Short Stories

This article also appeared in the Fall 1999 issue of Space Front, a publication of the Space Frontier Foundation.

The Case for Space

By Mike Combs

[email protected]

Copyright © 1999

Sunlight Available 24x7
Convenient Access to Zero Gravity
All Gravitational Options Available
Living at the Top of a Well
No Weather, Save What We Make for Ourselves
Warning! Warning! Meteor Storm!
Convenient Communication with the Homeworld
Convenient Travel to the Homeworld
Making A Living
A Staging Post for the Belt?
Room To Grow
Location Options
Spreading Interstellar
Getting What You Need
Robert Zubrin's Views


From the very beginnings of both science fiction and serious scientific speculation, most concepts for future colonization beyond the Earth have targeted the planet Mars. The reason why is easy to see. Of all the planets in our solar system, it's the one most like Earth.

But starting in 1969, Princeton University's Professor Gerard O'Neill began looking in a different direction: toward artificial habitats constructed in orbit from materials already in space. He had started by asking the question, "Is the surface of the Earth really the right place for an expanding technological civilization?" Some study seemed to indicate the answer was "No". Calculations revealed that orbital habitats could be surprisingly large and Earth-like, and would have many advantages over any planetary home.

O'Neill's findings made us realize there was an unspoken and unquestioned assumption underlying the logic that pointed toward Mars: In order to create colonies beyond Earth, we must first find a planet on which to build them.

In the years since the initial enthusiasm following the publication of O'Neill's "The High Frontier", interest in the concept of space colonies seems to have abated, while interest in Mars settlement and ultimate terraformation seems to be at a peak. Viewed from the High Frontier perspective, the construction of orbital habitats instead of Mars settlements would still seem to have better returns on shorter time scales, much greater long-term benefits, and numerous other advantages.

Before we begin, it should be noted that many of the benefits cited for Mars colonization (e.g. not keeping all of our eggs in one planetary basket, stimulation to technical advancement, multiplication of human opportunities and potentials for freedom, etc.) are in fact equally good arguments for either settlements on Mars or in orbit, and thus will be outside the scope of this article.

Also, when comparing the relative merits of orbital space versus the surface of Mars for settlement, such comparisons should only be made between proposals with similar returns. Some have compared Dr. Robert Zubrin's Mars Direct strategy with Gerard O'Neill's High Frontier proposals. Mars Direct should only be compared with a program placing an equivalent number of people in High Earth Orbit, along with an equivalent amount of infrastructure geared toward supporting future settlement plans. Other valid comparisons would be: comparing the construction of the first Island One or Stanford Torus habitat with the construction of the first domed community on Mars large enough to independently support a population of 10,000, and comparing the terraformation of Mars with an orbital settlement program resulting in the creation of habitats sufficient to equal the land area of Mars.

Sunlight Available 24x7

The reason we've not been able to convert our Earthly economy over to solar energy is that it's intermittent here on the surface. The sun is blocked every night, and is filtered by clouds. In a sufficiently high orbit, an O'Neill colony will be in sunlight around 99% of the time. This means solar energy can be relied upon not only for life-support and agriculture, but also for electrical utilities and process heating for industrial operations such as the smelting of ore. On Mars, as on any celestial body, solar energy will be unavailable half the time. We don't tend to see this as a disadvantage because it's the present situation here on Earth, and we're used to it. But locating in orbit opens up the possibility of a vast civilization powered by cheap, clean, plentiful, continuous solar power. And given the direct relationship between energy usage and living standard, this space civilization can be expected to exceed both Mars and Earth in this respect.

Nuclear fission reactors are a possibility for Mars, but have become politically unpalatable. Nuclear fusion is presently not possible, and given its history, probably cannot be counted upon in the near term.

Constant access to sunlight means the climate of an O'Neill settlement can be whatever we choose.

It's not presently known which of our food crops (if any) can tolerate 24-hour-a-day sunlight. But if some or all should prove able to take advantage of this, then unceasing sunlight will be an option for the food growing areas of space habitats.

