Great ships will arrive carrying essential technology from Earth, new families of settlers, scarce resources. It is hard to know, on the basis of our limited knowledge of Mars, whether they will go home empty—or whether they will carry with them something found only on Mars, something considered very valuable on Earth. Initially much of the scientific investigation of samples of the Martian surface will be done on Earth. But in time the scientific study of Mars (and its moons Phobos and Deimos) will be done from Mars.

Eventually—as has happened with virtually every other form of human transportation—interplanetary travel will become accessible to people of ordinary means: to scientists pursuing their own research projects, to settlers fed up with Earth, even to venturesome tourists. And of course there will be explorers.

If the time ever came when it was possible to make the Martian environment much more Earth-like—so protective garments, oxygen masks, and domed farmlands and cities could be dispensed with—the attraction and accessibility of Mars would be increased many-fold. The same, of course, would be true for any other world which could be engineered so that humans could live there without elaborate contrivances to keep the planetary environment out. We would feel much more comfortable in our adopted home if an intact dome or spacesuit weren’t all that stood between us and death. (But perhaps I exaggerate the dangers. People who live in the Netherlands seem at least as well adjusted and carefree as other inhabitants of Northern Europe; vet their dikes are all that stand between them and the sea.

Recognizing the speculative nature of the question and the limitations in our knowledge, is it nevertheless possible to envision terraforming the planets?

We need look no further than our own world to see that humans are now able to alter planetary environments in a profound way. Depletion of the ozone layer, global warming from an increased greenhouse effect, and global cooling from nuclear war are all ways in which present technology can significantly alter the environment of our world—and in each case as an inadvertent consequence of doing something else. If we had intended to alter our planetary environment, we would be fully able to generate still greater change. As our technology becomes more powerful, we will be able to work still more profound changes.

But just as (in parallel parking) it’s easier to get out of a parking place than into one, it’s easier to destroy a planetary environment than to move it into a narrowly prescribed range of temperatures, pressures, compositions, and so on. We already know of a multitude of desolate and uninhabitable worlds, and—with very narrow margins—only one green and clement one. This is a major conclusion from early in the era of spacecraft exploration of the Solar System. In altering the Earth, or any world with an atmosphere, we must be very careful about positive feedbacks, where we nudge an environment a little bit and it takes of on its own—a little cooling leading to runaway glaciation, as may have happened on Mars, or a little warming to a runaway greenhouse effect, as happened on Venus. It is not at all clear that our knowledge is sufficient to this purpose.

As far as I know, the first suggestion in the scientific literature about terraforming the planets was made in a 1961 article I wrote about Venus. I was pretty sure then that Venus had a surface temperature well above the normal boiling point of water, produced by a carbon dioxide/water vapor greenhouse effect. I imagined seeding its high clouds with genetically engineered microorganisms that would take CO2, N2, and H2O out of the atmosphere and convert them into organic molecules. The more CO2 removed, the smaller the greenhouse effect and the cooler the surface. The microbes would be carried down through the atmosphere toward the ground, where they would be fried, so water vapor would be returned to the atmosphere; but the carbon from the CO2 would be converted irreversibly by the high temperatures into graphite or some other involatile form of carbon. Eventually, the temperatures would fall below the boiling point and the surface of Venus would become habitable, dotted with pools and lakes of warm water.

The idea was soon taken up by a number of science fiction authors in the continuing dance between science and science fiction—in which the science stimulates the fiction, and the fiction stimulates a new generation of scientists, a process benefiting both genres. But as the next step in the dance, it is now clear that seeding Venus with special photosynthetic microorganisms will not work. Since 1961 we’ve discovered that the clouds of Venus are a concentrated solution of sulfuric acid, which makes the genetic engineering rather more challenging. But that in itself is not a fatal flaw. (There are microorganisms that live out their lives in concentrated solutions of sulfuric acid.) Here’s the fatal flaw: In 1961 I thought the atmospheric pressure at the surface of Venus N\-as a few “bars,” a few times the surface pressure on Earth. We now know it to be 90 bars, so that if the scheme worked, the result would be a surface buried in hundreds of meters of fine graphite, and an atmosphere made of 65 bars of almost pure molecular oxygen. Whether we would first implode under the atmospheric pressure or spontaneously burst into flames in all that oxygen is an open question. However, long before so much oxygen could build up, the graphite would spontaneously burn back into CO2, short-circuiting the process. At best, such a scheme can carry the terraforming of Venus only partway.

Let’s assume that by the early twenty-second century we have comparatively inexpensive heavy-lift vehicles, so we can carry large payloads to other worlds; abundant and powerful fusion reactors; and well-developed genetic engineering. All three assumptions are likely, given current trends. Could we terraform the planets?[37] James Pollack of NASA’s Ames Research Center and I surveyed this problem. Here’s a summary of what we found:

Venus: Clearly the problem with Venus is its massive greenhouse effect. If we could reduce the greenhouse effect almost to zero, the climate might be balmy. But a 90-bar CO2 atmosphere is oppressively thick. Over every postage stamp-sized square inch of surface, the air weighs as much as six professional football players, piled one on top of another. Making all that go away will take some doing.

Imagine bombarding Venus with asteroids and comets. Each impact would blow away some of the atmosphere. To blow away almost all of it, though, would require using up more big asteroids and comets than there are—at least in the planetary part of the Solar System. Even if that many potential impactors existed, even if we could make them all collide with Venus (this is the overkill approach to the impact hazard problem), think what we would have lost. Who knows what wonders, what practical knowledge they contain? We would also obliterate much of Venus’ gorgeous surface geology—which we’ve just begun to understand, and which may teach us much about the Earth. This is an example of bruteforce terraforming. I suggest we want to steer entirely clear of such methods, even if someday we’ll be able to afford them (which I very much doubt). We want something more elegant, more subtle, more respectful of the environments of other worlds. A microbial approach has some of those virtues, but does not do the trick, as we’ve just seen.

We can imagine pulverizing a dark asteroid and spreading the powder through the upper atmosphere of Venus, or carrying such dust up from the surface. This would be the physical equivalent of nuclear winter or the Cretaceous-Tertiary post-impact climate. If the sunlight reaching the ground is sufficiently attenuated, the surface temperature must fall. But by its very nature, this option plunges Venus into deep gloom, with daytime light levels perhaps only as bright as on a moonlit night on Earth. The oppressive, crushing 90-bar atmosphere would remain untouched. Since the emplaced dust would sediment out every few years, the layer would have to be replenished in the same period of time. Perhaps such an approach would be acceptable for short exploratory missions, but the environment generated seems very stark for a self-sustaining human community on Venus.