Kim Stanley Robinson's asteroid terrariums probably take inspiration from the O'Neill Cylinder, pictured above. Credit: Wikipedia

For most of us, the word "terrarium" conjures images of fish tanks filled with an assortment of plants, rocks and soil, maybe a little pond and a couple of salamanders. Whatever mental image comes to mind, it's probably not one of a massive, hollowed-out asteroid, careening through space supporting a miniature replica of the African savannah. Unless, that is, you're Kim Stanley Robinson. Or you've happened to read his recent novel, 2312.

Of Robinson's many imaginative concepts for humanity's future, this one takes the cake for me. Three centuries from now, the human race has found creative ways to survive in hostile extraterrestrial environments, from scorching Mercury to frozen Titan. Not limiting our sights to planets and moons, we've also started terraforming asteroids: hollowing out the centers, seeding them with nutrients and engineering everything – atmosphere, soil, gravity, day length, inhabitants- from the ground up. The result? A custom-designed, self-contained ecosystem. But also an astounding confluence of physics and art. Imagine if Earth's ecosystems, rather than stretching across the outer surface of our planet, somehow grew inward to face the center. Our sense of up and down would no longer be oriented by the sky, but by the Earth's core. Robinson describes this situation eloquently:

"Always the entire landscape will be curving up around you, rising on both sides and meeting overhead, so that the look of the land will envelop you like a work of art- a Goldsworthy inscribed on the inside of a rock, like a geode or a Faberge egg."


Could this awesomely crazy idea actually work? As we start to colonize planets throughout our solar system and beyond, could we also scatter our ecosystems among the lifeless rocks in between? And would these isolated pockets of life be worth the effort?

It's when pondering deep questions like this that I'm especially grateful to have my brother Nick Stone, astrophysicist at Columbia University, to turn to for guidance.

Gravity, light and movement

To start imagining how an asteroid terrarium might work, Robinson uses the analogy of a ship. Start with the ship's stern. This is where you bore into the asteroid, leaving a single hole through which you eject your carved out rock material (Robinson recommends keeping some of this material on-hand for mountain building and other landscaping activities later). The bow of the ship attaches to a "forward unit" which serves to navigate the craft.


The concept of turning an asteroid into a spacecraft is not entirely new: conceptual designer Bryan Versteeg has been working with the Mars One Team to imagine how we could use large asteroids for mining, transportation and habitat. Credit:

But the forward unit performs another essential role. It gets the asteroid spinning. Spinning creates a centrifugal force that anybody standing on the inside of the asteroid will experience as gravity: it will pull you toward the ground. (If you've ever been on a spinning ride like a Wipeout or Rotor at an amusement park, it's the exact same idea). The strength of that centrifugal force is proportional to the radius of the asteroid and its rotational rate. Using simple physics, one can calculate the exact rotational rate required to match the gravitational pull we experience on Earth.


"It'll take some energy to start up, probably a significant amount", says Nick. "But once you get it started it'll keep spinning forever. Remember, space is an empty vacuum, and objects in motion stay in motion until something- some sort of drag force- slows them down."

An important fact here is that every point on the inner surface will experience this artificial gravity. This means that in no part of the terrarium would you be "on the ceiling". The landscape would curve up and overhead to envelop you, making everything from trees to migratory birds appear "upside down" on the other side of the world.

Then there's the issue of sunlight and energy. According to Robinson, we'll use artificial "sunlines" to light up our asteroid terrariums:

"String the axis of the cylinder with your terrarium's sunline…the lit portion of the line, appropriately bright, then traverses the terrarium from stern to bow, taking usually the same time as a Terran day, as measured by the latitude of your biome on Earth."


One interesting aspect of the artificial sunline concept is that by varying the speed of your "sun", you would be able to create shorter or longer days, simulating the different photoperiods experienced at different Earth-latitudes. This would be important for accurately mimicking environmental conditions in, say, a Canadian boreal forest versus an Ecuadorian tropical forest.

But let's be real here. How much energy is this going to take? What, if not the sun, could we possibly use to generate enough energy to power an entire ecosystem?


Space based solar panels may not be far off: the Solaren Corporation is already working on it. Credit:

Well, maybe we can use the sun after all- indirectly. Why not cover the outer surface of the asteroid with solar panels, funnel that solar energy inside and regurgitate it? Sure, some energy will be lost along the way, but I'd like to think we'll have have substantially improved the efficiency of solar cells by the 23rd century.

However, this solar power plan assumes our asteroid is close enough to the sun to receive a decent amount of solar radiation. That could be difficult if, as Robinson imagines, our terrarium doubles as transportation, ferrying people across the solar system.


For solar panels, the energy flux per unit area falls as 1/r^2, with r being the distance from the sun. This means Jupiter, at about five Earth-distances from the sun, receives 1/25th the solar power we enjoy on Earth.

"Once you go beyond Mars, solar power isn't very useful," says Nick. "In the outer solar system you'd need some form of nuclear power to keep things going. Current NASA probes use fission, although in the future it could be fusion."

Getting around efficiently will pose similar problems to keeping your terrarium lit and heated. "Again, if you have fusion reactors, can probably get around pretty inexpensively," says Nick. "Otherwise, you might want to consider a more passive mechanism like a solar sail. This will be much slower, but it'll require very little energy."


Artist's conception of a solar sail. Credit: Space Services Holdings, Inc.

A solar sail is exactly what it sounds like. Paper-thin and occupying an enormous surface area, solar sails absorb momentum from the sun's radiation. "By tilting it like a sail on a sailboat, you could essentially move yourself around the solar system for free", Nick says. The first spacecraft to successfully demonstrate solar-sail technology, IKAROS, was launched by the Japan Aerospace Exploration Agency in 2010.