Convenient Access to Zero Gravity

Aside from the obvious entertainment possibilities, zero gravity enables the construction of vast, gossamer-thin space mirrors several miles across. Such mirrors can enable Earth-like conditions inside of orbital habitats, and concentrating mirrors can provide prodigious amounts of heat for ore processing. Such flimsy mirrors would not be practical on the surface of Mars due to the presence of gravity and winds.

By the same token, heat radiators for habitats can have both enormous area and very thin construction. Getting rid of waste heat is admittedly more difficult in space than in a planetary environment, but the superior kind of heat radiators which it will be possible to construct in zero G should provide a solution to this problem.

All Gravitational Options Available

Orbital habitats will simulate gravity by rotation. This means complete control over the gravitational environment. A full one-G can easily be provided if it should happen that nothing less than this will maintain normal muscular physique. If 1/3 G does turn out to be acceptable, we can have that too.

The unfortunate situation with planets is that you have to take the gravity you get, and it's frequently the wrong amount. It's possible children born on Mars may not be able to visit Earth. At best, they certainly couldn't visit Earth without considerable discomfort. Children born in an O'Neill habitat under a full one-G of centrifugal force shouldn't experience any problems in this regard.

Rotating a habitat on the surface of Mars to bring the gravity up to one G is probably not an option due to excessive loading on the bearings, and air drag. But in zero G and vacuum, making a habitat spin and keeping it spinning are much easier to accomplish.

Thus, in terms of both the levels of sunlight and gravity, orbital settlements may provide a much more Earth-like environment than even a completely-terraformed Mars.

Living at the Top of a Well

We here on the Earth are the "gravitationally disadvantaged", living as we are at the bottom of a steep gravity well. It's the reason space flight is so expensive and difficult for us. The gravity well of Mars is less deep than that of Earth, but it's still much deeper than that of the Moon, and enormous compared to that of an asteroid. This is why mining asteroids to bring materials back to Earth is just barely a possibility for the future, whereas Mars mining would probably not be able to compete due to the tariff which gravity imposes.

Settlers living near the top of Earth's gravity well will be ideally positioned for departures to Mars, or any other destination elsewhere in space.

No Weather, Save What We Make for Ourselves

Planets with atmospheres have weather, and we must consider it when designing and building structures. On Mars, there are globe encircling dust storms. Photographs from the Mars Global Surveyor have recently been released showing the shadows of 5-mile-high dust devils.

In orbit, coasting silently in vacuum, there's no weather. On the other hand, inside sufficiently large habitats, it should be possible to create our own natural weather, complete with cloudscapes and rainstorms. But since we'll be making the weather, we'll be in control of it. Thus there's every reason to expect the climate in any man-made habitat to be vastly superior to all but the most desirable climates found on Earth.

Some type of weather control might be possible inside of domed enclosures on Mars. But a thoroughly terraformed Mars will have weather systems beyond human control. Again, we tend not to recognize this as a liability, because it's the situation we're already well used to here on Earth.

Warning! Warning! Meteor Storm!

One widely perceived advantage of Martian settlements over orbital ones is that they'll have the atmosphere of Mars to protect them from meteors. But it turns out the danger of meteor strikes in space is much more modest than "Lost In Space" has led us to believe. O'Neill estimated his Island Three model (biggest cross-sectional area means greatest risk) might expect to get hit by a one ton meteor once every million years. One should expect a meteor the weight of a tennis ball to come along about every three years. Even if it did penetrate the hull, such a puncture would mean a routine, minor repair, not an emergency.

Some express a concern about man-made orbital debris, which is in fact becoming a serious problem for space stations. But most of this debris is in Low Earth Orbit, with some around Geosynchronous Earth Orbit. The nearest O'Neill habitats would orbit far higher; at least halfway to lunar orbit.

Convenient Communication with the Homeworld

The signal delay time due to the speed of light for High Earth Orbit is less than a second. Apollo astronauts were able to communicate (albeit with slight awkwardness) as far away as the Moon. The speed of light delay between Earth and Mars ranges from a bit over 4 minutes to 21 minutes. This obviously makes real-time conversations impossible. Martians would have to settle for "video letters". But much more significantly, this time delay makes the use of telepresence from Earth impossible on Mars. On the other hand, mining operations on the moon, and refining and fabrication operations in High Earth Orbit might get a boost at the onset by extensive teleoperation, reducing the initial manpower requirements. Being able to prime the industrial pump remotely (should advanced telepresence technologies become available) would certainly have significant advantages.

