are naturally eager to find additional signs of vulcanism on other worlds. On Europa, the second of the Galilean moons of Jupiter and Io’s neighbor, there are no volcanic mountains at all; but molten ice—liquid water—seems to have gushed to the surface through an enormous number of crisscrossing dark markings before freezing. And further out, among the moons of Saturn, there are signs that liquid water has gushed up from the interior and wiped away impact craters. Still, we have never seen anything that might plausibly be an ice volcano in either the Jupiter or Saturn systems. On Triton, we may have observed nitrogen or methane vulcanism.

The volcanos of other worlds provide a stirring spectacle. They enhance our sense of wonder, our joy in the beauty and diversity of the Cosmos. But these exotic volcanos perform another service as well: They help us to know the volcanos of our own world—and perhaps will help one day even to predict their eruptions. If we cannot understand what’s happening in other circumstances, where the physical parameters are different, how deep can our understanding be of the circumstance of most concern to us? A general theory of vulcanism must cover all cases. When we stumble upon vast volcanic eminences on a geologically quiet Mars; when we discover the surface of Venus wiped clean only yesterday by floods of magma; when we find a world melted not by the heat of radioactive decay, as on Earth, but by gravitational tides exerted by nearby worlds; when we observe sulfur rather than silicate vulcanism; and when we begin to wonder, in the moons of the outer planets, whether we might be viewing water, ammonia, nitrogen, or methane vulcanism—then we are learning what else is possible.

*The eruption of a nearby submarine volcano and the rapid construction of a new island in 197 B.C. are described by Strabo in the epigraph to this chapter.

*Even with its mountains and submarine trenches, our planet is astonishingly smooth. If the Earth were the size of a billiard ball, the largest protuberances would be less than a tenth of a millimeter in size—on the threshold of being too small to see or feel.

*The age of the Venus surface, as determined by Magellan radar imagery, puts an additional nail in the coffin of the thesis of Immanuel Velikovsky—who around 1950 proposed, to surprising media acclaim, that 3,500 years ago Jupiter spat out a giant “comet” which made several grazing collisions with the Earth, causing various events chronicled in the ancient books of many peoples (such as the Sun standing still on Joshua’s command), and then transformed itself into the planet Venus. There are still people who take these notions seriously.

*Io’s volcanos are also the copious source of electrically charged atoms such as oxygen and sulfur that populate a ghostly, doughnut-shaped tube of matter that surrounds Jupiter.