The Challenge of Recreating Earth

I became obsessed with terraforming at age seven. I just didn't know we called it that.

I was playing SimEarth, a relic from the early days of Maxis's sim franchises in which the player controls the evolution of an entire biosphere. As you work through the simulation, you make choices about your planet's atmosphere, geology, and climate that in turn shape its biomes and life forms. If you're lucky, these decisions will eventually lead to advanced sentient life. But more often, you run out of resources, your time expires, or you hit a snafu and your little pixelated world spirals out of control.

To be fair, SimEarth is an extremely simplistic depiction of what it takes to build a living planet. But the game's core challenge hits an undeniable truth: Recreating Earth's biosphere is really hard. And yet, our long term survival as a species may hinge on doing just that.

So far, we've tried, we've failed, and we've learned some important lessons along the way. It's these lessons, together with our growing knowledge of the Earth system, that may one day help humanity pull a SimEarth, and rebuild our biosphere on other worlds.

Before we can even talk about simulating the biosphere, we need to understand conceptually what it actually is. Simply put, Earth's biosphere is the sum of all its ecosystems—from rainforests and savannas to coral reefs and deserts—along with the air, water, soil and climate that sustain them. All told, the biosphere is about 12 miles thick from the deepest ocean trenches to the highest mountain peaks.

This thin film of life did not just spontaneously appear. It's been growing and evolving for nearly four billion years, in a manner that's deeply integrated with that of Earth's non-living systems. Our biosphere is dependent on everything from the molten core that drives tectonic activity to the invisible veneer of atmospheric ozone that shields us from harmful radiation. Just looking at what makes our biosphere tick, it starts to become clear why this system is so hard to replicate.

Still, some intrepid humans have striven to do just that. The most famous attempt is Biosphere 2, a 3.14 acre Earth-system research complex located just outside of Tucson, Arizona. Constructed in the early 1990s as a materially closed ecological system that could simulate future space habitats, this Earth-in-a-bottle includes replica biomes from across the planet, including a rainforest, a savannah, a fog desert, an ocean, and a mangrove swamp. The original design also included an agricultural system and a human habitat. Heated and powered by the sun's radiation, with water recycled through internal plumbing, Biosphere 2 remains to this day an outstanding engineering feat and one of the most ambitious ecological experiments ever conceived.

To determine whether human life can be sustained in a closed habitat, two manned missions were carried out inside the complex in the early 90s. Both were a spectacular fiasco.

"Because there wasn't a scientific community behind this, it exploded in their faces," Peter Troch, the current director of science for Biosphere 2, told me over the phone.

During Mission 1, which ran from 1991 to 1993, the eight Biosphere 2 inhabitants reported continual hunger. CO2 levels within the habitat fluctuated wildly16:5%3C597::AID-JOC28%3E3.0.CO;2-J/abstract), most vertebrate life and all pollinating insects died, cockroach and ant populations exploded. Filtration systems clogged, unanticipated condensation dampened the desert, and morning glories took over the rainforest, choking out other plant life.

Worst of all for the humans, the oxygen levels inside the facility fell steadily over the first 16 months. The reason was unclear at the time, but researchers have since presented evidence that O2 was rapidly consumed by soil microorganisms. By January of 1993, oxygen levels had fallen to under 15 percent—equivalent to those found at an elevation of 13,000 feet. Since the inhabitants were starting to suffer from sleep apnea and fatigue, the management group was forced to intervene and artificially boost oxygen levels. The crew managed to eke it out another 9 months.

Mission 2 only lasted 6 months, ending prematurely due to a bitter dispute among Biosphere 2's leaders that eventually ousted the management company Space Biosphere Ventures.

Biosphere 2 was an enormous undertaking, and, according to Troch, the people that put it together should have anticipated unexpected outcomes.

"If you take a very complex, nonlinear system, and force that system into an initial condition that's out of equilibrium, and let it run, it'll go in a non-predictable way," he said.

Which seems to be exactly what happened. Still, the Biosphere 2 debacle became an important cautionary tale, clearly demonstrating myriad ways in which a complex system can quickly spiral out of control.