Convenient Travel to the Homeworld

Most NASA estimates for the trip time to Mars place it at eight to ten months. Making various estimates regarding advanced space drives, this can be brought down by several months. For example, Robert Zubrin proposes a nuclear engine-augmented heavy lift launch vehicle which could transport colonists to Mars in seven months. The political acceptability of launching nuclear engines is presently uncertain (whether or not such fears are scientifically justifiable is largely irrelevant).

Space habitats in an orbit roughly halfway to the moon will have travel times from Earth of less than a week even with present chemical rockets. A Closed Ecology Life Support System (CELSS) may be required for journeys to Mars, but should be unnecessary for travel to and from habitats orbiting the Earth.

Making A Living

Although a Martian economy may someday become possible, it seems likely to remain a local economy. There seem to be no marketable products that Martians could sell to Earth which would be worth lifting out of the gravity well of Mars. Martians might sell real estate to Earthlings, provided there was some compelling reason to want to live there, and living conditions on Mars were reasonably pleasant.

The residents of orbital colonies, in addition to building additional colonies for sale to immigrants from Earth, would certainly also be constructing Solar Power Satellites (SPS) for sale to countries in need of additional electrical capacity, and gigantic communication platforms for geosynchronous orbit. Admittedly, one doesn't need large, Earth-like habitats in order to use space resources to create these products, but having a permanent workforce nearby certainly helps. Provided that Earth was still footing the bill for continued space exploration at this point, we could even add manned exploratory ships for travel elsewhere in the Solar System to the list of products which would help balance the sheets.

All of the above would represent returns to the economy of Earth. However dedicated Mars enthusiasts may be to creating a self-sustaining Martian economy, it's difficult to see how the process can get started in the first place without expectations of a return on Earth-originated investments. SPS and advanced communication platforms are products marketable to Earth. By the same token, unlike the natural resources of Mars, the resources of the Moon and Near Earth Objects (NEO's) are sufficiently close by that their usage could have returns to the terrestrial economy.

An earlier argument was made that Mars was the place to go to because, unlike the Moon, it has plentiful supplies of hydrogen, carbon, and nitrogen. The recent discovery of ice at the lunar poles by the Lunar Prospector probe has blunted this argument somewhat. It's also worth mentioning that all the elements we need are available in most Near Earth Asteroids.

A Staging Post for the Belt?

It's sometimes argued that communities on Mars are necessary to support further expeditions out into the asteroid belt (which is expected to be a treasure trove of needed resources). It may be that this perception comes from the Earth-bound truism that being closer to a place makes it easier to get to. But in space, distance does not count nearly so much as delta V. Inhabitants on Mars would be closer to the Belt than we are, but they'll be at the bottom of a gravity well. In terms of delta V, residents of a habitat in a high enough Earth orbit will already be half-way to the belt, and can use low-thrust, high-efficiency drives the entire way.

Distance can be a significant factor in space travel with regard to travel times, especially in relation to life-support system requirements. In fact, this can legitimately be used to argue that going to Mars is still a more "difficult" journey than returning to the Moon, even if aerobraking and in-situ fuel production combine to make it "easier" to get to Mars from a strictly fuel-oriented standpoint. But one imagines by the time closed-ecology habitats are circling the Earth, a CELSS for a trip to the asteroid belt would be available.

It seems likely that only NEO's will be utilized in the near term, with Main Belt asteroids being used later, as humanity spreads outward from Earth. Even after the NEO's are exhausted, it'll always be possible to build new space habitats in the Belt itself, close to the source of raw materials.

From several standpoints, settlers in orbital space will quite literally be in a better position to exploit asteroids than anyone on a planetary surface.

Room To Grow

A completely terraformed Mars would give us approximately the land area of the Earth. Earth is much bigger, but is mostly covered with oceans. If oceans are desired on Mars (or if they turn out to be necessary to sustain an Earth-like global climate), then deduct accordingly.