For a slightly faster, but still passive locomotion option, one might consider a magnetic sail. Our sun emits a stream of charged particles- protons and neutrons- that are collectively known as the stellar wind. The trajectories of these particles are bent by magnetic fields. "In theory, you can use stellar wind as a drag force, allowing you to move pretty quickly throughout the solar system," says Nick. "For this to work, you'd need to maintain a magnetic field, which will cost some energy."

The sky above

The atmosphere is a crucial, life-sustaining components of any biosphere. Here, I feel like Robinson's terrarium concept could use a bit more fleshing out. Robinson glosses over the issue of atmosphere, simply suggesting that you can "aerate the interior to the gas mix and pressure you desire… in something like the Terran mix of gases with perhaps a bit more oxygen."


Sounds reasonable enough. The thing is, once you start introducing life, that atmospheric composition can get out of whack really quickly.

My initial concern here was climate change. Would rapid fluctuations in the atmosphere's greenhouse gas concentrations cause the terrarium's climate to spiral out of control? Would a boomer crop of trees or algae suck all the heat-trapping CO2 out of the atmosphere, turning our little terrarium into an ice box? In Earth's geologic past we've gone through snowball periods and sweltering greenhouses due to small changes in atmospheric carbon dioxide and methane concentrations.

But it turns out this particular apocalypse scenario probably won't be a major concern inside an asteroid, for the very same reason I was worried: confinement. On Earth, a large fraction of the ~340 Watts per square meter of energy we receive from the sun shoots back into space as long-wave radiation. The exact amount of our sun's energy that ends up escaping Earth is highly sensitive to concentrations of greenhouse gases in the atmosphere. More greenhouse gases, less of the sun's energy escapes, and the surface of our planet heats up. That's global warming, in a very simplified nutshell.


But inside an asteroid, the atmosphere isn't open to space. It's encased in meters to kilometers of solid rock. This will effectively prevent any of the sunline's energy from escaping. Greenhouse gases may still trap heat in the terrarium's atmosphere, but that heat wasn't heading anywhere, anyway. It's a moot point.

An artistic conception of snowball earth. Credit: Walter Myers

Runaway climate change aside, the potential for trace gas concentrations to increase or decrease rapidly under confinement poses other biological threats. At concentrations of 1%, CO2 can start having negative physiological effects on most animals, while concentrations of 7-10% are toxic. For CO2's deadlier cousin carbon monoxide, this toxicity threshold is < 0.1%. We probably wouldn't want a fire-prone ecosystem in our terrarium, unless we had really good method for scrubbing extra carbon monoxide and dioxide out of the atmosphere.


The ground below

Soil is in the most diverse habitat on earth, performing many essential ecosystem functions. Credit: National Geographic


And finally, there's the matter of soil. The recycling factory and nutrient reservoir for all land-bound life, soil is an extraordinarily complex mess, composed of a mix of rock-derived minerals, dead organic matter and living microorganisms. To build soil on a lifeless rock, Robinson first proposes seeding the ground with microbes. These little guys will get the soil forming process started, eating away at solid substrate while building up pockets of nutrients and organic matter. "To make it even more welcoming", Robinson says, "scrape the interior wall of your cylinder, then crumble the rock scrapings finely, to a consistency ranging from large gravel to sand." Add some water, crank up the heat a bit, and our mixture "will rise like yeasted dough as it becomes the most delicious and rare substance, soil."

What Robinson is actually describing here is accelerated weathering: speeding up the natural process of rock crumbling. Adding heat and water alone start causing rocks to dissolve. By pre-crushing some of the asteroid's inner surface, we can create additional surface area for weathering reactions to occur, as well as habitats for our microbes.

A word of caution here: try as we might to speed things along, soil is going to take a really long time to build up. We're talking hundreds to thousands of years at least, given that we're starting from scratch. Perhaps we'd be able to watch the ground rising like "yeasted dough" if we had a couple thousand years of time-lapse footage. But I wouldn't hold my breath for fresh garden veggies in five years. This is a long-term investment.


My diagnosis?

There are perhaps millions of asteroids in our solar system. Do these represent tiny islands for us to colonize? Credit: NASA


Lots of open questions remain, but this is an exciting idea. I think for me the big issue here is not one of how we might build an asteroid terrarium but why. Because however you slice it, building an ecosystem from the ground up is going to require a lot of resources and time.

If goal is to build economy, whether that means food, natural resources, or even tourism, then revenues need to make up for investments, over some meaningful timescale. If we're growing food, we'd need to take into account not only the time and resources involved in building soil and maintaining artificial sunlight, but transportation costs: shuttling the food grown in our terrarium to human populations. I've no doubt there are a thousand other costs- from irrigation to waste recycling to insect control to ship repairs- that I haven't covered here.

But what if our terrariums are intended for conservation and species preservation? In this case, perhaps the question of cost is less important. But we'd need governments that cared enough about conservation to make the investment. Are we willing to pay the price needed to scatter droplets of Earth throughout the stars, simply to preserve our heritage or push the limits of our existence? If you're an Octavia Butler fan, you might see this as a sort of manifest destiny and call it by another name: Earthseed. At the end of the day, will these seeds of Earth help us survive as a species? If the answer is yes, then perhaps the costs, resources and time are all irrelevant, because at the end of the day, survival is our ultimate evolutionary goal.


Madeleine Stone is the author of this blog and The Lonely Spore. Follow her on Twitter.