"Take, for instance, the rainforest," Troch said. "When that was first put together, they used very nutrient rich soil, because they wanted the plants to survive. But then, when they closed it off, that soil started reacting with the atmosphere and doing strange things to the oxygen levels. This is a perfect example of what happens when you create initial conditions that are incompatible with each other."

Troch, who came on board when Biosphere 2 was acquired by the University of Arizona in 2007, is now refashioning the facility as a center for state-of-the-art Earth systems science research. Instead of trying to study an entire biosphere in confinement, Biosphere 2's new Landscape Evolution Observatory (LEO) will model the evolution of ecosystems from the ground up.

To start, Troch and his colleagues have replaced the agricultural land with barren artificial landscapes consisting of crushed basalt rock. In the coming years, researchers will use these landscapes to study the carbon and water cycles, slowly adding microbes and plants to build complexity over time. The entire system is being carefully monitored with hundreds of sensors that detect fluctuations in moisture, temperature, CO2 concentrations, nutrients and oxygen.

"When we study Earth system processes in the real world, we have a problem in that what we observe now has the full complexity of many millennia of development," Troch told me. "Here, we're starting at time zero. As we observe the evolution of an inert, dead landscape, we are asking, what are the interacting processes that lead to a healthy ecosystem?"

This bottom-up approach will allow the scientists to ask deep questions about how Earth's thin, life sustaining surface came to be. Insights from LEO could also prove incredibly valuable to future humans hoping to terraform other worlds.

"This will no doubt prove useful [to terraformers]," Troch told me. "It's a similar situation—you have an inert system, and you want to know what will happen as you add water, CO2, microbes and plants. What we are doing and learning every day will provide basic knowledge if we ever want to do this on another planet."

Terraforming an entire planet still sounds like a distant dream. And yet, in the past 20 years, the discussion of how we might terraform Mars—arguably the most Earth-like planet we know—has shifted from the realm of sci-fi writers into the discourse of a small but vocal group of planetary scientists.

"In our solar system, Mars is the only world for which we can entertain a scientific discussion about this idea," Chris McKay of NASA's Ames Research Center told me over the phone. To McKay, who has been leading the scientific discussion on Martian terraforming since the 90s, the idea of giving the red planet a green makeover is not nearly as crazy as it sounds. Still, the challenge will prove astronomically greater than building a biosphere in a lab.

Eons ago, Mars might have had a massive ocean, tectonic activity, and a thick, life-supporting atmosphere. But Earth's little brother couldn't hold its heat for long. When Mars's core grew cold, its plate tectonics shut down, its atmosphere fizzled away, and its surface settled into a deep freeze. Today, all that remains of Mars's ancient waterways is locked up as underground ice. Temperatures on the ruddy surface average around -63 degrees Celsius, and the thin atmosphere that remains is almost entirely carbon dioxide.

In other words, Mars as it stands today is a barren, inhospitable wasteland.

Still, Mars is geologically similar to Earth, and key ingredients for life, including carbon, water and oxygen, seem to be abundantly present. A biosphere might just be possible, if we kickstart a massive global warming campaign. McKay believes that the key to cranking up the heat on Mars lies in perfluorocarbons (PFCs)—carbon and fluorine-based molecules that trap solar energy with a thousand times the efficiency of CO2. As a first step, McKay envisions deploying hundreds of small PFC factories across the Martian surface to generate these super-greenhouse gases from elements already present in the crust.

In his seminal Nature paper on terraforming Mars, McKay showed that by adding just a few parts per million of PFCs to the Martian atmosphere, surface temperatures could be hiked from -60 to -40 degrees Celsius in a matter of decades. While -40 doesn't sound very spring-like, it'd be warm enough to trigger the release of frozen CO2 from the ground into the atmosphere. That CO2 would act as a positive feedback, accelerating the greenhouse effect even further.

Based on estimated CO2 inventories within the Martian regolith, McKay calculates that within a century of warming, Mars could have a CO2-rich atmosphere with enough pressure to support plant and microbial life. By the end of that century, according to McKay, the surface would be averaging 15 degrees Celsius—the same average temperature as the Earth—and the water ice frozen deep underground would begin to thaw.