We earlier stated that only an orbital settlement construction program which resulted in new land area equal to that offered by Mars should be compared to a Mars terraforming project. Is it really possible to build this many space communities? Can they be built on a time scale competitive with terraforming?

Let's do some simple calculations. The 1975 NASA-Ames study resulted in a space habitat design known as the Stanford Torus. It was designed to provide 670,000 square meters of living space. The surface area of Mars is around 145 million square kilometers. Let's assume no Martian oceans, and ignore the fact that even on a totally Earth-like Mars the Polar Regions would certainly remain inhospitable. In this comparison, Mars roughly equals 216 million Stanford Toruses.

NASA's study indicated that the first independent orbital habitat could be completed 22 years after initiation of the program, and that a habitat could build a duplicate habitat in 2 years. Let's conservatively assume it'll take 50 years to build the first one, and assign 5 years to the doubling time. We'll also ignore the possibility that the more space communities we build, the better (and faster) at it we'll get, as well as the efficiencies which might be later gained by building smaller numbers of larger habitats with greater land area.

By the time we had achieved 28 doublings, we would have begun to exceed the total land area of Mars. By the pessimistic timetable above, this would take 190 years. (Please note it's not being predicted that this will indeed be the actual growth rate seen, merely that this is the maximum rate allowed by technological constraints.) It's the rare terraforming proposal that's optimistic enough to promise a Mars made over in Earth's image in less than two centuries. Another important point is that near the beginning of the O'Neill settlement construction program, the workforce begins enjoying living conditions that may be even more Earth-like than those possible on Mars only at the very end of the terraforming effort.

Lest it be argued there's not enough raw material available to build this many orbital settlements, the resources of the asteroid Ceres alone would allow us to do this five hundred times over. Space habitats represent an incredible economy of mass in comparison to planets.

O'Neill was interested in shattering the limits to growth which were widely (if somewhat incorrectly) perceived in the 1960's. He once calculated that even if we limited ourselves to the resources of the asteroid belt (merely the most convenient source of materials, and not the only one), we could still build, in the form of orbital habitats, over 3,000 times the livable surface area of the Earth.

Thus, Mars holds out the promise of becoming home to a planetary civilization which might rival that of Earth. Orbital colonies can form a space-based civilization that far surpasses Earth's in both size and diversity.

Location Options

Martian settlements would be a home on Mars. Orbital settlements could be homes anywhere in the solar system (and perhaps even beyond) that we care to be, as long as we're not frequently eclipsed by a planet. One can locate further from the sun simply by making the habitat mirrors bigger, and slightly concave, so as to concentrate the sunlight up to Earthly levels. Even the apparent diameter of the sun in the sky would be the same as we're presently used to. Habitat orbits beyond Pluto are not out of the question. Unfortunately, this same solution cannot be used to raise the sunlight levels for colonies on the surface of Mars, due to the impracticality of such flimsy mirrors in an environment of gravity and weather.

Some space-based societies will doubtless elect to remain close to Earth, enjoying both real-time communication with, and speedy travel to, the homeworld. Other groups, wishing to remain forever apart from Earth-centered civilization, could choose to become far more remote than even Martians ever could.

Significantly, orbital territories would be mobile. This is yet another clear advantage that would scarcely even occur to us Earth-bound folk, since it's an option we've never enjoyed here. If an orbital community found itself next door to another that it simply couldn't stand, there would be a solution short of "ethnic cleansing". They could simply attach engines, and move.

Spreading Interstellar

At this point we cast our gaze even further afield, toward the distant stars. It's true the perfection of our terraforming skills on Mars might ultimately make the settlement of selected planets in certain other solar systems possible. But O'Neill habitats make every solar system a candidate for settlement, regardless of the presence or absence of suitable planets, or indeed any planets at all. Many solar systems may lack terrestrial planets due to gravitational disruptions from superjovian planets or multiple suns. But from our present understanding of the dynamics of solar system formation, systems without asteroids or comets seem unlikely.

Getting What You Need

Thus far the discussion may seem a little one-sided. Aren't there any criteria by which building settlements on Mars may have advantages over building them in free space? There's one I'm aware of. The High Frontier scenario is dependent on the retrieval of resources to High Earth Orbit, either from the lunar surface or from NEO's. The cost of transporting these raw materials must be factored into the cost of establishing space habitats. On Mars, the ores needed are literally underfoot. Both carbon and oxygen can be generated from the atmosphere using technology already demonstrated in the laboratory. The Martian atmosphere, while thin, is accessible anywhere on the planet. This is a significant advantage.

There are two questions then that must be answered. Number one is: Are the costs of transporting the needed resources to Earth orbit so large that they exceed the additional expenses of establishing colonies at the much more distant location of Mars? The answer may not be presently clear. Mass-driver technology holds out the promise of significant cost savings in the area of resource retrieval. A mass-driver erected on the lunar surface can act as a catapult for launching ores into space for pennies a pound. Less well known, a mass-driver can also function as a highly efficient reaction engine for an asteroid ore transporter. Such an engine would require nothing other than solar energy for power, and dirt for reaction mass. However, while demonstration models of mass-drivers have validated that the desired payload accelerations are possible, the required speeds have not yet been demonstrated.

The other question is: Even if it should happen that ore transportation costs were to make orbital settlements somewhat more expensive than Mars settlements, would the difference completely overwhelm the many advantages of orbital space we've discussed?

Robert Zubrin's Views

Since the passing of Gerard O'Neill in 1992, Dr. Robert Zubrin, founder of Pioneer Astronautics and author of "The Case For Mars" has emerged as the foremost public advocate of colonization beyond the Earth (the only other possible contender might be Marshall Savage). Zubrin has designed strategies for voyaging to Mars that are much less expensive than any previous proposals, and believes such journeys will lay the foundation for a vast future Martian civilization. He's been quoted as referring to O'Neill's concepts for constructing orbital habitats and Solar Power Satellites from space resources as "absurd".

In his paper "The Economic Viability of Mars Colonization", Zubrin makes the following remark:

But the biggest problem with the Moon, as with all other airless planetary bodies and proposed artificial free-space colonies (such as those proposed by Gerard O'Neill) is that sunlight is not available in a form useful for growing crops. This is an extremely important point and it is not well understood.

This point is not only "not well understood", but is quite surprising to one viewing things from the High Frontier perspective, which holds that solar energy is more available in High Earth Orbit than on any planetary body. But Zubrin proceeds to explain his logic. He reminds us that crops must be protected from space radiation, and calculates this would require glass walls 10 cm thick, which is assumed to be "prohibitively expensive". Then, apparently aware that this is in fact not the solution proposed by O'Neill, he goes on to say:

Use of reflectors and other light-channeling devices would not solve this problem, as the reflector areas would have to be enormous, essentially equal in area to the crop domains, creating preposterous engineering problems if any significant acreage is to be illuminated.

What these preposterous engineering problems specifically are, he does not indicate. It's true we're not normally accustomed to discussing the construction of mirrors miles across. But we normally view things from the perspective of the Earth's surface, where such mirrors would have to support their own weight, and withstand winds and other weather conditions. Space mirrors will face no such requirements.

In this same paper, Zubrin confesses that, following several decades of atmospheric density buildup, the normal processes of photosynthesis might take a millennia to add sufficient oxygen to the Martian atmosphere to make it breathable. Thus, he anticipates that more high-tech methods will be employed to speed up this process. One method he discusses is nanotechnology, which he estimates might cut the time down to a mere thirty years.

Zubrin chooses to consider mirrors the size of cities preposterous. A proponent of High Frontier concepts might similarly choose to view self-replicating machines the size of molecules as preposterous. So is that it? Is this debate ultimately reduced to dueling incredulities? Perhaps. But this can be said: The technology to create miles-scale mirrors in zero gravity and vacuum would certainly seem to be more in hand than molecular nanotechnology.

Some other methods Zubrin cites to possibly speed up terraforming efforts are terrawatt-sized fusion reactors, space-based lasers, and space-based reflectors; the latter of which is the very technology which he will not allow may make independent orbital communities possible.