Once the Martian hydrosphere starts to activate, we could begin introducing the first life forms. Certain UV-tolerant plants and microorganisms might survive on a CO2-rich Mars, and McKay imagines it'd be up to them to start oxygenating the atmosphere for the rest of us. Still, the process of greening Mars could take centuries or more, as the planet's deeply frozen regolith slowly thaws out. Suggestions to speed the wetting process along include using orbital mirrors to deflect extra sunlight to Mars's poles, spreading dark materials on the surface to increase heat absorption, or using a couple of good old fashioned thermonuclear detonations.

Even if we were able to fully green up Mars, the atmosphere wouldn't be breathable for humans and most animals. We'll need a lot of oxygen, and we'll also want that CO2 diluted with an inert gas, such as the nitrogen that comprises 78 percent of our own atmosphere. And therein lies a problem: More than a million billion tons of nitrogen would be needed to produce an Earth-like atmosphere on Mars. We have no idea whether such a massive reservoir exists.

"We now know that nitrates are there in Martian soil, but we're not sure of their abundance," McKay told me."For years, I've said this is biggest uncertainty in terraforming Mars."

Nitrates are a form of nitrogen that can be converted into dinitrogen gas, either by microbial metabolism or via industrial processes. Even if enough nitrates were present in the soil to fill the atmosphere, they'd take an enormous amount of energy to liberate, and extracting them would not be nearly as quick as rustling up the planet's CO2.

Concept of a terraformed Mars over the tharsis region. Image:

Daein Ballard / Wikimedia

Which brings us to the final ingredient we'll need to make an Earth-like biosphere on Mars: time. Nitrogen aside, McKay figures it'll take somewhere in the neighborhood of 100,000 years for plant photosynthesis to build up enough oxygen for humans to breathe.

"Plants are the best oxygen factories we've got, but they still produce it with very low efficiency," McKay told me. "That final step is a long, distant haul."

Even if we genetically engineer plants to double their natural photosynthetic capacity, or increase it tenfold, we're still talking many, many generations before future humans will be able to picnic on Mars without an O2 mask. Committing ourselves to such a long-term endeavor may be the hardest part of building the next biosphere. Especially when weighed against another, inescapable truth: Our Martian biosphere will, eventually, fail.

"Without plate tectonics to recycle carbon, all the CO2 we put into the atmosphere will eventually outgas into space," McKay said. Which, in turn, will cause all plant life to die. "It'll take about a hundred million years, but this biosphere won't stick around forever."

A hundred million years of life on Mars seems like a pretty good deal when weighed against a hundred thousand years of terraforming. And seeding other worlds with Earth-like biospheres might be necessary for our long term survival—after all, Earth itself has another billion years of habitability tops. Still, the time and resource investment we're talking about here is mind-bogglingly vast, and, as our early forays into simulated environments on Earth have shown, there's plenty of room for things to go horribly wrong.

After pondering all this complexity and time, I found an old copy of SimEarth on an abandonware site and decided to boot it up for kicks. I went for the Martian terraforming scenario and set the difficulty to easy, knowing I'd be rusty. Because SimEarth was released before we had strong evidence of widespread water ice on Mars, the game has you hurl ice meteors at the planet to wet it up. After nuking Mars with a handful of these, I was suddenly out of "omega," the currency needed to do just about everything. Impatient for my omega to refill, I set the game to fast mode and walked away for a few minutes. When I returned, a text box had popped up to inform me that I'd lost. What the hell? It was the 200th year of my Martian terraforming effort, and my time, apparently, had run out.

Two hundred years. If future Martian terraformers are lucky, that'll be barely enough time for the first hardy grasses to put down roots. Still, I'm kind of okay with SimEarth's absurdly skewed sense of time. I'll never get to walk through a Terran forest on Mars, but within the boundaries of this 640-pixel wide screen, I could green up half the galaxy.

Perfect Worlds is a series on Motherboard about simulations, imitations, and models. Follow along here.