In his paper on the economic viability of Mars, Zubrin foresees a triangle trade amongst Earth, Mars, and the Asteroid Belt. But this interplanetary economy is predicated on the assumption that asteroid miners will be unable to grow their own foodstuff. This proceeds from Zubrin's dismissal of the concept of space mirrors as big as cornfields. If the High Frontier concept should prove correct, asteroid miners can live permanently in the belt in Earth-like habitats perfectly capable of growing their own crops. In such a situation, Mars would seem to have little to sell them.


Gerard O'Neill's findings prompted Isaac Asimov to coin a new phrase: "planetary chauvinism". This refers to our natural tendency to assume activities elsewhere in space are best done on the surface of a planet. But as seen here, almost any way you look at it, planets are inconvenient things.

When making serious proposals for colonization beyond the Earth, we're obliged to set aside the romance of Burroughs and Bradbury, and ask what strategies return the greatest benefits most quickly for the least investment.

To my mind, space is the place.

Postscript: After publication of the above article, I started a thread on the Usenet newsgroup sci.space.policy named "Which is better, Mars settlements or space settlements?" in which I invited counter-points to the above. Discussion continued in another thread entitled "Why Mars Now?".

One poster made the case that on a terraformed Mars, the ecology would be "wilder". It's doubtless true that the ecology in an orbital habitat will have to be much more closely managed than any planetary ecology. And it may very well be that orbital settlements will for a very long time remain too "park-like" for some folk's tastes.

Two others saw advantages for Mars where incremental expansion of living space was concerned. One raised the possibility of a pressurized brickwork habitat for Mars made out of indigenous materials.  Making bricks is certainly a simpler materials process than making steel plate, but I'm uncertain about the relative levels of labor involved.  If we care to make it a comparison between orbital settlements and terraforming Mars, I think orbital settlements are the obvious winners in the incremental expansion category.

One poster, after urging that all debate on this subject cease, said he would only state one advantage of Mars:  Mars has more resources.  While Mars certainly masses more than the belt or even the moon, this argument doesn't consider the issues of ease of access, or costs of exporting resources.

One of the more frequent posters to the threads stated that Mars is more popular, and expressed the opinion this single fact overrides all other considerations.

Perhaps one of the best points made by someone on the newsgroups was that I was being somewhat unfair in my depiction of Robert Zubrin's views. It was pointed out that the section of "The Case For Mars" from which I pulled the quote concerning the impracticality of large space mirrors was from the near-term-future section of the book. The reference to using orbital mirrors to aid in Mars terraforming efforts was from a later section dealing with a much more distant future (one in which the engineering difficulties of enormous space mirrors had presumably been worked out).

It was almost certainly wrong of me to leave the reader with the impression that Zubrin's points are inconsistent. In fact, in his later book, "Entering Space", Zubrin, while arguing forcefully that space settlements will never be built in support of SPS, later says that a Type III (interstellar) civilization will build orbital habitats in asteroids and Oort cloud objects "with many of the features envisioned by O'Neill".

But I think the point remains that Zubrin mentions High Frontier proposals only to compare them to Mars Direct (a much more modest, and hence realizable, near-term goal). When discussing much more future eras, High Frontier gets little or no mention, and no comparison is ever made with proposed Mars terraforming efforts. Humanity is assumed to still be working the Martian surface in an attempt to make it more Earthlike. This seems to assume that no better alternative will be available even then.

I don't think it's incorrect to say that more than one reader of Zubrin has come away with an impression that orbital habitats are forever impossible, given his stressing of the engineering difficulties over the short term, and his nearly exclusive discussion of planetary engineering for later eras. The quote "...human beings will never settle Earth orbit, because there is nothing there to settle" seems open-ended, and without qualification with regard to time lines.

I would agree that High Frontier should not be proposed as an alternative to Mars Direct, since (as stated near the beginning of this article) they are not comparable projects. I still see the way clear to propose space settlements as an alternative to the large scale settlement and/or terraformation of Mars, as by the time we've gained the technical experience needed to re-engineer other worlds, there could be no doubt we could also engineer large orbital structures as well.

Return to Articles


This is the archive version of the Mike Combs Space Settlement web site and is provided as a courtesy of the National Space Society.